www.aaojournal.

org

Ophthalmology, the flagship journal of the American Academy of Ophthalmology, serves society by publishing clinical and basic
science research and other relevant manuscripts that relate to the sense of sight. Excellence is pursued through unbiased peer-
review, the advancement of innovation and discovery, and the promotion of lifelong learning.

Editor-in-Chief Editorial Board

Stephen D. McLeod, MD Anita Agarwal, MD Sue Lightman, MD, PhD

San Francisco, CA San Francisco, CA London, England

Distinguished Senior Editors Iqbal Ike K. Ahmed, MD Joan W. Miller, MD

Toronto, Canada Boston, MA
Daniel M. Albert, MD, MS
Portland, OR Tin Aung, FRCS(Ed), PhD Sanjay V. Patel, MD
George B. Bartley, MD Singapore Rochester, MN
Rochester, MN
Balwantray C. Chauhan, PhD Michael X. Repka, MD
Creig Hoyt, MD, MA Halifax, Canada Baltimore, MD
San Francisco, CA
Emily Y. Chew, MD SriniVas R. Sadda, MD
Andrew P. Schachat, MD Bethesda, MD Los Angeles, CA
Cleveland, OH

Associate Editors Michael F. Chiang, MD Carla J. Siegfried, MD

Portland, OR St. Louis, MO
Henry D. Jampel, MD
Baltimore, MD Mary Frances Cotch, PhD Sharon D. Solomon, MD
Bethesda, MD Baltimore, MD
Bennie H. Jeng, MD, MS
Baltimore, MD Laura Downie, BOptom, PhD Angelo P. Tanna, MD
Melbourne, Australia Chicago, IL
Susan Vitale, PhD, MHS
Bethesda, MD Anat Galor, MD, MSPH John T. Thompson, MD
Miami, FL Baltimore, MD
Editorial Staff
Martine J. Jager, MD, PhD Daniel Ting, MD, PhD
Dale E. Fajardo, EdD, MBA Singapore
Vice President, Education Leiden, the Netherlands

May B. Piotrowski Rachel W. Kuchtey, MD, PhD Russell N. Van Gelder, MD, PhD
Executive Editor Nashville, TN Seattle, WA

Meredith C. Jones Thomas M. Lietman, MD Marco A. Zarbin, MD, PhD

Senior Editor San Francisco, CA Newark, NJ
Greg Pelkofski
Web Production Coordinator
Jeanine M. Beisel
Editorial Specialist

Former Editors-in-Chief

Clarence Loeb, 1919e1929 W. Howard Morrison, 1969e1975 Donald S. Minckler, 1995e2002

Arthur W. Proetz, 1930e1931 Stanley M. Truhlsen, 1976e1979 Andrew P. Schachat, 2003e2012
William P. Wherry, 1932e1939 Paul Henkind, 1980e1986 George B. Bartley, 2013e2017
William L. Benedict, 1940e1968 Paul R. Lichter, 1987e1994

Inquiries or comments may be directed to the editorial office at [email protected].

Ophthalmology (ISSN: 0161-6420) is published monthly for the American Academy of Ophthalmology, Inc. by Elsevier Inc., 230 Park Avenue, Suite 800, New York, NY 10169-0901,
USA. Business and Editorial Offices: 1600 John F. Kennedy Boulevard, Suite 1800, Philadelphia, PA 19103-2899. Periodicals postage paid at New York, NY and additional mailing
offices. POSTMASTER: Send address changes to Ophthalmology, Elsevier, Journal Returns, 1799 Highway 50 East, Linn, MO 65051, USA.
 Ophthalmology®
¼ ONLINE-ONLY MATERIALS AVAILABLE ¼ VIDEO

Sign-up for Table of Contents and Articles In Press alerts at www.aaojournal.org.

Also available online is an article sharing feature, citation alerts, and downloadable
images as PowerPoint files. Follow us on Twitter @AAOjournal

Contents
Volume 127, Number 1, January 2020

This Issue at a Glance 1

Sandeep Ravindran, PhD

Ophthalmology Retina Abstracts 2

Ophthalmology Glaucoma Abstracts 7

Editorial: Trachoma: Time to Talk Eradication 11

Thomas M. Lietman, MD, Catherine E. Oldenburg, PhD, MPH, Jeremy D. Keenan, MD, MPH

Original Articles

Topical Recombinant Human Nerve Growth Factor (Cenegermin) 14

for Neurotrophic Keratopathy: A Multicenter Randomized
Vehicle-Controlled Pivotal Trial
Stephen C. Pflugfelder, MD, Mina Massaro-Giordano, MD, Victor L. Perez, MD,
Pedram Hamrah, MD, Sophie X. Deng, MD, PhD, Ladan Espandar, MD, MS,
C. Stephen Foster, MD, John Affeldt, MD, John A. Seedor, MD, Natalie A. Afshari, MD,
Wendy Chao, PhD, Marcello Allegretti, PhD, Flavio Mantelli, MD, PhD, Reza Dana, MD, MPH
In this pivotal trial conducted in neurotrophic keratopathy patients, topical cenegermin
(recombinant human nerve growth factor) demonstrated efficacy and safety in healing
persistent epithelial defects (with or without stromal thinning).

(Continued)

On the cover: “Anterior Segment Fluorescein Angiography in Viral Anterior Uveitis” by

Arun Kapil (Photographer), Aniruddha Agarwal, MD, Vishali Gupta, MS (Advanced Eye
Center, Department of Ophthalmology, Post Graduate Institute of Medical Education and
Research, Chandigarh, India). The photograph depicts inflamed iris vessels in a 29-year-old
woman with herpes simplex viral anterior uveitis who has undergone trabeculectomy for
increased intraocular pressures (secondary glaucoma). Equipment: Heidelberg Spectralis
(Heidelberg Engineering GmbH, Heidelberg, Germany).
 Ophthalmology®
Contents
(Continued)

A Randomized Noninferiority Trial of Wearing Adjustable Glasses 27

versus Standard and Ready-made Spectacles among Chinese
Schoolchildren: Wearability and Evaluation of Adjustable Refraction III
Congyao Y. Wang, MD, Guoshan Zhang, MM, Bobby Tang, MB, Ling Jin, MS,
Wenyong Huang, MD, Xiuqin Wang, MD, Tingting Chen, MD, Wenhui Zhu, MD,
Baixiang Xiao, MM, Jun Wang, MD, Zhongqiang Zhou, MD, Zhizheng Tang, MB,
Yan Liang, MD, Mabel Crescioni, LLM, DrPH, David Wilson, PhD, BEc,
Helen McAneney, PhD, Joshua D. Silver, PhD, Bruce Moore, OD, Nathan Congdon, MD, MPH
This randomized trial of self-refraction versus ready-made glasses versus standard refraction
showed that self-refraction was not inferior to standard refraction among Chinese school-aged
children with regard to accuracy, vision outcomes, and acceptability of glasses.

Diagnostic Accuracy of Technology-based Eye Care Services: 38

The Technology-based Eye Care Services Compare Trial Part I
April Y. Maa, MD, Charles M. Medert, MD, Xiaoqin Lu, MD, Rabeea Janjua, MD,
Ashley V. Howell, MPH, Kelly J. Hunt, PhD, Sarah McCord, MD, Annette Giangiacomo, MD,
Mary G. Lynch, MD
The Technology-based Eye Care Services protocol is comparable to an in-person screening
examination in terms of diagnostic accuracy and is one valid ophthalmologic telemedicine
tool to provide access to eye care for patients.

Methods for Intraocular Lens Power Calculation in Cataract Surgery 45

after Radial Keratotomy
Andrew M.J. Turnbull, BM, FRCOphth, Geoffrey J. Crawford, MD, FRANZCO,
Graham D. Barrett, MD, FRANZCO
Seven methods of intraocular lens power calculation after radial keratotomy are compared in
a large series of eyes with complete refractive history from 1 center. The Barrett True K
[History] formula delivers the greatest accuracy.

A Prospective Randomized Trial Comparing Hydrus and iStent 52

Microinvasive Glaucoma Surgery Implants for Standalone Treatment of
Open-Angle Glaucoma: The COMPARE Study
Iqbal Ike K. Ahmed, MD, Antonio Fea, MD, PhD, Leon Au, MBBS, Robert E. Ang, MD,
Paul Harasymowycz, MD, Henry D. Jampel, MD, Thomas W. Samuelson, MD,
David F. Chang, MD, Douglas J. Rhee, MD, on behalf of the COMPARE Investigators
A randomized trial comparing the Hydrus Microstent to 2 first-generation iStents for
treatment of standalone open-angle glaucoma shows that the Hydrus Microstent is associated
with reduced intraocular pressure and medication use compared with the iStent.

(Continued)
 Ophthalmology®
Contents
(Continued)

Comparison of Associations with Different Macular Inner Retinal 62

Thickness Parameters in a Large Cohort: The UK Biobank
Anthony P. Khawaja, PhD, FRCOphth, Sharon Chua, PhD, Pirro G. Hysi, PhD,
Stelios Georgoulas, FRCOphth, Hannah Currant, BSc, Tomas W. Fitzgerald, PhD,
Ewan Birney, PhD, Fang Ko, MD, Qi Yang, PhD, Charles Reisman, MSc,
David F. Garway-Heath, MD, FRCOphth, Chris J. Hammond, FRCOphth, Peng T. Khaw, PhD,
Paul J. Foster, PhD, Praveen J. Patel, FRCOphth, MD(Res), Nicholas Strouthidis, PhD, for the
UK Biobank Eye and Vision Consortium
A study of macular inner retinal anatomic features in 42 044 British adults found strongest
associations with the ganglion celleinner plexiform layer, suggesting this parameter to be the
superior biomarker for macular pathophysiologic processes.

HAWK and HARRIER: Phase 3, Multicenter, Randomized, 72

Double-Masked Trials of Brolucizumab for Neovascular
Age-Related Macular Degeneration
Pravin U. Dugel, MD, Adrian Koh, MD, FRCS, Yuichiro Ogura, MD, Glenn J. Jaffe, MD,
Ursula Schmidt-Erfurth, MD, David M. Brown, MD, Andre V. Gomes, MD, PhD,
James Warburton, MBBS, Andreas Weichselberger, PhD, Frank G. Holz, MD,
on behalf of the HAWK and HARRIER Study Investigators
Brolucizumab 6 mg (every[q] 12 weeks[w]/q8w) demonstrated noninferior best-corrected
visual acuity and superior anatomic outcomes versus aflibercept q8w aflibercept, with >50%
of brolucizumab 6 mg patients maintained on a q12w interval over 48 weeks.

Development and Validation of Deep Learning Models for Screening 85

Multiple Abnormal Findings in Retinal Fundus Images
Jaemin Son, MSc, Joo Young Shin, MD, MSc, Hoon Dong Kim, MD, MSc,
Kyu-Hwan Jung, PhD, Kyu Hyung Park, MD, PhD, Sang Jun Park, MD, MSc
A deep learning-based screening system that detects a broad range of ophthalmologic findings
was validated quantitatively and qualitatively with external datasets that consist of multiple
ethnicities.

Commentary: Using Deep Learning Models to Characterize Major Retinal 95

Features on Color Fundus Photographs
Cecilia S. Lee, MD, MS, Ryan T. Yanagihara, MD, Aaron Y. Lee, MD, MSCI

(Continued)
 Ophthalmology®
Contents
(Continued)

The Relative Impact of Patient, Physician, and Geographic Factors on 97

Variation in Primary Rhegmatogenous Retinal Detachment
Management
Daniel Vail, BA, Suzann Pershing, MD, MS, Mary-Grace Reeves, BA,
Armin R. Afshar, MD, MBA
Among 12 779 commercially insured patients with incident rhegmatogenous retinal
detachment managed between 2008 and 2016, patient-level variation contributed most to
repair likelihood, physician-level variation contributed most to repair type and geographic
variation contributed minimally.

Quantitative Assessment of the Retina Using OCT and Associations 107

with Cognitive Function
Yoshikazu Ito, MD, Mariko Sasaki, MD, PhD, Hiroki Takahashi, MD, Shoko Nozaki, MD, PhD,
Shinichiro Matsuguma, MD, PhD, Kaoru Motomura, MD, Rihito Ui, MD, Ryo Shikimoto, MD,
Ryo Kawasaki, MD, PhD, Kenya Yuki, MD, PhD, Norie Sawada, MD, PhD,
Masaru Mimura, MD, PhD, Kazuo Tsubota, MD, PhD, Shoichiro Tsugane, MD, PhD
Dementia was associated with macular thickness, but not with peripapillary retinal nerve fiber
layer (ppRNFL) thickness in 975 Japanese participants, suggesting that macula thickness
could be superior to ppRNFL thickness in assessing neurodegenerative changes.

Commentary: Retinal Imaging in Alzheimer’s Disease: In Search of 119

the Holy Grail
Amani A. Fawzi, MD, Sandra Weintraub, PhD, Waleed Fawzi, MRCPsych, MD

Secondary Prevention of Retinoblastoma Revisited: Laser 122

Photocoagulation of Invisible New Retinoblastoma
Sameh E. Soliman, MD, Cynthia VandenHoven, BAA, CRA, Leslie D. MacKeen, BSc,
Brenda L. Gallie, MD, FRCSC
Secondary prevention of familial retinoblastoma was possible with timely accurate
OCT-guided localization and laser photocoagulation of 11 new tumors in 7 eyes of 5 children
without further local or systemic treatment burden.

Ophthalmic Technology Assessment: Autologous Serum-Based 128

Eye Drops for Treatment of Ocular Surface Disease: A Report
by the American Academy of Ophthalmology
Roni M. Shtein, MD, MS, Joanne F. Shen, MD, Anthony N. Kuo, MD,
Kristin M. Hammersmith, MD, Jennifer Y. Li, MD, Mitchell P. Weikert, MD
Autologous serum-based eye drops may be an effective and safe treatment for dry eyes and
nonhealing epithelial defects, but conclusions are limited owing to the absence of controlled
trials.

(Continued)
 Ophthalmology®
Contents
(Continued)

Reports

Redundancy of Progress Notes for Serial Office Visits 134

Michelle R. Hribar, PhD, Adam Rule, PhD, Abigail E. Huang, MD, Haley Dusek, BA,
Isaac H. Goldstein, BA, Brad Henriksen, MD, Wei-Chun Lin, MD, Austin Igelman, BS,
Michael F. Chiang, MD

Effects of Omega-3 Supplementation on Exploratory Outcomes in 136

the Dry Eye Assessment and Management Study
Marko Oydanich, MS, Maureen G. Maguire, PhD, Maxwell Pistilli, MS,
Pedram Hamrah, MD, Jack V. Greiner, DO, PhD, Meng C. Lin, OD, PhD,
Penny A. Asbell, MD, MBA, for the Dry Eye Assessment and Management Study Research
Group

Computer-Based Quality Improvement for Management of 138

Functionally Monocular Patients
Kyle A. Rogers, MD, Andrew A. Wilson, MD, Mark A. Wyckoff, R. Michael Siatkowski, MD

Corrigendum 139

Preferred Practice PatternÒ

(Full Text Available at aaojournal.org)

Age-Related Macular Degeneration Preferred Practice PatternÒ

Christina J. Flaxel, MD, Ron A. Adelman, MD, MPH, MBA, FACS, Steven T. Bailey, MD,
Amani Fawzi, MD, Jennifer I. Lim, MD, Gurunadh A. Vemulakonda, MD,
Gui-shang Ying, MD, PhD

Diabetic Retinopathy Preferred Practice PatternÒ

Christina J. Flaxel, MD, Ron A. Adelman, MD, MPH, MBA, FACS, Steven T. Bailey, MD,
Amani Fawzi, MD, Jennifer I. Lim, MD, Gurunadh A. Vemulakonda, MD,
Gui-shang Ying, MD, PhD

Posterior Vitreous Detachment, Retinal Breaks, and Lattice

Degeneration Preferred Practice PatternÒ
Christina J. Flaxel, MD, Ron A. Adelman, MD, MPH, MBA, FACS, Steven T. Bailey, MD,
Amani Fawzi, MD, Jennifer I. Lim, MD, Gurunadh A. Vemulakonda, MD,
Gui-shang Ying, MD, PhD

(Continued)
 Ophthalmology®
Contents
(Continued)

Adult Strabismus Preferred Practice PatternÒ

Linda R. Dagi, MD, Federico G. Velez, MD, Jonathan M. Holmes, MD, Stacy L. Pineles, MD,
Steven M. Archer, MD, Mitchell B. Strominger, MD, Matthew Simon Pihlblad, MD,
Evelyn A. Paysse, MD, David R. Stager, Sr., MD, David Stager, Jr., MD,
Sarah E. MacKinnon, CO, COMT, Burton J. Kushner, MD, Hatice Tuba Atalay, MD,
Brian N. Campolattaro, MD, Hilda Capo, MD, Natalie C. Kerr, MD

Correspondence (available at www.aaojournal.org/current) e4

Re: Heindl et al.: The eyes of Oetzi: the Tyrolean iceman mummy
(Ophthalmology. 2019;126:530)
Francesco M. Galassi, MD, Elena Varotto, MA, PGD, Andreas G. Nerlich, MD, PhD,
Dong Hoon Shin, MD, PhD, Raffaella Bianucci, PhD

Reply
Stephanie Panzer, MD, Patrizia Pernter, MD, Ludwig M. Heindl, MD,
Albert R. Zink, PhD

Re: Lambert et al.: Intraocular lens implantation during early

childhood: A Report by the American Academy of Ophthalmology
(Ophthalmology. 2019;126:1454-1461)
Savleen Kaur, MS, Vijay Sharma, MS, Jaspreet Sukhija, MS, Jagat Ram, MS

Reply
Scott R. Lambert, MD, Deborah K. Vanderveen, MD, Stephen J. Kim, MD

Re: Lambert et al.: Intraocular lens implantation during early

childhood: a report by the American Academy of Ophthalmology
(Ophthalmology. 2019;126:1454-1461)
Ameenat Lola Solebo, PhD, FRCOphth, Ian Christopher Lloyd, FRCOphth,
Jugnoo Rahi, PhD, FRCOphth

(Continued)
 Ophthalmology®
Contents
(Continued)

Reply
Scott R. Lambert, MD, Deborah K. Vanderveen, MD,
Vinay K. Aakalu, MD, MPH, Stephen J. Kim, MD

Re: Lee et al.: Longitudinal changes in peripapillary retinal

nerve fiber layer thickness in high myopia: a prospective,
observational study (Ophthalmology. 2019;126:522-528)
Philip H. Wright, BMBS, BMedSci, Omar A. Mahroo, PhD, FRCOphth,
Christopher J. Hammond, MD, FRCOphth

Reply
Min-Woo Lee, MD, Ju-Mi Kim, MD, Yong-Il Shin, MD,
Young-Joon Jo, MD, PhD, Jung Yeul Kim, MD, PhD

Pictures & Perspectives

Progression from Classical Oguchi Disease to Retinitis Pigmentosa after 51

50 Years
Koji M. Nishiguchi, MD, PhD, Yoshihisa Oguchi, MD, PhD, Toru Nakazawa, MD, PhD

Intracorneal Migration of Silicon Band 61

Vinod Kumar, MD, Abhidnya Surve, MD

Fingerprint Macula Artifact on Optos Fundus Imaging in Nystagmus 96

Robert A. Sisk, MD, FACS, Prashant K. Parekh, MD, Christopher D. Riemann, MD

A Baby with Raccoon Eyes 106

Wei Xiao, MD, Chaochao Xu, MD, Yuxiang Mao, MD

Pembrolizumab-Related Enophthalmos 121

Alexandra Manta, MD, Robert A. Goldberg, MD

The journal’s Guide for Authors and Guide for Reviewers are
available at www.aaojournal.org.
 This Issue at a Glance

Ophthalmology ®

Volume 127, Number 1, January 2020

complex (GCC) thickness, but not in peripapillary retinal nerve

Comparing the Efficacy of Microinvasive fiber layer (ppRNFL) thickness. Ganglion cell complex and
macular thicknesses were associated with the presence of de-
Glaucoma Surgery Devices
mentia, but ppRNFL thickness was not. Ganglion cell-inner
Ahmed et al (p. 52) conducted a prospective, multicenter, plexiform layer, GCC, and full macular thicknesses were all
randomized clinical trial to compare the efficacy of 2 micro- associated with the presence of dementia in the inferior sectors.
invasive glaucoma surgery (MIGS) devicesdThe Hydrus The results indicate that macular thickness, but not ppRNFL,
Microstent and the iStent Trabecular Micro-Bypass Stent Sys- was associated with the presence of dementia, and the authors
temdfor reducing intraocular pressure (IOP) and medications in suggest that OCT measurements of the macula could be superior
open-angle glaucoma. A total of 152 eyes from 152 patients in assessing neuro-degenerative changes.
were randomized 1:1 to standalone MIGS, consisting of either 1
Hydrus Microstent or 2 iStent devices. In the 148 subjects who
completed 12-month follow-up, the Hydrus resulted in a higher
Evaluating the Diagnostic Accuracy of
surgical success rate and fewer medications used compared with Technology-Based Eye Care Services
the 2-iStent procedure, and both devices had similar safety Maa et al (p. 38) conducted a prospective comparison between
profiles. The study suggests that MIGS devices could help the standard protocol of the Veterans Affairs Healthcare System’s
manage IOP and demonstrates some advantages of the Hydrus Technology-based Eye Care Services (TECS) and face-to-face
Microstent. (FTF) examination to identify aspects of the process that could
be refined to improve accuracy. A total of 256 patients with no
Brolucizumab vs Aflibercept for Treating known history of significant ocular disease were recruited. The
researchers found that TECS readers showed substantial agree-
Neovascular Age-Related Macular
ment for cataract and diabetic retinopathy, moderate to substan-
Degeneration tial agreement for glaucoma/glaucoma suspect, and moderate
In 2 similarly designed randomized, double-masked, multicenter agreement for age-related macular degeneration compared with
phase 3 trials (HAWK and HARRIER), Dugel et al (p. 72) an FTF examination. The TECS protocol showed high percent
compared brolucizumab, a single-chain antibody fragment that agreements and good sensitivity and excellent specificity
inhibits vascular endothelial growth factor-A, with aflibercept to compared with an FTF eye examination for the detection of
treat neovascular age-related macular degeneration. A total of common eye diseases. The authors conclude that the standard
1817 patients were randomized to intravitreal brolucizumab 3 mg TECS protocol is comparable with an FTF examination, and its
(HAWK only) or 6 mg or aflibercept 2 mg. In both trials, each diagnostic capabilities could be enhanced by including additional
brolucizumab arm demonstrated noninferiority versus aflibercept sophisticated ophthalmic testing.
in best-corrected visual acuity change. More than 50% of bro-
lucizumab 6 mg-treated eyes were maintained on an every-12- Topical Recombinant Human Nerve Growth
weeks dosing through Week 48. At the matched Week 16
assessment, fewer brolucizumab 6 mg-treated eyes had disease
Factor for Neurotrophic Keratopathy
activity versus aflibercept in both studies. Anatomic retinal fluid Pflugfelder et al (p. 14) conducted a multicenter, randomized,
outcomes favored brolucizumab over aflibercept, and overall double-masked vehicle-controlled trial to evaluate the safety and
safety with brolucizumab was similar to aflibercept. The authors efficacy of topical recombinant human nerve growth factor
conclude that brolucizumab is noninferior to aflibercept in visual (cenegermin) in patients with neurotrophic keratopathy. Forty-
function at Week 48 and suggest that it could serve as an alter- eight patients were enrolled and randomized 1:1 to receive either
native treatment option. cenegermin 20 mg/ml or vehicle eye drops. Cenegermin treat-
ment showed higher rates of corneal healing, with statistically
significant differences between treatment groups at week 8.
Investigating the Associations between
Cenegermin-treated patients showed significantly reduced lesion
Retinal Thickness and Cognitive Function sizes compared with vehicle-treated patients, and fewer
In a cross-sectional, population-based survey, Ito et al set cenegermin-treated patients experienced disease progression over
(p. 107) out to determine the association between retinal thick- the 8-week treatment course compared with vehicle-treated pa-
ness and cognitive function. A total of 1293 Japanese persons tients. The researchers found that cenegermin was well tolerated.
aged 65 to 86 years underwent comprehensive ophthalmic They conclude that cenegermin treatment could potentially serve
assessment, as well as detailed mental health examinations to as a safe and noninvasive treatment for neurotrophic keratopathy.
designate them into 3 groups: normal, having mild cognitive
impairment, or having dementia. Significant differences were
found in the 3 groups in all retinal layers and ganglion cell Sandeep Ravindran, PhD

1
 Abstracts

Volume 3, Number 12, December 2019

www.ophthalmologyretina.org

Efficacy of Photodynamic Therapy for

Polypoidal Choroidal Vasculopathy Secondary and Exploratory Outcomes of
Associated with and without Pachychoroid the Subthreshold Nanosecond Laser
Phenotypes Intervention Randomized Trial in Age-
Masayuki Hata, MD, PhD, Miho Tagawa, MD, Akio Oishi, MD, Related Macular Degeneration: A LEAD
PhD, Yu Kawashima, MD, Isao Nakata, MD, PhD, Yumiko Akagi- Study Report
Kurashige, MD, PhD, Kenji Yamashiro, MD, PhD, Sotaro Ooto, Zhichao Wu, BAppSc (Optom), PhD, Chi D. Luu, PhD, Lauren
MD, PhD, Hiroshi Tamura, MD, PhD, Manabu Miyata, MD, PhD, A.B. Hodgson, MPH, Emily Caruso, BOrth, Kate H. Brassington,
Masahiro Miyake, MD, PhD, Naoko Ueda-Arakawa, MD, PhD, MPH, Nicole Tindill, BBus(BIS), Khin Zaw Aung, MBBS, Colin A.
Ayako Takahashi, MD, PhD, Akitaka Tsujikawa, MD, PhD Harper, MBBS, FRANZCO, Sanjeewa S. Wickremasinghe,
DMedSc, FRANZCO, Sukhpal S. Sandhu, MD, FRANZCO, Myra
Purpose: To compare the efficacy of photodynamic therapy
B. McGuinness, MBiostat, PhD, Fred K. Chen, MBBS (Hons),
(PDT) in eyes with polypoidal choroidal vasculopathy (PCV)
PhD, Usha Chakravarthy, MD, PhD, Jennifer J. Arnold, MBBS
associated with and without pachychoroid phenotypes (pachych-
(Hons), Wilson J. Heriot, MBBS, FRANZCO, Shane R. Durkin,
oroid PCV and nonpachychoroid PCV, respectively).
MBBS (Hons), FRANZCO, Maximilian W.M. Wintergerst, MD,
Design: Retrospective chart review.
Shekoufeh Gorgi Zadeh, MSc, Thomas Schultz, PhD, Robert P.
Participants: Patients previously diagnosed with PCV and
Finger, MD, PhD, Amy C. Cohn, MBBS, MMed, Elizabeth K.
initially treated with PDT.
Baglin, BOrth (Hons), Pyrawy Sharangan, BOrth, Robyn H. Guy-
Methods: Patients were classified as having pachychoroid- or
mer, MBBS, PhD, for the LEAD Study Group
nonpachychoroid-driven conditions. The long-term visual outcome
and its associated factors were investigated. Purpose: To evaluate the secondary and exploratory outcomes
Main Outcome Measures: Visual acuity (VA) outcomes at 1, of the Laser Intervention in Early Stages of Age-Related Macular
2, 3, 4, and 5 years after initial PDT in pachychoroid and non- Degeneration (LEAD) study, a 36-month trial of a subthreshold
pachychoroid PCV. nanosecond laser (SNL) treatment for slowing the progression to
Results: Of the 158 eyes, 88 (55.7%) met the criteria for late age-related macular degeneration (AMD) in its early stages.
pachychoroid PCV; 70 (44.3%) did not (nonpachychoroid PCV). In Design: Multicenter, randomized, sham-controlled trial.
cases of pachychoroid PCV, VA improved significantly at 1 year Participants: Two-hundred ninety-two patients with bilateral
(P ¼ 0.042) and maintained baseline level at 5 years (P ¼ 0.38). By large drusen.
contrast, VA continued to deteriorate in the nonpachychoroid PCV Methods: Participants were randomly assigned to receive SNL
group during the follow-up period and had already declined signif- or sham treatment to the study eye at 6-month intervals.
icantly by the second year (P ¼ 0.022, compared with baseline). Main Outcome Measures: The secondary outcome measure of
Despite no difference in baseline VA between pachychoroid and the LEAD study was the time to development of late AMD, defined
nonpachychoroid PCV groups (P ¼ 0.11), the VA at 5 years was by multimodal imaging in the nonestudy eye. The exploratory
significantly better in the pachychoroid PCV group compared with outcome measures were the rate of change in best-corrected visual
the nonpachychoroid PCV group (0.540.47 vs. 0.930.63, acuity (BCVA), low-luminance visual acuity, microperimetric mean
respectively; P ¼ 0.23  10-3). The incidence of massive sub- sensitivity, drusen volume in the study and nonestudy eyes, and
macular hemorrhage (SMH) or vitreous hemorrhage (VH) was not participant-reported outcomes based on the Night Vision Question-
different between groups at 5 years (P ¼ 0.67), and their occurrence naire and Impact of Vision Impairment questionnaire.
was associated with decreased VA in both the nonpachychoroid and Results: Progression to late AMD in the nonestudy eye was
pachychoroid PCV groups (coefficient b, 0.361 and 0.481; P ¼ not significantly delayed with SNL treatment (hazard ratio, 0.83;
0.59  10-3 and P < 1.0  10-5, respectively). 95% confidence interval, 0.40e1.71; P ¼ 0.611). There was no
Conclusions: Five years after PDT treatment, VA was main- evidence of effect modification based on the coexistence of retic-
tained at the baseline level in the pachychoroid PCV group but not ular pseudodrusen; interaction P ¼ 0.065). There was no signifi-
in the nonpachychoroid PCV group. Massive SMH or VH during cant difference between study groups in the rate of change of low-
the follow-up period affected the final visual outcomes in both luminance visual acuity, microperimetric mean sensitivity, and
conditions. drusen volume in the study or nonestudy eyes, and Night Vision
Ophthalmology Retina 2019;3:1016-1025. Questionnaire and Impact of Vision Impairment questionnaire

2
scores (all P  0.167). The rate of BCVA decline was slightly
higher for participants in the SNL group compared with the sham The Evolution of Fibrosis and Atrophy and
treatment group in the study eye (e0.54 and 0.23 letters/year, Their Relationship with Visual Outcomes in
respectively; P < 0.001) but not the nonestudy eye (e0.48 and Asian Persons with Neovascular Age-
e0.56 letters/year, respectively; P ¼ 0.628).
Conclusions: Subthreshold nanosecond laser treatment of one
Related Macular Degeneration
eye did not have an effect on delaying progression to late AMD in the Chui Ming Gemmy Cheung, FRCOphth, FAMS, Dilraj S. Grewal,
fellow eye and did not, in general, have an impact on the exploratory MBBS, Kelvin Yi Chong Teo, MBBS, Alfred Gan, BSc, Aditi Mohla,
structural, functional, and participant-reported outcomes. MBBS, Usha Chakravarthy, MD, PhD, Tien Yin Wong, MBBS,
Ophthalmology Retina 2019;3:1026-1034. Glenn J. Jaffe, MD

Outer Retinal Thickness and Fundus Purpose: To describe the rate of development and progression
of fibrosis and macular atrophy (MA) and their relationship with 1-
Autofluorescence in Geographic Atrophy
year visual outcomes in Asian participants with neovascular age-
Diane L. Wang, MD, Julia Agee, MD, Marco Mazzola, MD, Ric- related macular degeneration (nAMD).
cardo Sacconi, MD, Giuseppe Querques, MD, PhD, Alan D.
Design: Review of images collected from a prospectively
Weinberg, MS, R. Theodore Smith, MD, PhD
recruited observational cohort.
Purpose: Most studies of fundus autofluorescence (FAF) in Participants: Participants with treatment-naïve nAMD.
geographic atrophy (GA) have been nonquantitative, with inade- Methods: All participants underwent multimodality imaging at
quate registration of image modalities. Furthermore, as pointed out baseline and month 12 and were treated according to standard of care.
in the recent Consensus Definition for Atrophy Associated with Retinal specialists evaluated color fundus photographs fluorescein
Age-Related Macular Degeneration on OCT, it is unclear whether and indocyanine green angiography images to determine the sub-
decreased FAF would be correlated exclusively with a single types according polypoidal choroidal vasculopathy (PCV) and non-
category of OCT-defined atrophy. We sought to determine how PCV subtypes. An independent reading center graded qualitative and
FAF intensity in eyes with GA correlates with structural changes of quantitative morphologic features on spectral-domain OCT. Fibrosis
the outer retina and choroid as seen on co-registered spectral and MA were determined based on multimodal imaging.
domain OCT (SD-OCT) images. Main Outcome Measures: Incidence of fibrosis and MA and
Design: Retrospective cross-sectional. their impact on visual outcome at 1 year.
Participants: Twenty eyes of 11 patients with GA secondary to Results: We included 93 eyes (48.4% PCV). Between baseline
non-neovascular age-related macular degeneration (AMD). and month 12, visual acuity (VA) improved from 0.810.56 log-
Methods: Spectral domain OCT and FAF images for each eye arithm of the minimum angle of resolution (logMAR; Snellen
were co-registered using MATLAB (Math-Works Inc, Natick, MA). equivalent, approximately 20/126) to 0.710.61 logMAR (Snellen
On B-scans, the choroid, retinal pigment epithelium (RPE), photore- equivalent, approximately 20/100; P ¼ 0.007), and mean retinal
ceptor (PR) layer, and outer nuclear layer (ONL) were segmented. thickness decreased from 471.1 mm to 343.4 mm (P < 0.001).
Regions of interest (ROIs) including all atrophic and border regions Between baseline and month 12, prevalence of fibrosis and MA
were selected manually on the FAF scans. Regions of interest were increased from 13.0% to 37.8% (P < 0.001) and 9.7% to 17.2%
subdivided into quartiles of FAF level to correlate with retinal thick- (P ¼ 0.008), respectively. Worse baseline VA and presence of
ness measurements taken along the B-scans. Mean choroid, RPE, PR, subretinal hyperreflective material (SHRM) at month 12 were
and ONL thicknesses were compared across quartiles using an anal- associated with worse VA at month 12 after adjusting for multiple
ysis of variance factorial design testing for interaction effects, adjusted factors, whereas PCV subtype was associated with better VA at
for repeated measures (on both eyes) with a within-subjects factor. month 12. At month 12, the predominant composition of SHRM
Results: Seventeen eyes of 10 patients were selected for anal- was fibrosis (82.5%).
ysis. The mean choroidal thicknesses were not significantly Conclusions: We describe significant development of fibrosis
different across FAF quartiles, but the overall differences in mean and MA in Asian nAMD eyes and show that fibrosis is the most
RPE, PR layer, and ONL thicknesses across quartiles were statis- important predictor of outcomes. These results highlight the need
tically significant (analysis of variance, P < 0.001, P < 0.001, and for therapy beyond antievascular endothelial growth factor inhi-
P ¼ 0.015, respectively). Post hoc analysis demonstrated signifi- bition to address fibrosis in Asian nAMD.
cant differences in thickness among quartiles 1, 2, and 3 for the Ophthalmology Retina 2019;3:1045-1055.
RPE and PR layers (Tukey, P < 0.01 in each case). The FAF
quartiles within GA did not correlate exclusively with single cat-
egories of Consensus Definition for Atrophy Associated with Age- Higher-Order Assessment of OCT in
Related Macular Degenerationedefined atrophy. Diabetic Macular Edema from the VISTA
Conclusions: Not only RPE but also PR layer thickness on SD- Study: Ellipsoid Zone Dynamics and the
OCT varies significantly with FAF levels in GA. This suggests that Retinal Fluid Index
although the RPE cells are losing thickness and function, evi-
denced by decreased FAF from fluorophores, delicate PR cells also Justis P. Ehlers, MD, Atsuro Uchida, MD, PhD, Ming Hu, PhD,
succumb early in the disease process. These relationships should be Natalia Figueiredo, MD, Peter K. Kaiser, MD, Jeffrey S. Heier, MD,
pursued as a possibly better-detailed mechanism in GA. David M. Brown, MD, David S. Boyer, MD, Diana V. Do, MD,
Ophthalmology Retina 2019;3:1035-1044. Andrea Gibson, PhD, Namrata Saroj, OD, Sunil K. Srivastava, MD

3
 Purpose: To investigate retinal fluid features and ellipsoid zone Main Outcome Measures: The mean number of micro-
(EZ) integrity dynamics on spectral-domain OCT (SD-OCT) in aneurysms and NPA evaluated as the ISI.
eyes with diabetic macular edema (DME) treated with intravitreal Results: At baseline, the mean CRT was 485.790.6 mm. After
aflibercept injection (IAI) in the VISTA-DME study. treatment, the mean CRT was reduced significantly to 376.981.6
Design: A post hoc subanalysis of a phase III, prospective mm (P ¼ 0.1  10-5, repeated analysis of variance). The mean number
clinical trial. of microaneurysms was decreased significantly from 49.633.2 at
Participants: Eyes received either IAI 2 mg every 4 weeks baseline to 24.818.1 at 3 months after the initial treatment. This was a
(2q4) or every 8 weeks after 5 initial monthly doses (2q8). 50.421.2% reduction (P ¼ 0.3  10-5, paired t test). The mean ISI
Methods: All eyes from the VISTA Phase III study in the IAI was also decreased significantly from 55.520.4% at baseline to
groups imaged with the Cirrus HD-OCT system (Zeiss, Oberkochen, 28.816.8% after treatment (P ¼ 0.3  10-5, paired t test). This was a
Germany) were included. The OCT macular cube datasets were reduction of 43.328.5%. A significant correlation was found be-
evaluated using a novel software platform to generate retinal layer tween the CRT and number of microaneurysms at both baseline (r ¼
and fluid boundary lines that were manually corrected for assessment 0.56; P ¼ 0.004) and after treatment (r ¼ 0.53; P ¼ 0.006). A sig-
of change in EZ parameters and volumetric fluid parameters from nificant correlation was found between CRT and ISI at baseline (r ¼
baseline. The retinal fluid index (i.e., proportion of the retinal volume e0.39; P ¼ 0.03) but not after treatment (r ¼ e0.06; P ¼ 0.79).
consisting of cystic fluid) was also calculated at each time point. Conclusions: The reduction in the number of microaneurysms
Main Outcome Measures: The feasibility of volumetric was correlated with reduction in CRT.
assessment of higher-order OCT-based retinal parameters and its Ophthalmology Retina 2019;3:1067-1075.
correlation with best-corrected visual acuity (BCVA).
Results: Overall, 106 eyes of 106 patients were included.
Specifically, 52 eyes of 52 patients were included in the IAI 2q4
Intravitreal Aflibercept for Retinal
arm, and 54 eyes of 54 patients were included in the IAI 2q8 arm. Nonperfusion in Proliferative Diabetic
Ellipsoid zone integrity metrics significantly improved from Retinopathy: Outcomes from the
baseline to week 100, including central macular mean EZ to retinal Randomized RECOVERY Trial
pigment epithelium (RPE) thickness (2q4: 26.6 mm to 31.6 mm, P
< 0.001; 2q8: 25.2 mm to 31.4 mm, P < 0.001). At week 100, Charles C. Wykoff, MD, PhD, Muneeswar G. Nittala, PhM, Brenda
central macular intraretinal fluid volume was reduced by >65% (P Zhou, BS, Wenying Fan, MD, PhD, Swetha Bindu Velaga, BS,
< 0.001) and central macular subretinal fluid volume was reduced Shaun I.R. Lampen, BS, Alexander M. Rusakevich, BA, Justis P.
by >99% in both arms (P < 0.001). Central macular retinal fluid Ehlers, MD, Amy Babiuch, MD, David M. Brown, MD, Michael S.
index (RFI) significantly improved in both arms (2q4: 17.9% to Ip, MD, SriniVas R. Sadda, MD, for the RECOVERY Study Group
7.2%, P < 0.001; 2q8: 19.8% to 4.2%, P < 0.001). Central Purpose: Evaluate the impact of intravitreal aflibercept (Eylea;
macular mean EZ-RPE thickness (i.e., a surrogate for photore- Regeneron, Tarrytown, NY) on retinal nonperfusion (RNP) in eyes
ceptor outer segment length) and central RFI were independently with proliferative diabetic retinopathy (PDR).
correlated with BCVA at multiple follow-up visits. Design: Prospective, randomized clinical trial.
Conclusions: Intravitreal aflibercept injection resulted in sig- Participants: Eyes with treatment-naïve PDR and extensive
nificant improvement in EZ integrity and quantitative fluid metrics RNP without diabetic macular edema.
in both 2q4 and 2q8 arms and correlated with visual function. Methods: Patients were randomized 1:1 to intravitreal 2 mg
Ophthalmology Retina 2019;3:1056-1066. aflibercept every 4 weeks (monthly) or every 12 weeks (quarterly).
Main Outcome Measures: The primary outcome measure was
Multiple Effects of Intravitreal Aflibercept change in total RNP area (in square millimeters) from baseline to
on Microvascular Regression in Eyes with year 1. Secondary outcomes included ischemic index (ISI), diabetic
Diabetic Macular Edema retinopathy severity scale (DRSS) scores, visual acuity, central
retinal thickness, and adverse events. The mean and 95% confi-
Masahiko Sugimoto, MD, PhD, Atushi Ichio, MD, Daiki Mochida,
dence interval were calculated for each outcome.
Yumiho Tenma, MD, Ryohei Miyata, MD, Hisashi Matsubara, MD,
Results: Through 1 year, the monthly (n ¼ 20) and quarterly (n ¼
PhD, Mineo Kondo, MD, PhD
20) cohorts received 11.0 and 3.95 mean aflibercept injections, and
Purpose: To evaluate the effects of intravitreal aflibercept DRSS scores improved 2 steps or more in 74% and 67% of patients,
(IVA) on the number of microaneurysms and sizes of nonperfused respectively. Among all patients through 1 year, mean total area of
areas (NPAs) in eyes with diabetic macular edema (DME). RNP increased from 235 mm2 to 266 mm2 (P ¼ 0.18) and ISI
Design: Interventional, prospective study. increased from 25.8% to 31.9% (P ¼ 0.004). Retinal nonperfusion
Participants: Twenty-five eyes of 25 DME patients (average outcomes favored monthly dosing. Mean total RNP increased from
age, 64.08.8 years) were treated with 3 consecutive monthly IVA 207 mm2 at baseline to 268 mm2 (P ¼ 0.01) at 1 year in the quarterly
injections. cohort and remained stable at 264 mm2 at baseline and 1 year (P ¼
Methods: Fluorescein angiography (FA) and OCT were per- 0.70) in the monthly cohort (P ¼ 0.05, monthly vs. quarterly cohorts).
formed before the IVA injections (baseline) and at 1 week after the Although many eyes demonstrated increased areas of RNP longitu-
IVA treatment. The number of microaneurysms and the ischemic dinally (n ¼ 24 [66.7%]), this was more common with quarterly
index (ISI), a measure of NPA, were determined. The correlations dosing (n ¼ 14 [77.8%]), and a proportion of eyes (n ¼ 12 [33.3%])
between central retinal thickness (CRT) and number of micro- demonstrated localized areas of apparent reperfusion of nonperfused
aneurysms and the ISI were also determined. retina, more commonly in the monthly cohort (n ¼ 8 [44.4%]).

4
 Conclusions: Widespread evidence of retinal reperfusion with
aflibercept dosing of PDR eyes with extensive RNP was not Choroidal Vascular Pattern in Cases of
identified, and therefore the primary outcome of the current study Sturge-Weber Syndrome
was not met. Nevertheless, zones of apparent reperfusion were
Abhidnya Surve, MD, Shorya Azad, MS, Pradeep Venkatesh, MD,
detected in some patients, and a dose response was identified with a
Vinod Kumar, MD, Rohan Chawla, MD, Viney Gupta, MD, Rajpal
reduction of RNP progression with monthly compared to quarterly
Vohra, MD
dosing.
Ophthalmology Retina 2019;3:1076-1086. Purpose: To study the choroidal vascular pattern in patients with
Sturge-Weber syndrome (SWS) using swept-source OCT (SS-OCT).
Epiretinal Membrane Formation after Design: Prospective comparative observational study.
Treatment of Retinal Breaks Participants: All patients with SWS with no history of prior
treatment for posterior segment pathology were included.
Cryoretinopexy versus Laser Retinopexy
Methods: Both eyes of all patients were studied using fundus
Barton L. Blackorby, MD, Abdallah M. Jeroudi, MD, Kevin J. imaging, SS-OCT, fundus fluorescein angiography (FFA), and
Blinder, MD, Gaurav K. Shah, MD indocyanine green angiography (ICG) by 2 independent observers.
Purpose: The purpose of the study was to study the prevalence Main Outcome Measures: The FFA and ICG were screened
of macular epiretinal membrane (ERM) formation for retinal tears for any vascular abnormalities. The SS-OCT was evaluated for
treated with laser retinopexy and cryoretinopexy. The study sought choroidal changes.
to identify whether there is a difference in ERM formation prev- Results: A total of 34 eyes of 17 patients with diagnosed SWS
alence between these 2 treatments. in the age group 9 to 26 years were studied. The FFA and ICG in 7
Design: Retrospective, single-center, chart review study. and 11 patients, respectively, showed some vascular abnormalities.
Participants: Patients seeking treatment at a private practice SS-OCT was performed in all patients. The diffuse choroidal
institution (The Retina Institute, St. Louis, Missouri) over a 10-year hemangioma (DCH) was characterized by loss of the choroidal
period between 2006 and 2016 for the evaluation and treatment of vascular pattern, increase in the choroidal thickness and loss of
a retinal tear. visualization of the sclerochoroidal interface. Based on the FFA,
Methods: A chart review was conducted comprising all ICG, and SS-OCT imaging, there were 3 patients with no DCH, 5
patients undergoing procedures for Current Procedural Termi- with bilateral DCH, and the remaining 9 patients had unilateral
nology codes 67141 (prophylaxis of retinal detachment, cryo- DCH. The detection rate was 50% clinically: 52.94% with FFA,
therapy) and 67145 (prophylaxis of retinal detachment, 82.35% with ICG, and 86.36% with SS-OCT. There was sub-
photocoagulation) and patients with an International Classifica- stantial agreement between the 2 observers for all 3 investigations.
tion of Diseases, Ninth Edition, Clinical Modification, diagnosis Conclusion: SS-OCT is a reliable noninvasive imaging mo-
code of 362.56 (macular puckering), who underwent procedures dality for early diagnosis and follow-up of DCH over time.
identified with Current Procedural Terminology codes 67141 Ophthalmology Retina 2019;3:1091-1097.
and 67145.
Main Outcome Measures: Epiretinal membrane development, Retinal Vascular Abnormalities in
time between treatment and ERM development, and surgical Phakomatosis Pigmentovascularis
intervention for ERM progression. Aristomenis Thanos, MD, Tor Shwayder, MD, Thanos D. Papakos-
Results: A total of 2257 eyes underwent treatment for retinal tas, MD, Giulia Corradetti, MD, Antonio Capone Jr., MD, David
breaks with 1655 eyes treated by laser retinopexy and 602 eyes Sarraf, MD, Carol L. Shields, MD, Michael T. Trese, MD
treated by cryoretinopexy. The mean age of the cryoretinopexy
group was 59.41.5 years and in the laser retinopexy group was Purpose: To describe the spectrum of retinal vascular abnor-
61.40.8 years. A total of 74 patients (3.2%) demonstrated an malities in patients with phakomatosis pigmentovascularis (PPV).
ERM after treatment for a retinal tear during an 11-year period Design: Multicenter, retrospective, noncomparative, consecu-
(2006e2016). A total of 26 cryoretinopexy eyes (4.32%) and 48 tive case series.
laser retinopexy eyes (2.90%) demonstrated an ERM after Methods: Eligible patients underwent detailed retinal exami-
treatment of retinal breaks (P ¼ 0.094). The average time to nation including indirect ophthalmoscopy. Ultra-widefield fundus
ERM development was 11.5 months for the cryoretinopexy imaging, including color fundus photography and angiography,
group and 12 months in the laser retinopexy group (P ¼ 0.878). was performed using standardized protocols, and findings were
Seven ERMs progressed to requiring surgical intervention: 2 in recorded and reviewed and analyzed.
the cryoretinopexy group and 5 in the laser retinopexy group. Participants: Three patients with a clinical diagnosis of PPV
There was no statistically significant difference between the are presented.
groups with regard to ERM progression resulting in surgical Results: Evaluation of all patients (n ¼ 6 eyes of 3 patients) with
intervention (P ¼ 0.707). widefield fluorescein angiography showed several retinal vascular
Conclusions: Treatment of retinal breaks with either cryor- abnormalities, including peripheral retinal nonperfusion (n ¼ 3 eyes),
etinopexy or laser retinopexy showed no statistically significant peripheral vascular leakage (n ¼ 3 eyes), aberrant retinal vessels (n ¼ 1
difference in the incidence, timing, or severity of ERM formation eyes), vascular tortuosity (n ¼ 1 eyes), and disruption of the foveal
between these treatment methods. avascular zone including fovea plana (n ¼ 3 eyes). In addition, 2 eyes
Ophthalmology Retina 2019;3:1087-1090. demonstrated peripheral retinal vascular straightening and leakage

5
similar to the features of familial exudative vitreoretinopathy. One of patients with PPV. Considering the association of GNA11 patho-
the patients was a carrier of a somatic GNA11 R183C pathogenic genic variants with PPV and allied disorders, these observations may
variant that has been associated with PPV. suggest a role of guanine-binding proteins (G-proteins) in retinal
Conclusions: Fluorescein angiography, especially with wide- vascular development.
field capability, reveals numerous retinal vascular abnormalities in Ophthalmology Retina 2019;3:1098-1104.

6
 Abstracts

Volume 2, Number 6, November/December 2019

www.ophthalmologyglaucoma.org

Prioritizing Outcome Preferences in Although IOP is an outcome that researchers often measure in
Patients with Ocular Hypertension and glaucoma clinical trials, we found that patients also prioritized
outcomes related to the ability to perform vision-dependent activ-
Open-Angle Glaucoma Using BesteWorst ities (e.g., driving) as important.
Scaling Ophthalmology Glaucoma 2019;2:367-373.
Jimmy T. Le, MA, ScD, Amanda K. Bicket, MD, MSE, Ellen M.
Janssen, PhD, Davinder Grover, MD, MPH, Sunita Radhakrishnan, Identifying Outcomes That Are Important
MD, Steven Vold, MD, Michelle E. Tarver, MD, PhD, Malvina
to Patients with Ocular Hypertension or
Eydelman, MD, John F.P. Bridges, PhD, Tianjing Li, MD, PhD
Primary Open-Angle Glaucoma:
Purpose: To quantify patients’ preferences for glaucoma out- A Qualitative Interview Study
comes and use this information to prioritize outcomes that are
Jimmy T. Le, MA, ScD, Kareshma Mohanty, ScM, Amanda K.
important to patients.
Bicket, MD, MSE, Michelle E. Tarver, MD, PhD, Malvina B.
Design: Cross-sectional study using besteworst scaling (BWS)
Eydelman, MD, Tianjing Li, MD, PhD
object case.
Participants: Two hundred seventy-four participants newly Purpose: To explore patients’ perspectives and experiences
diagnosed with ocular hypertension or mild to moderate open- living with glaucoma and identify important benefits and risks that
angle glaucoma from 3 private practices and 1 academic medical patients consider before electing for new glaucoma treatments,
center in the United States. such as minimally invasive glaucoma surgery (MIGS) devices.
Methods: We designed a preference-elicitation survey based on Design: Semi-structured, in-person qualitative interviews with
findings from qualitative interviews with patients with glaucoma patients seen at the Johns Hopkins Wilmer Eye Institute.
(reported separately). The survey asked participants to rate the Participants: Adults older than 21 years of age with suspected
importance of 13 glaucoma outcomes on a 5-point scale as a warm- or who were diagnosed with ocular hypertension or mild to mod-
up exercise followed by completion of 13 BWS tasks. For each erate primary open-angle glaucoma (POAG) (e.g., eligible for
task, we presented participants with a subset of 4 outcomes from treatment with a MIGS procedure) presenting to a glaucoma clinic
the possible 13, and participants chose the most important and least in Baltimore, Maryland, between May and December 2016.
important outcome. Outcomes included in the survey pertain to Methods: We conducted in-person interviews with patients
maintaining ability to perform vision-dependent activities of daily recently diagnosed with ocular hypertension or POAG. We focused
living (e.g., driving), maintaining visual function and perception, on considerations patients take into account when deciding be-
minimizing need to take glaucoma drops, not experiencing ocular tween different treatments. We used the framework approach to
surface symptoms, and having adequate control of intraocular code and analyze the qualitative data. Considerations of special
pressure (IOP). We administered the survey online and analyzed interest to us were those that can be translated into outcomes
response patterns using conditional logistic regression to determine (or end points) in clinical trials.
the relative importance of outcomes. Main Outcome Measures: Patients’ perspectives concerning
Main Outcome Measure: Ordinal ranking of glaucoma out- outcomes that matter to them when managing ocular hypertension
comes based on preference weights. or POAG.
Results: Between September 1, 2017, and February 28, 2018, we Results: A total of 10 male and 15 female patients participated
invited 1035 patients to complete the survey, of whom 274 (26%) in our study. The median participant age was 69 years (range,
responded. Most participants were older than 65 years (146/274 47e82 years). We identified outcomes that patients expressed as
[53%]) and at the time were taking IOP-lowering drops (179/274 important, which we grouped into 4 thematic categories: (1) limi-
[65%]). Participants identified that outcomes with the largest relative tations in performing specific vision-dependent activities of daily
importance weight were having “adequate IOP control” and ability to living; (2) problems with general visual function or perceptions; (3)
“drive a car during the day,” and outcomes with the smallest relative treatment burden, including ocular adverse events; and (4) intra-
importance weights were “maintaining appearance of the eye” and ocular pressure (IOP). All 25 participants expressed some concerns
“reducing the number of IOP-lowering drops.” with their ability to perform vision-dependent activities, such as
Conclusions: Determining the relative importance of glaucoma reading and driving. All participants also had an opinion about
outcomes to patients can help researchers to design studies that IOP, and among those currently taking ocular hypotensive eye
may inform clinical and regulatory decision-making better. drops, all recognized the relationship between eye drops and IOP.

7
 Conclusions: We have identified outcomes that matter to pa- Purpose: To compare the long-term safety and efficacy of
tients who are deciding between different treatments for ocular amniotic membraneeumbilical cord (AMeUC) and pericardium
hypertension and POAG, such as the ability to drive or maintain patch grafts in reducing glaucoma shunt tube exposure.
mobility outside the home. These outcomes will be important in Design: Multicenter, prospective, randomized clinical trial.
future evaluations of new treatments for glaucoma. Participants: Adults with uncontrolled glaucoma undergoing
Ophthalmology Glaucoma 2019;2:374-382. glaucoma drainage device (GDD) implantation.
Methods: Patients were randomized to receive GDD with either
Long-Term Functional Outcomes of AM-UC or pericardium patch grafts to cover GDD tubes. Patients
were followed up clinically with anterior segment (AS) OCT to
Glaucoma Tube Shunt Revision Surgery
assess patch graft stability and hostetissue integration prospectively.
Samuel Weinreb, AB, Nur Cardakli, AB, Joan Jefferys, MS, Harry Main Outcome Measures: Tube exposure, graft thinning, and
Quigley, MD graft-related complications.
Purpose: To characterize long-term outcomes of tube shunt Results: A total of 81 eyes of 81 patients (50 women, 31 men)
revision surgeries and identify factors associated with their failure. with a mean age of 6713 years underwent GGD implantation
Design: Retrospective chart review. using Baerveldt (n ¼ 72) or Ahmed valve (n ¼ 9). Tubes were
Participants: One eye from each of 179 patients who under- inserted in the anterior chamber (n ¼ 71), sulcus (n ¼ 6), or pars
went tube shunt revision surgery at the Wilmer Eye Institute be- plana (n ¼ 4). Tube ligation was performed with Baerveldt GDD
tween 2004 and 2015 with a minimum follow-up of 6 weeks. along with fenestration (n ¼ 51) or orphan trabeculectomy (n ¼
Methods: Eligible eyes were identified from billing records and 21). Tubes were covered with AMeUC (n ¼ 41) or pericardium
data related to their care were extracted from electronic medical (n ¼ 40). The mean follow-up time was 298 months (range,
records. Eyes were analyzed in aggregate and by indication for 13e40 months). Tube exposure occurred in 1 eye (2%) in the
revision, including hypotony, high intraocular pressure (IOP), tube AMeUC group at 3 months and in 2 eyes (5%) in the pericardium
reposition, and tube exposure. group at 2 and 6 months (P ¼ 0.54). Sequential AS OCT showed
Main Outcome Measures: Surgical failure, defined as a need better hostetissue integration and significantly less graft thinning
for further tube shunt revision or other glaucoma surgery, unsat- in the AMeUC group. Early graft thinning ( 3 months) occurred
isfactory IOP at last follow-up, or both. Secondary outcomes in 5 eyes (12%) in the AMeUC group and in 17 eyes (43%) in the
included postoperative infection and functional visual impairment. pericardium group (P ¼ 0.002). Late thinning occurred in 2 eyes
Results: With a median follow-up of 4.2 years, 126 failures (5%) and 11 eyes (28%) in the AMeUC and pericardium groups,
occurred among 179 eyes. By Kaplan-Meier analysis, the cumu- respectively (P ¼ 0.007). Graft translucency and cosmetic
lative rates of surgical failure at 1, 2, and 5 years after revision were appearance of the AMeUC graft were superior to those of the
49%, 59%, and 74%, respectively. Most revision failures (105/126) pericardium graft. No evidence of graft rejection or infection was
were the result of the need for additional surgery, whereas 11 eyes associated with the patch grafts in either group.
showed IOP above target levels and 10 eyes showed dysfunc- Conclusions: Amniotic membraneeumbilical cord grafts are
tionally low IOP at the last follow-up. Factors associated with well tolerated and offer an alternative to pericardium for safe and
failure in a stepwise regression model were revision for hypotony stable tube shunt coverage. Its high-tensile strength, low immu-
(hazard ratio [HR], 6.79; P ¼ 0.002), different surgeons perform- nogenicity, and excellent hostetissue integration significantly
ing the original and revision surgeries (HR, 2.80; P ¼ 0.002), reduced graft thinning.
longer duration of symptoms before revision (P ¼ 0.01), revision Ophthalmology Glaucoma 2019;2:392-401.
of a right eye (HR, 1.92; P ¼ 0.03), and presumed preoperative
infection (HR, 2.47; P ¼ 0.04). In univariate analysis, success
varied significantly by prior surgeon (P ¼ 0.01), but not by revi-
Micropulse Transscleral Diode Laser
sion surgeon. There was a 16% cumulative incidence of post- Cyclophotocoagulation in Refractory
operative infection, with the highest risk in those with presumed Glaucoma: Short-Term Efficacy, Safety, and
preoperative infections (P ¼ 0.01) and in persons of non-African, Impact of Surgical History on Outcomes
non-European derivation (P ¼ 0.03). Giancarlo A. Garcia, MD, Christine V. Nguyen, MD, Aleksandr
Conclusions: The estimated rate of failure of tube shunt revi- Yelenskiy, MD, Goichi Akiyama, MD, Brett McKnight, MD, Vikas
sion is 75% by 5 years, most often because of a need for further Chopra, MD, Kenneth Lu, MD, Alex Huang, MD, PhD, James
surgery. The major potentially modifiable feature associated with C.H. Tan, MD, PhD, Brian A. Francis, MD, MS
success is tube shunt revision being performed by the original
surgeon. Purpose: To assess the short-term efficacy and safety of
Ophthalmology Glaucoma 2019;2:383-391. micropulse transscleral diode laser cyclophotocoagulation
(MP-TSCPC) in the management of refractory glaucoma and to
Outcomes of the Shunt Tube Exposure compare outcomes based on prior glaucoma surgeries.
Design: Retrospective analysis.
Prevention Study: A Randomized Clinical
Participants: Patients with refractory glaucoma who under-
Trial went MP-TSCPC at a single institution by 1 of 4 surgeons.
Hosam Sheha, MD, PhD, Celso Tello, MD, FACS, Lama A. Methods: Chart review of cases of MP-TSCPC using the Iridex
Al-Aswad, MD, MPH, Mohamed S. Sayed, MD, Richard K. Lee, Cyclo G6 (Mountain View, CA) laser with standard parameters
MD, PhD and laser duration at the discretion of each treating physician.

8
 Main Outcome Measures: Probability of postoperative suc- processes. By comparison, the TCP specimens showed erratic
cess was estimated by the Kaplan-Meier method. Success param- treatment of the ciliary processes with overlap into the pars plicata
eters included intraocular pressure (IOP) 6 to 21mmHg with or of the ciliary body. Microscopic analysis revealed significant
without topical antihypertensive therapy, 20% or more IOP disruption of the cells throughout the ciliary processes and loss of
vessels within the stroma.
reduction from baseline for any 2 consecutive visits after 3 post-
Conclusions: Treatment with ECP results in less overall
operative months, and no subsequent glaucoma surgery. tissue destruction and a targeted effect on the pigmented
Results: One hundred sixteen eyes of 116 patients (mean age, ciliary epithelium of the ciliary processes when compared with
65.8_16.9 years) were included. Baseline IOP was 22.27.9 TCP.
mmHg, and mean postoperative follow-up time was 6.33.4 Ophthalmology Glaucoma 2019;2:413-421.
months (range, 3e12 months.) Postoperative IOP at the final
follow up was 15.36.6 mmHg (P < 0.01), corresponding to a
Machine Learning Models for Diagnosing
reduction of approximately 6.9 mmHg (31.1%). Most eyes (66.4%)
underwent at least 6 months of follow-up. Short-term probability of Glaucoma from Retinal Nerve Fiber Layer
success was 93.1% at 3 months and 74.3% at 6 months. Eyes that Thickness Maps
had undergone prior traditional glaucoma surgery (trabeculectomy, Peiyu Wang, MS, Jian Shen, MS, Ryuna Chang, BS, Maemae
tube shunt, excessive pressure-regulating shunt system miniature Moloney, BS, Mina Torres, PhD, Bruce Burkemper, PhD, MPH,
glaucoma shunt [Alcon, Fort Worth, TX], or a combination Xuejuan Jiang, PhD, Damien Rodger, MD, PhD, Rohit Varma, MD,
thereof) demonstrated a higher probability of success (67.6%) MPH, Grace M. Richter, MD, MPH
compared with eyes that had not (41.4%; P ¼ 0.014). The most
common complications were decline in best-corrected visual acuity Purpose: To assess the diagnostic accuracy of multiple ma-
(7.8%) and hypotony (1.7%). chine learning models using full retinal nerve fiber layer (RNFL)
Conclusions: Micropulse transscleral diode laser cyclo- thickness maps in detecting glaucoma.
photocoagulation has a significant short-term ocular hypotensive Design: Case-control study.
effect and favorable safety profile in eyes with refractory glau- Participants: A total of 93 eyes from 69 patients with glau-
coma. The probability of successful outcome was greater in eyes coma and 128 eyes from 128 age- and sex-matched healthy con-
that had undergone prior traditional glaucoma surgery. trols from the Los Angeles Latino Eye Study (LALES), a large
Ophthalmology Glaucoma 2019;2:402-412. population-based, longitudinal cohort study consisting of Latino
participants aged 40 years residing in El Puente, California.
Methods: The 66-mm RNFL thickness maps centered on the
Endoscopic Cyclophotocoagulation and optic nerve head (Cirrus 4000; Zeiss, Dublin, CA) were supplied to
Other Cyclodestructive Methods: 4 different machine learning algorithms. These models included 2
Histopathologic Comparison of In Vivo conventional machine learning algorithms, Support Vector Ma-
Treatment in Humans and Monkeys chine (SVM) and K-Nearest Neighbor (KNN), and 2 convolutional
Brian A. Francis, MD, MS, Brian Flowers, MD, Anna Dastiridou, neural nets, ResNet-18 and GlaucomaNet, which was a custom-
MD, Aleksandr Yelenskiy, MD, Vikas Chopra, MD, Jorge A. made deep learning network. All models were tested with 5-fold
Alvarado, MD cross validation.
Main Outcome Measures: Area under the curve (AUC) sta-
Purpose: To compare the histologic effects of endoscopic tistics to assess diagnostic accuracy of each model compared with
cyclophotocoagulation (ECP) with other ciliary body ablative conventional average circumpapillary RNFL thickness.
procedures. A secondary aim was to correlate these findings with
Results: All 4 models achieved similarly high diagnostic ac-
historical clinical success and complication rates.
Design: Prospective, qualitative comparison of histopathologic curacies, with AUC values ranging from 0.91 to 0.92. These values
tissue analysis. were significantly higher than those for average circumpapillary
Participants: Two eyes of two patients who had undergone RNFL thickness, which had an AUC of 0.76 in the same patient
ECP for open-angle glaucoma were studied. Two eyes from a population.
healthy monkey were treated with ECP and studied. For compar- Conclusions: Superior diagnostic performance was achieved
ison, 1 eye each of patients who had undergone contact and with both conventional machine learning and convolutional neural
noncontact neodymium: yttriumealuminumegarnet transscleral net models compared with circumpapillary RNFL thickness. This
cyclophotocoagulation (TCP) for open-angle glaucoma were supports the importance of the spatial structure of RNFL thickness
analyzed. map data in diagnosing glaucoma and further efforts to optimize
Methods: The human globes were enucleated and submitted for
our use of this data.
analysis by gross examination and light and electron microscopy.
Monkey eyes were studied by gross examination and light mi- Ophthalmology Glaucoma 2019;2:422-428.
croscopy 1 week, 2 weeks, and 1 month after ECP.
Main Outcome Measures: Gross and histopathologic speci-
mens of all eyes were analyzed with respect to location and type of
Reproducibility of 5 Methods of Ocular
tissue effects and degree of collateral damage. Tonometry
Results: The gross analysis of all ECP specimens showed a Anne E. Kutzscher, BA, Rajesh S. Kumar, MBBS, MS, B. Ramgo-
uniform anterior-to-posterior whitening of each treated ciliary pal, MBBS, MS, Mahalakshmi V. Rackenchath, MBBS, DO, Sathi
process. This correlated with microscopic evidence of loss of Devi, MBBS, DOMS, Sriharsha Nagaraj, MBBS, MPH, Caitlin A.
pigmentation from the pigmented ciliary epithelial cells seen with Moe, MPH, Dionna M. Fry, MPH, Robert L. Stamper, MD, Jeremy
mostly preserved architecture and vasculature of the ciliary D. Keenan, MD, MPH

9
 Purpose: To evaluate the agreement between Goldmann Purpose: To compare the agreement in intraocular pressure
applanation tonometry (GAT; Haag-Streit, Bern, Switzerland) and (IOP) measurements obtained with the Tono-Pen (Reichert, Inc,
several portable tonometers that could be used for glaucoma Depew, NY) and Goldmann applanation tonometry (GAT). The
screening programs. influence of central corneal thickness (CCT) on IOP agreement
Design: Evaluation of a diagnostic test.
between the 2 tonometry methods also was evaluated.
Participants: Three hundred twenty-one eyes of 168 partici-
pants seeking treatment at the glaucoma clinic of a tertiary eye Design: Database study.
hospital in India. Participants: A total of 898 patients from the Research,
Methods: Participants underwent intraocular pressure (IOP) Innovation and Experimentation Database spanning 1999
measurement with GAT and 4 index tests: the Icare TA01i rebound through 2016.
tonometer (Icare USA, Raleigh, NC), a noncontact tonometer Methods: A total of 898 IOP measurements of right eye ob-
(NCT; Topcon CT-80; Topcon; Tokyo, Japan), a pneumaton- tained with GAT and the Tono-Pen were extracted from a glau-
ometer (Model 30; Reichert Technologies; Depew, NY), and the coma, glaucoma suspect, and ocular hypertension clinical database.
Tono-Pen AVIA (Reichert Technologies). The agreement between IOP measurements obtained by the 2
Main Outcome Measures: Estimates of reproducibility of IOP methods was analyzed using a Bland-Altman plot in Microsoft
measurements between each index test and GAT, including the
Excel (Microsoft, Redmond, WA). Their relationship to CCT was
intraclass correlation coefficient (ICC) and the Bland-Altman 95%
limits of agreement. evaluated using linear regression analysis.
Results: The Icare showed the greatest agreement with GAT Main Outcome Measures: Agreement between GAT and
(ICC, 0.69; 95% confidence interval, 0.61e0.74), followed by the Tono-Pen measurements of IOP and influence of CCT on tonom-
NCT (ICC, 0.65; 95% confidence interval, 0.58e0.71), the Tono- etry methods.
Pen (ICC, 0.51; 95% confidence interval, 0.43e0.59), and the Results: The correlation between both tonometry devices was
pneumatonometer (ICC, 0.36; 95% confidence interval, 0.76 (P < 0.001). The mean difference (Tono-Pen minus GAT) in
0.27e0.44). The pneumatonometer and Tono-Pen tended to IOP measurement was e0.15 mmHg (95% confidence limits,
overestimate IOP relative to GAT, with a mean difference of 3.4 6.8 mmHg). The Tono-Pen underestimated GAT measurements
mmHg (95% limits of agreement [LOA], e7.3 to 14.1 mmHg) for at IOP of more than 16.8 mmHg and overestimated at IOP of less
the pneumatonometer and 3.2 mmHg (95% LOA, e6.1 to 12.6
than 16.8 mmHg. Larger differences were associated with higher
mmHg) for the Tono-Pen. In contrast, measurements from the Icare
and NCT were on average within 1 point of those for GAT (mean IOPs. An average difference of more than 3 mmHg by the Tono-
difference, e0.4 mmHg [95% LOA, e8.4 to 7.6 mmHg] and e0.5 Pen was observed for IOPs of 29 mmHg or more. A thicker
mmHg [95% LOA, e8.7 to 7.6 mmHg], respectively). cornea also was associated with higher IOP. A change of 0.16
Conclusions: The Icare and NCT both demonstrated good mmHg for a 10-mm increase in CCT was appreciated for GAT,
agreement with GAT across a wide range of IOPs and could be and one of 0.15 mmHg per 10 mm was appreciated for the Tono-
considered for glaucoma screening programs. Pen (all P < 0.001).
Ophthalmology Glaucoma 2019;2:429-434. Conclusions: In a large cohort of more than 898 measure-
ments, the average difference between IOP measured by GAT and
Tono-Pen versus Goldmann Applanation the Tono-Pen was less than 1 mmHg, although the limits of
Tonometry: A Comparison of 898 Eyes agreement were wide. Higher IOPs were associated with larger
differences.
Bo Bao, MD, Vlad Diaconita, MD, David C. Schulz, MD, Cindy Ophthalmology Glaucoma 2019;2:435-439.
Hutnik, MD, PhD

10
 Editorial

Trachoma: Time to Talk Eradication

Thomas M. Lietman, MD - San Francisco, California
Catherine E. Oldenburg, PhD, MPH - San Francisco, California
Jeremy D. Keenan, MD, MPH - San Francisco, California

Trachoma was once the leading cause of blindness, endemic Inc. (New York, NY), have donated nearly 1 billion doses of
in nearly every country of the world.1 As recently as the azithromycin to the cause. After 20 years of distributions,
1990s, trachoma still ranked second only to cataract as a districts that remain endemic now can be divided into those
cause of blindness.2 Two decades ago, the World Health where infection will disappear regardless of any future
Organization (WHO) initiated a program to control the MDA, those where infection will disappear with continued
disease by 2020.3 Although that goal will not be met on annual MDA, and those where infection will not disappear
time, recent developments suggest that even loftier goals without more intensive intervention. The vast majority of
could be possible in the near future. Global eradication of endemic districts now fall into the first 2 categories.
the strains of Chlamydia that cause trachoma can be The WHO relies on the clinical signs of trachoma to
achieved, and sooner than previously thought, if we step declare that a district has obtained control.20 Specifically,
up interventions in the most affected areas. control requires reducing the prevalence of follicular
Most trachoma programs follow a WHO strategy that trachoma in the upper conjunctiva (TF) to be less than 5% in
aims for control, defined as bringing infection to a low children. But TF can linger long after infection has been
enough level that resulting disease is not a public health cleared. Thus, the community-level prevalence of TF is a
concern.4 Elimination indicates that infection, or at least lagging indicator.21 In communities with a low prevalence of
transmission of infection, is brought to zero in a TF, infection often is impossible to find even with polymerase
geographical area. Eradication chain reaction analysis. Moreover,
implies elimination of infection As infection is eliminated in more any association between TF and
worldwidedat least outside of actual infection decreases after
the laboratory. Smallpox remains countries, the argument that MDA.22 Previously endemic
the only infectious disease of eradication is impossible becomes countries, such as Nepal, Mexico,
humans eradicated by a public more difficult to make. Ghana, Uganda, and the Gambia,
health program. A number of recently performed population-
other infectious diseases are now on the ropes, including based surveys as part of their dossiers to declare trachoma
Guinea worm, polio, and onchocerciasis (river control. In each, polymerase chain reactionedetermined prev-
blindness).5e7 Trachoma has several characteristics that alence of chlamydia in children was less than the false-positive
make eradication at least feasible. Humans are the only host. rate of the test.2e4 Ironically, by the time a country has
Antibiotics are effective against Chlamydia, and no antibiotic been certified as controlled, elimination already may have
resistance to azithromycin has yet emerged. Perhaps most occurred.
importantly, trachoma benefits from an enormous secular What about districts where more intensive intervention will
trend. Trachoma disappeared in many regions without the be required? Approximately 30 districts in the Amhara region
benefit of specific trachoma programs. Where monitored of Ethiopia remain far above control targets after a decade of
longitudinally without active intervention, trachoma seems to MDA.23 Perhaps an equal number of communities outside of
be disappearing. Many agree that trachoma eventually will Amhara will prove similar. Circumstantial evidence has
disappear with or without a public health program, although suggested that water, sanitation, and hygiene measures could
in the latter case, perhaps not for decades.8e13 be complementary to MDA.24 Unfortunately, no water,
The cornerstone of the WHO program is mass treatment sanitation, and hygiene intervention has ever been shown to
with a single-dose of oral azithromycin. Azithromycin was have any measurable effect on ocular chlamydial infectiond
shown to be effective in clearing ocular chlamydial infection not face washing, not latrine construction, and not water
from most individuals.14,15 Three weekly mass drug programs.24,25 In fact, the only intervention ever proven to
administrations (MDAs) appeared to be as effective as reduce infection more than annual MDA is more frequent
6 weeks of the then standard-of-care topical tetracycline.16 MDA. Biannual distributions may reduce infection more
Subsequent community-randomized trials confirmed efficacy rapidly than annual distributions and have completely
of mass distribution with a single dose of azithromycin.17,18 eliminated infection from some of the most severely affected
Mathematical models suggested that annual distribution communities ever studied.26e29 Quarterly distributions to
eventually could eliminate infection in most communities children were proven superior to annual MDA in a community-
worldwide, although some may require more frequent randomized trial.18 That study was designed as a proof of
19
treatment. The International Trachoma Initiative and Pfizer, principle, to show that elimination of infection in children

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.11.001 11

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

reduced infection in adults through an indirect, herd-like effect. a population-based survey in Malawi. Ophthalmic Epidemiol.
At the time, no one realistically expected programs to distribute 2001;8(2e3):145e153.
antibiotics quarterly to the thousands of endemic districts. But 12. Jha H, Chaudary J, Bhatta R, et al. Disappearance of trachoma
now, quarterly distributions to the few remaining problem in western Nepal. Clin Infect Dis. 2002;35(6):765e768.
districts may be more palatable than a second decade of annual 13. House J, Gaynor B, Taylor H, Lietman TM. The real chal-
lenge: can we discover why trachoma is disappearing before
MDA. Eliminating chlamydia as quickly as possible also may it’s gone? Int Ophthalmol Clin. 2007;47(3):63e76.
be the optimal way to address concerns about antimicrobial 14. Dawson CR, Schachter J, Sallam S, et al. A comparison of oral
resistance and program fatigue. azithromycin with topical oxytetracycline/polymyxin for the treat-
Why not more enthusiasm for trachoma eradication? As ment of trachoma in children. Clin Infect Dis. 1997;24(3):363e368.
infection is eliminated in more countries, the argument that 15. Bailey RL, Arullendran P, Whittle HC, Mabey DC. Rando-
eradication is impossible becomes more difficult to make. mised controlled trial of single-dose azithromycin in treatment
However, some argue that eradication is not necessary, that a of trachoma. Lancet. 1993;342(8869):453e456.
low level of infection will never cause enough scarring to lead 16. Schachter J, West SK, Mabey D, et al. Azithromycin in control
to blindness. That may well be true. But if infection is not of trachoma. Lancet. 1999;354(9179):630e635.
eliminated in the problem areas, then the only proven way to 17. Chidambaram JD, Alemayehu W, Melese M, et al. Effect of a
single mass antibiotic distribution on the prevalence of infec-
maintain infection at a safe low level is continued MDA.30 The tious trachoma. JAMA. 2006;295(10):1142e1146.
enthusiasm for another decade or more of annual MDA may 18. House JI, Ayele B, Porco TC, et al. Assessment of herd pro-
wane, and cessation of the program could result in resurgence tection against trachoma due to repeated mass antibiotic dis-
of infection. Although infection eventually will disappear on tributions: a cluster-randomised trial. Lancet. 2009;373(9669):
its own, it is hard to blame WHO for not pushing a target of 1111e1118.
eradication. Failures to reach overly ambitious goals for 19. Lietman T, Porco T, Dawson C, Blower S. Global elimination
malaria and leprosy are still remembered. But WHO and of trachoma: how frequently should we administer mass
fellow stakeholders may now believe that they face a difficult chemotherapy? Nat Med. 1999;5(5):572e576.
decision: continue annual MDA for another decade and hope 20. Thylefors B, Dawson CR, Jones BR, et al. A simple system for
for the best, or intensify efforts in problem areas to accelerate the assessment of trachoma and its complications. Bull World
Health Org. 1987;65(4):477e483.
the disappearance, and be the first to eradicate a bacterial 21. Keenan JD, Lakew T, Alemayehu W, et al. Slow resolution of
disease? For us, that’s an easy choice. clinically active trachoma following successful mass antibiotic
treatments. Arch Ophthalmol. 2011;129(4):512e513.
References 22. Keenan JD, See CW, Moncada J, et al. Diagnostic character-
istics of tests for ocular Chlamydia after mass azithromycin
distributions. Invest Ophthalmol Vis Sci. 2012;53(1):235e240.
1. Trachoma Taylor HR. A blinding scourge from the bronze age to 23. Nash SD, Stewart AEP, Zerihun M, et al. Ocular Chlamydia tra-
the twenty-first century. Centre for Eye Research Australia; 2008. chomatis infection under the Surgery, Antibiotics, Facial Clean-
2. Thylefors B, Négrel AD, Pararajasegaram R, Dadzie KY. liness, and Environmental Improvement Strategy in Amhara,
Global data on blindness. Bull World Health Org. 1995;73(1): Ethiopia, 2011e2015. Clin Infect Dis. 2018;67(12):1840e1846.
115e121. 24. Emerson PM, Cairncross S, Bailey RL, Mabey DC. Review of the
3. World Health Organization. Report of the Third Meeting of the evidence base for the ‘F’ and ‘E’ components of the SAFE
W.H.O. Alliance for the Global Elimination of Trachoma. strategy for trachoma control. Trop Med Int Health. 2000;5(8):
Ouarzazate, Morocco: World Health Organization; 1999. 515e527.
WHO/PBLGET/99.3 English. 25. Stoller NE, Gebre T, Ayele B, et al. Efficacy of latrine pro-
4. Dowdle WR. The principles of disease elimination and erad- motion on emergence of infection with ocular Chlamydia
ication. Bull World Health Org. 1998;76(Suppl 2):22e25. trachomatis after mass antibiotic treatment: a cluster-
5. Hopkins DR, Ruiz-Tiben E, Weiss AJ, et al. Progress toward randomized trial. Int Health. 2011;3(2):75e84.
global eradication of dracunculiasisdJanuary 2017eJune 2018. 26. Gebre T, Ayele B, Zerihun M, et al. Comparison of annual
MMWR Morb Mortal Wkly Rep. 2018;67(45):1265e1270. versus twice-yearly mass azithromycin treatment for hyper-
6. Sauerbrey M, Rakers LJ, Richards FO. Progress toward endemic trachoma in Ethiopia: a cluster-randomised trial.
elimination of onchocerciasis in the Americas. Int Health. Lancet. 2012;379(9811):143e151.
2018;10(Suppl 1):i71ei78. 27. Melese M, Alemayehu W, Lakew T, et al. Comparison of
7. Moffett DB, Llewellyn A, Singh H, et al. Progress toward polio- annual and biannual mass antibiotic administration for elimi-
virus containment implementationdworldwide, 2018e2019. nation of infectious trachoma. JAMA. 2008;299(7):778e784.
MMWR Morb Mortal Wkly Rep. 2019;68(38):825e829. 28. Biebesheimer JB, House J, Hong KC, et al. Complete local
8. Keenan JD, Hotez PJ, Amza A, et al. Elimination and eradication elimination of infectious trachoma from severely affected
of neglected tropical diseases with mass drug administrations: a communities after six biannual mass azithromycin distribu-
survey of experts. PLoS Negl Trop Dis. 2013;7(12):e2562. tions. Ophthalmology. 2009;116:2047e2050.
9. Chidambaram JD, Bird M, Schiedler V, et al. Trachoma 29. Gill DA, Lakew T, Alemayehu W, et al. Complete elimination
decline and widespread use of antimicrobial drugs. Emerg is a difficult goal for trachoma programs in severely affected
Infect Dis. 2004;10(11):1895e1899. communities. Clin Infect Dis. 2008;46:564e566.
10. Dolin PJ, Faal H, Johnson GJ, et al. Reduction of trachoma in 30. Keenan JD, Tadesse Z, Gebresillasie S, et al. Mass azi-
a sub-Saharan village in absence of a disease control pro- thromycin distribution for hyperendemic trachoma following a
gramme. Lancet. 1997;349(9064):1511e1512. cluster-randomized trial: a continuation study of randomly
11. Hoechsmann A, Metcalfe N, Kanjaloti S, et al. Reduction of reassigned subclusters (TANA II). PLoS Med. 2018;15(8):
trachoma in the absence of antibiotic treatment: evidence from e1002633.

12
 Editorial

Footnotes and Financial Disclosures

Financial Disclosure(s): The author(s) have no proprietary or commercial Correspondence:
interest in any materials discussed in this article. Thomas M. Lietman, MD, University of California at San Francisco, Room
Supported by the National Institutes of Health, Bethesda, Maryland (grant S309, 513 Parnassus Ave, San Francisco, CA 94143. E-mail: tom.lietman@
nos.: UG1Y028088, R01 EY025350, and U10 EY023939); and Research to ucsf.edu.
Prevent Blindness, Inc., New York, New York.

13
Topical Recombinant Human Nerve Growth
Factor (Cenegermin) for Neurotrophic
Keratopathy
A Multicenter Randomized Vehicle-Controlled Pivotal Trial
Stephen C. Pflugfelder, MD,1 Mina Massaro-Giordano, MD,2 Victor L. Perez, MD,3 Pedram Hamrah, MD,4
Sophie X. Deng, MD, PhD,5 Ladan Espandar, MD, MS,6,7 C. Stephen Foster, MD,8,9 John Affeldt, MD,10
John A. Seedor, MD,11 Natalie A. Afshari, MD,12 Wendy Chao, PhD,13 Marcello Allegretti, PhD,13
Flavio Mantelli, MD, PhD,13 Reza Dana, MD, MPH9,14

Purpose: To evaluate the efficacy and safety of topical cenegermin (recombinant human nerve growth factor)
in patients with neurotrophic keratopathy.
Design: Multicenter, randomized, double-masked, vehicle-controlled trial.
Participants: Patients with neurotrophic persistent epithelial defect with or without stromal thinning.
Methods: The NGF0214 trial, conducted among 11 sites in the United States, randomized 48 patients 1:1 to
cenegermin 20 mg/ml or vehicle eye drops, 6 drops daily for 8 weeks of masked treatment. Follow-up was 24
weeks. Safety was assessed in all patients who received study drug. Efficacy was assessed by intention to treat.
Main Outcome Measures: The primary end point was healing of the neurotrophic lesion (persistent epithelial
defect or corneal ulcer) after 8 weeks of masked treatment. Masked central readers measured neurotrophic le-
sions in randomized clinical pictures, then assessed healing status conventionally (<0.5 mm of fluorescein
staining in the greatest dimension of the lesion area) and conservatively (0-mm lesion staining and no other re-
sidual staining). Secondary variables included corneal healing at 4 weeks of masked treatment (key secondary
end point), overall changes in lesion size, rates of disease progression, and changes in visual acuity and corneal
sensitivity from baseline to week 8.
Results: Conventional assessment of corneal healing showed statistically significant differences at week 8:
compared to 7 of 24 vehicle-treated patients (29.2%), 16 of 23 cenegermin-treated patients (69.6%) achieved less
than 0.5 mm of lesion staining (þ40.4%; 95% confidence interval [CI], 14.2%e66.6%; P ¼ 0.006). Conservative
assessment of corneal healing also reached statistical significance at week 8: compared to 4 of 24 vehicle-treated
patients (16.7%), 15 of 23 cenegermin-treated patients (65.2%) achieved 0 mm of lesion staining and no other
residual staining (þ48.6%; 95% CI, 24.0%e73.1%; P < 0.001). Moreover, the conservative measure of corneal
healing showed statistical significance at week 4 (key secondary end point). Compared to vehicle, cenegermin-
treated patients showed statistically significant reductions in lesion size and disease progression rates during
masked treatment. Cenegermin was well tolerated; adverse effects were mostly local, mild, and transient.
Conclusions: Cenegermin treatment showed higher rates of corneal healing than vehicle in neurotrophic
keratopathy associated with nonhealing corneal defects. Ophthalmology 2020;127:14-26 ª 2019 by the American
Academy of Ophthalmology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.
org/licenses/by-nc-nd/4.0/).

Supplemental material available at www.aaojournal.org.

Ocular surface integrity relies on the corneal nerves, which estimated prevalence of 1.6 to 4.2 cases per 10 000
not only enable sensory-mediated reflexes (such as blinking persons.1,2 Various conditions (such as ocular herpetic
and tearing), but also mediate production of trophic factors infection, ocular or neurologic surgery, trauma, diabetes,
that critically help maintain corneal epithelium and the and dry eye disease) are associated with neurotrophic
nerves themselves.1 Corneal nerve damage may disrupt this keratopathy3,4; however, correlations between underlying
homeostasis and lead to neurotrophic keratopathy (also etiologies and neurotrophic keratopathy severity or clinical
known as neurotrophic keratitis or neurotrophic outcomes (such as visual acuity and corneal sensitivity)
keratoconjunctivitis), a degenerative disease with an remain unclear.5

14 ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.08.020

This is an open access article under the CC BY-NC-ND license ISSN 0161-6420/19
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Published by Elsevier Inc.
 Pflugfelder et al 
Cenegermin for Neurotrophic Keratopathy

Clinical interventions for neurotrophic keratopathy (BCDVA) score of 75 Early Treatment Diabetic Retinopathy Study
(reviewed elsewhere)2,4,6 vary widely and generally are (ETDRS) letters or fewer (þ0.2 logarithm of the minimum angle
based on disease severity. Until recently, neurotrophic ker- of resolution, 20/32 Snellen, or 0.625 decimal fraction) in the
atopathy interventions have been limited to nonstandardized study eye; and no objective clinical evidence of improvement of
the neurotrophic lesion within 2 weeks preceding study enrollment.
treatments, such as preservative-free artificial tears, serum
Patients with stage 3 neurotrophic keratopathy that featured severe
drops, and therapeutic lenses. Surgical interventions (e.g., thinning (beyond two thirds of the cornea) and impending perfo-
tarsorrhaphy, amniotic membrane transplantation, conjunc- ration were not included in the study. The other main exclusion
tival flap, and corneal transplantation) generally are reserved criteria were active ocular infection or inflammation unrelated to
for refractory cases but tend to be invasive and themselves neurotrophic keratopathy and other ocular disease or severe vision
may compromise vision. loss in the affected eye(s). Unless it was the underlying etiology of
Cenegermin 0.002% ophthalmic solution (Oxervate; neurotrophic keratopathy, ocular surgery (including laser or
Dompé Farmaceutici SpA, Milan, Italy) recently received refractive surgical procedures) was not allowed within 3 months
approval from the European Commission, the United States before study enrollment. Ocular surgery also was not allowed
Food and Drug Administration, and other health authorities for during the study treatment period, and elective ocular surgery
procedures could not be planned for the follow-up period. For
the treatment of neurotrophic keratopathy. Developed based on
complete inclusion and exclusion criteria, see Appendix 2
studies with murine nerve growth factor (NGF) that showed (available at www.aaojournal.org).
promise in treating corneal neurotrophic ulcers,7,8 cenegermin
is a recombinant human NGF (rhNGF) produced in Escher-
ichia coli. Phase 1 randomized, double-masked, vehicle- Treatment
controlled studies showed topical cenegermin treatment to be
safe in healthy volunteers9 and in patients with neurotrophic The test product (cenegermin 20 mg/ml) or reference product
keratopathy.10 The approved indication for cenegermin was (vehicle), each containing methionine as an antioxidant excipient,
based on the results of NGF0212/REPARO (Latin for was provided by the sponsor and was self-administered by patients
"repair"), a phase 2, randomized, double-masked, vehicle- at a dosage of 1 eye drop (approximately 35 ml) 6 times daily. The
duration of masked treatment was 8 weeks. Of patients randomized
controlled, dose-ranging study conducted in Europe and re- to vehicle treatment, those who did not achieve corneal healing
ported previously,11 and NGF0214, a pivotal trial conducted in (<0.5 mm of fluorescein staining in the greatest dimension of the
the United States. Both trials demonstrated the efficacy and lesion area) by the end of masked treatment were eligible (at the
safety of cenegermin in promoting corneal healing in investigator’s discretion) to receive cenegermin during an 8-week
patients with neurotrophic keratopathy associated with open-label treatment period. If any patients healed under masked
persistent epithelial defect (with or without stromal thinning). or open-label cenegermin treatment and then experienced disease
However, in REPARO, effects on other clinically relevant recurrence at any point during follow-up, they could receive open-
end points (such as corneal sensitivity, visual acuity, and label recurrence treatment with cenegermin for 8 weeks at the in-
disease progression) remained unclear. We sought to define vestigator’s discretion. Any patient could receive a maximum of 2
further the role of cenegermin on these and other clinical 8-week cenegermin treatment courses before continuing follow-up
for 24 weeks.
metrics in the NGF0214 study. Herein we report the results On enrollment, patients were required to discontinue use of
of this pivotal trial. all previous topical ophthalmic medications, bandage contact
lenses, or both. During either masked or open-label treatment
Methods periods, patients were limited to the study medication (provided
by the sponsor) and topical ophthalmic medications allowed by
the study protocol (not provided by the sponsor). Allowed
Clinical Trial Design medications included preservative-free topical antibiotics or
NGF0214 was a multicenter, double-masked, randomized, vehicle- topical antiviral medications prescribed at the investigator’s
controlled, parallel-group, pivotal trial that evaluated the efficacy discretion. For patients who achieved corneal healing (<0.5 mm
and safety of cenegermin 20-mg/ml eye drops in patients with of lesion staining) after either masked or open-label treatment,
neurotrophic keratopathy. Study sites, investigators, and co- investigators could prescribe preservative-free artificial tears as
investigators are listed in Appendix 1 (available at needed during follow-up.
www.aaojournal.org). Figure 1 depicts the overall design of the If, during the 8-week masked treatment phase, the investigator
trial, which was registered at ClinicalTrials.gov (identifier, considered any patient to be at imminent risk of deterioration
NCT02227147). (lesion size increase of 1 mm, progression to corneal melting or
perforation, BCDVA decrease of >5 ETDRS letters, or onset of
Patients infection), the patient could use preservative-free topical antibi-
otics, preservative-free topical antivirals (not provided by the
The study enrolled adult patients (18 years of age) with neuro- sponsor), or both in addition to the assigned study medication.
trophic keratopathy in one or both eyes. For patients with bilateral In patients who achieved corneal healing (<0.5 mm of lesion
neurotrophic keratopathy, the worse affected eye at baseline was staining), preservative-free artificial tears (not provided by the
designated the study eye. The main inclusion criterion was neu- sponsor) could be prescribed as needed at the investigator’s
rotrophic keratopathy classified as stage 2 (persistent epithelial discretion during the 24-week follow-up period. If any patients
defect) or stage 3 (corneal ulcer) according to published criteria12 randomized to cenegermin treatment did not achieve corneal
and refractory to 1 or more conventional nonsurgical treatments. healing (<0.5 mm of lesion staining) by the end of the 8-week
The other main inclusion criteria were decreased corneal masked treatment, they could receive (at the investigator’s
sensitivity within the corneal lesion and in at least 1 corneal discretion) any nonexperimental treatment for neurotrophic kerat-
quadrant outside the lesion; best-corrected distance visual acuity opathy through the 24-week follow-up period.

15
 Ophthalmology Volume 127, Number 1, January 2020

Figure 1. Diagram showing NGF0214 study design and overall patient disposition. The NGF0214 study screened 52 patients and enrolled 48 with neu-
rotrophic keratopathy of severity stage 2 (persistent epithelial defect) or stage 3 (corneal ulcer). Patients were randomized 1:1 to cenegermin (20 mg/ml) or
vehicle for an 8-week masked treatment period and 24 weeks of follow-up. *According to the study protocol, patients who did not heal during 8-week
masked vehicle treatment were eligible (at the investigator’s discretion) to receive cenegermin in an 8-week open-label treatment period before pro-
ceeding to follow-up. yOne patient gave informed consent and was randomized to cenegermin treatment but withdrew the next day. No study treatment was
administered, and no baseline measures for primary efficacy were recorded; thus, this patient was excluded from the last observation carried forward analyses.
z
After 8 weeks of either masked or open-label cenegermin treatment, patients who healed (<0.5 mm of lesion staining) and then experienced a recurrence of
persistent epithelial defect or corneal ulcer were eligible (at the investigator’s discretion) for one recurrence treatment course of 8 weeks. A maximum of 2
8-week cenegermin treatment courses could be administered to any patient before continuing follow-up for 24 weeks. Of patients originally randomized to
vehicle who then completed the open-label treatment period, 4 experienced recurrence; however, 1 patient continued to the end of follow-up without
receiving recurrence treatment, so only 3 patients entered the recurrence treatment period. xOne patient died 163 days after the last dose of the study drug;
the cause of death was unknown and was assessed as unrelated to study treatment.

Efficacy Assessments efficacy end point, corneal healing (yes or no) was assessed by
masked central readers according to a more conservative threshold
The primary efficacy end point was healing of the neurotrophic (0 mm of staining in the lesion area and no other persistent staining
lesion (persistent epithelial defect or corneal ulcer) after 8 weeks of outside of the lesion area).
masked treatment. Corneal healing (defined as <0.5 mm of fluo- A key prespecified secondary end point was corneal healing
rescein staining in the greatest dimension of the lesion area) was (yes or no) at week 4 of masked treatment based on masked central
assessed in clinical images as a yes-or-no binary variable by central readings. Other prespecified secondary end points included per-
readers at Cologne Ophthalmological Reading and Image Analysis centage change from baseline of lesion size (reading center mea-
Center (Cologne, Germany), who were masked to the randomized surements of greatest dimension of fluorescein staining); change in
treatment and image dates. As an additional prespecified primary BCDVA from baseline to week 8, measured at each visit in

16
 Pflugfelder et al 
Cenegermin for Neurotrophic Keratopathy

ETDRS letters at 4 m (13 feet); change in corneal sensitivity from could receive any nonexperimental treatment for neurotrophic
baseline to week 8, measured in each of 4 corneal quadrants keratopathy, and vehicle-treated patients were eligible for open-
outside the lesion area using the Cochet-Bonnet esthesiometer; label cenegermin treatment (Fig 1). Unmasking for remaining
Schirmer testing without anesthesia (wetting distance measured in patients was allowed only after database lock (after the last
millimeters at 5 minutes); percentage of patients experiencing randomized patient completed 4 weeks of follow-up) for final
deterioration (yes or no) from baseline to week 8; and duration of statistical analysis.
corneal healing in patients who achieved less than 0.5 mm of lesion Sample size calculations were based on interim results of the
staining (based on investigator’s assessment on slit lamp) after phase 1/2 REPARO trial,10,11 which estimated that 70% of
cenegermin treatment. cenegermin-treated patients (vs. 30% of vehicle-treated patients)
would achieve the primary efficacy end point of corneal healing
Safety Assessments (<0.5 mm of lesion staining) after 8 weeks of masked treatment.
According to these assumptions and presumed dropout rates be-
Safety exploratory end points were adverse events, ocular tolera- tween 10% and 20%, the study randomized 48 patients to yield at
bility, intraocular pressure, fundus ophthalmoscopy results, he- least 38 evaluable patients to achieve 80% power of detecting a
matologic analysis results, clinical chemistry analysis results, and 40% difference in the primary efficacy variable.
immunogenicity results. All adverse events reported herein were There were no planned interim analyses or multiple compari-
treatment emergent (i.e., conditions that either arose or worsened in sons for this study; thus, the primary efficacy end point of corneal
intensity or frequency after initiation of study treatment). An healing after 8 weeks of masked treatment (assessed as a binary
adverse event was considered treatment related if its relationship to yes-or-no variable by masked central readers as described previ-
the study drug was recorded as possible, probable, or highly ously) was analyzed by 22 chi-square testing and reported with a
probable. If a relationship was missing, the adverse event also was 95% confidence interval (CI). Analyses were performed on the
considered to have a possible relationship to study treatment. intention-to-treat population with missing assessments of corneal
Serious adverse events were events that were life threatening or healing (yes or no) imputed based on the last observation carried
resulted in death, initiation or prolongation of hospitalization, forward. A sensitivity analysis was performed using nonresponder
persistent or significant incapacity, substantial disruption of activ- imputation, which considers all missing observations as failures,
ities of daily living, or a congenital anomaly or birth defect. and subjected to the chi-square test and Fisher exact test (1- and 2-
Adverse events were analyzed during each study period and sided). Additional sensitivity analyses included observed-case
summarized by preferred term and coded system organ class ac- analysis (removing patients who discontinued before week 4 of
cording to the Medical Dictionary for Regulatory Activities, masked treatment) and multiple imputation procedures MI and
version 19.0. Ocular tolerability was recorded by patients on a MIANALYZE in SAS software (SAS Institute, Cary, NC). Base-
visual analog scale and calculated as described previously.11 line variables, such as treatment assignment, demographics (e.g.,
Intraocular pressure was measured either by Goldmann age, gender), and neurotrophic keratopathy history (e.g., baseline
applanation tonometry or a handheld applanation tonometer (e.g., lesion size, time since diagnosis), were examined post hoc with
Tono-Pen, Reichert Technologies, Depew, NY) after instillation multiple logistic regression for potential effects on corneal healing.
of topical anesthesia. Dilated ophthalmoscopy examinations Changes from baseline in continuous variables, such as neu-
included the retina, macula, choroid, and optic nerve head. Blood rotrophic lesion size (greatest dimension of lesion staining in
samples were collected for hematologic, clinical chemistry, and millimeters), visual acuity (BCDVA in ETDRS letters), reflex
immunogenicity assessments (anti-NGF antibodies), which were tearing (Schirmer wetting distance in millimeters per 5 minutes),
performed as described previously9 and forwarded to a central and corneal sensitivity (Cochet-Bonnet esthesiometry in centime-
laboratory (Envigo, formerly Harlan Laboratories, ters), were analyzed as secondary and exploratory end points.
Cambridgeshire, United Kingdom) for masked analysis. Missing data were imputed using last postbaseline measurements
carried forward according to the statistical analysis plan.
Masking and Statistical Analysis For changes in corneal lesion size from baseline to week 4 and
week 8, reading center measurements of the greatest dimension of
Patients, investigators, and site or sponsor staff were masked to the fluorescein staining (last postbaseline measurement carried for-
randomized treatment. Kits for dispensing study treatment (cen- ward) were summarized using descriptive statistics according to the
egermin or vehicle) were identical in appearance and assigned statistical analysis plan. Reading center measurements at week 4
randomly according to codes generated in SAS statistical software and week 8 were assessed post hoc by an analysis of covariance
version 9.3 (SAS Institute, Cary, NC) by programmers not directly (ANCOVA) using treatment as a factor and both continuous (e.g.,
involved in study analysis. A contract research organization age) and categorical (e.g., gender) baseline variables. The contrast
maintained the masked database (Oracle Clinical software version of marginal linear predictions between treatment groups was re-
4.6.4, Oracle Corporation, Redwood Shores, CA) and performed ported with standard error, 95% CI, and P value.
prespecified statistical analyses. The sponsor was not involved in Change in BCDVA from baseline to week 8 was a prespecified
data collection for primary efficacy analyses, and central readers analysis by an ANCOVA model using randomized treatment group
were masked to randomized treatment assignment and duration of as a factor and controlling for baseline BCDVA score, time since
treatment (visit number). Immunologic data from the masked diagnosis of neurotrophic keratopathy (months), and baseline
central laboratory were provided to statisticians after official Schirmer values (millimeters). Changes in corneal lesion size were
unmasking of the study. According to protocol, unmasking was analyzed post hoc using an ANCOVA with randomized treatment
allowed if knowledge of the randomized treatment assignment was as a factor and baseline lesion size as a continuous covariate.
required to provide appropriate care for medical emergencies Deterioration (defined as decrease in BCDVA by >5 ETDRS
(including deterioration as defined previously). At the end of the 8- letters, onset of infection, or disease progression) was recorded as a
week masked treatment period, patients who did not achieve yes-or-no variable on the electronic case report form at each visit of
corneal healing (<0.5 mm of lesion staining) were unmasked and the masked treatment period and then analyzed by chi-square tests
eligible for treatment (at the investigator’s discretion) with medi- according to the statistical analysis plan. Disease progression
cations allowed by the study protocol: cenegermin-treated patients (defined as an increase in lesion size of 1 mm, progression in

17
 Ophthalmology Volume 127, Number 1, January 2020

lesion depth to corneal melting or perforation, or both), also other residual staining) yielded statistically significant differences
recorded on electronic case report forms, was assessed post hoc. between treatments at both time points (Fig 2B). At week 4,
The overall number of patients experiencing deterioration, disease corneal healing (0 mm of lesion staining and no other residual
progression, or both were based on tabulations of electronic case staining) was achieved in 5 of 24 patients receiving vehicle
report forms recorded on scheduled visits, then verified post hoc (20.8%), compared with 13 of 23 cenegermin-treated patients
with manual counts (including deterioration events recorded as (56.5%) who achieved healing (þ35.7%; 95% CI, 9.7%e61.7%;
adverse events during unscheduled visits) and analyzed post hoc by P ¼ 0.012). At week 8, 4 of 24 vehicle-treated patients (16.7%)
chi-square testing. achieved 0 mm of lesion staining and no other residual staining,
Corneal healing after open-label treatment and recurrence compared 15 of 23 cenegermin-treated patients (65.2%) who
treatment was assessed as less than 0.5 mm of lesion staining achieved corneal healing (þ48.6%; 95% CI, 24.0%e73.1%; P <
(according to slit-lamp assessment) and is presented using 0.001).
descriptive statistics. The duration of corneal healing (defined as Table 2 summarizes sensitivity analyses of the primary efficacy
time to intake of recurrence retreatment) also is presented with data using nonresponder imputation (all missing observations
descriptive statistics. imputed as failures) and analyses by 1- and 2-sided Fisher exact
tests in addition to chi-square testing. The results of nonresponder
Study Oversight imputation produced the same conclusions as last observation
carried forward imputation (Fig 2), which supports the robustness
Institutional review board approval was obtained from each partici- of the primary efficacy analyses at week 8 and key secondary
pating site (Appendix 1, available at www.aaojournal.org) for the analyses at week 4 using conventional and conservative
study protocol, amendments, and all study-related documents definitions of corneal healing. Observed-case and multiple impu-
(including informed consent forms). The study complied with the tation analyses produced similar results (not shown).
tenets of the Declaration of Helsinki, International Conference of To explore potential effects of baseline demographics and dis-
Harmonization Tripartite Guidelines for Good Clinical Practice, cur- ease characteristics, we performed multiple logistic regression on the
rent international and national regulations, the study protocol, and binary outcome of corneal healing (both conventional and conser-
respective standard operating procedures of the participating sites, vative definitions) at week 4 and week 8 of masked treatment (last
sponsor, and contract research organization. Written informed consent observation carried forward). Potential explanatory variables (in
was obtained before any study-related procedures, and study monitors addition to the randomized treatment) included demographic factors
verified compliance during onsite visits. (age, gender), disease characteristics (baseline lesion size, neuro-
trophic keratopathy severity, time since diagnosis), and underlying
etiologies (Table 1). Randomized treatment (cenegermin vs. vehicle)
Results and baseline lesion size (greatest dimension of fluorescein staining)
were variables with significant predictive values in logistic
Patients and Treatment regression models and are summarized in Table 3 with respective
Between May 1, 2015, and December 8, 2015, 52 patients from 11 odds ratios (ORs), CIs, and P values. Other potential variables in
sites (Appendix 1) provided informed consent and were screened our study population, such as underlying etiology (e.g., herpetic
for the study; 48 were enrolled and randomized 1:1 to receive vs. nonherpetic etiologies) and interactions between treatment and
active treatment or vehicle. Table 1 summarizes patient demographic variables, did not reach statistical significance in
demographics and baseline characteristics. Consistent with logistic regression (data not shown).
previous reports,3e5,8,10,11 the most common underlying etiol- Consistent with the primary efficacy analyses (Fig 2), the
ogies were herpetic eye disease (19 patients, including 2 with treatment variable (cenegermin vs. vehicle) showed a positive
multiple etiologies), ocular surgery (8 patients, including 1 with correlation overall with the outcome of corneal healing (OR, >1).
multiple etiologies), and dry eye disease (6 patients). Three For the conventional definition of corneal healing (<0.5 mm of
enrolled patients demonstrated bilateral neurotrophic keratopathy: lesion staining), the treatment variable did not reach statistical
2 randomized to cenegermin (dry eye disease and stem cell defi- significance at week 4 (OR, 3.13; 95% CI, 0.83e11.8; P ¼ 0.091),
ciency) and 1 randomized to vehicle (unknown etiology). Figure 1 but was significant at week 8 (OR, 7.31; 95% CI, 1.86e28.8; P ¼
presents an overview of patient disposition, including reasons for 0.004). For the conservative definition of corneal healing (0 mm of
withdrawal, recurrence treatment rates, and follow-up. lesion staining and no other persistent staining), the treatment
variable was a significant predictor of corneal healing at both week
Efficacy Outcomes 4 (P ¼ 0.006) and week 8 (P ¼ 0.001).
Conversely, baseline lesion size (greatest diameter of lesion
Figure 2 summarizes the primary efficacy analysis of corneal staining in millimeters) showed a negative association (OR, <1)
healing after 8 weeks of masked treatment, assessed by the with corneal healing (i.e., larger baseline lesion measurements
reading center (last observation carried forward). Also shown is predicted lower probability of corneal healing), independently of
the key secondary analysis of corneal healing at week 4, based randomized treatment. However, in our study population, this as-
on reading center assessments. One patient randomized to sociation was statistically significant at week 4 only; by the end of
cenegermin did not receive any baseline or postbaseline reading treatment (week 8), randomized treatment assignment (cenegermin
center assessments to carry forward and was excluded from these vs. vehicle) was the only significant predictor of corneal healing.
efficacy analyses. Based on our study population, there were no significant as-
The conventional definition of corneal healing (<0.5 mm of sociations among patient demographics, neurotrophic keratopathy
lesion staining) showed statistically significant differences between stage at baseline, or time since diagnosis. Although underlying
treatment groups at week 8 (Fig 2A), with healing in 7 of 24 etiologies did not show significant associations with healing
vehicle-treated patients (29.2%) compared with 16 of 23 outcome in logistic regression, the rarity of some etiologies in our
cenegermin-treated patients (69.6%) achieving healing (þ40.4%; patient population (such as corneal dystrophy, diabetes, and ocular
95% CI, 14.2%e66.6%; P ¼ 0.006). The more conservative surface injuries) preclude the ability to make conclusions on po-
definition of corneal healing (0 mm of lesion staining and no tential associations with clinical outcomes.

18
 Pflugfelder et al 
Cenegermin for Neurotrophic Keratopathy

Table 1. Baseline Demographics and Characteristics

Characteristics Cenegermin (N [ 24) Vehicle (N [ 24) Overall (N [ 48)

Age (yrs)
Mean (SD) 65.9 (13.85) 64.5 (14.15) 65.2 (13.87)
Median (minimumemaximum) 66.5 (33e94) 65.0 (35e92) 65.5 (33e94)
Age group (yrs), no. (%)
<50 2 (8.33) 3 (12.5) 5 (10.4)
50e64 9 (37.5) 8 (33.3) 17 (35.4)
65e74 7 (29.2) 7 (29.2) 14 (29.2)
75e84 4 (16.7) 4 (16.7) 8 (16.7)
85þ 2 (8.33) 2 (8.33) 4 (8.33)
Age 65 or older, no. (%) 13 (54.2) 13 (54.2) 26 (54.2)
Female, no. (%) 14 (58.3) 15 (62.5) 29 (60.4)
Ethnicity, no. (%)
Hispanic, Latino, or Spanish 0 1 (4.17) 1 (2.08)
Not Hispanic, Latino, or Spanish 20 (83.3) 19 (79.2) 39 (81.3)
N/A 4 (16.7) 4 (16.7) 8 (16.7)
Race, no. (%)
Asian 1 (4.17) 0 (0.0) 1 (2.08)
Black 3 (12.5) 2 (8.33) 5 (10.4)
Native Hawaiian or other Pacific Islander 0 (0.0) 1 (4.17) 1 (2.08)
White 20 (83.3) 20 (83.3) 40 (83.3)
Other 0 1 (4.17) 1 (2.08)
Primary neurotrophic keratopathy diagnosis, no. (%)
Stage 2 15 (62.5) 18 (75.0) 33 (68.8)
Stage 3 9 (37.5) 6 (25.0) 15 (31.3)
Bilateral neurotrophic keratopathy, no. (%) 2 (8.3) 1 (4.17) 3 (6.25)
Neurotrophic lesion size at baseline (maximum diameter; mm)*
Mean (SD) 3.51 (1.98) 3.02 (1.88) 3.26 (1.93)
Median (minimumemaximum) 3.10 (0.53e8.23) 2.99 (0.23e6.10) 3.10 (0.23e8.23)
Time since diagnosis of neurotrophic keratopathy (mos)
Mean (SD) 31.1 (108.34) 33.0 (73.83) 32.1 (91.72)
Median (minimumemaximum) 4.0 (0e535) 13.0 (0e366) 6.5 (0e535)
Time since diagnosis of neurotrophic keratopathy stage 2/3 (mos)
Mean (SD) 7.5 (14.51) 7.9 (8.59) 7.7 (11.80)
Median (minimumemaximum) 3.0 (0e71) 3.5 (0e28) 3.0 (0e71)
Underlying neurotrophic keratopathy etiology, no. (%)
Corneal dystrophy 1 (4.17) 1 (4.17) 2 (4.17)
Diabetes mellitus 0 1 (4.17) 1 (2.08)
Dry eye disease 3 (12.50) 3 (12.50) 6 (12.50)
Herpetic eye disease 9 (37.50) 8 (33.33) 17 (35.42)
Herpes simplex 5 (20.83) 4 (16.67) 9 (18.75)
Herpes zoster 2 (8.33) 4 (16.67) 6 (12.50)
Herpetic keratitis/unspecified 2 (8.33) 0 2 (4.17)
Multifactorial 2 (8.33) 0 2 (4.17)
Herpetic eye disease (herpes simplex); ocular surgery 1 (4.17) 0 1 (2.08)
(penetrating keratoplasty)
Herpetic eye disease (herpes zoster); diabetes mellitus 1 (4.17) 0 1 (2.08)
Neurosurgical procedure 1 (4.17) 1 (4.17) 2 (4.17)
Trigeminal ablation 1 (4.17) 0 1 (2.08)
Unspecified 0 1 (4.17) 1 (2.08)
Ocular surface injury/inflammation 2 (8.33) 1 (4.17) 3 (6.25)
Chemical burn/multiple transplant surgeries 1 (4.17) 0 1 (2.08)
Unspecified 1 (4.17) 1 (4.17) 2 (4.17)
Ocular surgery or procedure 3 (12.50) 4 (16.67) 7 (14.58)
Anterior and posterior surgeries 0 1 (4.17) 1 (2.08)
Cataract surgery 1 (4.17) 0 1 (2.08)
Conjunctival lesion excision 0 1 (4.17) 1 (2.08)
Unspecified 2 (8.33) 2 (8.33) 4 (8.33)
Other 2 (8.33) 5 (20.83) 7 (14.58)
Infectious keratitis (unspecified) 1 (4.17) 0 1 (2.08)
Radiation (unspecified) 0 1 (4.17) 1 (2.08)
Stem cell deficiency 1 (4.17) 0 1 (2.08)
Stevens-Johnson syndrome 0 1 (4.17) 1 (2.08)

(Continued)

19
 Ophthalmology Volume 127, Number 1, January 2020

Table 1. (Continued.)

Characteristics Cenegermin (N [ 24) Vehicle (N [ 24) Overall (N [ 48)

Trigeminal infiltration (metastatic cancer) 0 1 (4.17) 1 (2.08)
Unknown origin 0 1 (4.17) 1 (2.08)
Unspecified 0 1 (4.17) 1 (2.08)
Topical medication (glaucoma medication) 1 (4.17) 0 1 (2.08)

N/A ¼ not available (ethnicity and race were not reported for all patients); SD ¼ standard deviation.
*One patient randomized to the cenegermin group did not undergo a baseline measurement by the reading center, and so was excluded from the summary of
neurotrophic lesion size at baseline.

For changes in corneal lesion size from baseline, reading center values (millimeters), but did not reach statistical significance
measurements at weeks 4 and 8 of masked treatment (last post- (P ¼ 0.745). Visual acuity also was assessed as 15-letter gains (yes
baseline measurement carried forward) were analyzed using or no) at week 8 (last observation carried forward), which was
descriptive statistics according the study protocol and are presented achieved in 3 of 23 cenegermin-treated patients (13.0%) versus 4
alongside baseline measurements in Figure 3. Lesion of 24 vehicle-treated patients (16.7%). The difference between
measurements in cenegermin-treated patients seemed to be treatment groups was not statistically significant (P ¼ 0.727, chi-
reduced substantially from baseline compared with those of pa- square test).
tients receiving vehicle. Therefore, we conducted a post hoc Secondary efficacy variables included measurements of corneal
analysis of change in corneal lesion size from baseline using an sensitivity (measured within the lesion by Cochet-Bonnet esthesi-
ANCOVA with treatment (cenegermin or vehicle) as a factor and ometer) and reflex tearing (assessed using Schirmer testing). Mean
baseline lesion size as a continuous covariate. Cenegermin-treated Cochet-Bonnet esthesiometer measures at baseline were 0.81
patients showed significantly reduced lesion sizes compared with (standard deviation [SD], 1.187) in the cenegermin group versus
vehicle-treated patients; contrasts of marginal linear predictions 0.65 (SD, 0.699) in the vehicle group (n ¼ 24 per group); in
(cenegermin vs. vehicle treatment) were e1.146 at week 4 (stan- patients with available postbaseline observations, Cochet-Bonnet
dard error, 0.431; 95% CI, e2.018 to e0.274; P ¼ 0.011) and measurements were 2.91 (SD, 2.144) in cenegermin-treated
1.503 at week 8 (standard error, 0.448; 95% CI, 2.409e0.597; patients (n ¼ 18) versus 1.83 (SD, 1.952) in vehicle-treated
P ¼ 0.002). Effects of other baseline variables (age, gender, dis- patients (n ¼ 15) by week 8. Comparisons between treatment
ease stage, time since diagnosis, and etiology) were not significant groups did not reach statistical significance through an ANCOVA
in the ANCOVA model. performed with treatment as a factor and controlling for baseline
Visual acuity outcomes at week 8 (secondary efficacy end Cochet-Bonnet esthesiometer measures, Schirmer values, and time
point) were assessed using BCDVA in ETDRS letters. At baseline, since diagnosis of neurotrophic keratopathy (P ¼ 0.207).
mean BCDVA in patients randomized to cenegermin was 8.3 In Schirmer tear tests at week 8, cenegermin-treated patients
ETDRS letters (Snellen equivalent, 20/800), compared with 17.6 (n ¼ 17) showed a least-squares mean increase from baseline of 6.1
ETDRS letters (Snellen equivalent, 20/500) in patients randomized mm (95% CI, 1.3e10.9 mm), whereas vehicle-treated patients
to vehicle. By week 8, increases from baseline were observed in showed a small decrease (e0.1 mm; 95% CI, e5.3 to 5.0 mm). The
both treatment groups (last observation carried forward). Best- treatment difference (þ6.3 mm; 95% CI, e1.1 to 13.7 mm) was
corrected distance visual acuity changes from baseline were analyzed using an ANCOVA with treatment as factor and con-
assessed by an ANCOVA with randomized treatment as a factor trolling for baseline Schirmer test results and time since diagnosis
and controlling for baseline BCDVA, time since diagnosis of of neurotrophic keratopathy (months), as well as by Wilcoxon
neurotrophic keratopathy (months), and baseline Schirmer test rank-sum test results. Results suggested a trend favoring

Figure 2. Bar graphs showing primary efficacy analysis of corneal healing during masked treatment. Percentage of patients achieving corneal healing with
cenegermin treatment (gray) or vehicle (white). Corneal healing was defined (A) conventionally as less than 0.5 mm of lesion staining or (B) more
conservatively as 0 mm of lesion staining and no other persistent staining. Patients were randomized to treatment with cenegermin (n ¼ 24) or vehicle
(n ¼ 24), and corneal healing was assessed as a yes-or-no variable at week 4 and week 8 (missing assessments imputed by last observation carried forward).
See text for details. *P < 0.05, **P < 0.01, and ***P < 0.001 (22 chi-square analysis).

20
 Pflugfelder et al 
Cenegermin for Neurotrophic Keratopathy

Table 2. Sensitivity Analysis of Corneal Healing (Nonresponder Imputation)

Week 4 Week 8
Cenegermin (N ¼ 24) Vehicle (N ¼ 24) Cenegermin (N ¼ 24) Vehicle (N ¼ 24)
<0.5-mm maximum diameter of lesion 13/24 (54.2) 9/24 (37.5) 15/24 (62.5) 6/24 (25.0)
staining, no. (%)
Difference (cenegermin minus vehicle), % 16.7 37.5
95% confidence interval e11.1 to 44.5 11.5e63.5
P value
22 chi-square test 0.247 0.006
Fisher exact test, 1 sided 0.193 0.006
Fisher exact test, 2 sided 0.385 0.009
0 mm of lesion staining and no other residual 13/24 (54.2) 5/24 (20.8) 14/24 (58.3) 3/24 (12.5)
staining, no. (%)
Difference (cenegermin minus vehicle), % 33.3 45.8
95% confidence interval 7.62e59.1 22.1e69.6
P value
22 chi-square test 0.017 < 0.001
Fisher exact test, 1 sided 0.018 < 0.001
Fisher exact test, 2 sided 0.036 0.002

Missing values were imputed as failures.

cenegermin treatment at week 8 that approached statistical signif- chi-square analysis (P ¼ 0.092; 95% CI, 0.5078671e0.0296063).
icance both in the ANCOVA model (P ¼ 0.09) and the rank-sum Events assessed only as disease progression (increase in lesion size
test (P ¼ 0.07). of 1 mm, progression in lesion depth to corneal melting or
The percentage of patients experiencing deterioration from perforation, or both) also were analyzed post hoc. Overall, 5 of 23
baseline to week 8 was assessed as a secondary end point and is patients (21.7%) receiving cenegermin experienced 5 events of
presented in Table 4. Deterioration was defined as a decrease in disease progression, whereas 12 of 24 patients receiving vehicle
BCDVA by more than 5 ETDRS letters, onset of infection, (50%) experienced 19 events of disease progression. The treatment
disease progression (increase in lesion size of 1 mm, difference (28.3%) reached statistical significance in post hoc
progression in lesion depth to corneal melting or perforation, or analysis (P ¼ 0.044; 95% CI, 0.5442013 to e0.0210161). After
both), or a combination thereof. During the masked treatment week 4 of masked treatment, 4 events of deterioration or disease
period, lower rates of deterioration occurred in cenegermin- progression occurred in 3 patients receiving vehicle. In contrast, no
treated patients compared with vehicle-treated patients. Overall, 6 progression or deterioration events occurred in cenegermin-treated
events classified as deterioration occurred in 6 of 23 patients patients.
receiving cenegermin (26.1%), compared with 19 events in 12 of Per protocol, if patients healed after either masked or open-label
24 patients receiving vehicle (50%); the treatment difference cenegermin treatment (<0.5 mm of lesion staining) and then
(23.9%) suggests fewer deterioration events in cenegermin-treated experienced recurrence of persistent epithelial defect or corneal
patients, but statistical significance was not achieved in post hoc ulcer during follow-up, they were eligible for an additional 8-week

Table 3. Logistic Regression Analysis of Corneal Healing

Week 4 (N [ 47)* Week 8 (N [ 47)*

y z
Explanatory Variables Treatment Lesion Size Treatmenty Lesion Sizez
Outcome variable
<0.5-mm maximum diameter of lesion staining
Odds ratio 3.13 0.61 7.31 0.74
95% confidence interval 0.83e11.8 0.42e0.89 1.86e28.8 0.52e1.05
P value
Explanatory variable 0.091 0.010 0.004 0.089
Overall model 0.0076 0.0041
0-mm lesion staining and no other residual staining
Odds ratio 8.59 0.59 13.0 0.72
95% confidence interval 1.86e39.8 0.39e0.90 2.83e60.0 0.49e1.06
P value
Explanatory variable 0.006 0.014 0.001 0.094
Overall model 0.0008 0.0005

*All randomized patients with evaluable observations in the study (last observation carried forward). One patient (randomized to cenegermin) withdrew
before receiving treatment and did not have a baseline reading center measurement, and so was excluded from analysis.
y
Randomized treatment received during masked treatment (cenegermin ¼ 1; vehicle ¼ 0).
z
Lesion size at baseline (maximum diameter in millimeters), reading center measurement.

21
 Ophthalmology Volume 127, Number 1, January 2020

Figure 3. Box-and-whisker plots showing corneal lesion size change from baseline to week 8. Patients were randomized to 8 weeks of masked treatment with
(A) cenegermin or (B) vehicle. Reading center measurements of corneal lesions (greatest diameter of fluorescein staining in millimeters) are represented in a
boxplot at each time point. Missing values at weeks 4 and 8 were imputed as last postbaseline measurements carried forward. In the cenegermin group, 1
patient (who withdrew shortly after randomization) did not receive study treatment or a baseline reading center measurement, and thus was excluded from
the baseline summary. Five patients (3 randomized to cenegermin and 2 randomized to vehicle) did not demonstrate any postbaseline values to carry forward
and were excluded from summaries at weeks 4 and 8. Cenegermin-treated patients showed reduced lesion sizes overall compared with vehicle-treated
patients, reaching statistical significance in an analysis of covariance using treatment as a factor and baseline measurements as covariates (see text for
details). The box represents interquartile range (IQR); the midline represents the median value; the cross represents the mean value; the whiskers represent
the local minimum and maximum; and the circle represents the individual data point. Circles beyond the whiskers are outliers according to the Tukey
standard (1.5 times the IQR).

recurrence treatment period with cenegermin (Fig 1). Corneal follow-up. Of 13 patients (originally randomized to vehicle) who
healing after recurrence treatment was assessed by the received open-label cenegermin treatment, 9 (69.2%) achieved less
investigator as less than 0.5 mm of lesion staining and is than 0.5 mm of lesion staining. Of these patients, 4 of 9 (44.4%)
presented using descriptive statistics. Of the 16 patients who experienced recurrence. One patient experienced recurrence 104
achieved less than 0.5 mm of lesion staining after masked days after the last visit of the open-label cenegermin treatment
cenegermin treatment, 2 (12.5%) underwent re-treatment for period but was not retreated with cenegermin (investigator deci-
recurrence: 1 patient on the day after completing masked treatment sion) and did not achieve less than 0.5 mm of lesion staining by the
and 1 patient 39 days after the last visit of the masked treatment end of follow-up. The remaining 3 patients received cenegermin
period. Both patients (100%) healed after recurrence treatment and retreatment for recurrence. One patient (33.3%) began cenegermin
maintained less than 0.5 mm of lesion staining through the end of re-treatment 7 days after the last visit of open-label treatment but

Table 4. Secondary Efficacy Analysis of Deterioration and Disease Progression during Masked Treatment

Deterioration* Disease Progressiony

Cenegermin (N ¼ 24) Vehicle (N ¼ 24) Cenegermin (N ¼ 24) Vehicle (N ¼ 24)
z
Day 4, no. (%) 1/23 (4.3) N/A 1/23 (4.3) N/A
Week 1, no. (%) 2/22 (9.1) 4/24 (16.7) 2/22 (9.1) 4/24 (16.7)
Week 2, no. (%) 0/21 (0.0) 5/23 (21.7) 0/21 (0.0) 5/23 (21.7)
Week 3, no. (%) 1/20 (5.0) 5/20 (25.0) 1/20 (5.0) 5/20 (25.0)
Week 4, no. (%) 2/20 (10.0) 1/17 (5.9) 1/20 (5.0) 1/17 (5.9)
Week 6, no. (%) 0/19 (0.0) 2/16 (12.5) 0/19 (0.0) 2/16 (12.5)
Week 8, no. (%) 0/18 (0.0) 2/15 (13.3) 0/18 (0.0) 2/15 (13.3)
Total events, no.x 6 19 5 19
Total patients, no. (%)k 6/23 (26.1) 12/24 (50.0) 5/23 (21.7) 12/24 (50.0)
Difference (cenegermin minus vehicle), % e23.9 e28.3
95% confidence interval e0.5078671 to 0.0296063 e0.5442013 to e0.0210161
P value (chi-square test) 0.092 0.044

*Deterioration was defined as decrease in best-corrected distance visual acuity by >5 Early Treatment Diabetic Retinopathy Study letters, onset of infection,
disease progression (see below), or a combination thereof.
y
Defined as increase in lesion size of 1 mm, progression in lesion depth to corneal melting or perforation, or both.
z
One patient receiving cenegermin treatment experienced an adverse event recorded as “disease progression (worsening of neurotrophic keratitis)” and
withdrew on day 4 (unscheduled visit). All other deterioration or disease progression events were recorded on scheduled visits (weeks 1, 2, 3, 4, 6, and 8).
x
Includes multiple deterioration or disease progression events experienced by individual patients.
k
Patients experiencing multiple deterioration or disease progression events during masked treatment were counted only once for the purposes of this analysis.

22
 Pflugfelder et al 
Cenegermin for Neurotrophic Keratopathy

did not achieve corneal healing. The other 2 patients (66.7%) thinning) in patients with neurotrophic keratopathy. The
began cenegermin retreatment 35 days and 37 days after the last dosage (cenegermin 20 mg/ml) was based on the European
visit of open-label treatment; both healed and maintained less than REPARO phase II study,11 which compared cenegermin 10
0.5 mm of lesion staining through follow-up. mg/ml, cenegermin 20 mg/ml, and vehicle formulations
under the same treatment regimen as the NGF0214 study
Safety Outcomes reported herein. Although the European REPARO phase 2
Table 5 summarizes treatment-related adverse events, which are study did not show statistically significant differences
presented with patients grouped according to original randomized between the 2 cenegermin doses in any prespecified
treatment assignment. Overall, the most frequently reported efficacy parameters, their safety profiles were similar;
treatment-related adverse events were in the system organ class of furthermore, the 20-mg/ml formulation exhibited better
eye disorders according to the Medical Dictionary for Regulatory trends of efficacy in post hoc analyses, including reduction
Activities, version 19.0. During the masked treatment period, the in lesion size and reassessment of the primary efficacy
most common adverse event was eye pain, reported 5 times by 4 of
parameter of corneal healing (<0.5 mm of lesion staining)
47 patients (8.5%) overall: 4 events reported by 3 of 23
cenegermin-treated patients (13.0%) and 1 event reported by 1 of using the more conservative measure (0 mm of lesion
24 patients (4.2%) treated with vehicle. With the exception of joint staining and no other persistent staining). Therefore, the
swelling (reported by 1 patient in the cenegermin group), all other 20-mg/ml dose was selected for the formulation used in the
treatment-related adverse events in the masked treatment period United States NGF0214 study, which sought to assess
were ocular in nature (paresthesia, in the system organ class ner- prospectively the conservative definition of corneal healing
vous system disorders, represented transient tingling in the study (0 mm of lesion staining and no other persistent staining) as
eye in 1 cenegermin-treated patient). A total of 8 serious adverse a prespecified efficacy end point and to define further the
events occurred in 7 of 47 patients (14.9%) during the masked clinical outcomes of cenegermin treatment in neurotrophic
treatment period: 3 of 23 cenegermin-treated patients (13.0%) and keratopathy with metrics not used in the previous European
4 of 24 vehicle-treated patients (16.7%). No serious adverse events
REPARO study.
reported during the masked treatment period was considered
related to study treatment. It is well documented in the medical and scientific
Of 13 patients originally randomized to vehicle treatment who literature that palliative treatments for neurotrophic kerat-
received cenegermin in the open-label treatment period, 3 patients opathy (including tarsorrhaphy, preservative-free lubricants,
reported 1 eye disorder each. No serious adverse events were re- therapeutic contact lenses, and close monitoring) can pro-
ported during the open-label treatment period. mote corneal epithelial regrowth over a neurotrophic
In the follow-up period documented in Table 5, the cenegermin lesion.1 However, the cosmetic impact of some surgical
group includes patients who received recurrence treatment, and the interventions (such as tarsorrhaphy and conjunctival flap)
vehicle group includes patients who received cenegermin treatment may be undesirable to patients; furthermore, palliative
during open-label and recurrence treatment periods. Three patients treatments for neurotrophic keratopathy may pose a higher
originally randomized to cenegermin (2 of whom were retreated for
risk of disease recurrence, because these treatments do not
recurrence) reported eye pain, and 1 of these patients (with neu-
rotrophic keratopathy as the cause of herpetic keratitis) also address the underlying deficits that drive pathophysiologic
experienced ocular herpes simplex during follow-up. In the vehicle features. To this point, more vehicle-treated patients
group, 2 patients experienced treatment-related adverse events exhibited corneal healing at week 4 compared with week 8
during follow-up; both received cenegermin during the open-label (Fig 2), suggesting temporary epithelial proliferation
and recurrence treatment periods. One of these patients experi- followed by disease recurrence under palliative vehicle
enced disease progression and was withdrawn from the study. The treatment. In contrast, cenegermin treatment consistently
other patient experienced eye pain and worsening of a pre-existing demonstrated highest corneal healing rates at week 8,
cataract. suggesting steady improvement over time. These healing
During follow-up, serious adverse events occurred in 5 patients profiles withstood multiple sensitivity analyses (Table 2),
(3 in the cenegermin group and 2 in the vehicle group). No serious
supporting the robustness of the primary efficacy data.
adverse events reported during follow-up were considered related
to study treatment. In neurotrophic keratopathy patients, corneal healing was
No deaths were reported in either the masked or open-label defined conventionally as less than 0.5 mm of lesion
treatment periods. During follow-up, 1 patient originally random- staining, which is based on the lower limit of reliable slit-
ized to cenegermin treatment died of unknown causes; this lamp assessment. This conventional definition of corneal
occurred 163 days after the last dose of study drug and was healing may overlook small neurotrophic lesions at less than
considered unrelated to study treatment. the limits of slit-lamp detection and may be more subject to
Analyses of vital signs, ophthalmic parameters, and laboratory interoperator variability among central readers; thus, we also
parameters did not reveal any clinically significant patterns in used a conservative measure of 0 mm of lesion staining and
patients treated with cenegermin or vehicle. Consistent with results no other residual staining, assessed by central readers.
of the REPARO phase 2 study,11 no anti-NGF antibodies were
Although both measures of corneal healing showed statis-
detected at any testing time point.
tically significant differences in the primary efficacy
analysis (Fig 2), the requirement for lack of lesion staining
Discussion and other persistent staining suggests that it is a clinically
more meaningful measure of corneal healing.
In this pivotal trial conducted in the United States, topical In this study, we examined potential effects of baseline
cenegermin treatment effectively and safely promoted heal- variables (such as patient demographics and disease pa-
ing of persistent epithelial defects (with or without stromal rameters) on the clinical outcome of corneal healing. Based

23
 Ophthalmology Volume 127, Number 1, January 2020

Table 5. Treatment-Related Adverse Events during Masked Treatment, Open-Label Treatment, and Follow-up Periods

Cenegermin (N [ 23)y Vehicle (N [ 24)y Overall (N [ 47)y

z x x
System Organ Class* No. (%) Events No. (%) Events No. (%) Eventsx
Masked treatment period
Any treatment-related adverse eventsk 10 (43.5) 26 8 (33.3) 15 18 (38.3) 41
Eye disorders 9 (39.1) 20 8 (33.3) 13 17 (36.2) 33
Eye pain 3 (13.0) 4 1 (4.2) 1 4 (8.5) 5
Photophobia 1 (4.3) 1 2 (8.3) 2 3 (6.4) 3
Visual acuity reduced 1 (4.3) 1 2 (8.3) 3 3 (6.4) 4
Corneal epithelium defect 2 (8.7) 2 0 0 2 (4.3) 2
Eye irritation 0 0 2 (8.3) 2 2 (4.3) 2
Anterior chamber inflammation 1 (4.3) 1 0 0 1 (2.1) 1
Corneal deposits 1 (4.3) 1 0 0 1 (2.1) 1
Corneal neovascularization 0 0 1 (4.2) 2 1 (2.1) 2
Corneal thinning 0 0 1 (4.2) 1 1 (2.1) 1
Eye discharge 1 (4.3) 1 0 0 1 (2.1) 1
Eye inflammation 1 (4.3) 1 0 0 1 (2.1) 1
Eye pruritus 0 0 1 (4.2) 1 1 (2.1) 1
Eyelid pain 1 (4.3) 2 0 0 1 (2.1) 2
Foreign body sensation 1 (4.3) 1 0 0 1 (2.1) 1
Hyphema 1 (4.3) 1 0 0 1 (2.1) 1
Keratitis 1 (4.3) 1 0 0 1 (2.1) 1
Lacrimation increased 1 (4.3) 1 0 0 1 (2.1) 1
Ocular hyperemia 1 (4.3) 1 0 0 1 (2.1) 1
Posterior capsule opacification 1 (4.3) 1 0 0 1 (2.1) 1
Vision blurred 0 0 1 (4.2) 1 1 (2.1) 1
General disorders and administration site conditions 2 (8.7) 2 2 (8.3) 2 4 (8.5) 4
Disease progression 1 (4.3) 1 1 (4.2) 1 2 (4.3) 2
Foreign body sensation 1 (4.3) 1 1 (4.2) 1 2 (4.3) 2
Investigations 1 (4.3) 1 0 0 1 (2.1) 1
Intraocular pressure increased 1 (4.3) 1 0 0 1 (2.1) 1
Musculoskeletal and connective tissue disorders 1 (4.3) 2 0 0 1 (2.1) 2
Joint swelling 1 (4.3) 1 0 0 1 (2.1) 1
Nervous system disorders 1 (4.3) 1 0 0 1 (2.1) 1
Paresthesia 1 (4.3) 1 0 0 1 (2.1) 1
Open-label treatment period
Any treatment-related adverse eventsz N/A N/A 3 (23.1) 3 3 (23.1) 3
Eye disorders 3 (23.1) 3 3 (23.1) 3
Eye pain 1 (7.7) 1 1 (7.7) 1
Eyelid pain 1 (7.7) 1 1 (7.7) 1
Visual acuity reduced 1 (7.7) 1 1 (7.7) 1
Follow-up period{
Any treatment-related adverse eventsz 3 (13.0) 4 2 (8.3) 3 5 (10.6) 7
Eye disorders 3 (13.0) 3 1 (4.2) 2 4 (8.5) 5
Eye pain 3 (13.0) 3 1 (4.2) 1 4 (8.5) 4
Cataract (worsening) 0 0 1 (4.2) 1 1 (2.1) 1
General disorders and administration site conditions 0 0 1 (4.2) 1 1 (2.1) 1
Disease progression 0 0 1 (4.2) 1 1 (2.1) 1
Infections and infestations 1 (4.3) 1 0 0 1 (2.1) 1

AE ¼ adverse event.
*Medical Dictionary for Regulatory Activities, version 19.0, preferred term. Body system and preferred terms are sorted by descending frequency of event
count in the “Overall” column.
y
Number of patients who received the randomly assigned treatment in masked treatment period (safety population).
z
Number of patients with treatment-related adverse events counted on a per-patient basis (if a patient reported the same event repeatedly, the event was
counted only once). Percentages were calculated using the safety population of each treatment group as the denominator.
x
Number of observed events.
k
Treatment-emergent conditions having a relationship to study treatment recorded as possible, probable, highly probable, or missing.
{
Includes any recurrence treatment (see text for details).

on multiple logistic regression modeling in our patient are consistent with the greater healing rates (Fig 2) and
population, the effect of treatment (cenegermin vs. vehicle) overall reductions in corneal lesion size in cenegermin-
was the only significant predictor of healing at week 8, treated patients versus vehicle-treated patients (Fig 3) and
suggesting that a full treatment cycle achieves the thera- support maintaining cenegermin treatment over the full
peutic effect independent of initial lesion size. These results 8-week course.

24
 Pflugfelder et al 
Cenegermin for Neurotrophic Keratopathy

Other variables (such as age, disease stage, and time NGF, which is the basis for the recombinant test product
since diagnosis) did not have significant effects on healing that is identical in amino acid sequence to native human
status. Underlying etiologies also did not show significant NGF. No immunogenicity to NGF was detected in this
correlations with clinical outcomes; however, certain etiol- study, consistent with previous findings in healthy
ogies (such as diabetes and chemical burns) that may have volunteers9 and patients with neurotrophic keratopathy.10,11
divergent outcomes in clinical practice are not well repre- The current study clearly defined favorable benefit-to-
sented in our patient population. Thus, it is difficult to form risk ratios for topical cenegermin in patients with neuro-
conclusions on the relative efficacy of cenegermin treatment trophic persistent epithelial defects (with or without stromal
on cases of neurotrophic keratopathy of different etiologies. thinning). Future studies may decipher the precise patho-
In other clinically relevant end points, some statistically logic processes modulated by cenegermindparticularly
significant improvements and favorable trends were observed. corneal denervationdand the potential therapeutic efficacy
For example, fewer cenegermin-treated patients experienced of cenegermin in other neurodegenerative diseases. In
disease progression over the 8-week treatment course summary, cenegermin ophthalmic solution represents a safe,
compared with vehicle-treated patients. Although few events novel, and noninvasive pharmacologic treatment for neu-
of disease progression or deterioration occurred overall, it is rotrophic keratopathy and can become part of the treatment
interesting to note that after week 4 of the masked treatment algorithm for this often difficult to manage disease with a
period, only vehicle-treated patients exhibited signs of disease high need for targeted and effective pharmacotherapies.
progression or deterioration. Taken together, these results
suggest that cenegermin may prevent disease progression more
effectively than vehicle treatment. References
In line with the previously reported REPARO phase 2
study,11 this trial did not yield statistically significant
improvements in corneal sensitivity measured by Cochet- 1. Mastropasqua L, Massaro-Giordano G, Nubile M,
Bonnet esthesiometer, yet reflex tearing (which may Sacchetti M. Understanding the pathogenesis of neurotrophic
keratitis: the role of corneal nerves. J Cell Physiol.
reflect corneal sensitivity or nerve function not detectable by
2017;232(4):717e724.
Cochet-Bonnet esthesiometer) exhibited trends favoring 2. Sacchetti M, Lambiase A. Diagnosis and management of
cenegermin treatment. Conclusive data may require larger neurotrophic keratitis. Clin Ophthalmol. 2014;8:571e579.
sample sizes and longer follow-up. Also in line with the 3. Chang BH, Groos EB. Neurotrophic keratitis. In:
REPARO phase 2 study, we assessed BCDVA as a sec- Krachmer JH, Mannis MJ, Holland EJ, eds. Cornea. 3rd ed.
ondary efficacy end point, although visual acuity does not St. Louis: Mosby/Elsevier; 2011:1101e1108.
necessarily correlate with neurotrophic keratopathy severity 4. Mantelli F, Nardella C, Tiberi E, et al. Congenital corneal
or healing status. For example, persistent epithelial defects anesthesia and neurotrophic keratitis: diagnosis and manage-
may be relatively transparent (and have little to no effect on ment. Biomed Res Int. 2015;2015:805876.
vision), whereas the haze commonly associated with a 5. Hsu HY, Modi D. Etiologies, quantitative hypoesthesia, and
clinical outcomes of neurotrophic keratopathy. Eye Contact
healing corneal epithelium may cloud vision temporarily,
Lens. 2015;41(5):314e317.
particularly in the central or paracentral cornea. Therefore, it 6. Sacchetti M, Lambiase A. Neurotrophic factors and corneal
is not surprising that neither the phase 2 REPARO trial nor nerve regeneration. Neural Regen Res. 2017;12(8):
the current NGF0214 study showed statistically significant 1220e1224.
improvements in visual acuity measures. 7. Lambiase A, Rama P, Bonini S, et al. Topical treatment with
There were few events of disease recurrence recorded nerve growth factor for corneal neurotrophic ulcers. N Engl J
during follow-up. A total of 6 patients experienced recur- Med. 1998;338(17):1174e1180.
rence, and 5 of these patients received recurrence treatment 8. Bonini S, Lambiase A, Rama P, et al. Topical treatment with
with cenegermin. The recurrence rates (and healing rates after nerve growth factor for neurotrophic keratitis. Ophthalmology.
cenegermin retreatment) are too small to form any conclu- 2000;107(7):1347e1351. discussion 51-52.
9. Ferrari MP, Mantelli F, Sacchetti M, et al. Safety and phar-
sions; logistic regression analyses examining correlations
macokinetics of escalating doses of human recombinant nerve
between baseline variables (including etiologies) with growth factor eye drops in a double-masked, randomized
recurrence rates were inconclusive (data not shown). How- clinical trial. BioDrugs. 2014;28(3):275e283.
ever, it is interesting to note that 4 of 5 cenegermin-retreated 10. Bonini S, Lambiase A, Rama P, et al. Phase I trial of recom-
patients (80%; i.e., patients who received two 8-week courses binant human nerve growth factor for neurotrophic keratitis.
of cenegermin) achieved corneal healing, which was main- Ophthalmology. 2018;125(9):1468e1471.
tained through the end of the follow-up period. 11. Bonini S, Lambiase A, Rama P, et al. Phase II randomized,
No obvious safety concerns arose; none of the serious double-masked, vehicle-controlled trial of recombinant human
adverse events was considered related to study treatment, nerve growth factor for neurotrophic keratitis. Ophthalmology.
and nearly all treatment-related adverse events were ocular 2018;125(9):1332e1343.
12. Mackie IA. Neuroparalytic keratitis. In: Fraunfelder FT,
in nature (Table 5). The most common treatment-related
Roy FH, Grove J, eds. Current Ocular Therapy. 4th ed.
adverse events (e.g., eye pain, foreign body sensation, and Philadelphia: Saunders; 1995:452e454.
tingling) suggest nociceptor sensitization, which is associ- 13. Jankowski MP, Koerber HR. Neurotrophic factors and noci-
ated commonly with NGF in preclinical studies.13 ceptor sensitization. In: Kruger L, Light AR, eds. Trans-
Therefore, most of the ocular adverse events may lational Pain Research: From Mouse to Man. Boca Raton, FL:
represent known mechanisms of action of endogenous CRC Press/Taylor & Francis; 2010:31e49.

25
 Ophthalmology Volume 127, Number 1, January 2020

Footnotes and Financial Disclosures

Originally received: May 11, 2019. N.A.A.: Consultant e Aescula Tech, Dompé Farmaceutici SpA (Milan,
Final revision: July 15, 2019. Italy), Spark Therapeutics, Trefoil Biotherapeutics.
Accepted: August 19, 2019. W.C.: Employee e Dompé Farmaceutici SpA (Milan, Italy).
Available online: August 26, 2019. Manuscript no. 2019-1013.
M.A.: Employee e Dompé Farmaceutici SpA (Milan, Italy).
1
Department of Ophthalmology, Baylor College of Medicine, Houston,
F.M.: Employee e Dompé Farmaceutici SpA (Milan, Italy).
Texas.
2 R.D.: Consultant e Dompé Farmaceutici SpA (Milan, Italy), GSK, Kala,
Scheie Eye Institute, University of Pennsylvania, Philadelphia,
Novaliq, Oculis, Santen; Financial support e Cambium, Allergan; Equity
Pennsylvania.
owner e Claris Biotherapeutics, Aramis Biosciences.
3
Bascom Palmer Eye Institute, University of Miami, Plantation, Florida.
Victor L. Perez is currently affiliated the Department of Ophthalmology,
4
New England Eye Center, Tufts Medical Center, Tufts University School Duke University School of Medicine, Durham, North Carolina.
of Medicine, Boston, Massachusetts.
Supported by Dompé Farmaceutici SpA, Milan, Italy. Dompé Farmaceutici
5
Stein Eye Institute, University of California, Los Angeles, Los Angeles, SpA participated in the design and conduct of the study; data collection for
California. immunogenicity assessments; management, analysis, and interpretation of
6
Department of Ophthalmology, University of Pittsburgh, Pittsburgh, the data; and preparation and review of the manuscript. The sponsor was not
Pennsylvania. involved in efficacy data collection for masked central analysis.
7
Sightline Ophthalmic Associates, Pittsburgh, Pennsylvania. HUMAN SUBJECTS: Human subjects were included in this study. Insti-
8
Massachusetts Eye Research and Surgery Institution, Waltham, tutional Review Board Approval was obtained from each participating site
Massachusetts. for the study protocol. All research adhered to the tenets of the Declaration
9
Department of Ophthalmology, Harvard Medical School, Boston, of Helsinki. All participants provided informed consent.
Massachusetts. No animal subjects were included in this study.
10
Loma Linda University Eye Institute, Loma Linda, California. Author Contributions:
11
New York Eye and Ear Infirmary of Mt. Sinai School of Medicine, New Conception and design: Pflugfelder, Massaro-Giordano, Perez, Hamrah,
York, New York. Deng, Espandar, Foster, Affeldt, Seedor, Afshari, Allegretti, Mantelli, Dana
12
Shiley Eye Institute, University of California, San Diego, La Jolla, Analysis and interpretation: Pflugfelder, Massaro-Giordano, Perez, Hamrah,
California. Deng, Espandar, Foster, Affeldt, Seedor, Afshari, Chao, Allegretti, Man-
13
Dompé Farmaceutici SpA, Milan, Italy. telli, Dana
14
Department of Ophthalmology, Massachusetts Eye and Ear Infirmary, Data collection: Pflugfelder, Massaro-Giordano, Perez, Hamrah, Deng,
Boston, Massachusetts. Espandar, Foster, Affeldt, Seedor, Afshari, Chao, Allegretti, Mantelli, Dana
Presented in part at: European Society of Ophthalmology 2017 Congress, Obtained funding: Pflugfelder, Massaro-Giordano, Perez, Hamrah, Deng,
June 2017, Barcelona, Spain. Afshari, Chao, Allegretti, Mantelli, Dana
Financial Disclosure(s): Overall responsibility: Pflugfelder, Massaro-Giordano, Perez, Hamrah,
The author(s) have made the following disclosure(s): S.C.P.: Consultant e Deng, Espandar, Foster, Affeldt, Seedor, Afshari, Chao, Allegretti, Man-
Allergan, Dompé Farmaceutici SpA (Milan, Italy), Kala, Shire, Senju; telli, Dana
Financial support e Allergan. Abbreviations and Acronyms:
M.M.-G.: Consultant e PRN, GSK, Celularity. ANCOVA ¼ analysis of covariance; ANOVA ¼ analysis of variance;
V.L.P.: Consultant e Alcon, Dompé Farmaceutici SpA (Milan, Italy), BCDVA ¼ best-corrected distance visual acuity; CI ¼ confidence interval;
EyeGate, Kala, Novaliq, Shire; Equity owner e EyeGate; Financial ETDRS ¼ Early Treatment Diabetic Retinopathy Study; NGF ¼ nerve
support e Dompé Farmaceutici SpA. growth factor; OR ¼ odds ratio; rhNGF ¼ recombinant human nerve
growth factor; SD ¼ standard deviation.
P.H.: Consultant e Dompé Farmaceutici SpA (Milan, Italy), Tissue-Tech,
Noveome, Kala, Novaliq, Santen, Shire, Allergan, Clementia; Financial Correspondence:
support e Dompé Farmaceutici SpA, Allergan, Shire, Tissue-Tech, Bausch Reza Dana, MD, MPH, Department of Ophthalmology, Massachusetts Eye
& Lomb. and Ear, Harvard Medical School, 245 Charles Street, Boston, MA 02114.
S.X.D.: Consultant e W. L. Gore & Associates, Inc., Kowa Research E-mail: [email protected].
Institute, Inc.

26
A Randomized Noninferiority Trial of
Wearing Adjustable Glasses versus
Standard and Ready-made Spectacles
among Chinese Schoolchildren
Wearability and Evaluation of Adjustable Refraction III
Congyao Y. Wang, MD,1 Guoshan Zhang, MM,2 Bobby Tang, MB,3 Ling Jin, MS,2 Wenyong Huang, MD,2
Xiuqin Wang, MD,4 Tingting Chen, MD,1 Wenhui Zhu, MD,1,2 Baixiang Xiao, MM,2 Jun Wang, MD,2
Zhongqiang Zhou, MD,5 Zhizheng Tang, MB,6 Yan Liang, MD,7 Mabel Crescioni, LLM, DrPH,8
David Wilson, PhD, BEc,9 Helen McAneney, PhD,3 Joshua D. Silver, PhD,10,11 Bruce Moore, OD,12
Nathan Congdon, MD, MPH2,3

Purpose: To compare wear of standard, adjustable, and ready-made glasses among children.
Design: Randomized, controlled, open-label, noninferiority trial.
Participants: Students aged 11 to 16 years with presenting visual acuity (VA)
6/12 in both eyes, correctable
to 6/7.5, subjective spherical equivalent refractive error (SER)
e1.0 diopters (D), astigmatism and anisome-
tropia both <2.00 D, and no other ocular abnormalities.
Methods: Participants were randomly allocated (1:1:1) to standard glasses, ready-made glasses, or
adjustable glasses based on self-refraction. We recorded glasses wear on twice-weekly covert evaluation by
head teachers (primary outcome), self-reported and investigator-observed wear, best-corrected visual acuity
(BCVA) (not prespecified), children’s satisfaction, and value attributed to glasses.
Main Outcome Measure: Proportion of glasses wear on twice-weekly covert evaluation by head teachers
over 2 months.
Results: Among 379 eligible participants, 127 were allocated to standard glasses (mean age, 13.7 years;
standard deviation [SD], 1.0 years; 54.3% were male), 125 to ready-made (mean age, 13.6; SD, 0.83; 45.6%), and
127 to adjustable (mean age, 13.4 years; SD, 0.85; 54.3%). Mean wear proportion of adjustable glasses was
significantly lower than for standard glasses (45% vs. 58%; P ¼ 0.01), although the adjusted difference (90%
confidence interval [CI], e19.0% to e3.0%) did not meet the prespecified inferiority threshold of 20%. Self-
reported (90.2% vs. 84.8%, P ¼ 0.64) and investigator-observed (44.1% vs. 33.9%, P ¼ 0.89) wear did not
differ between standard and adjustable glasses, nor did satisfaction with (P ¼ 0.97) or value attributed to study
glasses (P ¼ 0.55) or increase in quality of life (5.53 [SD, 4.47] vs. 5.68 [SD, 4.34] on a 100-point scale, P > 0.30).
Best-corrected visual acuity with adjustable glasses was better (P < 0.001) than with standard glasses. Change in
power of study lenses at the end of the study (adjustable: 0.65 D, 95% CI, 0.52e0.79; standard, 0.01 D; 95% CI,
e0.006 to 0.03, P < 0.001) was greater for adjustable glasses, although interobserver variation in power mea-
surements may explain this. Lens scratches and frame damage were more common with adjustable glasses,
whereas lens breakage was less common than for standard glasses.
Conclusions: Proportion of wear was lower with adjustable glasses, although VA was better and measures of
satisfaction and quality of life were not inferior to standard glasses. Ophthalmology 2020;127:27-37 ª 2019 by the
American Academy of Ophthalmology

The implementation of the Millennium Developmental moderate and severe vision impairment.2 Subsequently,
Goals in 2000 led to notable improvements in global eye the adoption of the Sustainable Development
health.1 However, between 1990 and 2010, uncorrected Goals (SDGs) by United Nations member states in 20153
refractive error remained the most common cause of represented a renewed commitment to a just and healthy
visual impairment, contributing to more than 50% of all world for all. Reducing the burden of uncorrected

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.08.002 27

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

refractive error can contribute directly to achieving SDG 3 have been designed to evaluate self-refraction as a tool to
“Ensure healthy lives and promote well-being” and SDG 4 increase access to accurate spectacles among children in
“Ensure inclusive and equitable quality education.”4 underserved areas of China.19,20 In the current noninferiority
More than 12 million children worldwide have visual trial, school-going children in rural China were randomly
impairment due to uncorrected refractive errors, half of them assigned to receive standard spectacles or ready-made
living in China.5 The Refractive Error Study in Children glasses (a widely used alternative in low-resource areas)
estimated the prevalence of myopia more than a decade based on refraction by an experienced practitioner or
ago at 78% and 39% among urban and rural Chinese adjustable glasses with power determined by self-refraction
children, respectively, the highest of any group studied.6,7 carried out by the child. The main outcome of the current
A national, population-based study of vision in Chinese trial is spectacle wear, as assessed regularly by head teachers
children conducted at 5-year intervals before, during, and (primary outcome), by self-report, and by study personnel at
after the Refractive Error Study in Children has suggested unannounced inspection visits at the final examination.
that rates of visual impairment due to uncorrected refractive Secondary outcomes include children’s self-reported satis-
error are increasing significantly across China.8 faction with and value attributed to the glasses, best-
Despite the fact that it is safe and inexpensive to correct corrected visual acuity (BCVA) while wearing glasses (not
refractive errors with accurate spectacles,9 population prespecified), change in self-reported visual function after
studies show that the majority (80%e85%) of school- receipt of spectacles, and stability of power and robustness
aged children needing glasses do not have them in under- of lenses and frames of the different spectacle designs. The
served areas of urban and rural China.4,10 This is particu- study hypothesis was that wear and satisfaction rates would
larly concerning given that the provision of free spectacles not be significantly inferior in the adjustable compared with
in this setting has been shown to significantly improve the standard glasses group.
children’s educational outcomes.4 An important barrier to
achieving higher rates of correction is the lack of
affordable, high-quality refractive service in rural China: Methods
Two-thirds of refractionists in rural western China have a The protocol for this study was approved in full by the Ethics
high school or lower education, and half of children whom Committee of the Zhongshan Ophthalmic Center, Sun Yat-sen
they identify as having visual impairment uncorrectable University (Guangzhou, China) and by local Boards of Educa-
with glasses can be improved by more experienced practi- tion in Guangdong Province. Written informed consent was ob-
tioners.11 Among children owning glasses in significantly tained from at least 1 parent of each participating child, and the
more developed Guangdong province, approximately half principles of the Declaration of Helsinki were followed throughout.
had spectacles with power inaccurate by >1 diopter (D) Teachers were told to immediately report any problems with study
and approximately one-fifth had spectacles with power glasses, and all children not randomized to receive standard glasses
inaccurate by >2 D.12 were given such glasses at study closeout. The adjustable glasses
used in the study were approved for local use before the trial
Self-refraction with adjustable spectacles has been sug-
(Guangdong Supervision and Inspection Laboratory for Spectacle
gested as a means of addressing the shortfall of high-quality Quality, #1400709, August 1, 2014).
refractive services for children in underserved areas of
China and other low- and middle-income countries. In fact,
Participating Schools
these glasses have already been distributed en masse to
adults in developing countries such as Ghana.13 Previous This was a randomized, noninferiority, open-label, parallel-group
studies among urban14 and rural15 children found that study conducted in all 120 classes at 14 junior high schools
virtually all were capable of carrying out self-refraction, selected at random in Gaozhou and Xinyi Counties, Guangdong
whereas >90% achieved visual acuity (VA) of 6/7.5. Province, Southern China. Seven of the schools were situated in
Accuracy of self-refraction in rural Guangdong province, Xinyi County (population 9.1 million, area 3102 km2, and per
capita Gross Domestic Product of US dollar [USD] 7110 in 2016)
southern China, was superior to that of noncycloplegic and 7 schools in Gaozhou (population 18 million, area 3276 km2,
autorefraction, an alternative modality widely used in un- and per capita Gross Domestic Product of USD 4370 in 2016). For
derserved settings.16 Furthermore, an adequately powered reference, these values fall below those of Guangdong Province
noninferiority trial has shown that visual quality of life did (USD 11140) and the People’s Republic of China (USD 8130) for
not differ significantly between children wearing glasses the same year.21
with power determined by self-refraction compared with
those based on cycloplegic refraction by an experienced Baseline Examination and Eligibility and
optometrist.17 Exclusion Criteria
A remaining concern about adjustable glasses for daily
use has been the cosmetic appearance of the round, thick Students with parental consent in grades 7 and 8 (age 11e16 years)
frames, which children have consistently ranked as unat- at the selected schools underwent baseline VA screening by study
nurses and optometrists from September to October 2015. Uncor-
tractive in both urban and rural Chinese settings.18 Because
rected and presenting VA with children’s own spectacles, if
of the lack of any published studies on children’s routine use available, were assessed separately for each eye at a distance of 4 m
of adjustable glasses, data on adherence with wear, using Early Treatment Diabetic Retinopathy Study charts (Preci-
robustness of liquid-filled lenses, and stability of power sion Vision, La Salle, IL)22 in a well-lit indoor area of the school.
under conditions of daily use are not readily available. The All students with presenting vision
6/12 in both eyes were pro-
Wearability and Evaluation of Adjustable Refraction trials visionally eligible for the study and underwent refraction.

28
 Wang et al 
Trial of Adjustable Glasses in Children

Refraction and Ocular Examinations

For children with parental consent for cycloplegia, 2 drops of 1%
cyclopentolate were administered 5 minutes apart to each eye. A
third drop was given if the pupillary reflex persisted after 15 mi-
nutes. After cycloplegia, automated refraction (Topcon KR 8900,
Topcon Corporation, Tokyo, Japan) was carried out 5 times in each
eye, with the mean value recorded, and subjective refraction was
performed by a senior optometrist. The starting point for each eye
was the mean cycloplegic autorefraction, and the end point was the
least myopic power providing best VA at a distance of 4 m. Visual
acuity was measured with lenses of the indicated power placed into
trial frames and an Early Treatment Diabetic Retinopathy Study
chart with an identical layout but different sequence of optotypes
from that used in screening. Eligible children without parental
consent for cycloplegic refraction underwent noncycloplegic
autorefraction and retinoscopy by a senior optometrist before
having VA measured as described. For all children, an eye ex- Figure 1. Sample adjustable glasses.
amination was then performed by an ophthalmologist using direct
and indirect ophthalmoscopy. Children having any of the following
in either eye were excluded from the study and referred for further zero and the lens surfaces were clean. If VA still did not improve,
care as appropriate: subjective spherical equivalent refractive error the protocol was repeated again with the pumps aligned initially
>e1.0 D, best-corrected VA <6/7.5, astigmatism or anisometropia at þ6.00 D instead of zero (to detect possible hyperopia or far-
>2.00 D, or ocular abnormalities such as cataract or amblyopia. sightedness, rare in this setting). The refractive powers of both
lenses of all study glasses were measured by automated lensometry
Randomization and Intervention (Topcon CL-2500, Topcon Corp) at the conclusion of children’s
self-refraction and at the end of the study, 2 months later.
The independent statistician generated a randomization list using
an online random number generator (Randomization.com). Par- Outcome Assessment and Questionnaires
ticipants were individually randomized to receive standard, ready-
made, or adjustable glasses by stratified (cycloplegic vs. non- Details of participants’ baseline characteristics and prestudy
cycloplegic refraction) block randomization, with a block size of 6. glasses wear were collected with self-administered questionnaires
The 6 subjects in a block were assigned to the 3 study groups in at the beginning of the study. The main trial outcome was the
1:1:1 ratio in each stratum at each school, and allocation codes proportion in each group wearing assigned spectacles during the
were sealed in sequentially numbered opaque envelopes. Exam- study period. This was assessed in 3 ways. First, students’ wear
iners were masked to group allocation. All study glasses were of the assigned study glasses in class was monitored and recorded
provided free of charge. Standard glasses were customized to the covertly by head teachers twice a week during roll call at the
child’s refractive error with inter-pupillary distance measured using beginning of class for 2 months (primary outcome; several as-
standard techniques. Ready-made glasses were available in 0.50 D sessments were during holidays, so there was an expected total of
steps between e1.00 and e7.00 D and 1.00 D steps between e7.00 12 assessments). The study coordinator carried out regular
and e10.00 D with the same power in each eye. In choosing quality checks on the teachers. Students also self-reported their
glasses power, the spherical equivalent of the worse-seeing eye on own use of study glasses (“never wear,” “wear only for study-
subjective refraction was used, rounded down to the nearest ing,” “usually wear”) at the end of the study, and an unan-
available power as needed. Available inter-pupillary distances were nounced inspection of glasses wear by study personnel was
50, 55, 60, and 65 mm. The power of adjustable glasses (Fig 1) was carried out at study closeout 2 months after initial distribution of
determined by children themselves according to a previously assigned glasses. Children not observed wearing their glasses at
described protocol.14,15 baseline or at the end of the trial were asked to select the rea-
Briefly, self-refraction was supervised by project researchers son(s) for nonwear from a prespecified list.
who had previously attended a 90-minute training workshop. The Secondary outcomes included children’s self-reported satisfac-
optical power of each lens, from e6.00 to þ6.00 D, was deter- tion with and value attributed to the glasses, self-reported visual
mined by the curvature of its surfaces, controlled independently by function, BCVA wearing the study glasses (not prespecified), and
varying the volume of liquid in the lens with user-controlled stability of power and robustness of lenses and frames in the
pumps, one mounted on each side (these were removed after the different spectacle designs. A previously validated16 Mandarin
refraction process). Adjustable glasses were initially aligned to 0.0 version of the National Eye Institute Refractive Error Quality of
D on both sides, with each eye tested separately. The child was Life-42 (NEI RQL-42) questionnaire23-25 was used to evaluate
instructed to turn the dial backwards (creating a minus power lens) visual function-related quality of life at baseline and after 2 months
until the letters on the vision chart became as clear as possible and of spectacle wear at the final examination. Value attached to the
VA was measured. glasses and participant satisfaction with glasses were assessed on
Participants were then asked to make small adjustments in 5-point scales (0 least to 4 most) as described previously26,27 at the
either direction to refine the VA, which was measured again. final examination. A form, modified from an instrument used in
Finally, participants were directed to turn the dial forward glasses provision programs by an eye health nongovernmental
(reducing minus power) until the smallest visible line began to blur organization (Brien Holden Vision International, Sydney,
slightly. The VA was measured a third time, and if there was no Australia), was used to perform a detailed assessment of damage to
decrease from the previous step, it was accepted as the final value. the study glasses.
In the event that the VA did not improve over unaided acuity, study An additional testeretest study was conducted in August 2017
personnel would check again that the plunger was aligned with in Guangzhou, Guangdong, with 15 pairs of adjustable glasses and

29
 Ophthalmology Volume 127, Number 1, January 2020

Figure 2. Flowchart for study enrollment.

10 pairs of standard glasses worn by students in the trial. By using continuous variables and frequency (percentage) for categoric
a randomized sequence, 3 examiners tested each pair of glasses variables. Family wealth was estimated by summing the value, as
with the automated lensometer used in the main study (Topcon CL- reported in the China Rural Household Survey Yearbook
2500) on 2 separate occasions to assess intra- and inter-user (Department of Rural Surveys, National Bureau of Statistics of
measurement variability. All glasses were covered by a paper China, 2013), of household items owned by the family from a
frame so that the lens type (standard vs. adjustable) would be previously defined list of 13 common objects. Baseline knowledge
indistinguishable to examiners. One examiner (C.Y.W.) also and attitudes about myopia included 6 items on a 5-point scale: 1
measured the power of the 15 pairs of adjustable glasses after they (strongly agree) to 5 (strongly disagree). The proportion agreeing
had been stored for 120 minutes in a warm environment (34 C) and or strongly agreeing was calculated.
again after storage for 120 minutes in a cold setting (5 C). Baseline comparisons between the Standard Glasses Group and
other 2 groups were performed by linear regression for continuous
Sample Size and Statistical Methods variables with normal distribution, ordinal logistic regression for
ordinal categoric variables, and logistic regression for binary var-
The sample size was calculated on the basis of repeated head iables. Change in NEI RQL-42 scores after wearing study glasses
teacher assessment of the proportion of spectacle wear as primary was compared between the Standard Glasses group and other 2
outcome. Under the assumption that the ratio of the proportion of groups using the Wilcoxon rank-sum test for individual items and
wearing adjustable versus standard spectacles during the study 2-sample t test for Total Score. The proportion of subjects with
period would be 0.95 with a coefficient of variation of 0.45, and best-corrected VA 6/6 with study spectacles was compared be-
taking 20% as the noninferiority margin, the study was designed to tween the Standard Glasses group and other 2 groups, adjusting for
enroll 324 subjects (n ¼ 108 in each of 3 groups), resulting in baseline presenting VA in the better-seeing eye using logistic
power of 90% for a 1-sided noninferiority test at a significance regression.
level of 5%. The sample size was calculated using PASS 11 (NCSS Study group and all variables significant with P < 0.05 in
Statistical Software, Kaysville, UT). simple regression analyses were included in the multiple regression
Baseline characteristics of participants were presented as mean model to identify associations between potential prognostic in-
(standard deviation [SD]) or median (interquartile range) for dicators and outcomes. Both primary and secondary outcomes

30
 Table 1. Baseline Characteristics of Participating Students by Group (N ¼ 379)

Standard Ready-made Adjustable P Value P Value

Glasses Glasses Glasses Comparing Comparing
Characteristic Total n [ 379 n [ 127 (1) n [ 125 (2) n [ 127 (3) (1) with (2)# (1) with (3)#
Age (yrs), Mean (SD)* 13.6 (0.90) 13.7 (1.00) 13.6 (0.83) 13.4 (0.85) 0.22 0.004
Male sex, n (%) 195 (51.5) 69 (54.3) 57 (45.6) 69 (54.3) 0.17 1.00
Owned glasses at baseline, n (%) 111 (29.3) 42 (33.1) 33 (26.4) 36 (28.4) 0.25 0.42
Attends rural school, n (%) 236 (62.3) 78 (61.4) 79 (63.2) 79 (62.2) 0.77 0.90
Only child in family, n (%)* 32 (8.49) 5 (3.97) 14 (11.2) 13 (10.3) 0.04 0.06
At least 1 parent wears glasses, n (%)y 85 (22.9) 23 (18.6) 24 (19.4) 38 (30.7) 0.87 0.028
1 or both parents with 12 yrs of 144 (38.2) 48 (37.8) 45 (36.6) 51 (40.2) 0.84 0.70
education, n (%)*
Both parents out-migrated for 95 (25.1) 27 (21.3) 36 (28.8) 32 (25.2) 0.17 0.46

Wang et al
work, n (%)
Self-reported study time each day
after school
<1 hour 241 (63.6) 81 (63.8) 79 (63.2) 81 (63.8) Reference Reference
1 hour 138 (36.4) 46 (36.2) 46 (36.8) 46 (36.2) 0.92 1.00


Classroom teaching using the blackboard 0.35 0.71
(as opposed to textbooks), n (%)z

Trial of Adjustable Glasses in Children

Less than half 67 (17.7) 23 (18.1) 24 (19.4) 20 (15.7)
Half of teaching 130 (34.4) 41 (32.3) 47 (37.9) 42 (33.1)
More than half 181 (47.9) 63 (49.6) 53 (42.7) 65 (51.2)
Child reports blackboard clearly 152 (40.2) 53 (42.1) 45 (36.0) 54 (42.5) 0.33 0.94
visible, n (%)z
Child aware of having myopia, n (%){ 348 (92.8) 117 (92.9) 114 (92.7) 117 (92.9) 0.96 1.00
Child rating of glasses appearance (in
general, not with reference to study
spectacles), n (%)*
Very ugly/ugly 109 (28.9) 35 (27.6) 39 (31.5) 35 (27.8) 0.50 0.97
Moderate/attractive/very attractive 268 (71.1) 92 (72.4) 85 (68.5) 91 (72.2) Reference Reference
Child reports having been laughed at by 53 (14.1) 18 (14.3) 19 (15.3) 16 (12.6) 0.82 0.69
classmates for wearing
glasses, n (%)*
Child has friends who wear 310 (82.4) 102 (81.6) 101 (81.4) 107 (84.2) 0.98 0.58
glasses, n (%)x
Family wealth, Median (IQR), USD 0.36 0.09
Bottom tercile (n ¼ 125, 33.2%) 2110 (1800e2670) 2080 (1520e22 700) 2110 (1820e2770) 2130 (1810e2450)
Middle tercile (n ¼ 125, 33.2%) 3850 (3400e4110) 3880 (3460e4110) 3630 (3360e4110) 3760 (3410e4090)
Top tercile (n ¼ 126, 33.5%) 13 800 (12 500e14 100) 13 700 (10 600e14 100) 13 600 (12 700e14 100) 13 900 (12 500e14 200)
Spherical equivalent with
e2 D in the 181 (47.9) 58 (46.0) 66 (52.8) 57 (44.9) 0.28 0.85
better eye, n (%)z
Baseline knowledge about
myopia, n (%)**
Eye exercises can reduce myopia 140 (36.9) 46 (36.2) 46 (36.8) 48 (37.8) 0.55 0.86
Wearing glasses will worsen 111 (29.3) 31 (24.4) 37 (29.6) 43 (33.9) 0.42 0.02
children’s vision
Myopia can be treated by wearing 144 (38.0) 47 (37.0) 46 (36.8) 51 (40.2) 0.77 0.28
glasses

(Continued)
31
 Ophthalmology Volume 127, Number 1, January 2020

were modeled by linear regression. To satisfy the requirements of

Linear regression for age, ordinal logistic regression for classroom teaching on the blackboard, family wealth and baseline knowledge about myopia, and logistic regression for other variables were used to
(1) with (3)#
Comparing
P Value intention-to-treat analysis that all randomized participants be

0.24

0.51

0.86
included in the regression analysis, we used multiple imputation in
Stata27 to impute missing data using linear regression for
continuous variables, logistic regression for binary variables, and
ordinal logistic regression for ordinal variables. We used
different models for each variable, selecting the independent
(1) with (2)#
Comparing

variables based on predictive value and availability of data. The

P Value

0.54

0.92

0.82 multiple imputation approach created 20 copies of the data, in

which missing values were imputed by chained equations. Final
results were obtained by averaging these 20 datasets using
Rubin’s rules, which ensured that the standard errors for all
regression coefficients took into account uncertainty in the
imputations and uncertainty in the estimation.28

**Baseline knowledge test included 6 items on 5-point scales: 1 (strongly agree) to 5 (strongly disagree). The proportion of strongly agree and agree is presented.
A 1-sided noninferiority test at alpha ¼ 0.05 of whether the
n [ 127 (3)
Adjustable

difference in primary outcome between other groups and the

Glasses

Standard Glasses group was less than the prespecified non-

61 (48.0)

100 (79.4)

108 (85.0)

inferiority margin was equivalent to testing whether a 2-sided 90%

confidence interval (CI) around the differences lay entirely below
the noninferiority margin.29 Both crude and adjusted wear rate and
90% CIs were calculated. In the additional testeretest analyses,
both interobserver and intraobserver repeatability were assessed
using the BlandeAltman method30 and intra-class correlation co-
efficients (ICCs) were calculated, with estimates >0.75 indicative
of good reliability. A paired-samples t test was used to assess the
n [ 125 (2)
Ready-made

differences in mean power measurements at warm and cold tem-

Glasses

peratures. All analyses were performed using Stata 12.0 (StataCorp

57 (45.6)

90 (72.0)

103 (83.1)

LP, College Station, TX).

Table 1. (Continued.)

Results
Among 6203 students who underwent baseline screening at the
selected schools, 5105 (82.3%) did not meet the eligibility criteria
and 640 (10.3%) declined to participate in the study. Of the
n [ 127 (1)

remaining 458 students (7.4%), 55% (n ¼ 252) underwent sub-

Standard
Glasses

jective refraction with cycloplegia and 45% (n ¼ 206) without

53 (41.7)

92 (72.4)

103 (81.1)

cycloplegia, according to parental preference. Subsequently, 79

D ¼ diopters; IQR ¼ interquartile range; SD ¼ standard deviation; USD ¼ US dollar.

students (17.2%) were excluded from the study: 49 (62.0%) with

compare the Ready-made and Adjustable groups with the Standard Glasses Group.

subjective refractive power >e1.00 D in either eye, 19 (24.1%)

with subjective BCVA <6/7.5 in either eye, 8 (10.1%) with
astigmatism
e2.00 D in either eye or anisometropia 2.00 D,
and 3 (3.80%) with other ocular comorbidity (Fig 2).
In total, 379 students (6.11%) were randomly allocated to study
Total n [ 379

groups: 127 (33.5%) to standard glasses, 125 (33.0%) to ready-

made glasses, and 127 (33.5%) to adjustable glasses. Intention-
to-treat analysis was performed on all these participants, with
171 (45.1)

282 (74.6)

314 (83.1)

multiple imputation as required. At the end of the study, 6 students

(1.58%) were lost to follow-up: Four students transferred schools,
1 student dropped out of school, and 1 student dropped out of the
*A total of 2 students (0.53%) had missing values.

A total of 4 students (1.06%) had missing values.

A total of 7 students (1.85%) had missing values.

A total of 3 students (0.79%) had missing values.

study. An additional participant with follow-up but without data is

reported as missing in Tables 1 to 3 below (Fig 2).
One student (0.26%) had a missing value.

At baseline, the ready-made glasses group had a higher pro-

Wearing glasses can improve clarity of
It is unnecessary to wear glasses when

portion of students from 1-child families compared with the stan-

Failure to wear glasses when needed

dard glasses group (11.2% vs. 3.97%; P ¼ 0.04) (Table 1).

the degree of myopia is modest

can affect school performancez

material on the blackboardz

Students in the adjustable glasses group had a lower mean age

(13.4 vs. 13.7 years; P ¼ 0.004), were more likely to have a
Characteristic

parent wearing glasses (30.7% vs. 18.6%; P ¼ 0.03), and more

frequently believed that wearing glasses would worsen their
vision (33.9% vs. 24.4%; P ¼ 0.02), all compared with the
standard glasses group. Other characteristics, including gender,
glasses ownership, rural school location, parental education,
parental out-migration, study time, classroom variables, percep-
tion of glasses, family wealth, and other aspects of myopia
knowledge, did not differ between randomization groups at base-
line (Table 1).
{
#
y
z
x

32
 Wang et al 
Trial of Adjustable Glasses in Children

Table 2. Distribution of Visual Acuity of Better-Seeing Eye Expressed as n (%), before Receiving and after Wearing Study Spectacles
(N ¼ 379)

Baseline Presenting VA before Receiving the Study Spectacles Corrected VA with Study Spectacles
Ready-made Ready-made
Standard Glasses Glasses Adjustable Glasses Standard Glasses Glasses Adjustable Glasses
VA N ¼ 127 N ¼ 125 N ¼ 127 N ¼ 111y N ¼ 109y N ¼ 104z
‡6/6 d d d 23 (20.7) 18 (16.5) 63 (60.6)
6/7.5 d d d 43 (38.7) 38 (34.9) 28 (26.9)
6/9.5 d d d 36 (32.4) 29 (26.6) 11 (10.6)
6/12 31 (24.4) 29 (23.2) 25 (19.7) 7 (6.31) 20 (18.4) 1 (0.96)
6/15 27 (21.3) 27 (21.6) 40 (31.5) 2 (1.80) 4 (3.67) 0 (0.00)
6/19 27 (21.3) 21 (16.8) 19 (15.0) 0 (0.00) 0 (0.00) 1 (0.96)
6/24 18 (14.2) 25 (20.0) 24 (18.9) d d d
6/30 16 (12.6) 11 (8.80) 11 (8.66) d d d
6/38 5 (3.94) 10 (8.00) 5 (3.94) d d d
6/48 3 (2.36) 2 (1.60) 3 (2.36) d d d
6/60 d d d d d d
6/75 d d d d d d
Mean VA (logMAR) 6/19 (0.491) 6/19 (0.500) 6/18 (0.487) 6/8 (0.12) 6/8 (0.14) 6/6 (0.03)
P value for comparison / 0.639* 0.844* / 0.281x <0.001x
with standard glasses group

logMAR ¼ logarithm of the minimum angle of resolution; VA ¼ visual acuity.

*Linear regression was used for the comparison.
y
A total of 16 students (13%) had missing data.
z
A total of 23 students (18%) had missing data.
x
Logistic regression adjusting for baseline presenting VA (logarithm of the minimum angle of resolution) was used to compare the Ready-made and
Adjustable groups with the Standard Glasses Group on the proportion of participants having best-corrected VA with study spectacles 6/7.5.

The mean presenting VA did not differ between groups at group was associated with an 11% reduction in wear proportion
baseline (Table 2). However, after receiving study spectacles, a compared with the standard glasses group (90% CI, e18.7 to
significantly higher proportion of students in the adjustable e2.85). This did not meet the threshold for inferiority because
glasses group achieved BCVA 6/7.5 compared with the the upper limit of the 1-sided 90% CI for difference in wear
standard glasses group (87.5% vs. 59.4%; P < 0.001). proportions did not cross the prespecified noninferiority margin
Table 3 shows the glasses wear proportions and attitudes of of 20%, although the CI for the unadjusted mean difference
students toward study spectacles. For the primary outcome based (e12.7%, 90% CI, e20.8 to e4.57) did cross the noninferiority
on head teachers’ covert assessment, those in the standard margin. Other predictors of wear proportion included parental
glasses group had a significantly greater mean wear proportion glasses wear and disliking the appearance of study glasses that
compared with the adjustable glasses group (58% vs. 45%; 90% had a e12.5% (90% CI, e20.4 to e4.66) and 16.5% (90% CI,
CI, 4.76e20.9; P ¼ 0.01). Between the groups, there were no e23.7 to e9.27) reduction in compliance, respectively.
significant differences in wear proportions based on unannounced Mean change in quality of life scores (NEI RQL-42, 0-100
inspection (44.1% vs. 33.9%, P ¼ 0.89) or students’ self-report scale) at study closeout did not differ significantly between stan-
(90.2% vs. 84.8%, P ¼ 0.64). Between standard and ready-made dard glasses (5.53; SD, 4.47), ready-made glasses (5.27; SD, 4.82),
glasses groups, there were no significant differences in wear pro- and adjustable glasses (5.68; SD, 4.34) groups. In multiple
portions assessed by head teachers’ covert assessment (58% vs. regression models, the factors associated with a greater increase in
52%, P ¼ 0.24), unannounced inspection (44.1% vs. 43.2%, P ¼ quality of life included baseline glasses ownership (b 1.22; 95%
0.89), or self-report (83.3% vs. 93.9%, P ¼ 0.18). The expected CI, 0.38e2.06), awareness of own myopia status (b 2.34; 95% CI,
number of covert teacher assessments was 12, the mean number of 0.79e3.88), and baseline knowledge that myopia can be treated
assessments was 11, and 83.6% of children had 10 assessments. with glasses (b 1.32; 95% CI, 0.53e2.11) (data not shown).
Regarding the reasons for not wearing study glasses, a higher Systematic assessment of study glasses at the end of the study
proportion of the adjustable glasses group cited poor cosmetic indicated that several types of damage were more common in
appearance as their primary reason for not wearing glasses adjustable compared with standard glasses (scratches: 32.7% vs.
compared with the standard glasses group (23.8% vs. 6.67%, P ¼ 18.3%, P < 0.05; frame damage: 18.6% vs. 2.61%, P < 0.001; and
0.005). With regard to the proportion of participants reporting missing screws 9.73% vs. 1.74%, P < 0.001) but that other modes
being satisfied or very satisfied with the appearance of study of damage were less common in the adjustable glasses (broken
glasses, there were no significant differences between the standard lenses: 0.88% vs. 13.0%, P < 0.001; chipped lenses: 0.0% vs.
glasses and ready-made glasses groups (P ¼ 0.30) or between 7.83%, P < 0.01; damaged temples: 3.77% vs. 12.2%, P < 0.05).
standard glasses and adjustable glasses (P ¼ 0.968). Likewise, no The mean spherical equivalent refractive power of adjustable
significant differences were observed in the value attributed to glasses increased by 0.65 D (95% CI, 0.52e0.79 D; P < 0.001)
study glasses between these groups (P ¼ 0.69 for standard vs. over 2 months of wear during the study, whereas the power of
ready-made and P ¼ 0.553 for standard vs. adjustable glasses, standard lenses and ready-made lenses showed no significant
respectively). change. Interobserver and intraobserver repeatability of evaluation
In the multiple regression model that adjusted for baseline of lens power by lensometry among study personnel were signif-
myopia knowledge (Table 4), allocation to the adjustable glasses icantly worse for adjustable glasses (ICC, 0.09 and 0.10; 95% CI,

33
 Ophthalmology Volume 127, Number 1, January 2020

Table 3. Wear of and Attitudes toward Study Spectacles by Group

Standard Ready-made Adjustable P Value P Value

Glasses Glasses Glasses Comparing Comparing
Category N [ 127 (1) N [ 125 (2) N [ 127 (3) (1) with (2)* (1) with (3)*
Repeated teacher assessment of wearing study glasses, 58% (39) 52% (41) 45% (37) 0.24 0.01
Mean proportion (SD)y
Observed wearing study spectacles at unannounced 56 (44.1) 54 (43.2) 43 (33.9) 0.89 0.10
inspection, n (%)
Self-reported wear of study spectacles at end line
(N ¼ 379), nz (%)
Never wear 12 (9.76) 10 (8.06) 19 (15.2) Reference Reference
Wear for study only/usually wear 111 (90.2) 114 (91.9) 106 (84.8) 0.64 0.20
Self-reported wear of own spectacles at baseline
(N ¼ 111), n (%)
Never wear 7 (16.7) 2 (6.06) 7 (19.4) Reference Reference
Wear for study only/usually wear 35 (83.3) 31 (93.9) 29 (80.6) 0.18 0.75
Reason for not wearing study glasses at end line
(N ¼ 379), n (%)
Fear worsening VA from wear 15 (20.0) 20 (26.3) 8 (9.52) 0.36 0.07
Laughed at by classmates 6 (8.00) 5 (6.58) 12 (14.3) 0.74 0.22
No perceived need 15 (20.0) 11 (14.5) 9 (10.7) 0.37 0.11
Not convenient for daily activities 21 (28.0) 31 (40.8) 23 (27.4) 0.10 0.93
Not comfortable 10 (13.3) 8 (10.5) 8 (9.52) 0.60 0.45
No improvement in vision 2 (2.67) 0 (0.00) 1 (1.19) / 0.51
Wear only on special occasions 17 (22.7) 23 (30.3) 19 (22.6) 0.29 0.99
Don’t like the appearance of the glasses 5 (6.67) 4 (5.26) 20 (23.8) 0.72 0.005
Glasses broken 6 (8.00) 4 (5.26) 11 (13.1) 0.50 0.30
Other 9 (12.0) 7 (9.21) 6 (7.14) 0.58 0.30
Satisfaction with study spectacles 0.30 0.97
(N ¼ 379), n (%)z
Very satisfied 10 (8.13) 6 (4.84) 10 (8.00)
Satisfied 61 (49.6) 62 (50.0) 64 (51.2)
Neither satisfied nor dissatisfied 44 (35.8) 40 (32.3) 40 (32.0)
Dissatisfied 6 (4.88) 13 (10.5) 9 (7.20)
Very dissatisfied 2 (1.63) 3 (2.42) 2 (1.60)
Value attributed to the study spectacles 0.69 0.55
(N ¼ 379), n (%)z
Most valued possession 8 (6.50) 10 (8.06) 7 (5.60)
High value 29 (23.6) 27 (21.8) 30 (24.0)
Moderate value 64 (52.0) 59 (47.6) 59 (47.2)
Some value 19 (15.5) 26 (21.0) 27 (21.6)
No value 3 (2.44) 2 (1.61) 2 (1.60)

SD ¼ standard deviation; VA ¼ visual acuity.

*Linear regression was used to compare rates of wearing study glasses by continuous teacher assessment, ordinal logistic regression for satisfaction with study
glasses and value attributed to the study spectacles, and logistic regression for other variables.
y
The repeated teacher assessment of wearing study glasses was calculated as the number of times wearing glasses/total number of separate observations.
z
There were 7 children (1.85%) with missing values: 4 (3.15%) in the Standard Glasses Group, 1 (0.80%) in the Ready-made Glasses Group, and 2 (1.57%)
in the Adjustable Glasses Group.

e0.38 to 0.47 and e0.38 to 0.41, respectively) compared with standard spectacles (45% vs. 58%, P ¼ 0.01). Nevertheless,
standard glasses (ICC, 1.0 and 1.0, 95% CI, 1.0e1.0 and 1.0e1.0, both of these rates represent a substantial improvement on
respectively, P < 0.001 for both), which may have explained the 15% to 20% of wear reported among rural and urban
apparent power instability in the adjustable lenses. No significant migrant Chinese children needing glasses in our population
power change was observed in the adjustable lenses with heating or
studies,4,10 suggesting that there may be a useful role for
cooling (data not shown). No incidents of children being injured or
otherwise harmed by wear of study glasses were reported by self-refraction in such low-resource settings, where refrac-
teachers during the study. tive capacity is limited. The unadjusted, but not the adjusted,
difference met the prespecified definition for inferiority. The
secondary measures of wear did not differ significantly be-
Discussion tween standard glasses and both adjustable and ready-made
glasses.
On the basis of the primary outcome of head teachers’ Self-reported satisfaction and value attributed to the
covert assessments, adjustable glasses were worn by a sta- study glasses did not differ between groups, although chil-
tistically significantly smaller proportion of children than dren did find the adjustable glasses less attractive and

34
 Wang et al 
Trial of Adjustable Glasses in Children

Table 4. Intention-to-Treat Analysis of the Effect of Potential Predictors Glasses Wear (Continuous Teacher Assessment) during the
Study

Simple Regression Multiple Regressiony

Parameter b (90% CI) P Value b (90% CI) P Value
Study group
Standard glasses Reference Reference
Ready-made glasses e5.46% (e13.6% to 2.71%) 0.27 e4.53% (e12.5% to 3.41%) 0.35
Adjustable glasses e12.7% (e20.8% to e4.57%) 0.01 e10.8% (e18.7% to e2.85%)* 0.03
Age (yrs) e1.53% (e5.33% to 2.28%) 0.51
Male sex 1.91% (e4.76% to 8.58%) 0.64
Owning glasses at baseline e0.57% (e7.89% to 6.74%) 0.90
Rural school e0.001% (e6.88% to 6.88%) 1.00
Only child in family 3.38% (e8.65% to 15.4%) 0.64
At least 1 parent wears glasses e12.8% (e20.8% to e4.82%) 0.008 e12.5% (e20.4% to e4.66%) 0.009
1 or both parents with 12 yrs of education e3.76% (e10.6% to 3.11%) 0.37
Both parents away from the home the majority of time e2.85% (e10.6% to 4.89%) 0.54
Self-reported study time each day after school
<1 hour Reference
1 hour 4.89% (e2.10% to 11.9%) 0.25
Classroom teaching on the blackboard
Less than half of teaching Reference
More than half e1.52% (e10.3% to 7.27%) 0.78
Clarity of blackboard e2.10% (e8.93% to 4.72%) 0.61
Aware of having myopia 6.17% (e6.94% to 19.3%) 0.44
Feels study glasses are ugly or very ugly e16.6% (e23.8% to e9.36%) <0.001 e16.5% (e23.7% to e9.27%) <0.001
Laughed at by classmates e6.91% (e16.6% to 2.79%) 0.24
Has friends owning glasses 0.77% (e8.18% to 9.71%) 0.89
Family wealth
Bottom tercile Reference
Middle tercile 0.51% (e7.71% to 8.73%) 0.92
Top tercile e7.66% (e15.8% to 0.51%) 0.12
Baseline knowledge about myopia
Eye exercises can treat the problem of myopia 2.49% (e4.47% to 9.44%) 0.56
Wearing glasses will worsen children’s vision 0.57% (e6.80% to 7.95%) 0.90
Myopia can be treated by wearing glasses 3.07% (e3.83% to 9.97%) 0.46
It is unnecessary to wear glasses when the degree of e8.47% (e15.2% to e1.78%) 0.04 e5.78% (e12.4% to 0.80%) 0.15
myopia is modest
Failure to wear glasses when needed will affect grades 6.08% (e1.58% to 13.7%) 0.19
Wearing glasses can improve clarity of material on the 3.14% (e5.74% to 12.0%) 0.56
blackboard

b ¼ parameter estimate; CI ¼ confidence interval.

Boldface indicates results significant at the P < 0.05 level.
*Linear regression model was used for continuous assessment of wearing rate.
y
Study group and other variables with P < 0.05 in the simple regression were included in the multiple regression.

finding the glasses unattractive was a significant determinant subjective cycloplegic refraction.11 This phenomenon may
of wear. This is consistent with findings from a similar have accounted for the less than optimal visual results in
population who indicated that the round lens design of the standard glasses group in the current study. Second,
adjustable glasses was unpopular with both parents and our testeretest protocol suggests that measuring the refrac-
students.17 tive power of lenses using auto-lensometry may induce
The BCVA was significantly better with adjustable significant inaccuracies, leading to a poorer visual outcome
glasses, in contrast to similar trials in which BCVA 6/7.5 when using the results of self-refraction to dispense standard
was achieved by 99.1% to 99.8% of students undergoing glasses, as done previously.16 Children wore the adjustable
subjective refraction by experienced optometrists and 92.4% glasses directly in the current study, and so this reduction in
to 96.9% of students using adjustable glasses.14,15 More- accuracy was avoided.
over, in a trial by Zhou et al,16 which used self-refraction as This study also assessed the durability and power stability
a tool for dispensing standard glasses, 20.7% of the self- of adjustable glasses under conditions of actual wear. Dura-
refraction group failed to achieve BCVA 6/7.5. bility results were mixed in the current setting: Lens scratches
There are 2 potential reasons to account for the better and frame damage were more common in the adjustable
BCVA found in our adjustable glasses group: First, previous glasses group, whereas more severe damage to the lens,
research has indicated that refractionists in China tend to including breakage, was less common than for standard
undercorrect children with myopia, even when using glasses. Adjustable lenses had significantly greater apparent

35
 Ophthalmology Volume 127, Number 1, January 2020

power instability compared with standard lenses, but 2. Bourne RRA, Stevens GA, White RA, et al. Causes of vision
testeretest assessment suggests this may have been caused by loss worldwide 1990e2010: a systematic analysis. Lancet
significant intra-user and inter-user variability in reading the Glob Health. 2013;1:339e349.
power of the adjustable lenses. Variation in temperature be- 3. United Nations Sustainable. Development Goals. https://sus-
tween baseline and study end appeared unlikely to explain the tainabledevelopment.un.org/?menu¼1300; 2015. Accessed
January 8, 2017.
apparent power instability in adjustable lenses. 4. Ma X, Zhou Z, Yi H, et al. Effect of providing free glasses on
This study is the first of which we are aware to assess children’s educational outcomes in China: cluster-randomized
the use of adjustable glasses by children in real-world controlled trial. BMJ. 2014;349:g5740.
circumstances. 5. Resnikoff S, Pascolini D, Mariotti SP, et al. Global magnitude
of visual impairment caused by uncorrected refractive errors in
Study Strengths and Limitations 2004. Bull World Health Organ. 2008;86:63e70.
6. He M, Zeng J, Liu Y, et al. Refractive error and visual
Strengths of the current study include the randomized impairment in urban children in southern China. Invest Oph-
controlled design and the setting of the trial in rural China, thalmol Vis Sci. 2004;45:793e799.
where refractive capacity is known to be insufficient. Limi- 7. Zhao J, Pan X, Sui R, et al. Refractive error study in children:
tations must also be acknowledged. The small geographic results from Shunyi district, China. Am J Ophthalmol.
2000;129:427e435.
area limits wider generalizability, and the study protocol,
8. Sun HP, Li A, Xu Y, et al. Secular trends of reduced visual
which did not include BCVA testing at study closeout, made acuity from 1985 to 2010 and disease burden projection for
it difficult to assess if there was any meaningful visual impact 2020 and 2030 among primary and secondary school students
of the apparent power instability of adjustable glasses. Some in China. JAMA Ophthalmol. 2015;133:262e268.
statistically significant differences existed between study 9. Ma X, Congdon N, Yi H, et al. Safety of spectacles for chil-
groups at baseline, although these were generally unlikely to dren’s vision: a cluster-randomized controlled trial. Am J
be of clinical importance (e.g., 13.7 vs. 13.4 years). We report Ophthalmol. 2015;160:897e904.
both adjusted and unadjusted analyses of our main outcome, 10. Wang X, Yi H, Lu L, et al. Population prevalence of need for
as called for in the revised Consolidated Standards of spectacles and spectacle ownership among urban migrant chil-
Reporting Trials guidelines,31 but did not prespecify the use dren in eastern China. JAMA Ophthalmol. 2015;133:1399e1406.
11. Zhou Z, Zeng J, Ma X, et al. Accuracy of rural refractionists in
of adjustment, as should have been done under
western China. Invest Ophthalmol Vis Sci. 2014;55:154e161.
Consolidated Standards of Reporting Trials. Despite this, 12. Zhang M, Lu H, Gao Y, et al. Visual morbidity due to inac-
the advantages of reported adjusted outcomes remain curate spectacles among school children in rural China: the See
substantial.32 Although cost-effectiveness is a crucially Well to Learn Well Project, report 1. Invest Ophthalmol Vis
important issue in areas of low resources, the current study Sci. 2009;50:2011e2017.
was not designed to assess this outcome. The authors thought 13. Karnani A, Garrette B, Kassalow J, et al. Better vision for the
that given the novelty of adjustable glasses, which have not poor. Stanford Social Innovation Review. 2011;9:66e71.
previously been tested in children under actual conditions of 14. He M, Congdon N, MacKenzie G, et al. The child self-
wear to the best of our knowledge, the first priority was to refraction study: results from urban Chinese children in
assess basic aspects of wearability (e.g., compliance, satis- Guangzhou. Ophthalmology. 2011;118:1162e1169.
15. Zhang M, Zhang R, He M, et al. Self-correction of refractive
faction, stability of power). Further studies are needed to
error among young people in rural China: results of cross
assess cost-effectiveness, particularly once adjustable glasses sectional investigation. BMJ. 2011;343:d4767.
are in regular production and a reliable price can be assigned. 16. Zhou Z, Chen T, Jin L, et al. Self-refraction, ready-made
The different appearance of study glasses did not allow for the glasses and quality of life among rural myopic Chinese chil-
possibility of masking of teachers to children’s study group dren: a non-inferiority randomized trial. Acta Ophthalmol.
assignment, although teachers were unaware of the study 2017;95:567e575.
hypothesis. Finally, regarding the choice of the noninferiority 17. Zhou Z, Kecman M, Chen T, et al. Spectacle design prefer-
margin, we have chosen 20% as the degree of difference that ences among Chinese primary and secondary students and
would be expected to be clinically meaningful. It has been their parents: a qualitative and quantitative study. PLoS One.
suggested to use results of previous trials to support a non- 2014;9:e88857.
18. Douali MG, Silver JD. Self-optimised vision correction with
inferiority margin,33 but that was not possible in this case
adaptive spectacle lenses in developing countries. Ophthalmic
because other trials of children’s wear of adjustable glasses Physiol Opt. 2004;24:234e241.
could not be located in our literature search. Despite these 19. Zeng Y, Keay L, He M, et al. A randomized, clinical trial
limitations, this study adds significant new information on a evaluating ready-made and custom spectacles delivered via a
novel modality for addressing uncorrected refractive error school-based screening program in China. Ophthalmology.
that could be of value in settings that lack adequate 2009;116:1839e1845.
refractive capacity. 20. Brady CJ, Villanti AC, Gandhi M, et al. Visual function after
correction of distance refractive error with ready-made and
custom spectacles: a randomized clinical trial. Ophthalmology.
References 2012;119:2014e2020.
21. Historical data. 2016 GDP per capita for Chinese administra-
tive divisions. https://en.wikipedia.org/wiki/List of Chinese
1. Fall HB. Millennium development goals and eye health. In- administrative divisions by GD P_per_capita#2016. Accessed
dian J Ophthalmol. 2012;60:411e415. September 10, 2018.

36
 Wang et al 
Trial of Adjustable Glasses in Children

22. Ferris III FL, Kassoff A, Bresnick GH, et al. New visual acuity 28. Royston Patrick. Multiple imputation of missing values. Stata
charts for clinical research. Am J Ophthalmol. 1982;94:91e96. J. 2004:227e241.
23. Berry S, Mangione CM, Lindblad AS, et al. Development of 29. Mulla SM, Scott IA, Jackevicius CA, et al. How to use a
the National Eye Institute refractive error correction quality of noninferiority trial: users’ guides to the medical literature.
life questionnaire: focus groups. Ophthalmology. 2003;110: JAMA. 2012;308:2605e2611.
2285e2291. 30. Altman DG, Bland JM. Measurement in medicine: the anal-
24. Hays RD, Mangione CM, Ellwein L, et al. Psychometric ysis of method comparison studies. Statistician. 1983;32:
properties of the National Eye Institute-Refractive Error Quality 307e317.
of Life instrument. Ophthalmology. 2003;110:2292e2301. 31. Altman DG, Schulz KF, Moher D, et al, for the CONSORT
25. Hays RD, Spritzer KL. National Eye Institute Refractive Error Group: the revised CONSORT statement for reporting ran-
Quality of Life Instrument (NEI RQL-42), Version 1.0: A domized trials: explanation and elaboration. Ann Intern Med.
Manual for Scoring and Use. Los Angeles, CA: Rand Health 2001;134:663e694.
Sciences; 2002. 32. Yu L, Chan A, Hopewell S, et al. Reporting on covariate
26. Nutt D, Allgulander C, Lecrubier Y, et al. Establishing non- adjustment in randomised controlled trials before and after
inferiority in treatment trials in psychiatry: guidelines from revision of the 2001 CONSORT statement: a literature review.
an Expert Consensus Meeting. J Psychopharmacol. 2008;22: Trials. 2010;11:59.
409e416. 33. Althunian TA, de Boer A, Groenwald RHH, Klungel OH.
27. Jones B, Jarvis P, Lewis JA, et al. Trials to assess equivalence: Defining the non-inferiority margin and analysing non-inferiority:
the importance of rigorous methods. BMJ. 1996;313:36e39. an overview. Br J Clin Pharmacol. 2017;83:1631e1642.

Footnotes and Financial Disclosures

Originally received: March 21, 2019. Funded by the Chinese government (grant no. 953) under the Thousand
Final revision: July 31, 2019. Man Plan program and a grant to the Nuffield Laboratory of Ophthal-
Accepted: August 4, 2019. mology (University of Oxford) from the Partnership for Child Development
Available online: August 14, 2019. Manuscript no. 2019-625. (Imperial College of Science, Technology and Medicine) under the World
1
Department of Ophthalmology, First Affiliated Hospital of Sun Yat-sen Bank’s FY2009 Development Grant Facility Window 1. The adjustable
University, Guangzhou, Guangdong Province, China. spectacles used in this study were provided free of charge by Adaptive
Eyecare, a company involved in the production of adjustable glasses.
2
State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Centre,
ClinicalTrials.gov Identifier: The study was registered at ClinicalTrials.gov,
Sun Yat-sen University, Guangzhou, Guangdong Province, China.
registration number NCT02529540, Date of Trial registration: August 16,
3
Centre for Public Health, Queen’s University Belfast, Belfast, United 2015, Study commencement: September 1, 2015.
Kingdom. HUMAN SUBJECTS: Human subjects were included in this study. The
4 human ethics committees at the Zhongshan Ophthalmic Center, Sun Yat-
Department of Ophthalmology, Hospital of Guangdong Medical Univer-
sity, Zhanjiang, Guangdong Province, China. sen University (Guangzhou, China), and local Boards of Education in
5 Guangdong Province approved the study. All research adhered to the tenets
Department of Ophthalmology, Henan Provincial People’s Hospital,
of the Declaration of Helsinki. All participants provided informed consent.
Zhengzhou, Henan Province, China.
6
No animal subjects were used in this study.
Department of Ophthalmology, Gaozhou Traditional Chinese Medicine
Author Contributions:
Hospital, Maoming, Guangdong Province, China.
7
Conception and design: Wang, Zhang, Tang, Jin, Huang, Wang, Chen,
Department of Ophthalmology, Xinyi Traditional Chinese Medicine Xiao, Zhou, Wilson, Silver, Moore, Congdon
Hospital, Maoming, Guangdong Province, China.
Data collection: Wang, Zhang, Tang, Wang, Chen, Zhu, Xiao, Wang, Zhou,
8
University of Arizona, Department of Ophthalmology and Vision Science, Tang, Liang, Congdon
Tucson, Arizona. Analysis and interpretation: Wang, Tang, Jin, Xiao, McAneney, Silver,
9
Brien Holden Vision Institute, Sydney, Australia. Moore, Congdon
10
Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, Obtained funding: Congdon
Oxford, United Kingdom. Overall responsibility: Wang, Zhang, Tang, Jin, Huang, Wang, Chen, Zhu,
11 Xiao, Wang, Zhou, Tang, Liang, Crescioni, Wilson, McAneney, Silver,
Centre for Vision in the Developing World Charitable Foundation, St.
Moore, Congdon
Catherine’s College, Oxford, United Kingdom.
12
New England College of Optometry, Boston, Massachusetts. Abbreviations and Acronyms:
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval;
Financial Disclosure(s):
D ¼ diopters; ICC ¼ intra-class correlation coefficient; NEI RQL-
The author(s) made the following disclosres:
42 ¼ National Eye Institute Refractive Error Quality of Life-42;
N.C.: Support e Thousand Man Plan grant from the Chinese government SD ¼ standard deviation; SDG ¼ Sustainable Development Goal;
and by the Ulverscroft Foundation (UK); Employed by Orbis International, USD ¼ US dollar; VA ¼ visual acuity.
a non-governmental organisation involved in distribution of glasses to
children, as Director of Research. The study sponsor played no role in study Correspondence:
design and the collection, analysis, and interpretation of data, the writing of Nathan Congdon, MD, MPH, Centre for Public Health, Royal Victoria
the article or the decision to submit it for publication. Hospital, Queen’s University Belfast, Belfast, 274 Grosvenor Rd, Belfast
UK BT12 6BA; and State Key Laboratory of Ophthalmology, Zhongshan
J.D.S.: Shareholder and director of Adaptive Eyecare Ltd; Shareholder in
Ophthalmic Center, Sun Yat-sen University, 54 S. Xianlie Road, Guangz-
Adlens Ltd., a company involved in the development and commercializa-
hou, People’s Republic of China. E-mail: [email protected].
tion of adjustable lenses.

37
Diagnostic Accuracy of Technology-based
Eye Care Services
The Technology-based Eye Care Services Compare Trial
Part I
April Y. Maa, MD,1,2 Charles M. Medert, MD,3 Xiaoqin Lu, MD,1,2 Rabeea Janjua, MD,1
Ashley V. Howell, MPH,4 Kelly J. Hunt, PhD,4 Sarah McCord, MD,5 Annette Giangiacomo, MD,1
Mary G. Lynch, MD2,6

Purpose: Ophthalmologic telemedicine has the ability to provide eye care for patients remotely, and many
countries have used screening tele-ophthalmology programs for several years. One such initiative at the Veterans
Affairs (VA) Healthcare System is Technology-based Eye Care Services (TECS). The TECS services are located in
primary care clinics and provide basic screening eye care, including vision, refraction, and retinal photography. Eye
care providers (“readers”) review the clinical data and recommend appropriate follow-up. One of the most common
referrals from TECS has been for glaucoma, and this study was powered for glaucoma/glaucoma suspect detection.
The current study was undertaken to identify aspects of the protocol that could be refined to enhance accuracy.
Design: Prospective comparison between the standard TECS protocol versus a face-to-face (FTF) exami-
nation on 256 patients, all of whom had no known history of significant ocular disease.
Participants: Patients with no known ocular disease who were scheduled for an in-person eye appointment
at the Atlanta VA. Patients underwent screening through the TECS protocol and received an FTF examination on
the same day (gold standard). The TECS readers were masked to the results of the FTF examination.
Main Outcome Measures: Percent agreement, kappa, sensitivity, and specificity were calculated for the
TECS readers’ interpretations versus the FTF examination.
Results: The TECS readers showed substantial agreement for cataract (k
0.71) and diabetic retinopathy
(k
0.61) and moderate to substantial agreement for glaucoma/glaucoma suspect (k
0.52) compared with an
FTF examination. Age-related macular degeneration (AMD) showed moderate agreement (k
0.34). Percent
agreement with the TECS protocol was high (84.3%e98.4%) for each of the disease categories. Overall sensi-
tivity and specificity were
75% and
55%, respectively, for any diagnosis resulting in referral. Inter-reader and
intra-reader agreement was substantial for most diagnoses (k > 0.61) with percent agreements ranging from 66%
to 99%.
Conclusions: Our results indicate that the standard TECS protocol is accurate when compared with an FTF
examination for the detection of common eye diseases. The inclusion of additional testing such as OCT could
further enhance diagnostic capability. Ophthalmology 2020;127:38-44 Published by Elsevier on behalf of the
American Academy of Ophthalmology

Supplemental material available at www.aaojournal.org.

Telemedicine is defined as care given to patients when common uses of store and forward ophthalmologic tele-
the provider and patient are separated by distance, time, medicine is diabetic teleretinal imaging (TRI). Teleretinal
or both. Ophthalmology is an ideal specialty for tele- screening for diabetes is used worldwide to reduce
medicine because diagnoses made during face-to-face blindness from diabetic retinopathy (DR).1-3 Diabetic
(FTF) visits are often based on pattern recognition and TRI is well validated, and many studies have illustrated
the use of multiple imaging modalities. Images and that other common ocular diseases such as glaucoma,
clinical information such as vision and eye pressure can cataract, and age-related macular degeneration (AMD)
be collected remotely and then transmitted electronically can also be incidentally detected with these photo-
to a physician stationed at another location for interpre- graphs.2-6 This knowledge has led to the expansion of
tation. This form of telemedicine is called “store various tele-ophthalmology programs to use fundus
and forward” or “asynchronous,” and one of the most photographs to screen for other common eye conditions.

38 Published by Elsevier on behalf of the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.07.026

ISSN 0161-6420/19
 Maa et al 
Diagnostic Accuracy of the TECS Protocol

The Veterans Affairs (VA) Health Administration has a refraction with a phoropter was performed using the auto-refractor’s
particular interest in novel telemedicine interventions prescription as a starting point. Distance and near best-corrected
because the VA Health Administration is one of the largest spectacle visual acuity were recorded. Pupils, intraocular pressure
integrated healthcare systems in the United States with more (iCare tonometer; Raleigh, NC), central corneal thickness (Accutome
Pachpen; Malvern, PA), and anterior chamber depth (using a Finhoff
than 5.5 million patients,7 many of whom live in rural
transilluminator; Welch Allyn, Skaneateles Falls, NY) were
communities. The VA has long been at the forefront of measured. The patient’s eyes were dilated using 1% tropicamide
using telemedicine tools to decrease health disparities of drops. Once dilated, a Canon CX-1 camera (Canon Inc, Tokyo,
the medically underserved because barriers to Japan) was used to collect for each eye 1 external and 3 non-
telemedicine, such as reimbursement and licensure, are stereoscopic, 45 field, color fundus photographs according to the
mitigated in a single integrated healthcare system. Since VA diabetic teleretinal protocol10 (Fig S1, available at
2006, the VA has used a national diabetic TRI program to www.aaojournal.org). Finally, the patient received an FTF
screen for DR. In 2015, the Atlanta VA developed examination by a comprehensive ophthalmologist (A.Y.M.). The
Technology-based Eye Care Services (TECS), an exten- FTF examiner would indicate whether the patient needed a follow-
sion of the TRI program that provides broader eye screening up visit to the Eye Clinic for further testing or initiate treatment.
At the end of each patient’s visit, the FTF physician completed a
and eyeglasses to all eligible Veterans regardless of diabetic
standardized reporting form specifically detailing whether there was a
status. The most common referral from TECS has been for surgical cataract (defined as best-corrected vision worse than 20/40 or
glaucoma suspect or frank glaucoma.8,9 The current study glare vision worse than 20/40), glaucoma suspect/glaucoma, AMD,
was undertaken to further investigate the TECS protocol and or DR if the patient was diabetic.
identify aspects of the process that could be further refined Study patients were assigned a code by research staff. Each
to enhance accuracy. patient’s history, clinical data, and ocular photographs were
de-identified and placed into a secure research database
(REDCap).11 The de-identified information was transmitted to 2
Methods ophthalmologists (reader 1, R.J.; reader 2, X.L.) who individually
reviewed the information and provided interpretations in accor-
This project was approved by the Emory University Institutional
dance with established TECS reading guidelines.8 Readers did not
Review Board and the VA Research and Development Committee.
know the patient’s true identity, had never met the patient in
This project conformed to the tenets in the Declaration of Helsinki
person, and did not have access to the patient’s Computerized
and was Health Insurance Portability and Accountability Act
Patient Record System medical chart. Readers were also masked
compliant. The study was registered at clinicaltrials.gov under the
to the examining physician’s findings and each other’s
identifier NCT02558712. This research was partially funded by the
interpretations. The reading physicians interpreted the TECS
Atlanta Clinical and Translational Science Institute; however, no
information and documented their findings on a REDCap case
conflict of interest exists for any of the authors.
report form that was identical to the FTF physician’s form. Three
Participants were recruited over a 2-year period, from March
months after completion of enrollment, each reader had 150
2015 to December 2017. Power calculations were based on the
patients randomly selected for a second read. Studies were
expected prevalence of glaucoma suspect/glaucoma in the Veteran
re-read by the original reader. On the second interpretation,
population. The trial was powered for glaucoma detection, because
readers were masked to their initial read and repeated the same
this is a common disease that is asymptomatic in its earliest stages
procedure described and re-documented their findings on REDCap
and presents the greatest challenge for obtaining consensus because
case report forms.
the disease is not diagnosed on the basis of visual criteria alone
Data were analyzed using SAS statistical software (SAS
(unlike AMD or DR).8 A sample size of 250 produces a 2-sided
Institute Inc, Cary, NC). Five diagnostic categories were created:
95% confidence interval with widths equal to 0.127, 0.117, and
surgical cataracts, glaucoma suspect/glaucoma, AMD, DR, and
0.078 for kappa statistics of 0.5, 0.7, and 0.9, respectively.
any condition requiring referral. Each diagnosis category was
The Atlanta VA Eye Clinic offers routine appointments in the
recorded as present or not present. We measured concordance
“New Comprehensive Clinic” (NCC) for patients who have not had
between diagnoses obtained from the TECS protocol with those
an examination for 2 or more years. These patients have no known
obtained from FTF visits using percent agreement and Cohen’s
ocular disease and are presenting for a baseline assessment.
kappa statistics. The screening performance of the TECS protocol
Recruitment letters were mailed to patients who were already
was assessed with sensitivity and specificity measures, using the
scheduled in the NCC informing them of the study, and patients self-
FTF visits as the gold standard. We also calculated percent
selected to participate in the trial. Once patients agreed to participate,
agreement and kappa statistics to compare diagnostic classifica-
their Computerized Patient Record System chart was reviewed to
tions performed by the 2 readers (inter-reader agreement) and for
confirm that there was no known history of macular degeneration
the same reader 90 days apart (intra-reader agreement). All sta-
(AMD), glaucoma, visually significant cataract, moderate-to-severe
tistical tests were 2-sided and considered significant at an alpha
DR, or macular edema. Patients with “glaucoma suspect” history
0.05 level.
were excluded if they had documented visual field changes or history
of therapy. On the day of their NCC visit, informed consent was
obtained from eligible participants and a full TECS screening pro-
tocol was initiated. The TECS protocol included a detailed chief Results
complaint, ocular, medical, social, and family history. Distance vision
with present correction (if available) was assessed using a Marco A total of 256 patients were recruited in the 2-year period. Table 1
ARK-1S auto-refractor (Jacksonville, FL) in both eyes. The auto- illustrates the demographics of the study population. Most patients
refractor was used to obtain an auto-refraction, and the vision was enrolled in the study were male (86.7%) and African American
reassessed through the Marco unit with the auto-refraction in place. (61.3%). One-quarter of the subjects had a history of eye trauma
Then the patient was brought to a standard eye lane, and manifest or a family history of significant eye diagnoses or blindness.

39
 Ophthalmology Volume 127, Number 1, January 2020

Table 1. Characteristics of Study Participants (N ¼ 256) Discussion

Participant Characteristics Statistic
The results demonstrate that the TECS protocol had high
Age, meanSD 60.011.6 percent agreement with moderate to substantial kappa
Males, n (%) 222 (86.7) values when compared with an FTF examination for the 4
Race-ethnicity, n (%) most common causes of visual loss in the Veteran
White 98 (38.3)
population.
Black 157 (61.3)
Asian 1 (0.4) For the purposes of this analysis, we used the definition
Eye trauma, n (%)* 69 (27.6) of kappa in Landis and Koch12: k ¼ 0.0e0.20 none to slight
Family history of eye diagnoses or blindness, n (%)* 63 (25.2) agreement, k ¼ 0.21e0.40 fair agreement, k ¼ 0.41e0.60
Smoking history, n (%)* moderate agreement, k ¼ 0.61e0.80 substantial
Never 100 (41.7) agreement, and k > 0.80 near perfect agreement. Table 5
Former 71 (29.6) is a summary table that reports the results from the TECS
Current 69 (28.8)
trial alongside other published literature. Values that are
missing indicate the authors did not publish that calculation.
SD ¼ standard deviation.
*Missing: eye trauma (n ¼ 6), family eye history (n ¼ 6), and smoking Cataract
history (n ¼ 16).
There are few studies in the literature that directly compare
photographs with an FTF examination for the diagnosis of
Table 2 indicates the percent agreement, kappa statistics, cataract. Our study results for sensitivity and kappa are
sensitivity, and specificity of the TECS protocol between the 2 consistent with both Gupta et al13 and Conlin et al.14 The
readers and the FTF examination. According to the FTF TECS protocol had better specificity than in the study by
provider, the prevalence of surgical cataracts was 3.9% in our Gupta et al in the diagnosis of cataract.
study population, 26.6% in the glaucoma suspect/glaucoma
population, 2.3% in the AMD population, and 3.1% in the DR Macular Degeneration
population, and the presence of any condition resulting in
referral was 43.8%. By using the TECS protocol, readers Although there was a high percent agreement with the FTF
diagnosed more patients with cataracts (6.3% and 5.9% for examination, the lowest kappa overall in the study for both
reader 1 and reader 2, respectively) and any condition requiring readers 1 and 2 were for AMD. Our results are difficult to
referral (48.1% and 59.0% for reader 1 and reader 2, interpret because there is low prevalence of AMD in our
respectively) compared with the FTF physician and diagnosed specific Veteran population, resulting in a low number of
fewer patients with glaucoma (25.4% and 14.5% for reader 1 AMD cases in the study, causing imprecise estimates of
and reader 2, respectively). Percent agreement between the sensitivity, specificity, and kappa. Nevertheless, TECS re-
diagnostic classifications obtained from FTF visits and the TECS
sults were similar to 3 other studies comparing photographs
protocol ranged from 68.4% to 98.4%, with the lowest level of
agreement observed in the compound variable “any diagnosis with an FTF examination for AMD (Table 5).14-16
resulting in referral” (75.4% and 68.4% for reader 1 and reader
2, respectively). Diagnostic concordance with the FTF visits was Diabetic Retinopathy
higher for reader 1 than for reader 2, with kappa statistics Several studies have compared fundus images for DR
between 0.51 and 0.77 for reader 1 and between 0.34 and 0.71
detection with a retinal examination. The TECS kappa was
for reader 2. Specificity for the TECS protocol was generally
high. Specificity measures for cataracts, glaucoma, macular similar to studies comparing a retinal examination with
degeneration, and DR were between 0.91 and 0.99 for both photographs (Conlin et al14 and Kerr et al17), with TECS
readers, whereas specificity estimates for any diagnosis resulting having a better percent agreement than Cavallerano et al18
in referral were 0.74 and 0.58 for reader 1 and reader 2, and Gomez-Ulla et al.19 One reason for the differences in
respectively. Sensitivity estimates exhibited more variation with the reported data might be study design or DR
values ranging from 0.50 to 1.00 for reader 1 and 0.47 to 0.90 classification scheme. For example, Cavallerano et al18
for reader 2. performed an FTF examination approximately 30 days
Tables 3 and 4 illustrate inter-reader and intra-reader variability, postimaging, and Gomez-Ulla et al19 used a modified
respectively. Inter-reader agreement was highest for cataracts (l ¼ Airlie House classification, whereas TECS uses Early
0.83), followed by glaucoma (l ¼ 0.62), DR (l ¼ 0.61), and AMD
Treatment Diabetic Retinopathy Study classification.
(l ¼ 0.46). The readers differed most often in their categorization
of “any diagnosis resulting in referral” (l ¼ 0.33). Reader 1
diagnosed more patients with glaucoma than reader 2, whereas Glaucoma/Glaucoma Suspect
reader 2 was more likely to diagnose patients with AMD compared The TECS trial was powered for glaucoma and glaucoma
with reader 1. According to the intra-reader agreement calculations, suspect detection. Glaucoma is one of the most difficult
reader 2’s diagnostic classifications were slightly more consistent
disease entities to consistently diagnose because multiple
over time. Kappa statistics for diagnoses made 90 days apart
ranged from 0.59 to 0.87 for reader 2 and 0.39 to 0.70 for reader 1. factors are considered when making the diagnosis. Not
Notably, reader 1 diagnosed 1 patient with AMD at the initial surprisingly then, kappa values for TECS readers were
TECS assessment and zero patients at the 90-day TECS assess- slightly lower for glaucoma (compared with cataract or DR)
ment, so we were unable to calculate the kappa statistic for this but still reflected moderate to substantial agreement with the
category for reader 1. FTF examination. Furthermore, TECS had a higher percent

40
 Maa et al 
Diagnostic Accuracy of the TECS Protocol

Table 2. Prevalence of Ophthalmologic Diagnoses among Study Participants and Agreement, Sensitivity, and Specificity for Diagnoses
Obtained from Face-to-Face Examinations Compared with Those Obtained Using the Technology-based Eye Care Services Protocol (N ¼ 256)

Percent Kappa Sensitivity Specificity

FTF* n (%) TECS n (%) Agreement (95% CI) (95% CI) (95% CI)
Diagnosis FTF Reader 1 Reader 1 Compared with FTF
Cataracts referred for surgery 10 (3.9) 16 (6.3) 97.7 0.77 (0.57e0.94) 1.00 (0.69e1.00) 0.98 (0.95e0.99)
Glaucoma and glaucoma suspect 68 (26.6) 65 (25.4) 86.3 0.65 (0.54e0.75) 0.72 (0.60e0.82) 0.91 (0.87e0.95)
Macular degeneration 6 (2.3) 5 (2.0) 98.1 0.54 (0.18e0.90) 0.50 (0.12e0.88) 0.99 (0.97e1.00)
DR (any) 8 (3.1) 8 (3.1) 98.4 0.74 (0.50e0.99) 0.75 (0.35e0.97) 0.99 (0.97e1.00)
Any diagnosis resulting in referral 112 (43.8) 123 (48.1) 75.4 0.51 (0.40e0.61) 0.77 (0.68e0.84) 0.74 (0.66e0.81)

FTF Reader 2 Reader 2 Compared with FTF

Cataracts referred for surgery 10 (3.9) 15 (5.9) 97.3 0.71 (0.50e0.91) 0.90 (0.56e1.00) 0.98 (0.95e0.99)
Glaucoma and glaucoma suspect 68 (26.6) 37 (14.5) 84.0 0.52 (0.40e0.64) 0.47 (0.35e0.60) 0.97 (0.94e0.99)
Macular degeneration 6 (2.3) 16 (6.3) 94.5 0.34 (0.08e0.60) 0.67 (0.22e0.96) 0.95 (0.92e0.98)
DR (any) 8 (3.1) 8 (3.1) 97.7 0.61 (0.33e0.90) 0.63 (0.25e0.92) 0.99 (0.97e1.00)
Any diagnosis resulting in referral 112 (43.8) 151 (59.0) 68.4 0.38 (0.27e0.49) 0.81 (0.73e0.88) 0.58 (0.50e0.66)

CI ¼ confidence interval; DR ¼ diabetic retinopathy; FTF ¼ face-to-face; TECS ¼ Technology-based Eye Care Services.
*A single FTF examination was done with TECS reader 1 and TECS reader 2 being compared with the single FTF examination.

agreement than Gupta et al,13 kappa was similar to 3 other period. Kappa statistics were in the substantial to near-
studies, and reader 1’s estimates were comparable to the perfect range (0.70e0.87), and percent agreements were
large meta-analysis by Thomas et al20 with regard to tele- 89% to 99%.
glaucoma sensitivity and specificity.
Overall Assessment of Technology-based Eye
Intraobserver and Interobserver Variability of Care Services
Technology-based Eye Care Services
Overall, TECS has good sensitivity and excellent specificity
The data show that the TECS protocol allowed for sub- when compared with an FTF eye examination. Given that
stantial to near-perfect agreement between readers 1 and the trial was powered for glaucoma/glaucoma suspect,
2, with k of 0.61 (DR and glaucoma) to 0.83 (cataract). readers were 47% to 72% sensitive when compared with the
The only value that was slightly lower was AMD at 0.46, FTF provider in detecting cases of glaucoma/glaucoma
and the k is less reliable because of the low number of suspect. These glaucoma detection percentages make TECS
cases. In addition, the percent agreement was high, useful as a screening tool because it allows for up to three-
ranging from 87% to 98% between the readers. Most quarters of asymptomatic patients to be identified and is
important, interobserver agreement for glaucoma/glau- used in a population who might not otherwise receive care
coma suspect was substantial (0.62) and percent agree- and therefore go undiagnosed.
ment was high (>80%). These results are consistent with The high specificity of TECS indicates that when the
previously published literature for glaucoma suspect/ readers do not find a problem, there is a high chance of the
glaucoma (0.50e0.68)21-24; TECS was even on par with patient being truly free of abnormalities. Limitations in
inter-reader data obtained between glaucoma sensitivity, however, suggest that patients should still
specialists.22 receive FTF examinations at some interval, supporting the
Intra-reader variability was minimal because both readers TECS protocol that does not permit patients to continue
1 and 2 had substantial to near-perfect agreement when they telemedicine screening indefinitely. These data also
reviewed the same information after the 90-day wash-out emphasize the importance of ensuring screened patients

Table 3. Inter-reader Agreement between Reader 1 and Reader 2 Using the Technology-based Eye Care Services Protocol (N ¼ 256)

Diagnosis Reader 1 n (%) Reader 2 n (%) Percent Agreement Kappa (95% CI)
Cataracts referred for surgery 16 (6.3) 15 (5.9) 98.1 0.83 (0.68e0.98)
Glaucoma and glaucoma suspect 65 (25.4) 37 (14.5) 87.5 0.62 (0.50e0.73)
Macular degeneration 5 (2.0) 16 (6.3) 95.7 0.46 (0.20e0.72)
DR 8 (3.1) 8 (3.1) 97.7 0.61 (0.33e0.90)
Any diagnosis resulting in referral 123 (48.1) 151 (59.0) 66.4 0.33 (0.22e0.45)

CI ¼ confidence interval; DR ¼ diabetic retinopathy.

41
 Ophthalmology Volume 127, Number 1, January 2020

Table 4. Intra-reader Agreement of Diagnoses Obtained 90 Days Apart Using the Technology-based Eye Care Services Protocol (N ¼ 150)

Diagnosis Day 0 TECS n (%) Day 90 TECS n (%) Percent Agreement Kappa (95% CI)
Reader 1
Cataracts referred for surgery 9 (6.0) 5 (3.3) 97.3 0.70 (0.43e0.98)
Glaucoma and glaucoma suspect 40 (26.7) 28 (18.7) 89.3 0.70 (0.56e0.83)
Macular degeneration 1 (0.7) 0 (0.0) 99.3 *
DR 4 (2.7) 3 (2.0) 98.0 0.56 (0.12e1.00)
Any diagnosis resulting in referral 71 (47.3) 58 (38.7) 70.0 0.39 (0.25e0.54)
Reader 2
Cataracts referred for surgery 8 (5.3) 8 (5.3) 98.7 0.87 (0.69e1.00)
Glaucoma and glaucoma suspect 21 (14.0) 34 (22.7) 90.0 0.67 (0.52e0.82)
Macular degeneration 6 (4.0) 4 (2.7) 97.3 0.59 (0.22e0.95)
DR 3 (2.0) 3 (2.0) 98.7 0.66 (0.22e1.00)
Any diagnosis resulting in referral 84 (56.0) 89 (59.3) 84.7 0.69 (0.57e0.80)

CI ¼ confidence interval; DR ¼ diabetic retinopathy; TECS ¼ Technology-based Eye Care Services.

*Kappa statistic not calculated because of zero cells.

receive follow-up care and stress the importance of an Eye reads over time. Finally, the TECS data show similar kappa
Clinic using telemedicine to appropriately plan resources to values, percent agreements, sensitivity, and specificity as
accommodate follow-up patients.25 Moreover, the high other published trials, such as Sperduto et al26 and Conlin
kappa and percent agreement for inter-reader and intra- et al,14 confirming their findings and conclusions that a
reader variability support the premise that the TECS “Technology Assisted Exam” such as TECS is comparable
protocol promotes equal quality of care across sites, to, but not a substitute for, an FTF examination for the
concordance between different readers, and consistency of detection of cataract, glaucoma, DR, and AMD.

Table 5. Comparison of Technology-based Eye Care Services Protocol with Other Telehealth Studies

Diagnosis Percent Agreement Kappa (95% CI) Sensitivity (95% CI) Specificity (95% CI)
Cataract
TECS (reader 1 and reader 2) 97.7 0.77 (0.57e0.94) 1.00 (0.69e1.00) 0.98 (0.95e0.99)
97.3 0.71 (0.50e0.91) 0.90 (0.56e1.00) 0.98 (0.95e0.99)
Gupta et al13 93.0 0.68 0.98 (0.89e0.99) 0.63 (0.26e0.90)
Conlin et al14 99.0 0.71
Macular Degeneration
TECS (reader 1 and reader 2) 98.1 0.54 (0.18e0.90) 0.50 (0.12e0.88) 0.99 (0.97e1.00)
94.5 0.34 (0.08e0.60) 0.67 (0.22e0.96) 0.95 (0.92e0.98)
15
Pirbhai et al 80.0 0.59 (0.49e0.70) 0.82 (0.72e0.90) 0.79 (0.71e0.86)
Duchin et al16 0.84 0.94
Conlin et al14 97.0 0.59 0.67 (0.31e0.91) 0.98 (0.96e0.99)
Diabetic Retinopathy
TECS (reader 1 and reader 2) 98.4 0.74 (0.50e0.99) 0.75 (0.35e0.97) 0.99 (0.97e1.00)
97.7 0.61 (0.33e0.90) 0.63 (0.25e0.92) 0.99 (0.97e1.00)
18
Cavallerano et al 89.3
Gomez-Ulla et al19 94.0 0.92 (0.90e0.95)
Kerr et al17 82.0e94.0 0.64
Conlin et al14 97.0 0.68 0.75 (0.42e0.93) 0.98 (0.96e0.99)
Glaucoma and Glaucoma Suspect
TECS (reader 1 and reader 2) 86.3 0.65 (0.54e0.75) 0.72 (0.60e0.82) 0.47 (0.35e0.60) 0.91 (0.87e0.95)
84.3 0.52 (0.40e0.64) 0.97 (0.94e0.99)
Thomas et al20 0.83 0.79
Conlin et al14 94.0 0.80 0.83 (0.71e0.91) 0.96 (0.92e0.98)
Gupta et al13 67.0 0.52 0.72 (0.57e0.83) 0.81 (0.47e0.97)
Any Disease
TECS (reader 1 and reader 2) 75.0 0.51 (0.40e0.61) 0.77 (0.68e0.84) 0.74 (0.66e0.81)
6.4 0.38 (0.27e0.49) 0.81 (0.73e0.88) 0.58 (0.50e0.66)
Sperduto et al26 71.0 0.61 (0.43e0.78)
Conlin et al14 84.0 0.67 0.86 (0.77e0.92) 0.84 (0.78e0.88)

CI ¼ confidence interval; TECS ¼ Technology-based Eye Care Services.

42
 Maa et al 
Diagnostic Accuracy of the TECS Protocol

Study Limitations Acknowledgments

There were several limitations to our study. The sample This project would not have been possible without the support of
size, although adequately powered for glaucoma suspect/ the Atlanta VA Ophthalmology Chief, Dr. Steven Urken, the
glaucoma, did not have a high enough number of cases of Atlanta VA Eye Clinic staff, the reading physicians, the medical
the other disease entities such as AMD. This may help students, and the ophthalmic technicians. A special thanks to the
project’s Research Coordinator, Deirdre Dixon, whose hard work
explain why, despite a high percent agreement, the kappa and commitment to the project were truly invaluable to the study’s
values were lower and sensitivity and specificity are more success.
difficult to calculate reliably. In addition, the Veteran
population is different from the greater US population,4
possibly limiting generalizability. Recruitment strategies References
(patients self-volunteered for the study) may have intro-
duced selection bias. The potential to receive free addi-
tional imaging studies may have prompted sicker patients 1. Nathoo N, Ng M, Rudnisky CJ, Tennant MT. The prevalence
of diabetic retinopathy as identified by teleophthalmology in
to volunteer at higher rates compared with healthy coun-
rural Alberta. Can J Ophthalmol. 2010;45:28e32.
terparts. Finally, the study was based on the presumption 2. Ng M, Nathoo N, Rudnisky CJ, Tennant MT. Improving ac-
that the FTF examination is 100% accurate and 100% cess to eye care: teleophthalmology in Alberta, Canada.
consistent and represents a standardized modality for the J Diabetes Sci Technol. 2009;3:289e296.
diagnosis of all diseases of interest. Having only 1 FTF 3. Boucher MC, Desroches G, Garcia-Salinas R, et al. Tele-
examiner may have introduced bias related to individual ophthalmology screening for diabetic retinopathy through
practice patterns and skill level. Calculations might change mobile imaging units within Canada. Can J Ophthalmol.
if differences between the 2 readers and FTF physician 2008;43:658e668.
were adjudicated to arrive at the “truth,” and both the FTF 4. Ong HS, Levin S, Vafidis G. Glaucoma detection using
examiner and reader were compared with the “truth.” Re- optic disc images from the English national screening
programme for diabetic retinopathy. J Glaucoma. 2013;22:
sults may also change if both TECS and the FTF examiner
496e500.
are compared with the patient’s actual clinical outcome. 5. Paul PG, Raman R, Rani PK, et al. Patient satisfaction levels
Specifically, for glaucoma/glaucoma suspect, the trial data during teleophthalmology consultation in rural South India.
compare the initial TECS examination with the initial FTF Telemed J E Health. 2006;12:571e578.
examination, but the FTF examination may eventually 6. Rosengren D, Blackwell N, Kelly G, et al. The use of tele-
reveal false-positives (overcalls) when patients are found medicine to treat ophthalmological emergencies in rural
not to have glaucoma (physiologic cupping) after ancillary Australia. J Telemed Telecare. 1998;4(Suppl 1):97e99.
testing is completed. 7. Knoblauch H. Focused ethnography. Qual Soc Res. 2005;6:44.
Future studies can address some of these issues. 8. Maa AY, Evans C, DeLaune WR, et al. A novel tele-eye
Adding multiple FTF examiners and adjudicating their protocol for ocular disease detection and access to eye care
services. Telemed J E Health. 2014;20:318e323.
diagnoses may reduce variation and help form a more
9. Maa AY, Patel S, Chasan JE, et al. Retrospective evaluation of
reliable gold standard. Having the study data read by a teleretinal screening program in detecting multiple nondia-
more readers, including a glaucoma or retina specialist, betic eye diseases. Telemed J E Health. 2017;23:41e48.
may change the kappa or sensitivity/specificity, espe- 10. Cavallerano AA, Conlin PR. Teleretinal imaging to screen for
cially for the glaucoma or AMD/DR diagnostic group. diabetic retinopathy in the Veterans Health Administration.
Finally, comparing TECS and FTF with the long-term J Diabetes Sci Technol. 2008;2:33e39.
clinical outcome may allow for better assessment of the 11. Harris PA, Taylor R, Thielke R, et al. Research electronic data
performance of TECS for the diagnosis of glaucoma/ capture (REDCap)–a metadata-driven methodology and
glaucoma suspect. workflow process for providing translational research infor-
In conclusion, Part I of the TECS Compare trial matics support. J Biomed Inform. 2009;42:377e381.
12. Landis JR, Koch GG. The measurement of observer agreement
demonstrated high percent agreements, substantial kappa
for categorical data. Biometrics. 1977;33:159e174.
agreement, and sensitivity and specificity equal or 13. Gupta SC, Sinha SK, Dagar AB. Evaluation of the effective-
potentially better than previously published literature for ness of diagnostic & management decision by tele-
the detection of common ocular disease. The inclusion of ophthalmology using indigenous equipment in comparison
additional, sophisticated ophthalmic testing such as OCT, with in-clinic assessment of patients. Indian J Med Res.
visual fields, or contrast sensitivity may improve diag- 2013;138:531e535.
nostic agreement and sensitivity, especially for AMD or 14. Conlin PR, Asefzadeh B, Pasquale LR, et al. Accuracy of a
glaucoma/glaucoma suspect, and will be analyzed in part technology-assisted eye exam in evaluation of referable dia-
II of this trial. The current TECS protocol is accurate betic retinopathy and concomitant ocular diseases. Br J Oph-
when compared with an FTF examination, especially thalmol. 2015;99:1622e1627.
15. Pirbhai A, Sheidow T, Hooper P. Prospective evaluation of
with regard to glaucoma/glaucoma suspect, and allows
digital non-stereo color fundus photography as a screening tool
for correct identification of abnormal patients with high in age-related macular degeneration. Am J Ophthalmol.
precision and reliability. TECS can serve as a beneficial 2005;139:455e461.
tool to help address the growing need for accessible eye 16. Duchin KS, Asefzadeh B, Poulaki V, et al. Teleretinal imaging
care in the VA healthcare system and potentially in the for detection of referable macular degeneration. Optom Vis Sci.
private sector. 2015;92:714e718.

43
 Ophthalmology Volume 127, Number 1, January 2020

17. Kerr D, Cavan DA, Jennings B, et al. Beyond retinal screening: 22. Varma R, Steinmann WC, Scott IU. Expert agreement in
digital imaging in the assessment and follow-up of patients with evaluating the optic disc for glaucoma. Ophthalmology.
diabetic retinopathy. Diabet Med. 1998;15:878e882. 1992;99:215e221.
18. Cavallerano AA, Cavallerano JD, Katalinic P, et al. Use of 23. Nicolela MT, Drance SM, Broadway DC, et al. Agreement
Joslin Vision Network digital-video nonmydriatic retinal im- among clinicians in the recognition of patterns of optic disk
aging to assess diabetic retinopathy in a clinical program. damage in glaucoma. Am J Ophthalmol. 2001;132:
Retina. 2003;23:215e223. 836e844.
19. Gomez-Ulla F, Fernandez MI, Gonzalez F, et al. Digital retinal 24. Abrams LS, Scott IU, Spaeth GL, et al. Agreement among
images and teleophthalmology for detecting and grading dia- optometrists, ophthalmologists, and residents in evaluating
betic retinopathy. Diabetes Care. 2002;25:1384e1389. the optic disc for glaucoma. Ophthalmology. 1994;101:
20. Thomas SM, Jeyaraman MM, Hodge WG, et al. The effec- 1662e1667.
tiveness of teleglaucoma versus in-patient examination for 25. Chasan JE, Delaune B, Maa AY, Lynch MG. Effect of a tel-
glaucoma screening: a systematic review and meta-analysis. eretinal screening program on eye care use and resources.
PloS One. 2014;9:e113779. JAMA Ophthalmol. 2014;132:1045e1051.
21. Breusegem C, Fieuws S, Stalmans I, Zeyen T. Agreement and 26. Sperduto RD, Hiller R, Podgor MJ, et al. Comparability of
accuracy of non-expert ophthalmologists in assessing glau- ophthalmic diagnoses by clinical and Reading Center exam-
comatous changes in serial stereo optic disc photographs. iners in the Visual Acuity Impairment Survey Pilot Study. Am
Ophthalmology. 2011;118:742e746. J Epidemiol. 1986;124:994e1003.

Footnotes and Financial Disclosures

Originally received: May 2, 2019. HUMAN SUBJECTS: Human subjects were included in this study. The
Final revision: July 1, 2019. human ethics committees at the Emory University and the VA Research and
Accepted: July 26, 2019. Development Committee approved the study. All research adhered to the
Available online: August 13, 2019. Manuscript no. 2019-471. tenets of the Declaration of Helsinki. All participants provided informed
1
TECS Division, Regional Telehealth Services, Atlanta Veterans Affairs consent.
Health Care System, Atlanta, Georgia. No animal subjects were used in this study.
2
Emory University School of Medicine, Atlanta, Georgia. Author Contributions:
3 Conception and design: Maa, Hunt, Lynch
Bascom Palmer Eye Institute, University of Miami, Miami, Florida.
Data collection: Maa, Medert, Lu, Janjua, McCord
4
Charleston Health Equity and Rural Outreach Innovation Center (HE-
Analysis and interpretation: Maa, Howell, Hunt, Lynch
ROIC), Ralph A. Johnson Department of Veterans Affairs Medical Center,
Obtained funding: Maa, Lynch
Charleston, South Carolina.
5 Overall responsibility: Maa, Lu, Janjua, Howell, Hunt, Giangiacomo,
New York Eye and Ear Infirmary, New York, New York.
Lynch
6
Ophthalmology Division, Surgical Services, Atlanta Veterans Affairs Abbreviations and Acronyms:
Health Care System, Atlanta, Georgia. AMD ¼ age-related macular degeneration; DR ¼ diabetic retinopathy;
Financial Disclosure(s): FTF ¼ face-to-face; NCC ¼ New Comprehensive Clinic;
The author(s) have no proprietary or commercial interest in any materials TECS ¼ Technology-based Eye Care Services; TRI ¼ teleretinal imaging;
discussed in this article. VA ¼ Veterans Affairs.
Partially funded by an Atlanta Clinical and Translational Institute Trans- Correspondence:
lational Research grant. The funding source had no role in the design or April Y. Maa, MD, 1670 Clairmont Road MC 112E, Decatur, GA 30033.
conduct of this study. E-mail: [email protected].

44
Methods for Intraocular Lens Power
Calculation in Cataract Surgery after Radial
Keratotomy
Andrew M.J. Turnbull, BM, FRCOphth,1,2,3 Geoffrey J. Crawford, MD, FRANZCO,2,4
Graham D. Barrett, MD, FRANZCO2,4

Purpose: To compare methods of calculating the required intraocular lens (IOL) power for patients under-
going cataract surgery after radial keratotomy (RK), including the 2016 update of the True K formula.
Design: Retrospective case series.
Participants: A total of 52 eyes of 34 patients who had sequential RK and cataract surgery performed in the
same institution by 1 of 2 surgeons.
Methods: Seven IOL calculation formulae were evaluated: True K [History], True K [Partial History], True K
[No History], Double-K Holladay 1 (DK-Holladay-IOLM), Potvin-Hill, Haigis, and Haigis with a
0.50 diopter (D)
offset. Biometry was obtained with the IOLMaster 500 (Carl Zeiss Meditec AG, Jena, Germany) and Pentacam
(OCULUS Inc, Arlington, WA) devices. Subjective refraction was performed at 4 to 6 weeks postoperatively. The
achieved spherical equivalent outcome was compared with the target outcome to calculate the absolute error for
each eye with each formula.
Main Outcome Measures: Median absolute error (MedAE) and mean absolute error (MAE), and percentage
of patients within 0.50 D, 0.75 D, and 1.00 D of refractive target. Mean error (ME) was also calculated to
demonstrate whether a formula tended toward more myopic or hyperopic outcomes.
Results: Best results were achieved with the True K [History]. The MedAE was higher (0.382 vs. 0.275) with
the True K [Partial History], but a similar percentage of patients (75.0%e76.6%) achieved within 0.50 D of target.
Of the methods that do not require refractive history, the True K [No History] and unadjusted Haigis were most
accurate (69.2% within 0.50 D of target), with the True K [No History] returning the lowest MedAE but also more
of a tendency toward hyperopia (ME þ0.269 vs.
0.006 for Haigis). The DK-Holladay-IOLM and Potvin-Hill
methods were the least accurate.
Conclusions: Knowledge of the refractive history significantly improves the accuracy of IOL calculations in
patients undergoing cataract surgery after previous RK. The post-RK refraction appears to be the most important
parameter, with inclusion of the pre-RK refraction offering a further slight improvement in MedAE. When no
refractive history is available, the True K [No History] and Haigis formulae both perform well, with the added
advantage of not requiring data from separate biometric devices. Ophthalmology 2020;127:45-51 ª 2019 by the
American Academy of Ophthalmology

Surgeons have long been aware of the difficulties that radial postoperative corneal instability, early hyperopic shift, and
keratotomy (RK) poses for intraocular lens (IOL) calcula- subsequent myopic regression further compound the
tions when patients later present for cataract surgery. First, difficulties in achieving an accurate refractive result with
RK alters the normal corneal curvature and cataract surgery.
anterioreposterior relationship in a different manner than This study evaluates the outcomes of cataract surgery in
that caused by laser refractive surgery.1,2 Second, the optical post-RK eyes and compares the efficacy of several different
zone is often less than 3.0 mm,3 meaning that standard formulae and methods, including 3 options for the recently
keratometry measures the region representing the updated (2016) RK algorithm of the Barrett True K formula.
inflection point between the incised cornea and the
indirectly flattened central zone. This leads to
overestimation of corneal power and consequently a risk Methods
of hyperopic refractive error after cataract surgery when
Institutional Review Board approval for chart review and IOL
standard formulae are used. Furthermore, post-RK corneas formulae comparison was obtained, and the study was conducted
are frequently irregular with multiple steep and flat areas, according to the tenets of the Declaration of Helsinki. Individual
making determination of the net steep and flat meridians patient consent was not required because this was an anonymized
challenging.1 Diurnal variations in refraction can also database analysis. Consecutive patients with a history of RK pre-
occur.4 Intraoperative opening of the radial incisions and senting to the Lions Eye Institute, Perth, Australia, for cataract

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.08.019 45

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

surgery between 2005 and 2018 were retrospectively analyzed. All Instead of IOLMaster or LENSTAR keratometry, the Potvin-
RK surgeries were performed by a single surgeon (G.J.C.), and Hill formula uses the mean sagittal front keratometry value at a
most patients had complete pre- and post-RK refractive data 4.0-mm zone (PWR_SF_40) from the Power Distribution display
available. of the Pentacam in the DK-Holladay 1 formula.9 The thinnest local
Biometry before cataract surgery was measured with the IOL- pachymetry measurement (CT_MIN), is obtained from the
Master 500 (Carl Zeiss Meditec AG, Jena, Germany). Corneal Pachymetric display of the Pentacam and used as a further
tomography was obtained using the Pentacam device (OCULUS independent variable in the following formula9:
Inc, Arlington, WA). Cataract surgery was performed by 1 of 2 IOL Power (Potvin-Hill Method) ¼ Holladay 1 IOL Calcula-
experienced surgeons (GJC or GDB), using a temporal clear tion (PWR_SF_40)
4.4554 þ 0.0084  CT_MIN
corneal incision, phacoemulsification, and implantation within the The Haigis formula does not use keratometry values to predict
capsular bag of an Acrysof SN60WF or SN6ATx IOL (Alcon, Fort ELP,1 and this may have advantages when dealing with atypical
Worth, TX). No intraoperative or postoperative complications corneas. In 2015, Geggel4 reported the successful use of the
occurred, and all cases were included in the analysis. Subjective Haigis formula and IOLMaster keratometry in post-RK eyes by
refraction was obtained by a highly experienced optometrist setting the target for
1.00 when a final outcome of
0.50 was
between 4 and 6 weeks postoperatively. desired. In this study, emmetropia was the intended outcome, so
The IOL power calculations were performed retrospectively the Haigis formula was used with a “target” of
0.50 spherical
using 7 methods, 4 of which are available on the American Society equivalent (“Haigis [
0.50 offset]”). The Haigis formula was also
of Cataract and Refractive Surgeons (ASCRS) online RK calcu- evaluated without any offset, simply targeting emmetropia.
lator:5 the Barrett True K (RK algorithm: “history” and “no To calculate the refractive prediction error (PE) with each for-
history” options), the double-K modified Holladay 1 using IOL- mula for each eye, the predicted refraction based on the implanted
Master keratometry (DK-Holladay-IOLM), and the Potvin-Hill IOL power was subtracted from the actual postoperative refraction.
formula. The DK-Holladay-IOLM and Potvin-Hill formulae are Following this, the mean absolute error (MAE) and median abso-
referred to as the “IOLMaster” and “Pentacam” formulae, respec- lute error (MedAE) in refraction, as well as the percentage of eyes
tively, on the ASCRS RK calculator.5 The standard Haigis formula within 0.50 D, 0.75 D, and 1.00 D of target, were calculated
was also evaluated, both with and without a
0.50 diopter (D) for each formula.
offset as suggested in a previous study.4 To perform the analysis, In IOL power calculation analyses of normal eyes, mean error
it was necessary to embed the formulae in a Microsoft Excel (ME) should be reduced to 0 via a process of IOL constant optimi-
spreadsheet (Microsoft Corp, Redmond, WA). A random sample zation for the dataset. However, given the high heterogeneity of
of 5 cases was cross-checked against the ASCRS online RK unusual subsets such as post-RK or post-LASIK eyes in clinical
calculator to ensure consistency of output with the spreadsheet. practice, it is considered inappropriate to derive optimized IOL
The True K is a theoretical formula for postrefractive surgery constants for these special cases because they are unlikely to be
eyes derived from the Barrett Universal II formula.6,7 The Barrett accurate or applicable to future cases. A more appropriate solution is
Universal II uses a complex algorithm to predict the effective lens for the formulae designed for these cases to inherently and reliably
position (ELP) and IOL power using up to 5 biometric parameters allow for this variability, without relying on artificially modified IOL
(axial length [AL], keratometry, anterior chamber depth, lens constants. Thus, most surgeons use IOL constants previously opti-
thickness, and horizontal white-to-white distance), and continues to mized for normal eyes.10 Accordingly, the True K, DK-Holladay-
evolve via data-driven enhancements. The True K incorporates a IOLM, and Haigis formulae in this study included standard IOL
modified double-K method and can be applied to eyes that have constants optimized for normal eyes, and a further process of dataset
undergone RK or myopic or hyperopic laser refractive surgery IOL constant optimization was not performed. This is consistent
(with different algorithms for each), both with and without with previous work in this field.10 The ME indicates the direction of
knowledge of the prerefractive surgery metrics. The RK algorithm refractive bias for each formula, either myopic (
) or hyperopic (þ).
of the True K was modified in 2016 to include a partial history Statistical analysis was performed using Microsoft Excel
option, which can use the refraction after RK even when the pre- (Microsoft Corp, Redmond, WA). The nonparametric 2-tailed
RK refractive data are unknown. To take advantage of this func- Wilcoxon signed-rank test was used to compare the absolute er-
tion, the post-RK refraction should be entered and the box for the rors obtained by different methods. The Bonferroni correction was
pre-RK refraction should be left blank. This and other updates were used to control for multiple comparisons and reduce the risk of a
subsequently applied to the ASCRS RK calculator and made type 1 statistical error (family-wise error), with an adjusted P value
available via the Asia Pacific Association of Cataract and Refrac- of < 0.004 being considered statistically significant.
tive Surgeons website.8 A formula guide to assist users is available
via the Asia Pacific Association of Cataract and Refractive
Surgeons website when accessing the True K calculator.8 Results
In this study, the True K was investigated using the full
refractive history (True K [History]), without any refractive history A total of 52 eyes of 34 patients were included. Further de-
(True K [No History]) or using just the post-RK refraction (True K mographic details are shown in Table 1. Complete refractive history
[Partial History]). For the latter, the most hyperopic refraction was available for 47 eyes, with 4 eyes having no refractive history
recorded for the patient was used. This was to allow for the pro- and 1 eye having only the post-RK refraction available.
gressive hyperopic drift that occurs after RK, but also to control for Detailed refractive outcomes for each formula are shown in
any myopic shift caused by nuclear sclerosis. The patients included Table 2. The lowest MedAE and the highest percentage of eyes
in this study were not used in the original derivation of the True K achieving their refractive target were achieved by the True K
formula or the 2016 update. [History]. There was a small hyperopic bias in the ME
The DK-Holladay 1 formula uses a standard K value of 43.86 D of þ0.125 (95% confidence interval [CI], þ0.008 to þ0.242).
based on a population mean to predict the ELP, and mean kera- Similar results were obtained with the True K [Partial History],
tometry and AL measured by the IOLMaster or LENSTAR (Haag but this time with a small myopic ME (
0.083; 95% CI,
0.209
Streit, Koniz, Switzerland) to determine the required IOL power.5 to þ0.043). With the True K [History] and True K [Partial
Only IOLMaster keratometry was available for all eyes in this History], 76.6% and 75.0% patients were within 0.50 D and
study (“DK-Holladay-IOLM”). 97.9% and 95.8% were within 1.00 D of target, respectively.

46
 Turnbull et al 
IOL Calculations after RK

Table 1. Demographic Details

Gender 21 Female and 13 Male

26 Left Eyes and 26 Right Eyes
Laterality Mean Standard Deviation Range
Age at time of cataract surgery (yrs) 65 8 48e80
No. of RK incisions 8 3 4e16
AL (mm) 24.98 0.87 22.79e26.87
Anterior chamber depth (mm) 3.42 0.38 2.63e4.53
Mean K (IOLMaster, Carl Zeiss Meditec AG, Jena, Germany) (D) 38.38 2.70 32.09e44.23
PWR_SF_40 (Pentacam, OCULUS Inc, Arlington, WA) (D) 38.9 2.1 35.3e43.9
CT_min (Pentacam) (mm) 543 43 410e624
IOL power implanted (D) 24.7 2.8 19e30

AL ¼ axial length; D ¼ diopters; IOL ¼ intraocular lens; RK ¼ radial keratotomy.

The Haigis formula targeting emmetropia resulted in the lowest post-RK led to poor results. The Clinical History Method
ME and outperformed the Haigis [
0.50 offset] method. The DK- was confounded by the progressive hyperopia that occurs
Holladay-IOLM produced similar outcomes to the Haigis [
0.50 after RK and nuclear sclerosis.4 Lyle and Jin11 suggested
offset], but with a tendency toward myopic errors (ME,
0.355; subtracting 1.00 D from the measured keratometry, but
95% CI,
0.531 to
0.179). The Potvin-Hill had the highest MAE
this led to a wide range of hyperopic PE from þ0.50 D
and MedAE, with a large range of PE from
2.631 to þ4.611.
This resulted in the lowest percentage of eyes achieving their to þ3.00 D.3
refractive target of all methods studied. Although there have been some reports of the efficacy of
The True K [History] resulted in a significantly lower MedAE intraoperative aberrometry in post-RK cataract surgery,
than the DK-Holladay-IOLM (P ¼ 0.005), the Potvin-Hill (P < others have demonstrated its unreliability.12 In post-RK
0.0001), and the Haigis [
0.50 offset] (P ¼ 0.024). Although the eyes, the Optiwave Response Analyzer has been shown to
True K [History] had a lower MedAE than both the other True K be prone to substantial hyperopic errors due to intra-
options and the standard Haigis formula, the differences were not operative fluctuations in corneal curvature, anterior chamber
statistically significant (True K [Partial History], P ¼ 0.703; True depth, and AL, the parameters on which Optiwave Response
K [No History], P ¼ 0.619; Haigis, P ¼ 0.172). However, this may Analyzer predictions are based, even at the recommended
represent a type II error due to the study having insufficient power
IOP of 21 mmHg.13 Because sophisticated technology
to detect a statistical difference.
Likewise, the True K [Partial History] method had significantly cannot be relied upon in this context, it is vital that
lower MedAE than the DK-Holladay-IOLM (P <0.0001), Potvin- accurate IOL calculation formulae are available and used,
Hill (P < 0.0001), or Haigis [
0.50 offset] (P ¼ 0.002). The True and continue to be improved.
K [Partial History] had a lower MedAE compared with the Haigis The “double K” method was proposed by Aramberri14 to
formula, but the difference did not reach statistical significance tackle the problem of postrefractive IOL calculations,
(P ¼ 0.07). whereby the prerefractive surgery keratometry is used to
With regard to the “no history” methods, the True K [No predict the ELP and the postrefractive surgery keratometry
History] had significantly lower MedAE than the DK-Holladay- is used to determine the corneal power to be used in a
IOLM and Potvin-Hill (P ¼ 0.024 and P < 0.0001, respec- vergence formula such as SRK/T or Holladay 1. The
tively). The difference between the True K [No History] and the
double-K principle has subsequently helped to develop
Haigis [
0.50 offset] (P ¼ 0.041) approached but did not reach the
Bonferroni adjusted level of statistical significance. Although there several postrefractive IOL calculation formulae, including
was no statistically significant difference between the True K [No some that can be used for post-RK eyes.5 The optimum
History] and the standard Haigis formula (P ¼ 0.655), the True K source of post-RK keratometry data is yet to be
[No History] returned a lower MedAE and standard deviation of established, with authors having investigated a variety of
the ME. The range of PE was less with the Haigis secondary to the biometric and topographic devices.4,5,9,15-17
presence of a single outlier with the True K [No History]: a patient Double-K formulae for use post-RK include the True K
with an AL of 27 mm, a mean keratometry of 32 D, and a history (RK algorithm), DK-Holladay-IOLM, and Potvin-Hill,
of 8-cut RK. Without this outlier, the range of PE would have been which are evaluated in the current study. Other methods
similar for both methods. The True K [No History] and Haigis currently available on the ASCRS RK calculator include the
formulae achieved similar percentages of eyes within each refrac-
Effective Refractive Power,15 Average Central Corneal
tive bracket. The ME with the Haigis was close to 0, whereas the
True K [No History] tended toward hyperopic outcomes Power,16 and Atlas 1-4,5 all of which use the DK-
(ME þ0.269; 95% CI, þ0.140 to þ0.398). Holladay 1 formula with different sources of keratometry,
and the OCT-based method that requires corneal parameters
measured on the RTVue device (Optovue Inc, Fremont, CA)
Discussion as well as AL and anterior chamber depth from the IOL-
Master.17 The EyeSys Effective Refractive Power, Atlas 1-
The IOL calculation in eyes with previous RK has proved a 4, and OCT-based methods were not evaluated in this study
considerable challenge. Initial attempts at IOL calculation because the relevant imaging devices were not available.

47
 Ophthalmology Volume 127, Number 1, January 2020

The EyeSys Effective Refractive Power was shown to

% ±1.00 D
have good efficacy when originally reported, with 80% of

97.9
95.8
92.3
92.3
84.6
82.7
69.2
eyes within 0.50 D of target, but this included eyes that
had undergone RK, hexagonal keratotomy, or laser thermal
keratoplasty.15 Only 14 post-RK eyes were studied, with a
hyperopic ME of þ0.27 0.51 D. This method has not
% ±0.75 D been widely adopted or reported, partly because of the
95.7
93.8
84.6
86.5
73.1
71.2
53.8
limited uptake of the EyeSys tomography device.
The Average Central Corneal Power method produced
promising results in its initial description using the Topo-
% ±0.50 D

graphic Modeling System device (Tomey Corp, Phoenix,

AZ), with 87.5% within 0.50 D of target.16 In a
Table 2. Refractive Accuracy for Each Method of Intraocular Lens Calculation after Previous Radial Keratotomy

76.6
75.0
69.2
69.2
50.0
46.2
40.4
subsequent small study of 15 eyes by DeMill et al,18 the
Average Central Corneal Power using Pentacam data
resulted in MAE of 1.030.92 and a hyperopic ME
of þ0.541.28 (range,
1.54 to þ2.72). This was
MedAE

CI ¼ confidence interval; D ¼ diopters; MAE ¼ mean absolute error; ME ¼ mean error; MedAE ¼ median absolute error; SD ¼ standard deviation.
0.275
0.382
0.330
0.406
0.504
0.569
0.632

slightly superior to using the ASCRS average output

(ME 0.811.09; range,
0.88 to þ2.79; MAE
1.020.87), and both were significantly better than the
Atlas 1-4 method (ME 1.071.30; range,
1.82
(0.303e0.435)
(0.302e0.446)
(0.323e0.513)
(0.346e0.504)
(0.477e0.715)
(0.452e0.692)
(0.664e1.130)
MAE (95% CI)

to þ2.85; MAE 1.530.64).18 DeMill et al18 reported

60%, 47%, and 0% of eyes within 0.50 D of target
using the Average Central Corneal Power, ASCRS
Average, and Atlas 1-4 methods, respectively.
0.369
0.374
0.418
0.425
0.596
0.572
0.897

This study reports the first published outcomes of the

updated RK algorithm of the True K formula (version 2.0,
2016). In 2016, Ma et al10 reported a comparison of the
Range of Prediction Error

True K [No History] with the DK-Holladay-IOLM, DK-

Holladay-Atlas, OCT formula, and ASCRS average output
þ1.020
þ0.735

þ1.085
þ0.807
þ0.585
þ4.611

in 65 patients. In their study, the True K [No History]

þ1.88

tended toward more myopic outcomes than the other

methods, but overall there were no significant differences
to
to
to
to
to
to
to

among the formulae. The original True K [No History] for

0.795

1.245

0.845

1.318

2.193

1.818

2.631

post-RK eyes yielded outcomes comparable to the OCT

formula and the DK-Holladay-IOLM, on the order of 40%
within 0.50 D and 60% within 1.00 D of target at 1
month.10 MedAE at 1 month (n ¼ 65) was 0.78, 0.74, and
SD of ME
0.410
0.446
0.474
0.515
0.648
0.515
1.226

0.60 for the DK-Holladay-IOLM, OCT, and True K [No

History], respectively, and at 4 months (n ¼ 26) was
0.34, 0.65, and 0.69, respectively. With the DK-Holladay-
IOLM, 54% and 77% of eyes achieved within 0.50 D
þ0.242)

þ0.398)
þ0.134)

0.179)

0.366)
þ0.485)

and 1.00 D of target, respectively, at 4 months.10 These

0.043)

are similar to the results obtained for this formula at 4 to 6

ME (95% CI)

weeks in the current study.

to
to
to
to
to
to
to

Around the same time as Ma et al10 reported their

(
0.008
(
0.209
(þ0.140
(
0.146
(
0.531
(
0.646
(
0.257

findings, the RK version of the True K was undergoing

an update that was subsequently uploaded to the ASCRS
0.125

0.083
0.269

0.006

0.355

0.506
0.188

online IOL Calculator and the Asia Pacific Association

of Cataract and Refractive Surgeons website. This was
communicated to the authors at the time (Barrett GD,
Koch DD, Wang L, personal correspondence, January
47
48
52
52
52
52
52
n

21, 2017). Thus, the results presented are from a

substantially modified and improved version of the
True K [Partial History]

formula.
Haigis [
0.50 offset]
True K [No History]

DK-Holladay-IOLM

The present study is unique in that the complete

True K [History]
Formula

refractive history was available for nearly all patients (47/

52 eyes). This enabled a comparison to be made between
Potvin-Hill

the True K [History] (n ¼ 47) and a variety of no history

(n ¼ 52) or partial history (n ¼ 48) options. The True K
Haigis

(RK) formulae are derived from a theoretical model that

48
 Turnbull et al 
IOL Calculations after RK

uses changes in refraction to calculate the preoperative History] were almost identical, with 75.0% to 76.6% of eyes
keratometry and predict the contribution of the posterior within 0.50 D and 95.8% to 97.9% within 1.00 D of
cornea. With the True K [History], the magnitude of RK target. This demonstrates the value of obtaining details of
correction is known. Conversely, with the True K [No previous refractions, even if pre-RK data are unavailable.
History], the change in refraction is predicted from the AL, For the post-RK refraction, we advocate using the most
anterior chamber depth, and current keratometry, enabling hyperopic refraction recorded for the eye. This allows for
an estimation of the preoperative keratometry. In the case of the progressive hyperopic shift associated with RK, while
the True K [Partial History], the post-RK refractive status is controlling for any subsequent index myopia caused by
used to improve the prediction. This estimation of the pre- nuclear sclerosis. Although the difference in absolute error
operative corneal curvature is used to derive the ELP, with between the True K [History] and True K [No History]
the IOL power being derived from the current biometry. In methods was not statistically significant, this may represent
this way, the “double-K” issue is addressed, rather than a type II error. The difference in MedAE (0.275 vs. 0.330)
relying on a population mean. The True K (RK) is different and percentage within 0.50 D of target (76.6% vs. 69.2%)
than the True K algorithms used for myopic or hyperopic are clinically important and using the full or partial refrac-
LASIK/PRK, whereas the normal relationship between the tive history is strongly recommended when it is available.
anterior and posterior cornea is altered by RK. The change is The results with the True K [History] and True K [Partial
less fundamental than occurs after laser vision correction. History] approach the outcomes achievable with modern
Of the no history methods analyzed, the True K [No formulae in virgin eyes. In a recently published study of
History] performed significantly better than the DK- normal eyes comparing state-of-the-art formulae including
Holladay-IOLM, Potvin-Hill, or Haigis [
0.5 offset], but the updated Hill 2.0, Olsen, Barrett Universal II, and newly
similarly to the standard Haigis formula. described Kane formula, MAE and MedAE ranged from
By using the Potvin-Hill formula, the percentage of pa- 0.329 to 0.381 and from 0.231 to 0.311, respectively. Some
tients within 0.50 D and 1.00 D of target in the current 72.1% to 77.9% and 95.5% to 96.6% eyes achieved within
study (40% and 69%, respectively) closely replicated those 0.50 D and 1.00 D of target, respectively.19
reported in the original publication (42% and 76%).9 Using the average output from the ASCRS calculator,
Our results with the Haigis formula contrast with those rather than favoring 1 formula, is a popular option with
reported previously. Geggel4 studied 26 eyes of 20 patients some surgeons. This method has been investigated.10,18 We
and evaluated the Haigis formula with keratometry did not specifically study this, because taking the average of
measured by a range of devices. By using the Haigis multiple formulae will dilute the overall performance of
formula with a
1.00 offset (aiming for an outcome more accurate methods. The ASCRS average output has
of
0.50 D) and IOLMaster keratometry, the PE ranged improved as more advanced formulae have been added to
from
2.11 to þ0.25, with ME of
0.660.61 and MAE the calculator, but it will always be limited by the less
of 0.51. Some 69% and 88% of eyes were within 0.50 precise methods. We therefore advise against this approach.
D and 1.00 D of target, respectively.4 Although other A limitation of the current study is the relatively short
evaluated options led to smaller MAE and ME, high follow-up period of 4 to 6 weeks, which has been suggested
hyperopic surprises were more likely, which led to the as the minimum in this context.1 It is known that an initial
recommendation of the Haigis [
0.50 offset] as the best hyperopic overshoot due to RK incision swelling and
compromise. Geggel4 also investigated the utility of the temporary corneal flattening, followed by partial myopic
postrefractive Haigis-L formula, which was designed for regression, is observed after cataract surgery in eyes with
use after myopic or hyperopic laser refractive surgery but previous RK.1,10,20 This phenomenon has been docu-
not RK. Although this was successful in eliminating hy- mented to stabilize by 3 months.11,21,22 Therefore, an
peropic surprises, higher myopic errors were encountered outcome of emmetropia or low hyperopia at 4 weeks is not
(
4.80 to
0.12).4 considered problematic, because although the eye may gain
In our study, the True K [No History] performed simi- some additional myopia that should be well tolerated by
larly to the standard Haigis formula with no offset. Possible most patients, it is unlikely to become more hyperopic.
explanations for this difference include variations in the Both eyes of some patients were included. This was
cataract surgical technique, refraction protocol, follow-up considered appropriate because of the relative scarcity of
period, number of RK incisions, and presence or other- post-RK eyes with complete refractive history undergoing
wise of arcuate keratotomies for astigmatism. Although the cataract surgery to avoid “wastage” of valuable data.23 This
Haigis formula was not designed for post-RK eyes, its replicates the methodology of other key publications on this
reliance on data other than keratometry to predict the ELP topic.4,10 However, an element of between-eye correlation is
has advantages. According to our results, the True K [No acknowledged as a potential limitation of the analysis.
History] and Haigis formulae are both useful options for In conclusion, the RK algorithm of the True K formula
patients in whom refractive history is unavailable or produces refractive outcomes that approach the accuracy of
unreliable. IOL calculations in virgin eyes. Although it is preferable to use
The inclusion of refractive history led to improvements the True K [History] or True K [Partial History], good results
in the percentage of eyes within each refractive bracket. The are obtainable in eyes in which no refractive history is available
results of the True K [History] and the True K [Partial by using the True K [No History] or Haigis formula. Another

49
 Ophthalmology Volume 127, Number 1, January 2020

advantage of this approach is that biometric data from multiple radial keratotomy. Invest Ophthalmol Vis Sci. 2016;57:
devices are not required. The most important factor in the OCT162eOCT168.
refractive history appears to be the post-RK refraction, which is 11. Lyle AW, Jin GJ. Intraocular lens power prediction in patients
taken to be the most hyperopic refraction on record for a who undergo cataract surgery following previous radial kera-
particular eye. The inclusion of pre-RK data provides a small totomy. Arch Ophthalmol. 1997;115:457e461.
12. Canto AP, Chhadva P, Cabot F, et al. Comparison of IOL
additional benefit but is not essential. power calculation methods and intraoperative wavefront
aberrometer in eyes after refractive surgery. J Refract Surg.
References 2013;29:484e489.
13. Zhang F. Optiwave Refractive Analysis may not work well in
patients with previous radial keratotomy. Am J Ophthalmol
1. Alio JL, Adbelghany A, Abdou AA, Maldonado MJ. Cataract Case Rep. 2018;10:163e164.
surgery on the previous corneal refractive surgery patient. Surv 14. Aramberri J. Intraocular lens power calculation after corneal
Ophthalmol. 2016;61:769e777. refractive surgery: double-K method. J Cataract Refract Surg.
2. Camellin M, Savini G, Hoffer K, et al. Scheimpflug camera 2003;29:2063e2068.
measurement of anterior and posterior corneal curvature in 15. Packer M, Brown LK, Hoffman RS, Fine IH. Intraocular lens
eyes with previous radial keratotomy. J Refract Surg. 2012;28: power calculation after incisional and thermal keratorefractive
275e279. surgery. J Cataract Refract Surg. 2004;30:1430e1434.
3. Chen L, Mannis MJ, Salz JJ, et al. Analysis of intraocular lens 16. Awwad ST, Dwarakanathan S, Bowman W, et al. Intraocular
power calculation in post-radial keratotomy eyes. J Cataract lens power calculation after radial keratotomy: estimating the
Refract Surg. 2003;29:65e70. refractive corneal power. J Cataract Refract Surg. 2007;33:
4. Geggel HS. Intraocular lens power selection after radial ker- 1045e1050.
atotomy: topography, manual, and IOLMaster keratometry 17. Huang D, Tang M, Wang L, et al. Optical coherence
results using Haigis formulas. Ophthalmology. 2015;122: tomography-based corneal power measurement and intraocular
897e902. lens power calculation following laser vision correction. Trans
5. Hill W, Wang L, Koch DD. IOL calculator for eyes with prior Am Ophthal Soc. 2013;111:57e68.
RK. http://iolcalc.ascrs.org/wbfrmCalculator3.aspx. Accessed 18. DeMill DL, Hsu M, Moshirfar M. Evaluation of the American
May 11, 2019. Society of Cataract and Refractive Surgery intraocular lens
6. Barrett GD. True-K formula: new approach to biometry after calculator for eyes with prior radial keratotomy. Clin Oph-
LASIK. Presented at: the American Society of Cataract and thalmol. 2011;5:1243e1247.
Refractive Surgeons’ Annual Meeting, April 3-8, 2009, San 19. Connell BJ, Kane JX. Comparison of the Kane formula with
Francisco, California. existing formulas for intraocular lens power selection. BMJ
7. Barrett GD. An improved universal theoretical formula for Open Ophthalmol. 2019;4:e000251.
intraocular lens power prediction. J Cataract Refract Surg. 20. Koch DD, Liu JF, Hyde LL, et al. Refractive complications of
1993;19:713e720. cataract surgery after radial keratotomy. Am J Ophthalmol.
8. Barrett GD. Barrett True K Formulaefor prior myopic or hy- 1989;108:676e682.
peropic LASIK/PRK/RK. Version 2.0. http://calc.apacrs.org/ 21. Hoffer K. Intraocular lens power calculation for eyes after
Barrett_True_K_Universal_2105. Accessed May 11, 2019. refractive keratotomy. J Refract Surg. 1995;11:490e493.
9. Potvin R, Hill W. New algorithm for post-radial keratotomy 22. Gimbel H, Sun R, Kaye GB. Refractive error in cataract sur-
intraocular lens power calculations based on rotating gery after previous refractive surgery. J Cataract Refract Surg.
Scheimpflug camera data. J Cataract Refract Surg. 2013;39: 2000;26:142e144.
358e365. 23. Murdoch IE, Morris SS, Cousens SN. People and eyes: sta-
10. Ma JX, Tang M, Wang L, et al. Comparison of newer tistical approaches in ophthalmology. Br J Ophthalmol.
IOL power calculation methods for eyes with previous 1998;82:971e973.

Footnotes and Financial Disclosures

Originally received: June 29, 2019. HUMAN SUBJECTS: No human subjects were included in this study.
Final revision: July 27, 2019. Institutional Review Board approval for chart review and IOL formulae
Accepted: August 15, 2019. comparison was previously obtained, and the study was conducted ac-
Available online: August 23, 2019. Manuscript no. 2019-1424. cording to the tenets of the Declaration of Helsinki.
1
Sir Charles Gairdner Hospital, Perth, WA, Australia. No animal subjects were used in this study.
2
Lions Eye Institute, Perth, WA, Australia. Author Contributions:
3 Conception and design: Turnbull, Barrett
Royal Bournemouth Hospital, Bournemouth, United Kingdom.
4
Data collection: Turnbull, Crawford, Barrett
Centre for Ophthalmology and Visual Science, University of Western
Analysis and interpretation: Turnbull, Barrett
Australia, WA, Australia.
Obtained funding: N/A
Presented at: the Australasian Society of Cataract and Refractive Surgeons,
Overall responsibility: Turnbull, Barrett
July 20, 2019, Queenstown, New Zealand.
Financial Disclosure(s):
The author(s) have made the following disclosure(s): G.D.B. is the author
of the True K formulae and has licensed these to several manufacturers.

50
 Turnbull et al 
IOL Calculations after RK

Abbreviations and Acronyms: Correspondence:

AL ¼ axial length; ASCRS ¼ American Society of Cataract and Refractive Andrew M. J. Turnbull, BM, FRCOphth, Sir Charles Gairdner Hospital,
Surgeons; CI ¼ confidence interval; D ¼ diopters; ELP ¼ effective lens Eye Clinic, E Block, Hospital Avenue, Nedlands, WA 6018, Australia.
position; IOL ¼ intraocular lens; ME ¼ mean error; MAE ¼ mean ab- E-mail: [email protected].
solute error; MedAE ¼ median absolute error; PE ¼ prediction error;
RK ¼ radial keratotomy.

Pictures & Perspectives

Progression from Classical Oguchi Disease to Retinitis Pigmentosa after 50 Years

Fundus photograph of a 77-year-old Japanese man with retinitis pigmentosa (RP) and homozygous c.636delT in S-antigen (SAG)
(Fig A). The patient’s condition had been reported as Oguchi disease (Fig B) with Mizuo-Nakamura phenomenon at the 20th International
Congress of Ophthalmology held at Munich in 1966. Fundus photograph of the same patient taken around that time (age w23) showed
metallic sheen covering the posterior pole with no sign of retinal degeneration. These photographs mark the first example of progression of
classical juvenile Oguchi disease to RP. The patient was recruited through the Japanese Retinitis Pigmentosa Registry Project. Ethical
approval for this study was granted by the Institutional Review Board of Tohoku University School of Medicine. The old fundus
photograph showing classical Oguchi disease was recovered from the articles of Dr. Takehisa Oguchi (deceased). (Magnified version of Fig
A-B is available online at www.aaojournal.org).
KOJI M. NISHIGUCHI, MD, PHD1,2
YOSHIHISA OGUCHI, MD, PHD3
TORU NAKAZAWA, MD, PHD1,2
1
Department of Advanced Ophthalmic Medicine, Tohoku University Graduate School of Medicine, Sendai, Japan; 2Department of Ophthalmology,
Tohoku University Graduate School of Medicine, Sendai, Japan; 3The Mizumachi Hoken Clinic, Tokyo, Japan

Footnotes and Financial Disclosures

Footnotes and Financial support: Supported in part by the Japan Agency for Medical Research and Development (grant no. 19ek0109213h0001 [to K.M.N.]).

51
A Prospective Randomized Trial Comparing
Hydrus and iStent Microinvasive Glaucoma
Surgery Implants for Standalone Treatment
of Open-Angle Glaucoma
The COMPARE Study
Iqbal Ike K. Ahmed, MD,1 Antonio Fea, MD, PhD,2 Leon Au, MBBS,3 Robert E. Ang, MD,4
Paul Harasymowycz, MD,5 Henry D. Jampel, MD,6 Thomas W. Samuelson, MD,7 David F. Chang, MD,8
Douglas J. Rhee, MD,9 on behalf of the COMPARE Investigators

Purpose: To compare the efficacy of different microinvasive glaucoma surgery (MIGS) devices for reducing
intraocular pressure (IOP) and medications in open-angle glaucoma (OAG).
Design: Prospective, multicenter, randomized clinical trial.
Participants: One hundred fifty-two eyes from 152 patients aged 45 to 84 years with OAG, Shaffer angle
grade IIIeIV, best-corrected visual acuity (BCVA) 20/30 or better, and IOP 23 to 39 mmHg after washout of all
hypotensive medications. Eyes with secondary glaucoma other than pseudoexfoliative or pigmentary glaucoma,
angle closure, previous incisional glaucoma surgery, or any significant ocular pathology other than glaucoma
were excluded.
Intervention: Study eyes were randomized 1:1 to standalone MIGS consisting of either 1 Hydrus Microstent
(Ivantis, Inc, Irvine, CA) or 2 iStent Trabecular Micro Bypass devices (Glaukos Inc, San Clemente, CA). Follow-up
was performed 1 day, 1 week, and 1, 3, 6, and 12 months postoperatively.
Main Outcome Measures: Within-group and between-group differences in IOP and medications at 12
months and complete surgical success defined as freedom from repeat glaucoma surgery, IOP 18 mmHg or less,
and no glaucoma medications. Safety measures included the frequency of surgical complications, changes in
visual acuity, slit-lamp findings, and adverse events.
Results: Study groups were well matched for baseline demographics, glaucoma status, medication use, and
baseline IOP. Twelve-month follow-up was completed in 148 of 152 randomized subjects (97.3%). At 12 months, the
Hydrus had a greater rate of complete surgical success (P < 0.001) and reduced medication use (difference ¼ 0.6
medications, P ¼ 0.004). More Hydrus subjects were medication free at 12 months (difference ¼ 22.6% P ¼ 0.0057).
Secondary glaucoma surgery was performed in 2 eyes in the 2-iStent group (3.9%) and in none of the Hydrus eyes.
Two eyes in the Hydrus group and 1 in the 2-iStent group had BCVA loss of 2 lines.
Conclusion: Standalone MIGS in OAG with the Hydrus resulted in a higher surgical success rate
and fewer medications compared with the 2-iStent procedure. The 2 MIGS devices have similar safety
profiles. Ophthalmology 2020;127:52-61 ª 2019 by the American Academy of Ophthalmology

Supplemental material available at www.aaojournal.org.

Glaucoma is the second-leading cause of blindness world- Topical medications have a proven record of efficacy and
wide.1 Although intraocular pressure (IOP) is not a direct are used as first-line therapy at all stages of glaucoma and
measure of structural or functional optic neuropathy, it is ocular hypertension. Though chronic medication use is
the only risk factor that can be modified, and reduction of generally safe, efficacy is frequently undermined by high
IOP has been shown to reduce glaucoma progression and rates of patient noncompliance,6e8 which increase when
visual field loss.2 IOP can be lowered medically or multiple medications are concurrently prescribed. Medica-
surgically, and the treatment modality is usually based on tions are also associated with side effects and may exacerbate
the severity of visual field impairment and rate of dry eye and ocular surface disease.9 Furthermore, chronic
progression.3e5 medication use may reduce the success rate of subsequent

52 ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.04.034

Published by Elsevier Inc. ISSN 0161-6420/19
 Ahmed et al 
Randomized Comparison of Hydrus and iStent

glaucoma filtration surgery.10 Surgical management lowers stratified by site and prepared in advance by the study statistician
IOP more than medical management but is typically so as to provide balanced study groups. Baseline and postoperative
reserved for more advanced disease because of the potential tonometry was performed according to the 2-operator method to
for sight-threatening complications.11 provide masking as described in the Ocular Hypertension Treat-
ment Study.25 All medications were discontinued before surgery,
A new class of microinvasive glaucoma surgery
and reintroduction at IOP <19 mmHg required clinical
(MIGS) devices12 has been developed that shunt aqueous justification. Additionally, study subjects were masked to their
into the Schlemm canal,13 the suprachoroidal,14 or treatment assignment.
subconjunctival15 space. MIGS devices may be safer than The study protocol was approved by the Medical Ethics
conventional filtering surgery, can eliminate or reduce Committees at each site. The study was conducted according to the
adjunctive topical medication therapy, and do not impede principles described in the Declaration of Helsinki and all study
or preclude subsequent filtration surgery.16 In randomized subjects provided written informed consent before participation in
clinical trials, 3 different MIGS devices implanted in the trial. The study was registered in the National Library of
combination with phacoemulsification were shown to be Medicine database (clinicaltrials.gov NCT02023242). The study
effective in reducing IOP and medication use.17e20 data were 100% source document verified by independent clinical
monitors with funding provided by the study sponsor. The data set
The purpose of this study was to compare 2 different
was audited and the final statistical analyses were conducted using
Schlemm canalebased MIGS devices for standalone use in SAS (software version 9.3; SAS Institute Inc, Cary, NC).
phakic or pseudophakic patients with OAG. The Hydrus
Microstent (Ivantis, Inc, Irvine, CA) and the iStent
Study Patients
Trabecular Micro-Bypass Stent System (Glaukos Corpora-
tion, San Clemente, CA) both provide a conduit for aqueous The risks and benefits of the 2 procedures were described to po-
fluid into the Schlemm canal through the trabecular mesh- tential subjects, and signed informed consent was required for in-
work (TM), which is thought to be the primary location of clusion in the study. Only 1 eye per patient was eligible for
outflow obstruction in open-angle glaucoma (OAG).21e23 treatment, although both eyes could be screened for inclusion.
The iStent provides a trabecular bypass and penetrates Entry criteria included phakic and pseudophakic eyes with a
diagnosis of mild-to-moderate OAG (primary OAG, pseudoexfo-
approximately 1 mm into the Schlemm canal, and the
liative glaucoma, or pigmentary glaucoma) confirmed by optic
Hydrus device provides a trabecular bypass and dilates nerve examination, characteristic visual field deficits on automated
approximately 3 clock hours of the Schlemm canal, thereby perimetry, and a history of hypotensive medication use. Study
providing direct aqueous access to a quadrant of collector candidates had to be capable of safely undergoing medication
channels. washout, in the opinion of the investigator.
Previous laboratory experiments have shown mechanistic Clinical exclusion criteria included angle closure glaucoma,
differences in the 2 approaches; in paired cadaveric eyes, the secondary glaucoma (except for pseudoexfoliative and pigmen-
Hydrus Microstent increased outflow facility by a factor of 2 tary glaucoma), exudative age-related macular degeneration,
to 2.5 when compared with eyes implanted with 2 iStent proliferative diabetic retinopathy, or significant risk of glaucom-
devices.24 Based on the mechanistic differences and atous vision loss owing to washout of IOP-lowering medications.
Anatomic exclusion criteria were narrow anterior chamber angle
laboratory results, we hypothesized that a single Hydrus
(Shaffer grade IeII) or other angle abnormality visible upon
Microstent lowers IOP and decreases number of gonioscopy, clinically significant corneal dystrophy, and central
medications more than 2 iStent devices. corneal thickness of less than 480 or more than 620 mm. Patients
with prior corneal surgery, cycloablation, or any incisional
glaucoma procedure such as trabeculectomy, tube shunts, deep
Methods sclerectomy, or canaloplasty were also excluded. Prior selective
laser trabeculoplasty (SLT) was allowed, but not argon laser
Study Design trabeculoplasty.
Subjects who met eligibility criteria had to have a diurnal IOP
The COMPARE study was a prospective, multicenter, randomized, (DIOP) of 23 to 39 mmHg after washout of ocular hypotensive
single-masked clinical trial. The efficacy measures were chosen to medications. The duration of washout was a minimum of 4 weeks
evaluate the ability of each device to lower IOP while reducing or for prostaglandin analogues or b-blockers and 2 weeks for carbonic
eliminating hypotensive medications. The study was conducted at 12 anhydrase inhibitors or a adrenergic agonists. The DIOP value was
investigational sites in 9 countries (see listing in online Appendix; obtained by averaging 3 Goldmann tonometry measurements taken
supplementary materials available at www.aaojournal.org). Follow- 4 hours apart between 8AM and 4PM. Two readings were taken at
up was completed at 1 year, at which time the outcomes were assessed. each time point. If the difference in the 2 measurements was more
Because this study compares 2 types of surgically implanted than 2 mmHg, a third measurement was taken. The average of 2
devices, prior experience with both devices was required of the measurements (or the median value of 3) was used for the time
investigators. The iStent had been in commercial distribution point, and the DIOP was the average of all 3 time points on the
internationally for at least 3 years before the start of the study, and visit day. Subjects whose DIOP met entry criteria were scheduled
so the investigators began this study with comparatively more for surgery.
experience with the iStent device.
The study protocol was designed to minimize the potential for Study Devices
bias in determining the efficacy end points. Baseline washout of
medications was required to ensure inclusion of subjects within a The Hydrus Microstent is a metallic microstent designed to bypass
known range of IOP. Randomization was performed in the oper- the TM and dilate approximately 3 clock hours of the Schlemm
ating room by opening a sequentially numbered envelope. The canal. The details of the device construction and preclinincal
allocation was determined by a computer-generated sequence testing have been described previously.20 The Hydrus was first

53
 Ophthalmology Volume 127, Number 1, January 2020

approved for international use in 2011 and is currently approved in intraoperative complications and the prevalence of ocular adverse
the United States for use in combination with cataract surgery. The events.
iStent is an “L”-shaped titanium device, with a single fluid inlet The study was originally designed to include washout at the
port and extension that is designed to extend approximately 1 12-month follow-up visit, thereby allowing for direct comparison of
mm into the Schlemm canal.26 Like the Hydrus, iStent is IOP reduction associated with each device without the confounding
approved internationally for general use but in the United States effect of concomitant medications. However, among the first 40
in combination with cataract surgery only. randomized study subjects to reach 12 months follow-up, in-
Prior comparative studies suggest the Hydrus may be superior vestigators were unwilling to wash out approximately 20% of eyes in
to SLT27 and canaloplasty.28 The effectiveness of 1 or more iStents the 2-iStent group owing to persistent IOP elevation despite appli-
for reduction of IOP has been reported in cadaveric outflow cation of maximum tolerated medical therapy. Therefore, in the in-
studies,29 in single-center clinical studies in combination with terest of patient safety, the study protocol was modified to eliminate
cataract surgery,30 and as a standalone procedure with and without the 12-month washout requirement. The modified study plan was
adjunctive hypotensive medication.31,32 Published data indicate fully implemented across study sites by the time the 64th subject
that placement of 2 iStents may increase IOP-lowering effective- reached the 12-month visit. As a result of this change, the ability to
ness compared with a single iStent implant in standalone directly compare washed-out IOP was lost, but comparisons asso-
procedures.33 ciated with medicated IOP and medication reduction were preserved.
As a surrogate for washout, categorical end points evaluating un-
Surgical Technique medicated eyes with a 20% or more reduction in IOP, or an IOP of
18 mmHg, were added to the protocol.
The surgical microscope was positioned and the head tilted to allow a
clear view of the angle structures with a surgical gonioprism. Subjects
Statistical Analysis
were randomized for treatment with a single Hydrus Microstent or 2
iStent devices after intraoperative gonioscopy confirmed the angle The study was originally designed to detect a 2 mm difference in 12-
structures could be adequately visualized. Viscoelastic was injected month washed-out IOP with >80% power and 0.05 significance
through a 1- to 1.5-mm clear corneal incision for chamber mainte- level; approximately 60 subjects per study arm were required. After
nance and better angle visualization. The assigned device(s) was elimination of the follow-up washout, an additional 15 subjects per
introduced into the anterior chamber through the incision and arm were added to provide >80% power to detect a 25% difference
implanted through the TM in the nasal hemisphere of the Schlemm in the proportion of eyes that were medication free with IOP 18
canal. The iStents were implanted in the nasal hemisphere approxi- mmHg or less at 12 months with 0.05 significance level. Analyses of
mately 2 clock hours apart, facing opposite directions, according to the the efficacy outcomes were performed using the intention-to-treat
technique described by Ahmed.34 iStent placement was targeted to principle. Patients who underwent glaucoma surgery or IOP-
areas with reflux of blood or greater TM pigmentation where lowering procedures of any kind (including trabeculoplasty or
apparent, as these are thought to represent a locus of collector cataract surgery) after the index procedure were counted as failures
channel outflow. Hydrus implantation was done as described in categorical efficacy measures and assigned last observed IOP and
previously by Pfeiffer et al.18 Upon visual confirmation of device(s) medication values for subsequent IOP and medication averages.
position in the canal, the delivery system was withdrawn, Categorical variables were reported as counts and percentages and
viscoelastic removed, and the anterior chamber was inflated with differences were tested using Fisher exact test for binary variables.
balanced salt solution to achieve physiologic IOP. Patients were Means and standard deviations of continuous variables are presented
administered antibiotics (moxifloxacin 0.5% or equivalent, 1 drop 4 according to treatment group. Within-group and between-group
times per day, commencing the day of the procedure and continued differences were tested using unpaired 2-sided t tests.
for approximately 1 week postoperatively) and anti-inflammatory
medications (prednisolone acetate 1.0% or equivalent) with initial
frequency of 1 drop 4 times per day, tapered over 4 weeks Results
postoperatively.
Preoperative Characteristics
Follow-up Examinations
A total of 152 eyes from 152 patients were randomized from March
Follow-up examinations were conducted at 1 day, 1 week, 1 month, 2013 to May 2015. The study population included 75 Hydrus eyes
3 months, 6 months, and 12 months postoperatively. At each and 77 2-iStent eyes. The 12-month IOP assessment was
scheduled visit, examinations included slit-lamp biomicroscopy, completed in 73 of 75 (97.3%) Hydrus and 75 of 77 (97.4%)
ophthalmoscopy, manifest refraction, visual acuity using the Early 2-iStent eyes; 1 subject in each group was lost to follow-up and 1
Treatment of Diabetic Retinopathy Study (ETDRS) system, and subject from each group missed the 12-month visit and were
measurement of IOP using Goldmann applanation tonometry. excluded from the analyses.
Automated achromatic visual field testing using the 24-2 SITA The 2 randomized study arms were well balanced for de-
standard strategy using a Humphrey Visual Field Analyzer (Carl mographic and baseline characteristics (Table 1). In the intention-
Zeiss Meditech, Jena, Germany) at 3 and 12 months. Ocular hypo- to-treat population, the mean age was 66 years and 55% of study
tensive medications could be added during follow-up at the subjects were women; 65.3% and 62.3% were phakic in the Hydrus
investigator’s discretion. At 12 months, DIOP was measured using and 2 iStent groups, respectively. Primary OAG was the predom-
the same tonometry techniques as the baseline visit. inant diagnosis. At the screening visit, mean IOP was 19.03.9
mmHg and 19.13.6 mmHg on 2.50.7 and 2.70.8 medications
End Points in the Hydrus and 2-iStent groups, respectively. Almost all patients
were on multiple medications at baseline. The most frequently used
Clinical outcome measures included changes in mean IOP and medications were prostaglandin analogues and b-blockers, or a
medications. Surgical success was defined as freedom from sec- combination of both. After medication washout, baseline DIOP
ondary surgery, IOP 18 mmHg or less, and discontinuation of all was 27.54.4 mmHg and 27.34.2 mmHg in the Hydrus and
ocular hypotensive medications. Safety end points included 2-iStent groups, respectively.

54
 Ahmed et al 
Randomized Comparison of Hydrus and iStent

Table 1. Baseline Demographics and Ocular Characteristics significantly reduced medication use at 12 months compared with
preoperative; however, the reduction in the Hydrus group was
Hydrus 2 iStents significantly greater (difference in change ¼ e0.6 medications;
Parameter N [ 75 N [ 77 P Value 95% confidence interval [CI], e0.9 to e0.2; P ¼ 0.004). IOP
Age (years) 66.910.0 66.59.5 0.9
distribution stratified by IOP interval are shown in Table 2. Even
Female 54.7% 58.4% 1.0 with fewer medicaitons, there were significant reductions in the
European 65.3% 63.6% 0.9 number of eyes at IOP  21, 18, and 15 mmHg at 12 months
Latin American 18.7% 16.9% 0.8 compared to preoperative in the Hydrus group and no significant
Asian 14.7% 14.3% 1.0 changes in the 2 iStent group.
African ancestry 1.3% 3.9% 0.6 At 12 months 46.6% of Hydrus patients and 24.0% of 2-iStent
Glaucoma status patients were medication free (P ¼ 0.006) (Table 3 and Figure 3).
POAG 96.0% 92.2% 0.5 The percentage of eyes reaching 18 mmHg without medications
PXG/PDG 4.0% 7.8% 0.5 was greater in the Hydrus group (30.1% vs. 9.3%, P ¼ 0.002), as
MD, dB 6.25.4 6.26.5 1.0 was the percentage of eyes reaching a 20% or more reduction in
PSD, dB 5.53.5 5.13.3 0.9 IOP from washed-out baseline without medications (39.7% vs.
Ocular status 13.3%, P <0.001). The mean IOP for eyes without medications
Phakic 65.3% 62.3% 0.7 was 17.33.3 in the Hydrus group and 19.22.4 in the 2-iStent
Pseudophakic 34.7% 37.7% 0.7 group (mean change e8.2 mmHg vs. e5.1 mmHg, difference in
BCVA (Snellen) 20/25 20/24 0.9 change, e3.1 mmHg; 95% CI, e5.4 to e0.8 mmHg; P ¼ 0.003).
Pachymetry (mm) 54236 54134 1.0 Before surgery, 50.7% of Hydrus eyes and 58.5% of 2 iStent
Mean vertical C:D 0.650.16 0.670.18 0.9 eyes were on 3 or more medications. At 12 months, the percentage
Previous SLT 14.7% 15.6% 1.0
of eyes on 3 medications was lower in the Hydrus vs. 2-iStent
Preoperative IOP and medication
group (8.2% vs. 29.3%, P ¼ 0.001). Conversely, the number of
Mean IOP, mm Hg 19.03.9 19.13.6 0.8
Mean medications 2.50.7 2.70.8 0.2
subjects with no change or an increase in medication use was lower
WO DIOP (mmHg) 27.54.4 27.34.2 0.8 in the Hydrus group (17.8% vs. 38.7%, P ¼ 0.006), and the
Medication distribution class number of patients with a 3-medication reduction was higher
PGA 68% 68% 0.9 (23.3% vs. 5.3%, P ¼ 0.002) in the Hydrus group. (Changes in
BB 23% 31% 0.3 medication use from baseline to follow-up are presented in
CAI 44% 45% 0.9 Table S1, available at www.aaojournal.org).
AA 15% 17% 0.8 Because the follow-up washout was eliminated from the pro-
BB þ PGA* 19% 23% 0.6 tocol after approximately 42% of subjects completed the 12-month
BB þ CAI* 21% 23% 0.8 visit, the medication-prescribing behavior of study investigators
BB þ AA* 13% 12% 0.8 was assessed to determine if the study groups were treated uni-
formly. During the first year of follow-up, there were more than
AA ¼ a adrenergic agonist; BB ¼ b-blocker; BCVA ¼ best-corrected
600 IOP assessments. The mean IOP for a medication increase
visual acuity; CAI ¼ carbonic anhydrase inhibitor; C:D ¼ cup-to-disc decision was 21.34.9 mmHg in the Hydrus group and 21.85.9
ratio; DIOP ¼ diurnal intraocular pressure; IOP ¼ intraocular pressure; mmHg in the 2-iStent group (P ¼ 0.61). The mean IOP for all
MD ¼ mean deviation; PDG ¼ Pigmentary Glaucoma; PGA ¼ prosta- decisions to leave medications unchanged or to decrease by 1 or
glandin analogue; POAG ¼ primary open-angle glaucoma; PSD ¼ pattern more was 16.93.4 mmHg in the Hydrus group and 17.53.8
standard deviation; PXG ¼ pseudoexfoliative glaucoma; SLT ¼ selective mmHg in the 2-iStent group (P ¼ 0.09). The intent of the protocol
laser trabeculoplasty; WO ¼ Washed Out. was to leave eyes unmedicated if IOP was <19 mmHg. The fre-
*Combination medication. quency at which medication was added at this IOP range as a
percentage of all medication change decisions was evenly distrib-
uted between groups at 28.6% in the Hydrus group and 28.8% in
the 2-iStent group. The medication change decisions were also
Procedure Outcomes evenly distributed at various IOP intervals above 19 mmHg
Implantation was successful in 100% of Hydrus eyes and 97.4% of (Fig S1, available at www.aaojournal.org).
2-iStent eyes; in 2 cases only 1 iStent could be implanted. These Surgical success through 12 months was assessed using
subjects were included in the intent-to-treat analysis. All subjects KaplaneMeier event-free survival analysis. Failure was defined as
underwent the assigned procedure. There were no instances of lost repeat glaucoma surgery or any IOP-lowering procedure (including
or migrating devices or corneal touch in either group, and there cataract surgery) at any postoperative time, IOP >18 mmHg, or use
were no surgical complications associated with either device. of glaucoma medications for 2 consecutive visits after 1 month. As
shown in Figure 4, the 12-month cumulative event-free survival
Efficacy Measures rate was 35.6% in the Hydrus group and 10.5% in the 2-iStent
group (P ¼ 0.001, log-rank test).
Mean IOP and medication use at each follow up visit are shown in This study included both phakic and pseudophakic eyes.
Figure 1A-B. IOP was uniformly lower throughout the 12 month Compared with phakic eyes, there were more eyes on 0 medica-
follow up, and although there were no significant between-group tions in the Hydrus pseudophakic subgroup (56% vs. 42%) but
differences in IOP at any time point, medication use was signifi- fewer in the 2-iStent group (14% vs. 30%), although the differ-
cantly lower in the Hydrus group at all visits after 1 month. The ences were not significant. The proportion of eyes with 0 medica-
results of the analysis for change in IOP and medications from tions and IOP 18 mmHg was also higher in the Hydrus
preoperative to 12 months are presented in Figure 2A-B. Although pseudophakic eyes compared with phakic eyes (40% vs. 25%), as
the mean IOP in the Hydrus group was lower compared with was the proportion of eyes with 0 medications and a 20% IOP
preoperative IOP, the between-group difference was not signifi- reduction (56% vs. 31%). In the 2-iStent group, the proportion of
cant (difference in change ¼ e0.7 mmHg, P ¼ 0.3). Both groups eyes with 0 medications and IOP 18 mmHg was similar in the

55
 Ophthalmology Volume 127, Number 1, January 2020

Figure 1. A, Intraocular pressure (IOP). B, Medications. There were no significant differences in IOP between groups at any time point. There was a
significant differences in mean medication count at all time points 90 days. The error bars are 95% confidence intervals.

pseudophakic and phakic subgroups (11% pseudophakic vs. 9% from baseline to 12 months was e6.05.4 mmHg in the Hydrus
phakic), as was the proportion of eyes with 0 medications and 20% group (n ¼ 30) and e4.05.6 in the 2-iStent group (n ¼ 24).
IOP reduction (11% pseudophakic vs. 15% phakic). The influence of key demographic and preoperative factors on the
Before the washout requirement was removed, 64 subjects probability that the subject would be medication free at 12 months
reached the 12-month end point, 32 in each study arm. The number was assessed using logistic regression The model included the
of washout exceptions was proportionally higher in the 2-iStent subject’s treatment group, ethnicity, age, number of baseline medi-
group (7 vs. 1), and so the number of eyes available for analysis cations (2 or 3), visual field mean deviation (e6 dB or > e6
was not evenly distributed between groups. In the Hydrus group, dB), baseline washed-out DIOP, and surgeon experience (more than
30 of 32 eyes were available (94%); 1 eye was not washed out for or fewer than 100 cases before initiation of the study). The analysis
safety reasons and 1 was lost to follow-up. In the 2-iStent group, 24 showed that age, ethnicity, number of baseline medications, and
of 32 eyes were available (75%); 1 eye had trabeculectomy at surgeon experience did not significantly affect the end point. The
month 4 and 7 eyes were not washed out for safety reasons. For strongest predictor of freedom from medication was treatment with
eyes completing washout, the mean change in washed-out IOP the Hydrus device (odds ratio [OR], 4.08; 95% CI, 1.7e9.90;

Figure 2. A, Intraocular pressure (IOP). B, Assessment of IOP and medication at 12 months. Compared with preoperative IOP, there was a significant
reduction in IOP in the Hydrus group but not in the 2-iStent group. A, The difference in change between groups was not significant. Both groups
significantly reduced medication use compared with preoperative values. B, The difference in change was significantly greater in the Hydrus group. Note:
within-group P value is calculated for the 12-month vs. medicated preoperative IOP. The error bars are 95% confidence intervals.

56
 Ahmed et al 
Randomized Comparison of Hydrus and iStent

Table 2. Medicated Mean Intraocular Pressure and Stratified

Distribution

Preoperative 12 Months P Value*

Hydrus
N 75 73 e
Mean (SD) IOP, mmHg 19.02.5 17.33.7 0.009
% 21 mmHg 25.3 8.2 0.008
% 21 mmHg 74.7 91.8 0.008
% 18 mmHg 41.3 64.4 0.006
% 15 mmHg 17.3 24.7 0.31
2 iStents
N 77 75 e
Mean (SD) IOP, mmHg 19.13.6 18.13.7 0.10
% 21 mmHg 27.3 16.0 0.12
% 21 mmHg 72.7 84.0 0.12
% 18 mmHg 44.2 57.3 0.11
% 15 mmHg 14.3 20.0 0.31

IOP ¼ intraocular pressure; SD ¼ standard deviation.

*P value: within-group preoperative vs. 12 months.

P ¼ 0.001). The model also showed that lower baseline DIOP (OR,
0.73; 95% CI, 0.63e0.84; P < 0.0001) and milder visual field
severity (OR, 1.16; 95% CI, 1.04e1.29; P ¼ 0.0065) were signifi-
cant predictors of 0 medication count. (A summary of the analysis is
provided in Table S2, available at www.aaojournal.org.)

Safety
Slit-lamp findings were limited to mild anterior chamber cell and
flare and mild corneal edema, apparent in a minority of patients in
the first postoperative month. New cataracts were reported in 2 Figure 3. The percentage of eyes using 0, 1, 2, or 3 medications at 12
Hydrus eyes and 1 2-iStent eye by 12 months, and cataract surgery months in Hydrus and 2-iStent groups.
combined with trabeculectomy was performed in 1 eye in the 2-
iStent group. Mild posterior capsular opacification was reported
in approximately 5% of pseudophakic patients in both groups at or other tissue adhesions were more common in the 2-iStent group
screening and remained stable throughout follow-up. There were and obstructions owing to peripheral anterior synechiae (PAS)
no significant changes in fundus appearance and cup-to-disc were more common in the Hydrus group. There were no reports of
assessments. hypotony, device migration, or dislocation in either group. In the 2-
There were few adverse events in study eyes from either group iStent group, 2 subjects required subsequent glaucoma surgery
within the 12-month follow-up period, and there were no signifi- owing to uncontrolled IOP despite maximum medical therapy;
cant differences between groups (Table 4). In the 2-iStent group, there were no instances of incisional glaucoma surgery in the
there was 1 case of best-corrected visual acuity (BCVA) loss >2 Hydrus group, although there was 1 yttriumealuminumegarnet
lines and 4 cases of IOP elevation >10 mmHg over baseline. In the laser treatment for tissue adhesion near the device inlet.
Hydrus group, there were 2 cases of BCVA loss >2 lines and 3
cases of IOP elevation >10 mmHg. Device obstructions occurred
at similar rates between groups, although obstructions related to iris
Discussion

Table 3. Effectiveness Assessment This prospective, multicenter, randomized, single-masked

trial demonstrated an advantage in favor of the Hydrus
Hydrus 2 iStents Microstent over 2-iStent Trabecular Bypass devices in
N [ 73 N [ 75 P Value
reducing medication use and surgical success in OAG pa-
Patients on 0 medications at 12 months tients at 1 year postoperatively. Medication use was reduced
N (%) 34 (46.6%) 18 (24.0%) 0.006 by a greater margin (difference ¼ e0.6 medications/eye,
Mean (SD) IOP, mmHg 17.33.7 19.22.4 0.037 P ¼ 0.004) or eliminated completely more frequently in the
Mean (SD) DIOP, mmHg 8.23.7 5.12.9 0.003
 20% IOP reduction* 39.7% 13.3% <0.001
Hydrus group (46.6% vs. 24.0%, P ¼ 0.006) than in the
IOP 18 mmHg 30.1% 9.3% 0.002 2-iStent group. Among eyes without medications, Hydrus
achieved an IOP 18 mmHg more often (30.1% vs. 9.3%,
P <0.001). At 12 months, mean IOP was reduced in the
IOP ¼ intraocular pressure; SD ¼ standard deviation.
*Compared with washed-out baseline diurnal IOP.
Hydrus group concurrently with elimination of 1.6 medi-
cations; in the 2-iStent group IOP was maintained at

57
 Ophthalmology Volume 127, Number 1, January 2020

We observed substantially lower levels of IOP reduction

for the standalone 2-iStent procedure compared with a pre-
viously published series of single-center studies for the same
procedure33,35,36 These studies report 12-month mean IOPs in
the range of 13 to 14 mmHg, with and without medications at
12 months. Of these previous studies, 1 reported an IOP of
17.12.2 mmHg after medication washout at 12 months for
standalone treatment of OAG with 2 iStents.36 Treatment
efficacy rates >90% at 12 months were consistently
reported among these studies, whether defined as 20%
decrease in IOP or IOP 18 mmHg with or without
medications. Although MIGS devices are thought to lower
IOP more when combined with cataract surgery,37 the
previously reported standalone findings are superior to the
results for any published combined study using 1 or more
Figure 4. Twelve-month KaplaneMeier event-free survival: complete iStent devices. The 12-month IOP and medication counts
success. Failure was defined as any secondary glaucoma surgery, intraocular reported in these studies are consistent with published out-
pressure (IOP) >18 mmHg, or use of hypotensive medications on 2 comes for filtration surgery.38
consecutive visits after the 1-month follow-up visit. Patients lost to follow-up The differences in outcome between our study and pre-
were censored. viously published results do not appear to be related to
demographics or inclusion criteria. Although our study was
ethnically heterogeneous and multicenter, whereas the
preoperative levels concurrently with reduction of 1.0
aforementioned studies were conducted in a primarily white
medication.
population at a single center located in Armenia, multivar-
This is the first study to directly compare the efficacy of
iate analysis of the COMPARE study data showed ethnicity
different MIGS devices for use in glaucoma management
and site location did not affect our outcomes. Although
without concurrent cataract surgery. Multiple surgeons with
preoperative visual fields were not reported in the previous
significant prior experience with both devices participated in
studies, the average patient in the current study was using
the study. The study population was ethnically heteroge-
more medications compared with prior iStent studies, which
neous, had moderate disease severity, and was using mul-
were grouped by 0, 1, or 2 preoperative medications.
tiple medications to control IOP preoperatively. The study
Nonetheless, the mean washout IOPs were similar, and
groups were well matched for age, demographics, and key
multivariate analysis showed that the Hydrus group reached
baseline ocular characteristics such as IOP, medication use,
0 medications more frequently than the 2-iStent group.
BCVA, cup-to-disc ratio, and visual field defect.
Factors associated with reaching 0 medications included
This study suggests that trabecular MIGS devices may
lower baseline IOP and medication counts and for mild
play an important role in managing IOP and reducing the
disease severity as assessed by visual field. In addition,
need for hypotensive medications. In the 40% to 50% of
multivariate analysis did not show surgeon experience to be
glaucoma patients dependent on multiple medications to
a significant factor in the outcome.
control IOP, a procedure that reliably reduces medication
The only published report of double iStent standalone
number to 1 or 0 is desirable because patient adherence
surgery performed outside of the MIGS Study Group was a
drops to 44% when more than 1 medication is prescribed.35
10-patient pilot study conducted at a single center in
Poor adherence increases the risk of visual loss in glaucoma
Japan.39 In that study, the mean preoperative IOP was
patients.36
22.23.0 mmHg and each patient was on 3 medications
(prostaglandin analogues, b-blockers, and carbonic
Table 4. Safety Findings anhydrase inhibitors) per protocol throughout the 6-month
course of the study. At 6 months follow-up, mean IOP
Hydrus 2 iStents
N [ 74 N [ 76
was 16.93.6 mmHg on 3 medications, and the response
rate, defined as the percentage of eyes reaching 18 mmHg,
BCVA loss >2 lines at 12 months, n (%) 2 (2.7) 1 (1.3) was 44%. (We observed that 57.3% of 2-iStent eyes reached
IOP spike >10 mmHg, n (%) 3 (4.1) 4 (5.2)
this level on 1.6 medications.) Although this was a small
New cataract, n (%) 2 (2.6) 1 (1.3)
Device obstruction, n (%) series with limited follow-up, these results are more similar
Iris/other tissue 4 (5.4) 10 (13.2) to the 2-iStent results in our study than the response rates of
PAS 5 (6.8) 0 (0) 90% or more reported in the other cited studies.
Secondary surgical intervention, n (%) The clinical results of this study were predicted in pre-
Glaucoma surgery (trabeculectomy/GDD) 0 (0) 2 (2.6) clinical human cadaveric anterior segment perfusion
Stent migration/repositioning/removal 0 (0) 0 (0) models, which compared 2 iStents vs. the Hydrus device.
Cataract surgery 0 (0) 1 (1.3)
Hays et al24 showed that aqueous outflow facility (i.e.,
drainage) was increased by 74% by the Hydrus device
BCVA ¼ best-corrected visual acuity; GDD ¼ glaucoma drainage device; compared with 34% by 2 iStents. In prior cadaveric
IOP ¼ intraocular pressure; PAS ¼ peripheral anterior synechiae.
laboratory studies, the Hydrus device improved outflow

58
 Ahmed et al 
Randomized Comparison of Hydrus and iStent

facility and corresponded histologically to a dilated Acknowledgments

Schlemm canal and open and unobstructed collecting
The authors thank the COMPARE study investigators and clinical
channels.26,30 It is possible that stretching the TM by the research teams for their dedication and support.
Hydrus scaffold may be the basis of the improved aqueous
drainage, which was also seen with the 1-year results for
canaloplasty as a standalone procedure; the eyes that had
greater trabecular distension, as measured by high- References
resolution ultrasound biomicroscopy, had greater IOP
reduction compared with eyes that had minimal or no 1. Quigley HA, Broman AT. The number of people with glau-
trabecular distension (40% vs. 24%, respectively), even if coma worldwide in 2010 and 2020. Br J Ophthalmol. 2006;90:
there had been successful intracanalicular suture place- 262e267.
ment.40 TM stretch is one of the mechanisms of action of 2. Heijl A, Leske MC, Bengtsson B, et al. Reduction of intra-
pilocarpine.41,42 In live mice, which are the most robust ocular pressure and glaucoma progression: results from the
models aside from certain monkey species to study outflow, Early Manifest Glaucoma Trial. Arch Ophthalmol. 2002;120:
pilocarpine-induced stretch prevents the collapse of the 1268e1279.
3. American Academy of Ophthalmology Glaucoma Panel.
Schlemm canal in situations of elevated IOP.43 It is possible
Preferred Practice Pattern Guidelines. Primary Open Angle
that it is the TM stretch, induced by the Hydrus and in those Glaucoma. San Francisco, CA: American Academy of
cases of canaloplasty, that is preventing collapse of the Ophthalmology; 2010.
Schlemm canal, leading to the observed superiority of 4. Lichter PR, Musch DC, Gillespie BW, et al. Interim clinical
outflow facility in experimental studies and the clinical outcomes in the Collaborative Initial Glaucoma Treatment
results of this investigation. Study comparing initial treatment randomized to medications
The safety observations in this study are of particular or surgery. Ophthalmology. 2001;108:1943e1953.
importance given the generally moderate disease severity in 5. The Advanced Glaucoma Intervention Study (AGIS): 7. The
the study cohort. There was no difference in visual acuity relationship between control of intraocular pressure and visual
between the 2 study groups throughout the 12-month post- field deterioration. The AGIS Investigators. Am J Ophthalmol.
2000;130:429e440.
operative follow-up period. We observed 2 cases of sec-
6. Tsai JC, McClure CA, Ramos SE, et al. Compliance barriers in
ondary glaucoma surgery in the 2-iStent group, 1 of which glaucoma: a systematic classification. J Glaucoma. 2003;12:
was combined with cataract surgery; otherwise, ocular 393e398.
adverse findings in both groups were similar in frequency to 7. Tsai JC. A comprehensive perspective on patient adherence to
those reported for cataract surgery and were mild and topical glaucoma therapy. Ophthalmology. 2009;116:S30eS36.
transient. Typical safety risks for traditional transscleral 8. Tsai T, Robin AL, Smith 3rd JP. An evaluation of how
glaucoma surgery, such as hypotony, significant vision loss, glaucoma patients use topical medications: a pilot study. Trans
infections, and bleb-related complications, were absent from Am Ophthalmol Soc. 2007;105:29e35.
both treatment groups, as expected for a MIGS procedure. 9. Baudouin C, Renard JP, Nordmann JP, et al. Prevalence and risk
We did observe new cataracts in 3 subjects (2 in the Hydrus factors for ocular surface disease among patients treated over the
long term for glaucoma or ocular hypertension. Eur J Oph-
group and 1 in the 2-iStent group), which may or may not
thalmol. 2012 Jun 11;0. https://doi.org/10.5301/ejo.5000181.
have been related to the study procedure. Additional follow- 10. Broadway DC, Chang LP. Trabeculectomy, risk factors for
up is ongoing. failure and the preoperative state of the conjunctiva.
J Glaucoma. 2001;10:237e249.
11. Lichter PR, Musch DC, Gillespie BW, et al; CIGITS Study
Limitations Group. Interim clinical outcomes in the Collaborative Initial
This study was powered to detect differences in efficacy and Glaucoma Treatment Study comparing initial treatment ran-
although safety outcomes were carefully recorded, the domized to medications or surgery. Ophthalmology. 2001;108:
1943e1953.
sample size is too small to fully evaluate safety differences. 12. Samples JR, Ahmed IK, eds. Surgical Innovations in Glau-
The reluctance of investigators to conduct 12-month coma. USA: Springer; 2013.
washout in a high proportion of 2-iStent eyes was a devia- 13. Spiegel D, Kobuch K. Trabecular meshwork bypass tube shunt:
tion from the original study design, and it led to a protocol initial case series. Br J Ophthalmol. 2002;86:1228e1231.
change that eliminated the ability to directly compare 14. Hoeh H, Ahmed IK, Grisanti S, et al. Early postoperative
device-related IOP reductions and limits our ability to reach safety and surgical outcomes after implantation of a supra-
definitive conclusions about the efficacy of the 2 devices. choroidal micro-stent for the treatment of open-angle glau-
Patient follow-up is limited to 12 months, although extended coma concomitant with cataract surgery. J Cataract Refract
follow-up is ongoing. The study population was limited to Surg. 2013;39:431e437.
subjects with elevated washed-out IOP. Although the study 15. Sheybani A, Dick B, Ahmed IK. Early clinical results of a
novel ab interno gel stent for the surgical treatment of open-
incorporated design elements intended to minimize bias, the angle glaucoma. J Glaucoma. 2016;25:e691ee696.
investigator at each study site was not masked to treatment 16. Jea SY, Mosaed S, Vold SD, Rhee DJ. Effect of a failed tra-
randomization during follow-up examinations. Finally, bectome on subsequent trabeculectomy. J Glaucoma.
although balanced evenly between groups by randomiza- 2012;21(2):71e75.
tion, the study was not large enough to determine the in- 17. Samuelson TW, Katz LJ, Wells JM, et al. Randomized evaluation
fluence of lens status or prior SLT on study outcomes. of the trabecular micro-bypass stent with phacoemulsification in

59
 Ophthalmology Volume 127, Number 1, January 2020

patients with glaucoma and cataract. Ophthalmology. 2011;118: trabecular micro-bypass stents in open angle glaucoma. Clin
459e467. Ophthalmol. 2015;9:2057e2065.
18. Pfeiffer N, Garcia Feijoo JG, Martinez JM, et al. A randomized 32. Vold SD, Voskanyan L, Tetz M, et al. Newly diagnosed pri-
trial of a Schlemm’s canal microstent with phacoemulsification mary open-angle glaucoma randomized to 2 trabecular bypass
for reduction of intraocular pressure in open-angle glaucoma. stents or prostaglandin: outcomes through 36 months. Oph-
Ophthalmology. 2015;122:1283e1293. thalmol Ther. 2016;5(2):161e172.
19. Vold S, Ahmed I, Craven ER, et al. Two-year COMPASS trial 33. Katz LJ, Erb C, Guillamet AC, et al. Prospective, randomized
results: supraciliary microstenting with phacoemulsification in study of one, two, or three trabecular bypass stents in open-
patients with open-angle glaucoma and cataracts. Ophthal- angle glaucoma subjects on topical hypotensive medication.
mology. 2016;123:2103e2112. Clin Ophthalmol. 2015;9:2313e2320.
20. Samuelson TW, Marquis R, Jampel H, et al. A Schlemm’s 34. Ahmed IK, Katz LJ, Chang DF, et al. Prospective evaluation
canal microstent for intraocular pressure reduction in primary of microinvasive glaucoma surgery with trabecular micro-
open angle glaucoma and cataract: the HORIZON trial. bypass stents and prostaglandin in open-angle glaucoma.
Ophthalmology. 2019;126(1):29e37. J Cataract Refract Surg. 2014;40:1295e1300.
21. Grant WM. Further studies on facility of flow through the 35. Siani SD, Schoenfeld P, Kaulback K, Dubinsky MC. Effect of
trabecular meshwork. Arch Ophthalmol. 1958;60:523e533. dosing frequency on adherence in chronic diseases. Am J
22. Maepea O, Bill A. Pressures in the juxtacanalicular tissue and Manag Care. 2009;15(6):22e33.
Schlemm’s canal in monkeys. Exp Eye Res. 1992;54: 36. Sleath B, Blalock S, Covert D, et al. The relationship between
879e883. glaucoma medication adherence, eye drop technique, and visual
23. Seiler T, Wollensak J. The resistance of the trabecular mesh- field defect severity. Ophthalmology. 2011;118(12):2398e2402.
work to aqueous humor outflow. Graefes Arch Clin Exp 37. Singh K. Things go better with cataract surgery. Ophthal-
Ophthalmol. 1985;223:88e91. mology. 2014;121(1).
24. Hays CI, Gulati V, Toris C. Improvement in outflow facility 38. Gedde SJ, Schiffman JC, Feuer WJ, et al. Three-year follow-
by two novel microinvasive glaucoma surgery implants. Invest up of the Tube versus trabeculectomy study. Am J Oph-
Ophthalmol Vis Sci. 2014;55:1893e1900. thalmol. 2009;148:670e684.
25. Gordon MO, Kass MA. The Ocular Hypertension Treatment 39. Shiba D, Hosoda S, Yaguchi S, et al. Safety and efficacy of
Study: design and baseline description of the participants. Arch two trabecular micro-bypass stents as the sole procedure in
Ophthalmol. 1999;117:573e583. Japanese patients with medically uncontrolled primary open
26. Buznego C. Trabecular micro-bypass stent for glaucoma. angle glaucoma: a pilot case series. J Ophthalmol. 2017;2017:
Techniques in Ophthalmology. 2009;7:21e24. 9605461.
27. Fea AM, Ahmed IK, Lavia C, et al. Hydrus microstent 40. Lewis RA, von Wolf K, Tetz M, et al. Canaloplasty: circum-
compared to selective laser trabeuloplastly in primary open ferential viscodilation and tensioning of Schlemm’s canal us-
angle glaucoma: one year results. Clin Exp Ophthalmol. ing a flexible microcatheter for the treatment of open-angle
2017;45:120e127. glaucoma in adults: interim clinical study analysis. J Catract
28. Gandolfi SA, Ungaro N, Ghirardini S, et al. Comparison of Refract Surg. 2007;33:1217e1266.
surgical outcomes between canaloplasty and a Schlemm’s 41. Skaat A, Rosman MS, Park SC, et al. Effect of pilocarpine
canal scaffold at 24 months follow-up. J Ophthalmol. hydrochloride on the Schlemm canal in healthy eyes and eyes
2016;2016:3410469. with open-angle glaucoma. JAMA Ophthalmol. 2016;134:
29. Bahler C, Smedley G, Zhou J, Johnson D. Trabecular bypass 976e981.
stents decrease intraocular pressure in cultured human anterior 42. Grierson I, Lee WR, Abraham S. Effects of pilocarpine on the
segments. Am J Ophthalmol. 2004;138:988e994. morphology of the human outflow apparatus. Br J Oph-
30. Belovay GW, Naqi A, Chan BJ, et al. Using multiple trabec- thalmol. 1978;62:302e313.
ular micro-bypass stents in cataract patients to treat open-angle 43. Li G, Farsui S, Chiu SJ, et al. Pilocarpine-induced dilation of
glaucoma. J Cataract Refract Surg. 2012;38(11):1911e1917. Schlemm’s canal and prevention of lumen collapse at elevated
31. Donnenfeld ED, Solomon KD, Voskanyan L, et al. intraocular pressures in living mice visualized by OCT. Invest
A prospective 3-year follow up trial of implantation of two Ophthalmol Vis Sci. 2014;55:3737e3746.

Footnotes and Financial Disclosures

9
Originally received: June 27, 2018. University Hospitals, Case Western Reserve University, Cleveland, Ohio.
Final revision: April 4, 2019. Financial Disclosure(s):
Accepted: April 19, 2019. The author(s) have made the following disclosure(s): R.E.A.: Support e
Available online: April 26, 2019. Manuscript no. 2018-1478. Ivantis; Consulting, honoraria, and travel expenses e Glaukos; Study
1
University of Toronto, Toronto, Canada. grants, honoraria, and travel expenses e Ivantis.
2
Dipartimento di Scienze Chirurgiche, Universita’ di Torino, Torino, Italy. D.J.R.: Study grants e Ivantis.
3
Manchester Royal Eye Hospital, Manchester Academic Health Science T.W.S.: Consulting fees e Ivantis, Glaukos.
Centre, Manchester, United Kingdom. D.F.C.: Honoraria for public speaking e Ivantis
4
Asian Eye Institute, Makati City, Philippines. P.H.: Grants, honoraria, and travel expenses e Ivantis, Inc; Honoraria e
5 Glaukos.
University of Montreal, Bellevue Ophthalmology Clinic, Montreal,
Canada. I.I.K.A.: Support e Ivantis, Glaukos, Alcon.
6
Wilmer Eye Institute, Johns Hopkins University, Baltimore, Maryland. H.J.: Consulting e Ivantis.
7 A.F.: Study grants e Ivantis; Consulting e Glaukos.
Minnesota Eye Consultants, Minneapolis, Minnesota.
8
Altos Eye Physicians, Los Altos, California.

60
 Ahmed et al 
Randomized Comparison of Hydrus and iStent
L.A.: Grants and consulting fees e Ivantis; Investigation and honoraria for Obtained funding: Ahmed, Fea, Au, Ang, Harasymowycz, Jampel,
public lectures e Glaukos. Samuelson, Chang, Rhee
Supported by Ivantis, Inc, Irvine, CA. Overall responsibility: Ahmed, Fea, Au, Harasymowycz, Jampel, Samuel-
HUMAN SUBJECTS: Human subjects were included in this study. The son, Chang, Rhee
study protocol was approved by the Medical Ethics Committees at each site. Abbreviations and Acronyms:
The study was conducted according to the principles described in the BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval;
Declaration of Helsinki and all study subjects provided written informed DIOP ¼ diurnal intraocular pressure; IOP ¼ intraocular pressure;
consent before participation in the trial. The study was registered in the MIGS ¼ microinvasive glaucoma surgery; OAG ¼ open-angle glaucoma;
National Library of Medicine database (clinicaltrials.gov NCT02023242). OR ¼ odds ratio; SLT ¼ selective laser trabeculoplasty; TM ¼ trabecular
No animal subjects were used in this study. meshwork.
Author Contributions: Correspondence:
Conception and design: Ahmed, Fea, Jampel, Samuelson, Rhee Douglas J. Rhee, MD, Case Western Reserve University, UH Hospitals
Analysis and interpretation: Ahmed, Fea, Jampel, Samuelson, Chang, Rhee Cleveland, 11100 Euclid Ave, Cleveland, OH 44106. E-mail: dougrhee@
Data collection: Fea, Au, Ang, Harasymowycz aol.com.

Pictures & Perspectives

Intracorneal Migration of Silicon Band

A 40-year-old-man presented with a large ciliary staphyloma
due to secondary glaucoma in his left eye. The patient had
undergone scleral buckling surgery with intravitreal per-
fluoropropane gas injection for rhegmatogenous retinal detach-
ment 6 years previously. Anterior-segment slit-lamp (Fig A and
B) examination showed intracorneal migration of silicone band
(Fig A and B). Infrared imaging (Fig C) showing the linear area
through which anterior-segment OCT (AS-OCT) is captured.
The AS-OCT imaging (Fig D) showed extension of the band into
the cornea. Intracorneal migration of silicone band may be
attributed to ciliary staphyloma resulting from raised intraocular
pressure leading to enlargement of eyeball along with scleral
thinning. (Magnified version of Fig 1A-D is available online at
www.aaojournal.org).

VINOD KUMAR, MD1

ABHIDNYA SURVE, MD1
1
Vitreo-retina, Trauma & Uvea Services, Dr Rajendra Prasad
Centre for Ophthalmic Sciences, All India Institute of Medical
Sciences, Ansari Nagar, New Delhi, India

61
Comparison of Associations with Different
Macular Inner Retinal Thickness Parameters
in a Large Cohort
The UK Biobank
Anthony P. Khawaja, PhD, FRCOphth,1 Sharon Chua, PhD,1 Pirro G. Hysi, PhD,2,3
Stelios Georgoulas, FRCOphth,1 Hannah Currant, BSc,4 Tomas W. Fitzgerald, PhD,4 Ewan Birney, PhD,4
Fang Ko, MD,1 Qi Yang, PhD,5 Charles Reisman, MSc,5 David F. Garway-Heath, MD, FRCOphth,1
Chris J. Hammond, FRCOphth,2 Peng T. Khaw, PhD,1 Paul J. Foster, PhD,1
Praveen J. Patel, FRCOphth, MD(Res),1 Nicholas Strouthidis, PhD,1,6 for the UK Biobank Eye and Vision
Consortium

Purpose: To describe and compare associations with macular retinal nerve fiber layer (mRNFL), ganglion cell
complex (GCC), and ganglion celleinner plexiform layer (GCIPL) thicknesses in a large cohort.
Design: Cross-sectional study.
Participants: We included 42 044 participants in the UK Biobank. The mean age was 56 years.
Methods: Spectral-domain OCT macular images were segmented and analyzed. Corneal-compensated
intraocular pressure (IOPcc) was measured with the Ocular Response Analyzer (Reichert, Corp., Buffalo, NY).
Multivariable linear regression was used to examine associations with mean mRNFL, GCC, and GCIPL thick-
nesses. Factors examined were age, sex, ethnicity, height, body mass index (BMI), smoking status, alcohol
intake, Townsend deprivation index, education level, diabetes status, spherical equivalent, and IOPcc.
Main Outcome Measures: Thicknesses of mRNFL, GCC, and GCIPL.
Results: We identified several novel independent associations with thinner inner retinal thickness. Thinner
inner retina was associated with alcohol intake (most significant for GCIPL: e0.46 mm for daily or almost daily
intake compared with special occasion only or never [95% confidence interval (CI), 0.61e0.30]; P ¼ 1.110e8),
greater social deprivation (most significant for GCIPL: e0.28 mm for most deprived quartile compared with least
deprived quartile [95% CI, e0.42 to e0.14]; P ¼ 6.610e5), lower educational attainment (most significant for
mRNFL: e0.36 mm for less than O level compared with degree level [95% CI, e0.45 to 0.26]; P ¼ 2.310e14), and
nonwhite ethnicity (most significant for mRNFL comparing blacks with whites: e1.65 mm [95% CI, e1.86 to
e1.43]; P ¼ 2.410e50). Corneal-compensated intraocular pressure was associated most significantly with
GCIPL (e0.04 mm/mmHg [95% CI, e0.05 to e0.03]; P ¼ 4.010e10) and was not associated significantly with
mRNFL (0.00 mm/mmHg [95% CI, e0.01 to 0.01]; P ¼ 0.77). The variables examined explained a greater
proportion of the variance of GCIPL (11%) than GCC (6%) or mRNFL (7%).
Conclusions: The novel associations we identified may be important to consider when using inner retinal
parameters as a diagnostic tool. Associations generally were strongest with GCIPL, particularly for IOP. This
suggests that GCIPL may be the superior inner retinal biomarker for macular pathophysiologic processes and
especially for glaucoma. Ophthalmology 2020;127:62-71 ª 2019 by the American Academy of Ophthalmology

Supplemental material available at www.aaojournal.org.

Damage to macular retinal ganglion cells occurs early in plexiform layer (GCIPL; GCL þ IPL). Both GCC and
glaucoma,1 and spectral-domain (SD) OCT measurements GCIPL thickness have been reported to be comparable with
of the inner retina at the macula have been shown to be circumpapillary retinal nerve fiber layer (cRNFL) thickness
useful for detecting glaucoma.2,3 Different commercially at diagnosing glaucoma.4,5 Macular GCC and GCIPL
available SD OCT devices report different segments of inner measurements have been shown to be helpful in the detec-
retinal macular thickness; commonly reported segments are tion of glaucoma progression6,7 and may be superior to
the ganglion cell complex (GCC; macular retinal nerve fiber cRNFL measurements at detecting progression in severe
layer [mRNFL] þ ganglion cell layer [GCL] þ inner disease.8e10 A meta-analysis reports similar accuracy of
plexiform layer [IPL]) and the ganglion celleinner GCC and GCIPL measurements for glaucoma diagnosis,11

62 ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.08.015

Published by Elsevier Inc. ISSN 0161-6420/19
 Khawaja et al 
Macular Inner Retinal Morphology in UK Biobank

which is in agreement with studies that conducted head-to- the inner and outer retinal surfaces.15 Quality control to exclude
head comparisons of GCC and GCIPL diagnostic accuracy images of poor quality was described in detail previously.16 We
within the same study participants.12e14 excluded scans with an image quality score (signal strength) less
Understanding epidemiologic associations with macular than 45. Additionally, several segmentation indicators were
calculated that also identified poor scan quality or segmentation
inner retinal measurements is important to help define
failures; we excluded the poorest 20% of images for each of
normal ranges in population subgroups and may shed light these indicators. The inner limiting membrane indicator was a
on pathophysiologic mechanisms underlying glaucoma. measure of the minimum localized edge strength around the
Comparing strengths of associations among mRNFL, GCC, inner limiting membrane boundary across the entire scan; this is
and GCIPL measurements may provide insight into their useful for identifying blinks, scans that contain regions of severe
relative potential as biomarkers. Using data from a very signal fading, and segmentation errors. The validity count
large adult cohort, the UK Biobank, we aimed to describe indicator is used to identify scans with a significant degree of
and compare basic demographic, socioeconomic, anthro- clipping in the OCT scan’s z-axis dimension. The motion
pometric, lifestyle, and ocular associations with mRNFL, indicators use both the nerve fiber layer and the full retinal
GCC, and GCIPL. thicknesses, from which Pearson correlations and absolute
differences between the thickness data from each set of
consecutive B-scans are calculated. The lowest correlation and
Methods the highest absolute difference in a scan serve as the resulting
indicator scores and identify blinks, eye motion artifacts, and
UK Biobank segmentation failures. The image quality score and the
aforementioned indicators usually are highly correlated. We used
The UK Biobank is a very large multisite cohort study established average thickness parameters derived from the macula 6 grid.
by the Medical Research Council, Department of Health, Well- Participant-level mRNFL, GCC, and GCIPL thicknesses (in mi-
come Trust medical charity, Scottish Government, and Northwest crometers) were calculated as the mean of right and left eye values
Regional Development Agency. Detailed study protocols are for each participant with good-quality images available for both
available online (http://www.ukbiobank.ac.uk/resources/ and eyes. If data were available only for 1 eye, we considered that value
http://biobank.ctsu.ox.ac.uk/crystal/docs.cgi). A baseline ques- for the participant.
tionnaire, physical measurements, and biological samples were Participant intraocular pressure (IOP; in millimeters of mer-
undertaken in 22 assessment centers across the United Kingdom cury) was measured once for each eye using the Ocular Response
between 2006 and 2010. All United Kingdom residents 40 to 69 Analyzer (ORA; Reichert, Corp., Buffalo, NY). Participants who
years of age who were registered with the National Health Service reported eye surgery within the previous 4 weeks or participants
and living up to 25 miles from a study center were invited to reporting an eye infection were precluded from having IOP
participate. The study was conducted with the approval of the measured. The ORA is a noncontact tonometer that measures the
North-West Research Ethics Committee (reference, 06/MRE08/ force required to flatten the cornea using a jet of air. Unlike
65), in accordance with the principles of the Declaration of Hel- conventional noncontact tonometry, the ORA measures 2 pres-
sinki, and all participants gave written informed consent. This sures: first, when the cornea flattens on inward motion, and
research has been conducted using the UK Biobank Resource second, when the cornea is flattened on outward motion. The
under application number 2112. average of these 2 pressures has been calibrated to derive a
Participants completed a touch-screen self-administered Goldmann-correlated IOP (IOPg), and the difference between
questionnaire and underwent physical examination at a baseline these 2 pressures has been shown to be related to the biome-
assessment. Table 1 summarizes the ascertainment of the chanical properties of the cornea.17 A linear combination of these
baseline assessment variables used in the current study. Body 2 pressures has been developed to derive a corneal-compensated
mass index (BMI) was calculated as weight in kilograms per IOP (IOPcc).18 We used IOPcc for our primary analyses because
height in square meters. We selected these variables a priori to it is thought to provide the most accurate assessment of true IOP
examine the basic descriptive epidemiologic features of inner and to be least affected by corneal properties.19 To handle
retinal morphologic characteristics, including demographic, extreme values of IOP, we excluded the top and bottom 0.5%
socioeconomic, anthropometric, and basic lifestyle factors. We of IOP measurements. We excluded participants with a history
additionally examined diabetes status as a potentially important of laser or surgery for glaucoma, eye injury, corneal graft
confounder, given that diabetes is relatively common and has surgery, or refractive laser surgery because these participants
known retinal sequalae. are likely to have IOP that has been altered from physiologic
levels. For patients using IOP-lowering medication (n ¼ 1151),
Ophthalmic Assessment we imputed pretreatment IOP by dividing by 0.7 based on the
mean IOP reduction achieved by medication.20 This approach has
Ophthalmic assessment was not part of the original baseline been used successfully in genome-wide association studies of
assessment and was introduced as an enhancement in 2009 for 6 IOP.21,22 Additionally, in sensitivity analyses, we used IOPg with
assessment centers that are spread across the United Kingdom imputed pretreatment IOP and also IOPg and IOPcc after
(Liverpool and Sheffield in North England, Birmingham in the exclusion of participants using IOP-lowering medication.
Midlands, Swansea in Wales, and Croydon and Hounslow in Refractive status of both eyes was measured by autorefraction
greater London). Spectral-domain OCT imaging of both eyes was (Tomey RC5000; Erlangen-Tennenlohe). Spherical equivalent
performed using the Topcon 3D OCT-1000 Mark II (Topcon, Inc., was calculated as the sphere þ 0.5  cylinder. We excluded
Tokyo, Japan) in a dark room without pupil dilation using the 3- participant eyes with high refractive error (<e6 diopters [D] or
dimensional 66-mm2 macular volume scan mode (512 A scans >þ6 D). We calculated participant-level IOP and spherical
per B scan; 128 horizontal B scans in a raster pattern). The right equivalent as the mean of right and left eye values if data were
eye was imaged first. Version 1.6.1.1 of the Topcon Advanced available for both or as either the right or left eye value if data
Boundary Segmentation (TABS) algorithm was used to delineate were available for only 1 eye.

63
 Ophthalmology Volume 127, Number 1, January 2020

Table 1. Baseline Assessment Details for Variables Analyzed in the Current Study

Variable Assessment Details

Touch-screen self-administered questionnaire
Ethnicity Response options included: white (English/Irish or other white background), Asian or British Asian (Indian/Pakistani/
Bangladeshi or other Asian background), black or black British (Caribbean, African, or other black background),
Chinese, mixed (white and black Caribbean or African, white and Asian, or other mixed background), or other
ethnic group (not defined).
Smoking status Determined by response to “Do you smoke tobacco now?” and “In the past, how often have you smoked tobacco?”
Alcohol intake Determined by response to “About how often do you drink alcohol?” (options: daily or almost daily, 3 or 4 times a week,
once or twice weekly; 1 to 3 times monthly; special occasions only; never).
Townsend deprivation Determined according to participant post code at recruitment and the corresponding output area from the preceding
index national census. The index was calculated based on the output area’s employment status, home and car ownership,
and household condition; the higher and more positive the index, the more deprived an area.
Education level Determined by response to “Which of the following qualifications do you have (you can select more than one)?”
(options: college or university degree; A levels/AS levels or equivalent; O levels/GCSEs or equivalent; CSEs or
equivalent; NVQ or HND or HNC or equivalent; other professional qualifications, e.g., nursing, teaching; none of the
above). For analyses, qualifications lower than “O level/GCSEs or equivalent” were considered together as “less than
O level.”
Diabetes status Determined as those who answered yes to “Has a doctor ever told you that you have diabetes?”
Physical measures
Weight BV-418 MA body composition analyzer (Tanita, Arlington Heights, IL)
Height Seca 202 height measure (Seca, Birmingham, United Kingdom)

A level ¼ Advanced level; AS level ¼ Advanced subsidiary level; CSE ¼ Certificate of Secondary Education; GCSE ¼ General Certificate of Secondary
Education; HNC ¼ Higher National Certificate; HND ¼ Higher National Diploma; NVQ ¼ National Vocational Qualification.

Statistical Analyses Results

Demographic, systemic, and ocular characteristics for included In total, SD OCT images from 67 310 UK Biobank participants
participants were described, stratified by sex. Comparisons be- were available at the time of this analysis. After image segmen-
tween men and women for each of the variables were made using tation and quality control, 45 815 participants with mRNFL, GCC,
the independent sample t test for continuous variables and the chi- and GCIPL thickness measurements remained. Complete data were
square test for categorical variables. We first examined crude as- available for all exposure (age, sex, ethnicity, weight, height, BMI,
sociations with mRNFL, GCC, and GCIPL thicknesses using smoking status, alcohol intake, deprivation score, diabetes status,
univariable linear regression. Variables significantly associated education level, spherical equivalent, IOP) and retinal thickness
with any of mRNFL, GCC, or GCIPL thicknesses at a P < 0.01 variables for 42 044 participants; all analyses were conducted using
level then were considered together in a multivariable linear these participants. The mean age of included participants was 56
regression model for each of mRNFL, GCC, and GCIPL thickness. years and 53% were women. Table 2 summarizes mean mRNFL,
Given that weight and BMI are correlated highly, we considered GCC, and GCIPL thicknesses as well as demographic, systemic,
only 1 of the parameters in multivariable analyses; we selected and ocular factors for included participants.
BMI based on stronger univariable associations. Similarly, given Univariable associations with mRNFL, GCC, and GCIPL
that IOPg and IOPcc are correlated highly, we considered only 1 of thicknesses are shown in Table 3. Significant associations were
the parameters in multivariable analyses; IOPcc was selected on the found with at least 1 of mRNFL, GCC, and GCIPL thicknesses
basis that it better reflects true physiologic IOP.19 To determine for all examined variables except for height, which was not
whether the associations we identified were driven primarily by associated significantly with any thickness parameter (all P >
participants with established glaucoma, we carried out the same 0.30; Table 3). Therefore, height was not carried forward for the
multivariable analyses after exclusion of participants with self- multivariable analyses. Both weight and BMI were associated
reported (touch-screen questionnaire) or hospital admission coded significantly with all 3 thickness parameters; given their
(International Classification of Disease, Tenth Edition) glaucoma, collinearity, only BMI was carried forward for multivariable
and excluding any participants using IOP-lowering medication analyses, as described in Methods. Both IOPg and IOPcc were
(n ¼ 41 449 after exclusion of 595 participants). We also con- associated significantly with GCC and GCIPL thicknesses (both
ducted sensitivity analyses for the associations of mRNFL, GCC, P < 0.001) but not with mRNFL thickness (both P  0.12).
and GCIPL thicknesses with IOP for primary analyses included Given the collinearity between IOPg and IOPcc, only IOPcc was
IOP measurements that were imputed for pretreatment levels in carried forward for multivariate analyses, as detailed in “Methods.”
participants using glaucoma medication. First, rather than imputing Multivariable associations with mRNFL, GCC, and GCIPL
pretreatment IOP, we conducted analyses using current IOP, even thicknesses are shown in Table 4. Age was associated strongly with
if the patient was using IOP-lowering medication, and additional a thinner mRNFL, GCC, and GCIPL; the association appeared
analyses excluding participants using IOP-lowering medication. stronger for GCC and GCIPL thickness than for mRNFL
Second, we conducted further analyses using IOPg rather than thickness. Related to the strength of association and the very
IOPcc. To examine how much of mRNFL, GCC, and GCIPL large sample size, the P values for associations with age were
thickness variances are explained by the factors we examined, we extremely small; for GCC and GCIPL thickness, the P values
calculated R2 statistics for the multivariate regression models. Stata were smaller than can be handled by most modern statistical
software version 15.1 (StataCorp LP, College Station, TX) was software (P < 10e300). Men had significantly thinner mRNFL
used for all statistical analyses. and GCC than women (both P  7.110e23) and thinner GCIPL

64
 Khawaja et al 
Macular Inner Retinal Morphology in UK Biobank

Table 2. Demographic, Systemic, and Ocular Characteristics for with self-reported diabetes had thinner mRNFL, GCC, and
Included Participants (n ¼ 42 044) GCIPL (all P < 0.003). There were very strong and highly sig-
nificant associations between thickness parameters and spherical
Characteristic Data equivalent, and these were in different directions for mRNFL
compared with GCC and GCIPL. A more myopic refraction was
Age, mean (SD) 56.4 (8.1)
associated with a thicker mRNFL (P ¼ 1.110e251) but a thinner
Gender, no. (%)
Female 22 332 (53)
GCC (P ¼ 1.210e93) and GCIPL (P < 10e300). Corneal-
Male 19 712 (47) compensated IOP was not associated with mRNFL but was asso-
Ethnicity, no. (%) ciated negatively with both GCC (P ¼ 5.810e5) and GCIPL (P ¼
White 38 509 (92) 4.0  10e10) thickness. Of the 3 multivariable models, the R2 was
Asian 1163 (3) greatest for the GCIPL model, indicating that the explanatory
Black 1234 (3) variables we assessed explained more of the variance of GCIPL
Other/mixed 1138 (3) thickness (11%) than mRNFL (7%) or GCC (6%) thickness
Weight (kg), mean (SD) 78.1 (15.3) (Table 4). The same multivariable analyses also were conducted
Height (cm), mean (SD) 169.0 (9.2) after exclusion of participants using glaucoma medication, self-
Body mass index (kg/m2), mean (SD) 27.3 (4.4) reported glaucoma, or both or hospital International Classifica-
Smoking status, no. (%) tion of Disease, Tenth Edition, coded glaucoma (Table S1,
Never 23 052 (55) available at www.aaojournal.org). Associations were very similar
Previous 14 869 (35) for all variables apart from IOP, for which the associations were
Current 4123 (10) less significant. There was no longer a significant association
Alcohol intake between IOP and GCC, and the association between IOP and
Never or special occasion only 8237 (20) GCIPL was less significant (P ¼ 3.910e5).
1e3 times per month 4771 (11) We also conducted sensitivity analyses for the associations
1e2 times per week 10 557 (25)
between IOP and inner retinal thickness measures, as described in
3e4 times per week 9610 (23)
Methods. Results were similar when we examined IOPg instead of
Daily or almost daily 8869 (21)
Townsend deprivation index, no. (%)
IOPcc (Table 5). Also, results were similar if we either excluded
Least deprived quartile 10 721 (25) participants using IOP-lowering medication or used current
Second quartile 10 748 (26) treated IOP, rather than imputing the pretreatment IOP (Table 5).
Third quartile 10 487 (25) For all analyses, IOP was not associated with mRNFL and was
Most deprived quartile 10 088 (24) associated more significantly with GCIPL than GCC. Again, the
Education level, no. (%) model R2 value was greatest for the GCIPL models (Table 5).
Less than O level 13 215 (31)
O level 8919 (21)
A Level 5007 (12) Discussion
Degree 14 903 (35)
Diabetes (self-report), no. (%) Our study is the largest to date examining the epidemiologic
No 39 955 (95)
Yes 2089 (5)
features of macular inner retinal anatomic characteristics, to
Spherical equivalent (diopters), mean (SD) 0.001 (1.9) the best of our knowledge. We confirmed previously re-
IOPg (mmHg), mean (SD) 15.66 (3.59) ported associations with age, sex, BMI, diabetes, and
IOPcc (mmHg), mean (SD) 15.83 (3.60) refractive error and identified multiple novel associations
mRNFL (mm), mean (SD) 28.91 (3.83) with thinner inner retina at the macula, including nonwhite
GCC (mm), mean (SD) 104.15 (7.56) ethnicity, frequent alcohol intake, greater social deprivation,
GCIPL (mm), mean (SD) 75.24 (5.19)
lower educational attainment, and higher IOP. Our study
also examines how epidemiologic associations vary between
GCC ¼ ganglion cell complex; GCIPL ¼ ganglion celleinner plexiform different inner retinal parameters, namely mRNFL, GCC,
layer; IOPcc ¼ corneal-compensated intraocular pressure; IOPg ¼ and GCIPL thickness.
Goldmann-correlated intraocular pressure; mRNFL ¼ macular retinal
nerve fiber layer; SD ¼ standard deviation. Older age was associated strongly with thinner inner
retinal thickness, in agreement with previous studies.23e26
This association was apparent for all 3 inner retinal pa-
of borderline significance (P ¼ 0.042). Asian and black rameters but was strongest and most significant for GCC
participants showed thinner mRNFL, GCC, and GCIPL than and GCIPL thickness. Although it is not possible to infer a
white participants. Participants with higher BMI had thinner causal effect of inner retinal thinning because of aging from
mRNFL, GCC, and GCIPL (all P  1.510e8). Daily or almost a cross-sectional study, it is unlikely an association this
daily alcohol intake was associated with thinner mRNFL, GCC, strong is the result of a cohort effect. Comparing the age
and GCIPL when compared with participants who drank least coefficients (Table 4) with the mean thickness values in our
(never or special occasions only). There was no significant study (Table 2) derives a yearly percentage decline in
difference in thickness parameters for participants reporting less thickness of 0.14% for mRNFL, 0.18% for GCC, and
frequent alcohol intake (Table 4). Participants in the most
0.20% for GCIPL; this is also in keeping with previous
deprived quartile of the Townsend deprivation index had
significantly thinner GCC and GCIPL (both P  1.210e4) than studies.23e26
participants in the least deprived quartile; the difference was We found men to have thinner macular inner retinas, and
borderline significant only for mRNFL thickness (P ¼ 0.012). this was most apparent for mRNFL. Thinner GCIPL in men
There was evidence of progressively thicker mRNFL, GCC, and was reported previously in a multiethnic volunteer study of
GCIPL with higher educational attainment (Table 4). Participants 282 healthy participants.23 Other studies found no

65
 Ophthalmology Volume 127, Number 1, January 2020

Table 3. Univariable Associations with Average Macular Retinal Nerve Fiber Layer, Ganglion Cell Complex, and Ganglion CelleInner
Plexiform Layer Thicknesses

Macular Retinal Nerve Fiber Ganglion Cell Complex Ganglion CelleInner Plexiform
Layer Thickness (mm) Thickness (mm) Layer Thickness (mm)
b Coefficient b Coefficient b Coefficient
(95% Confidence Interval) P Value (95% Confidence Interval) P Value (95% Confidence Interval) P Value
Age (per decade) e0.57 (e0.61 to e0.52) <0.001 e1.73 (e1.82 to e1.64) <0.001 e1.17 (e1.22 to e1.10) <0.001
Gender
Female Reference Reference Reference
Male e0.70 (e0.77 to e0.63) <0.001 e1.00 (e1.14 to e0.85) <0.001 e0.30 (e0.39 to e0.19) <0.001
Ethnicity
White Reference Reference Reference
Asian e0.75 (e0.97 to e0.53) <0.001 e1.28 (e1.72 to e0.84) <0.001 e0.53 (e0.83 to e0.22) 0.001
Black e1.46 (e1.68 to e1.24) <0.001 e1.20 (e1.62 to e0.77) <0.001 0.27 (e0.03e0.55) 0.08
Other/mixed e0.21 (e0.43, 0.01) 0.07 0.64 (0.19e1.08) 0.005 0.84 (0.53e1.14) <0.001
Weight (per 10 kg) e0.12 (e0.14 to e0.09) <0.001 e0.21 (e0.25 to e0.15) <0.001 e0.08 (e0.11 to e0.05) <0.001
Height (per 10 cm) 0.02 (e0.02, 0.05) 0.37 0.04 (e0.03 to 0.11) 0.30 0.02 (e0.03 to 0.07) 0.40
Body mass index (per 5 kg/m2) e0.26 (e0.30 to e0.22) <0.001 e0.46 (e0.53 to e0.37) <0.001 e0.19 (e0.25 to e0.13) <0.001
Smoking status
Never Reference Reference Reference
Previous e0.22 (e0.29 to e0.13) <0.001 e0.45 (e0.60 to e0.29) <0.001 e0.24 (e0.34 to e0.13) <0.001
Current e0.28 (e0.40 to e0.15) <0.001 0.04 (e0.21 to 0.29) 0.75 0.32 (0.14e0.49) <0.001
Alcohol intake
Never or special occasion only Reference Reference Reference
1e3 times per month 0.26 (0.12e0.39) <0.001 0.42 (0.15e0.69) 0.002 0.16 (e0.02 to 0.34) 0.09
1e2 times per week 0.24 (0.12e0.34) <0.001 0.40 (0.18e0.61) <0.001 0.16 (0.01e0.31) 0.032
3e4 times per week 0.16 (0.04e0.27) 0.005 0.07 (e0.15 to 0.29) 0.54 e0.09 (e0.24 to 0.06) 0.25
Daily or almost daily e0.02 (e0.13 to 0.09) 0.68 e0.62 (e0.84 to e0.39) <0.001 e0.60 (e0.75 to e0.44) <0.001
Townsend deprivation index
Least deprived quartile Reference Reference Reference
Second quartile e0.06 (e0.16 to 0.04) 0.27 e0.03 (e0.23 to 0.17) 0.76 0.03 (e0.11 to 0.16) 0.71
Third quartile e0.08 (e0.18 to 0.02) 0.12 e0.10 (e0.30 to 0.10) 0.32 e0.02 (e0.16 to 0.11) 0.76
Most deprived quartile e0.20 (e0.30 to e0.09) <0.001 e0.19 (e0.39 to 0.01) 0.07 0.02 (e0.12 to 0.15) 0.83
Test of trend e0.06 (e0.09 to e0.03) <0.001 e0.06 (e0.12 to 0.00) 0.06 0.00 (e0.04 to 0.04) 0.99
Education level
Less than O level Reference Reference Reference
O level 0.47 (0.37e0.57) <0.001 0.69 (0.48e0.89) <0.001 0.22 (0.08e0.35) 0.002
A Level 0.85 (0.72e0.96) <0.001 1.11 (0.86e1.35) <0.001 0.26 (0.09e0.42) 0.003
Degree 0.83 (0.74e0.92) <0.001 0.99 (0.81e1.16) <0.001 0.16 (0.03e0.28) 0.010
Diabetes status
No Reference Reference Reference
Yes e0.95 (e1.12 to e0.78) <0.001 e1.92 (e2.25 to e1.59) <0.001 e0.97 (e1.19 to e0.74) <0.001
Spherical equivalent (diopters) e0.40 (e0.41 to e0.38) <0.001 0.20 (0.16e0.23) <0.001 0.60 (0.57e0.62) <0.001
IOPg (mmHg) e0.01 (e0.01 to 0.00) 0.15 e0.11 (e0.12 to e0.08) <0.001 e0.10 (e0.11 to e0.08) <0.001
IOPcc (mmHg) e0.01 (e0.01 to 0.00) 0.12 e0.11 (e0.13 to e0.09) <0.001 e0.10 (e0.11 to e0.09) <0.001

IOPcc ¼ corneal-compensated intraocular pressure; IOPg ¼ Goldmann-correlated intraocular pressure.

Results are from linear regression models (n ¼ 42 044). P values less than 0.0020 appear in boldface and represent significance at a Bonferroni-corrected
threshold for 25 statistical tests.

significant association between inner retinal thickness and similar strength for mRNFL, GCC, and GCIPL thicknesses,
sex,24,26 and 1 study from a subset of the Singapore Chi- suggesting the association is not related specifically to
nese Eye Study found women to have thinner inner ret- retinal ganglion cells. Also in agreement with this is the
inas.25 Although it is possible that the relationship between previously reported association of higher BMI with
sex and macular inner retinal thickness varies between thinner macular total retina thickness in the UK Biobank.16
populations, it is more likely that the variation in results is We observed thinner inner retinas in participants with
stochastic because of the smaller sample sizes and diabetes; this association was more significant for mRNFL
resultant statistical power of previous studies. Our finding thickness than for GCC or GCIPL thickness. This is in
of a thinner inner retina in men may be aligned with the agreement with small case-control studies that have reported
greater susceptibility to glaucoma reported in men.27 thinner inner retinas in participants with diabetes compared
Higher BMI was associated with thinner inner retina, in with controls28e30 and has led to the hypothesis that diabetic
agreement with a study of British twins that reported thinner peripheral neuropathy and inner retinal thinning may share
GCC with higher BMI.24 We observed the association with common biological pathways.31 Interestingly, laser

66
 Khawaja et al 
Macular Inner Retinal Morphology in UK Biobank

Table 4. Multivariate Associations with Average Macular Retinal Nerve Fiber Layer, Ganglion Cell Complex, and Ganglion CelleInner
Plexiform Layer Thicknesses

Macular Retinal Nerve Ganglion Cell Complex Ganglion CelleInner Plexiform

Fiber Layer Thickness (mm) Thickness (mm) Layer Thickness (mm)
b Coefficient b Coefficient b Coefficient
(95% Confidence Interval) P Value (95% Confidence Interval) P Value (95% Confidence Interval) P Value
e70 e300
Age (per decade) e0.43 (e0.47 to e0.38) 5.5 3 10 e1.90 (e1.99 to e1.80) <1 3 10 e1.51 (e1.58 to e1.45) <1 3 10e300
Gender
Female Reference Reference Reference
Male e0.66 (e0.73 to e0.58) 1.8 3 10e68 e0.73 (e0.88 to e0.59) 7.1 3 10e23 e0.10 (e0.20 to 0.00) 0.042
Ethnicity
White Reference Reference Reference
Asian e0.97 (e1.19 to e0.75) 9.9 3 10e18 e1.98 (e2.42 to e1.54) 1.2 3 10e18 e1.09 (e1.38 to e0.79) 3.9 3 10e13
Black e1.65 (e1.86 to e1.43) 2.4 3 10e50 e1.95 (e2.38 to e1.52) 5.7 3 10 e19
e0.38 (e0.67 to e0.10) 0.009
Other/mixed e0.48 (e0.70 to e0.26) 2.1 3 10e5 e0.12 (e0.56 to 0.32) 0.58 0.39 (0.10e0.68) 0.009
Body mass index e0.12 (e0.16 to e0.08) 1.5 3 10e8 e0.29 (e0.37 to e0.21) 5.9 3 10e12 e0.18 (e0.24 to e0.13) 1.1 3 10e10
(per 5 kg/m2)
Smoking status
Never Reference Reference Reference
Previous 0.02 (e0.06 to 0.10) 0.65 0.02 (e0.14 to 0.18) 0.80 0.00 (e0.10 to 0.10) 0.99
Current e0.06 (e0.18 to 0.07) 0.37 e0.07 (e0.32 to 0.18) 0.57 e0.01 (e0.18 to 0.15) 0.89
Alcohol intake
Never or special Reference Reference Reference
occasion only
1e3 times per month 0.02 (e0.11 to 0.16) 0.72 0.08 (e0.19 to 0.34) 0.57 0.07 (e0.11 to 0.24) 0.45
1e2 time per week 0.04 (e0.07 to 0.15) 0.43 0.07 (e0.15 to 0.29) 0.53 0.04 (e0.11 to 0.18) 0.60
3e4 times per week e0.09 (e0.21 to 0.02) 0.11 e0.24 (e0.47 to e0.02) 0.036 e0.14 (e0.29 to 0.02) 0.08
Daily or almost daily e0.19 (e0.31 to e0.07) 0.002 e0.63 (e0.87 to e0.40) 1.4 3 10e7 e0.46 (e0.61 to e0.30) 1.1 3 10e8
Townsend
deprivation index
Least deprived quartile Reference Reference Reference
Second quartile e0.02 (e0.12 to 0.08) 0.69 e0.02 (e0.22 to 0.18) 0.84 0.00 (e0.14 to 0.13) 0.95
Third quartile e0.12 (e0.22 to e0.02) 0.019 e0.25 (e0.44 to e0.05) 0.016 e0.13 (e0.26 to 0.00) 0.06
Most deprived quartile e0.13 (e0.24 to e0.03) 0.012 e0.41 (e0.61 to e0.20) 1.2 3 10e4 e0.28 (e0.42 to e0.14) 6.6 3 10e5
Education level
Less than O level Reference Reference Reference
O level 0.15 (0.05e0.25) 0.004 0.30 (0.10e0.50) 0.003 0.16 (0.02e0.29) 0.022
A Level 0.35 (0.22e0.47) 3.3 3 10e8 0.61 (0.37e0.86) 8.4 3 10e7 0.26 (0.09e0.42) 0.002
Degree 0.36 (0.26e0.45) 2.3 3 10e14 0.65 (0.47e0.83) 1.7 3 10e12 0.30 (0.18e0.42) 8.4 3 10e7
Diabetes status
No Reference Reference Reference
Yes e0.54 (e0.71 to e0.38) 1.5 3 10e10 e0.83 (e1.16 to e0.50) 8.5 3 10e7 e0.34 (e0.56 to e0.12) 0.003
Spherical e0.33 (e0.35 to e0.31) 1.1 3 10e251 0.40 (0.36e0.43) 1.2 3 10e93 0.73 (0.71e0.76) <1 3 10e300
equivalent (diopters)
IOPcc (mmHg) 0.00 (e0.01e0.01) 0.77 e0.03 (e0.05 to e0.01) 5.8 3 10e5 e0.04 (e0.05 to e0.03) 4.0 3 10e10
Model R2 value 6.7% 5.6% 11.2%

IOPcc ¼ cornealecompensated intraocular pressure.

Results are presented for 3 multivariate models with all explanatory variables presented together in the same model (n ¼ 42 044 for each model). P values
less than 0.0024 appear in boldface and represent significance at a Bonferroni-corrected threshold for 21 statistical tests.

treatment for proliferative diabetic retinopathy without accurately. The grid within which the SD OCT measure-
macular edema has been shown to cause an increase in ments are made will cover different absolute amounts of the
GCIPL thickness.32 macula depending on the refractive status of the eye.
We observed very strong associations of spherical Because of this, the foveal pit will take up a different pro-
equivalent with inner retinal thickness, and strikingly, the portion of the grid simply as a result of refractive error-
associations were in a different direction for mRNFL than induced magnification effects. In longer, myopic eyes, the
for GCC and GCIPL. Our finding of thinner GCC and grid will cover a larger proportion of the macula than in
GCIPL with increasing myopia is in agreement with pre- shorter, emmetropic eyes. This will result in the thickest
vious reports.23e26 Our finding of a thicker mRNFL with parts of the inner retina proportionally covering less of the
increasing myopia is novel, to the best of our knowledge. grid in myopic eyes, potentially explaining the thinner GCC
Analyzing the relationship between refractive error and and GCIPL. Additionally, the foveal pit will make up pro-
retinal thickness is extremely difficult because of the issue portionally less of the imaged and analyzed grid in myopic
of magnification effects, which cannot be accounted for eyes than emmetropic eyes. If the foveal pit affects the

67
 Ophthalmology Volume 127, Number 1, January 2020

retinal nerve fiber layer more than the GCC or GCIPL, then

Value (%)

Results are from linear regression models adjusted for age, gender, and spherical equivalent. Six different associations with intraocular pressure are presented: either IOPg or IOPcc and 3 methods for handling
intraocular pressure values for patients taking intraocular pressure-lowering treatment (including participants taking intraocular pressure-lowering medication and imputing pretreatment intraocular pressure,
including participants taking intraocular pressure-lowering medication and using current intraocular pressure rather than imputing intraocular pressure, and excluding participants using intraocular pressure-
Table 5. Associations of Intraocular Pressure with Average Macular Retinal Nerve Fiber Layer, Ganglion Cell Complex, and Ganglion CelleInner Plexiform Layer Thicknesses

this may explain why the retinal nerve fiber layer is thicker

10.8
10.8

10.8
10.8

10.7
10.6
R2
in SD OCT images of myopic eyes. With the current data in
Ganglion CelleInner Plexiform

our study, we do not believe it is possible to determine true

e0.06 (e0.08 to e0.05) 2.7 3 10e20

e0.05 (e0.06 to e0.03) 1.1 3 10e11

e0.05 (e0.07 to e0.04) 4.4 3 10e15

e0.05 (e0.07 to e0.04) 2.8 3 10e14

e0.04 (e0.05 to e0.02) 1.4 3 10e7

e0.03 (e0.05 to e0.02) 1.2 3 10e6

differences in inner retinal thickness by refractive error
Layer Thickness (mm)

P Value
because we were unable to distinguish the contribution as a
result of magnification.
We identified several novel associations with inner
Value (%) (95% Confidence Interval)

retinal thickness parameters. Asian and black participants

had thinner mRNFL, GCC, and GCIPL than white partici-
b Coefficient

pants. Although this in part may reflect a greater suscepti-

lowering medications [see “Methods”]). P values less than 0.008 appear in boldface and represent significance at a Bonferroni-corrected threshold for 6 statistical tests.
bility to glaucomatous processes in nonwhite people, as
suggested from epidemiologic data,27 this more likely
reflects ethnically determined differences in baseline
retinal anatomic features. This highlights the importance
of taking ethnicity into account when defining normal
ranges for diagnostic tests for glaucoma.
Frequent alcohol intake was associated with thinner
4.8
4.8

4.8
4.7

4.7
4.6
R2
Ganglion Cell Complex Thickness (mm)

mRNFL, GCC, and GCIPL compared with rare or no

alcohol intake. This is in agreement with a study examining
e0.07 (e0.09 to e0.05) 7.8 3 10e13
e0.05 (e0.07 to e0.03) 1.4 3 10e6

3.5 3 10e8

1.6 3 10e7

cRNFL (i.e., circumpapillary rather than macular mea-

P Value

0.003

0.012

sures)33; Lamparter et al33 reported thinner cRNFL in

participants of the Gutenberg Health Study whose alcohol
intake was high according to World Health Organization
guidelines (10 g/day for women; 20 g/day for men).
(95% Confidence Interval) P Value Value (%) (95% Confidence Interval)

e0.06 (e0.08 to e0.04)

e0.03 (e0.05 to e0.01)

e0.05 (e0.07 to e0.03)

e0.03 (e0.05 to e0.01)

Our findings support the assertion that retinal nerve fiber

layer thinning may occur as a result of chronic alcohol
b Coefficient

intake and in a dose-dependent manner.33 It is not

possible to determine from our study what mechanisms
may be underlying the association with alcohol. Potential
IOPcc ¼ corneal-compensated intraocular pressure; IOPg ¼ Goldmann-correlated intraocular pressure.

mechanisms may include direct effects of alcohol on

retinal ganglion cells or indirect effects via dehydration.
We found participants who were more socially deprived
to have thinner inner retinas, particularly for GCC and
5.6
5.6

5.6
5.6

5.6
5.6
R2

GCIPL measurements. This is consistent with the previously

Macular Retinal Nerve Fiber

reported association of social deprivation with self-reported

Layer Thickness (mm)

glaucoma in the UK Biobank.34 We also found less

0.042
0.55

0.56
0.31

0.72
0.15

educated participants to have thinner mRNFL, GCC, and

GCIPL. This is consistent with a scanning laser
e0.01 (e0.02 to 0.00)
0.00 (e0.01 to 0.01)

0.00 (e0.01 to 0.01)

0.01 (e0.01 to 0.02)

0.00 (e0.01 to 0.01)

ophthalmoscopy study of cRNFL in participants of

0.01 (0.00e0.02)

another, independent United Kingdom cohort of older

b Coefficient

adults.35 Interestingly, the association of a thicker GCC

and GCIPL in more educated participants is strong enough
to outweigh the expected thinner GCC and GCIPL we
may expect to see given the association between education
and myopia,36 even in unadjusted analyses (Table 3).
From our cross-sectional study, it is not possible to know
IOP-lowering medication (n ¼ 42 044)

if less educated participants demonstrated thinner inner

retinas at baseline, or whether this is something that
developed over time as a result of lack of education.
IOP imputed if using IOP-lowering

Current IOP analyzed regardless of

Another study of UK Biobank participants reported that

Excluded if using IOP-lowering
medication (n ¼ 42 044)

medication (n ¼ 41 770)

baseline mRNFL predicted future cognitive decline.37 If

inner retinal thickness is associated causally with cognitive
health, this may explain the relationship with education
that we observed with more cognitively able people with
IOPcc (mmHg)

IOPcc (mmHg)

IOPcc (mmHg)
IOPg (mmHg)

IOPg (mmHg)

thicker inner retinas being more likely to pursue education

IOPg (mmHg)

for longer periods.

Typically, in epidemiologic studies, if a significant as-
sociation is not found, it may be the case that a true asso-
ciation does not exist or that the study was underpowered to

68
 Khawaja et al 
Macular Inner Retinal Morphology in UK Biobank

detect a true association. With the huge sample size in our definitively which factors were or were not associated with
study, it is unlikely that a biologically meaningful associa- inner retinal thickness. Limitations of our study include the
tion will not be identified if it truly exists. Strong associa- reliance on automated segmentation of the retina. Although
tions in our study (e.g., age and spherical equivalent) were we applied strict quality control criteria and manually
so statistically significant that the P value was so small such checked a proportion of scans,16 it was not feasible to check
that the statistical software could not distinguish it from all scans manually for accurate segmentation. Additionally,
0 (P < 10e300). We did not find associations between inner it was not possible to segment reliably the boundary
retinal thickness and height or smoking status. Given the between the ganglion cell layer and inner plexiform layer,
statistical power, our study provides good evidence for no meaning we could not examine these layers individually.
true association between inner retinal thickness and height Another limitation of the UK Biobank is that it is a
or smoking. The lack of association with smoking suggests volunteer cohort, and participants are likely healthier than
that inflammatory mechanisms do not have a prominent role the general population. Furthermore, our quality control
in pathophysiologic processes underlying variation in inner process excluded participants, and this also could lead to
retinal thickness. selection bias. This may limit the generalizability of our
The effect sizes for the associations we report are modest results, although it seems unlikely that that directions of
in magnitude but important when considered in the context association with inner retinal thickness would be
of the standard deviation of the retinal thickness parameters differential by selection.
and when compared with the association with age (a well- In summary, we present a very large epidemiologic
established important association of inner retinal thickness study of inner retinal anatomic characteristics. We identi-
that is corrected for in diagnostic tests). For example, the fied novel associations with thinner inner retina, including
thinner mRNFL observed in men had a magnitude of 17% non-European ethnicity, frequent alcohol intake, greater
of the standard deviation of mRNFL and is equivalent to the social deprivation, and lower educational attainment.
magnitude of thinner mRNFL observed in participants who These associations were statistically independent from
were 15 years older. Similarly, the thinner GCC seen in each other and warrant further investigation to help
black compared with white participants had a magnitude of determine if they are causal and what the underlying
26% of the standard deviation of GCC, equivalent to being mechanisms may be. Stronger associations were seen with
10 years older. Collectively, the predictor variables we GCIPL compared with mRNFL or GCC, particularly for
examined explained a considerable proportion of the total IOP, suggesting that GCIPL may be a superior biomarker
variance of inner retinal thickness: 6.7% for mRNFL, 5.6% for macular pathophysiologic processes and especially for
for GCC, and 11.2% for GCIPL (Table 4). glaucoma.
We found higher IOP to be associated with a thinner GCC
and GCIPL. If we consider glaucoma as a complex disease
with multiple underlying causes and with a phenotypic References
spectrum from normal to severe disease, it is likely that
variation in inner retinal anatomic features in a population
1. Hood DC, Raza AS, de Moraes CGV, et al. Glaucomatous
may be reflecting the preclinical disease spectrum and may be damage of the macula. Prog Retin Eye Res. 2013;32:1e21.
secondary to these causes. Therefore, determinants of inner 2. Kim KE, Park KH. Macular imaging by optical coherence
retinal thickness variation also may be determinants of the tomography in the diagnosis and management of glaucoma. Br
glaucomatous process. On this basis, we would expect to see J Ophthalmol. 2018;102:718e724.
an association between IOP and inner retinal thickness, given 3. Oddone F, Lucenteforte E, Michelessi M, et al. Macular versus
the strength of IOP as a risk factor for glaucoma.38 The retinal nerve fiber layer parameters for diagnosing manifest
association with IOP in our study was most significant for glaucoma: a systematic review of diagnostic accuracy studies.
GCIPL, potentially suggesting GCIPL as a superior Ophthalmology. 2016;123:939e949.
biomarker for glaucomatous processes. We found no 4. Mwanza JC, Durbin MK, Budenz DL, et al. Glaucoma diag-
nostic accuracy of ganglion cell-inner plexiform layer thick-
significant association between IOP and mRNFL,
ness: comparison with nerve fiber layer and optic nerve head.
potentially suggesting mRNFL to be a less effective Ophthalmology. 2012;119:1151e1158.
biomarker for glaucomatous processes. This is in contrast to 5. Tan O, Chopra V, Lu ATH, et al. Detection of macular
the well-established role of cRNFL as a biomarker in the ganglion cell loss in glaucoma by Fourier-domain optical
management of glaucoma.39 Our data suggest that, at the coherence tomography. Ophthalmology. 2009;116(12):
macula, variation in mRNFL within a population may be 2305e2314.
more influenced by factors other than glaucomatous 6. Shin JW, Sung KR, Park SW. Patterns of progressive ganglion
processes. celleinner plexiform layer thinning in glaucoma detected by
Overall, the predictor variables we examined explained OCT. Ophthalmology. 2018;125(10):1515e1525.
twice more of the variance of GCIPL than of either GCC or 7. Hou HW, Lin C, Leung CKS. Integrating macular ganglion
cell inner plexiform layer and parapapillary retinal nerve fiber
mRNFL. This suggests that GCIPL is better reflecting the
layer measurements to detect glaucoma progression.
biological processes that these variables contribute to and Ophthalmology. 2018;125:822e831.
therefore may be a superior biomarker for pathophysiologic 8. Hammel N, Belghith A, Weinreb RN, et al. Comparing the
processes influencing retinal ganglion cell health in general. rates of retinal nerve fiber layer and ganglion celleinner
The major strength of our study is the very large sample plexiform layer loss in healthy eyes and in glaucoma eyes.
size, which afforded sufficient power to determine Am J Ophthalmol. 2017;178:38e50.

69
 Ophthalmology Volume 127, Number 1, January 2020

9. Lavinsky F, Wu M, Schuman JS, et al. Can macula and optic complex in a healthy aging British cohort. JAMA Ophthalmol.
nerve head parameters detect glaucoma progression in eyes 2017;135:31e38.
with advanced circumpapillary retinal nerve fiber layer dam- 25. Koh VT, Tham YC, Cheung CY, et al. Determinants of gan-
age? Ophthalmology. 2018;125(12):1907e1912. glion cell-inner plexiform layer thickness measured by high-
10. Zhang X, Dastiridou A, Francis BA, et al. Comparison of definition optical coherence tomography. Invest Ophthalmol
glaucoma progression detection by optical coherence tomog- Vis Sci. 2012;53:5853e5859.
raphy and visual field. Am J Ophthalmol. 2017;184:63e74. 26. Kim NR, Kim JH, Lee J, et al. Determinants of perimacular
11. Kansal V, Armstrong JJ, Pintwala R, Hutnik C. Optical inner retinal layer thickness in normal eyes measured by
coherence tomography for glaucoma diagnosis: an evidence fourier-domain optical coherence tomography. Invest Oph-
based meta-analysis. PLoS One. 2018;13:e0190621. thalmol Vis Sci. 2011;52:3413e3418.
12. Kim HJ, Park KH, Kim YK, Jeoung JW. Evaluation of layer- 27. Tham Y-C, Li X, Wong TY, et al. Global prevalence of
by-layer segmented ganglion cell complex thickness for glaucoma and projections of glaucoma burden through 2040: a
detecting early glaucoma according to different macular grids. systematic review and meta-analysis. Ophthalmology.
J Glaucoma. 2017;26:712e717. 2014;121:2081e2090.
13. Pazos M, Dyrda AA, Biarnés M, et al. Diagnostic accuracy of 28. van Dijk HW, Kok PHB, Garvin M, et al. Selective loss of
Spectralis SD OCT automated macular layers segmentation to inner retinal layer thickness in type 1 diabetic patients with
discriminate normal from early glaucomatous eyes. Ophthal- minimal diabetic retinopathy. Invest Ophthalmol Vis Sci.
mology. 2017;124:1218e1228. 2009;50:3404e3409.
14. Chien JL, Ghassibi MP, Patthanathamrongkasem T, et al. 29. Pekel E, Tufaner G, Kaya H, et al. Assessment of optic disc
Glaucoma diagnostic capability of global and regional mea- and ganglion cell layer in diabetes mellitus type 2. Med
surements of isolated ganglion cell layer and inner plexiform (United States). 2017;96:1e4.
layer. J Glaucoma. 2017;26:208e215. 30. Ng DSK, Chiang PPC, Tan G, et al. Retinal ganglion cell
15. Keane PA, Grossi CM, Foster PJ, et al. Optical coherence neuronal damage in diabetes and diabetic retinopathy. Clin
tomography in the UK Biobank studydrapid automated Exp Ophthalmol. 2016;44:243e250.
analysis of retinal thickness for large population-based studies. 31. Dehghani C, Srinivasan S, Edwards K, et al. Presence of pe-
PLoS One. 2016;11:1e15. ripheral neuropathy is associated with progressive thinning of
16. Patel PJ, Foster PJ, Grossi CM, et al. Spectral-domain optical retinal nerve fiber layer in type 1 diabetes. Invest Ophthalmol
coherence tomography imaging in 67 321 adults: associations Vis Sci. 2017;58:BI0234eBI0239.
with macular thickness in the UK Biobank Study. Ophthal- 32. Kim JJ, Im JC, Shin JP, et al. One-year follow-up of macular
mology. 2016;123:829e840. ganglion cell layer and peripapillary retinal nerve fibre layer
17. Luce DA. Determining in vivo biomechanical properties of the thickness changes after panretinal photocoagulation. Br J
cornea with an ocular response analyzer. J Cataract Refract Ophthalmol. 2014;98:213e217.
Surg. 2005;31:156e162. 33. Lamparter J, Schmidtmann I, Schuster AK, et al. Association
18. Luce D. Methodology for corneal compensated IOP and of ocular, cardiovascular, morphometric and lifestyle param-
corneal resistance factor for an Ocular Response Analyzer. eters with retinal nerve fibre layer thickness. PLoS One.
Invest Ophthalmol Vis Sci. 2006;47:2266. 2018;13:1e11.
19. Chan MPY, Grossi CM, Khawaja AP, et al. Associations with 34. Shweikh Y, Ko F, Chan MPY, et al. Measures of socioeco-
intraocular pressure in a large cohort: results from the UK nomic status and self-reported glaucoma in the U.K. Biobank
Biobank. Ophthalmology. 2016;123:771e782. cohort. Eye (Lond). 2015;29:1360e1367.
20. van der Valk R, Webers CAB, Schouten JSAG, et al. Intra- 35. Khawaja AP, Chan MPY, Broadway DC, et al. Laser scanning
ocular pressure-lowering effects of all commonly used glau- tomography in the EPIC-Norfolk Eye Study: principal com-
coma drugs: a meta-analysis of randomized clinical trials. ponents and associations. Invest Ophthalmol Vis Sci. 2013;54:
Ophthalmology. 2005;112:1177e1185. 6638e6645.
21. Hysi PG, Cheng C-Y, Springelkamp H, et al. Genome-wide 36. Mountjoy E, Davies NM, Plotnikov D, et al. Education and
analysis of multi-ancestry cohorts identifies new loci influ- myopia: assessing the direction of causality by mendelian
encing intraocular pressure and susceptibility to glaucoma. Nat randomisation. BMJ. 2018;361:k2022.
Genet. 2014;46:1126e1130. 37. Ko F, Muthy ZA, Gallacher J, et al. Association of retinal
22. Khawaja AP, Cooke Bailey JN, Wareham NJ, et al. Genome- nerve fiber layer thinning with current and future cognitive
wide analyses identify 68 new loci associated with intraocular decline: a study using optical coherence tomography. JAMA
pressure and improve risk prediction for primary open-angle Neurol. 2018;75(10):1198e1205.
glaucoma. Nat Genet. 2018;50:778e782. 38. de Voogd S, Ikram MK, Wolfs RCW, et al. Incidence
23. Mwanza JC, Durbin MK, Budenz DL, et al. Profile and pre- of open-angle glaucoma in a general elderly population:
dictors of normal ganglion cell-inner plexiform layer thickness the Rotterdam Study. Ophthalmology. 2005;112:
measured with frequency-domain optical coherence tomogra- 1487e1493.
phy. Invest Ophthalmol Vis Sci. 2011;52:7872e7879. 39. Tatham AJ, Medeiros FA. Detecting structural progression in
24. Bloch E, Yonova-Doing E, Jones-Odeh E, et al. Genetic and glaucoma with optical coherence tomography. Ophthal-
environmental factors associated with the ganglion cell mology. 2017;124:S57eS65.

Footnotes and Financial Disclosures

1
Originally received: February 20, 2019. NIHR Biomedical Research Centre, Moorfields Eye Hospital NHS
Final revision: August 7, 2019. Foundation Trust and UCL Institute of Ophthalmology, London, United
Accepted: August 13, 2019. Kingdom.
Available online: August 21, 2019. Manuscript no. 2019-387. 2
Department of Ophthalmology, King’s College London, St. Thomas’
Hospital, London, United Kingdom.

70
 Khawaja et al 
Macular Inner Retinal Morphology in UK Biobank
3
Department of Twin Research & Genetic Epidemiology, King’s College Hospital (ST 12 09), and the Department of Health and Social Care
London, St. Thomas’ Hospital, London, United Kingdom. through the award made by the NIHR Biomedical Research Centre at
4
European Molecular Biology Laboratory, European Bioinformatics Moorfields Eye Hospital NHS Foundation Trust (BRC2_009). The
Institute (EMBL-EBI), Wellcome Genome Campus, Hinxton, Cambridge, funding organizations had no role in the design or conduct of this research.
United Kingdom. HUMAN SUBJECTS: Human subjects were included in this study. The
5
Topcon Advanced Biomedical Imaging Laboratory, Oakland, New Jersey. human ethics committees at the North-West Research Ethics Committee
6
Discipline of Clinical Ophthalmology and Eye Health, University of approved the study. All research adhered to the tenets of the Declaration of
Helsinki. All participants provided informed consent.
Sydney Medical School, Sydney, Australia.
Presented at: American Academy of Ophthalmology Annual Meeting, No animal subjects were included in this study.
Chicago, Illinois, October 2018. Author Contributions:
Financial Disclosures: Conception and design: Khawaja, Khaw, Foster, Patel
The author(s) have made the following disclosures: A.P.K.: Consultant e Analysis and interpretation: Khawaja, Chua, Hysi, Georgoulas, Currant,
Allergan, Novartis, Santen, Aerie; Lecturer e Grafton Optical; Financial Fitzgerald, Birney, Ko, Yang, Reisman, Garway-Heath, Hammond, Khaw,
support e Thea. Foster, Patel, Strouthidis
Q.Y.: Employee e Topcon Healthcare Solutions, Inc. Data collection: Khawaja, Khaw, Foster, Patel
E.B.: Consultant e Oxford Nanopore, Dovetail Ltd., GSK plc. Obtained funding: Khawaja, Yang, Birney, Reisman, Garway-Heath,
C.R.: Employee e Topcon Healthcare Solutions. Khaw, Foster
D.F.G.-H.: Consultant e Carl Zeiss Meditec; Lecturer e OptoVue; Roy- Overall responsibility: Khawaja, Chua, Hysi, Georgoulas, Currant, Fitz-
alties e Heidelberg Engineering; Equipment support e Heidelberg Engi- gerald, Birney, Ko, Yang, Reisman, Garway-Heath, Hammond, Khaw,
neering, Topcon, Carl Zeiss Meditec. Foster, Patel, Strouthidis
P.T.K.: Consultant e Radiance Tx, Santen Pharmaceuticals, Aerie Phar- Abbreviations and Acronyms:
maceuticals, Novartis, Alcon, Valid Insight, ISARNA, Thea Pharmaceuti- BMI ¼ body mass index; cRNFL ¼ circumpapillary retinal nerve fiber
cals, Glaukos, Belkin, GLG; Lecturer e Allergan; Patent e Müller cells; layer; D ¼ diopter; GCC ¼ ganglion cell complex; GCIPL ¼ ganglion
Equity owner e Radiance. celleinner plexiform layer; IOP ¼ intraocular pressure; IOPcc ¼ corneal-
P.J.F.: Consultant e InHouse Research, AlphaSights, GLG, 90Ten, Simon compensated intraocular pressure; IOPg ¼ Goldmann-correlated intraoc-
Kucher, DeepMind/Google. ular pressure; mRNFL ¼ macular retinal nerve fiber layer; ORA ¼ Ocular
Response Analyzer; SD ¼ spectral-domain; TABS ¼ Topcon Advanced
The UK Biobank (Appendix 1, available at www.aaojournal.org) was
Boundary Segmentation.
established by the Wellcome Trust medical charity; the Medical Research
Council, Department of Health, Scottish Government; and the Northwest Correspondence:
Regional Development Agency. It has also had funding from the Welsh Anthony P. Khawaja, PhD, FRCOphth, Moorfields Eye Hospital NHS
Assembly Government, British Heart Foundation, and Diabetes United Foundation Trust, 162 City Road, RDCEC (2nd Floor), London EC1V 2PD,
Kingdom. Dr. Foster is supported by the Richard Desmond Charitable United Kingdom. E-mail: [email protected].
Trust via Fight for Sight (1956), the Special Trustees of Moorfields Eye

71
HAWK and HARRIER: Phase 3, Multicenter,
Randomized, Double-Masked Trials of
Brolucizumab for Neovascular Age-Related
Macular Degeneration
Pravin U. Dugel, MD,1 Adrian Koh, MD, FRCS,2 Yuichiro Ogura, MD,3 Glenn J. Jaffe, MD,4
Ursula Schmidt-Erfurth, MD,5 David M. Brown, MD,6 Andre V. Gomes, MD, PhD,7 James Warburton, MBBS,8
Andreas Weichselberger, PhD,8 Frank G. Holz, MD,9 on behalf of the HAWK and HARRIER Study Investigators*

Purpose: Two similarly designed phase 3 trials (HAWK and HARRIER) compared brolucizumab, a single-
chain antibody fragment that inhibits vascular endothelial growth factor-A, with aflibercept to treat neovascular
age-related macular degeneration (nAMD).
Design: Double-masked, multicenter, active-controlled, randomized trials.
Participants: Patients (N ¼ 1817) with untreated, active choroidal neovascularization due to age-related
macular degeneration in the study eye.
Intervention: Patients were randomized to intravitreal brolucizumab 3 mg (HAWK only) or 6 mg or aflibercept
2 mg. After loading with 3 monthly injections, brolucizumab-treated eyes received an injection every 12 weeks
(q12w) and were interval adjusted to every 8 weeks (q8w) if disease activity was present; aflibercept-treated eyes
received q8w dosing.
Main Outcome Measures: The primary hypothesis was noninferiority in mean best-corrected visual acuity
(BCVA) change from baseline to Week 48 (margin: 4 letters). Other key end points included the percentage of
patients who maintained q12w dosing through Week 48 and anatomic outcomes.
Results: At Week 48, each brolucizumab arm demonstrated noninferiority to aflibercept in BCVA change
from baseline (least squares [LS] mean, þ6.6 [6 mg] and þ6.1 [3 mg] letters with brolucizumab vs. þ6.8 letters
with aflibercept [HAWK]; þ6.9 [brolucizumab 6 mg] vs. þ7.6 [aflibercept] letters [HARRIER]; P < 0.001 for each
comparison). Greater than 50% of brolucizumab 6 mgetreated eyes were maintained on q12w dosing through
Week 48 (56% [HAWK] and 51% [HARRIER]). At Week 16, after identical treatment exposure, fewer brolucizumab
6 mgetreated eyes had disease activity versus aflibercept in HAWK (24.0% vs. 34.5%; P ¼ 0.001) and HARRIER
(22.7% vs. 32.2%; P ¼ 0.002). Greater central subfield thickness reductions from baseline to Week 48
were observed with brolucizumab 6 mg versus aflibercept in HAWK (LS mean 172.8 mm vs. 143.7 mm;
P ¼ 0.001) and HARRIER (LS mean 193.8 mm vs. 143.9 mm; P < 0.001). Anatomic retinal fluid outcomes
favored brolucizumab over aflibercept. Overall, adverse event rates were generally similar with brolucizumab and
aflibercept.
Conclusions: Brolucizumab was noninferior to aflibercept in visual function at Week 48, and >50% of
brolucizumab 6 mgetreated eyes were maintained on q12w dosing interval through Week 48. Anatomic out-
comes favored brolucizumab over aflibercept. Overall safety with brolucizumab was similar to aflibercept (Clin-
icalTrials.gov; NCT02307682, NCT02434328). Ophthalmology 2020;127:72-84 ª 2019 by the American Academy
of Ophthalmology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

Supplemental material available at www.aaojournal.org.

Age-related macular degeneration (AMD) is a chronic, anticipated increased prevalence of patients with AMD
progressive disease and a leading cause of vision loss.1e3 portend a scenario that is not sustainable.1,2,7 Factors lead-
Pivotal trials validated intravitreally administered ing to treatment nonadherence include travel to appoint-
antievascular endothelial growth factor A (VEGF-A) ther- ments, patient dissatisfaction, and the burden of numerous
apy for neovascular AMD (nAMD) treatment, which has visits, which may contribute to suboptimal vision out-
greatly improved patient outcomes.4e6 However, the need comes.8e10 Real-world studies across several countries
for frequent clinic and injection visits coupled with the reveal lower treatment frequencies and poorer vision

72 ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.04.017

This is an open access article under the CC BY-NC-ND license ISSN 0161-6420/19
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Published by Elsevier Inc.
 Dugel et al 
Brolucizumab for Neovascular AMD

outcomes versus phase 3 trials.11 An ongoing challenge is to 408 sites in North, Central, and South America; Europe; Asia;
maintain nAMD treatment efficacy while reducing clinic Australia; and Japan. All patients provided written informed con-
visits.7,12,13 sent before screening or initiation of any study-related procedures.
Single-chain antibody fragments (scFv) are the smallest Protocols were approved by an Independent Ethics Committee/
Institutional Review Board. Trials were conducted in accordance
functional unit of an antibody, allowing delivery of a greater
with principles of the Declaration of Helsinki, International Con-
molar dose compared with larger molecules and the potential ference on Harmonization E6 Good Clinical Practice Consolidated
for more effective tissue penetration,14e16 attributes designed Guideline, and other regulations as applicable and were compliant
to increase duration.17 Brolucizumab (formerly ESBA1008 with the Health Insurance Portability and Accountability Act of
and RTH258) was developed by grafting complementarity- 1996. The trials were designed by a committee of investigators and
determining regions of a novel antieVEGF-A antibody to a the sponsor. All investigators collected data, and the sponsor
human scFv scaffold, thus circumventing the production, analyzed data. All authors had full 1-year data access. An inde-
solubility, stability, and in vivo activity issues encountered pendent data monitoring committee was established to monitor the
over the last quarter century of scFv development.16e21 Pre- safety of the trial participants, ensure that the trials were conducted
clinical data demonstrated a 2.2- and 1.7-fold higher exposure with the highest scientific and ethical standards, and make appro-
priate recommendations based on the safety data reviewed. Spe-
in the retina and retinal pigment epithelium (RPE)/choroid,
cifically, the data monitoring committee was tasked to make
respectively, with brolucizumab compared with ranibizu- recommendations to amend treatment rules, inclusion and exclu-
mab,16 suggesting the potential for better intraretinal fluid sion criteria, and adverse event definitions and grading scales based
(IRF), subretinal fluid (SRF), and sub-RPE fluid control on planned and unplanned safety data reviews. Medical writers
across retinal layers. (paid by the sponsor) provided editorial assistance, including first
HAWK and HARRIER are 2 similarly designed phase 3 draft development with input from all authors. All authors
trials comparing brolucizumab with aflibercept to treat contributed to data interpretation and manuscript writing, reviewed
nAMD. The design of these studies was informed by and provided feedback on all drafts, and collectively decided to
exploratory analyses of previous studies with other anti- publish the results; the sponsor reviewed and approved the
VEGF agents, as well as the results of the phase 2 brolu- manuscript. All authors vouch for the completeness and accuracy
of the data and analyses and affirm the trial was conducted and
cizumab OSPREY study. Analyses from the PIER and
reported in agreement with the protocol.
EXCITE studies have shown that visual and anatomic
response during and for the 12 weeks after the loading phase
are associated with visual acuity outcomes over the Trial Participants
remainder of the first year of treatment.22e25 In addition, Eligible patients were aged 50 years and had untreated, active
recent analyses from the EXCITE study have shown that choroidal neovascularization lesions secondary to AMD affecting
patients who lose vision during the initial loading phase will the central subfield (the circular area within 1 mm diameter around
have better visual outcomes with more frequent treatment the foveal center on imaging); choroidal neovascularization lesions
versus patients who follow an every 12 weeks (q12w) (including classic and occult), as assessed by fluorescein angiog-
treatment regimen.26 Analyses from CATT and EXCITE raphy, comprising >50% of total lesion area; IRF and/or SRF
affecting the central subfield as assessed on spectral-domain OCT;
have shown that new IRF/intraretinal cysts, and to a lesser
BCVA between 78 and 23 Early Treatment Diabetic Retinopathy
degree central subfield thickness (CST) increase, are Study letters (inclusive; Snellen equivalents, approximately 20/32
associated with later visual acuity decline.27e29 Finally, to 20/400); and no fibrosis or geographic atrophy affecting the
the selection of q12w and every 8 weeks (q8w) dosing was central subfield. Patients could not have received any approved or
based on results of the OSPREY study, in which a head-to- investigational nAMD treatment at any time (study eye). Full in-
head comparison of the q8w treatment regimen between clusion/exclusion criteria are provided in Supplemental Materials
brolucizumab 6 mg and aflibercept 2 mg showed anatomic (available at www.aaojournal.org). Final anatomic eligibility
advantages with brolucizumab while reaching noninferiority determination was made by a central reading center (Duke
in best-corrected visual acuity (BCVA).14 In the same study, Reading Center for HAWK and Vienna Reading Center for
brolucizumab-treated patients were subsequently challenged HARRIER).
with a q12w dosing interval, with an outcome suggesting
that approximately 50% of patients were adequately treated. Randomization and Treatment
The primary objective of both HAWK and HARRIER Eyes were randomized (Interactive Response Technology [IRT])
was to demonstrate that brolucizumab (q12w/q8w) is non- 1:1:1 to brolucizumab 3 mg, brolucizumab 6 mg, or aflibercept 2
inferior to fixed-dose aflibercept with respect to the change mg (HAWK) or 1:1 to brolucizumab 6 mg or aflibercept 2 mg
in BCVA from baseline to Week 48. In these studies, a (HARRIER). The unmasked injecting investigator or his/her
q12w/q8w regimen allows for treatment interval assignment unmasked delegate contacted the IRT after confirming that the
guided by assessment of individual disease activity using patient met all the inclusion and none of the exclusion criteria. The
functional and anatomic parameters. IRT assigned a randomization number to the patient that was used
to link the patient to a treatment arm and specified a unique
medication number for the first package of study treatment to be
Methods administered to the patient. The randomization number was not
communicated to unmasked staff. The randomization numbers
Trial Description and Oversight were generated using the following procedure to ensure that
treatment assignment was unbiased and concealed from patients
HAWK (NCT02307682) and HARRIER (NCT02434328) were 2- and masked study center personnel. A member of the Statistical
year, randomized, double-masked, multicenter trials conducted in Programming group who was not part of the study team generated

73
 Ophthalmology Volume 127, Number 1, January 2020

the randomized allocation for the study treatment assignment based LOCF imputation as well as observed data for the full analysis set
on a randomization plan that provided study-specific criteria for and per-protocol analysis set. A 2-sided 95% confidence interval
randomization, including block size and randomization ratio. (CI) for the treatment difference was derived from an analysis of
Treatment was assigned to patients through the IRT. Each patient variance model with treatment, baseline BCVA categories (55,
number was associated with a treatment arm, according to the 56e70, and 71 letters), and age categories (<75 and 75 years)
randomized allocation generated using the computer software SAS as fixed effects. To demonstrate noninferiority, the lower limit of
version 9.2 (PROC PLAN; SAS Institute Inc, Cary, NC). Patients the 95% CI was required to be greater than 4 letters. Supportive
were assigned numbers sequentially according to the time of analyses were based on mixed-model repeated measures analysis
randomization. models using observed data.
After injections at Weeks 0, 4, and 8 (loading phase), broluci- If each BCVA-related noninferiority hypothesis (Supplemental
zumab was injected q12w unless disease activity was identified, Materials, available at www.aaojournal.org), tested hierarchically
resulting in permanent adjustment to q8w (collectively reported as by dose (6 mg before 3 mg) and end point (primary before
a q12w/q8w regimen); aflibercept was injected q8w, per label at secondary), reached statistical significance, additional
the time of study initiation30 (Fig S1, available at confirmatory superiority testing of brolucizumab versus
www.aaojournal.org). On the basis of the assumption of stable aflibercept was prespecified in HAWK (based on HARRIER
treatment need,31 subsequent monitoring of the adequacy of the learnings) with parallel testing in the categories of CST
brolucizumab q12w treatment interval was assessed by masked reductions, presence of IRF and/or SRF, and disease activity at
investigators at Week 16 and at scheduled q12w treatment visits Week 16 using a global 1-sided alpha of 0.025, which was split
(disease activity assessments at Weeks 20, 32, and 44). In into local 1-sided significance levels of 0.005, 0.01, and 0.01,
HARRIER, additional assessments were performed 8 weeks after respectively. Within each end point category, multiplicity was
every scheduled q12w brolucizumab injection (Weeks 28 and controlled by hierarchical testing according to a prespecified order
40) based on health authority feedback. Briefly, in addition to of doses and time points. The probabilities for maintaining q12w
BCVA testing, the masked investigator assessed nAMD disease status were derived from time-to-event analyses (first disease ac-
activity to identify eyes that required more frequent treatment at tivity/q8w need). In case of informative censoring (lack of efficacy
a q8w interval. To rapidly identify those patients with a higher or safety), q8w need was imputed.
anti-VEGF treatment need after loading, the protocol guidance at A sample size of 297 eyes per arm allowed noninferiority
Week 16 provided specific criteria for CST and IRF status assessed determination of brolucizumab versus aflibercept regarding BCVA
by spectral domain OCT (Table S1; available at change from baseline to Week 48 at a 1-sided alpha level of 0.025
www.aaojournal.org). Thereafter, guidance was based on BCVA with a power of approximately 90%, assuming equal efficacy and
decline due to nAMD activity compared with Week 12. an SD of 15 letters. To account for a 10% dropout rate, 330 eyes
Ultimately, the masked investigator made the final treatment were planned to be randomized to each arm.
decisions based on clinical judgment. To maintain masking,
aflibercept-treated eyes underwent the same disease activity as-
sessments as brolucizumab-treated eyes. Treatment exposure was Results
identical up to Week 16, allowing a matched comparison of bro-
lucizumab and aflibercept up to 8 weeks after loading. In both Study Patients
studies, patients received a complete ophthalmic exam (including
BCVA and anatomic assessments [IRF/SRF/sub-RPE fluid and Overall, 1082 patients were randomized in HAWK between
CST]) and were evaluated for adverse events every 4 weeks (Table December 2014 and May 2016, and 743 patients were randomized
S2, available at www.aaojournal.org). in HARRIER between June 2015 and April 2016 (Fig S2, available
at www.aaojournal.org). No clinically meaningful differences in
End Points, Statistical Analyses, and Sample demographics and baseline ocular characteristics were observed
in either trial (Tables S3 and S4, available at
Size Determination
www.aaojournal.org). Mean baseline BCVA was 60.6 (HAWK)
Primary and key secondary end point analyses of BCVA change and 61.2 (HARRIER) letters, and approximately 25% of study
and noninferiority margins were established in discussion with the eyes had BCVA 71 letters at baseline, which is reflective of
US Food and Drug Administration, European Medicines Agency, current clinical practice and in line with BCVA inclusion criteria.
and Pharmaceutical and Medical Device Agency. The primary end In HAWK, 91.4%, 89.8%, and 87.3% of patients treated with
point was mean BCVA change from baseline to Week 48. Key brolucizumab 3 mg, brolucizumab 6 mg, and aflibercept,
secondary end points were BCVA change from baseline averaged respectively, completed study treatment up to Week 48. In
over the period of Week 36 through Week 48 (to account for HARRIER, 93.3% and 93.5% of brolucizumab 6 mge and
differences in timing of treatment), q12w treatment status at Week aflibercept-treated patients, respectively, completed study treat-
48 (brolucizumab only), and q12w treatment status at Week 48 ment up to Week 48. In both trials, the primary reasons for
among eyes with no q8w need during the first q12w cycle (to discontinuation of study treatments were withdrawal by patient and
evaluate the predictive value of the first q12w cycle; brolucizumab adverse events. In HAWK, zero patients treated with brolucizumab
only). Additional secondary efficacy end points included, for each (both doses) and 3 patients (0.8%) treated with aflibercept dis-
post-baseline visit, changes from baseline in BCVA (including continued the study treatment before Week 48 because of lack of
BCVA gain/loss of 15 letters) and CST, status of SRF/IRF and efficacy. In HARRIER, 1 (0.3%) brolucizumab 6 mgetreated pa-
sub-RPE fluid, and presence of disease activity at Week 16. Safety tient and 2 (0.5%) aflibercept-treated patients discontinued the
end points included incidence and characteristics of treatment- study treatment before Week 48 because of lack of efficacy.
emergent adverse events and treatment-emergent changes in
ocular and nonocular parameters. Best-Corrected Visual Acuity
Primary and key secondary end points were analyzed on the
basis of the full analysis set with last observation carried forward In both trials, each brolucizumab arm demonstrated noninferiority
(LOCF) imputation of missing values. Supportive analyses of the versus aflibercept in least squares (LS) mean BCVA change from
primary end point were conducted using the per-protocol set with baseline to Week 48 (Table 1). In HAWK, brolucizumab 3 mge and

74
 Dugel et al 
Brolucizumab for Neovascular AMD

Table 1. Primary End Point and Secondary End Points (Full Analysis Set, Last Observation Carried Forward)

HAWK HARRIER
Brolucizumab 3 mg Brolucizumab 6 mg Aflibercept 2 mg Brolucizumab 6 mg Aflibercept 2 mg
Outcome (N ¼ 358) (N ¼ 360) (N ¼ 360) (N ¼ 370) (N ¼ 369)
Primary end point
Change in BCVA from baseline to
Week 48
Letters, LS mean (SE) 6.1 (0.69) 6.6 (0.71) 6.8 (0.71) 6.9 (0.61) 7.6 (0.61)
LS mean difference (brolucizumab e aflibercept)
Difference (SE) 0.6 (0.98) 0.2 (1.00) d 0.7 (0.86) d
95% CI for treatment difference 2.5 to 1.3 2.1 to 1.8 d 2.4 to 1.0 d
P value for noninferiority (1-sided; margin: 4 letters) <0.001 <0.001 <0.001
Key secondary end point
Average change in BCVA from baseline
over the period of Weeks 36-48
Letters, LS mean (SE) 6.2 (0.67) 6.7 (0.68) 6.7 (0.68) 6.5 (0.58) 7.7 (0.58)
LS mean difference (brolucizumab e aflibercept)
Difference (SE) 0.5 (0.95) 0.0 (0.96) d 1.2 (0.82) d
95% CI for treatment difference 2.4 to 1.3 1.9 to 1.9 d 2.8 to 0.5 d
P value for noninferiority (1-sided; margin: 4 letters) <0.001 <0.001 <0.001
Secondary end point
Patients with 15 letter gain from 25.2 33.6 25.4 29.3 29.9
baseline to Week 48, %

BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; LS ¼ least squares; SE ¼ standard error.

brolucizumab 6 mgetreated eyes gained þ6.1 and þ6.6 letters, and slightly increased further up to Week 48 (Fig 1A and B). The
respectively, versus þ6.8 letters among aflibercept-treated eyes (LS proportion of study eyes that gained 15 letters of vision from
mean; 95% CI for treatment difference, 2.5 to 1.3; P value for baseline to Week 48 was 25.2% (brolucizumab 3 mg), 33.6%
noninferiority <0.001 and 95% CI for treatment difference, 2.1 to (brolucizumab 6 mg), and 25.4% (aflibercept 2 mg) in HAWK and
1.8; P value for noninferiority <0.001, respectively). In HARRIER, 29.3% and 29.9% (brolucizumab 6 mg and aflibercept 2 mg,
brolucizumab 6 mgetreated eyes gained þ6.9 letters versus þ7.6 respectively) in HARRIER.
letters among aflibercept-treated eyes (LS mean; 95% CI for treat-
ment difference, 2.4 to 1.0; P value for noninferiority <0.001). In Every 12-Week Dosing Maintenance Over 48
general, these outcomes were not affected by baseline BCVA or age Weeks
(Fig S3, available at www.aaojournal.org). As an alternative to the
LOCF approach, an analysis of the observed data (based on the For brolucizumab-treated eyes, the probabilities (KaplaneMeier
mixed-model repeated-measures analysis) in HAWK revealed an estimates) for exclusively maintaining q12w dosing after loading
LS mean BCVA change from baseline to Week 48 of þ6.4 (3 mg) through Week 48 were 49.4% (3 mg; 95% CI for KaplaneMeier
and þ6.6 (6 mg) letters with brolucizumab compared with þ7.3 estimate, 43.9% to 54.6%) and 55.6% (6 mg; 95% CI for
letters with aflibercept (95% CI for treatment difference, 2.8 to 1.0; KaplaneMeier estimate, 50.2% to 60.8%) in HAWK, and 51.0% (6
P value for noninferiority <0.001 and 95% CI for treatment mg; 95% CI for KaplaneMeier estimate, 45.7% to 56.1%) in
difference, 2.7 to 1.3; P value for noninferiority <0.001, respec- HARRIER (Fig 1C). Under the condition that a brolucizumab-
tively); in HARRIER, LS mean BCVA change from baseline to treated eye did not show disease activity during the first q12w in-
Week 48 was þ7.2 letters with brolucizumab (6 mg) versus þ7.7 terval, the probabilities for remaining on q12w dosing up to Week 48
letters with aflibercept (95% CI for treatment difference, 2.1 to 1.2; increased to 80.9% (3 mg; 95% CI for KaplaneMeier estimate,
P value for noninferiority <0.001; Table S5A, available at 74.5% to 85.7%) and 85.4% (6 mg; 95% CI for KaplaneMeier es-
www.aaojournal.org). Brolucizumab was also noninferior to timate, 79.9% to 89.5%) in HAWK and 81.7% (6 mg; 95% CI for
aflibercept in LS mean BCVA change from baseline averaged over KaplaneMeier estimate, 75.8% to 86.3%) in HARRIER.
the period of Week 36 through Week 48 in both trials (HAWK:
brolucizumab 3 mg vs. aflibercept 2 mg, þ6.2 vs. þ6.7 letters; Disease Activity Assessment (Week 16,
95% CI for treatment difference, 2.4 to 1.3; P value for Matched) and Anatomic Outcomes
noninferiority <0.001 and brolucizumab 6 mg vs. aflibercept 2
mg, þ6.7 vs. þ6.7 letters; 95% CI for treatment difference, 1.9 Each of the 4 BCVA-related noninferiority hypotheses of HAWK
to 1.9; P value for noninferiority <0.001; HARRIER: reached statistical significance (1-sided P < 0.025); therefore,
brolucizumab 6 mg vs. aflibercept 2 mg, þ6.5 vs. þ7.7 letters; additional confirmatory superiority testing was conducted in
95% CI for treatment difference, 2.8 to 0.5; P value for HAWK to assess the superiority of brolucizumab regarding CST
noninferiority <0.001). Noninferiority of BCVA outcomes was reduction, presence of IRF and/or SRF, and presence of disease
confirmed on the basis of the per-protocol analysis (Table S5A activity at Week 16 (Table S6, available at www.aaojournal.org).
and S5B, available at www.aaojournal.org). In all treatment arms, The period up to Week 16 allowed for a masked, dosing
LS mean BCVA gains were observed during the loading phase frequencyematched (all treatment arms had 3 monthly loading

75
 Ophthalmology Volume 127, Number 1, January 2020

Figure 1. Best-corrected visual acuity (BCVA) over time in (A) HAWK and (B) HARRIER, and (C) KaplaneMeier analysis of every 12 weeks (q12w)
treatment status (time to first every 8 weeks [q8w] need). A, B, Full analysis set; last observation carried forward (LOCF). C, Full analysis set; efficacy/safety
approach (for KaplaneMeier curve). *In the case of informative censoring (lack of efficacy or safety), disease activity/q8w need was imputed (efficacy
approach: in case of lack of efficacy; efficacy/safety approach: in case of lack of efficacy and/or safety). yBased on 220 patients. CI ¼ confidence interval; LS ¼
least squares; q8w ¼ every 8 weeks; q12w ¼ every 12 weeks.

doses followed by 8 weeks before the next possible treatment) 52.2%; 95% CI for treatment difference, 25.3% to 10.9%; P <
assessment. Fewer brolucizumab 6 mgetreated eyes had disease 0.001) and HARRIER (29.4% vs. 45.1%; 95% CI for treatment
activity versus aflibercept in HAWK (24.0% vs. 34.5%; 95% CI for difference, 22.9% to 9.0%; P < 0.001); similar results were
treatment difference, 17.1% to 3.5%; P ¼ 0.001) and HAR- observed at Week 48 in HAWK (3 mg, 34.1% vs. 44.7%; 95% CI
RIER (22.7% vs. 32.2%; 95% CI for treatment difference, 15.8% for treatment difference, 17.4% to 3.3%; P ¼ 0.002 and 6 mg,
to 3.1%; P ¼ 0.002; Fig 2A), thus revealing a formal 31.2% vs. 44.6%; 95% CI for treatment difference, 20.7%
demonstration of superiority versus aflibercept in HAWK to 6.1%; P < 0.001) and HARRIER (25.8% vs. 43.9%; 95% CI
regarding duration of effect. for treatment difference, 24.9% to 11.8%; P < 0.001; Fig 2B),
Greater CST reductions from baseline to Week 16 were with formal demonstration of statistical superiority versus
observed among eyes treated with brolucizumab 6 mg versus aflibercept in HAWK. At Week 48, sub-RPE fluid was present in
aflibercept in HAWK (LS mean; 161.4 vs. 133.6 mm; 95% CI fewer brolucizumab 6 mgetreated eyes than aflibercept-treated
for treatment difference, 45.1 to 10.5; P < 0.001) and HAR- eyes in HAWK (13.5% vs. 21.6%; 95% CI for treatment
RIER (LS mean; 174.4 vs. 134.2 mm; 95% CI for treatment difference, 13.6% to 2.7%; P ¼ 0.004) and HARRIER (12.9%
difference, 58.9 to 21.6; P < 0.001); similar results were vs. 22.0%; 95% CI for treatment difference, 13.8% to 3.9%; P
observed at Week 48 in HAWK (LS mean; 172.8 vs. 143.7 < 0.001; Fig 2C). Analyses of IRF and/or SRF, as well as sub-RPE
mm; 95% CI for treatment difference, 47.6 to 10.4; P ¼ 0.001) fluid presence between Weeks 36 and 48 also supported better fluid
and HARRIER (LS mean; 193.8 vs. 143.9 mm; 95% CI for control with brolucizumab 6 mg (Tables S8 and S9, available at
treatment difference, 68.9 to 30.9; P < 0.001; Fig 3), with www.aaojournal.org).
formal significance demonstrated versus aflibercept in HAWK.
The CST reduction difference from baseline averaged over the Safety
period of Week 36 through Week 48 between brolucizumab 6
mg, and aflibercept was numerically higher for brolucizumab in Brolucizumab was generally well tolerated; overall ocular and
both studies without formally reaching significance in HAWK nonocular adverse event rates were similar to those with aflibercept
(Table S7, available at www.aaojournal.org). within each trial (Table 2). The most common ocular adverse
Intraretinal fluid/SRF was present in fewer brolucizumab- events were conjunctival hemorrhage (brolucizumab 3 and 6 mg;
treated eyes versus aflibercept-treated eyes at Week 16 in HAWK) and reduced visual acuity (aflibercept; HAWK, both
HAWK (3 mg, 41.8% vs. 52.0%; 95% CI for treatment treatments; HARRIER; Table 3). Adverse events of interest
difference, 17.3% to 2.5%; P ¼ 0.003 and 6 mg, 33.9% vs. included uveitis and iritis (2.2% for each) with brolucizumab 6

76
 Dugel et al 
Brolucizumab for Neovascular AMD

Figure 1. Continued

77
 Ophthalmology Volume 127, Number 1, January 2020

Figure 2. (A) Disease activity at Week 16. Full analysis set; analysis conducted based on the efficacy/safety approach. *The 95% confidence interval (CI) for
treatment difference, e13.2 to 0.3; P ¼ 0.033. y95% CI for treatment difference, e17.1 to e3.5; P ¼ 0.001. z95% CI for treatment difference, e15.8 to
e3.1; P ¼ 0.002. 1-sided P values versus aflibercept. (B) Presence of intraretinal (IRF) and/or subretinal fluid (SRF) at Weeks 16 and 48. Full analysis set;
LOCF. *95% CI for treatment difference, e17.3 to e2.5; P ¼ 0.003. y95% CI for treatment difference, e25.3 to e10.9; P < 0.001. z95% CI for treatment
difference, e17.4 to e3.3; P ¼ 0.002. x95% CI for treatment difference, e20.7 to e6.1; P < 0.001. k95% CI for treatment difference, e22.9 to e9.0; P <
0.001. {95% CI for treatment difference, e24.9 to e11.8; P < 0.001. 1-sided P values versus aflibercept. (C) Presence of sub-RPE fluid at Weeks 16 and 48.
Full analysis set; LOCF. * 95% CI for treatment difference, 11.8 to 1.1; P ¼ 0.027. y95% CI for treatment difference, 14.4 to 2.9; P ¼ 0.003. z95% CI
for treatment difference, 9.4 to 1.4; P ¼ 0.15. x95% CI for treatment difference, 13.6 to 2.7; P ¼ 0.004. k95% CI for treatment difference, 13.0 to
2.7; P ¼ 0.004. {95% CI for treatment difference, 13.8 to 3.9; P < 0.001. 2-sided P values vs aflibercept. Aflib ¼ aflibercept; Brol ¼ brolucizumab;
LOCF ¼ last observation carried forward; RPE ¼ retinal pigment epithelium.

mg versus 0.3% and 0%, respectively, with aflibercept in HAWK; nonocular arterial thromboembolic events and death was
corresponding rates in HARRIER were <1% in both arms consistent across treatment arms within each trial (Table 2).
(Table 3). Approximately 90% of the uveitis and iritis cases were Nonocular adverse events and nonocular serious adverse events
mild to moderate and treated with a course of topical are summarized in Tables S10 and S11, respectively (available at
corticosteroids/anti-infectives; most resolved with no sequelae. www.aaojournal.org).
Incidence of increased intraocular pressure was similar with bro-
lucizumab and aflibercept (2.5%e3.2% and 2.2%e2.4%, respec-
tively; Table 3). The incidence of serious ocular adverse events was Discussion
low in both trials; no event occurred in >1% of eyes (Table 4). An
imbalance of uveitis serious adverse events between brolucizumab
and aflibercept was observed in both trials, and an imbalance of HAWK and HARRIER, the phase 3 trials evaluating bro-
endophthalmitis serious adverse events was observed in HAWK; lucizumab on a q12w/q8w regimen versus q8w aflibercept,
however, there was a small number of reports (Table 4). The met the primary end point of noninferiority in BCVA of
proportion of eyes with a 15-letter loss at Week 48 was brolucizumab versus aflibercept, with >50% of brolucizu-
balanced across all treatments (Table 2). The incidence of mab 6 mg patients being maintained on a q12w interval

78
 Dugel et al 
Brolucizumab for Neovascular AMD

Figure 3. Central subfield thickness (CST) over time in (A) HAWK and (B) HARRIER. Full analysis set; LOCF. Aflib ¼ aflibercept; Brol ¼ brolucizumab;
CI ¼ confidence interval; LOCF ¼ last observation carried forward; LS ¼ least squares.

through Week 48. Moreover, anatomic retinal fluid out- by exudation from abnormally growing blood vessels in
comes favored brolucizumab over aflibercept. Overall the macula, resulting in progressive degeneration of the
adverse event rates were generally similar with brolucizu- photoreceptors and RPE.1,33 The course of the disease is
mab and aflibercept. that VEGF increases, causing increased retinal fluid accu-
Neovascular AMD is a variable disease with regard to mulation, which then leads to edema and functional deteri-
individual treatment needs; for example, patients with early oration.34,35 Thus, presence of retinal fluid and increased
persistent retinal fluid have better outcomes with more CST are indicators of disease activity, and disease activity
frequent treatment.32 Therefore, a goal of nAMD can be identified more rapidly through analysis of fluid on
management is to determine therapeutic needs on an OCT compared with BCVA-based indicators.1,33,36 Clinical
individual basis and treat accordingly to achieve an practice guidelines from the American Academy of
optimal visual outcome with minimal clinic visits and Ophthalmology, The Royal College of Ophthalmology, and
intravitreal injection burden. The disease is characterized EURETINA state that fluid on OCT is an indication of

79
 Ophthalmology Volume 127, Number 1, January 2020

Table 2. Safety Summary (Safety Analysis Set)

HAWK HARRIER
Brolucizumab 3 mg Brolucizumab 6 mg Aflibercept 2 mg Brolucizumab 6 mg Aflibercept 2 mg
Adverse Event (N ¼ 358) (N ¼ 360) (N ¼ 360) (N ¼ 370) (N ¼ 369)
Patients with 1 adverse event, n (%)*
Ocular 175 (48.9) 179 (49.7) 170 (47.2) 122 (33.0) 119 (32.2)
Nonocular 242 (67.6) 232 (64.4) 258 (71.7) 219 (59.2) 211 (57.2)
Patients with 1 serious adverse event, total,
n (%)*
Ocular 5 (1.4) 11 (3.1) 3 (0.8) 9 (2.4) 4 (1.1)
Nonocular 47 (13.1) 47 (13.1) 68 (18.9) 35 (9.5) 43 (11.7)
Patients with 15-letter loss from baseline at 5.9 6.4 5.5 3.8 4.8
Week 48, %y
Death, n (%) 4 (1.1) 4 (1.1) 6 (1.7) 3 (0.8) 4 (1.1)
Patients with 1 nonocular arterial 11 (3.1) 6 (1.7) 10 (2.8) 6 (1.6) 8 (2.2)
thromboembolic event, n (%)*

Medical Dictionary for Regulatory Activities version 20.1 has been used for the reporting of adverse events.
*Adverse events with a start date on or after the date of first study treatment administration were counted. A patient with multiple occurrences of an adverse
event for a preferred term or system organ class was counted only once in each specific category.
y
Last observation carried forward.

active disease and recommend retreatment when fluid is treatment has led to the emergence of pro re nata (PRN,
present.1,34,37 or “as needed”) regimens, whereby patients are monitored
Previous investigations of fixed q12w ranibizumab monthly and treated only if signs of active disease are
dosing without interval adjustment for disease activity present.38,39 Clinical trial data in CATT and HARBOR
showed inferiority of visual acuity outcomes to monthly demonstrated effective BCVA improvements with PRN
dosing.25 The resulting need to provide individualized treatment;38,39 however, the monthly monitoring need does

Table 3. Ocular Adverse Events by Preferred Term in Study Eye (2% of Eyes in any Treatment Group of any Study; Safety Analysis Set)

HAWK HARRIER
Brolucizumab 3 mg Brolucizumab 6 mg Aflibercept 2 mg Brolucizumab 6 mg Aflibercept 2 mg
Preferred Term, n (%)* (N ¼ 358) (N ¼ 360) (N ¼ 360) (N ¼ 370) (N ¼ 369)
Patients with 1 event 175 (48.9) 179 (49.7) 170 (47.2) 122 (33.0) 119 (32.2)
Conjunctival hemorrhage 30 (8.4) 23 (6.4) 20 (5.6) 7 (1.9) 12 (3.3)
Visual acuity reduced 23 (6.4) 19 (5.3) 24 (6.7) 20 (5.4) 20 (5.4)
Vitreous floaters 24 (6.7) 18 (5.0) 11 (3.1) 11 (3.0) 3 (0.8)
Eye pain 21 (5.9) 16 (4.4) 15 (4.2) 10 (2.7) 12 (3.3)
Dry eye 11 (3.1) 14 (3.9) 15 (4.2) 8 (2.2) 6 (1.6)
Retinal hemorrhage 10 (2.8) 13 (3.6) 16 (4.4) 5 (1.4) 2 (0.5)
Retinal pigment epithelial tear 5 (1.4) 12 (3.3) 4 (1.1) 6 (1.6) 4 (1.1)
Vitreous detachment 16 (4.5) 10 (2.8) 13 (3.6) 7 (1.9) 5 (1.4)
Eye irritation 8 (2.2) 10 (2.8) 8 (2.2) 3 (0.8) 1 (0.3)
Intraocular pressure increased 11 (3.1) 9 (2.5) 8 (2.2) 12 (3.2) 9 (2.4)
Posterior capsule opacification 5 (1.4) 9 (2.5) 7 (1.9) 5 (1.4) 1 (0.3)
Uveitis 5 (1.4) 8 (2.2) 1 (0.3) 3 (0.8) 0
Blepharitis 4 (1.1) 8 (2.2) 7 (1.9) 8 (2.2) 3 (0.8)
Iritis 1 (0.3) 8 (2.2) 0 0 1 (0.3)
Cataract 10 (2.8) 7 (1.9) 8 (2.2) 4 (1.1) 12 (3.3)
Visual field defect 7 (2.0) 7 (1.9) 3 (0.8) 1 (0.3) 0
Conjunctivitis 2 (0.6) 7 (1.9) 3 (0.8) 10 (2.7) 3 (0.8)
Vision blurred 11 (3.1) 6 (1.7) 5 (1.4) 1 (0.3) 2 (0.5)
Visual impairment 10 (2.8) 6 (1.7) 10 (2.8) 0 2 (0.5)
Punctate keratitis 5 (1.4) 6 (1.7) 8 (2.2) 1 (0.3) 3 (0.8)
Corneal abrasion 5 (1.4) 5 (1.4) 8 (2.2) 0 1 (0.3)
Lenticular opacities 6 (1.7) 0 3 (0.8) 8 (2.2) 7 (1.9)

Medical Dictionary for Regulatory Activities version 20.1 has been used for the reporting of adverse events.
*Adverse events with a start date on or after the date of first study treatment administration were counted. A patient with multiple occurrences of an adverse
event for a preferred term or system organ class was counted only once in each specific category.

80
 Dugel et al 
Brolucizumab for Neovascular AMD

Table 4. Ocular Serious Adverse Events by Preferred Term in Study Eye (Safety Analysis Set)

HAWK HARRIER
Brolucizumab 3 mg Brolucizumab 6 mg Aflibercept 2 mg Brolucizumab 6 mg Aflibercept 2 mg
Preferred Term, n (%)* (N ¼ 358) (N ¼ 360) (N ¼ 360) (N ¼ 370) (N ¼ 369)
Patients with 1 event 5 (1.4) 11 (3.1) 3 (0.8) 9 (2.4) 4 (1.1)

Uveitis 1 (0.3) 2 (0.6) 0 3 (0.8) 0

Retinal detachment 1 (0.3) 1 (0.3) 0 0 1 (0.3)
Retinal pigment epithelial tear d d d 2 (0.5) 0
Visual acuity reduced 0 1 (0.3) 2 (0.6) 1 (0.3) 1 (0.3)
Macular hole 0 1 (0.3) 1 (0.3) d d
Cataract 0 1 (0.3) 0 d d
Retinal artery embolism d d d 1 (0.3) 0
Retinal artery occlusion 2 (0.6) 0 0 0 1 (0.3)
Retinal artery thrombosis 0 1 (0.3) 0 1 (0.3) 0
Retinal depigmentation 0 1 (0.3) 0 d d
Retinopathy proliferative 0 1 (0.3) 0 d d
Vitritis 0 1 (0.3) 0 d d
Anterior chamber inflammation d d d 1 (0.3) 0
Dry age-related macular degeneration d d d 0 1 (0.3)

Endophthalmitis 3 (0.8) 2 (0.6) 0 1 (0.3) 0

Cataract traumatic d d d 1 (0.3) 0

Medical Dictionary for Regulatory Activities version 20.1 has been used for the reporting of adverse events.
*Serious adverse events with a start date on or after the date of first study treatment administration were counted. A patient with multiple occurrences of an
adverse event for a preferred term or system organ class was counted only once in each specific category. A dash indicates the event was not reported in the
trial.

not alleviate the overall burden to patients, clinics, and the treatment course and may offer a novel and reliable para-
healthcare system. digm for efficient and individualized long-term nAMD
As an alternative to PRN, “treat-and-extend” regimens management.
gradually extend treatment intervals in patients without On the basis of this treatment concept of assessment
active disease, and the approach has been investigated during the initial q12w cycle and potential adjustments at
recently in randomized trials using ranibizumab.40,41 scheduled q12w injection visits, the probability for main-
Standard “treat-and-extend” regimens in these studies taining on q12w dosing throughout Year 1 was estimated to
extend the treatment interval in 2-week increments, thereby be >50% for eyes treated with brolucizumab 6 mg. Eyes
requiring 36 weeks of successive extensions postloading to treated with a maintenance regimen of q12w dosing corre-
observe the initial q12w interval and, on average, result in sponds to a reduction of 2 injections per year compared with
9 to 10 injections in the first year.40,41 Thus, a treatment a q8w maintenance regimen.
providing comparable efficacy to fixed dosing but in a In both HAWK and HARRIER, disease activity was
regimen in which the monitoring and dosing interval is assessed at each disease activity assessment visit. At the
based on an individual’s anti-VEGF need soon after matched Week 16 assessment, corresponding to 8 weeks
loading may alleviate the treatment burden associated with after completion of the loading phase in all patients, fewer
nAMD. brolucizumab 6 mgetreated eyes had disease activity versus
HAWK and HARRIER are the first multinational nAMD aflibercept in both studies, with formal demonstration of
registration trials to use masked investigator identification of superiority in HAWK, suggesting a prolonged duration of
disease activity after the loading phase to identify a suitable effect. In both studies, this advantage of brolucizumab 6 mg
maintenance dose interval based on individual treatment versus aflibercept was also reflected in the anatomic as-
need. This approach differs from previous studies evaluating sessments at Week 16, again with formal demonstration of
q12w dosing intervals24,25 by providing the opportunity for superiority in HAWK, regarding reductions of CST and
masked physicians to adjust to q8w dosing during the study, presence of IRF and/or SRF. These advantages in anatomic
if needed. As a result, q8w allocation was not randomized parameters support the underlying hypothesis that a lower
but driven by disease activity, making comparative analyses molecular weight combined with a higher concentration
of eyes treated exclusively with a q12w interval versus eyes gradient between the vitreous and retina increase the drug
adjusted to a q8w interval not valid. Time-to-event analyses distribution to the target site, resulting in more effective
revealed that most q8w treatment need was identified during control of anatomic disease activity. Collectively, the data
the first q12w interval (Weeks 16 and 20). Thus, these data suggest greater treatment duration and thus reduced treat-
support the predictive value of dynamic changes early in the ment need with brolucizumab.

81
 Ophthalmology Volume 127, Number 1, January 2020

In previous nAMD registration trials, the mean visual 5. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for
acuity improvements after 1 year of treatment were 6.5e7.2 neovascular age-related macular degeneration. N Engl J Med.
(ranibizumab, MARINA), 8.5e11.3 (ranibizumab, AN- 2006;355:1419e1431.
CHOR), and 6.9e10.9 (aflibercept, VIEW) letters compared 6. Heier JS, Brown DM, Chong V, et al. Intravitreal aflibercept
with 6.1e6.9 letters with brolucizumab (current report).4e6 (VEGF trap-eye) in wet age-related macular degeneration.
Ophthalmology. 2012;119:2537e2548.
The difference in the magnitude in BCVA change be- 7. Freund KB, Mrejen S, Gallego-Pinazo R. An update on the
tween the present trials and previous registration trials can pharmacotherapy of neovascular age-related macular
be explained by differences in baseline BCVA. The higher degeneration. Expert Opin Pharmacother. 2013;14:
baseline BCVA value (resulting from the upper limit of 78 1017e1028.
letters for the BCVA inclusion criterion) compared with 8. Boulanger-Scemama E, Querques G, About F, et al. Ranibi-
previous registrational trials (upper limit of 73 letters in the zumab for exudative age-related macular degeneration: a five
ANCHOR, MARINA, and VIEW 1/2 trials4e6) is in line year study of adherence to follow-up in a real-life setting. J Fr
with current disease management. Ophtalmol. 2015;38:620e627.
The 48-week results of the trials showed robust visual 9. Kiss S, Liu Y, Brown J, et al. Clinical monitoring of patients
acuity gains with brolucizumab dosed with a q12w/q8w with age-related macular degeneration treated with intravitreal
bevacizumab or ranibizumab. Ophthalmic Surg Lasers Imag-
regimen that were noninferior to aflibercept dosed q8w, ing Retina. 2014;45:285e291.
while >50% of brolucizumab 6 mgetreated eyes were 10. Ehlken C, Helms M, Bohringer D, et al. Association of
estimated to maintain on q12w dosing immediately after the treatment adherence with real-life VA outcomes in AMD,
loading phase through Week 48. Superior anatomic out- DME, and BRVO patients. Clin Ophthalmol. 2017:
comes regarding retinal fluid and retinal thickness with 1213e1220.
brolucizumab 6 mg versus aflibercept could be concluded 11. Kim LN, Mehta H, Barthelmes D, et al. Metaanalysis of real-
from HAWK and HARRIER at Weeks 16 and 48 in both world outcomes of intravitreal ranibizumab for the treatment of
studies. The predictive value of the behavior during the first neovascular age-related macular degeneration. Retina.
q12w interval allows physicians to confidently determine 2016;36:1418e1431.
which patients are suitable to continue on q12w dosing. 12. Holz FG, Schmitz-Valckenberg S, Fleckenstein M. Recent
developments in the treatment of age-related macular degen-
Overall safety of brolucizumab was similar to aflibercept eration. J Clin Invest. 2014;124:1430e1438.
and consistent with other antieVEGF-A agents approved 13. Holz FG, Tadayoni R, Beatty S, et al. Multi-country real-life
for nAMD treatment.6,42e44 The 96-week results will pro- experience of anti-vascular endothelial growth factor therapy
vide additional insight into the safety and efficacy of q12w/ for wet age-related macular degeneration. Br J Ophthalmol.
q8w brolucizumab versus that of q8w aflibercept. 2015;99:220e226.
In conclusion, the HAWK and HARRIER studies suc- 14. Dugel PU, Jaffe GJ, Sallstig P, et al. Brolucizumab versus
cessfully evaluated an alternative treatment option, aflibercept in participants with neovascular age-related macu-
combining the prolonged duration of effect of brolucizumab lar degeneration: a randomized trial. Ophthalmology.
with an individualized treatment regimen, allowing for 2017;124:1296e1304.
favorable efficacy, effective treatment scheduling, and 15. Holz FG, Dugel PU, Weissgerber G, et al. Single-chain anti-
body fragment VEGF inhibitor RTH258 for neovascular age-
minimal monitoring burden. related macular degeneration: a randomized controlled study.
Ophthalmology. 2016;123:1080e1089.
Acknowledgments 16. TBC. Brolucizumab: a unique single-chain antibody fragment
The authors thank Aaron Runkle, PhD, and Joelle Suchy, PhD inhibitor of vascular endothelial growth factor.
(Nucleus Global, Hamilton, NJ), for assistance with medical 17. Escher D, Schmidt A, Steiner P, et al. Single-chain antibody
writing (funded by Novartis Pharma AG). fragments in ophthalmology. Presented at: 15th EURETINA
Congress; September 17e20, 2015; Nice, France.
18. Miller BR, Demarest SJ, Lugovskoy A, et al. Stability engi-
neering of scFvs for the development of bispecific and multi-
References valent antibodies. Protein Eng Des Sel. 2010;23:549e557.
19. Nelson AL. Antibody fragments: hope and hype. MAbs.
2010;2:77e83.
1. Schmidt-Erfurth U, Chong V, Loewenstein A, et al. Guidelines 20. Huston JS, Levinson D, Mudgett-Hunter M, et al. Protein
for the management of neovascular age-related macular engineering of antibody binding sites: recovery of specific
degeneration by the European Society of Retina Specialists activity in an anti-digoxin single-chain fv analogue produced
(EURETINA). Br J Ophthalmol. 2014;98:1144e1167. in escherichia coli. Proc Natl Acad Sci U S A. 1988;85:
2. Wong WL, Su X, Li X, et al. Global prevalence of age-related 5879e5883.
macular degeneration and disease burden projection for 2020 21. Bird RE, Hardman KD, Jacobson JW, et al. Single-chain
and 2040: a systematic review and meta-analysis. Lancet Glob antigen-binding proteins. Science. 1988;242:423e426.
Health. 2014;2:e106ee116. 22. Brown DM, Tuomi L, Shapiro H, et al. Anatomical measures
3. Yonekawa Y, Miller JW, Kim IK. Age-related macular as predictors of visual outcomes in ranibizumab-treated eyes
degeneration: advances in management and diagnosis. J Clin with neovascular age-related macular degeneration. Retina.
Med. 2015;4:343e359. 2013;33:23e34.
4. Brown DM, Kaiser PK, Michels M, et al. Ranibizumab versus 23. Eldem MB, Bartz-Schmidt KU, Schlingemann RO, et al. Vi-
verteporfin for neovascular age-related macular degeneration. sual acuity response profiles in patients with neovascular age-
N Engl J Med. 2006;355:1432e1444. related macular degeneration treated quarterly with

82
 Dugel et al 
Brolucizumab for Neovascular AMD

ranibizumab in the EXCITE trial. Invest Ophthalmol Vis Sci. rcophth.ac.uk/wp-content/uploads/2014/12/2013-SCI-318-

2009;50:2374. RCOphth-AMD-Guidelines-Sept-2013-FINAL-2.pdf. Pub-
24. Regillo CD, Brown DM, Abraham P, et al. Randomized, lished September 2013. Accessed November 28, 2018.
double-masked, sham-controlled trial of ranibizumab for 35. Campochiaro PA, Soloway P, Ryan SJ, et al. The pathogenesis
neovascular age-related macular degeneration: PIER study of choroidal neovascularization in patients with age-related
year 1. Am J Ophthalmol. 2008;145:239e248. macular degeneration. Mol Vis. 1999;5:34.
25. Schmidt-Erfurth U, Eldem B, Guymer R, et al. Efficacy and 36. Arnold JJ, Markey CM, Kurstjens NP, et al. The role of sub-
safety of monthly versus quarterly ranibizumab treatment in retinal fluid in determining treatment outcomes in patients
neovascular age-related macular degeneration: the EXCITE with neovascular age-related macular degeneration–a phase IV
study. Ophthalmology. 2011;118:831e839. randomised clinical trial with ranibizumab: the FLUID study.
26. Chong V, Amoaku W, Alsop J, et al. Effectiveness of BMC Ophthalmol. 2016;16:31.
quarterly (Q3M) versus monthly (QM) ranibizumab ac- 37. American Academy of Ophthalmology. Age-related macular
cording to initial gains in visual acuity - an analysis of 12 degeneration: preferred practice pattern. https://www.aao.org/
month data from the EXCITE study. Invest Ophthalmol Vis Assets/db935a77-1997-4d60-b850-71b7602f46e2/635582143
Sci. 2013;54:3816. 853270000/age-related-macular-degeneration-ppp-pdf. Upda-
27. Kim BJ, Ying GS, Huang J, et al. Sporadic visual acuity ted January 2015. Accessed November 28, 2018.
loss in the Comparison of Age-Related Macular Degenera- 38. Busbee BG, Ho AC, Brown DM, et al. Twelve-month efficacy
tion Treatments Trials (CATT). Am J Ophthalmol. and safety of 0.5 mg or 2.0 mg ranibizumab in patients with
2014;158:128e135.e10. subfoveal neovascular age-related macular degeneration.
28. Ying GS, Kim BJ, Maguire MG, et al. Sustained visual Ophthalmology. 2013;120:1046e1056.
acuity loss in the comparison of age-related macular 39. CATT Research Group, Martin DF, Maguire MG, et al.
degeneration treatments trials. JAMA Ophthalmol. 2014;132: Ranibizumab and bevacizumab for neovascular age-related
915e921. macular degeneration. N Engl J Med. 2011;364:
29. Simader C, Ritter M, Bolz M, et al. Morphologic parameters 1897e1908.
relevant for visual outcome during anti-angiogenic therapy of 40. Wykoff CC, Croft DE, Brown DM, et al. Prospective trial of
neovascular age-related macular degeneration. Ophthal- treat-and-extend versus monthly dosing for neovascular age-
mology. 2014;121:1237e1245. related macular degeneration: TREX-AMD 1-year results.
30. EYLEA (aflibercept) Injection [package insert]. Tarrytown, Ophthalmology. 2015;122:2514e2522.
NY: Regeneron Pharmaceuticals, Inc; July 2014. 41. Silva R, Berta A, Larsen M, et al. Treat-and-extend versus
31. Mantel I, Deli A, Iglesias K, et al. Prospective study monthly regimen in neovascular age-related macular degen-
evaluating the predictability of need for retreatment with eration: results with ranibizumab from the TREND study.
intravitreal ranibizumab for age-related macular degenera- Ophthalmology. 2018;125:57e65.
tion. Graefes Arch Clin Exp Ophthalmol. 2013;251: 42. Kitchens JW, Do DV, Boyer DS, et al. Comprehensive review
697e704. of ocular and systemic safety events with intravitreal afli-
32. Jaffe GJ, Kaiser PK, Thompson D, et al. Differential response bercept injection in randomized controlled trials. Ophthal-
to anti-VEGF regimens in age-related macular degeneration mology. 2016;123:1511e1520.
patients with early persistent retinal fluid. Ophthalmology. 43. EYLEA (aflibercept) Injection [prescribing information]. Tar-
2016;123:1856e1864. rytown, NY: Regeneron Pharmaceuticals, Inc; 2017.
33. Ambati J, Fowler BJ. Mechanisms of age-related macular 44. European Medicines Agency. Eylea (aflibercept) assessment
degeneration. Neuron. 2012;75:26e39. report. https://www.ema.europa.eu/documents/assessment-report/
34. The Royal College of Ophthalmologists. Age-related macular eylea-epar-public-assessment-report_en.pdf. Published September
degeneration: guidelines for management. https://www. 20, 2012. Accessed June 7, 2018.

Footnotes and Financial Disclosures

Originally received: December 3, 2018. *A list of members of the HAWK and HARRIER Study Investigators is
Final revision: February 20, 2019. available online in Supplemental Materials (www.aaojournal.org).
Accepted: April 9, 2019. Financial Disclosure(s):
Available online: April 12, 2019. Manuscript no. 2018-2572. The author(s) have made the following disclosure(s): P.U.D.: Scientific
1
Retinal Consultants of Arizona, Phoenix, Arizona; University of Southern advisory boards e Alcon Surgical (RACII), Genentech, MacuSight,
California, Los Angeles, California. Novartis, NeoVista, ArticDX, Alcon Pharmaceutical, AMO, Thrombo-
2 genics, Santen, Ophthotech, Lux BioScience, Digisight, Roche, Acucela,
Eye & Retina Surgeons, Singapore.
Stealth Biotherapeutics, Lutronic, Avalanche, TrueVision, Alimera Sci-
3
Nagoya City University Graduate School of Medical Sciences, Nagoya, ences, Orbis International, Annidis, Neurotech, Aerpio, DOSE Medical,
Japan. Omeros, Shire Human Genetics, Opthea, Graybug Vision, CDR-Life Inc.,
4
Duke Eye Center, Durham, North Carolina. Clearside Biomedical; Consultancy e Bausch & Lomb Pharma, Genentech,
5 Alcon Surgical, Alcon Pharmaceutical, NeoVista, MacuSight, ArticDx,
Medical University of Vienna, Vienna, Austria.
ORA, Novartis, Allergan, Santen, Inc., Thrombogenics, Ophthotech, Lux
6
Retina Consultants of Houston, Houston, Texas. BioScience, DigiSight, Roche, TopCon, Acucela, Pentavision, ORA,
7 Stealth Biotherapeutics, Annidis, Clearside Biomedical, Optovue, Penta-
University of Sao Paulo, Sao Paulo, Brazil.
8 vision, Neurotech, Lutronic, Alimera Sciences, DOSE Medical, Aerpio,
Novartis Pharma AG, Basel, Switzerland.
Omeros, Shire Human Genetics, Opthea, Spark Thereapeutics, Graybug
9
University of Bonn, Bonn, Germany. Vision, Zeiss Group, Irenix, ByeOnics, Clearside Biomedical, PanOptica,
Presented at: the American Academy of Ophthalmology Annual Meeting, Chengdu Kanghong Biotechnology, SciFluor Life Sciences, Boehringer
November 11e14, 2017, New Orleans, Louisiana. Ingelheim, Kodiak Sciences Oculis SA, pSivida Corporation Amgen, Aerie

83
 Ophthalmology Volume 127, Number 1, January 2020
Pharmaceutical; Stock e Alimera Sciences, Aerpio, Annidis, ArctixDx, HUMAN SUBJECTS: Human subjects were included in this study. All
Digisight, Irenix, Ophthotech, Clearside Biomedical, PanOptica. patients provided written informed consent prior to screening or initiation of
Y.O.: Grants and personal fees e Novartis Pharma, during the conduct of any study related procedures. Protocols were approved by an Independent
the study; Personal fees e Bayer, Senju, Kowa, Wakamoto, Hoya; Grants Ethics Committee/Institutional Review Board. Trials were conducted in
and personal fees e Santen, outside the submitted work. accordance with tenets of the Declaration of Helsinki, International Con-
G.J.J.: Research funding e Alcon/Novartis, during the conduct of the study; ference on Harmonization E6 Good Clinical Practice Consolidated Guide-
line, and other regulations, as applicable, and were compliant with the
Personal fees e Sanofi, Heidelberg Engineering, Novartis, pSivida,
Regeneron, outside the submitted work. Health Insurance Portability and Accountability Act of 1996.
U.S.-E.: Grants e Novartis during the conduct of the study; Personal fees e No animal subjects were used in this study.
Genentech, Novartis, Boehringer, Roche, outside the submitted work. Author Contributions:
D.B.: Grants and personal fees e Alcon, during the conduct of the study; Conception and design: Warburton, Weichselberger
Grants and personal fees e Novartis, Regeneron, Clearside Biomedical, Data collection: Dugel, Koh, Ogura, Jaffe, Schmidt-Erfurth, Brown,
Santen, Allergan, Samsung, Thrombogenics, Heidelberg, Chengdu Kan- Gomes, Warburton, Weichselberger, Holz
ghong Biotechnology Co., Ltd, Adverum, Regenxbio, OHR, Tyrogenix; Analysis and interpretation: Dugel, Koh, Ogura, Jaffe, Schmidt-Erfurth,
Personal fees e Bayer, Aerpio, Optos, Google/Verily, Carl Zeiss Meditec, Brown, Gomes, Warburton, Weichselberger, Holz
Coda Therapeutics, Janssen, Johnson & Johnson, Notal Vision, Optovue,
Obtained funding: N/A
Pfizer, Senju Pharmaceuticals, Stealth Biotherapeutics; Grants e Oph-
thotech, National Eye Institute, Allegro, outside the submitted work. Overall responsibility: Dugel, Koh, Ogura, Jaffe, Schmidt-Erfurth, Brown,
Gomes, Warburton, Weichselberger, Holz
A.V.G.: Consultant e Alcon, during the conduct of the study; Consultant e
Novartis, Allergan, Bayer, Roche, outside the submitted work. Abbreviations and Acronyms:
J.W.: Full-time employee e Novartis Pharmaceuticals; Patent e AMD ¼ age-related macular degeneration; BCVA ¼ best-corrected visual
PAT056526-US-PSP pending. acuity; CI ¼ confidence interval; CST ¼ central subfield thickness;
IRF ¼ intraretinal fluid; IRT ¼ Interactive Response Technology;
A.W.: Employee e Novartis Pharmaceuticals. LOCF ¼ last observation carried forward; LS ¼ least squares;
F.H.: Grants and personal fees e Alcon/Novartis, during the conduct of the nAMD ¼ neovascular age-related macular degeneration; PRN ¼ pro re
study; Consultant e Heidelberg Engineering, Zeiss, Acucela, Genentech/ nata; q8w ¼ every 8 weeks; q12w ¼ every 12 weeks; RPE ¼ retinal
Roche, Allergan, Boehringer-Ingelheim, Bayer Healthcare, LIN Bioscience, pigment epithelium; scFv ¼ single-chain antibody fragments;
Pixium; Grants/grants pending paid to his institution e Nightstar, Optos, SRF ¼ subretinal fluid; VEGF-A ¼ vascular endothelial growth factor A.
Heidelberg Engineering, Carl Zeiss Meditec, Allergan, Roche/Genentech,
Correspondence:
Pixium; Lectures fees including service on speakers bureaus e Roche/
Pravin U. Dugel, MD, Retinal Consultants of Arizona, Roski Eye Institute,
Genentech, Zeiss, Heidelberg Engineering, Bayer Healthcare, outside the
Keck School of Medicine-University of Southern California. E-mail: pdugel@
submitted work.
gmail.com.
Financial support was provided by Novartis Pharma AG (Basel,
Switzerland). The sponsor or funding organization participated in the
design of the study; management, analysis, and interpretation of the data;
preparation, review, and approval of the manuscript.

84
Development and Validation of Deep
Learning Models for Screening Multiple
Abnormal Findings in Retinal Fundus Images
Jaemin Son, MSc,1,* Joo Young Shin, MD, MSc,2,* Hoon Dong Kim, MD, MSc,3 Kyu-Hwan Jung, PhD,1
Kyu Hyung Park, MD, PhD,4 Sang Jun Park, MD, MSc4

Purpose: To develop and evaluate deep learning models that screen multiple abnormal findings in retinal
fundus images.
Design: Cross-sectional study.
Participants: For the development and testing of deep learning models, 309 786 readings from 103 262
images were used. Two additional external datasets (the Indian Diabetic Retinopathy Image Dataset and
e-ophtha) were used for testing. A third external dataset (Messidor) was used for comparison of the models with
human experts.
Methods: Macula-centered retinal fundus images from the Seoul National University Bundang Hospital
Retina Image Archive, obtained at the health screening center and ophthalmology outpatient clinic at Seoul
National University Bundang Hospital, were assessed for 12 major findings (hemorrhage, hard exudate, cotton-
wool patch, drusen, membrane, macular hole, myelinated nerve fiber, chorioretinal atrophy or scar, any vascular
abnormality, retinal nerve fiber layer defect, glaucomatous disc change, and nonglaucomatous disc change) with
their regional information using deep learning algorithms.
Main Outcome Measures: Area under the receiver operating characteristic curve and sensitivity and
specificity of the deep learning algorithms at the highest harmonic mean were evaluated and compared with the
performance of retina specialists, and visualization of the lesions was qualitatively analyzed.
Results: Areas under the receiver operating characteristic curves for all findings were high at 96.2% to
99.9% when tested in the in-house dataset. Lesion heatmaps highlight salient regions effectively in various
findings. Areas under the receiver operating characteristic curves for diabetic retinopathy-related findings tested
in the Indian Diabetic Retinopathy Image Dataset and e-ophtha dataset were 94.7% to 98.0%. The model
demonstrated a performance that rivaled that of human experts, especially in the detection of hemorrhage, hard
exudate, membrane, macular hole, myelinated nerve fiber, and glaucomatous disc change.
Conclusions: Our deep learning algorithms with region guidance showed reliable performance for detection
of multiple findings in macula-centered retinal fundus images. These interpretable, as well as reliable, classifi-
cation outputs open the possibility for clinical use as an automated screening system for retinal fundus
images. Ophthalmology 2020;127:85-94 ª 2019 by the American Academy of Ophthalmology. This is an open
access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Supplemental material available at www.aaojournal.org.
See Commentary on page 95.

Macula-centered retinal fundus images may be used for found on fundus examination, a deep learning algorithm that
screening potential vision-threatening conditions, including identifies multiple disease conditions may be more ideal for
diabetic retinopathy (DR),1e3 age-related macular degener- clinical application. Also, these deep learning algorithms do
ation (AMD), and glaucoma.4 To maximize accessibility not reveal how the decisions are made for the diagnoses,
and mitigate cost, automatic algorithms have been limiting interpretation of the outputs of these algorithms and
developed in the past decade to streamline the process for discouraging potential clinical use. Ophthalmologists usu-
the diagnosis of DR5,6 and glaucoma.7e9 Recently, deep ally determine diagnoses in retinal fundus images by
neural networks10 have revolutionized the field of medical observing certain findings (e.g., hemorrhage, exudate,
image analysis, and the diagnoses of DR, AMD, and cotton-wool patches, etc.) that are associated commonly
possible glaucoma with these deep learning algorithms with the diagnosis (e.g., DR, glaucoma, etc.). This stepwise
have demonstrated discriminative abilities comparable process is not embedded explicitly in deep learning algo-
with those of ophthalmologists.11e20 However, because a rithms that are trained in an end-to-end manner to generate
diverse spectrum of abnormal findings and diseases can be outputs regarding diagnoses directly from an input image.

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.05.029 85

This is an open access article under the CC BY-NC-ND license ISSN 0161-6420/19
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Published by Elsevier Inc.
 Ophthalmology Volume 127, Number 1, January 2020

Previous deep learning algorithms have focused on reaching hemorrhage only with simple fundus imaging, were also included
the proper diagnosis rather than drawing the diagnosis from in the category of hemorrhage. We collected 3 independent
detection of specific findings. Deep learning algorithms that readings from 3 different readers for each image. Details on the
are able to detect multiple abnormalities within a fundus annotation process and reading system are presented in a
previous publication elsewhere.21
image with considerable accuracy while providing the basis
After completion of annotation, all datasets from the 3 reading
of the decisions to the user in an interpretable way may systems were aggregated regardless of the subspecialties of the
allow easier clinical application. readers. Then, 10% of the aggregate dataset was set aside for the
In this article, we present our deep learning algorithms test dataset, and the remaining 90% was divided again into the
that screen for and show lesion heatmaps of multiple training and validation sets in a 9:1 ratio. The training set was used
abnormal findings (hemorrhage, hard exudate, cotton-wool to optimize learnable weights of the network, whereas the valida-
patch, drusen, membrane, macular hole, myelinated nerve tion set was used to tune hyperparameters of the network. In the
fiber, chorioretinal atrophy or scar, any vascular abnormal- training and validation sets, positive labels were given for each
ity, retinal nerve fiber layer [RNFL] defect, glaucomatous finding based on the majority rule (at least 2 of 3 readers noted the
disc change, and nonglaucomatous disc change) on retina target finding) and negative labels based on the unanimity rule
(all 3 readers found no abnormality regarding the target funding).
fundus images. We evaluated the algorithms in external
This labeling scheme was designed to ensure that fundus images
datasets as well as in our in-house dataset under various labeled negative are unlikely to include the target finding, whereas
experimental settings to identify possible influence of the images labeled positive would surely have the target finding.
acquisition protocol and annotation quality on the perfor- For the test dataset, along with this majority rule test set (2 or more
mance of the model. readers noted the target finding), another separate test dataset was
constructed by assigning positive labels with the unanimity rule (all
3 readers agreed on the existence of the target finding) to collect
Methods less controversial cases. Region information was not considered in
labeling the images with the presence of findings; that is, an image
Datasets was considered positive for a certain finding regardless of the
location of that finding. All images not meeting the positive or
This study was approved by the institutional review board at Seoul negative results criteria were discarded in the experiments. The
National Bundang Hospital (institutional review board identifier, final number of images in the in-house dataset used for the training,
B-1508-312-107) and was conducted in accordance with the tenets validation, and 2 test settings is described in Table 1.
of the Declaration of Helsinki. Written consent was waived by the We also used external datasetsdthe Indian Diabetic Retinop-
institutional review board because of the retrospective nature of the athy Image Dataset (IDRiD; https://idrid.grand-challenge.org/
study, conducted on a preconstructed de-identified dataset. We Rules/)22 and e-ophtha dataset23dto evaluate the performance of
used macula-centered retinal fundus images from the Seoul Na- the algorithms outside of our in-house dataset. The IDRiD data-
tional University Bundang Hospital Retina Image Archive, which set consists of 144 images collected in India with a Kowa VX-10a
contains images obtained at the health screening center and fundus camera with a 50
field of view and provides annotations
ophthalmology outpatient clinic at Seoul National University for findings related to DR (microaneurysm, hemorrhage, hard
Bundang Hospital obtained from June 1, 2003, through June 30, exudate, cotton-wool patch). The e-ophtha dataset was collected
2016, using various fundus cameras (CF60Uvi and CR6-45NM from a telemedical DR screening system, which provides annota-
[Canon, Utsunomiya, Japan]; VX-10, VX-10a, nonmyd 7, and tions of microaneurysm and exudates annotated by specialists in
GENESIS-D [Kowa Optimed, Tokyo, Japan]). All images were de- 434 images. Microaneurysms were not discriminated from dot
identified by randomly assigning 10 digits to patient identifications hemorrhages in our database; thus, microaneurysms in the IDRiD
and removing all personal data (e.g., patient name, birth date, and and e-ophtha datasets were subsumed to hemorrhages in the
study date) except gender and age at the study date. We recruited evaluation.
57 licensed ophthalmologists (including 16 certified retina spe-
cialists, 9 certified glaucoma specialists, and 3 certified cornea Development of the Deep Learning Algorithms
specialists) as readers of macular-centered retinal fundus images.
We assigned 3 independent reading systems to readers in accor- We used a deep learning architecture that generates both a classi-
dance with their subspecialties: (1) general ophthalmologists and fication result regarding the presence of a target finding and a
cornea specialists, (2) retina specialists, and (3) glaucoma spe- heatmap that highlights lesions.24 Output of the classification result
cialists. They annotated observed abnormal findings (hemorrhage, ranges from 0 to 1, which corresponds to the predicted probability
hard exudate, cotton-wool patch, drusen, membrane, macular hole, of the existence of the finding. A lesion heatmap is a single-channel
myelinated nerve fiber, chorioretinal atrophy or scar, any vascular low-resolution image whose pixel values are normalized between
abnormality, RNFL defect, glaucomatous disc change, and non- 0 and 1. When this low-resolution image is superimposed onto the
glaucomatous disc change) and their corresponding lesion location original fundus image after resizing, lesions identified by the al-
information according to 8 nonoverlapping regions (macular, gorithms as positive for the target finding are highlighted with
temporal, superior temporal, inferior temporal, superior nasal, high-intensity values. The most important aspect of our deep
inferior nasal, superior disc, and inferior disc areas) using the learning algorithms is that the classification result is correlated
reading system previously described elsewhere.21 The findings directly with the intensity values in the lesion heatmap. A fundus
were selected based on the frequency observed in primary care image is classified to contain the target finding only when the
clinics and health screening centers rather than in clinics of eye cumulative intensity values detected in the lesion heatmap reach a
care professionals and were devised to include most clinically certain threshold. This way, the activation map shows the location
significant findings while maintaining simplicity. For example, on which the decision of the algorithms is based. A single network
all hemorrhages, whether preretinal, intraretinal, subretinal, and is dedicated to classifying a single finding; thus, 12 neural net-
so forth, were included in 1 category of “hemorrhage.” works exist in total in our screening system. Details of the deep
Microaneurysms, which may be difficult to discern from dot learning algorithms are provided in the supplemental material

86
 Son et al 
Deep Learning Detection of Fundus Findings

Table 1. Demographic Statistics of Seoul National University Bundang Hospital Datasets and 3 External Datasets (Indian Diabetic
Retinopathy Image Dataset, e-ophtha, and Messidor)

External Datasets
Indian Diabetic
Retinopathy
Seoul National University Bundang Hospital Datasets Image Dataset e-ophtha Messidor
Total no. of fundus images 103 262 143 434 1200
Total no. of gradable 95 350 (92.3) 143 (100.0) 434 (100.0) 1189 (99.1)
images (%)
Total no. of right eyes 47 586 70 219 601
Total no. of left eyes 47 764 73 215 588
No. of patients 47 022 N/A 203 N/A
Age, meanSD 51.525.4 N/A N/A N/A
No. of women (%) 22 808 (48.5) N/A N/A N/A
Location of hospital Korea India France France
Training and Majority Rule Unanimity Rule
Validation Set Test Set Test Set
(2 Agreements) (3 Agreements)
Hemorrhage 3846/78 315 (4.9) 429/8642 (5.0) 286/8499 (3.4) 53/143 (37.0) d 551/1062 (51.9)
Microaneurysms d d d 54/143 (37.7) 148/381 (38.8) d
Hard exudate 1930/79 071 (2.4) 234/8780 (2.7) 158/8704 (1.8) 54/143 (37.7) 47/82 (57.3) 254/1098 (23.1)
Cotton-wool patch 648/79 626 (0.8) 63/8858 (0.7) 30/8825 (0.3) 26/143 (18.1) d 142/1101 (12.9)
Drusen 7591/72 514 (10.5) 800/8125 (9.8) 376/7701 (4.9) d d 106/1077 (9.8)
Membrane 3110/78 202 (4.0) 338/8739 (3.9) 191/8592 (2.2) d d 23/1162 (2.0)
Macular hole 291/79 990 (0.4) 38/8859 (0.4) 16/8837 (0.2) d d 0/1189 (0.0)
Myelinated nerve fiber 259/80 275 (0.3) 16/8939 (0.2) 8/8931 (0.1) d d 4/1187 (0.3)
Chorioretinal atrophy 2642/77 749 (3.4) 295/8639 (3.4) 144/8488 (1.7) d d 42/1162 (3.6)
Vascular abnormality 596/79 226 (0.8) 76/8829 (0.9) 32/8785 (0.4) d d 14/1171 (1.2)
RNFL defect 1687/77 406 (2.2) 199/8603 (2.3) 66/8470 (0.8) d d 9/1173 (0.8)
Glaucomatous 3040/75 344 (4.0) 318/8363 (3.8) 111/8156 (1.4) d d 5/1171 (0.4)
disc change
Nonglaucomatous 962/77 509 (1.2) 104/8650 (1.2) 23/8569 (0.3) d d 5/1150 (0.4)
disc change

N/A ¼ not available; RNFL ¼ retinal nerve fiber layer; SD ¼ standard deviation.
Data are no. of images/total no. (%) unless otherwise indicated.

(Supplemental Material Fig S1, S2, and S3 and Table S1 available were used for the test set of glaucoma-related findings (RNFL
at www.aaojournal.org). defect, glaucomatous disc change, and nonglaucomatous disc
change).
Evaluation of the Algorithms
The performance of our deep learning algorithms was evaluated Comparison of the Model with Human Experts
in the 2 in-house test datasetsdthe majority rule and unanimity A separate dataset, the Messidor dataset,26 was used to compare the
rule test datasetsdand in the 2 external datasetsdthe IDRiD and algorithms’ performance with human experts. The Messidor
e-ophtha datasets. Areas under receiver operating characteristic dataset entails 1200 images obtained by a Topcon TRC NW6
curves (AUCs) were computed for each finding with 95% con- fundus camera with a 45
field of view (TOPCON, Tokyo,
fidence intervals computed by the exact Clopper-Pearson Japan). Because the Messidor dataset provides only the severity
method.25 In addition, we also estimated specificity and of DR and presence of diabetic macular edema, without any
sensitivity that yield the highest harmonic mean with the 95% annotations regarding individual findings, 3 retina specialists
confidence intervals. We used the metrics module in the Python (S.J.P., J.Y.S., H.D.K.) independently annotated all 1200 images
scikit-learn package version 0.18.2 (http://www.python.org) in using our reading system. Positive labels were given for each
the computation of the AUC and confusion matrix for sensitivity finding based on the majority rule, and negative labels were
and specificity calculation. given based on the unanimity rule, as in our in-house dataset.
We also devised another experimental setting to evaluate the The performance of each human retina expert was evaluated
influence of quality of annotation on the performance, in which against the majority and compared with the results from the deep
data annotated by the relevant specialists were reserved as the test learning models.
set, whereas the rest of the data were used as training and valida-
tion sets. For retina-related findings (hemorrhage, hard exudate,
cotton-wool patch, drusen, membrane, macular hole, myelinated Results
nerve fiber, chorioretinal atrophy or scar, and any vascular ab-
normality), the data annotated by retina specialists were used as a We collected annotations for 103 262 images during the study
test set, whereas the remaining data were used as training and period and excluded 7912 ungradable images. Images were
validation sets. Similarly, data annotated by glaucoma specialists excluded if judged not gradable by even 1 reader because of poor

87
 Ophthalmology Volume 127, Number 1, January 2020

visibility resulting from media opacity or lack of light in any region sensitivities and specificities at the highest harmonic mean for each
because of a small pupil. Two hundred eighty-six thousand fifty finding are presented in Table S4 (available at
annotations of 95 350 images (92.3%) were assessed. Sixty-three www.aaojournal.org).
thousand nine hundred fifteen images were annotated by general Receiver operating characteristic curves for retinal hemorrhage
ophthalmologists and certified cornea specialists, 22 226 images and hard exudates in the Messidor dataset with comparison with
were annotated by certified retina specialists, and the remaining human expert performance are shown in Figure 3. The zoomed
9209 images were annotated by certified glaucoma specialists. area shows that the performance of the model in the Messidor
Table 1 summarizes demographic statistics and the number of dataset was comparable with that of the human retina experts.
images for each of the 12 findings in our in-house and external The comparison of the performance of the model with human
datasets. Additional details on the dataset used in this study are experts in the Messidor dataset of other non-DR findings are pre-
available elsewhere.21 sented in Fig S4 (available at www.aaojournal.org). Careful
Table 2 shows the AUCs in our in-house and external datasets. interpretation is necessary because the number of cases is not
Areas under receiver operating characteristic curve were 96.2% to sufficient for meaningful interpretation for some findings (Table 1).
99.9% in the majority rule dataset and 97.9% to 99.9% in the
unanimity rule dataset. Figure 1 presents the corresponding
receiver operating characteristic curves for individual findings. Discussion
The AUC was higher in all findings on the unanimity rule in-
house dataset compared with that of the majority rule in-house We introduce our deep learning algorithms determining the
dataset. In addition to the AUC, we also measured sensitivity existence of abnormal findings and highlighting salient areas
and specificity at the highest harmonic mean for each finding of corresponding lesions in retinal fundus images. The al-
(Table S2, available at www.aaojournal.org). The performance of gorithms showed consistently reliable performance for
our algorithms in the external datasets was AUCs of 95.7% to hemorrhage, hard exudate, membrane, macular hole, and
98.0% for the IDRiD and 94.7% to 96.5 % for the e-ophtha.
Figure 2 presents the examples of activation maps for each
myelinated nerve fiber and moderate performance for
finding, accompanied by the corresponding original image. cotton-wool patch, drusen, chorioretinal atrophy, vascular
Lesion heatmaps highlight salient regions effectively even for abnormality, RNFL defect, glaucomatous disc change, and
lesions that could be observed anywhere in the fundus with nonglaucomatous disc change, even when tested on our in-
various shapes (e.g., hard exudate, hemorrhage, cotton-wool house test dataset of specialists and other external datasets.
patch, drusen), as well as in lesions that could be observed at The performance of the algorithms rivaled that of the retina
specific regions with distinct (e.g., macular hole) or various (e.g., experts, raising the possibility of their clinical use in primary
glaucomatous disc change) appearances. Furthermore, large lesions screening situations with low accessibility to specialists.
that span multiple regional subdivisions were also well detected Our approach represents an innovation on the previous
(e.g., RNFL defect, chorioretinal atrophy). approaches using deep learning algorithms12e19 in many
Areas under receiver operating characteristic curve of separate
test sets according to subspecialty of reader are listed in Table S3
aspects. Our model architecture can classify the presence of
(available at www.aaojournal.org). Areas under receiver operating individual findings observable in fundus examination and
characteristic curve for hemorrhage, hard exudate, cotton-wool visualize salient regions using a lesion heatmap, which can
patch, drusen, membrane, macular hole, myelinated nerve fiber, be combined by running multiple models to find multiple
and chorioretinal atrophy were constantly high, whereas AUCs for findings. Historically, researchers have devoted efforts to
vascular abnormality, RNFL defect, glaucomatous disc change, the localization of findings such as hemorrhage,27 drusen
and nonglaucomatous disc change decreased slightly. The deposits,28 hard exudate,29 and cotton-wool patch30 in

Table 2. Area under the Receiver Operating Characteristic Curve for Each Finding in the Seoul National University Bundang Hospital
Dataset and External Datasets

Seoul National University Bundang Hospital Dataset External Datasets

Majority Rule Test Set Unanimity Rule Test Set Indian Diabetic
(2 Agreements) (3 Agreements) Retinopathy Image Dataset e-ophtha
Hemorrhage 99.3 (96.7e99.9) 99.7 (97.3-100.0) 98.0 (81.0e99.9) 94.7 (84.9e98.9)
Hard exudate 99.8 (97.4e100.0) 99.9 (96.4e100.0) 98.0 (76.8e100.0) 96.5 (73.5e100.0)
Cotton-wool patch 99.3 (94.9e100.0) 99.7 (94.7e100.0) 95.7 (69.8e99.8) d
Drusen 98.2 (96.4e99.2) 99.0 (96.3e99.9) d d
Membrane 98.9 (96.1e99.7) 99.7 (96.9e100.0) d d
Macular hole 99.9 (92.9e100.0) 99.9 (88.8e100.0) d d
Myelinated nerve fiber 98.7 (89.7e100.0) 99.9 (92.3e100.0) d d
Chorioretinal atrophy 99.6 (97.1e100.0) 99.9 (96.5e100.0) d d
Vascular abnormality 96.2 (91.4e99.1) 97.9 (90.1e100.0) d d
RNFL defect 98.3 (95.3e99.5) 99.5 (94.9e100.0) d d
Glaucomatous disc change 98.2 (95.8e99.3) 99.5 (96.1e100.0) d d
Nonglaucomatous disc change 97.7 (94.8e99.4) 99.2 (93.6e100.0) d d

RNFL ¼ retinal nerve fiber layer.

Area under the receiver operating characteristic curve and 95% confidence interval is given in a format of mean (lower boundeupper bound). When more
than 100%, the upper bound is truncated to 100%.

88
 Son et al 
Deep Learning Detection of Fundus Findings

Figure 1. Receiver operating characteristic curves for all findings. IDRiD ¼ Indian Diabetic Retinopathy Image Dataset; RNFL ¼ retinal nerve fiber layer.

fundus images with heuristic methods. However, these focused on abnormal findings rather than final diagnoses,
methods have severe limitations in clinical application which are more clinically useful in covering a broader range
because they do not generalize in heterogeneous datasets. of diagnoses, including retinal vein occlusion, epiretinal
A recent study introduced the localization of retinal membrane, and macular hole, along with the afore-studied
findings of microaneurysm and hard exudates using 243 DR, glaucoma, and AMD. Also, lesion heatmaps provided
retinal fundus images in developing the algorithm and 148 by our method help physicians understand how the final
images in the testing algorithm.31 In our study, more than decision was made from the automatic algorithm. Moreover,
100 of thousands fundus images were used in the our deep learning algorithms export an activation map as an
development process and included thousands of abnormal auxiliary output to visualize where the algorithms attend to
fundus images for testing 12 abnormal findings that for generating the output for the existence of the target
subsume hemorrhage and hard exudate. Ting et al17 finding. Our model produces classification results by aver-
proposed a screening system based on deep learning aging the activation map by which the lesion heat map is
models targeting referable DR, vision-threatening DR, generated. This is crucial in making the output of the
possible glaucoma, and referable AMD. The work parallels algorithm interpretable, because we can see whether the
our approach in that we used neural networks to examine algorithm learned meaningful patterns or meaningless biases
multiple abnormalities on fundus images. However, we through the lesion heatmap. For example, if biases resulting

89
 Ophthalmology Volume 127, Number 1, January 2020

Figure 2. Activation map demonstrating representative findings, shown in pairs of original fundus image (left) and corresponding activation map (right) for
each target finding.

from certain image characteristics of a specific camera or surges during peak hours, short computation time is
institution are learned by the algorithm, biased activation imperative for clinician-assisting device systems. The fast
will appear in the lesion heatmap. Furthermore, our algo- computation speed of our algorithm (less than 15 ms for
rithm will not misclassify in severe cases where the lesions processing an image with an off-the-shelf graphics pro-
are obvious and widespread because intensity values in the cessing unit) is suitable for clinical settings and amenable to
lesion heatmap are proportional to the size. This also implies reading center environments, where servers can generate
that physicians can interpret the decision made by the model outputs for images sent in by multiple clients.
by inspecting the suggested lesions. With the lesion heat- A major obstacle in developing deep learning algorithms
map, our system may help to mitigate the mental burden of for automated screening of retinal fundus images is that no
ophthalmologists in examining retinal fundus images by gold standard annotation already exists for images. Massive
guiding attention to abnormal lesions. The system is also amounts of images have to be annotated for use in such
promising for screening the general population when algorithms, and the issue of interreader variability arises. In
maintained with high sensitivity32,33 to reduce the entire recent studies, it has been reported that nontrivial interreader
workload and medical costs by obliterating the need for variability exists in the assessment of retinal fundus images
inspection of evidently normal eyes.34 We believe that our among general ophthalmologists as well as retina special-
system, which can screen multiple findings, is preferable ists.35e37 Although readers with more experience tend to
to a system that diagnoses only a single, specific disease have less interreader variability,38 disagreement still exists
such as DR or possible glaucoma in health screening among specialists, especially in ambiguous findings that
centers. In real-world clinics where the number of patients resemble artifacts. However, further research is necessary

90
 Son et al 
Deep Learning Detection of Fundus Findings

Figure 3. Receiver operating characteristics curves and comparison with retinal experts for hemorrhage and hard exudate in the Messidor database.
IDRiD ¼ Indian Diabetic Retinopathy Image Dataset.

to investigate how this interreader variability affects the test set is homogenous to the training set, and thus the
performance of deep learning algorithms or how malicious collective patterns concur. In some findings, namely
effects, if they exist, can be minimized. Through our hemorrhage, hard exudate, cotton-wool patch, drusen,
experiments, we have witnessed several interesting membrane, macular hole, myelinated nerve fiber, and cho-
phenomena regarding annotation quality and performance. rioretinal atrophy, there was no decrease when tested on
First, the performance was better in the unanimity rule test specialist annotation sets. These may have been findings that
set, which is reasonable because the images with were more obvious and less ambiguous, and the readings
unanimous annotations tend to include findings that are were not affected by reader subspecialty.
less controversial. Second, the performance decreased in The fact that the training dataset decides the perceptive
some findings when the network trained with annotations patterns of deep learning algorithms has practical and clin-
of the nonspecialists was tested on annotations by the ical implications. The outputs of the algorithms will be
corresponding specialists (e.g., retinal abnormality for representative of the group of readers who assessed the
retina specialist), because specialists would have identified training dataset at best. When the test dataset is qualitatively
more subtle cases. Because fundus images were assigned dissimilar to the training dataset, the performance suffers.
randomly to each reader, the decrease in AUC is mainly Nonoverlapping readers as well as disparate image sources
attributed to mismatch in subspecialties of the readers between the training and test dataset result in an underesti-
between the training and test datasets. Deep learning mation of the performance. Conversely, when an algorithm
algorithms are trained to learn collective patterns inherent is tested on the same dataset as it was trained on, it may
to the training dataset. The AUC decreased when the show falsely high performance. Therefore, although an
collective patterns extracted from the training set did not algorithm is proven to perform well on a designated test
match well to those that reside in the test set of dataset, every algorithm should also be validated on separate
specialists. If the training and test sets were created under external datasets to evaluate its generalization performance
the same policy, the AUC would be higher because the in various clinical settings. Our algorithm showed extremely

91
 Ophthalmology Volume 127, Number 1, January 2020

high performance when tested on our in-house dataset but the majority decision tends to secure only apparent lesions
performed rather less well for some findings on the external and overlook subtle lesions. For instance, an image
datasets in which heterogeneous images were annotated by annotated as positive by 1 of 3 readers may actually include
other experts. It is possible that clinicians who are equally trivial abnormalities that 2 readers omitted. To prevent such
qualified may not agree with the decision of the algorithms, cases, data should be collected through adjudication,38
and this may cause problems when used in clinics. In such although gathering adjudicated annotations of the experts
cases, visual demonstration of the lesions in our algorithm incurs high costs and considerable time that are
can extricate such clashes by allowing ophthalmologists to prohibitive to the collection of large datasets. It is a
second-check on what the decision of the algorithm was technical challenge to fully exploit the discarded data in
based and overturn the decision when the lesions are the development processes so that the model generalizes
detected wrongly in their opinions. well even for minor lesions. Finally, the current algorithm
Another possible cause of the lower performance of our does not incorporate medical knowledge in detecting the
algorithm in the external datasets may be that they lesions. For instance, drusen and hard exudate may be
demonstrated different demographic statistical characteris- similar in appearance, although ophthalmologists can
tics, as seen in Table 1. The IDRiD and e-ophtha datasets differentiate them from the location and presence of
contained annotation information only for DR-related find- leakage from abnormal blood vessels such as
ings, and the Messidor dataset contained only a few images microaneurysms or macroaneurysms. Further research is
for certain findings such as macular hole, myelinated nerve required to overcome these limitations.
fiber, glaucomatous disc change, and nonglaucomatous disc Our deep learning network not only classifies a broad
change. After investigation of failure cases (Figs S5, S6, range of ophthalmologic findings with high accuracy but also
and S7, available at www.aaojournal.org), we suspect that highlights the salient regions on the images. In particular,
the slight decrease in the discriminative performance for lesion heatmaps of findings bridge machine learning models
the IDRiD and e-ophtha datasets may be mostly attributed and human clinical interpretation, which is an important next
to conflicting assessment criteria compared with our step in the application of machine learning models in clinical
annotation guidelines and image-quality issues, such as settings. We believe our approach of determining the exis-
dark regions in nonmydriatic images and blob-like artifacts tence of abnormal findings with lesion heatmaps is inclined to
on the lens. In the case of the e-ophtha dataset, extremely the deployment in clinics in that classification results are
minute lesions were segmented as well, although some of interpretable as well as reliable. In real-life clinical settings,
them were controversial in our evaluation. In addition, our the same network architecture can also be used to prescreen
in-house dataset included findings not only from DR but for ungradable images and media opacity before detection of
also from other diseases such as AMD, retinal vein occlu- various fundus abnormalities. This may contribute signifi-
sion, epiretinal membrane, macular hole, and glaucoma, cantly to reducing costs for fundus screening in areas where
because our model was also intended to detect findings in ophthalmologists are not readily available and may enable
various non-DR images, which might have affected the wider distribution of fundus screening, resulting in earlier
performance in a DR-specific dataset. For example, assess- detection of many potentially vision-threatening diseases.
ment criteria for hemorrhage in our in-house dataset Furthermore, the approach that detects findings first will help
included various types of hemorrhages (preretinal, retinal, to open new research opportunities for developing algorithms
subretinal, disc hemorrhage, and microaneurysms indistin- that diagnose diseases based on the discovered findings, just
guishable from dot hemorrhages) observed in AMD, retinal as human experts do. Prospective clinical studies to evaluate
vein occlusion, and glaucoma as well as DR, and these the safety and efficacy of such automated screening systems
various types of hemorrhages were simplified to 1 category are currently undergoing.
of hemorrhage in our annotation process. We suspect that
the lower performance in the DR-specific datasets may be
the result of the predominance of minute dot hemorrhages, References
which may be overlooked by our network, which included
learned collective patterns of hemorrhage with different
shapes, sizes, and ambiguity. Dot-like artifacts may also be 1. Zhang X, Saaddine JB, Chou CF, et al. Prevalence of diabetic
confused with dot hemorrhages because of inherent ambi- retinopathy in the United States, 2005e2008. JAMA.
guity. Despite these shortcomings, currently no other public 2010;304(6):649e656.
dataset provides multifaceted annotations to match our 2. Early Treatment Diabetic Retinopathy Study Research Group.
needs in validating our models. Such datasets would be of Grading diabetic retinopathy from stereoscopic color fundus
immense value for further studies to evaluate the perfor- photographsdan extension of the modified Airlie House
mance of similar models in the future. classification. ETDRS report number 10. Ophthalmology.
There are several limitations in our study. First, the 1991;98(5 Suppl):786e806.
3. Lin DY, Blumenkranz MS, Brothers RJ, Grosvenor DM.
algorithms need to be tested for rare findings with larger
The sensitivity and specificity of single-field nonmydriatic
external datasets. Collecting rare cases may be taxing; monochromatic digital fundus photography with remote
however, tests of various cases would increase reliability of image interpretation for diabetic retinopathy screening: a
the system and avert catastrophic failures when the algo- comparison with ophthalmoscopy and standardized mydri-
rithm is deployed in clinics. Also, positive labels were given atic color photography. Am J Ophthalmol. 2002;134(2):
by the majority rule for developing the algorithm. However, 204e213.

92
 Son et al 
Deep Learning Detection of Fundus Findings

4. Detry-Morel M, Zeyen T, Kestelyn P, et al. Screening for 22. Prasanna Porwal SPRK, Kokare M, Deshmukh G, et al. Indian
glaucoma in a general population with the non-mydriatic Diabetic Retinopathy Image Dataset (IDRiD). IEEE Dataport
fundus camera and the frequency doubling perimeter. Eur J (The Institute of Electrical and Electronics Engineers) Data-
Ophthalmol. 2004;14(5):387e393. port; 2018. Available at: https://idrid.grand-challenge.org/
5. Mookiah MRK, Acharya UR, Chua CK, et al. Computer-aided Rules/.
diagnosis of diabetic retinopathy: a review. Comput Biol Med. 23. Decencière E, Cazuguel G, Zhang X, et al. TeleOphta:
2013;43(12):2136e2155. machine learning and image processing methods for tele-
6. Winder RJ, Morrow PJ, McRitchie IN, et al. Algorithms for ophthalmology. IRBM. 2013;34(2):196e203.
digital image processing in diabetic retinopathy. Comput Med 24. Son J, Bae W, Kim S, et al. Classification of findings with
Imag Graphics. 2009;33(8):608e622. localized lesions in fundoscopic images using a regionally
7. Singh A, Dutta MK, ParthaSarathi M, et al. Image processing guided CNN. Computational Pathology and Ophthalmic
based automatic diagnosis of glaucoma using wavelet features Medical Image Analysis, Lecture Notes in Computer Science
of segmented optic disc from fundus image. Comput Methods 11039. 2018:176e184. Available at: https://link.springer.com/
Prog Biomed. 2016;124:108e120. chapter/10.1007/978-3-030-00949-6_21.
8. Bock R, Meier J, Nyúl LG, et al. Glaucoma risk index: auto- 25. Kottas M, Kuss O, Zapf A. A modified Wald interval for the
mated glaucoma detection from color fundus images. Medical area under the ROC curve (AUC) in diagnostic case-control
Image Analysis. 2010;14(3):471e481. studies. BMC Med Res Methodol. 2014;14:26.
9. Issac A, Partha Sarathi M, Dutta MK. An adaptive threshold 26. Decencière E, Zhang X, Cazuguel G, et al. Feedback on a
based image processing technique for improved glaucoma publicly distributed image database: the Messidor database.
detection and classification. Comput Methods Prog Biomed. Image Analysis & Stereology. 2014;33(3):231e234.
2015;122(2):229e244. 27. Bae JP, Kim KG, Kang HC, et al. A study on hemorrhage
10. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. detection using hybrid method in fundus images. J Digit Im-
2015;521(7553):436e444. aging. 2011;24(3):394e404.
11. Annan L, Jun C, Damon Wing Kee W, Jiang L. Integrating 28. Rapantzikos K, Zervakis M, Balas K. Detection and segmen-
holistic and local deep features for glaucoma classification. tation of drusen deposits on human retina: potential in the
Conf Proc IEEE Eng Med Biol Soc. 2016;2016:1328e1331. diagnosis of age-related macular degeneration. Med Image
12. Abràmoff MD, Folk JC, Han DP, et al. Automated analysis of Anal. 2003;7(1):95e108.
retinal images for detection of referable diabetic retinopathy. 29. Sasaki M, Kawasaki R, Noonan JE, et al. Quantitative mea-
JAMA Ophthalmol. 2013;131(3):351e357. surement of hard exudates in patients with diabetes and their
13. Gulshan V, Peng L, Coram M, et al. Development and vali- associations with serum lipid levels. Invest Ophthalmol Vis
dation of a deep learning algorithm for detection of diabetic Sci. 2013;54(8):5544e5550.
retinopathy in retinal fundus photographs. JAMA. 30. Kose C, Sevik U, Ikibas C, Erdol H. Simple methods for
2016;316(22):2402e2410. segmentation and measurement of diabetic retinopathy lesions
14. Abramoff MD, Lou Y, Erginay A, et al. Improved automated in retinal fundus images. Comput Methods Programs Biomed.
detection of diabetic retinopathy on a publicly available dataset 2012;107(2):274e293.
through integration of deep learning. Invest Ophthalmol Vis 31. Lam C, Yu C, Huang L, Rubin D. Retinal lesion detection with
Sci. 2016;57(13):5200e5206. deep learning using image patches. Invest Ophthalmol Vis Sci.
15. Gargeya R, Leng T. Automated identification of diabetic 2018;59(1):590e596.
retinopathy using deep learning. Ophthalmology. 2017;124(7): 32. Chakrabarti R, Harper CA, Keeffe JE. Diabetic retinopathy
962e969. management guidelines. Exp Rev Ophthalmol. 2012;7(5):
16. Burlina PM, Joshi N, Pekala M, et al. Automated grading of 417e439.
age-related macular degeneration from color fundus images 33. Marmor MF, Kellner U, Lai TY, et al. Revised recommen-
using deep convolutional neural networks. JAMA Ophthalmol. dations on screening for chloroquine and hydroxychloroquine
2017;135(11):1170e1176. retinopathy. Ophthalmology. 2011;118(2):415e422.
17. Ting DSW, Cheung CY, Lim G, et al. Development and 34. Tufail A, Rudisill C, Egan C, et al. Automated diabetic reti-
validation of a deep learning system for diabetic retinopathy nopathy image assessment software: diagnostic accuracy and
and related eye diseases using retinal images from multiethnic cost-effectiveness compared with human graders. Ophthal-
populations with diabetes. JAMA. 2017;318(22):2211e2223. mology. 2017;124(3):343e351.
18. Li Z, He Y, Keel S, et al. Efficacy of a deep learning system for 35. Scott IU, Bressler NM, Bressler SB, et al. Agreement be-
detecting glaucomatous optic neuropathy based on color fundus tween clinician and reading center gradings of diabetic
photographs. Ophthalmology. 2018;125(8):1199e1206. retinopathy severity level at baseline in a phase 2 study of
19. Chen X, Xu Y, Wong DWK, et al. Glaucoma detection based intravitreal bevacizumab for diabetic macular edema.
on deep convolutional neural network. 2015 37th Annual In- Retina. 2008;28(1):36.
ternational Conference of the IEEE. (The Institute of Electrical 36. Lytwyn A, Salit IE, Raboud J, et al. Interobserver agreement in
and Electronics Engineers) Engineering in Medicine and the interpretation of anal intraepithelial neoplasia. Cancer.
Biology Society (EMBC), Milan. 2015:715e718. 2005;103(7):1447e1456.
20. Xiangyu C, Yanwu X, Damon Wing Kee W, et al. Glaucoma 37. Ruamviboonsuk P, Teerasuwanajak K, Tiensuwan M, et al.
detection based on deep convolutional neural network. Conf Interobserver agreement in the interpretation of single-field
Proc IEEE Eng Med Biol Soc. 2015;2015:715e718. digital fundus images for diabetic retinopathy screening.
21. Park SJ, Shin JY, Kim S, et al. A novel fundus image reading Ophthalmology. 2006;113(5):826e832.
tool for efficient generation of a multi-dimensional categorical 38. Krause J, Gulshan V, Rahimy E, et al. Grader variability and
image database for machine learning algorithm training. the importance of reference standards for evaluating machine
J Korean Med Sci. 2018;33(43):e239. https://doi.org/10.3346/ learning models for diabetic retinopathy. Ophthalmology.
jkms.2018.33.e239. eCollection 2018 Oct 22. 2018;125(8):1264e1272.

93
 Ophthalmology Volume 127, Number 1, January 2020

Footnotes and Financial Disclosures

Originally received: February 11, 2019. HUMAN SUBJECTS: No human subjects were included in this study.
Final revision: May 3, 2019. Images were de-identified and written consent from patients was waived by
Accepted: May 24, 2019. the institutional review board due to the retrospective nature of the study.
Available online: May 31, 2019. Manuscript no. 2019-313. All research adhered to the tenets of the Declaration of Helsinki.
1
VUNO, Inc., Seoul, Korea. No animal subjects were included in this study.
2
Department of Ophthalmology, Seoul Metropolitan Government Seoul Author Contributions:
National University Boramae Medical Center, Seoul, Korea. Conception and design: Son, Shin, Jung, S. J. Park
3
Department of Ophthalmology, College of Medicine, Soonchunhyang Analysis and interpretation: Son, Shin, Jung, S. J. Park
University, Cheonan, Korea.
Data collection: Son, Shin, Kim, Jung, K.H.Park, S. J. Park
4
Department of Ophthalmology, Seoul National University College of
Obtained funding: Son, Jung, S. J. Park
Medicine, Seoul National University Bundang Hospital, Seongnam, Korea.
Overall responsibility: Son, Shin, Jung, S. J. Park
*Both authors contributed equally as first authors.
Financial Disclosure(s): Abbreviations and Acronyms:
The author(s) have made the following disclosure(s): J.S.: Employee e AMD ¼ age-related macular degeneration; AUC ¼ area under receiver
operating characteristic curve; DR ¼ diabetic retinopathy; IDRiD ¼ Indian
VUNO, Inc. (Seoul, Korea).
Diabetic Retinopathy Image Dataset; RNFL ¼ retinal nerve fiber layer.
K.-H.J.: Employee and Equity owner e VUNO, Inc. (Seoul, Korea).
S.J.P.: Equity owner e VUNO, Inc. (Seoul, Korea). Correspondence:
Sang Jun Park, MD, MSc, Department of Ophthalmology, Seoul National
Supported by the Intelligence Information Service Expansion Project, which is
University Bundang Hospital, 82, Gumi-ro 173 Beon-gil, Bundang-gu,
funded by National IT Industry Promotion Agency (grant no.: NIPA-C0202-
Seongnam-si, Gyeonggi-do, 13620, Korea. E-mail: [email protected];
17-1045) and the Small Grant for Exploratory Research of the National
and Kyu-Hwan Jung, PhD, Chief Technology Officer, Vuno, Inc., 6F, 507,
Research Foundation of Korea, which is funded by the Ministry of Science,
Gangnam-daero, Seocho-gu, Seoul, Republic of Korea. E-mail: khwan.
Information and Communications Technology, and Future Planning (grant no.:
[email protected].
NRF-2018R1D1A1A09083241). The sponsors or funding organizations had
no role in the design or conduct of this research. The authors alone are
responsible for the content and writing of the article.

94
 Commentary

Using Deep Learning Models to Characterize Major Retinal Features on Color

Fundus Photographs
Cecilia S. Lee, MD, MS - Seattle, Washington
Ryan T. Yanagihara, MD - Seattle, Washington
Aaron Y. Lee, MD, MSCI - Seattle, Washington

Exciting applications of deep learning have been made Furthermore, although Son et al present the results of 12
recently in the field of ophthalmology. In particular, several independent convolutional neural networks, it would be
algorithms have shown remarkable success in classifying interesting to compare the performance of the algorithms with
common diseases such as age-related macular degeneration a single ensembled model that integrates all 12 networks.
and diabetic retinopathy as well as rare pathologic Presumably, similar low-level features are used in common
features.1,2 In the current issue of Ophthalmology, Son et al3 across the 12 models, and the current architecture harbors
(see page 85) report a deep learning classification model redundancies. Future application of deep learning using cur-
for 12 different features from fundus photographs with rent study data could include segmentation or localization
impressive results. A total of 12 separate, binary models, where each feature would be segmented as in
classification convolutional neural networks were trained, bounding boxes or pixelwise masks. However, these would
validated, and tested using a dataset containing 286 050 require a large dataset with even more extensive annotation.
annotations from 95 350 fundus images. Lesion heatmaps Finally, the current model would learn only human-derived
were used to demonstrate the probable areas of salient features, given that training was based on the expert annota-
features. tions. Future attempts to train the model with a purely data-
One important strength of this study is its extensive driven method may allow novel, machine-discovered bio-
labeling with multiple human experts. A total of 57 markers in various ocular conditions.
ophthalmologists from various subspecialities annotated The study by Son et al demonstrates exciting advances in a
fundus images, and each image was labeled by 3 independent deep learning classification model in which multiple features
ophthalmologists. Ground truth was defined by either the can be identified rather than diagnosis alone, which would
majority or the unanimity rule. Another strength was external allow exploring novel feature-related hypotheses.5 Open
validation to assess the algorithm’s generalizability. Both sourcing the data and models for others to build on would
in-house datasets along with 2 external databases (the Indian be an important contribution to this field by allowing the
Diabetic Retinopathy image Database and e-ophtha) were creation of more reproducible methods, encouraging
analyzed using the presented deep learning algorithm. The researchers to share data, and creating a collaborative
areas under receiver operating characteristic curve for the 12 culture for deep learning research in ophthalmology.
fundus features ranged from 96.2% to 99.9% for the in-house
datasets, whereas performance was lower (94.7%e98.0%) in References
the external datasets, as expected. In addition, 3 retina
specialists labeled every image in a third external dataset
(Messidor) for all 12 features. Although detection accuracy 1. Ting DSW, Cheung CY-L, Lim G, et al. Development and
for most features was comparable with that of the retina validation of a deep learning system for diabetic retinopathy and
specialists, the evaluation of features not related to diabetic related eye diseases using retinal images from multiethnic
populations with diabetes. JAMA. 2017;318:2211e2223.
retinopathy, in particular, was limited given their low fre-
2. Kihara Y, Heeren TFC, Lee CS, et al. Estimating retinal
quencies in this diabetic retinopathy-focused dataset. sensitivity using optical coherence tomography with deep-
Although the areas under receiver operating character- learning algorithms in macular telangiectasia type 2. JAMA
istic curve were excellent, the training and validation data- Network Open. 2019;2:e188029.
sets in this study were not balanced for each classification 3. Son J, Shin JY, Kim HD, et al. Development and validation of
feature. Thus, reporting the area under precision recall curve deep learning models for screening multiple abnormal findings
would be more informative because the precision recall in retinal fundus images. Ophthalmology. 2019;126:85e94.
curve has been shown to be a more accurate measure of 4. Davis J, Goadrich M. The relationship between precision-recall
performance particularly in unbalanced test sets.4 Testing and ROC curves. Proceedings of the 23rd International Con-
the model’s performance in several external validation sets ference on Machine LearningdICML ’06 2006; 2006. https://
doi.org/10.1145/1143844.1143874. Accessed July 2, 2019.
such as was done in this study is a critical component of
5. Lee CS, Lee AY, Baughman D, et al. The United Kingdom
advancing this field. However, it is important to note that Diabetic Retinopathy Electronic Medical Record Users Group:
the validation sets did not contain any poor-quality images report 3: baseline retinopathy and clinical features predict pro-
that would be encountered in real-world deployment, which gression of diabetic retinopathy. Am J Ophthalmol. 2017;180:
ultimately may limit the generalizability of the model. 64e71.

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.07.014 95

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

Footnotes and Financial Disclosures

Financial Disclosure(s): The author(s) have made the following disclo- sponsors or funding organizations had no role in the design or conduct of
sure(s): A.Y.L.: Financial support e Carl Zeiss Meditec, Novartis, Topcon this research.
Corporation, Lowy Medical Research Institute. Correspondence:
Supported by the National Institutes of Health, Bethesda, Maryland (grant Aaron Y. Lee, MD, MSCI, Department of Ophthalmology, University of
nos.: K23EY024921 [C.S.L.] and K23EY029246 [A.Y.L.]); and Research Washington, Box 359608, 325 Ninth Avenue, Seattle, WA 98104. E-mail:
to Prevent Blindness, Inc., New York, New York (unrestricted grant). The [email protected].

Pictures & Perspectives

Fingerprint Macula Artifact on Optos Fundus Imaging in Nystagmus

A fingerprint appearance to the macula was observed in both eyes of a patient with foveal hypoplasia and pendular horizontal nystagmus
imaged with the Optos California (Marlborough, MA) fundus camera. (A, B) Examination under anesthesia with Natus Retcam 3
(San Carlo, CA) color fundus photography (C) and Bioptigen Envisu C2300 (Buffalo Grove, IL) portable spectral-domain OCT (D)
confirmed these to be artifactual. The Optos camera utilizes sweeping red and green lasers and multi-element optics with nonlinear surfaces
resulting in nonuniform scan velocities slowest across the fovea. Nystagmus causes compressed and expanded scanning producing the
central fingerprint artifact. (Magnified version of Fig A-D is available online at www.aaojournal.org).

ROBERT A. SISK, MD, FACS1,2,3

PRASHANT K. PAREKH, MD1,2,3
CHRISTOPHER D. RIEMANN, MD1,2
1
Retina Division, Cincinnati Eye Institute, Cincinnati, Ohio; 2University of Cincinnati Department of Ophthalmology, Cincinnati, Ohio; 3Abrahamson
Pediatric Eye Institute, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio

96
The Relative Impact of Patient, Physician,
and Geographic Factors on Variation in
Primary Rhegmatogenous Retinal
Detachment Management
Daniel Vail, BA,1 Suzann Pershing, MD, MS,1,2 Mary-Grace Reeves, BA,1 Armin R. Afshar, MD, MBA3

Purpose: To evaluate geographic variation and characterize the relative contributions of patient character-
istics, physician practice, and geographic region on variation in primary rhegmatogenous retinal detachment
(RRD) management.
Design: Retrospective claims-based analysis.
Participants: Commercially insured patients with incident RRD diagnosed between 2008 and 2016 (12 779
patients).
Methods: We determined whether patients underwent primary RRD repair within 60 days of diagnosis and
identified repair type. We characterized physicians using physician identifier variables and characterized geog-
raphy by Combined Statistical Areas or Core-Based Statistical Area. We used multilevel mixed effects logistic
regression models to evaluate patient-, physician-, and geographic-level variation in whether patients underwent
RRD repair and used multilevel mixed effects multinomial models to characterize variation in repair type. For each
model, we evaluated patient fixed effects and physician random effects nested within geographic random effects.
We estimated intraclass correlation coefficients and variance partition coefficients, respectively, to compare
relative contributions of patient, physician, and geography to overall variation.
Main Outcome Measures: Odds ratios for RRD repair and variation estimates for patient, physician, and
geography.
Results: Most incident RRD patients received treatment within 60 days post-diagnosis. Pars plana vitrec-
tomy was most common (49%), followed by laser barricade (23%), scleral buckle and pneumatic retinopexy (both
11%), and cryotherapy (5%). Physician-level variation showed greater impact on receipt of any treatment than
geographic-level variation (estimated variance coefficients of 1.09 and 0.32, respectively). Patient-level charac-
teristics represented approximately 82% of overall variation in receipt of any repair, versus 16% from physician-
level and 2% from geographic-level factors. Among RRD patients who underwent repair, estimated variance
coefficients were 0.07 for geography and 3.37 for physician. Physician-level factors represented approximately
50% of total variation in repair type, followed by patient-level (49%), and geographic-level (1%) factors.
Conclusions: Rhegmatogenous retinal detachment repair decisions are influenced by patient-level and
physician-level factors, less so by geographic variation. Patient characteristics account for most of the variation in
receipt of repair, and physician practice accounts for most of the variation in choice of procedure. These findings
indicate a need for additional studies to understand drivers behind differences in care and clinical outcomes and
to identify barriers in access to care. Ophthalmology 2020;127:97-106 Published by Elsevier on behalf of the
American Academy of Ophthalmology

Supplemental material available at www.aaojournal.org.

Rhegmatogenous retinal detachments (RRDs) are the most causing central vision loss.2,4 Five procedures are used
common type of retinal detachment managed by ophthal- commonly in RRD management: pars plana vitrectomy
mologists in the United States, with an estimated incidence (PPV); scleral buckle (SB), alone or combined with PPV;
of 13 cases per 100 000 persons per year and with a growing pneumatic retinopexy (PR); laser barricade; cryotherapy; or
incidence in an aging population.1e3 They occur when a combination thereof.
liquid vitreous passes through a full-thickness hole or tear in Previous studies have identified PPV as the most common
the retina and collects in the subretinal space, separating the RRD repair procedure among Medicare-insured and
neurosensory retina from the retinal pigment epithelium.1 commercially insured patients, increasingly chosen over the
These detachments may require surgical intervention, traditional SB approach.5,6 Pars plana vitrectomy and SB can
particularly when the macula is threatened or involved, have similar efficacy.7e10 However, although SB provides

Published by Elsevier on behalf of the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.04.019 97

ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

Table 1. Demographics and Clinical Characteristics among have investigated changes in surgical management of retinal
Patients with Incident Rhegmatogenous Retinal Detachment, detachments over time and have associated changes in
2008e2016 practice patterns with costs and patient characteristics.6
However, geographic variation in surgical management
No. (%)
remains poorly characterized. Some studies have surveyed
Surgery physicians about their use of repair techniques, and a 2012
None 4838 (37.86) article described regional variation in repair technique
Pneumatic retinopexy 871 (6.82) among Medicare beneficiaries, but these descriptive studies
Scleral buckle 871 (6.82)
Pars plana vitrectomy 3924 (30.71)
do not include patient-level or physician-level characteris-
Laser barricade 1849 (14.47) tics that may affect surgical management decisions.5,14 The
Cryotherapy 426 (3.33) relative importance of patient-, physician-, and geographic-
Gender level variables in accounting for variation in RRD manage-
Male 7175 (56.15) ment is not yet well understood.
Female 5604 (43.85) In this study, we characterized variation in practice pat-
Age (yrs)
terns for management of primary RRD across the United
<18 230 (1.80)
18e29 560 (4.38) States among a cohort of commercially insured patients with
30e39 786 (6.15) RRD who were diagnosed and managed between 2008 and
40e49 1855 (14.52) 2016. We compared the relative contributions of patient
50e59 5379 (42.09) characteristics, physician practice, and geographic region to
60e69 3969 (31.06) overall variation in care.
Clinical comorbidities
Myopia 594 (4.65)
Pseudophakia 1191 (9.32)
Vitreous hemorrhage 1666 (13.04) Methods
Lattice degeneration 2452 (19.19)
Diabetes mellitus 1610 (12.60) Data Source and Sample Selection
Year
2008 1829 (14.31) We used administrative insurance claims data from the Truven
2009 2003 (15.67) Health MarketScan Commercial Claims and Encounters database
2010 1147 (8.98) for this analysis (Truven Health Analytics, San Jose, CA). Mar-
2011 1205 (9.43) ketScan is a large database of more 150 million commercially
2012 1351 (10.57) insured patients from more than 350 insurance carriers and includes
2013 1247 (9.76) patients in all 50 states. This database has been used previously for
2014 1154 (9.03) ophthalmic studies as well as for studies of geographic variation in
2015 1263 (9.88) physician practice patterns.18,19 The MarketScan database comprises
2016 1580 (12.36) longitudinal patient claims data, with a follow-up period that ex-
tends until a patient discontinues insurance (i.e., becomes uninsured,
changes commercial insurers, or transitions to Medicare). The
database includes patient-level demographic information, diagnosis
structural support and is associated with lower costs and procedural codes for each patient encounter with a health care
(approximately $1400 for PPV, compared with $1100 for SB provider, and billing codes. It also includes individual physician
and $800 for a single PR procedure) as well as lower rates of identifiers and various geographic variables denoting the location of
postoperative cataract,11 with the advent of wide-angle a patient’s employer, physician, and hospital. This secondary anal-
viewing systems and transconjunctival sutureless cannulas, ysis of de-identified administrative data was determined to be
PPV may be preferred by some surgeons because of its su- exempt from institutional review board approval.
perior intraoperative visualization and quicker repair. Pars Our sample included all patients in the MarketScan database
plana vitrectomy can be beneficial in pseudophakic de- with a new diagnosis of RRD between January 1, 2008, and
December 31, 2016 (Fig S1, available at www.aaojournal.org). We
tachments and complex cases such as giant retinal tears and
limited our sample to patients with continuous enrollment in
minimizes postoperative myopic shift.7,11,12 Pneumatic reti- MarketScan, at least 60 days of continuous coverage after the first
nopexy has a lower success rate than PPV or SB but has the diagnosis of RRD, and at least 365 days of continuous coverage
advantages of being less costly when successful, less inva- before the first diagnosis of RRD. We excluded patients with any
sive, and able to be performed quickly in the outpatient observed procedure code for RRD repair during the minimum 1
setting without need for anesthesia services.12e15 It also has year of enrollment preceding a first diagnosis of RRD or with
been suggested that PR may yield better visual outcomes ocular comorbidities that might confound treatment decisions
compared with PPV.15 Cryotherapy or laser barricade (Table S1, available at www.aaojournal.org). We also excluded
typically are used in conjunction with other surgical repair patients with missing geographic information and limited our
procedures to barricade a retinal break and promote analysis to patients with nonmissing physician identifier variables
at the time of diagnosis and at the time of repair (if an RRD
chorioretinal adhesion.16 However, asymptomatic retinal
repair procedure was identified). Physician identifiers were drawn
breaks or small peripheral detachments may be managed by from MarketScan’s provider identification variable using the
either cryotherapy or laser barricade alone.17 encounter with a given patient’s index diagnosis of RRD and at
Given that none of the techniques for RRD repair is su- the encounter where patients received treatment, respectively.
perior in all respects to the alternatives, it is reasonable to Although we were not able to evaluate provider specialty using
expect substantial variation in use by physicians. Prior studies specialty type codes (missing for more than 99% of the patients

98
 Vail et al 
Variation in RRD Management

Table 2. Multilevel Mixed Effects Logistic Regression Models for Likelihood of Receiving Any Rhegmatogenous Retinal Detachment
Repair, 2008e2016

Odds Ratio (95% Confidence Interval)

Model 3 (n ¼ 4146)* Model 2 (n ¼ 4146)* Model 1 (n ¼ 4146)*
Fixed effects, patient characteristics
Gender (reference ¼ male)
Female 0.62 (0.48e0.80) 0.64 (0.51e0.80)
Age at diagnosis (1-yr increments) 1.02 (1.02e1.03) 1.02 (1.01e1.03)
Year of diagnosis (1-yr increments) 0.90 (0.85e0.95) 0.91 (0.86e0.95)
Comorbidities
Myopia 0.76 (0.45e1.26) 0.79 (0.49e1.26)
Pseudophakia 0.89 (0.60e1.32) 0.94 (0.65e1.35)
Vitreous hemorrhage 1.09 (0.76e1.55) 1.07 (0.77e1.49)
Lattice degeneration 1.32 (0.97e1.80) 1.31 (0.99e1.74)
Diabetes mellitus 0.62 (0.45e0.87) 0.64 (0.47e0.86)

Variance Estimate (95% Confidence Interval)

Random effects
Geography (CSA or CBSA) 0.32 (0.13e0.81) 0.43 (0.23e0.81) 0.49 (0.28e0.89)
Physician (nested within CSA or CBSA) 1.09 (0.61e1.97)

CBSA ¼ Core-Based Statistical Area; CSA ¼ Combined Statistical Area based on United States Census.
*Model 1 ¼ empty (geographic random effects only); model 2 ¼ 2-level, patient-level fixed effects þ geographic random effects; model 3 ¼ 3-level, patient-
level fixed effects þ geographic random effects þ physician random effects (nested within geographic region).

in our analysis), we estimated provider specialty and subspecialty received treatment within 60 days of diagnosis, a categorical var-
using procedure and diagnosis codes (Supplemental Appendix, iable indicating whether initial treatment was performed via PPV,
available at www.aaojournal.org). SB, PR, cryotherapy, or laser barricade. Patients who underwent
multiple procedures on the same date were categorized according
Predictor Variables to the most invasive procedure. Procedures were identified using
Current Procedural Terminology (CPT) codes (Table S2, available
Model covariates included patient gender, age at diagnosis and age at www.aaojournal.org). Patients who underwent PPV combined
at treatment, year of diagnosis and treatment, and diagnosis codes for with SB were classified as having undergone PPV, because CPT
relevant ocular comorbidities (myopia, pseudophakia, vitreous code 67108 encompasses RRD repair by vitrectomy with or
hemorrhage, lattice degeneration, and diabetes mellitus). Diagnoses without SB.
were identified using International Classification of Diseases, 9th
and 10th editions, diagnosis codes (Table S1). All comorbidities
Statistical Analysis
were identified by presence of any diagnosis code preceding the
index date of RRD diagnosis or any diagnosis code within 60 Our first outcome variable (receipt of any repair) was binary, and
days after the index date of RRD diagnosis (to account for our second (type of repair received) was a nonordered categorical
patients whose ocular comorbidities might not have been recorded variable. These different types of outcome variables require
until they sought treatment with RRD from an ophthalmologist). different model specifications. We used multilevel mixed-effects
To characterize provider-level variation in initial receipt of repair,
we used the physician identification variable at the time of a patient’s
first diagnosis of RRD. To characterize provider-level variation in
the type of repair intervention used, we used the physician identi-
fication variable from the encounter at which patients received
treatment, where applicable. To characterize geographic-level vari-
ation, we used patients’ locations within 1 of the Combined Statis-
tical Areas designed by the United States Census Bureau or within a
Core-Based Statistical Area for patients residing outside of a Com-
bined Statistical Area. For models that included patients with and
without repair, we used geographic variables drawn from the visit
with the index RRD diagnosis. For models that exclusively included
patients who underwent repair, we used geographic variables drawn
from the treatment visit.

Outcome Variables
Figure 1. Map showing percentage of incident rhegmatogenous retinal
We investigated variation in RRD management using 2 outcome detachment (RRD) patients undergoing surgical repair by state, 2008
variables: (1) a binary variable indicating whether a patient with through 2016. States with fewer than 100 patients diagnosed with RRD
newly diagnosed RRD underwent any form of repair within 60 excluded. Alaska and Hawaii are not shown; both showed fewer than 100
days of the index diagnosis of RRD and (2) among patients who patients with a diagnosis of RRD in the sample.

99
 Ophthalmology Volume 127, Number 1, January 2020

logistic models to characterize variation in whether patients un-

derwent repair for RRD and multilevel mixed-effects multinomial
models to characterize variation in type of repair received.
Receipt of Any Rhegmatogenous Retinal Detachment
Repair. Our multilevel mixed logistic model included patient-level
characteristics (such as age, gender, year of diagnosis, and comor-
bidities) as fixed effects and physician-level and geographic-level
characteristics as random effects. Thus, at the patient level, we
estimated the effects of specific patient characteristics on the prob-
ability of undergoing repair, and at the physician and geographic
levels, we characterized how variation affects practice patterns in
aggregate (as opposed to estimating the effect of an individual
physician’s identity on a patient’s probability of undergoing repair).
To avoid including patients whose first diagnosis of RRD repre-
sented the first report of an old diagnosis or patients whose first
diagnosis with RRD was a mistaken diagnosis that was recoded at a
subsequent visit (either of which would cause us to underestimate
the repair rate for new RRDs in this sample), we imposed additional
sample restrictions in these models. First, we limited our sample to
patients who were seen by a probable optometrist, ophthalmologist,
or retina specialist in the year preceding the index diagnosis and who
had no code for RRD at that visit. Second, we included only patients Figure 2. Pie chart showing relative contributions of patient-, provider-,
who underwent a follow-up visit with a probable optometrist, and geographic-level factors to variation in receipt of primary rhegmatog-
ophthalmologist, or retina specialist in the 60 days after the index enous retinal detachment repair based on intraclass correlation of model 3
diagnosis to avoid including patients who were lost to follow-up. in Table 2.
Finally, we restricted our sample to patients who had a diagnosis
code for RRD at both the index diagnosis visit and the follow-up eye
care visit to avoid including patients who were misdiagnosed with sample to the 4146 patients who were seen by a probable optometrist,
RRD and subsequently were referred to a specialist who concluded ophthalmologist, retina specialist, or combination thereof in the year
they had no RRD. preceding the index diagnosis; had a second appointment with a
We sequentially evaluated 3 models: first an empty model probable eye care provider within 60 days of the index diagnosis; and
(model 1) including no fixed effects and including geography as a had the diagnosis of RRD confirmed at the follow-up appointment.
random effect, second adding patient-level fixed effects (model 2), Among patients with identified RRD repair procedures, the most
and third adding a physician-level identifier nested within the common procedure performed was PPV (49%), followed by laser
geographic level as a second random effect (model 3). We used a barricade (23%), SB and PR (both 11%), and cryotherapy (5%). The
linear specification of our model to estimate an intraclass correla- sample cohort comprised 56% men, with a mean age of 53 years at
tion coefficient that summarizes the contribution of each level of diagnosis, and the most common ocular comorbidities were lattice
the model (patient, physician, and geographic factor) to explaining degeneration (19%) and vitreous hemorrhage (13%; Table 1). The
overall variation.20 mean age among women was slightly lower than that of men (51.9
Choice of Rhegmatogenous Retinal Detachment Repair years vs. 53.0 years), with slightly more women 30 to 39 years of
Procedure. For a categorical outcome (type of procedure age than men (7.6% vs. 5.0%) and slightly fewer women older than
received), we used generalized structural equation modeling to 60 years compared with men (28.4% compared with 33.2%). Nine
specify a multilevel mixed-effects model with a multinomial dis- thousand two hundred forty-four patients (72%) with an RRD diag-
tribution and a logistic link function, using the gsem command in nosis were diagnosed by a probable retinal specialist, and 92% were
Stata version 13.1 (Stata Corp., College Station, TX).21,22 We used diagnosed by a probable eye care provider (retina specialist, other
multilevel mixed-effects multinomial modeling to calculate the ophthalmologist, or optometrist). Patients who underwent docu-
probability of undergoing PR, SB, laser barricade, or cryotherapy mented RRD repair experienced a mean of 10.24 visits with a docu-
relative to PPV (the most common procedure in this sample). Es- mented RRD diagnosis, compared with a mean of 3.48 visits with an
timates of the contribution of different patient-level characteristics RRD diagnosis for patients who did not have documented RRD repair.
at predicting each type of repair procedure (relative to the reference
procedure) were reported as fixed effects. Because intraclass cor- Geographic, Physician, and Patient Effects on
relation coefficients cannot be estimated from a multinomial Receipt of Any Rhegmatogenous Retinal
multilevel mixed-effects model, we instead constructed a similar Detachment Repair
measure from the estimated variance by each level using variance
partition coefficients and used this to compare the relative contri- In our base model (model 1), we determined variance with
bution of each level of the model to explaining total model geographic random effects and no fixed effects (Table 2). Patient-
variation in practice patterns.23 level fixed effects were added in model 2, which slightly reduced
the estimated contribution of geography to overall model variance.
Inclusion of a physician-level identifier nested within the
Results geographic level as a second random effect (model 3) indicated that
both geography and physician factors contributed significantly to
Study Sample Characteristics overall variance; however, physician-level variation showed a
substantially greater impact on whether patients underwent RRD
Our sample consisted of 12 779 patients with incident RRD. Most treatment than geographic-level variation (estimated variance co-
patients had an observed record of treatment within 60 days of diag- efficients of 1.09 and 0.32, respectively). Our results also indicated
nosis (62%), but the rate of repair rose to 92% when we restricted our that women are significantly less likely to undergo repair (odds

100
 Table 3. Multilevel Mixed-Effects Multinomial Model of Procedure Type among Patients Who Underwent Rhegmatogenous Retinal Detachment Repair, 2008e2016 (n ¼ 9314)

Pneumatic Retinopexy* Scleral Buckle* Laser Barricade* Cryotherapy* Overall Variance

Estimate
b (95% CI) P Value b (95% CI) P Value b (95% CI) P Value b (95% CI) P Value (95% CI)
Fixed effects, patient
characteristics

Vail et al
Gender
(reference ¼ male)
Female e0.01 (e0.19 to 0.17) 0.90 0.12 (e0.04 to 0.28) 0.15 0.56 (0.41e0.72) <0.01 0.10 (e0.13 to 0.33) 0.41
Age at surgery 0.01 (0.00e0.02) 0.02 e0.05 (e0.05 to e0.04) <0.01 e0.02 (e0.03 to e0.01) <0.01 0.00 (e0.02 to 0.01) 0.50
(1-yr increments)


Year of surgery 0.07 (0.03e0.11) <0.01 e0.09 (e0.12 to e0.05) <0.01 0.09 (0.05e0.12) <0.01 e0.07 (e0.10 to e0.02) <0.01

Variation in RRD Management

(1-yr increments)
Clinical comorbidities
Myopia e0.33 (e0.81 to 0.14) 0.17 0.19 (e0.17 to 0.56) 0.30 e0.37 (e0.77 to 0.02) 0.06 0.21 (e0.83 to 0.41) 0.50
Pseudophakia e0.77 (e1.07 to e0.46) <0.01 e0.97 (e1.29 to e0.63) <0.01 e1.53 (e1.86 to e1.21) <0.01 e1.53 (e2.07 to e0.99) <0.01
Vitreous hemorrhage e0.64 (e0.90 to e0.37) <0.01 e1.07 (e1.38 to e0.76) <0.01 0.32 (0.11e0.52) <0.01 e0.33 (e0.67 to e0.00) 0.05
Lattice degeneration e0.40 (e0.63 to e0.16) <0.01 e0.02 (e0.17 to 0.21) 0.83 0.18 (0.00e0.37) 0.05 e0.34 (e0.65 to e0.03) 0.03
Diabetes mellitus e0.27 (e0.54 to 0.01) 0.06 e0.03 (e0.22 to 0.29) 0.79 0.01 (e0.22 to 0.25) 0.91 e0.17 (e0.54 to 0.19) 0.35
Random effects
Geography 1 (constrained) e1.61 (e3.05 to e0.16) 0.03 e2.40 (e4.40 to e0.39) 0.02 0.03 (e0.88 to 0.93) 0.96
(CSA or CBSA)
Physician (nested 1 (constrained) 0.18 (0.07e0.29) <0.01 1.18 (1.06e1.31) <0.01 1.15 (0.97e1.33) <0.01
within CSA or CBSA)
Variance by level for
entire model
Geography (CSA) 0.07
(0.02e0.26)
Physician (nested 3.37
within CSA) (2.65e4.28)

CBSA ¼ Core-Based Statistical Area; CI ¼ confidence interval; CSA ¼ Combined Statistical Area based on United States Census.
*All outcomes compared relative to likelihood of pars plana vitrectomy.
101
 Ophthalmology Volume 127, Number 1, January 2020

Figure 3. Map showing choice of primary rhegmatogenous retinal detachment (RRD) repair procedure by state, 2008 through 2016. Data are unadjusted for
provider and patient characteristics. States with fewer than 100 patients diagnosed with RRD are excluded. Alaska and Hawaii are not shown; both showed
fewer than 100 patients with a diagnosis of RRD in the sample.

ratio, 0.62; 95% confidence interval, 0.48e0.80). Interaction receipt of treatment than patient-level characteristics. Patient-level
between age and gender was not statistically significant Table S3, characteristics were estimated to represent approximately 82% of
available at www.aaojournal.org). Patients with diabetes were less overall variation, compared with 16% from physician-level factors
likely to undergo any repair (odds ratio, 0.62; 95% confidence and only 2% from geographic-level factors (Fig 2).
interval, 0.45e0.87).
A linear version of this model was estimated and used to Geographic, Physician, and Patient Effects on
calculate intraclass correlation coefficients at each level of the Choice of Rhegmatogenous Retinal Detachment
model. We found that although geography is a significant deter-
Repair Procedure
minant of undergoing treatment and unadjusted rates of treatment
do vary by state (Fig 1), the effect of geography is dramatically In multilevel mixed-effects multinomial regression modeling of
overshadowed by physician-level variation, and both geographic- patient-level fixed effects, we found that older age was associated
and provider-level factors contributed much less to determining with higher likelihood of PR than PPV and a higher likelihood of

102
 Vail et al 
Variation in RRD Management

PPV than of SB or laser barricade. Patients with a diagnosis of

vitreous hemorrhage were more likely to be managed with PPV
than with PR, SB, or cryotherapy, and pseudophakic patients were
more likely to be managed with PPV than with any other approach.
Patients with lattice degeneration were more likely to be managed
with PPV or laser barricade than PR or cryotherapy (Table 3).
Using geographic-level and physician-level random effects to
estimate variance by level for the aggregate model, we found
estimated variance coefficients of 0.07 for the geographic level and
3.37 for the physician level (Table 3). As found with receipt of any
RRD repair, type of repair received did vary by geography (Fig 3),
but physician-level and patient-level characteristics far outweighed
geographic-level characteristics. Physician-level factors accounted
for approximately 50% of the total variation, followed by patient-
level factors (49%) and geographic-level factors (1%; Fig 4).

Discussion

Using a large nationally representative database of 2008

through 2016 administrative claims data from commercial
insurers, we characterized variation in management of
incident RRD and evaluated the respective effects of
patient-, physician-, and geographic-level factors on Figure 4. Pie chart showing relative contributions of patient-, provider-,
explaining observed variation in RRD care. We found that and geographic-level factors to variation in type of primary rhegmatogenous
geographic variation is a small contributor to RRD man- retinal detachment repair based on variance partition coefficient of
agement overall. Patient characteristics account for most of multinomial model in Table 3.
the variation in whether patients initially receive treatment,
with less contribution from physician practice characteris- of repair, which is consistent with evidence that PPV is
tics. However, choice of specific repair procedure exhibits superior in these settings.8,9,15,32,33 Wide-angle viewing
greater variation at the physician level. systems in PPV enable better visualization of small pe-
Geographic variation is well studied in the medical ripheral pseudophakic breaks, and adequate visualization of
literature20,24e27 and has been evaluated in ophthalmology retinal breaks obscured by vitreous hemorrhage can be
for conditions such as cataract surgery, strabismus, and achieved only through PPV.
age-related macular degeneration treatment.28e31 However, Although decisions regarding RRD repair are determined
although prior analyses have described changes in RRD primarily by patient-level attributes such as comorbidities
management over time and have demonstrated geographic and clinical presentation, there are differences in RRD
differences in use of repair procedures,5,6 geographic vari- management that cannot be explained by patient character-
ation has not been well studied for RRD. Our results build istics alone. Physician-specific practice patterns and the
on prior work, including a 2012 study that found evidence prevailing practice patterns in a geographic area also are
of variation in RRD repair across 4 large United States important determinants of RRD management. We found that
regions using 2009 Medicare claims data, but did not physician-level factors had a much stronger effect on vari-
account for patient characteristics or physician preferences ation than geographic-level factors, both for presence of any
that might affect practice patterns in each region.5 In this RRD repair (16% of variance attributed to physician-level
analysis, we were able to tease apart the relative influence factors vs. 2% attributed to geographic-level factors) and
of patient, provider, and geographic factors (at the more for choice of specific type of repair procedure (50% of
granular Combined Statistical Area and Core-Based Statis- variance attributed to physician-level factors vs. 1% of
tical Area level) and to quantify the contribution of each of variance attributed to geographic-level factors). Physician
these factors to overall variation in RRD care over a 9-year practice patterns were the leading contributor to variation in
period. We found that RRD repairs are not performed in the choice of repair procedure.
same manner in every state, but that at the national level, the There are many potential drivers of these physician- and
vast majority of variation in practice is attributable to factors geographic-level differences. Likely the leading contributor
other than geography. to physician-level variation is differences in physician age
Our findings demonstrate that patient characteristics or training and different levels of comfort or experience in
predict much of the variation in care (49%e82%), sug- performing one type of procedure over another. However,
gesting that retinal surgeons are using evidence to inform physicians also may have access to different surgical
their patient care decisions and are responsive to new equipment, and density of qualified ophthalmologists or
research clarifying the efficacy of different procedures in vitreoretinal specialists may vary by region. Patients may
different patient populations. For example, in our analysis, live far away, lack easy transportation, have preferences
pseudophakic patients and patients with vitreous hemor- regarding surgery in the operating room versus a clinic
rhage were more likely to undergo PPV than any other type procedure, or a combination thereof. Also, providers may

103
 Ophthalmology Volume 127, Number 1, January 2020

demonstrate regional clustering in practice techniques follow-up, or patients who underwent RRD repair under a
because of stable patterns of recruitment from the same different health insurance plan. However, we minimized the
programs, because of influence by senior physicians or key impact of these factors by limiting our analysis of variation
opinion leaders in the region, or both. Geographic-level in RRD repair to patients who were seen by a probable eye
variation seems to play a slightly greater role for the care provider and had at least 1 other eye care visit with an
initial decision to perform repair than for the choice of RRD diagnosis within 60 days after the index diagnosis, and
specific RRD repair procedure, suggesting that access to we limited our analysis of variation in type of repair to
care, regional confluence in practice patterns, or both may patients with known RRD repair procedures. Despite the
have more influence on referral for repair than on the type of imprecision in claims data, the scale and duration of avail-
procedure performed. able clinical information provide a valuable view of real-
Another noteworthy finding from this study is lower odds world RRD management that cannot be replicated with
of RRD repair among women, even after adjusting for age smaller, if richer, single-institution datasets.
and other factors. This may reflect multiple considerations.
For example, older women are more likely to be widowed Conclusions and Implications
and live alone compared with men based on census and
survey data from 2014 (37% of women vs. 13% of men Rhegmatogenous retinal detachment can be managed suc-
widowed, and 32% of women vs. 18% of men living cessfully through a variety of approaches, some of which
alone).34 Therefore, older men may be more likely to have a may be preferred to others based on clinical presentation and
younger spouse able to take on the role of caregiver, patient characteristics. However, the influence of physician
including transportation for surgery and eye examinations and geographic factors is not well understood. In this
and assistance with preoperative and postoperative home analysis of more than 150 million commercially insured
care, whereas older women may outlive an older spouse United States patients, we evaluated 12 779 cases of incident
and lack readily available caregivers. Similarly, it is RRD and found that treatment decisions were responsive to
possible that women may be more likely to take on the patient-level characteristics but also were determined sub-
role of caregiver for another family member, including stantively by physician-level variation and less so by
spouse or parent, and have less time to devote to their geographic variation. These findings indicated a need for
own health care. Women also may be more subject to additional studies to understand drivers behind physician-
errors in coding. Female patients are known to be and geographic-level differences in care and associated
disproportionately affected by age-related macular degen- clinical outcomes as well as to identify existence of barriers
eration, for example, and some cases of wet age-related in access to care.
macular degeneration (e.g., subretinal fluid over an area of
choroidal neovascularization) may be coded mistakenly as a
retinal detachment in administrative claims.35e37 Other un- References
observed confounders, such as educational level, socioeco-
nomic status, and behavioral and environmental factors, also 1. Burton TC. Recovery of visual acuity after retinal detachment
may affect RRD repair rates. However, the important pos- involving the macula. Trans Am Ophthalmol Soc. 1982;80:
sibility of gender-related disparities in RRD management 475e497.
warrants further study. 2. Velez-Montoya R, Jacobo-Oceguera P, Flores-Preciado J,
et al. Primary repair of moderate severity rhegmatogenous
retinal detachment: a critical decision-making algorithm. Med
Strengths and Limitations Hypothesis Discov Innov Ophthalmol. 2016;5:18e31.
This analysis of administrative claims data allowed us to 3. Wilkes SR, Beard CM, Kurland LT, et al. The incidence of
study the real-world variation in management of RRD in a retinal detachment in Rochester, Minnesota, 1970e1978. Am J
Ophthalmol. 1982;94:670e673.
large population, but nonetheless is subject to inherent
4. Mahmoudi S, Almony A. Macula-sparing rhegmatogenous
limitations of claims databases. For example, we rely on retinal detachment: is emergent surgery necessary?
CPT codes to identify procedures and are thus unable to J Ophthalmic Vis Res. 2016;11:100e107.
distinguish PPV from PPV combined with SB (both falling 5. Hwang JC. Regional practice patterns for retinal detachment
under the same CPT 67108 code). Similarly, we cannot repair in the United States. Am J Ophthalmol. 2012;153:
evaluate the use of surgical techniques (such as buckle type 1125e1128.
and suture vs. scleral tunnel fixation), and we lack clinical 6. Reeves MG, Pershing S, Afshar AR. Choice of primary
data such as visual acuity, time to initial presentation, or rhegmatogenous retinal detachment repair method in US
clinical appearance of the detachment. We also are unable to commercially insured and Medicare Advantage patients,
account for coding or billing errors. Some unrepaired RRDs 2003e2016. Am J Ophthalmol. 2018;196:82e90.
7. Schwartz SG, Flynn HW. Pars plana vitrectomy for primary
in our analysis may represent miscoded cases or old, pre-
rhegmatogenous retinal detachment. Clin Ophthalmol. 2008;2:
viously repaired RRDs that are processed with International 57e63.
Classification of Diseases, 9th edition, codes for “recent” or 8. Heimann H, Bartz-Schmidt KU, Bornfeld N, et al. Scleral
“unspecified” RRDs instead of “old” ones. Other cases may buckling versus primary vitrectomy in rhegmatogenous retinal
represent detachments with poor chance of visual recovery detachment: a prospective randomized multicenter clinical
(e.g., blind eye with longstanding total RD), patients lost to study. Ophthalmology. 2007;114:2142e2154.

104
 Vail et al 
Variation in RRD Management

9. Brazitikos PD, Androudi S, Christen WG, Stangos NT. Pri- 25. Farrow DC, Hunt WC, Samet JM. Geographic variation in the
mary pars plana vitrectomy versus scleral buckle surgery for treatment of localized breast cancer. N Engl J Med. 1992;326:
the treatment of pseudophakic retinal detachment: a random- 1097e1101.
ized clinical trial. Retina. 2005;25:957e964. 26. McKellar MR, Landrum MB, Gibson TB, et al. Geographic
10. Schaal S, Sherman MP, Barr CC, Kaplan HJ. Primary retinal variation in quality of care for commercially insured patients.
detachment repair: comparison of 1-year outcomes of four Health Serv Res. 2017;52:849e862.
surgical techniques. Retina. 2011;31:1500e1504. 27. Keating NL, Landrum MB, Huskamp HA, et al. Dartmouth
11. Centers for Medicare and Medicaid Services. Physician fee Atlas area-level estimates of end-of-life expenditures: how
schedule. https://www.cms.gov/apps/physician-fee-schedule/ well do they reflect expenditures for prospectively identified
search/search-criteria.aspx; 2018. Accessed 10.04.18. advanced lung cancer patients? Health Serv Res. 2016;51:
12. Lumi X, Luznik Z, Petrovski G, et al. Anatomical success rate 1584e1594.
of pars plana vitrectomy for treatment of complex rhegma- 28. Erie JC, Barkmeier AJ, Hodge DO, Mahr MA. High variation
togenous retinal detachment. BMC Ophthalmol. 2016;16:216. of intravitreal injection rates and Medicare anti-vascular
13. Lira RP, Takasaka I, Arieta CE, et al. Cryotherapy vs laser endothelial growth factor payments per injection in the United
photocoagulation in scleral buckle surgery: a randomized States. Ophthalmology. 2016;123:1257e1262.
clinical trial. Arch Ophthalmol. 2010;128:1519e1522. 29. Kauh CY, Blachley TS, Lichter PR, et al. Geographic variation
14. Benson WE, Chan P, Sharma S, et al. Current popularity of in the rate and timing of cataract surgery among US commu-
pneumatic retinopexy. Retina. 1999;19:238e241. nities. JAMA Ophthalmol. 2016;134:267e276.
15. Day S, Grossman DS, Mruthyunjaya P, et al. One-year out- 30. Ehrlich JR, Anthopolos R, Tootoo J, et al. Assessing
comes after retinal detachment surgery among medicare ben- geographic variation in strabismus diagnosis among children
eficiaries. Am J Ophthalmol. 2010;150:338e345. enrolled in Medicaid. Ophthalmology. 2016;123:2013e2022.
16. Hillier RJ, Felfeli T, Berger AR, et al. The pneumatic reti- 31. Janetos TM, French DD, Beaumont JL, Tanna AP. Geographic
nopexy versus vitrectomy for the management of Primary and provider variations in ocular hypotensive medication
Rhegmatogenous Retinal Detachment Outcomes Randomized claims among Medicare Part D enrollees. J Glaucoma.
Trial (PIVOT). Ophthalmology. 2019;126:531e539. 2019;28:e29ee33.
17. Wilkinson CP. Interventions for asymptomatic retinal breaks 32. Sharma YR, Karunanithi S, Azad RV, et al. Functional and
and lattice degeneration for preventing retinal detachment. anatomic outcome of scleral buckling versus primary vitrec-
Cochrane Database Syst Rev. 2014:CD003170. tomy in pseudophakic retinal detachment. Acta Ophthalmol
18. Lykins J, Wang K, Wheeler K, et al. Understanding Toxo- Scand. 2005;83:293e297.
plasmosis in the United States through "Large Data" analyses. 33. Ahmadieh H, Moradian S, Faghihi H, et al. Anatomic and
Clin Infect Dis. 2016;63:468e475. visual outcomes of scleral buckling versus primary vitrectomy
19. Azad TD, Vail D, O’Connell C, et al. Geographic variation in in pseudophakic and aphakic retinal detachment: six-month
the surgical management of lumbar spondylolisthesis: char- follow-up results of a single operationdreport no. 1.
acterizing practice patterns and outcomes. Spine J. 2018;18: Ophthalmology. 2005;112:1421e1429.
2232e2238. 34. Stepler R. Smaller share of women ages 65 and older are
20. Aquina CT, Becerra AZ, Justiniano CF, et al. Surgeon, hos- living alone: more are living with spouse or children. Wash-
pital, and geographic variation in minimally invasive colec- ington, DC: Pew Research Center: Social & Demographic
tomy. Ann Surg. 2019;269(6):1109e1116. Trends; 2016.
21. Verkuilen J. Explanatory item response models: a generalized 35. 2010 U.S. Prevalent Cases of AMD (in thousands) by Age,
linear and nonlinear approach by P. de Boeck and M. Wilson Gender, and Race/Ethnicity. National Eye Institute Statistics
and Generalized Latent Variable Modeling: Multilevel, Lon- and Data, 2018. https://nei.nih.gov/eyedata/amd#4; July 2018.
gitudinal and Structural Equation Models by A. Skrondal and Accessed March 17, 2019.
S. Rabe-Hesketh. Psychometrika. 2006;71:415e418. 36. Vision Problems in the US: Age-Related Macular Degeneration.
22. StataCorp. Stata. 13 ed: StataCorp; 2013. Prevent Blindness America. 2012. http://www.visionproblems
23. Browne WJ, Subramanian SV, Jones K, Goldstein H. Variance us.org/news-resources/fact-sheets/2012-VPUS-Eye-Condi-
partitioning in multilevel logistic models that exhibit over- tion-Factsheets/2012_AMD_FS.pdf; 2012. Accessed March 17,
dispersion. J Royal Stat Society Series A (Statistics in Society). 2019.
2005;168:599e613. 37. Zambelli-Weiner A, Crews JE, Friedman DS. Disparities in
24. McDonald DC, Carlson K, Izrael D. Geographic variation in adult vision health in the United States. Am J Ophthalmol.
opioid prescribing in the U.S. J Pain. 2012;13:988e996. 2012;154(6 Suppl):S23eS30.

Footnotes and Financial Disclosures

Originally received: February 11, 2019. Financial Disclosure(s):
Final revision: April 6, 2019. The author(s) have made the following disclosure(s): S.P.: Consultant e
Accepted: April 9, 2019. Acumen, LLC (Burlingame, CA), Verana Health (San Francisco, CA).
Available online: April 12, 2019. Manuscript no. 2019-314. Supported by the Stanford University School of Medicine MedScholars
1
Byers Eye Institute, Department of Ophthalmology, Stanford University Fund, Palo Alto, California (D.V., M.-G.R.); the National Institutes of
School of Medicine, Palo Alto, California. Health, Bethesda, Maryland (grant no.: R03AG056453 [S.P.] and
2 K23EY027466 [A.R.A.]); Research to Prevent Blindness, Inc., New York,
Veterans Affairs Palo Alto Health Care System, Palo Alto, California. New York (departmental support to the Department of Ophthalmology,
3
Department of Ophthalmology, University of California, San Francisco, Stanford University School of Medicine [S.P], and University of California
San Francisco, California. San Francisco [A.R.A.]; and Career Development Award A130755
[A.R.A.]); the National Eye Institute, and National Institutes of Health,

105
 Ophthalmology Volume 127, Number 1, January 2020
Bethesda, Maryland (A.R.A.); and That Man May See, Inc., San Francisco, Obtained funding: Vail, Pershing, Reeves, Afshar
California (A.R.A.). Overall responsibility: Vail, Pershing, Reeves, Afshar
HUMAN SUBJECTS: Human subjects were included in this study. This Abbreviations and Acronyms:
secondary analysis of de-identified administrative data was determined to
CPT ¼ Current Procedural Terminology; PPV ¼ pars plana vitrectomy;
be exempt from institutional review board approval. PR ¼ pneumatic retinopexy; RRD ¼ rhegmatogenous retinal detachment;
No animal subjects were included in this study. SB ¼ scleral buckle.
Author Contributions: Correspondence:
Conception and design: Vail, Pershing, Reeves, Afshar Suzann Pershing, MD, MS, Byers Eye Institute, Department of Ophthal-
Analysis and interpretation: Vail, Pershing, Afshar mology, Stanford University School of Medicine, 2452 Watson Court, Palo
Data collection: Vail, Pershing Alto, CA 94304. E-mail: [email protected].

Pictures & Perspectives

A Baby with Raccoon Eyes

A 10-month-old boy was observed for 2 months’ history of recurrent bruising around the eyes. His parents denied any history of trauma.
On ophthalmic examination, bilateral periorbital ecchymosis (raccoon eyes) and subconjunctival hemorrhages were noted (Fig A).
Computed tomography shows bilateral extensive mass surrounding bony orbit with widespread skull bone erosion (Fig B). Positron
emission tomography-computed tomography revealed a large mass within the left adrenal gland region (Fig C). A biopsy of the abdominal
mass and histopathology revealed small round blue cell tumor with background neuropil (Fig D, 40, arrow). A final diagnosis of
neuroblastoma with orbital metastasis was thus made. (Magnified version of Fig A-D is available online at www.aaojournal.org.)

WEI XIAO, MD
CHAOCHAO XU, MD
YUXIANG MAO, MD
State Key Laboratory of Ophthalmology, Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China

Footnotes and Financial Disclosures

Financial Disclosure(s):
Supported by the Natural Science Foundation of Guangdong Province, China (2016A030310230, 2017A030313613)

106
Quantitative Assessment of the Retina
Using OCT and Associations with
Cognitive Function
Yoshikazu Ito, MD,1,* Mariko Sasaki, MD, PhD,1,2,3,* Hiroki Takahashi, MD,1,2 Shoko Nozaki, MD, PhD,4
Shinichiro Matsuguma, MD, PhD,1 Kaoru Motomura, MD,1 Rihito Ui, MD,1,2 Ryo Shikimoto, MD,4
Ryo Kawasaki, MD, PhD,5 Kenya Yuki, MD, PhD,1 Norie Sawada, MD, PhD,6 Masaru Mimura, MD, PhD,4
Kazuo Tsubota, MD, PhD,1 Shoichiro Tsugane, MD, PhD6

Purpose: To determine the association of retinal thickness with cognitive function in Japanese persons.
Design: Cross-sectional, population-based survey.
Participants: A total of 1293 Japanese persons aged 65 to 86 years who resided in the Saku area in the
Japan Public Health CentereBased Prospective Study participated in the eye and mental health screening.
Methods: Participants underwent comprehensive ophthalmic assessment, including fundus photography,
measurement of intraocular pressure, and determination of refraction status. We assessed the thickness of the
macular retinal nerve fiber layer (mRNFL), ganglion cell-inner plexiform layer (GC-IPL), and ganglion cell complex
(GCC, which includes the retinal nerve fiber layer and GC-IPL), and the full thickness in the macula and peri-
papillary retinal nerve fiber layer (ppRNFL) using spectral-domain (SD) OCT. Cognitive tests consisted of the Mini-
Mental State Examination, Wechsler Memory Scale Revised logical memory I/II subtest, clock drawing test, and
Clinical Dementia Rating Scale. These were used to designate the participants in the following 3 groups: Normal,
those with mild cognitive impairment (MCI), and those with dementia. Multivariable logistic regression models
were used to analyze associations between retinal thickness and cognitive function after adjusting potential
confounding factors.
Main Outcome Measures: Association of retinal thickness with cognitive function.
Results: Among the 1293 potential subjects, 114 were excluded for a diagnosis of depression, 64 were
excluded for retinal disease, and 140 were excluded for scanning errors or suboptimal OCT images. The
remaining 975 participants (mean age, 73.2 years) were included in this analysis. Significant differences were
found in the 3 groups in all layers and GCC thickness, but not in ppRNFL thickness. After adjusting for age, sex,
educational status, and refraction, full macular thickness and GCC thickness were inversely associated with the
presence of dementia, but ppRNFL thickness was not. Furthermore, GC-IPL, GCC, and full macular thicknesses
were all associated with the presence of dementia in the inferior sectors.
Conclusions: Macular thickness was associated with the presence of dementia, but ppRNFL was not. Our results
suggest that OCT measurements of the macula could be superior to those of the ppRNFL in assessing neurode-
generative changes and a potentially useful diagnostic biomarker of cognitive function. Ophthalmology 2020;127:107-
118 ª 2019 by the American Academy of Ophthalmology
Supplemental material available at www.aaojournal.org.
See Commentary on page 119.

As the elderly population is expanding at an alarming rate, region,4 assessment of retinal changes could indicate
the population of patients with dementia has been rapidly subclinical changes much earlier before the clinical onset
increasing. This population was estimated to be 24 million of AD.
worldwide in 2011 and to double every 2 decades.1 Case-control studies5,6 and a meta-analysis7 have
Alzheimer’s disease (AD), a leading cause of dementia, is reported that reduction in the thickness of the peripapillary
characterized by chronic inflammation, glial dysfunction, retinal nerve fiber layer (ppRNFL), including RGC axons,
and synaptic loss in the brain decades before clinical is related to cognitive impairment. The ppRNFL thickness
onset, due to abnormal processing of amyloid and tau in these studies was mainly assessed using time-domain
proteins.2 Of note, Ab plaques have also been found in OCT. OCT is an imaging technique based on low-
the inner retina of patients with AD, especially in retinal coherence interferometry, and it captures images with
ganglion cell (RGC) layers.3 Because RGC and optic micrometer resolution. Recently, OCT has advanced
nerve axonal loss occur before damage to the hippocampal remarkably from time-domain OCT to spectral domain (SD)

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.05.021 107

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

OCT, which provides higher resolution and more precise OCT Measurement
segmentation and measurement of the retina.8 These
OCT scanning was performed by 2 experienced examiners using the
advances have allowed SD OCT to measure specific RS-3000 Advance OCT with eye tracking. For macular thickness,
retinal layer thicknesses, including the thicknesses of the raster scanning over a 99-mm area centered on the fovea was
RNFL, ganglion cell-inner plexiform layer (GC-IPL), and performed at a scan density of 512 A-scans (horizontal) 128 B-
ganglion cell complex (GCC, which includes the RNFL and scans (vertical). For ppRNFL thickness, raster scanning over a 66-
GC-IPL) in the macula. The RGC body is 10 to 20 times mm area centered on the optic disc was performed at a scan density
larger than the diameter of its axon, and more than 50% of of 512 A-scans (horizontal) 128 B-scans (vertical).12 For OCT
RGCs reside in the macula;9 thus, macular thickness may measurements in the macula, 2 grids were used: the Early
have advantages over ppRNFL thickness as an indicator Treatment Diabetic Retinopathy Study (ETDRS) grid and the
of neurodegenerative changes in the retina. However, glaucoma analysis grid. The ETDRS grid consists of 3 concentric
these neurodegenerative changes were mostly investigated circles with diameters of 1, 3, and 6 mm. The inner and outer
rings were each divided into 4 quadrants (superior, inferior, nasal,
in hospital-based, case-control studies5-7 and assessed and temporal) (Fig 2A, upper right). The glaucoma analysis grid
ppRNFL thickness.5-7,10 No field studies, to our knowledge, consists of 2 concentric subfields with diameters of 1.5, 4.5, and
have compared ppRNFL thickness with macular thickness 9 mm (inner and outer) divided into 4 quadrants (superotemporal,
using SD OCT. inferotemporal, superonasal, and inferonasal for parafoveal) or 2
This study included 1293 residents located in the Saku sectors (superior and inferior) (Fig 2A, bottom right). The
area, Nagano, in the Japan Public Health CentereBased thickness of the ppRNFL was evaluated in a circle with a
Prospective Study. Participants underwent comprehensive diameter of 3.45 mm, which was divided into 4 quadrants
mental health and ophthalmic examinations including OCT (superior, inferior, nasal, and temporal) (Fig 2B). Thickness of the
measurements. The goal of this study was to determine the macular retinal nerve fiber layer (mRNFL) (the layer between the
association of ppRNFL and macular thickness, including all internal limiting membrane [ILM] and the outer border of the
RNFL), GC-IPL (the layer between the RNFL and the outer
layers and the GCC, with cognitive function and to inves- border of the IPL), and GCC (the layer between the ILM and the
tigate the influence of GP-IPL thickness changes on these outer border of the IPL), full macula thickness (the layer between
associations and the retinal area most closely associated the ILM and the outer border of the retinal pigment epithelium), and
with cognitive impairment. ppRNFL thickness (the same as mRNFL) were automatically pro-
vided by the built-in software.
Only scans with a Signal Strength Index greater than 5 were
Methods used in this analysis. The measurement of the right eye was used
if the Signal Strength Index was 5, whereas the left eye was
Study Population used if that of the right eye was <5. All scans were reviewed,
This study was a cross-sectional population-based survey con- and 2 co-authors (H.T. and Y.I.), who were masked to the
ducted in one of the Japan Public Health Center Study areas11 (the excluded poor-quality scans, corrected the centering of all B-
Saku Public Health Center catchment area, Nagano Prefecture, scan images manually if necessary. Participants with optimal
Japan). The present study population comprised residents in the images of both the macular and peripapillary areas were
area, and there were 12 219 participants (6172 men, 6047 included this analysis.
women) aged 40 to 59 years at the beginning of the study in
1990. We excluded those who moved out of the region, died, or Evaluation of Cognitive Decline
did not respond to subsequent questionnaires. Therefore, 8827
participants were invited to eye and mental health screening Cognitive function was assessed by experienced neuropsycholo-
between May 2014 and March 2016. Of these, 1299 responded, gists using the Mini-Mental State Examination,13 Wechsler
and 1293 underwent both comprehensive psychiatric assessment Memory Scale Revised logical memory I/II subtest,14 clock
and ophthalmic examination (Fig 1). Information regarding drawing test,15 and Clinical Dementia Rating Scale.16 Depressive
educational status and history of glaucoma was obtained using a symptoms were assessed using the Center for Epidemiologic
standardized self-administered questionnaire. Studies Depression Scale,17 and Patient Health Questionnaire-9.18
Refractive status and intraocular pressure were measured by We then categorized the participants’ cognitive function in
using an auto refractometer (Tonoref II, NIDEK Co. Ltd., Tokyo, accordance with the criteria used in the Japanese Alzheimer's Disease
Japan). Nonstereoscopic fundus photographs of both eyes were Neuroimaging Initiative (J-ADNI) project,19,20 in which mild
obtained using a 45
nonmydriatic fundus camera (AFC-330; cognitive impairment (MCI) was defined as amnestic MCI as orig-
NIDEK Co. Ltd.), and OCT images were obtained using the inally described by Petersen et al.21 Memory impairment was defined
RS-3000 Advance OCT (version 1.4.2.11.4.2; NIDEK Co. Ltd.) by a score below the education-adjusted cutoff level on the Wechsler
under dim light conditions without using pharmacologic dilating Memory Scale Revised Logical Memory II test. These cutoff values
agents. Images were centered on the optic disc and macula. The were as follows: 0e9 years of education, 2; 10e15 years, 4; and >15
presence of a nerve fiber layer defect (NFLD), the cup-to-disc years, 8. The Mini-Mental State Examination cutoff point for de-
(C/D) ratio, and retinal disease were determined by 4 ophthal- mentia was set at 23. All the participants were interviewed by trained
mologists (K.M., R.U., K.Y., and S.M.). Participants with retinal psychiatrists to determine the final diagnosis regarding cognitive
disease (i.e., epiretinal membrane, age-related macular degenera- decline and depressive disorder. Diagnosis was made according to
tion, retinal vein occlusion, and diabetic retinopathy) were the Diagnostic and Statistical Manual of Mental Disorders, 4th
excluded from this analysis. Edition criteria.22 Participants who were not demented but had
The study was approved by the Institutional Review Board of clinically significant memory impairment were categorized as
the National Cancer Center Japan and Keio University School of having MCI. Participants who received a diagnosis of depression
Medicine. Written informed consent was obtained from all par- were excluded from this analysis. For sensitivity analysis, patients
ticipants included in the study. with features suggestive of vascular dementia were differentiated

108
 Ito et al 
Assessment of Cognitive Function Using OCT

Figure 1. Flowchart describing subject selection from participants of the Japan Public Health CentereBased Prospective Study. JPHC ¼ Japan Public
Health Center.

from the other types of dementia using information on history of glaucoma. The first model was to determine associations between
stroke. the presence of dementia and retinal thickness with adjustment for
age and sex. The second model was adjusted for age, sex, educa-
tional status, and refraction.
Statistical Analysis Associations were expressed as odds ratios with 95% confi-
Baseline characteristics were compiled for overall samples and for dence intervals per standard deviation decrease. P values less than
groups stratified by cognitive function (normal, MCI, and 0.05 were considered statistically significant. All statistical ana-
dementia). Differences in basic characteristics and retinal mea- lyses were performed using SAS version 9.4 for Windows (SAS
surements between groups were assessed using analysis of variance Institute Inc., Cary, NC).
with the post hoc Tukey test for continuous variables and the chi-
square test or Fisher exact test for categoric variables.
Associations between each retinal measurement and the Results
3 groups were assessed using univariate multinomial logistic
regression models. Associations between each retinal measurement Among the 1293 potential subjects, 114 were excluded for a diag-
and the presence of dementia were assessed using univariate and nosis of depression, 64 were excluded for retinal disease, and 140
2 multivariable logistic regression models. In multivariate logistic were excluded for scanning errors or suboptimal OCT images. The
regression analysis, we excluded participants with features remaining 975 participants (mean age, 73.25.5 years; 427 male
suggestive of glaucoma, such as those with intraocular pressure [43.8%] and 548 female [56.2%]) were included in this analysis.
21 mmHg, a history of glaucoma, and both OCT markers Among them, 6 participants previously diagnosed as having de-
(GCC or RNFL thinning) and glaucomatous fundus changes mentia were included in this analysis; 5 were diagnosed with de-
(glaucomatous disc changes or NFLD) to remove the influence of mentia, and 1 was diagnosed with MCI. Also, participants with

109
 Ophthalmology Volume 127, Number 1, January 2020

Figure 2. Example of spectral-domain (SD) OCTegenerated retinal measurements. A, Macular OCT. Upper left: Color fundus. Upper center and right:
Macular thickness map and averaged macular thickness in the 9 Early Treatment Diabetic Retinopathy Study (ETDRS) grid sectors. Bottom left: Macular b-
scan. Bottom center and right: Macular thickness map and averaged macular thickness in the 8 glaucoma analysis grid sectors. B, Peripapillary OCT. Upper
left: Color fundus. Upper center and right: Peripapillary retinal nerve fiber layer (ppRNFL) thickness map and averaged ppRNFL thickness in 4 sectors.
Bottom left: Retinal b-scan. Bottom right: ppRNFL thickness graph surrounding optic disc.

features suggestive of glaucoma were included: 5 with intraocular Significant differences were observed among the 3 groups in
pressure 21 mmHg, 85 with a history of glaucoma, and 23 with terms of age, sex, refraction, and educational status. In
both OCT markers (GCC or RNFL thinning) and glaucomatous comparisons of OCT measurements in the ETDRS grid,
fundus changes (glaucomatous disc changes or NFLD). significant differences were found among the 3 groups in GCC
thickness in 3 of 8 sectors and in full macular thickness in 7 of
9 sectors. In the glaucoma analysis grid, significant differences
Characteristics and Ganglion Cell Complex, were seen among the 3 groups in GCC thickness in 5 of 10
Peripapillary Retinal Nerve Fiber Layer, and Full sectors and in full macular thickness in 9 of 11 sectors. In post
Macular Thickness of the Study Participants hoc analysis, these differences remained significant between the
Stratified by Cognitive Function dementia and normal groups, and between the dementia and
MCI groups, but not between the normal and MCI groups.
Table 1 presents the characteristics and GCC, ppRNFL, and full There were no significant differences for ppRNFL thickness in
macular thicknesses of the normal, MCI, and dementia groups. any groups.

110
 Ito et al 
Assessment of Cognitive Function Using OCT

Table 1. Characteristics and Ganglion Cell Complex, Peripapillary Retinal Nerve Fiber Layer, and Full Macular Thicknesses of the Study
Participants Stratified by Cognitive Function

Normal MCI Dementia

n [ 613 n [ 324 n [ 38 P Value
Age, y 72.4±5.3 74.2±5.6 77.2±5.6 <0.0001
Sex, male, % 40.0 50.8 44.7 0.007
Refraction, D L0.59±2.22 L0.10±2.16 0.02±1.37 0.003
NFLD, % 4.2 2.9 2.9 0.586
C/D ratio 0.7, % 8.7 9.1 8.8 0.983
IOP, mmHg 13.22.9 12.82.7 13.54.6 0.056
History of glaucoma, % 9.3 7.4 10.5 0.565
Educational status, y 11.7±2.0 11.5±2.2 9.7±2.0 <0.0001
ETDRS grid
Macular/GCC thickness, mm
Inner temporal 100.0 (94.4e105.5) 100.7 (93.7e105.9) 96.9 (86.9e102.6) 0.015
Inner superior 113.2 (106.1e118.6) 112.3 (103.6e119.0) 106.9 (102.2e116.2) 0.016
Inner nasal 111.4 (105.1e117.5) 110.3 (102.5e117.5) 108.0 (102.6e113.2) 0.112
Inner inferior 112.1 (104.8e119.2) 112.5 (104.3e119.2) 109.2 (101.6e118.6) 0.472
Outer temporal 87.4 (82.4e92.5) 87.7 (81.6e92.6) 83.8 (78.8e90.8) 0.077
Outer superior 99.2 (93.8e105.0) 98.8 (92.7e104.9) 98.3 (93.0e102.9) 0.517
Outer nasal 111.3 (105.1e117.3) 111.4 (103.7e117.7) 109.3 (101.2e116.0) 0.309
Outer inferior 97.5 (91.1e103.6) 97.2 (90.5e104.3) 94.4 (81.8e101.4) 0.004
Macular/full thickness, mm
Central 265.1 (251.9e278.1) 268.4 (253.1e282.6) 265.6 (246.3e281.5) 0.240
Inner temporal 324.9 (314.3e334.0) 324.8 (313.3e335.7) 315.6 (300.8e330.0) 0.002
Inner superior 337.4 (326.3e347.1) 336.0 (323.9e347.1) 328.6 (314.5e340.2) 0.019
Inner nasal 339.9 (328.3e351.3) 339.4 (327.4e351.5) 330.3 (317.7e346.6) 0.102
Inner inferior 333.3 (321.8e343.1) 332.8 (319.3e344.7) 321.7 (311.1e336.5) 0.013
Outer temporal 285.8 (276.5e296.1) 286.0 (274.1e296.3) 278.0 (271.2e290.0) 0.004
Outer superior 296.6 (286.6e306.9) 296.0 (286.3e307.7) 287.4 (278.6e302.9) 0.007
Outer nasal 310.4 (298.8e321.5) 311.3 (299.7e324.4) 302.1 (290.2e315.8) 0.023
Outer inferior 288.0 (276.3e299.7) 288.1 (275.7e301.3) 280.4 (263.6e293.0) 0.007
Glaucoma analysis grid
Macular/GCC thickness, mm
Inner superotemporal 103.8 (98.2e109.1) 102.9 (97.4e108.4) 101.3 (96.2e105.6) 0.110
Inner superonasal 115.1 (109.2e120.7) 114.7 (107.8e120.3) 110.1 (105.6e117.4) 0.076
Inner inferonasal 113.3 (108.3e119.2) 113.8(105.5 119.6) 108.6 (100.5e116.9) 0.013
Inner inferotemporal 105.7 (100.3e111.2) 105.1 (99.4e111.8) 100.9 (95.1e107.5) 0.025
Outer superotemporal 71.3 (67.0e76.1) 71.7 (66.9e76.8) 70.3 (65.8e74.8) 0.326
Outer superonasal 105.6 (98.1e113.2) 105.0 (96.5e113.2) 100.1 (90.8e112.0) 0.181
Outer inferonasal 109.9 (100.4e116.2) 109.7 (99.4e118.5) 103.8 (93.6e113.2) 0.055
Outer inferotemporal 72.6 (67.8e77.8) 73.7 (68.0e78.1) 69.3 (62.8e75.9) 0.009
Superior 92.7 (87.4e97.7) 92.0 (85.7e97.9) 88.9 (85.1e95.2) 0.211
Inferior 94.3 (88.3e99.5) 93.9 (87.3e101.0) 90.4 (80.5e96.3) 0.018
Macular/full thickness, mm
Central 290.3 (277.9e303.2) 290.2 (279.6e306.0) 289.4 (270.3e300.3) 0.372
Inner superotemporal 316.4 (306.9e327.0) 317.5 (305.3e327.3) 309.4 (299.0e319.3) 0.010
Inner superonasal 330.7 (321.1e341.0) 331.5 (318.4e343.4) 322.0 (312.5e336.3) 0.062
Inner inferonasal 326.4 (314.8e336.6) 325.6 (314.6e337.3) 315.7 (301.6e329.0) 0.003
Inner inferotemporal 317.2 (306.1e327.1) 316.6 (305.7e328.2) 307.7 (294.2e320.7) 0.007
Outer superotemporal 255.0 (247.3e263.2) 255.4 (246.6e262.9) 249.3 (240.5e256.8) 0.005
Outer superonasal 289.5 (279.0e299.3) 287.7 (277.1e299.0) 283.2 (265.9e295.7) 0.009
Outer inferonasal 290.8 (279.7e300.9) 291.3 (277.3e303.0) 282.8 (268.6e292.5) 0.016
Outer inferotemporal 251.7 (243.2e259.4) 252.0 (242.7e261.5) 244.1 (236.5e251.9) 0.009
Superior 285.0 (275.5e291.9) 283.5 (274.4e293.2) 277.1 (266.2e286.1) 0.005
Inferior 283.1 (273.3e292.3) 283.2 (272.6e294.3) 274.3 (263.2e285.7) 0.002
Peripapillary RNFL thickness, mm
Global 95.7 (85.9e104.3) 95.0 (87.1e103.1) 90.9 (84.4e104.4) 0.791
Temporal 70.7 (62.8e79.4) 71.1 (61.7e77.6) 69.7 (61.7e78.3) 0.153
Superior 116.7 (103.8e129.3) 116.5 (101.7e130.2) 113.0 (97.6e127.0) 0.641
Nasal 67.4 (56.9e80.0) 69.8 (58.6e81.6) 64.4 (58.0e77.1) 0.715
Inferior 120.5 (108.0e135.9) 122.8 (108.8e136.1) 117.0 (99.8e135.5) 0.910

Significant values appear in boldface.

C/D ¼ cup-to-disc ratio; D ¼ diopters; ETDRS ¼ Early Treatment Diabetic Retinopathy Study; GCC ¼ ganglion cell complex (including RNFL, GCL, and
IPL); GCL ¼ ganglion cell layer; IOP ¼ intraocular pressure; IPL ¼ inner plexiform layer; MCI ¼ mild cognitive impairment; NFLD ¼ nerve fiber layer
defect; RNFL ¼ retinal nerve fiber layer.
Data are the mean ( standard deviation [SD]) for normally distributed data or the median (interquartile range) for skewed data unless otherwise stated.

111
 Ophthalmology Volume 127, Number 1, January 2020

Table 2. Univariate Analysis of Associations of Participant Association of Ganglion Cell Complex,

Factors, and Ganglion Cell Complex, Peripapillary Retinal Nerve Peripapillary Retinal Nerve Fiber Layer, or Full
Fiber Layer, and Full Macular Thickness with the Presence of
Macular Thickness with the Presence of
Dementia
Dementia
OR 95% CI Table 2 shows associations of the presence of dementia with
Age, y 1.16 1.10e1.24 potential factors and OCT measurements according to a
Sex, female vs. male 0.82 0.43e1.59 univariate logistic regression model. Among these factors, age
Refraction, D 1.18 0.97e1.42 was positively associated, and educational status was negatively
NFLD 0.69 0.09e5.24 associated with the presence of dementia. In the ETDRS grid,
C/D ratio 0.7 1.01 0.30e3.43 the GCC thickness was inversely associated with the presence of
IOP, mmHg 1.03 0.92e1.14
dementia in 4 of 8 sectors and for full macular thickness in 7 of
History of glaucoma 1.15 0.39e3.35
Educational status, y 0.57 0.46e0.70
9 sectors. In the glaucoma analysis grid, the GCC thickness was
ETDRS grid inversely associated with the presence of dementia in the 6 of 10
Macular/GCC thickness, per 1 SD decrease sectors and for full macular thickness in 10 of 11 sectors.
Inner temporal 1.43 1.14e1.80 However, ppRNFL thickness was not significantly associated
Inner superior 1.78 1.33e2.38 with dementia.
Inner nasal 1.21 0.85e1.73 We excluded participants with features suggestive of glaucoma
Inner inferior 1.10 0.80e1.51 as described earlier and then adjusted for age, sex, refraction, and
Outer temporal 1.39 0.98e1.96 educational status. In the ETDRS grid, GCC thickness was
Outer superior 1.45 1.12e1.86 inversely associated with the presence of dementia in only the inner
Outer nasal 1.11 0.82e1.51 temporal sector and for full macular thickness in 6 of 9 sectors.
Outer inferior 1.36 1.02e1.82
Macular/full thickness, per 1 SD decrease
Associations with dementia were found for both GCC and full
Central 1.01 0.73e1.40 macular thicknesses in the inner temporal sector. In the glaucoma
Inner temporal 1.50 1.21e1.87 analysis grid, GCC thickness was inversely associated with the
Inner superior 1.47 1.12e1.93 presence of dementia in 7 of 10 sectors and for full macular
Inner nasal 1.23 0.96e1.58 thickness in 4 of 11 sectors. Associations with dementia were
Inner inferior 1.32 1.07e1.63 found for both GCC and full macular thicknesses in the inner in-
Outer temporal 1.61 1.23e2.12 ferotemporal, inner inferonasal, and inferior sectors (Table 3,
Outer superior 1.59 1.19e2.11 Model 2). Multinomial logistic regression analysis among the 3
Outer nasal 1.36 1.03e1.79 groups revealed an association between the normal group and
Outer inferior 1.65 1.26e2.17 dementia group, but no association was found between the MCI
Glaucoma analysis grid
group and any OCT measurements. The results were not
Macular/GCC thickness, per 1 SD decrease
Inner superotemporal 1.47 1.16e1.86 markedly changed with the inclusion of participants with
Inner superonasal 1.29 0.99e1.68 depression (data not shown).
Inner inferonasal 1.26 0.98e1.63 Further, we differentiated between participants with features
Inner inferotemporal 1.51 1.21e1.90 suggestive of vascular dementia from those with other types of
Outer superotemporal 1.38 1.04e1.82 dementia using information on stroke history, and 40 participants
Outer superonasal 1.32 0.98e1.79 were identified including 4 with dementia and 13 with MCI.
Outer inferonasal 1.30 0.97e1.75 Among them, significant differences were found between the
Outer inferotemporal 1.64 1.24e2.19 normal and dementia groups in GCC thickness in the inner
Superior 1.44 1.08e1.91 superotemporal (P ¼ 0.049) and outer superotemporal (P ¼ 0.045)
Inferior 1.56 1.19e2.03
quadrants. However, these associations were not maintained in the
Macular/full thickness, per 1 SD decrease
Central 1.20 0.88e1.64 logistic regression analysis. Also, after excluding these partici-
Inner superotemporal 1.50 1.15e1.96 pants, the results were not markedly changed.
Inner superonasal 1.33 1.02e1.73
Inner inferonasal 1.33 1.06e1.67
Inner inferotemporal 1.53 1.19e1.96
Association of Macular Retinal Nerve Fiber Layer
Outer superotemporal 1.48 1.12e1.96 or Ganglion Cell-Inner Plexiform Layer Thickness
Outer superonasal 1.59 1.18e2.14 with the Presence of Dementia
Outer inferonasal 1.42 1.07e1.87
Outer inferotemporal 1.45 1.14e1.86 We also analyzed the associations of the mRNFL or GC-IPL
Superior 1.65 1.22e2.24 thickness with the presence of dementia. Table S1 (available at
Inferior 1.64 1.23e2.19 www.aaojournal.org) presents the mRNFL and GC-IPL thickness
Peripapillary RNFL thickness, per 1 SD decrease of the normal, MCI, and dementia groups. In the ETDRS grid,
Global 0.94 0.71e1.26 significant differences were found among the 3 groups in mRNFL
Temporal 1.02 0.72e1.44
thickness in 1 of 8 sectors and in GC-IPL thickness in none of 8
Superior 0.92 0.69e1.22
Nasal 0.87 0.72e1.06
sectors. In the glaucoma analysis grid, significant differences were
Inferior 1.22 0.87e1.72 seen among the 3 groups in mRNFL thickness in 2 of 10 sectors
and in GC-IPL thickness in 3 of 10 sectors. In post hoc analysis,
these differences remained significant between the dementia and
Significant values appear in boldface. normal groups, and between the dementia and MCI groups, but not
C/D ratio ¼ cup-to-disc ratio; CI ¼ confidence interval; D ¼ diopters; between the normal and MCI groups.
ETDRS ¼ Early Treatment Diabetic Retinopathy Study; GCC ¼ ganglion cell
Table 4 shows associations of mRNFL or GC-IPL thickness
complex (including RNFL, GCL, and IPL); GCL ¼ ganglion cell layer; IOP ¼
intraocular pressure; IPL, inner plexiform layer; NFLD ¼ nerve fiber layer defect; with the presence of dementia according to a univariate logistic
OR ¼ odds ratio; RNFL ¼ retinal nerve fiber layer; SD ¼ standard deviation. regression model. In the ETDRS grid, mRNFL thickness was

112
 Ito et al 
Assessment of Cognitive Function Using OCT

Table 3. Multivariable Analysis of Associations of Ganglion Cell Complex and Full Macular Thickness with the Presence of Dementia

OR, per 1 SD Decrease Tertile 1 Tertile 2 Tertile 3

ETDRS grid
Macular/GCC thickness
Inner temporal
No. of noncases/cases 180/7 177/10 171/15
Thickness, mm 108.1 (105.7e111.9) 100.5 (98.9e102.3) 92.0 (87.4e95.1)
Model 1 1.33 (1.02e1.74) 0.71 (0.25e1.96) Reference 1.27 (0.54e2.97)
Model 2 1.36 (1.01e1.83) 0.68 (0.24e1.95) Reference 1.25 (0.52e2.98)
Macular/full thickness
Inner temporal
No. of noncases/cases 182/8 182/7 171/18
Thickness, mm 340.1 (334.3e346.6) 325.3 (322.1e327.9) 311.3 (301.7e315.0)
Model 1 1.49 (1.12e1.98) 1.29 (0.45e3.75) Reference 2.17 (0.87e5.45)
Model 2 1.44 (1.02e2.04) 1.44 (0.48e4.26) Reference 2.51 (0.98e6.45)
Inner nasal
No. of noncases/cases 180/9 185/5 170/19
Thickness, mm 356.9 (351.3e363.7) 340.8 (336.9e343.7) 324.4 (315.2e329.6)
Model 1 1.23 (0.91e1.64) 2.11 (0.68e6.56) Reference 3.11 (1.11e8.72)
Model 2 1.24 (0.91e1.68) 2.69 (0.84e8.63) Reference 3.45 (1.20e9.91)
Inner inferior
No. of noncases/cases 181/8 183/6 171/19
Thickness, mm 348.8 (343.6e354.9) 333.5 (330.0e335.9) 316.6 (308.2e322.7)
Model 1 1.35 (1.07e1.70) 1.63 (0.54e4.91) Reference 2.69 (1.03e7.05)
Model 2 1.39 (1.08e1.81) 2.18 (0.69e6.91) Reference 3.22 (1.19e8.68)
Outer temporal
No. of noncases/cases 182/7 184/6 169/20
Thickness, mm 300.5 (296.3e307.0) 286.2 (282.9e289.8) 274.4 (264.7e277.6)
Model 1 1.55 (1.12e2.15) 1.18 (0.38e3.78) Reference 2.67 (1.02e7.01)
Model 2 1.50 (1.06e2.11) 1.20 (0.38e3.75) Reference 2.69 (1.01e7.23)
Outer superior
No. of noncases/cases 180/9 184/7 171/17
Thickness, mm 311.4 (307.3e316.8) 296.6 (293.4e299.2) 282.8 (275.6e286.8)
Model 1 1.47 (1.03e2.10) 1.69 (0.60e4.77) Reference 2.11 (0.84e5.32)
Model 2 1.50 (1.05e2.15) 1.81 (0.62e5.29) Reference 2.47 (0.95e6.40)
Outer inferior
No. of noncases/cases 182/6 179/11 174/16
Thickness, mm 305.2 (300.3e312.7) 288.7 (285.4e293.0) 272.1 (264.7e277.6)
Model 1 1.57 (1.18e2.10) 0.58 (0.21e1.63) Reference 1.25 (0.55e2.83)
Model 2 1.61 (1.17e2.22) 0.60 (0.21e1.71) Reference 1.32 (0.57e3.07)
Glaucoma analysis grid
Macular/GCC thickness
Inner superotemporal
No. of noncases/cases 181/8 180/10 174/15
Thickness, mm 111.2 (109.5e115.3) 104.1 (102.6e105.7) 96.1 (91.5e98.8)
Model 1 1.33 (1.02e1.72) 0.84 (0.32e2.21) Reference 1.18 (0.50e2.76)
Model 2 1.34 (1.02e1.78) 0.86 (0.32e2.31) Reference 1.18 (0.49e2.84)
Inner superonasal
No. of noncases/cases 182/8 184/6 169/19
Thickness, mm 123.6 (121.0e127.9) 115.3 (113.7e117.0) 107.1 (100.8e109.5)
Model 1 1.17 (0.86e1.59) 1.56 (0.52e4.68) Reference 2.85 (1.10e7.43)
Model 2 1.18 (0.84e1.65) 1.72 (0.56e5.28) Reference 3.06 (1.14e8.22)
Inner inferonasal
No. of noncases/cases 180/9 184/5 171/19
Thickness, mm 122.1 (119.5e125.8) 113.6 (112.4e115.2) 105.7 (98.9e108.4)
Model 1 1.20 (0.90e1.59) 2.19 (0.71e6.76) Reference 3.35 (1.20e9.29)
Model 2 1.26 (0.92e1.71) 2.24 (0.71e7.06) Reference 3.38 (1.20e9.54)
Inner inferotemporal
No. of noncases/cases 181/8 180/8 174/17
Thickness, mm 113.6 (111.3e116.7) 106.0 (104.5e107.4) 97.3 (92.2e100.8)
Model 1 1.45 (1.14e1.86) 1.18 (0.42e3.27) Reference 1.87 (0.77e4.54)
Model 2 1.47 (1.12e1.93) 1.20 (0.42e3.38) Reference 1.90 (0.77e4.72)
Outer superonasal
No. of noncases/cases 182/8 183/7 170/18
Thickness, mm 116.7 (113.5e121.3) 106.2 (103.9e108.6) 94.8 (86.0e98.9)
Model 1 1.13 (0.79e1.60) 1.62 (0.56e4.71) Reference 2.64 (1.06e6.58)
Model 2 1.17 (0.81e1.68) 1.67 (0.56e5.00) Reference 2.93 (1.14e7.49)

(Continued)

113
 Ophthalmology Volume 127, Number 1, January 2020

Table 3. (Continued.)

OR, per 1 SD Decrease Tertile 1 Tertile 2 Tertile 3

Outer inferotemporal
No. of noncases/cases 182/8 182/8 171/17
Thickness, mm 80.2 (78.1e84.1) 72.8 (71.5e74.8) 65.4 (61.6e67.9)
Model 1 1.48 (1.09e2.00) 0.90 (0.32e2.48) Reference 1.70 (0.70e4.13)
Model 2 1.50 (1.20e2.06) 0.84 (0.30e2.35) Reference 1.85 (0.74e4.62)
Inferior
No. of noncases/cases 182/8 179/9 174/16
Thickness, mm 102.3 (99.8e105.6) 94.8 (92.9e96.2) 85.5 (80.8e88.9)
Model 1 1.43 (1.07e1.92) 1.03 (0.38e2.77) Reference 1.54 (0.65e3.64)
Model 2 1.50 (1.10e2.04) 0.97 (0.35e2.67) Reference 1.81 (0.74e4.43)
Macular/full thickness
Inner inferonasal
No. of noncases/cases 181/6 179/9 168/17
Thickness, mm 341.6 (336.5e347.4) 326.6 (323.5e330.5) 310.6 (301.1e315.3)
Model 1 1.35 (1.04e1.77) 0.71 (0.25e2.08) Reference 1.62 (0.69e3.82)
Model 2 1.39 (1.05e1.83) 0.75 (0.25e2.21) Reference 1.71 (0.71e4.15)
Inner inferotemporal
No. of noncases/cases 181/5 177/10 170/17
Thickness, mm 331.9 (327.2e338.9) 317.8 (314.5e320.9) 303.2 (295.4e307.4)
Model 1 1.57 (1.14e2.18) 0.48 (0.16e1.46) Reference 1.30 (0.56e3.06)
Model 2 1.59 (1.12e2.25) 0.50 (0.16e1.54) Reference 1.32 (0.55e3.18)
Superior
No. of noncases/cases 179/6 177/10 172/16
Thickness, mm 296.4 (292.6e300.9) 285.3 (281.9e287.3) 272.4 (266.7e276.1)
Model 1 1.42 (0.99e2.03) 0.73 (0.26e2.10) Reference 1.28 (0.55e2.96)
Model 2 1.46 (1.01e2.10) 0.68 (0.23e1.99) Reference 1.18 (0.49e2.82)
Inferior
No. of noncases/cases 182/5 178/9 168/18
Thickness, mm 295.6 (292.5e301.1) 283.3 (280.7e286.7) 269.6 (262.9e274.1)
Model 1 1.46 (1.06e2.03) 0.63 (0.21e1.96) Reference 1.78 (0.76e4.17)
Model 2 1.49 (1.06e2.08) 0.66 (0.21e2.08) Reference 1.99 (0.83e4.78)

Significant values appear in boldface.

ETDRS ¼ Early Treatment Diabetic Retinopathy Study; GCC ¼ ganglion cell complex (including RNFL, GCL, and IPL); GCL ¼ ganglion cell layer;
IPL ¼ inner plexiform layer; OR ¼ odds ratio; RNFL ¼ retinal nerve fiber layer; SD ¼ standard deviation.
Model 1: Adjusted for age and sex. Model 2: Adjusted for age, sex, refraction, and educational status.

inversely associated with the presence of dementia in 4 of 8 sectors was not, in our sample. The presence of MCI was not
and for GC-IPL thickness in 1 of 8 sectors. In the glaucoma associated with any OCT measurements. Although MCI is
analysis grid, mRNFL thickness was inversely associated with the generally regarded as a preceding stage of dementia, it may
presence of dementia in 5 of 10 sectors and for GC-IPL thickness include a wide range of spectrums. Furthermore, GC-IPL,
in 5 of 10 sectors.
GCC, and full macular thicknesses were all associated
We excluded participants with features suggestive of glaucoma
as described earlier and then adjusted for age, sex, refraction, and with the presence of dementia in the inferior sectors.
educational status. In the ETDRS grid, mRNFL thickness was A postmortem study showed that RGC loss in the macula
inversely associated with the presence of dementia in only the inner was associated with AD.23 Pathological examination of
superior sector and for GC-IPL thickness in the outer inferior sector retinas from patients with AD showed Ab plaques in the
(Table 5, Model 2). In the glaucoma analysis grid, mRNFL inner retina, especially in the RGC layers, which suggests
thickness was inversely associated with the presence of dementia that deposition of Ab plaques could induce inflammation
in 2 of 10 sectors and for GC-IPL thickness in 5 of 10 sectors. and lead to RGC degeneration.3 The RGC body is 10 to
Associations with dementia were found for both GC-IPL and GCC 20 times larger than the diameter of its axon, and more
thickness in the inner inferotemporal, inner inferonasal, outer in- than 50% of RGCs are in the macula.9 The structure of
ferotemporal, and inferior sectors. Moreover, in the inner infero-
the GCL and IPL, containing RGC bodies and their
temporal, inner inferonasal, and inferior sectors, GP-IPL, GCC,
and full macular thicknesses were all associated with the presence dendrites, respectively, differ less than the RNFL
of dementia. containing axons. These results suggest that retinal
measurements in the macula could be more useful than
those in the ppRNFL for assessing neurodegenerative
Discussion change in AD. In a large case-control study, Cheung
et al24 demonstrated that the reduction of the GC-IPL
We found that GCC and macular thicknesses were associ- thickness in the macula was more strongly associated with
ated with the presence of dementia, but ppRNFL thickness the presence of MCI and AD than the reduction of the

114
 Ito et al 
Assessment of Cognitive Function Using OCT

Table 4. Univariate Analysis of Associations of Macular Retinal reduction of the GCC and full macular thickness due to
Nerve Fiber Layer and Ganglion Cell-Inner Plexiform Layer neurodegeneration caused by dementia. Our findings sup-
Thickness with the Presence of Dementia port those of previous studies24 in which OCT
measurements of the macula were found to be superior to
OR 95% CI
those of the ppRNFL in assessing neurodegenerative
ETDRS grid changes in AD in a field study. The glaucoma analysis
Macular/mRNFL thickness, per 1 SD decrease grid might be useful because of the robust ability to detect
Inner temporal 1.47 1.01e2.15 associations of GP-IPL and GCC with the presence of
Inner superior 1.39 1.06e1.81
Inner nasal 1.26 0.97e1.64
dementia.
Inner inferior 1.19 0.90e1.57 Regarding the most sensitive area, Jones-Odeh and
Outer temporal 1.43 1.11e1.85 Hammond28 summarized 11 studies and concluded that
Outer superior 1.21 0.88e1.65 reduction in thickness was likely to appear in the superior
Outer nasal 1.19 0.89e1.59 and inferior sectors in patients with cognitive impairment.
Outer inferior 1.44 1.09e1.89 Postmortem studies showing that degeneration of the GCL
Macular/GC-IPL thickness, per 1 SD decrease
occurs in the superior, inferior, and temporal
Inner temporal 1.27 0.97e1.67
Inner superior 1.07 0.78e1.47 quadrants23,29 support these findings. In the present study,
Inner nasal 1.24 0.94e1.63 in the ETDRS grid, both GCC and full macular thicknesses
Inner inferior 1.20 0.91e1.58 were associated with the presence of dementia in the inner
Outer temporal 1.30 0.98e1.72 temporal sector. In the glaucoma analysis grid, GC-IPL,
Outer superior 0.97 0.69e1.35 GCC, and full macular thicknesses were all associated
Outer nasal 1.21 0.89e1.65 with the presence of dementia in the 3 inferior sectors,
Outer inferior 1.43 1.03e2.00
Glaucoma analysis grid
which was consistent with previous observations.
Macular/mRNFL thickness, per 1 SD decrease Meta-analysis30 of 25 studies (887 patients with AD, 216
Inner superotemporal 1.64 1.21e2.21 patients with MCI, and 864 healthy controls) showed that
Inner superonasal 1.43 1.09e1.88 ppRNFL thickness was decreased in those with AD and
Inner inferonasal 1.03 0.74e1.42 MCI compared with healthy controls, and similar results
Inner inferotemporal 1.44 1.08e1.93 were reported in another meta-analysis.7 In contrast, we
Outer superotemporal 1.33 0.96e1.85
Outer superonasal 1.31 1.00e1.71
observed no association between dementia or MCI and
Outer inferonasal 1.20 0.89e1.61 ppRNFL thickness or between MCI and thickness of any
Outer inferotemporal 1.45 1.05e2.01 part of the retina. We do not have any clear answer for
Superior 1.51 1.14e2.01 this discrepancy between studies; however, it may be
Inferior 1.33 0.99e1.79 because participants with dementia in our study may have
Macular/GC-IPL thickness, per 1 SD decrease had a more modest degree of cognitive impairment than
Inner superotemporal 1.30 0.99e1.71
those in previous case-control studies, and MCI may
Inner superonasal 1.09 0.80e1.49
Inner inferonasal 1.32 1.02e1.70 include wide range of spectrums.
Inner inferotemporal 1.45 1.14e1.84 Ab accumulation in the retina has also been observed in
Outer superotemporal 1.18 0.86e1.61 animal models of glaucoma and in postmortem studies of
Outer superonasal 1.09 0.78e1.51 patients with glaucoma.31,32 Many studies have reported that
Outer inferonasal 1.58 1.16e2.17 patients with AD have a higher risk of glaucoma and vice
Outer inferotemporal 1.39 1.02e1.90 versa;33-37 however, a meta-analysis showed an inverse as-
Superior 1.13 0.82e1.56
Inferior 1.60 1.19e2.16
sociation between AD and glaucoma.38 In the present study,
there was no association of dementia with signs or a history
of glaucoma. We could not draw a conclusion on this issue
Significant values appear in boldface.
CI ¼ confidence interval; ETDRS ¼ Early Treatment Diabetic Retinop-
because the visual field test was not performed. Given that
athy Study; GC-IPL ¼ ganglion cell-inner plexiform layer; glaucoma may not have a solitary pathogenesis but may
mRNFL ¼ macular retinal nerve fiber layer; OR ¼ odds ratio; share an underlying pathogenesis with AD, it might be
SD ¼ standard deviation. difficult to elucidate the association between these
2 diseases. Further studies are needed to clarify the
potential relationship between these 2 diseases.
ppRNFL thickness. Other studies showed the relationship
between the thickness of specific layers or full macular Study Strengths and Limitations
thickness and the presence of AD.25-27 In the present study,
the presence of dementia was associated with full macular The strengths of this study include use of detailed mental
thickness (all layers) and GCC thickness but not with health checks conducted by psychiatrists in a large sample in
ppRNFL thickness. Moreover, mRNFL thickness was a field. The OCT measurements were assessed using specific
associated with the presence of dementia in fewer sectors layers such as mRNFL, GC-IPL, and GCC, as well as full
than was GC-IPL thickness. In the glaucoma analysis grid, macular thickness, in 2 grids and ppRNFL. Also, we
GC-IPL, GCC, and full macular thicknesses were all asso- analyzed the associations between retinal thickness and the
ciated with the presence of dementia in the 3 inferior sectors, presence of dementia after excluding participants with fea-
which suggests that GC-IPL loss, in part, would cause tures suggestive of glaucoma to remove the influence of

115
 Ophthalmology Volume 127, Number 1, January 2020

Table 5. Multivariable Analysis of Associations of Macular Retinal Nerve Fiber Layer and Ganglion Cell-Inner Plexiform Layer Thickness
with the Presence of Dementia

OR, per
1 SD Decrease Tertile 1 Tertile 2 Tertile 3
ETDRS grid
Macular/mRNFL thickness
Inner superior
No. of noncases/cases 181/5 177/12 170/16
Thickness, mm 29.5 (28.1e31.7) 25.1 (23.5e26.2) 18.9 (14.7e20.9)
Model 1 1.60 (1.16 2.19) 0.39 (0.13e1.14) Reference 1.18 (0.53e2.63)
Model 2 1.57 (1.13e2.20) 0.38 (0.13e1.15) Reference 1.14 (0.50e2.60)
Macular/GC-IPL thickness
Outer inferior
No. of noncases/cases 180/6 176/11 171/15
Thickness, mm 68.8 (66.2e73.2) 60.0 (58.5e61.8) 53.2 (49.9e55.2)
Model 1 1.40 (0.97e2.01) 0.58 (0.21e1.62) Reference 1.40 (0.61e3.19)
Model 2 1.53 (1.04e2.25) 0.54 (0.19e1.54) Reference 1.49 (0.64e3.47)
Glaucoma analysis grid
Macular/mRNFL thickness
Inner superotemporal
No. of noncases/cases 182/5 176/12 170/16
Thickness, mm 18.3 (15.5e20.3) 24.6 (23.3e25.8) 29.0 (27.9e30.9)
Model 1 1.38 (1.01e1.90) 0.34 (0.11e0.99) Reference 1.09 (0.49e2.43)
Model 2 1.39 (1.03e1.94) 0.28 (0.09e0.87) Reference 1.05 (0.46e2.40)
Superior
No. of noncases/cases 179/7 175/11 169/15
Thickness, mm 41.4 (40.0e42.7) 36.7 (35.4e37.7) 30.6 (26.6e32.7)
Model 1 1.40 (1.02e1.93) 0.72 (0.27e1.93) Reference 1.17 (0.51e2.69)
Model 2 1.40 (1.02e1.94) 0.61 (0.22e1.69) Reference 1.05 (0.45e2.43)
Macular/GC-IPL thickness
Inner inferonasal
No. of noncases/cases 181/4 180/10 167/19
Thickness, mm 88.5 (86.4e91.3) 81.0 (79.6e82.9) 73.2 (69.1e75.2)
Model 1 1.28 (0.97e1.69) 0.48 (0.14e1.57) Reference 1.86 (0.83e4.19)
Model 2 1.38 (1.02e1.88) 0.48 (0.14e1.61) Reference 1.95 (0.85e4.51)
Inner inferotemporal
No. of noncases/cases 181/7 176/10 171/16
Thickness, mm 89.1 (86.3e91.9) 81.0 (79.3e82.7) 73.4 (68.5e75.8)
Model 1 1.44 (1.11e1.87) 0.71 (0.26e1.93) Reference 1.23 (0.53e2.86)
Model 2 1.48 (1.11e1.97) 0.73 (0.26e2.03) Reference 1.28 (0.54e3.07)
Outer inferonasal
No. of noncases/cases 179/8 178/9 171/16
Thickness, mm 60.1 (57.3e64.3) 52.2 (50.6e53.7) 44.6 (41.1e47.4)
Model 1 1.42 (0.99e2.02) 0.91 (0.34e2.44) Reference 1.52 (0.64e3.61)
Model 2 1.48 (1.03e2.13) 0.90 (0.33e2.48) Reference 1.69 (0.70e4.10)
Outer inferotemporal
No. of noncases/cases 179/7 176/13 173/13
Thickness, mm 53.5 (51.7e56.9) 46.9 (45.4e48.3) 40.3 (37.2e42.3)
Model 1 1.50 (1.08e2.08) 0.48 (0.18e1.26) Reference 0.99 (0.44e2.25)
Model 2 1.50 (1.07e2.09) 0.48 (0.18e1.28) Reference 1.14 (0.49 2.67)
Inferior
No. of noncases/cases 179/8 178/9 171/16
Thickness, mm 63.2 (61.3e66.0) 56.9 (55.5e58.3) 50.6 (47.6e52.4)
Model 1 1.56 (1.12e2.15) 0.95 (0.35e2.55) Reference 1.61 (0.68e3.82)
Model 2 1.62 (1.15e2.28) 0.91 (0.33e2.50) Reference 1.78 (0.73e4.31)

Significant values appear in boldface.

ETDRS ¼ Early Treatment Diabetic Retinopathy Study; GC-IPL ¼ ganglion cell-inner plexiform layer; mRNFL ¼ macular retinal nerve fiber layer;
OR ¼ odds ratio; SD ¼ standard deviation.
Model 1: Adjusted for age and sex. Model 2: Adjusted for age, sex, refraction, and educational status.

glaucoma. However, we recognize several limitations with types of dementia, which may have limited our ability to
our study. First, we used a cross-sectional observational detect an association between the cognitive function and the
design, which does not provide temporal information examined variables, such as retinal thickness. Specifically,
regarding associations. Second, we did not determine the the dementia group included individuals with AD together

116
 Ito et al 
Assessment of Cognitive Function Using OCT

with vascular dementia and dementia with Lewy bodies. We 9. Tan O, Chopra V, Lu AT, et al. Detection of macular gan-
reassessed the data after excluding participants with features glion cell loss in glaucoma by Fourier-domain optical
suggestive of vascular dementia; however, the results were coherence tomography. Ophthalmology. 2009;116:
not markedly changed. Finally, in the present study, the 2305e2314.e1-2.
participant rate was low and the number of participants with 10. Khawaja AP, Chan MP, Yip JL, et al. Retinal nerve fiber layer
measures and cognitive function in the EPIC-Norfolk Cohort
dementia was small, so we could not exclude the possibility Study. Invest Ophthalmol Vis Sci. 2016;57:1921e1926.
of a chance finding. Therefore, future studies to validate 11. Tsugane S, Sawada N. The JPHC study: design and some
these findings are needed in longitudinal large cohort studies findings on the typical Japanese diet. Jpn J Clin Oncol.
with advanced OCT. 2014;44:777e782.
In conclusion, we analyzed a Japanese cohort aged more 12. Cruz-Herranz A, Balk LJ, Oberwahrenbrock T, et al. The
than 65 years by using detailed mental health examinations APOSTEL recommendations for reporting quantitative optical
and OCT. Full macular thickness and GCC thickness were coherence tomography studies. Neurology. 2016;86:
associated with the presence of dementia, but ppRNFL 2303e2309.
thickness was not. Furthermore, GC-IPL, GCC, and full 13. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A
macular thicknesses were all associated with the presence of practical method for grading the cognitive state of patients for
the clinician. J Psychiatr Res. 1975;12:189e198.
dementia in the inferior sectors, which suggests that GC-IPL 14. Elwood RW. The Wechsler Memory Scale-Revised: psycho-
loss, in part, would cause reduction of GCC and full macular metric characteristics and clinical application. Neuropsychol
thickness due to neurodegeneration caused by dementia. Rev. 1991;2:179e201.
Although this method has moderate detecting power for 15. Agrell B, Dehlin O. The clock-drawing test. 1998. Age Ageing.
dementia and further studies are warranted to use in the 2012;41(Suppl 3):iii41eiii45.
clinical setting, OCT measurements of the macula could be 16. Hughes CP, Berg L, Danziger WL, et al. A new clinical scale for
superior to those of the ppRNFL in assessing neurodegen- the staging of dementia. Br J Psychiatry. 1982;140:566e572.
erative changes and a potentially useful diagnostic 17. Radloff LS. The CES-D Scale: a self-report depression scale
biomarker of cognitive function. for research in the general population. Appl Psychol Meas.
1977;1:385e401.
Acknowledgments 18. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a
brief depression severity measure. J Gen Intern Med. 2001;16:
The authors acknowledge Yasuhiro Makihara and Yukiko Miya- 606e613.
saka, other psychiatrists, and medical staff for technical assistance, 19. Fujishima M, Kawaguchi A, Maikusa N, et al. Sample size
and thank all staff members in the Saku area for extensive efforts to estimation for Alzheimer’s disease trials from Japanese ADNI
conduct the survey. Serial Magnetic Resonance Imaging. J Alzheimers Dis.
2017;56:75e88.
20. Iwatsubo T. Japanese Alzheimer’s Disease Neuroimaging
References Initiative: present status and future. Alzheimers Dement.
2010;6: 297-269.
21. Petersen RC, Smith GE, Waring SC, et al. Mild cognitive
1. Reitz C, Brayne C, Mayeux R. Epidemiology of Alzheimer impairment: clinical characterization and outcome. Arch
disease. Nat Rev Neurol. 2011;7:137e152. Neurol. 1999;56:303e308.
2. Gupta VK, Chitranshi N, Gupta VB, et al. Amyloid beta accu- 22. American Psychiatric Association. Diagnostic and Statistical
mulation and inner retinal degenerative changes in Alzheimer’s Manual of Mental Disorders. 4th ed. Washington, DC:
disease transgenic mouse. Neurosci Lett. 2016;623:52e56. American Psychiatric Association; 2000.
3. Doustar J, Torbati T, Black KL, et al. Optical coherence to- 23. Blanks JC, Torigoe Y, Hinton DR, Blanks RH. Retinal pa-
mography in Alzheimer’s disease and other neurodegenerative thology in Alzheimer’s disease. I. Ganglion cell loss in foveal/
diseases. Front Neurol. 2017;8:701. parafoveal retina. Neurobiol Aging. 1996;17:377e384.
4. Cheung CY, Ikram MK, Chen C, Wong TY. Imaging retina to 24. Cheung CY, Ong YT, Hilal S, et al. Retinal ganglion cell
study dementia and stroke. Prog Retin Eye Res. 2017;57:89e107. analysis using high-definition optical coherence tomography in
5. Parisi V, Restuccia R, Fattapposta F, et al. Morphological and patients with mild cognitive impairment and Alzheimer’s
functional retinal impairment in Alzheimer’s disease patients. disease. J Alzheimers Dis. 2015;45:45e56.
Clin Neurophysiol. 2001;112:1860e1867. 25. Bayhan HA, Aslan Bayhan S, Celikbilek A, et al. Evaluation
6. Iseri PK, Altinas O, Tokay T, Yuksel N. Relationship between of the chorioretinal thickness changes in Alzheimer’s disease
cognitive impairment and retinal morphological and visual using spectral-domain optical coherence tomography. Clin Exp
functional abnormalities in Alzheimer disease. Ophthalmol. 2015;43:145e151.
J Neuroophthalmol. 2006;26:18e24. 26. Marziani E, Pomati S, Ramolfo P, et al. Evaluation of retinal
7. Thomson KL, Yeo JM, Waddell B, et al. A systematic review nerve fiber layer and ganglion cell layer thickness in Alz-
and meta-analysis of retinal nerve fiber layer change in de- heimer’s disease using spectral-domain optical coherence
mentia, using optical coherence tomography. Alzheimers tomography. Invest Ophthalmol Vis Sci. 2013;54:
Dement (Amst). 2015;1:136e143. 5953e5958.
8. Bock M, Brandt AU, Dorr J, et al. Time domain and spectral 27. Polo V, Garcia-Martin E, Bambo MP, et al. Reliability and
domain optical coherence tomography in multiple sclerosis: a validity of Cirrus and Spectralis optical coherence tomography
comparative cross-sectional study. Mult Scler. 2010;16: for detecting retinal atrophy in Alzheimer’s disease. Eye
893e896. (Lond). 2014;28:680e690.

117
 Ophthalmology Volume 127, Number 1, January 2020

28. Jones-Odeh E, Hammond CJ. How strong is the rela- 33. Helmer C, Malet F, Rougier MB, et al. Is there a link between
tionship between glaucoma, the retinal nerve fibre layer, open-angle glaucoma and dementia? The Three-City-Alienor
and neurodegenerative diseases such as Alzheimer’s dis- cohort. Ann Neurol. 2013;74:171e179.
ease and multiple sclerosis? Eye (Lond). 2015;29: 34. Tamura H, Kawakami H, Kanamoto T, et al. High frequency
1270e1284. of open-angle glaucoma in Japanese patients with Alzheimer’s
29. Blanks JC, Hinton DR, Sadun AA, Miller CA. Retinal gan- disease. J Neurol Sci. 2006;246:79e83.
glion cell degeneration in Alzheimer’s disease. Brain Res. 35. Tsai CS, Ritch R, Schwartz B, et al. Optic nerve head and
1989;501:364e372. nerve fiber layer in Alzheimer’s disease. Arch Ophthalmol.
30. den Haan J, Verbraak FD, Visser PJ, Bouwman FH. Retinal 1991;109:199e204.
thickness in Alzheimer’s disease: a systematic review and 36. Bayer AU, Ferrari F, Erb C. High occurrence rate of glaucoma
meta-analysis. Alzheimers Dement (Amst). 2017;6:162e170. among patients with Alzheimer’s disease. Eur Neurol.
31. Guo L, Salt TE, Luong V, et al. Targeting amyloid-beta in 2002;47:165e168.
glaucoma treatment. Proc Natl Acad Sci U S A. 2007;104: 37. Bayer AU, Ferrari F. Severe progression of glaucomatous
13444e13449. optic neuropathy in patients with Alzheimer’s disease. Eye
32. Gupta V, You Y, Li J, et al. BDNF impairment is associated (Lond). 2002;16:209e212.
with age-related changes in the inner retina and exacerbates 38. Tsilis AG, Tsilidis KK, Pelidou SH, Kitsos G. Systematic
experimental glaucoma. Biochim Biophys Acta. 2014;1842: review of the association between Alzheimer’s disease and
1567e1578. chronic glaucoma. Clin Ophthalmol. 2014;8:2095e2104.

Footnotes and Financial Disclosures

Originally received: January 22, 2019. HUMAN SUBJECTS: Human subjects were included in this study. The
Final revision: April 26, 2019. human ethics committees at the National Cancer Center Japan and Keio
Accepted: May 10, 2019. University School of Medicine approved the study. All research adhered to
Available online: June 3, 2019. Manuscript no. 2018-2874. the tenets of the Declaration of Helsinki. All participants provided informed
1
Department of Ophthalmology, Keio University, Tokyo, Japan. consent.
2
Department of Ophthalmology, Tachikawa Hospital, Tokyo, Japan. No animal subjects were used in this study.
3
Department of Ophthalmology, Tokyo Medical Center, Tokyo, Japan. Author Contributions:
4
Department of Neuropsychiatry, Keio University School of Medicine, Conception and design: Sasaki
Tokyo, Japan. Data collection: Ito, Sasaki, Takahashi, Nozaki, Matsuguma, Motomura,
5
Department of Vision Informatics (Topcon), Osaka University Graduate Ui, Shikimoto, Yuki, Sawada, Mimura, Tsubota, Tsugane
School of Medicine, Osaka, Japan. Analysis and interpretation: Sasaki, Mimura, Kawasaki, Tsubota
6
Epidemiology and Prevention Group, Center for Public Health Sciences, Obtained funding: Sasaki
National Cancer Center, Tokyo, Japan. Overall responsibility: Ito, Sasaki
*Y.I. and M.S. contributed equally to this work as first authors.
Abbreviations and Acronyms:
Financial Disclosure(s): AD ¼ Alzheimer’s disease; ETDRS ¼ Early Treatment Diabetic Reti-
The author(s) have made the following disclosure(s): M.M.: Grants and nopathy Study; GCC ¼ ganglion cell complex; GC-IPL ¼ ganglion cell-
personal fees e Daiichi Sankyo, Takeda Yakuhin, Tsumura; Personal inner plexiform layer; ILM ¼ internal limiting membrane; MCI ¼ mild
fees e Dainippon-Sumitomo Pharma, Eisai, Eli Lilly, Fuji Film RI Pharma, cognitive impairment; mRNFL ¼ macular retinal nerve fiber layer;
Janssen Pharmaceutical, Mochida Pharmaceutical, MSD, Nippon Chemi- NFLD ¼ nerve fiber layer defect; ppRNFL ¼ peripapillary retinal nerve
pher, Novartis Pharma, Ono Yakuhin, Otsuka Pharmaceutical, Pfizer, fiber layer; RGC ¼ retinal ganglion cell; SD ¼ spectral domain.
Yoshitomi Yakuhin; Grants e Pfizer, Shionogi, Tanabe Mitsubishi, outside
Correspondence:
the submitted work.
Mariko Sasaki, MD, PhD, Department of Ophthalmology, Keio University
Supported in part by a Grant-in-Aid for Scientific Research from the
School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582
Ministry of Education, Culture, Sports, Science, and Technology (MEXT),
Japan. E-mail: [email protected].
Japan (KAKENHI, JP26460778 [to M.S.]). The cohort study was originally
supported by the National Cancer Center Research and Development Fund.

118
 Commentary

Retinal Imaging in Alzheimer’s Disease: In Search of the Holy Grail

Amani A. Fawzi, MD - Chicago, Illinois
Sandra Weintraub, PhD - Chicago, Illinois
Waleed Fawzi, MRCPsych, MD - London, United Kingdom

Alzheimer’s disease (AD) is the most common cause of mild cognitive impairment (MCI), not dementia. This subtle
dementia, accounting for 60% to 80% of cases worldwide.1 difference also highlights why the field could benefit greatly
The pathologic hallmark of the disease is the accumulation from additional noninvasive and quantitative biomarkers
of extracellular beta amyloid plaques and intracellular tau when evaluating patients with cognitive impairment.
filaments, ultimately leading to neuronal death. The Macular pathologic involvement in AD was described first
current approved treatments slow down the progression of in histopathologic specimens of human donors by Blanks
the disease, but do not stop the neurodegeneration caused et al5 in 1989. These authors showed that in the macula, the
by it. The past 2 decades have witnessed repeated failures ganglion cells were affected in the temporal parafoveal
of clinical trials using disease-modifying therapies target- region.6 The ganglion cell loss extended to the entire retina,
ing amyloid.2 These failures have prompted a major particularly superior and inferior mid-peripheral regions.7
rethinking of the approach, with the general consensus These studies also showed significant histopathologic
being that potential treatments would work better if started evidence of loss and degeneration of the ganglion cell
at an earlier, possibly preclinical, stage of the disease.3 axons in the optic nerve in the AD eyes.5 With that in
With the redoubled efforts to diagnose AD at earlier stages mind, the field of retinal imaging has been in search of the
that could be amenable to preventive therapy, the field has best tools to reveal these pathologic changes. Studies have
been in search of biomarkers, clues that aid in diagnosing the explored retinal blood flow and OCT of the nerve and
disease early and help differentiate AD from other causes of macula with varying success in different populations.
dementia. These biomarkers would be most useful if they A lack of finding of significant changes in the peripapillary
were noninvasive, widely available, reliable, sensitive, and retinal nerve fiber layer (RNFL), representing the axons of the
specific to AD. Furthermore, the ideal biomarker would allow ganglion cells, in the current study population is intriguing on
clinicians to make the diagnosis during the prodromal stages many levels. Recently, using the United Kingdom Biobank
of AD, before the onset of debilitating dementia, as it is these database, which is a prospective multi-center, community-
earlier stages that are now the focus of therapeutic efforts. based study of United Kingdom residents 40 to 69 years of
However, diagnosing patients at these stages of the disease age at enrollment, Ko et al8 found that thinner peripapillary
remains difficult and costly. Thus, if retinal imaging can be RNFL was associated with worse cognitive performance at
validated as a biomarker to confirm the clinical suspicion of baseline. Further, in a smaller subset of the population that
AD, this would be a highly attractive potential tool. underwent repeat cognitive testing (1251 participants), the
In this population-based, cross-sectional study in the Saku authors found that those with the thinnest quintile of RNFL
region of Japan (see page 107), persons 65 to 86 years of age thickness were more likely to perform worse on repeat
underwent retinal OCT imaging and neuropsychological cognitive testing, suggesting a correlation between baseline
testing to explore the association between cognitive impair- RNFL and future cognitive decline, even after controlling for
ment and retinal thickness.4 The study revealed an inverse potential confounders. The discrepancy between the current
association between macular ganglion cell complex study and others that found a correlation between AD and
thickness and the presence dementia after adjusting for RNFL thickness may be explained by several factors. For
potential confounders including age and refractive error. example, the correlation between baseline cognitive function
Additionally, in the macular inferior sectors, several OCT and RNFL data in the United Kingdom Biobank dataset
measures were associated with the presence of dementia, demonstrated a nonlinear, but rather curvilinear, correlation
defined as having a score of less than 23 on the Mini- with the lowest quintiles of RNFL (<45 mm and 45e50 mm)
Mental State Exam. The authors included psychiatric showing much stronger correlation than the higher quintiles.
assessment to exclude participants with depression and also This finding could explain in part the lack of association in
attempted to exclude participants with vascular dementia the current study, in which the authors considered the entire
based on their history of stroke. Importantly, they did not range of RNFL, rather than examining the different quartiles
report the functional impact of cognitive status on partici- of thickness. Another possibility is the different inclusion
pants’ activities of daily living. In clinical practice, a score of criteria, especially glaucoma, which was a definite exclusion
23 on the Mini-Mental State Exam alone does not establish in the United Kingdom Biobank study design, but not in the
the diagnosis of dementia because the same score can be current study. Another difference relates to the different
obtained by an individual with normal activities of daily cognitive tests used in these studies. The Mini-Mental State
living who then would belong in the diagnostic category of Exam used to categorize participants in the current study has

ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.07.025 119

Published by Elsevier Inc. ISSN 0161-6420/19
 Ophthalmology Volume 127, Number 1, January 2020

been used widely, but suffers from a lack of discrimination for both on positron emission tomography scans. Another
the milder range of cognitive impairment and may not be important area of investigation would be whether these retinal
sufficient to distinguish control participants from those with findings progressively worsen as an individual patient’s
MCI.9 In contrast, the United Kingdom Biobank study used a cognitive impairment and brain neuronal loss advance, which
set of psychological tests that were specifically more sensitive would be critical validation of the potential for the retina to
to these milder stages of impairments. Another difference serve as a surrogate marker for AD pathologic progression,
relates to the age of the populations, with the Asian/Japanese supplementing information from other biomarkers.
study having a distinctly older population, as well as a clear Overall, this is an exciting time for the field of retinal
difference in the racial background of the 2 populations, imaging in the arena of cognitive disorders of aging.
with the United Kingdom Biobank study being Accumulating data show that the retina could provide
overwhelmingly white, whereas the population in the current valuable insights into the diagnosis of AD and cognitive
was entirely Asian/Japanese. It is a possibility that different impairment. Pending the next level of rigorous multi-center
populations can differ in their retinal manifestations of AD, studies, it is fair to say that retinal imaging one day could
although we are not aware of any research to support this. serve as a potential biomarker in AD studies.
Another salient negative finding in the current study is
the lack of association of any of the OCT metrics with the
MCI category, an earlier stage of cognitive dysfunction
considered to be a precursor to dementia resulting from AD
References
and representing an important population for studies aim-
ing for earlier treatment to prevent the onset of dementia. 1. Alzheimer’s Association. 2018 Alzheimer’s disease facts and
As the authors stated, MCI encompasses a loosely defined figures. Alzheimers Dement. 2018;14(3):367e429.
clinical spectrum, with a range of cognitive deficits and a 2. Cummings JL, Morstorf T, Zhong K. Alzheimer’s disease
wide range of susceptibility to progression to AD. drug-development pipeline: few candidates, frequent failures.
Although it would be highly desirable to find an ocular Alzheimers Res Ther. 2014;6(4):1e7.
3. Khan TK. An algorithm for preclinical diagnosis of Alz-
parameter that correlates with this earlier, prodromal state,
heimer’s disease. Front Neurosci. 2018;12:275.
the current study was not designed with that goal in mind. 4. Ito Y, Sasaki M, Takahashi H, et al. Quantitative assessment of
Within the MCI clinical spectrum, amnestic MCI is a the retina using OCT and associations with cognitive function.
subtype that is most likely to be caused by AD. Thus, 50% Ophthalmology. 2019;127:107e118.
of these individuals progress to dementia resulting from 5. Blanks JC, Hinton DR, Sadun AA, Miller CA. Retinal gan-
AD within 30 months.10 A recent meta-analysis showed glion cell degeneration in Alzheimer’s disease. Brain Res.
the potential of the peripapillary RNFL to be affected 1989;501(2):364e372.
significantly in people with MCI when considering the 6. Blanks JC, Torigoe Y, Hinton DR, Blanks RH. Retinal pa-
total available evidence in 6 rigorously performed studies thology in Alzheimer’s disease. I. Ganglion cell loss in foveal/
with a total of 161 MCI patients from 5 different pop- parafoveal retina. Neurobiol Aging. 1996;17(3):377e384.
7. Blanks JC, Schmidt SY, Torigoe Y, et al. Retinal pathology in
ulations.11 More recently, OCT angiography has revealed
Alzheimer’s disease. II. Regional neuron loss and glial
that macular capillary density is decreased significantly in changes in GCL. Neurobiol Aging. 1996;17(3):385e395.
patients with amnestic MCI and early AD, providing 8. Ko F, Muthy ZA, Gallacher J, et al. Association of retinal
further evidence that the retina may be affected in well- nerve fiber layer thinning with current and future cognitive
defined populations with prodromal stages of AD.12 decline: a study using optical coherence tomography. JAMA
Confirming the diagnosis of AD relies on clinical sus- Neurol. 2018;75(10):1198e1205.
picion, coupled with confirmatory imaging and fluid 9. Mitchell AJ. A meta-analysis of the accuracy of the mini-mental
biomarker evidence, with AD diagnosis definitively state examination in the detection of dementia and mild cogni-
confirmed only at brain autopsy. Cerebrospinal fluid bio- tive impairment. J Psychiatr Res. 2009;43(4):411e431.
markers and amyloid positron emission tomography scans 10. Fischer P, Jungwirth S, Zehetmayer S, et al. Conversion from
subtypes of mild cognitive impairment to Alzheimer dementia.
remain the most widely studied and validated, although they
Neurology. 2007;68(4):288e291.
suffer from being relatively invasive and expensive, 11. Knoll B, Simonett J, Volpe NJ, et al. Retinal nerve fiber layer
respectively, which has restricted their wide clinical imple- thickness in amnestic mild cognitive impairment: case-control
mentation.13,14 Compared with these approaches, retinal study and meta-analysis. Alzheimers Dement (Amst). 2016;4:
OCT imaging presents a far less invasive, less costly, and 85e93.
higher resolution method that is more widely accessible 12. Zhang YS, Zhou N, Knoll BM, et al. Parafoveal vessel loss
globally. Therefore, an important need exists to investigate and correlation between peripapillary vessel density and
whether retinal OCT imaging parameters can be informative cognitive performance in amnestic mild cognitive impairment
as potential biomarkers of AD in the clinical arena. The AD and early Alzheimer’s disease on optical coherence tomogra-
field is in dire need of noninvasive biomarkers that would phy angiography. PLoS One. 2019;14(4):e0214685.
13. Olsson B, Lautner R, Andreasson U, et al. CSF and blood
aid in confirming or disproving the diagnosis of AD, when it
biomarkers for the diagnosis of Alzheimer’s disease: a sys-
is clinically suspected and before the onset of dementia. Future tematic review and meta-analysis. Lancet Neurol. 2016;15(7):
large-scale, multi-center, and multi-ethnic studies need to 673e684.
correlate the retinal (macular and optic nerve) parameters with 14. Henriques AD, Benedet AL, Camargos EF, et al. Fluid and
other, more pathologically relevant, brain biomarkers such as imaging biomarkers for Alzheimer’s disease: where we stand
cerebrospinal fluid amyloid and tau load, amyloid load, or and where to head to. Exp Gerontol. 2018;107:169e177.

120
 Commentary

Footnotes and Financial Disclosures

Financial Disclosure(s): The author(s) have no proprietary or commercial Correspondence:
interest in any materials discussed in this article. Amani A. Fawzi, MD, Department of Ophthalmology, Feinberg School of
Medicine, Northwestern University, 645 North Michigan Avenue, Chicago,
IL 60611. E-mail: [email protected].

Pictures & Perspectives

Pembrolizumab-Related Enophthalmos
Pembrolizumab is a highly selective immune checkpoint inhibitor that targets the programmed cell death protein (PD-1) receptor on
lymphocytes and has been approved for the treatment of malignant melanoma. A 59-year-old woman had been treated for 1 year with
pembrolizumab for melanoma metastatic to lung, spleen, and central nervous system. She noted changes of her facial appearance with
temporal wasting and sunken eyes while denying any decrease in body weight. Exophthalmometry was 11 mm for the right eye (OD) and
10 mm for the left eye (OS). (Fig A, B) Axial magnetic resonance imaging (MRI) scans performed 6 months apart (Fig C, D) demonstrated
a significant change in globe position with 4.7 mm recession on the right and 4.1 mm on the left. There were no signs of orbital metastasis
or inflammation. She had no other medical or iatrogenic factors to explain loss of orbital volume and enophthalmos. (Magnified version of
Fig A-D is available online at www.aaojournal.org).
ALEXANDRA MANTA, MD
ROBERT A. GOLDBERG, MD
Division of Orbital and Ophthalmic Plastic Surgery, Jules Stein Eye Institute, University of California, Los Angeles, California

121
Secondary Prevention of Retinoblastoma
Revisited
Laser Photocoagulation of Invisible New Retinoblastoma
Sameh E. Soliman, MD,1,2 Cynthia VandenHoven, BAA, CRA,1 Leslie D. MacKeen, BSc,1
Brenda L. Gallie, MD, FRCSC1,3,4,5

Purpose: Invisible retinoblastoma tumors are now detected with screening for retinal tumors in at-risk
neonates (those inheriting RB1 pathogenic alleles from affected parents) using handheld OCT. Laser
photocoagulation is challenging, requiring exact localization of a tumor invisible to indirect ophthalmoscopy and
standard imaging. We describe OCT-guided localization and photocoagulation of these invisible tumors with
1-year follow-up.
Design: Retrospective, noncomparative, single-institutional, observational case series.
Participants: Children with any clinically invisible retinoblastoma tumor that was detected on OCT posterior
pole screening.
Methods: OCT revealed round homogeneous invisible tumors within the inner nuclear layer. Software cali-
pers placed beside anatomic retinal landmarks (branched/curved vessels, fovea, or optic disc) mapped the tumor
location and extent. A single laser (532 nm) burn flagged the location, and OCT evaluated the tumorelaser burn
relationship; laser treatment was then continued in the correct location. Post-laser OCT ensured complete
treatment.
Main Outcome Measures: Accuracy (frequency of geographic miss and skip areas), effectiveness (recur-
rence rate), and burden (scar size and characteristics at final follow-up) of laser treatment.
Results: Eleven new invisible posterior pole tumors in 7 eyes of 5 children were treated by this technique.
Localization and tumorelaser burn relationships were accurate in 11 of 11 tumors (100%, 95% confidence in-
terval [CI], 49.9e100), and all showed swelling and hyper-reflectiveness of the tumor in post-laser OCT. Two
photocoagulation sessions (2 weeks apart) were sufficient to successfully manage 9 of 11 tumors (82%, 95% CI,
37.4e100) with resulting permanent flat scars. One tumor (9%, 95% CI, 0.2e50.6) developed OCT-detected
subclinical recurrences within 3 months, treated by 1 laser session. No treatment scar showed gliosis, foveal
involvement, or retinal traction at 1-year follow-up. Scar expansion occurred in 1 tumor (9%, 95% CI, 0.2e50.6),
and all scars (100%, 95% CI, 49.9e100) showed pigmentary changes.
Conclusions: The OCT-guided localization and photocoagulation technique is valuable in achieving preci-
sion results in managing invisible new retinoblastoma tumors. This technique shows a potential to improve
outcomes of secondary prevention screening for retinoblastoma. Ophthalmology 2020;127:122-127 ª 2019 by
the American Academy of Ophthalmology

Supplemental material available at www.aaojournal.org.

Cancer progression can be prevented at 3 levels: primary Clinically, small tumors can be detected by performing
(prevent exposure to carcinogens), secondary (early diag- wide fundus indirect ophthalmoscopy or imaging. OCT has
nosis and treatment), or tertiary (prevent tumor invasive- improved visualization of retinal layers rendering sub-
ness/spread and early rehabilitation).1 Retinoblastoma, the millimeter or subclinical invisible tumor detection
most common pediatric intraocular malignancy, is usually possible.9-12 We have previously described an OCT-guided
managed at the tertiary level to prevent local/systemic posterior pole screening technique to identify potential
tumor spread and loss of an eye or vision.2 Screening invisible tumors.12
offspring of retinoblastoma survivors is a step toward The concept of technology-enhanced procedure guidance
secondary prevention if tumors are detected early and is becoming prevalent in medicine, such as sonographic-
precision therapy achieves good visual outcomes with and radiologic-guided tissue biopsies, injections, and
minimal scarring.2-7 At birth, retinoblastoma tumors endovascular procedures.13 In ophthalmology, OCT was
threaten the macula,8 and laser scars can contribute to visual reported to guide complicated procedures such as
outcomes. intracameral air injection for iatrogenic Descemet’s

122 ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.08.011

Published by Elsevier Inc. ISSN 0161-6420/19
 Soliman et al 
Laser of Invisible Retinoblastoma

membrane detachment,14 complete excision of an ms] between shots) were used to flag the center of the rectangular
intracorneal epithelial cyst,15 and intraoperative OCT- area (i.e., the invisible tumor) followed by an OCT scan to ensure
guided vitrectomy.16 Our group reported handheld OCT to precision by validating laser location relative to tumor before
guide laser therapy of perifoveal tumors.17 further laser shots were applied, thus preventing inadvertent retinal
damage. Accuracy of localization was assessed by visible retinal
Laser treatment to an invisible tumor is challenging.
whitening by laser and immediate post-laser follow-up OCT scan.
Laser delivery systems (operating microscope or indirect In larger lesions, the post-laser OCT scan can identify areas of
ophthalmoscopy) depend on direct visualization to treat the skipped treatment so that further laser can be applied. A final post-
tumor.4 We developed an OCT-based protocol to guide laser OCT ensured complete tumor treatment with laser effect
localization and laser photocoagulation of newly discovered extending beyond the tumor edge into at least 1 previously
invisible tumors as a step further toward secondary uninvolved OCT frame on each side of the tumor.
prevention of retinoblastoma.
Technique Evaluation
Methods Accuracy of laser treatment based on OCT was evaluated by
assessing frequency of geographic miss and skip areas in initial
Study Design post-laser OCT. Effectiveness was evaluated by tumor recurrence
This study is a retrospective noncomparative interventional case rate. Burden was evaluated by final scar size (in relation to the
series of children treated in the Hospital for Sick Children (Sick- initial tumor size) and characteristics, including migration (defined
Kids), Toronto, Canada. The study complies with the Declaration of as expansion of resultant scar >25% over time), traction (defined
Helsinki and was approved by the institutional research ethics board. as tangential pulling of the scar on the adjacent retinal structures as
optic nerve, fovea, or blood vessels), gliosis (defined as superficial
white fibrous proliferation resembling tumor recurrence and veri-
Eligibility fied to be scar by OCT), and pigmentary changes. The 95% con-
Any retinoblastoma tumor was included that was (1) clinically fidence interval (CI) was calculated using the exact Poisson
invisible, (2) detected on OCT posterior pole screening as a round method. However, because the sample does not constitute inde-
homogenous tumor, and (3) treated with OCT-guided localization pendent events, the true CIs would be even wider taking lack of
and photocoagulation. Data collected were age at diagnosis, independence into account.
treatment of tumor, genetic status, eye staging (International
Intraocular Retinoblastoma Classification18 and the Tumor Node
Metastasis and Heritability staging19), OCT assessment of tumor Results
characters (shape, size, location, and associated signs), laser
treatment sessions, tumor recurrence, and treatment scar follow-up. OCT-guided laser was used to manage 11 invisible new tumors in
7 eyes of 5 children with retinoblastoma. Six eyes (4 children) had
Technique Description tumors discovered by OCT screening at examination under anes-
thesia (EUA) of an eye clinically free of tumor in a retinoblastoma-
The OCT software displays an infrared or sum voxel image pro- prone child (H1)19 with a family history of retinoblastoma (2
jection fundus image of the scanned volume cube with a line children had prenatal and 2 had postnatal RB1 pathogenic allele
representing each OCT frame. Software calipers (Fig 1) placed detection). Early-term delivery at 36/37 weeks was advised for
vertically on the OCT frame identify the axial point location on the 2 children (3 eyes) with prenatal diagnosis, whereas the
the associated fundus image. Horizontally placed calipers define remaining children (3 eyes) had normal full-term (38/39 weeks)
the lateral extent of tumor but are not quantitative, because the delivery. One eye had a new tumor several months after control of
lateral extent is affected by the axial length and refraction. its primary tumors by chemoreduction and focal consolidation.
Multiple different colored calipers can be used on the same Two tumors (Table 1, tumors 2 and 10) were missed in an earlier
image to specify the affected area.20 OCT and discovered, in retrospect at 45 (Fig 2) and 21 days, to
When an invisible tumor is detected by OCT screening under have been present in the earlier OCT scans. Eight tumors
general anesthesia, calipers (Fig 1) were used to identify the tumor showed rounded intraretinal homogenous appearance on OCT; 3
extension in each frame that showed the tumor in both vertical and were ovoid with a fishtail21 configuration. Initial vertical tumor
horizontal scans (12
12-mm cube). The invisible tumor location height showed wide variation in different scans of the same
and extent were mapped over the fundus image as a rectangular tumor because of OCT-scan inclination and vertical caliper sub-
area. The upper and lower boundaries of the rectangles were 1 jective placement. Two tumors were within 2 disc diameters of the
frame above and below the horizontal OCT frame(s) containing the foveal pit and considered visually threatening (Table 1, tumors 1
invisible tumors. The lateral boundaries were identified similarly in and 4). All other tumors were outside the vascular arcades.
vertical scans. If a vertical scan was insufficient or difficult to All tumors (11/11 tumors, 100%, 95% CI, 49.9e100) were
acquire, 2 vertical localizing calipers just outside the tumor accurately treated by laser without geographic misses (initial
boundaries were placed in OCT frame(s) showing tumor and tumor-laser burn relation) or skipped areas verified by post-laser
mapped to the fundus image. Once the suspicious area was iden- OCT. All tumors (11/11 tumors, 100%, 95% CI, 49.9e100)
tified, nearby anatomic characteristics such as vessel curving and showed post-laser signs on OCT, including tumor swelling, hyper-
branching, fovea, or optic disc were noted and plotted on reflectiveness, and shadowing extending to the retinal pigment
corresponding color fundus images. epithelium. All tumors (11/11 tumors, 100%, 95% CI, 49.9e100)
With the use of indirect ophthalmoscopy, the rectangular area were controlled with a median of 2 laser sessions (range, 1e3
and its surrounding anatomic landmarks are identified with the sessions) with a mean number of 24 laser spots (range, 8e38
upright and lateral image inversion in perspective. Shorter wave- spots). One tumor (9%, 95% CI, 0.2e50.6) (Table 1, tumor 1)
length laser was used (532 nm) to avoid deep laser burns and showed a subclinical recurrence within 84 days detected by OCT
provide precise localization because of its small spot size. Few (Fig 2) and treated by a single laser session. We estimated the
laser shots (low-power, 0.5-second duration, longer interval [500 technique effectiveness at 91% (10/11 tumors, 95% CI,

123
 Ophthalmology Volume 127, Number 1, January 2020

Figure 1. OCT-guided localization and photocoagulation of invisible tumor. Horizontal caliper identified the lateral extent of the tumor. Vertical caliper
identified point localization of the tumor. Point and lateral locations were identified in each of the pre-laser OCT frames that showed the tumor. Rectangular
area (blue star) was plotted over pre-laser fundus images considering nearby vascular bifurcations (yellow stars) and upper and lower limits (green lines)
corresponding to 1 frame below and above last limits (yellow box) with detected tumor. Post-laser fundus images show retinal whitening in response to
treatment. Post-laser OCT frames ensured complete tumor treatment.

43.6e100). One child (Table 1, child 4) received 1 cycle of the initial scar dimensions. One scar (9%, 95% CI, 0.2e50.6)
systemic chemotherapy for a suspicious growth22 in the other (Table 1, tumor 7) showed approximately 50% expansion and was
eye 4 months after laser therapy. considered to be true scar migration, although the final scar size
At 3 months post-laser completion, all tumor scars showed was approximately 1 disc diameter. No scar showed gliosis or
proportionate expansion in all directions that was less than 20% of traction (0/11, 0%, 95% CI, 0e33.5). All scars (11/11 tumors,

Table 1. Characteristics of Invisible Tumors at Diagnosis and Tumor Scars at Final Follow-up

Tumor at Diagnosis Tumor Scar Prognosis

Tumor Age (days) Age
Eye No. Screened (days) IIRC cT Shape Laser Rx Recurrence Migration Traction Gliosis Pigmentation
1 1 3 15 A* cT1a rounded 2 Y (84 d) N N N Y
2 3 99^ A rounded 2 N N N N Y
3 3 219 A rounded 2 N N N N Y
2 4 3 15 B* cT1b rounded 1 N N N N Y
3 5 632 706 D cT2a rounded 2 N N N N Y
4 6 89 92 A cT1a rounded 2 N N N N Y
7 89 92 rounded 1 N Y N N Y
5 8 89 92 B cT1b ovoid, fish-tail 1 N N N N Y
9 89 92 rounded 1 N N N N Y
6 10 1 118^ A cT1a ovoid, fish-tail 3 N N N N Y
7 11 147 299 A cT1a ovoid, fish-tail 3 N N N N Y

cT ¼ clinical tumor staging using 8th ed. TNM staging; IIRC ¼ International Intraocular Retinoblastoma Classification; N ¼ no; Y ¼ yes; ^ ¼ confirmed
diagnosis.
*Within 2 disc diameters of the foveal pit.

124
 Soliman et al 
Laser of Invisible Retinoblastoma

Figure 2. OCT detection of subclinical tumors and scar recurrence. Upper row (yellow box): A small rounded invisible tumor (*) was missed in initial OCT
scanning and diagnosed 45 days later when it attained a larger size. Laser was applied using the described technique and verified by post-laser OCT that
showed tumor swelling, hyper-reflectiveness, and shadowing. A flat scar can be seen within 2 months verified by OCT. Lower row (green box): A small
tumor scar (*) within 2 disc diameters of the foveal center (yellow arrowhead) showed initial stability after treatment with the described laser technique
followed by subclinical subtle recurrence 84 days later detected on OCT. It was treated with a single laser session verified by post-laser OCT. The scar is flat
with pigmentary changes 1 year post-laser.

100%, 95% CI, 49.9e100) showed pigmentary changes. However, examining the newborn at risk within the first few days
8 of 11 scars (73%, 95% CI, 31.4e100) showed pigmentation after birth followed by meticulous follow-up schedule to
involving >50% of the tumor scar at 1-year follow-up. detect tumors at the smallest possible size and stage.
Detection of the proband’s RB1 variant in the child confirms
the H1 status whether performed postnatal through cord or
Discussion peripheral blood or prenatal via amniocentesis.
Conventionally, screening tumors via indirect ophthal-
Retinoblastoma as the prototype of genetic cancers23 now moscopy or wide fundus imaging was performed. Suspi-
has heritability designation in the 8th edition American cious areas (minimally elevated area adjacent to vascular
Joint Committee TNM staging.19 H1 designates kinks or area of lighter coloration between prominent
individuals with 50% chance of transmitting their RB1 choroidal vessels especially in a light-colored fundus) were
pathogenic variant to their offspring (bilateral or trilateral followed every 2 to 3 weeks and designated as a new tumor
retinoblastoma, positive family history, and when RB1 if growth was documented.5,25 Now, OCT-guided posterior
gene pathogenic variants are documented in blood via pole screening for the first 6 months of life detects even
molecular diagnosis). With high survival due to improved invisible tumors.12 OCT shows that small retinoblastomas
treatments in the last 2 decades, more children with a (<1 mm) arise in the inner11 or outer nuclear layers and
family history of retinoblastoma are anticipated. This may show fishtail and shark-fin signs.21 Tumors in our
highlights the importance of secondary prevention to cohort were detected earlier, and the fishtail sign was
facilitate early diagnosis and treatment, so that these detected in only 3 tumors. These signs may arise later
children have preserved vision and avoid systemic when the tumor grows, pushes on the adjacent retinal
therapies and eye loss.2 Secondary prevention typically layers, and becomes clinically visible. In the current study,
starts with genetic counseling of survivors regarding risk we presumed all invisible tumors to be active new
to offspring.24 Molecular diagnosis of the exact retinoblastoma based on positive RB1 pathogenic status,
pathogenic RB1 variant in the proband improves risk concomitant multifocal tumors, detection of the first tumor
determination and standardizes guidelines for screening per eye within the first 6 months, or an invisible tumor
(clinic visits vs. EUA).6 Screening by OCT entails developed in a previously normal area by OCT.

125
 Ophthalmology Volume 127, Number 1, January 2020

Handheld OCT has become an important tool in evalu- OCT screening and localization can be performed for any
ating retinoblastoma, available in most tertiary level referral at-risk child, but the highest value would be if the initial
ocular oncology centers.25 OCT localization technique can screening examination were in a tumor-free eye. Early-term
be performed after OCT acquisition of an invisible tumor delivery for positive RB1 pathogenic neonates can reduce
allowing treatment to be performed in the same EUA the probability of eyes having tumors at birth to 21%
session with approximately 20 to 30 minutes prolongation. compared with 50% at full term, allowing higher opportu-
Attentive review of every frame is essential to identify nity of screening a tumor-free eye.3 In Canada, with
minute tumors that might appear only in a single frame. available resources and accurate molecular diagnosis,
Management of an invisible tumor necessitates many H1 retinoblastoma survivors currently in their
reproducible laser photocoagulation technique. Laser reproductive years have a 50% risk of having children
delivery requires visualization rendering localization of with retinoblastoma, for whom we suggest a 4-step stra-
treatment area essential for precise laser application.3 OCT tegic secondary prevention plan. The 4 steps are prenatal
software enables precise area determination, and post-laser molecular diagnosis for risk determination,23 planned late-
OCT confirms complete tumor treatment. A major limita- preterm or early-term delivery at 36/37 weeks of gestation
tion of this technique in routine screening is equipment for children at 100% risk,3 OCT-guided posterior pole
availability. Handheld OCT is relatively expensive and screening for invisible tumors,12 and OCT-guided
unavailable in many centers and even whole countries. localization and photocoagulation. Follow-up for each scar
Our study showed high accuracy and effectiveness in by OCT for at least 6 months after last treatment will
invisible tumor control with no misapplied or incomplete detect subclinical recurrence (Fig S3, available at
treatment. With the handheld OCT, assessment of tumor www.aaojournal.org).
extension and characteristics was more valuable than lateral In conclusion, precise localization of small tumors
and vertical measurements. Lateral measurements are resulted in small treatment scars that preserved normal retina
affected by probable angular inclination and subjective through avoiding misapplied laser burns with reduced
caliper placement, with confounding individual factors such treatment burden achieving good visual potential (no foveal
as axial length, lateral tumor location, and refractive error.26 involvement by tumor growth, tumor scar, or misapplied
Our group used lateral calipers in assessment of foveal- laser). The results are limited by the small sample size,
tumor distance in perifoveal tumors that are central and less retrospective nature, and absence of an adjustment for inter-
affected by inclination of OCT acquisition.17 For an invisible eye correlation, making CIs even wider.
tumor, initial accurate size was independent of the success of
the technique. In 11 of 11 tumors, the laser covered the whole
tumor regardless of its size. Follow-up of tumor scars using References
OCT can easily detect recurrences at a subclinical level that
then require only focal laser therapy.12,27 Avoidance of sys-
temic or intra-arterial chemotherapy in central tumors with 1. Meyskens Jr FL, Mukhtar H, Rock CL, et al. Cancer preven-
tion: obstacles, challenges and the road ahead. J Natl Cancer
their respective adverse effects by early laser control of the
Inst. 2016;108(2):1e8.
tumors further reduced treatment burden. 2. Dimaras H, Corson TW, Cobrinik D, et al. Retinoblastoma.
The resultant treatment scar size is an important Nat Rev Dis Primers. 2015;1:15021.
consideration when evaluating any laser technique, espe- 3. Soliman SE, Dimaras H, Khetan V, et al. Prenatal versus
cially in a perifoveal tumor in which a large scar has postnatal screening for familial retinoblastoma. Ophthal-
equivalent deleterious effect on vision. Expansion up to mology. 2016;123:2610e2617.
35% of chorioretinal scars is reported after diode laser 4. Soliman S, Kletke S, Roelofs K, et al. Precision laser therapy
transpupillary thermotherapy within the first year of treat- for retinoblastoma. Expert Rev Ophthalmol. 2018:1e11.
ment.28 Kim et al29 reported a 52% increase in scar size in 5. Rothschild PR, Levy D, Savignoni A, et al. Familial retino-
small tumors after photocoagulation (532 or 810 nm) and blastoma: fundus screening schedule impact and guideline
proposal. A retrospective study. Eye (Lond). 2011;25:
after at least 1 cycle of chemotherapy. No reports
1555e1561.
regarding scar expansion after the sole use of 532 nm 6. Skalet AH, Gombos DS, Gallie BL, et al. Screening children at
laser photocoagulation were published. A degree of scar risk for retinoblastoma: consensus report from the American
expansion is expected with rapid growth of infant’s eye. Association of Ophthalmic Oncologists and Pathologists.
The axial length increases approximately 0.16 mm/week Ophthalmology. 2018;125:453e458.
until full term and 1 mm/year for the first 2 years 7. Imhof SM, Moll AC, Schouten-van Meeteren AY. Stage of
estimating approximately 20% to 25% increase in the presentation and visual outcome of patients screened for fa-
ocular size during the first 2 years.30,31 We postulated milial retinoblastoma: nationwide registration in the
approximately 25% of scar expansion to be due to ocular Netherlands. Br J Ophthalmol. 2006;90:875e878.
growth regardless of the laser technique used. We noticed 8. Abramson DH, Mendelsohn ME, Servodidio CA, et al. Fa-
milial retinoblastoma: where and when? Acta Ophthalmol
minimal circumferential proportionate scar expansion within
Scand. 1998;76:334e338.
25% expansion in all scars except 1. Pigmentary change is a 9. Saktanasate J, Vongkulsiri S, Khoo CT. Invisible Retinoblas-
common secondary change after laser therapy due to retinal toma. JAMA Ophthalmol. 2015;133:e151123.
pigment epithelium hypertrophy and hyperplasia. Progres- 10. Berry JL, Cobrinik D, Kim JW. Detection and intraretinal
sion of scar pigmentation over 1 year further supports the localization of an ‘invisible’ retinoblastoma using optical
physiologic scar expansion postulate. coherence tomography. Ocul Oncol Pathol. 2016;2:148e152.

126
 Soliman et al 
Laser of Invisible Retinoblastoma

11. Rootman DB, Gonzalez E, Mallipatna A, et al. Hand-held tomograph STRATUS OCT 3000. Ophthalmic Physiol Opt.
high-resolution spectral domain optical coherence tomography 2008;28:21e28.
in retinoblastoma: clinical and morphologic considerations. Br 21. Welch RJ, Rao R, Gordon PS, et al. Optical coherence to-
J Ophthalmol. 2013;97:59e65. mography of small retinoblastoma. Asia Pac J Ophthalmol
12. Soliman SE, VandenHoven C, MacKeen LD, et al. Optical (Phila). 2018;7:301e306.
coherence tomography-guided decisions in retinoblastoma 22. Soliman SE, Halliday W, Shaikh F, et al. White orbital mass
management. Ophthalmology. 2017;124:859e872. after enucleation for retinoblastoma: the power of illusion.
13. Ward TJ, Goldman RE, Weintraub JL. Electromagnetic navi- Ophthalmic Genet. 2017;38:584e586.
gation with multimodality image fusion for image-guided 23. Soliman SE, Racher H, Zhang C, et al. Genetics and molecular
percutaneous interventions. Tech Vasc Interv Radiol. diagnostics in retinoblastoma–an update. Asia Pac J Oph-
2013;16:177e181. thalmol (Phila). 2017;6:197e207.
14. Yoon HY, Kim HS, Hwang HB. Guided intracameral air in- 24. Soliman SE, ElManhaly M, Dimaras H. Knowledge of genetics in
jection based on 3D reconstructed anterior optical coherence familial retinoblastoma. Ophthalmic Genet. 2017;38:226e232.
tomography images in iatrogenic Descemet’s membrane 25. Canadian Retinoblastoma Society. National Retinoblastoma
detachment: case report. Medicine (Baltimore). 2018;97: Strategy Canadian Guidelines for Care: strategie therapeutique
e13003. du retinoblastome guide clinique Canadien. Can J Oph-
15. Kim SW, Kim EK. Portable OCT-assisted surgical treatment thalmol. 2009;44(Suppl 2):S1eS88.
of intracorneal pre-Descemet epithelial cyst: a case report. 26. Vinekar A, Mangalesh S, Jayadev C, et al. Retinal imaging of
BMC Ophthalmol. 2017;17:160. infants on spectral domain optical coherence tomography.
16. Carrasco-Zevallos OM, Keller B, Viehland C, et al. Optical Biomed Res Int. 2015;2015:782420.
coherence tomography for retinal surgery: perioperative anal- 27. Gaillard MC, Houghton S, Stathopoulos C, Munier FL. OCT-
ysis to real-time four-dimensional image-guided surgery. guided management of subclinical recurrent retinoblastoma.
Invest Ophthalmol Vis Sci. 2016;57:OCT37e50. Ophthalmic Genet. 2018;39:338e343.
17. Soliman SE, VandenHoven C, Mackeen LD, Gallie BL. 28. Lee TC, Lee SW, Dinkin MJ, et al. Chorioretinal scar growth
Vision and visual potential for perifoveal retinoblastoma after 810-nanometer laser treatment for retinoblastoma.
after optical coherence tomographic-guided sequential laser Ophthalmology. 2004;111:992e996.
photocoagulation. Br J Ophthalmol. 2019;103:753e760. 29. Kim JW, Aziz HA, McGovern K, et al. Treatment outcomes
18. Murphree AL. Intraocular retinoblastoma: the case for a new of focal laser consolidation during chemoreduction for
group classification. Ophthalmol Clin North Am. 2005;18: group B retinoblastoma. Ophthalmol Retina. 2017;1:
41e53. 361e368.
19. Mallipatna A, Gallie BL, Chévez-Barrios P, et al. Retino- 30. Zha Y, Zhu G, Zhuang J, et al. Axial length and ocular
blastoma. In: Amin MB, Edge SB, Greene FL, eds. AJCC development of premature infants without ROP. J Ophthalmol.
Cancer Staging Manual. 8th edition. New York, NY: 2017;2017:6823965.
Springer; 2017:819e831. 31. O’Connor AR, Stephenson TJ, Johnson A, et al. Change of
20. Sanchez-Cano A, Baraibar B, Pablo LE, Honrubia FM. refractive state and eye size in children of birth weight less
Magnification characteristics of the optical coherence than 1701 g. Br J Ophthalmol. 2006;90:456e460.

Footnotes and Financial Disclosures

Originally received: June 11, 2019. HUMAN SUBJECTS: Human subjects were included in this study. This is
Final revision: August 8, 2019. a retrospective noncomparative interventional case series of children treated
Accepted: August 13, 2019. in the Hospital for Sick Children (SickKids), Toronto, Canada. The study
Available online: August 21, 2019. Manuscript no. 2019-1272. complies with the Declaration of Helsinki and was approved by the insti-
1 tutional research ethics board.
Department of Ophthalmology and Vision Sciences, Hospital for Sick
Children, Toronto, Canada. No animal subjects were used in this study.
2
Department of Ophthalmology, Faculty of Medicine, University of Author Contributions:
Alexandria, Alexandria, Egypt. Conception and design: Soliman, VandenHoven, MacKeen, Gallie
3
Department of Ophthalmology & Vision Sciences, Faculty of Medicine, Data collection: Soliman, VandenHoven, MacKeen
University of Toronto, Toronto, Ontario, Canada. Analysis and interpretation: Soliman, VandenHoven, Gallie
4
Departments of Molecular Genetics and Medical Biophysics, Faculty of Obtained funding: N/A
Medicine, University of Toronto, Toronto, Ontario, Canada. Overall responsibility: Soliman, VandenHoven, Gallie
5
Division of Visual Sciences, Toronto Western Research Institute, Toronto, Abbreviations and Acronyms:
Ontario, Canada. CI ¼ confidence interval; EUA ¼ examination under anesthesia.
Financial Disclosure(s): Correspondence:
The author(s) have no proprietary or commercial interest in any materials Sameh E. Soliman, MD, 555 University Avenue, Room 7265, Toronto,
discussed in this article. ON, M5G 1X8. E-mail: [email protected].

127
 Ophthalmic Technology Assessment

Autologous Serum-Based Eye Drops for

Treatment of Ocular Surface Disease
A Report by the American Academy of Ophthalmology
Roni M. Shtein, MD, MS,1 Joanne F. Shen, MD,2 Anthony N. Kuo, MD,3 Kristin M. Hammersmith, MD,4
Jennifer Y. Li, MD,5 Mitchell P. Weikert, MD6

Purpose: To describe the safety and effectiveness of using autologous serum-based eye drops for the
treatment of severe dry eye and persistent corneal epithelial defect.
Methods: Literature searches of the PubMed and Cochrane Library databases were conducted most
recently in March 2019. The searches identified 281 citations, which were reviewed in abstract form. Of these, 48
were selected for a full-text review, and 13 met the inclusion criteria and were assigned a quality-of-evidence
rating by the panel methodologist. Eight of these studies were rated level II and 5 were rated level III; there
were no level I studies.
Results: This analysis included 10 studies of the use of autologous serum-based eye drops for severe dry
eye disease and 4 studies of persistent epithelial defect. Several studies showed good effectiveness, with some
improvement in symptoms, signs, or both. Eight of the studies reported improved symptoms for severe dry eye
disease, and all noted improvement in at least 1 clinical sign. For persistent epithelial defects, all of the studies
showed improvement, with 3 of the 4 demonstrating an improvement rate of more than 90%. Adverse events
were rare.
Conclusions: Although autologous serum-based tears may be effective in the treatment of severe dry
eye and persistent epithelial defect, conclusions are limited owing to the absence of controlled trials.
Ophthalmology 2020;127:128-133 ª 2019 by the American Academy of Ophthalmology

The American Academy of Ophthalmology prepares using an ocular perfusion pump to deliver a variety of so-
Ophthalmic Technology Assessments to evaluate new and lutions, including serum or plasma, to the ocular surface.1 In
existing procedures, drugs, and diagnostic and screening 1984, the topical use of serum for the treatment of
tests. The goal of an Ophthalmic Technology Assessment is Sjögren’s-related dry eye was reported.2 In that study, 15
to review systematically the available research for clinical patients were given their own serum, diluted with
efficacy and safety. After review by members of the preservative-free normal saline to 33% concentration, as a
Ophthalmic Technology Assessment Committee, other tear substitute for a 3-week period. All of the patients noted
Academy committees, relevant subspecialty societies, and improved symptoms, and no adverse events were reported.
legal counsel, assessments are submitted to the Academy’s The composition of serum is similar to that of tears pro-
Board of Trustees for consideration as official Academy duced by the lacrimal gland and includes a variety of
statements. The purpose of this assessment by the components (epithelial growth factor, vitamin A, and others)
Ophthalmic Technology Assessment Committee Cornea and that may contribute to a beneficial effect of serum tears on
Anterior Segment Disorders Panel was to evaluate the safety the corneal epithelium.3,4 After further studies showed pa-
and effectiveness of using autologous serum-based eyedrops tient improvement and detailed the stability of some of the
for the treatment of ocular surface disease. biochemical compounds within the serum,3,5 the use of
serum tears for ocular surface disease became more widely
accepted. Over the past 20 years, treatment of ocular surface
Background disease using autologous serum tears has expanded. Other
topical treatments, including allogeneic serum, umbilical
Blood-derived products have been used for treatment of the cord serum, platelet-rich plasma, plasma rich in growth
ocular surface for decades. The first ophthalmic use reported factors, and nerve growth factor, have been investigated.
in the literature was in 1975 for the treatment of ocular burns Peer-reviewed studies of these topical treatments have been

128 ª 2019 by the American Academy of Ophthalmology https://doi.org/10.1016/j.ophtha.2019.08.018

Published by Elsevier Inc. ISSN 0161-6420/19
 Shtein et al

Ophthalmic Technology Assessment

carried out for severe dry eye, nonhealing or recurrent these, 13 met the following inclusion criteria: (1) the study
epithelial defect, and neurotrophic keratopathy. This had to have 1 month or more of follow-up, and (2) the study
assessment focused on autologous serum-based eye drops population had to include 20 or more patients treated for
and their use in severe dry eye and persistent corneal severe dry eye disease or 15 or more patients treated for
epithelial defect. nonhealing epithelial defect. The inclusion criteria for
nonhealing epithelial defect studies allowed fewer patients
Food and Drug Administration Status because large collections of patients with this condition
are rare.
The panel methodologist (R.M.S.) assigned a level of
Because autologous serum-based eye drops are a blood
evidence to each of the 13 articles that met the inclusion
product and not a pharmaceutical, they are not regulated by
criteria based on the rating scale developed by the British
the United States Food and Drug Administration. Currently,
Centre for Evidence-Based Medicine and adopted by the
there is no federal protocol and there are no requirements for
American Academy of Ophthalmology.6 A level I rating
the use or preparation of autologous serum-based eyedrops
was assigned to well-designed and well-conducted ran-
in the United States, although some states do have
domized clinical trials, a level II rating was assigned to well-
regulations.
designed case-control and cohort studies and lower-quality
randomized clinical trials, and a level III rating was
Resource Requirements assigned to case series, case reports, and lower-quality
cohort and case-control studies. Of the 13 studies, 8 were
There is no universal protocol for the production of autol- rated level II and 5 were rated level III. There were no level
ogous serum-based eye drops. In general, serum is obtained I-rated studies of autologous serum-based eyedrops found in
by routine blood draw with specialized serum-separating the literature.
tubes. The blood is allowed to clot, after which the serum
and solid components of the blood are separated by centri- Published Results
fugation. The serum then can be removed and diluted with
either a balanced salt solution, preservative-free normal sa-
Basic information about the 13 studies included in this
line, or another sterile, preservative-free, eye-compatible
analysis is summarized in Table 1. Table 2 provides more
solution at an appropriate concentration (the range of serum
detailed information on outcomes for patients with severe
concentration in the literature is 20%e100%). As soon as
dry eye disease or nonhealing corneal epithelial defect.
they are formulated, these drops must remain frozen until
None of the studies directly addressed the issues of cost
ready for use and be refrigerated while in use.
or accessibility of the treatment.
Compounding pharmacies with the appropriate facilities
can produce autologous serum eye drops, and some provide Severe Dry Eye Disease
mail service. In some areas of the United States, eye banks
have the facilities to compound autologous serum eye drops. Ten of the studies evaluated the use of autologous serum-
Some ophthalmology offices provide this as a service to based eye drops for the treatment of severe dry eye.7e16
their patients, although regulations differ by state, so These studies included patients with moderate to severe
adherence to local standards of care should be ensured. dry eye symptoms and signs, some of whom had Sjögren’s
Costs differ substantially between suppliers, and serum- syndrome or graft-versus-host disease, and all were resistant
based eye drops usually cost several hundred dollars for a to conventional dry eye treatment options. All of these
2- to 3-month supply. Insurance coverage varies greatly. studies provided level II or III evidence, and most were case
Accessibility and cost are substantial barriers to the use of series or case-control studies. There were no randomized
this treatment method. controlled studies found in the peer-reviewed literature. The
10 studies included patients with severe dry eyes resulting
Questions for Assessment from multiple causes, used a variety of different serum
concentrations, and included variable frequencies of eye
drop use (Table 1). Despite these variations, the available
The purpose of this assessment was to address the
evidence supports the effectiveness of topical autologous
following 2 questions: (1) Are autologous serum-based
serum eye drops. Patient symptoms were assessed
eyedrops safe and effective for the treatment of dry
subjectively using a survey questionnaire symptom score.
eyes? (2) Are autologous serum-based eyedrops safe and
There was statistically significant improvement in the
effective for the treatment of nonhealing corneal epithelial
subjective symptom score in 6 of 10 studies,7,8,10,13,14,16
defect?
and there was improvement that was not statistically sig-
nificant in 2 studies.9e11 One study reported no change in
Description of Evidence symptom scores,15 and the final study did not report on
patient symptoms.12 At least 1 clinical measure of ocular
Literature searches of the PubMed and Cochrane library surface disease (i.e., ocular surface staining, Schirmer
databases were conducted most recently in March of 2019. testing, tear film breakup time, or cytologic analysis)
They identified 281 citations, which were reviewed in ab- showed statistically significantly improvement after
stract form, and 48 were selected for a full-text review. Of treatment in 8 studies.7,8,10,12e16 Only 1 of the studies

129
 Ophthalmology Volume 127, Number 1, January 2020

Table 1. Summary of Included Studies

Age
(yrs; Mean ±
Level of Dilution No. of Standard Gender (%; Follow-up
Authors (Year) Title Evidence (%) Participants Deviation) Female/Male) Condition (mos)
Noda-Tsuruya Autologous serum eye drops for dry eye II 20 27 306 0/100 DED 6
et al15 (2006) after LASIK
Yoon et al16 Comparison of autologous serum and II 20 48 4011 52/48 DED 2
(2007) umbilical cord serum eye drops for dry
eye syndrome
Kim et al17 Effect of autologous platelet-rich plasma II 20 17 67 41/59 PED 1
(2012) on persistent corneal epithelial defect
after infectious keratitis
Cho et al8 Comparison of autologous serum eye II 100 85 NR 61/39 DED, PED 3
(2013) drops with different diluents
Celebi et al7 The efficacy of autologous serum eye II 20 20 568 190/10 DED 2
(2014) drops for severe dry eye syndrome: a
randomized double-blind crossover
study
Hussain et al9 Long-term use of autologous serum 50% II 50 63 6111 83/17 DED 12
(2014) eye drops for the treatment of dry eye
disease
Hwang et al10 Comparison of clinical efficacies of II 50 34 569 100/0 DED 1
(2014) autologous serum eye drops in patients
with primary and secondary Sjögren
syndrome
Lopez-Garcia Autologous serum eye drops diluted with II 20 52 5213 92/8 DED 2
13
et al (2014) sodium hyaluronate: clinical and
experimental comparative study
Liu et al12 Effectiveness of autologous serum eye III 20 28 5614 89/11 DED 42
(2015) drops combined with punctal plugs for
the treatment of Sjögren syndrome-
related dry eye
Lee and Chen11 Autologous serum in the management of III 20 23 6314 83/17 DED 17
(2008) recalcitrant dry eye syndrome
Lekhanont Topical 100% serum eye drops for III 100 181 6214 51/49 PED 3
et al18 (2013) treating corneal epithelial defect after
ocular surgery
Semeraro et al19 Evaluation of the efficacy of 50% III 50 15 4017 NR PED 4
(2014) autologous serum eye drops in different
ocular surface pathologies
Mahelkova Using corneal confocal microscopy to III 20 26 5114 73/27 DED 3
et al14 (2017) track changes in the corneal layers of
dry eye patients after autologous serum
treatment

DED ¼ dry eye disease; NR ¼ not reported; PED ¼ persistent epithelial defect.

reported visual acuity results with no changes noted after Nonhealing Corneal Epithelial Defect
treatment.13 Eight of the studies reported side effects or
adverse events,7e9,11e15 and only 1 study reported an Four studies evaluated the use of autologous serum-based
adverse event of microbial growth measured in an eye drop eye drops for the treatment of persistent corneal epithelial
bottle with no clinical sequelae.8 No patient-reported defect, 2 with level II evidence and 2 with level III
negative symptoms from the treatments were noted. One evidence.8,17e19 All showed substantial improvement in the
study compared serum concentrations of 50% versus 100% epithelial defects, and 3 showed a reduction of more than
and found that the 100% concentration was more effective 90% in the size of the defects.8,18,19 A study comparing 50%
in patients with Sjögren’s syndrome yet was not better for versus 100% serum concentration found faster closure of the
other types of severe dry eye.8 Another study compared the epithelial defects in patients using the 100% concentration
treatment effectiveness in primary versus secondary eye drops.8 In 1 study, a patient experienced a recurrence of
Sjögren’s syndrome and suggested that autologous serum- the epithelial defect when the serum-based eye drops were
based eye drops work better for patients with primary discontinued.19 In another study, 1 of the bottles of serum
Sjögren’s syndrome.10 A different study compared the eye drops was found to have microbial growth that
diluent (normal saline vs. sodium hyaluronate) and matched the bacterium (Serratia marcescens) cultured
showed no statistical difference.13 from the patient’s corneal lesion.8

130
 Table 2. Autologous Serum Results

Persistent
Tear Film Cytologic Epithelial
Level of Dilution Follow-up Visual Surface Schirmer’s Breakup Analysis Defect
Authors (Year) Evidence (%)* Condition (mos) Acuity Stain Test Results Time Results Healed (%) Symptoms Side Effects
Noda-Tsuruya II 20 DED 6 NR Improved Unchanged SSI NR NA Unchanged None

Shtein et al
et al15 (2006) but not SSI
Yoon et al16 (2007) II 20 DED 2 NR SSI Unchanged SSI SSI NA SSI NR
Cho et al8 (2013) II 50/100 DED 3 NR SSI SSI SSI NR NA SSI MGBy
Celebi et al7 (2014) II 20 DED 2 NR SSI NR SSI NR NA SSI None
Hussain et al9 (2014) II 50 DED 12 NR Improved Improved NR NR NA Improved None

but not SSI but not SSI but not SSI
DED (1 SS)

Ophthalmic Technology Assessment

Hwang et al10 (2014) II 50 1 NR SSI NR SSI NR NA SSI NR
50 DED (2 SS) 1 NR Improved NR Improved NR NA Improved NR
but not SSI but not SSI but not SSI
Lopez-Garcia II 20/saline DED 2 Unchanged SSI Improved SSI SSI NA SSI None
et al13 (2014) but not SSI
20/hyaluronidase DED 2 Unchanged SSI Improved SSI SSI NA SSI None
but not SSI
12
Liu et al (2015) II 20 DED 42 NR SSI Unchanged SSI NR NA NR None
Lee and III 20 DED 17 NR Improved NR NR NR NA Improved None
Chen11 (2008) but not SSI but not SSI
Mahelkova III 20 DED 3 NR SSI NR NR NR NA SSI None
et al14 (2017)
Kim et al17 (2012) II 20 PED 1 NR NR NR NR NR 71 NR NR
Cho et al8 (2013) II 50/100 PED 3 NR SSI SSI Unchanged NR 100 SSI MGBz
Lekhanont III 100 PED 3 NR NR NR NR NR 94 NR None
et al18 (2013)
Semeraro III 50 PED 4 SSI NR NR NR NR 100 SSI None
et al19 (2014)

DED ¼ dry eye disease; MGB ¼ microbial growth in bottle; NA ¼ not applicable; NR ¼ not reported; PED ¼ persistent epithelial defect; SS ¼ Sjögren’s syndrome; SSI ¼ statistically significantly improved.
*Percent autologous serum used.
y
No clinical infection.
z
Correlated with microbe in corneal lesion.
131
 Ophthalmology Volume 127, Number 1, January 2020

Conclusions 2. Fox RI, Chan R, Michelson JB, et al. Beneficial effect of

artificial tears made with autologous serum in patients with
Conclusive evidence on the safety and effectiveness of keratoconjunctivitis sicca. Arthritis Rheum. 1984;27(4):
459e461.
autologous serum-based tears is limited by the lack of 3. Tsubota K, Goto E, Fujita H, et al. Treatment of dry eye by
controlled studies and by the variability in components of the autologous serum application in Sjogren’s syndrome. Br J
study protocols. The limited accessibility and substantial Ophthalmol. 1999;83(4):390e395.
cost of autologous serum-based eye drops create challenges 4. Quinto GG, Campos M, Behrens A. Autologous serum for
for implementation, and therefore result in reserving their ocular surface diseases. Arq Bras Oftalmol. 2008;71(6 Suppl):
use either for more severe cases or for cases that have not 47e54.
improved using more readily available and less costly ther- 5. Tsubota K, Goto E, Shimmura S, Shimazaki J. Treatment of
apies. The results of the studies in the peer-reviewed litera- persistent corneal epithelial defect by autologous serum
ture suggest that this treatment is a reasonable option in application. Ophthalmology. 1999;106(10):1984e1989.
refractory cases of dry eyes or nonhealing epithelial defects. 6. Oxford Centre for Evidence-Based Medicine. Levels of evi-
dence. http://www.cebm.net/index.aspx?o¼1025; March 2009.
The primary safety consideration for autologous serum- Accessed 4.3.19.
based eye drops is the risk of microbial growth during 7. Celebi AR, Ulusoy C, Mirza GE. The efficacy of autologous
storage, because serum-based solutions essentially are serum eye drops for severe dry eye syndrome: a randomized
growth media. Care must be taken in the preparation of double-blind crossover study. Graefes Arch Clin Exp Oph-
these eye drops to ensure that they are prepared in a sterile thalmol. 2014;252(4):619e626.
manner, and proper care and use instructions must be fol- 8. Cho YK, Huang W, Kim GY, Lim BS. Comparison of
lowed by patients to minimize contamination. Compound- autologous serum eye drops with different diluents. Curr Eye
ing pharmacies and eye banks have the equipment necessary Res. 2013;38(1):9e17.
to reduce the risk of contamination during preparation. 9. Hussain M, Shtein RM, Sugar A, et al. Long-term use of
Microbial contamination remains a considerable risk in pa- autologous serum 50% eye drops for the treatment of dry eye
disease. Cornea. 2014;33(12):1245e1251.
tients who have a compromised ocular surface. Although no 10. Hwang J, Chung SH, Jeon S, et al. Comparison of clinical
patients in the dry eye studies included in this assessment efficacies of autologous serum eye drops in patients with pri-
experienced any reported clinical adverse events, bacterial mary and secondary Sjogren syndrome. Cornea. 2014;33(7):
growth was reported in the eye drops of 2 patients with no 663e667.
adverse clinical consequence. A patient in 1 of the non- 11. Lee GA, Chen SX. Autologous serum in the management of
healing epithelial defect studies demonstrated microbial recalcitrant dry eye syndrome. Clin Exp Ophthalmol.
infection of the corneal lesion, and the bacterium identified 2008;36(2):119e122.
in the eye drop solution was the same as the bacterium 12. Liu Y, Hirayama M, Cui X, et al. Effectiveness of autologous
identified from the corneal culture.8 serum eye drops combined with punctal plugs for the treatment
of Sjogren syndrome-related dry eye. Cornea. 2015;34(10):
1214e1220.
Future Research 13. López-Garcia JS, Garcia-Lozano I, Rivas L, et al. Autologous
serum eye drops diluted with sodium hyaluronate: clinical and
The existing peer-reviewed studies on the use of autologous experimental comparative study. Acta Ophthalmol.
serum-based eye drops are limited to observational designs 2014;92(1):e22ee29.
14. Mahelkova G, Jirsova K, Seidler Stangova P, et al. Using
and are mostly case series. A randomized controlled study
corneal confocal microscopy to track changes in the corneal
would make it possible to evaluate the effectiveness of the layers of dry eye patients after autologous serum treatment.
treatment and to assess the optimal concentration and fre- Clin Exp Optom. 2017;100(3):243e249.
quency of use of the drops for various ocular surface con- 15. Noda-Tsuruya T, Asano-Kato N, Toda I, Tsubota K. Autolo-
ditions. However, the low prevalence of persistent epithelial gous serum eye drops for dry eye after LASIK. J Refract Surg.
defects makes it unlikely that such a study will be con- 2006;22(1):61e66.
ducted. It is also possible that further research may help to 16. Yoon KC, Heo H, Im SK, et al. Comparison of autologous
improve accessibility of serum-based eye drops. Further serum and umbilical cord serum eye drops for dry eye syn-
research also is warranted to develop greater understanding drome. Am J Ophthalmol. 2007;144(1):86e92.
of the mechanism of action of serum-based eye drops and to 17. Kim KM, Shin YT, Kim HK. Effect of autologous platelet-rich
plasma on persistent corneal epithelial defect after infectious
target the specific components of serum that are most
keratitis. Jpn J Ophthalmol. 2012;56(6):544e550.
helpful to the ocular surface. 18. Lekhanont K, Jongkhajornpong P, Choubtum L,
Chuckpaiwong V. Topical 100% serum eye drops for treating
References corneal epithelial defect after ocular surgery. Biomed Res Int.
2013;2013:521315.
19. Semeraro F, Forbice E, Braga O, et al. Evaluation of the
1. Ralph RA, Doane MG, Dohlman CH. Clinical experience with efficacy of 50% autologous serum eye drops in different
a mobile ocular perfusion pump. Arch Ophthalmol. ocular surface pathologies. Biomed Res Int. 2014;2014:
1975;93(10):1039e1043. 826970.

132
 Shtein et al

Ophthalmic Technology Assessment

Footnotes and Financial Disclosures

Originally received: August 14, 2019. K.M.H.: Consultant and Lecturer e Shire.
Final revision: August 16, 2019. M.P.W.: Consultant e Alcon Laboratories, Inc., Carl Zeiss Meditec.
Accepted: August 16, 2019. Funded without commercial support by the American Academy of
Available online: September 24, 2019. Manuscript no. 2019-171. Ophthalmology.
1
Department of Ophthalmology and Visual Sciences, University of HUMAN SUBJECTS: This study did not use human subjects.
Michigan, Ann Arbor, Michigan.
2 No animal subjects were included in this study.
Mayo Clinic Arizona, Scottsdale, Arizona.
3 Author Contributions:
Department of Ophthalmology, Duke Eye Center, Duke University
Medical Center, Durham, North Carolina. Conception and design: Shtein, Shen, Kuo, Hammersmith, Li, Weikert
4 Analysis and interpretation: Shtein, Shen, Kuo, Hammersmith, Li, Weikert
Wills Eye Hospital, Philadelphia, Pennsylvania.
5 Data collection: Shtein, Shen, Kuo, Hammersmith, Li, Weikert
UC Davis Eye Center, University of California, Davis, Sacramento,
California. Obtained funding: N/A
6
Cullen Eye Institute, Baylor College of Medicine, Houston, Texas. Overall responsibility: Shtein, Shen, Kuo, Hammersmith, Li, Weikert
Prepared by the Ophthalmic Technology Assessment Committee Cornea Correspondence:
and Anterior Segment Disorders Panel and approved by the American Ali Al-Rajhi, PhD, MPH, American Academy of Ophthalmology, Quality
Academy of Ophthalmology’s Board of Trustees July 2, 2019. and Data Science, P. O. Box 7424, San Francisco, CA 94120-7424. E-mail:
Financial Disclosure(s): [email protected].
The author(s) have made the following disclosure(s): A.N.K.: Patent e
Leica.

133
 Reports

Redundancy of Progress Notes the provider’s 3 most common (November 1, 2016, through
for Serial Office Visits October 31, 2017). One author (A.E.H.) manually coded the first
note of each encounter report pair into SOAP and other sections
Adoption of electronic health records (EHRs) has transformed and a second author (B.H.) independently coded 4 representative
ophthalmology documentation. Whereas handwritten paper charts reports to assess interobserver agreement. We used document
encouraged brevity, documentation in EHRs can be composed with comparison software (Workshare Compare; Workshare, San
content-importing technologies such as templates and copy and Francisco, CA) to highlight new text in the second report of each
paste,1 resulting in longer electronic notes than their paper pair and counted new and total words per section. We used a
counterparts.2 Furthermore, electronic notes can look remarkably Kruskal-Wallis test with Bonferroni correction to assess overall
similar and may contain outdated or erroneous information that difference in redundancy between note sections and multiple post
contributes to medical errors.3,4 One step toward weighing the hoc Dunn tests to identify specific statistically different sections.
risks and benefits of content-importing technologies is under- All calculations for this manual review were performed in
standing similarity among notes. Although a few studies found spreadsheet software (Excel 2016; Microsoft, Redmond, WA)
rates of redundancy of approximately 30% in primary care notes,5,6 with XLSTAT add-in software (Addinsoft, New York, NY).
to our knowledge, no study has quantified similarity rigorously For the large-scale analysis with 12 355 progress note pairs, the
between subsequent outpatient ophthalmology notes. second note was on average 74.5% redundant with the first (95%
This study was conducted at Casey Eye Institute, the ophthal- confidence interval [CI], 74.2%e74.8%; Fig 1). That is, 74.5% of
mology department of Oregon Health and Science University note text was found verbatim in the progress note for the patient’s
(OHSU), a large academic medical center in Portland, Oregon, and previous office visit. The 24 710 progress notes in our sample
was approved by Oregon Health and Science University’s insti- averaged 610
563 words in length. By comparison, across the
tutional review board, which granted a waiver of informed consent 10 000 randomly paired progress notes, the second note was on
for analysis of EHR data. This study adhered to the tenants of the average 19.3% redundant with the first (95% CI, 19.1%e19.5%).
Declaration of Helsinki. We first performed a large-scale analysis The 120 manually reviewed encounter reports were, on average,
of note redundancy using natural language processing. We 75.4% redundant (95% CI, 73.5%e77.3%) and 1129
288 words
included all 48 faculty providers (42 ophthalmologists, 6 optom- long. Interobserver agreement for word count by section was
etrists) who saw patients between January 1, 2017, and December excellent (Spearman’s r ¼ 0.983; P < 0.0001). The Plan section
31, 2018, and collected progress notes for follow-up visits where had the highest proportion of new words (44.6%; 95% CI, 39.7%e
the primary visit diagnosis (parent International Classification of 49.6%). All other sections ranged from 22.7% to 27.4% new text
Diseases, Tenth Edition, diagnostic code) was 1 of the provider’s 3 (Table S1, available at www.aaojournal.org). The percentage of
most common based on billing data. We paired each note with the new words in the Plan section was significantly higher than in
progress note for the next office visit by the same patient to the the other 4 sections (P < 0.001, post hoc Dunn test). There were
same provider for the same diagnosis. We repeated this process to no statistically significant differences among other sections.
generate note pairs for all 12 355 patients who attended 2 or more To our knowledge, this is one of the first studies quantifying
follow-up appointments with a study provider during the study redundancy of ophthalmology notes. We found that most text in
period. For comparison, we collected 10 000 random pairs of notes, both progress notes and encounter reports does not change between
ensuring that each note represented a different patient. serial visits. Moreover, we found that the Plan section changed the
We used sequence alignment to assess note redundancy, using most between serial visits, with nearly half of Plan text being new,
the modified Levenshtein edit-distance algorithm5 to generate a possibly reflecting a greater propensity for providers to rethink and
master note containing an aligned sequence of words from either modify their plan at each visit. Together, these findings identify a
note. Redundancy was defined as the percentage of words from high level of redundancy in EHR documentation for ophthal-
the second note that aligned with words from the first note. mology (approximately 75%), especially when compared with the
Redundancy calculations were performed in Python version 3.7 19.1% redundancy in randomly paired progress notes. Although
(Python Software Foundation; https://www.python.org/). some of this redundancy may reflect helpful repeated structure
Next, to analyze redundancy by SOAP section (i.e., Subjec- (e.g., section headers) or stable objective findings, other redundant
tive, Objective, Assessment, Plan), we manually reviewed 120 text may be included for nonclinical purposes (e.g., billing attes-
pairs of encounter reports, which at our institution, are generated tations) or irrelevant to the visit (e.g., some medication lists). Most
automatically to include the progress note, examination findings, encounter reports (88/120) even had duplicate examination text
and any procedure notes for a visit. We analyzed encounter re- within the same report (Table S1). The American Academy of
ports because many providers leave objective findings out of Ophthalmology’s guideline that “the final patient note must be
their progress notes. We selected serial encounter reports for 15 edited carefully to ensure accuracy and relevance to the current
patients for each of 8 attending ophthalmology providers (2 each visit” after use of content-importing begins to address this issue.7
from cornea, neuro-ophthalmology, retina, and comprehensive However, more may be carried out prospectively to ensure that
subspecialties), randomly selecting the first report from a pool of documentation aids, such as templates, support concise and
follow-up office visits where the primary visit diagnosis was 1 of accurate documentation.

134
 Reports

Figure 1. Bar graph showing the distribution of ophthalmology progress note redundancy based on automated text analysis between 12 355 pairs of serial
notes by the same provider seeing the same patient. On average, 74.5% of text in the second note was redundant with the first note. Forty-eight providers
were included in this analysis.

Further study is needed to quantify rigorously the impact of No animal subjects were included in this study.
redundancy on provider efficiency and patient safety. Moreover, the
Author Contributions:
amount and type of redundant text that is helpful or harmful depends Conception and design: Hribar, Rule, Huang, Chiang
on specialty and end use, which additional, more targeted documen- Analysis and interpretation: Hribar, Rule, Huang, Dusek, Goldstein,
tation review may clarify. Still, this high level of redundancy may Henriksen, Lin, Igelman, Chiang
reflect documentation designed as an aid for billing, medicolegal Data collection: Hribar, Rule, Huang, Dusek, Goldstein, Chiang
defense, or the consolidation of information across a fragmented re- Obtained funding: Hribar, Chiang
cord, rather than concise communication of clinical reasoning. Overall responsibility: Hribar, Rule, Huang, Dusek, Goldstein, Hen-
riksen, Lin, Igelman, Chiang
MICHELLE R. HRIBAR, PHD1,2,*
Correspondence:
ADAM RULE, PHD1,*
Michelle R. Hribar, PhD, Department of Ophthalmology, Casey Eye
ABIGAIL E. HUANG, MD2,* Institute, Oregon Health and Science University, Mail Code: BICC,
HALEY DUSEK, BA1 3181 SW Sam Jackson Park Road, Portland, OR 97239. E-mail:
ISAAC H. GOLDSTEIN, BA1 [email protected].
BRAD HENRIKSEN, MD1
WEI-CHUN LIN, MD2
AUSTIN IGELMAN, BS1 References
MICHAEL F. CHIANG, MD1,2
1
Department of Ophthalmology, Casey Eye Institute, Oregon Health 1. Weis JM, Levy PC. Copy, paste, and cloned notes in electronic
and Science University, Portland, Oregon; 2Department of Medical health records. Chest. 2014;145(3):632e638.
Informatics and Clinical Epidemiology, Oregon Health and Science 2. Sanders DS, Lattin DJ, Read-Brown S, et al. Electronic health
University, Portland, Oregon record systems in ophthalmology: impact on clinical docu-
*
M.H.R., A.R., and A.E.H. contributed equally as first authors. mentation. Ophthalmology. 2013;120(9):1745e1755.
Financial Disclosure(s): The author(s) have made the following dis-
3. Tsou A, Lehmann C, Michel J, et al. Safe practices for copy and
closure(s): A.E.H.: Consultant e Advanced Clinical (Deerfield, IL).
paste in the EHR: systematic review, recommendations, and
M.F.C.: Consultant e Novartis (Basel, Switzerland); Financial
novel model for health IT collaboration. Applied Clinical
support e Genentech; Scientific advisory board e Clarity Medical
Informatics. 2017;26(01):12e34.
Systems (Pleasanton, CA); Equity owner e Inteleretina, LLC (Hono-
4. Singh H, Giardina TD, Meyer AND, et al. Types and origins of
lulu, HI); Patent (provisional) e I-ROP deep learning system.
diagnostic errors in primary care settings. JAMA Intern Med.
Supported by the National Institutes of Health, Bethesda, Maryland
2013;173(6):418e425.
(grant nos.: R00LM12238 and P30EY10572); Research to Prevent
5. Wrenn JO, Stein DM, Bakken S, Stetson PD. Quantifying
Blindness, Inc., New York, New York (unrestricted departmental
clinical narrative redundancy in an electronic health record.
funding); and Children’s Hospital of Philadelphia, Philadelphia, Penn-
J Am Med Inform Assoc. 2010;17(1):49e53.
sylvania (lecture honorarium [M.R.H.]). The funding organizations had
6. Cohen R, Elhadad M, Elhadad N. Redundancy in electronic
no role in the design or conduct of this research.
health record corpora: analysis, impact on text mining perfor-
mance and mitigation strategies. BMC Bioinformatics.
HUMAN SUBJECTS: No human subjects were included in this study. 2013;14(1):10.
The human ethics committees at Oregon Health and Science University 7. Silverstone DE, Lim MC. American Academy of Ophthal-
approved the study and granted a waiver of informed consent for mology Medical Information Technology Committee. Ensuring
analysis of electronic health records data. This study adhered to the information integrity in the electronic health record: the crisis
tenants of the Declaration of Helsinki. and the challenge. Ophthalmology. 2014;121(2):435e437.

135
 Ophthalmology Volume 127, Number 1, January 2020

Effects of Omega-3 mOsm/l (P ¼ 0.02; Table S2, available at www.aaojournal.org;

Supplementation on Fig 1A).
The baseline keratography measurements were similar between
Exploratory Outcomes in the
treatment groups (Table S1). The mean noninvasive keratographic
Dry Eye Assessment and tear film break-up time decreased by 0.5 second in each group
Management Study (P ¼ 0.97; Table S2; Fig 1B). The change in mean tear meniscus
height was near 0 in the active (0.00 mm) and placebo (e0.01 mm)
The Dry Eye Assessment and Management (DREAM) Study was a
groups (P ¼ 0.71; Table S2; Fig 1C). The mean change in bulbar
multicenter (27 sites), randomized, double-masked clinical trial for
conjunctival redness score was near 0 in the active (0.00) and
people with moderate to severe dry eye disease (DED).1 Between
placebo (e0.01) groups (P ¼ 0.81; Table S2; Fig 1D). The
October 2014 and July 2016, 535 participants were assigned in a
percentage of eyes with Pult scale scores indicating
2:1 ratio to either active omega-3 fatty acid daily supplements (2
improvement, stability, or worsening by 1 or more categories
g eicosapentaenoic acid and 1 g docosahexaenoic acid) or placebo
was similar for the upper lid (P ¼ 0.34) and lower lid
(5 g refined olive oil). One-year results showed no difference be-
(P ¼ 0.21; Table S2).
tween the omega-3 and placebo groups for the primary outcome of
At baseline, the MMP 9 test showed positive results for similar
symptoms, as measured by the Ocular Surface Disease Index, or
proportions of eyes in the active (33%) and placebo (30%) groups.
the traditional signs of DED (conjunctival and corneal staining, tear
Between baseline and 3 months, 10% of eyes in the active group
film break-up time, and Schirmer II test results).1
and 13% of eyes in the placebo group converted negative to pos-
Additional signs of DED acquired through use of devices were
itive results, and 13% of each group converted from positive to
assessed in the DREAM Study as exploratory outcome measures.
negative results (P ¼ 0.69; Table S2).
Clinical staff completed a certification program including review of
Results of analyses of the mean difference between active and
the protocol and instructional slides and a written test for each
placebo groups within subgroups are displayed in Table S3
device. Measurements were made according to the manufacturer’s
(available at www.aaojournal.org). None of the tests of
instructions.
interaction showed statistically significant results (all P
0.39).
Testing was performed on both eyes with the right eye first.
In this randomized, double-masked clinical trial, there were no
Tear osmolarity was measured using the TearLab Osmolarity
significant differences between daily supplementation with omega-
System (TearLab, San Diego, CA). The Keratograph 5M (Oculus,
3 versus refined olive oil supplementation in noninvasive kerato-
Arlington, WA) was used for noninvasive keratographic tear film
graphic tear film break-up time, tear meniscus height, bulbar
break-up time, tear meniscus height, bulbar conjunctival redness,
conjunctival redness, upper and lower lid meibography, and MMP
and meibomian gland imaging. The examiner everted each eyelid
9 positivity (all P > 0.21). Only the mean change in tear osmolarity
and used the keratograph’s infrared photography system to capture
yielded a statistically significant difference, with slight improve-
images of meibomian glands. Examiners graded meibomian gland
ment in the active treatment group (e0.7 mOsm/l) when compared
dropout on the Pult scale.2 When lid eversion or image quality was
with the worsening in the placebo treatment group (þ3.6 mOsm/l).
insufficient to judge dropout area, the result was classified as
The mean changes over time within each treatment group were
missing. Matrix metalloproteinase (MMP) 9 testing was
small for keratography measures, and the net change in classifi-
performed with the InflammaDry system (RPS Diagnostics,
cation of meibomian gland dropout and MMP 9 positivity was
Sarasota, FL). Keratography and tear osmolarity testing was
small. When subgroups were examined, there was no evidence of a
conducted only at centers equipped with the devices. Testing was
greater benefit of omega-3 supplementation among eyes with more
at baseline, 6 months, and 12 months, except for MMP 9 testing
abnormal values at baseline.
(carried out at screening and 3 months).
Although a small improvement was observed in the mean
Differences between treatment groups were estimated with
change in tear osmolarity for the active group and a worsening in
regression models using a generalized estimating equations
the placebo group, there was no difference between the active and
approach to account for intereye correlation. Subgroups were
placebo groups at 12 months (303.118.4 mOsm/l vs. 303.3 17.5
defined based on the baseline values of the measures for signs,
mOsm/l; P ¼ 0.90). These findings are difficult to interpret given
using category bounds to form tertiles or, for tear osmolarity, a
the high variability among readings from the TearLab system and
previously defined threshold for abnormal (
308 mOsm/l). Vari-
lack of correlation changes in tear osmolarity with changes in
ation in treatment effects across subgroups was assessed with tests
symptoms or corneal fluorescein staining.3,4
of interaction.
Although several clinical trials have tested the efficacy of omega-3
The DREAM Study protocol was approved by each center’s
in treating symptoms of DED, only 3 addressed the exploratory out-
institutional review board, complied with the Health Insurance
comes used in the DREAM study.5e7 Three studies measured tear
Portability and Accountability Act, and adhered to the tenets of the
osmolarity using the TearLab system, showing improvements relative
Declaration of Helsinki. Patients provided written informed con-
to placebo after shorter periods (90 days) with lower doses of omega-3
sent. The trial was registered on ClinicalTrials.gov (identifier,
supplementation than in the DREAM Study. Matrix metalloproteinase
NCT02128763).
9 positivity and bulbar conjunctival redness also were measured in 2
The baseline mean  standard deviation value of tear osmo-
small (n < 55) studies and showed improvement within 90 days of
larity in the active group (303.917.2 mOsm/l) was higher than in
omega-3 supplementation.6,7
the placebo group (300.614.5 mOsm/l; P ¼ 0.02; Table S1,
In conclusion, omega-3 supplementation was not beneficial
available at www.aaojournal.org). The mean change was a
relative to placebo for most of the exploratory measures. Although
decrease of 0.7 mOsm/l in the active group and an increase of
there was a difference in the mean change in tear osmolarity in
3.6 mOsm/l in the placebo group, yielding a difference of 4.3
favor or the omega-3 group, the clinical significance of the

136
 Reports

Figure 1. Graphs showing the mean level of continuous exploratory outcomes at baseline and through 12 months by treatment group: (A) tear osmolarity,
(B) keratograph tear film break-up time; (C) tear meniscus height, and (D) bulbar conjunctival redness score. Red line denotes the active group and blue
line denotes the placebo group. Vertical bars denote 95% confidence intervals.

difference is unclear. These findings are consistent with the results Presented at: Association for Research in Vision and Ophthalmology
of no difference between the omega-3 and placebo groups for the Annual Meeting, May 2019, Vancouver, Canada.
primary and secondary outcomes of the DREAM Study. Financial Disclosure(s): The author(s) have made the following dis-
closure(s): M.C.L.: Financial support e Amorphex Therapeutics
MARKO OYDANICH, MS1
(Andover, MA), CooperVision, Inc (Pleasanton, CA), Essilor USA
MAUREEN G. MAGUIRE, PHD2 (Dallas, TX), GLIA LLC (Madison, WI), Johnson & Johnson (Jack-
MAXWELL PISTILLI, MS2 sonville, FL), Leo Lens, Inc (San Diego, CA), Orinda Pharma (Orinda,
PEDRAM HAMRAH, MD3 CA), Verily Life Science (South San Francisco, CA), Viewpoint
JACK V. GREINER, DO, PHD4 Pharmaceuticals (South San Francisco, CA), Google X (Mountain
View, CA), Health Advances (Weston, MA), Novartis (Cambridge,
MENG C. LIN, OD, PHD5
MA), Shire (Lexington, MA).
PENNY A. ASBELL, MD, MBA6 P.A.A.: Financial support e Santen, Shire, Novartis, Medscape (New
FOR THE DRY EYE ASSESSMENT AND MANAGEMENT STUDY York, NY), MC2 Therapeutics (Copenhagen, Denmark), Valeant
RESEARCH GROUP* (Laval, Canada), Bausch & Lomb (Rochester, NY), Allergan (Dublin,
1
Department of Cell Biology and Molecular Medicine, RutgersdNew Ireland), ScientiaCME (Cambridge, England), Rtech (New York, NY),
Jersey Medical School, Newark, New Jersey; 2Department of Oculus (Menlo Park, CA), Miotech (Hold, MI), Shire, Santen (Goelta,
Ophthalmology, University of Pennsylvania, Philadelphia, CA), CLAO (Denver, CO), Vindico (Thorofare, NJ); Nonfinancial
Pennsylvania; 3Cornea Service, New England Eye Center, Department support e Valeant, Bausch & Lomb, Shire, Santen, CLAO.
of Ophthalmology, Tufts Medical Center, Tufts University School of Supported by the National Eye Institute, National Institutes of
Medicine, Boston, Massachusetts; 4Schepens Eye Research Institute, Health, Department of Health and Human Services, Bethesda,
Massachusetts Eye and Ear Infirmary, Department of Ophthalmology, Maryland (cooperative agreement nos.: U10EY022879 and
Harvard Medical School, Boston, Massachusetts; 5Clinical Research U10EY022881). Additional support provided by the Office of Di-
Center, Berkeley Optometry, University of California, Berkeley, etary Supplements National Institutes of Health, Department of
Berkeley, California; 6Hamilton Eye Institute, University of Tennessee Health and Human Services. The sponsor participated in overseeing
Health Science Center (UTHSC), Memphis Tennessee the conduct of the study. EPAX Norway AS (Alesund, Norway)
*
A complete listing of the members of the Dry Eye Assessment and contributed the fish oil and the Access Business Group; and LLC
Management (DREAM) Research Group is available at (Ada, MI) manufactured, packaged, and delivered study supplements
www.aaojournal.org. to the central pharmacy. OCULUS, Inc. (Arlington, WA), provided

137
 Ophthalmology Volume 127, Number 1, January 2020
discounted equipment, customized software, and user support for the Preferred Practice Patterns for Amblyopia2 recommends that
keratograph used in the study. RPS Diagnostics, Inc. (Sarasota, FL), individuals with unilateral vision impairment with vision of
provided InflammaDry Detector test kits to the clinical centers for 20/50 in the amblyopic eye should wear protective eyewear
testing matrix metalloproteinase 9 levels. TearLab Corporation (San
full time, even if they do not require optical correction. At Dean
Diego, CA) provided a discount to the study for testing materials for
McGee Eye Institute (DMEI), the preferred practice is to
their TearLab Osmolarity System.
recommend protective eyewear for all functionally monocular
HUMAN SUBJECTS: Human subjects were included in this study. individuals (best-corrected vision of 20/50 with qualifying
The human ethics committees at each participating institution diagnoses and >2 lines of interocular difference in visual acuity)
approved the study. All research complied with the Health Insurance (Tables S1 and S2, available at www.aaojournal.org).
Portability and Accountability (HIPPA) Act of 1996 and adhered to We embarked on large-scale quality improvement (QI) initia-
the tenets of the Declaration of Helsinki. All participants provided
tive for all providers at the DMEI/University of Oklahoma
informed consent.
Department of Ophthalmology over a 7-year period.
No animal subjects were included in this study. The project used an Institutional Review Boardeapproved retro-
spective chart review that adhered to the Declaration of Helsinki and
Author Contributions:
all federal and state laws. The Institutional Review Board determined
Conception and design: Maguire, Asbell
that patient and provider consent were not required. Baseline review
Analysis and interpretation: Oydanich, Maguire, Pistilli, Hamrah,
Greiner, Lin, Asbell examined charts of patients seen over a 12-month period (July 1,
Data collection: Oydanich, Maguire, Pistilli, Hamrah, Greiner, Lin, 2011eJune 30, 2012) to quantify the percentage of functionally
Asbell monocular new patients who were prescribed impact-resistant lenses
Obtained funding: Maguire, Asbell (IRLs) within 90 days of presentation. After this, all providers
Overall responsibility: Oydanich, Maguire, Pistilli, Hamrah, Greiner, completed a survey regarding barriers to prescription of IRLs, and
Lin, Asbell staged interventions were implemented to improve compliance. The
interventions were made on the first day of each data-collection
Correspondence:
Maureen G. Maguire, PhD, Department of Ophthalmology, Uni- period. Each individual patient was observed for 90 days. The first
versity of Pennsylvania, 3535 Market Street, Philadelphia, PA set of interventions included reminder pop-up windows within the
19104. E-mail: [email protected]. electronic medical record (EMR) indicating a patient’s monocular
status and stickers placed at work stations reminding providers to
References recommend IRLs. Analysis of patients seen over a subsequent 5-
month period (November 1, 2015eMarch 31, 2016) was performed
using updated International Classification of Diseases, 10th revision
1. The Dry Eye Assessment and Management Study Research codes that corresponded to the International Classification of Dis-
Group. n-3 fatty acid supplementation for treatment of dry eye
eases, 9th revision codes from the baseline data. A second inter-
disease. N Engl J Med. 2018;378(18):1681e1690.
2. Pult AH, Nichols JJ. A review of meibography. Optom Vis Sci. ventionddefaulting lens material to polycarbonate for certain
2012;89(5):760e769. providersdwas then implemented. After this intervention, analysis of
3. Szczesna-Iskander DH. Measurement variability of the TearLab new patients seen over another 2-month period (September 15,
Osmolarity System. Cont Lens Anterior Eye. 2016;39:353e358. 2016eNovember 15, 2017) was performed. A third interventiond
4. Amparo F, Jin Y, Hamrah P, et al. What is the value of defaulting lens material to polycarbonate material for all providersd
incorporating tear osmolarity measurement in assessing patient was then implemented, followed by analysis of patients seen over a
response to therapy in dry eye disease? Am J Ophthalmol. 6-month period (April 1, 2018eSeptember 30, 2018).
2014;157(1):69e77.e2. Baseline data showed that 169 of 374 qualifying patients
5. Chinnery HR, Naranjo Golborne C, Downie LE. Omega-3 sup- (45.2%) received a prescription for IRLs. Provider survey results
plementation is neuroprotective to corneal nerves in dry eye dis-
regarding barriers to IRL prescription indicated that 45% of pro-
ease: a pilot study. Ophthalmic Physiol Opt. 2017;37(4):473e481.
6. Deinema LA, Vingrys AJ, Wong CY, et al. A randomized, viders attributed forgetfulness as the main reason for lack of pre-
double-masked, placebo-controlled clinical trial of two forms of scription; 24% do not routinely prescribe glasses in their clinic;
omega-3 supplements for treating dry eye disease. Ophthal- 15% refer patients for refractions; and 39% did not prescribe IRLs
mology. 2017;124(1):43e52. because of the added cost, patients’ refusal to wear glasses, or
7. Epitropoulos AT, Donnenfeld ED, Shah ZA, et al. Effect of oral dislike of polycarbonate material.
re-esterified omega-3 nutritional supplementation on dry eyes. Baseline compliance rates among different subspecialties were
Cornea. 2016;35(9):1185e1191. glaucoma 0% (4 physicians), retina 2.4% (5 physicians), oculo-
plastics 5.7% (2 physicians), comprehensive 16.7% (5 physicians),
neuro-ophthalmology 33.3% (3 physicians), resident clinic 40%,
Computer-Based Quality optometry 89.8% (5 providers), and pediatrics 92.9% (3 physicians).
Improvement for Management After the first intervention (pop-up reminders and stickers at
provider terminals), data from patient charts from November 1,
of Functionally Monocular 2015, to March 31, 2016, showed 63 of 113 eligible patients
Patients (55.8%) received a prescription for IRLs (P ¼ 0.049, chi-square
There are an estimated 2.4 million eye injuries in the United States test, compared with baseline).
annually, approximately 90% of which can be prevented by the use After the second intervention (defaulting selected providers to
of protective eyewear.1 The American Academy of Ophthalmology polycarbonate lens material), 13 of 24 patients (54.1%) seen by the

138
 Ophthalmology Volume 127, Number 1, January 2020
discounted equipment, customized software, and user support for the Preferred Practice Patterns for Amblyopia2 recommends that
keratograph used in the study. RPS Diagnostics, Inc. (Sarasota, FL), individuals with unilateral vision impairment with vision of
provided InflammaDry Detector test kits to the clinical centers for
20/50 in the amblyopic eye should wear protective eyewear
testing matrix metalloproteinase 9 levels. TearLab Corporation (San
full time, even if they do not require optical correction. At Dean
Diego, CA) provided a discount to the study for testing materials for
McGee Eye Institute (DMEI), the preferred practice is to
their TearLab Osmolarity System.
recommend protective eyewear for all functionally monocular
HUMAN SUBJECTS: Human subjects were included in this study. individuals (best-corrected vision of
20/50 with qualifying
The human ethics committees at each participating institution diagnoses and >2 lines of interocular difference in visual acuity)
approved the study. All research complied with the Health Insurance (Tables S1 and S2, available at www.aaojournal.org).
Portability and Accountability (HIPPA) Act of 1996 and adhered to We embarked on large-scale quality improvement (QI) initia-
the tenets of the Declaration of Helsinki. All participants provided
tive for all providers at the DMEI/University of Oklahoma
informed consent.
Department of Ophthalmology over a 7-year period.
No animal subjects were included in this study. The project used an Institutional Review Boardeapproved retro-
spective chart review that adhered to the Declaration of Helsinki and
Author Contributions:
all federal and state laws. The Institutional Review Board determined
Conception and design: Maguire, Asbell
that patient and provider consent were not required. Baseline review
Analysis and interpretation: Oydanich, Maguire, Pistilli, Hamrah,
Greiner, Lin, Asbell examined charts of patients seen over a 12-month period (July 1,
Data collection: Oydanich, Maguire, Pistilli, Hamrah, Greiner, Lin, 2011eJune 30, 2012) to quantify the percentage of functionally
Asbell monocular new patients who were prescribed impact-resistant lenses
Obtained funding: Maguire, Asbell (IRLs) within 90 days of presentation. After this, all providers
Overall responsibility: Oydanich, Maguire, Pistilli, Hamrah, Greiner, completed a survey regarding barriers to prescription of IRLs, and
Lin, Asbell staged interventions were implemented to improve compliance. The
interventions were made on the first day of each data-collection
Correspondence:
Maureen G. Maguire, PhD, Department of Ophthalmology, Uni- period. Each individual patient was observed for 90 days. The first
versity of Pennsylvania, 3535 Market Street, Philadelphia, PA set of interventions included reminder pop-up windows within the
19104. E-mail: [email protected]. electronic medical record (EMR) indicating a patient’s monocular
status and stickers placed at work stations reminding providers to
References recommend IRLs. Analysis of patients seen over a subsequent 5-
month period (November 1, 2015eMarch 31, 2016) was performed
using updated International Classification of Diseases, 10th revision
1. The Dry Eye Assessment and Management Study Research codes that corresponded to the International Classification of Dis-
Group. n-3 fatty acid supplementation for treatment of dry eye
eases, 9th revision codes from the baseline data. A second inter-
disease. N Engl J Med. 2018;378(18):1681e1690.
2. Pult AH, Nichols JJ. A review of meibography. Optom Vis Sci. ventionddefaulting lens material to polycarbonate for certain
2012;89(5):760e769. providersdwas then implemented. After this intervention, analysis of
3. Szczesna-Iskander DH. Measurement variability of the TearLab new patients seen over another 2-month period (September 15,
Osmolarity System. Cont Lens Anterior Eye. 2016;39:353e358. 2016eNovember 15, 2017) was performed. A third interventiond
4. Amparo F, Jin Y, Hamrah P, et al. What is the value of defaulting lens material to polycarbonate material for all providersd
incorporating tear osmolarity measurement in assessing patient was then implemented, followed by analysis of patients seen over a
response to therapy in dry eye disease? Am J Ophthalmol. 6-month period (April 1, 2018eSeptember 30, 2018).
2014;157(1):69e77.e2. Baseline data showed that 169 of 374 qualifying patients
5. Chinnery HR, Naranjo Golborne C, Downie LE. Omega-3 sup- (45.2%) received a prescription for IRLs. Provider survey results
plementation is neuroprotective to corneal nerves in dry eye dis-
regarding barriers to IRL prescription indicated that 45% of pro-
ease: a pilot study. Ophthalmic Physiol Opt. 2017;37(4):473e481.
6. Deinema LA, Vingrys AJ, Wong CY, et al. A randomized, viders attributed forgetfulness as the main reason for lack of pre-
double-masked, placebo-controlled clinical trial of two forms of scription; 24% do not routinely prescribe glasses in their clinic;
omega-3 supplements for treating dry eye disease. Ophthal- 15% refer patients for refractions; and 39% did not prescribe IRLs
mology. 2017;124(1):43e52. because of the added cost, patients’ refusal to wear glasses, or
7. Epitropoulos AT, Donnenfeld ED, Shah ZA, et al. Effect of oral dislike of polycarbonate material.
re-esterified omega-3 nutritional supplementation on dry eyes. Baseline compliance rates among different subspecialties were
Cornea. 2016;35(9):1185e1191. glaucoma 0% (4 physicians), retina 2.4% (5 physicians), oculo-
plastics 5.7% (2 physicians), comprehensive 16.7% (5 physicians),
neuro-ophthalmology 33.3% (3 physicians), resident clinic 40%,
Computer-Based Quality optometry 89.8% (5 providers), and pediatrics 92.9% (3 physicians).
Improvement for Management After the first intervention (pop-up reminders and stickers at
provider terminals), data from patient charts from November 1,
of Functionally Monocular 2015, to March 31, 2016, showed 63 of 113 eligible patients
Patients (55.8%) received a prescription for IRLs (P ¼ 0.049, chi-square
There are an estimated 2.4 million eye injuries in the United States test, compared with baseline).
annually, approximately 90% of which can be prevented by the use After the second intervention (defaulting selected providers to
of protective eyewear.1 The American Academy of Ophthalmology polycarbonate lens material), 13 of 24 patients (54.1%) seen by the

138
 Reports

providers who had been defaulted to polycarbonate lens material providers’ concerns, patient concerns were not a factor in this
received a prescription for IRLs, in contrast to 4 of 15 functionally study. As of March 2019, no IRLs had been returned to our optical
monocular patients (26.6%) seen by providers without the default. shop, and there were no complaints from patients who purchased
Statistical significance was not reached because of the small polycarbonate lenses.
sample size. However, given the large difference between the 2 In conclusion, implementing computer-based decision support
groups, we moved quickly to default all providers to IRLs. After in a large-scale QI study improved overall physician compliance
this intervention, 56 of 81 functionally monocular patients (69.1%) for prescription of IRLs in functionally monocular patients.
received a prescription for IRLs (P ¼ 0.09, chi-square test,
compared with baseline). KYLE A. ROGERS, MD
Adherence of a multispecialty group to the Preferred Practice ANDREW A. WILSON, MD
Pattern2 that all individuals who are visually impaired in 1 eye MARK A. WYCKOFF
wear protective eyewear full-time has not been reported to date. R. MICHAEL SIATKOWSKI, MD
This project established a baseline level of compliance among all Dean McGee Eye Institute, University of Oklahoma Department of
providers and instituted interventions to improve the level of Ophthalmology, Oklahoma City, Oklahoma
compliance.
Presented at: The Association of University Professors of Ophthal-
Changes within the EMR reminding providers of a patient’s mology Annual Meeting, Educating the Educators, January 23, 2019,
functionally monocular status improved the level of compliance Fort Lauderdale, Florida.
from 45.2% to 55.8%. After the second intervention, compared
with baseline, there was an improvement in compliance rate in the Financial Disclosure(s): The author(s) have no proprietary or com-
defaulted provider group (54.1%), but there was a decrease among mercial interest in any materials discussed in this article.
Supported in part by an unrestricted grant from Research to Prevent
the provider group, which had not been defaulted to polycarbonate
Blindness, Inc., New York, New York.
material (26.6%). On the basis of data from the second interven-
tion, the decision was made to default all providers to IRL material. HUMAN SUBJECTS: The IRB/ethics committee at the University of
However, providers had the ability to change the lens material for Oklahoma approved this study. All research adhered to the tenets of the
each individual patient by unclicking the polycarbonate box in the Declaration of Helsinki.
EMR.
No animal subjects were used in this study.
Following the universal default, 69.1% of functionally
monocular patients received a prescription for IRLs, a 23.9 Author Contributions:
percentage-point absolute increase and a 152.8% relative Conception and design: Rogers, Wilson, Wyckoff, Siatkowski
improvement from the baseline of 45.2%. Data collection: Rogers, Wilson, Wyckoff, Siatkowski
Providers indicated forgetfulness as the most common reason Analysis and interpretation: Rogers, Wilson, Wyckoff, Siatkowski
Obtained funding: N/A
for not prescribing IRLs. This may be more common among
Overall responsibility: Rogers, Wilson, Wyckoff, Siatkowski
subspecialty services that do not routinely perform refractions such
as glaucoma, retina, and oculoplastics, which had the lowest rates Correspondence:
of compliance in the baseline data. Some providers also noted cost Kyle A. Rogers, MD, Dean McGee Eye Institute, 608 Stanton L. Young
as a limitation to prescription of IRLs. At DMEI, polycarbonate Boulevard, Oklahoma City, OK 73104. E-mail: [email protected].
lens material costs $70 more than the base material (CR39). Online
budget glasses distributers such as Zenni (San Francisco, CA) have
significantly lower upcharges for polycarbonate lenses; addition- References
ally, the base material of Warby Parker (New York, NY) glasses is
polycarbonate. 1. Eye Safety at Home j Prevent Blindness. Preventblindness.org.
Another possible option is recommending plano polycarbonate http://www.preventblindness.org/eye-safety-home. Published
lenses from stores such as Home Depot (Atlanta, GA), where such 2019. Accessed February 24, 2019.
lenses are available for $5 or less. Additionally, targeted refraction 2. Wallace D, Repka M, Lee K, et al. Amblyopia Preferred
clinics could be scheduled to serve these patients. Despite some Practice PatternÒ. Ophthalmology. 2018;125:P105eP142.

Corrigendum

The article entitled, “Nonsyndromic Retinitis Pigmentosa in the Ashkenazi Jewish Population: Genetic
and Clinical Aspects” (Ophthalmology. 2018;125:725e734), appeared in the May 2018 issue of
Ophthalmology. The authors noted that the following should have appeared in the financial disclosure:
This research was supported by the Legacy Heritage Biomedical Science Partnership Program of the
Israel Science Foundation (grant no. 1037/15).

139
 A Simple Vista en Este Número

Oftalmología ®

Aclaración: La American Academy of Ophthalmology no se responsabiliza de ningún error,

omisión u otra posible falla en la traducción.
Número de Enero 2020
Translation Editor: Dr. J. Fernando Arevalo

Comparación de la Eficacia de los

Investigación de las Asociaciones Entre el
Dispositivos de Cirugía Microinvasiva de
Grosor de la Retina y la Función
Glaucoma
Cognoscitiva
Ahmed y otros (p. 52) realizaron un estudio clínico aleatorio
En una encuesta poblacional de corte transversal, Ito y otros
multicentro para comparar la eficacia de dos dispositivos micro-
(p. 107) se propusieron determinar la asociación entre el grosor de
invasivos para cirugía de glaucoma (MIGS) d El Microstent
la retina y la función cognoscitiva. Un total de 1293 japoneses
Hydrus y el iStent del Sistema de Stent de Micro-Bypass Trabe-
entre los 65 y los 86 años fueron sometidos a evaluaciones oftál-
cular d para reducir la presión intraocular (IOP) y los medi-
micas completas así como a exámenes detallados para determinar
camentos para glaucoma de ángulo abierto. Se distribuyó al azar un
su salud mental a fin de asignarlos a tres grupos: normales, con
total de 152 ojos de 152 pacientes en una relación de 1:1 a recibir el
leves limitaciones cognoscitivas, o con demencia. Se encontraron
MIGS sólo, consistente en ya fuera 1 iStent Microstent Hydrus o 2
diferencias significativas en los tres grupos en el grosor de todas las
dispositivos iStent, entre marzo de 2013 y mayo de 2015. En los
capas de la retina y en el complejo de células ganglionares (GCC),
148 pacientes que completaron el seguimiento de doce meses, el
aunque no en el grosor de la capa de fibras nerviosas de la retina
Hydrus mostró la tasa más alta de éxito quirúrgico y un menor uso
peripapilar (ppRNFL). El grosor del complejo de células ganglio-
de medicamentos en comparación con el procedimiento de 2
nares y de la mácula se asoció con la presencia de demencia, pero
dispositivos iStent, y ambos dispositivos demostraron perfiles de
no así el grosor de la ppRNFL. La capa plexiforme interna de
seguridad similares. El estudio sugiere que los dispositivos MIGS
células ganglionares, el GCC y el grosor completo de la mácula se
podrían ayudar a manejar la IOP y que del Microstent Hydrus
asociaron con la presencia de demencia en los sectores inferiores.
ofrece algunas ventajas sobre los 2 dispositivos iStent de Bypass
Los resultados indican que el grosor de la mácula, y no así la
Trabecular.
ppRNFL, se asoció con la presencia de demencia y los autores
sugieren que las mediciones de OCT de la mácula podrían ser
El Brolucizumab vs el Aflibercept para el superiores a las de la ppRNFL para evaluar los cambios
Tratamiento de la Degeneración Macular neurodegenerativos.
Neovascular Relacionada con la Edad
En dos estudios fase 3 multicentro doble ciegos aleatorios de Evaluación de la Precisión Diagnóstica de
diseño similar (HAWK y HARRIER), Dugel y otros (p. 72) los Servicios de Atención de Salud Ocular
compararon el brolucizumab, un fragmento antibiótico de cadena Basados en Tecnología
única que inhibe el factor de crecimiento del endotelio vascular e Maa y otros (p. 38) realizaron una comparación prospectiva entre
A con el aflibercept para tratar la degeneración macular neo- el protocolo estándar de los Servicios de Salud Ocular Basados en
vascular relacionada con la edad. Un total de 1817 pacientes se Tecnología del Sistema (TECS) de Atención de Veteranos y los
distribuyeron aleatoriamente a recibir 3 mg de brolucizumab exámenes cara a cara (FTF) para identificar aspectos del proceso que
intravítreo (el estudio HAWK únicamente) o 6 mg o 2 mg de pudieran refinarse para mejorar la precisión. En el período com-
aflibercept. En ambos estudios, cada rama debrolucizumab prendido entre marzo de 2015 a diciembre de 2017 se reclutó un total
demostró no ser inferior al aflibercept en cuanto a agudeza visual de 256 pacientes sin historia conocida de enfermedades oculares
mejor corregida, cambio a partir de la línea de base hasta la significativas. Los investigadores encontraron que los resultados del
semana 48. Más de 50% de los ojos tratados con 6 mg de bro- TECS mostraban una concordancia sustancial para cataratas y reti-
lucizumab se mantuvieron en una dosificación de cada 12 semanas nopatía diabética, una concordancia de moderada a sustancial para
hasta la semana 48. En la evaluación equiparable de la semana 16, glaucoma/sospecha de glaucoma, y una concordancia moderada para
menos ojos tratados con 6 mg de brolucizumab presentaban degeneración macular relacionada con la edad en comparación con
enfermedad activa en comparación con el aflibercept en ambos un examen FTF. El protocolo de los TECS mostró un alto porcentaje
estudios. Los resultados de líquido anatómico retiniano favo- de concordancias y buena sensibilidad y una excelente especificidad
recieron el brolucizumab en comparación con el aflibercept, y la en comparación con los exámenes oculares FTF para la detección de
seguridad global con el brolucizumab fue similar a la del afli- enfermedades oculares comunes. Los autores concluyen que el
bercept. Los autores concluyen que el brolucizumab no fue inferior protocolo TECS estándar es comparable a un examen FTF para la
al aflibercept en cuanto a la función visual para la semana 48 y detección de enfermedades oculares comunes y sus capacidades de
sugieren que éste podría servir como opción de tratamiento alterno diagnóstico podrían mejorarse con la inclusión de pruebas oftálmicas
al aflibercept. adicionales más sofisticadas como la OCT.
 cicatrización corneal. La medida conservadora de la cicatri-
Factor de Crecimiento Nervioso Humano zación corneal mostró además una significancia estadística para
Recombinante Tópico para Queratopatía la semana 4. Los pacientes tratados con cenegermin mostraron
Neurotrófica una significativa reducción en el tamaño de las lesiones en
comparación con los pacientes tratados con el vehículo, y un
Pflugfelder y otros (p. 14) realizaron un estudio doble ciego, número menor de pacientes tratados con cenegermin expe-
aleatorio, multicentro, controlado con vehículo para evaluar la rimentó progresión de la enfermedad durante el curso del
seguridad y eficacia del factor de crecimiento nervioso humano tratamiento de 8 semanas en comparación con los pacientes
recombinante tópico (cenegermin) en pacientes con queratopatía tratados con el vehículo. Los investigadores encontraron que el
neurotrófica. Entre el 1 de mayo de 2015 y el 8 de diciembre de cenegermin fue bien tolerado y que sus efectos adversos fueron
2015, se inscribieron 48 pacientes y de distribuyeron en forma en la mayoría locales, leves y transitorios. Concluyen que el
aleatoria, en una relación de 1:1 a recibir ya fuera 20 mg/ml de tratamiento con cenegermin mostró tasas más altas de
cenegermin o vehículo de gotas oftálmicas. El tratamiento con cicatrización corneal que el vehículo en los pacientes estudiados
cenegermin mostró tasas más altas de cicatrización corneal con y que potencialmente podría servir como tratamiento seguro no
diferencias estadísticamente significativas entre los grupos de invasivo para la queratopatía neurotrófica.
tratamiento, para la semana 8, tanto por evaluaciones
convencionales como por evaluaciones conservadoras de la Sandeep Ravindran, PhD
 期刊一览

Ophthalmology ®

免责声明: 美国眼科学会对翻译中出现的任何错误、
遗漏或其他可能的缺陷不承担任何责任.
Volume 127, Number 1, January 2020

[GCC]）厚度，但未见于视盘周围视网膜神经纤维层（peri-
微创青光眼手术设备疗效对比 papillary retinal nerve fiber layer [ppRNFL]）厚度。节细胞复合
体和黄斑厚度与痴呆相关，但ppRNFL厚度则不然。下方象限
Ahmed (p. 52) 等进行了一项前瞻性多中心随机临床试验，以
节细胞内丛状层、GCC及全层黄斑厚度与痴呆相关。结果提
比较两种微创青光眼手术（microinvasive glaucoma surgery
示黄斑厚度，非ppRNFL，与痴呆相关，作者指出OCT黄斑区
[MIGS]）设备ddHydrus微型支架和iStent小梁网微导流支架
指标在评估神经退行性变化时可能优于ppRNFL指标。
系统dd对于开角型青光眼降低眼压（intraocular pressure
[IOP]）和减少药物使用的效果。共计152名患者的152例患眼
于2013年三月至2015年五月被1:1随机分配至单纯MIGS治疗 基于技术的眼保健服务诊断准确性评估
组，分别接受1 Hydrus微型支架或2 iStent设备植入。在完成 Maa (p. 38) 等进行了一项前瞻性对照研究，通过比较军人事
12个月随访的148名受试者中，与2-iStent手术相比，Hydrus手 务医疗保健系统基于科技的眼保健服务（Technology-based
术成功率较高，使用药物较少，两种设备安全性类似。该研 Eye Care Services [TECS]）和面对面（face-to-face [FTF]）检
究提示，MIGS设备可有效控制IOP，同时证明Hydrus微型支 查识别可优化的流程部分以提高准确性。自2015年三月至
架某些特征优于2-iStent小梁网微导流支架设备。 2017年十二月入选256位非合并重大眼疾的患者。研究者发现
与FTF检查相比，TECS读片员识别白内障和糖尿病性视网膜
溴珠单抗对比阿柏西普治疗新生血管性年龄相关 病变一致性高，青光眼/可疑青光眼一致性中等，年龄相关性
黄斑变性一致性中等。就常见眼病而言，TECS方案与FTF检
性黄斑变性
查相比一致性高、敏感度良好、特异性佳。作者总结标准
在两项设计相类似的随机、双盲、多中心3期临床试验 TECS方案与FTF检查对于常见眼病的甄别具有可比性，诊断
（HAWK和HARRIER）中， Dugel (p. 72) 等就一种抑制血管 能力可通过纳入精良的眼科检查如OCT以提高。
内皮生长因子-A的单链抗体片段溴珠单抗以及阿柏西普治疗
新生血管性年龄相关性黄斑变性进行比较。共计1817名患者
被随机分配至玻璃体注射溴珠单抗3 mg组（仅HAWK研究）
表面重组人神经生长因子治疗神经营养性角膜
或6 mg组，或阿柏西普2 mg组。两项试验中，自基线至48 病变
周，溴珠单抗治疗组最佳矫正视力变化均非劣于阿柏西普治 Pflugfelder (p. 14) 等进行了一项多中心、随机、双盲、空白对
疗组。在48周期间，超过50%溴珠单抗6 mg治疗眼维持在每 照试验以评估表面重组人神经生长因子（cenegermin）在神经
12周给药。在匹配的16周评估中，两项研究中溴珠单抗6 mg 营养性角膜病变患者中的应用。2015年5月1日至2015年12月8
治疗眼较阿柏西普治疗眼少出现病变活动。解剖学视网膜积 日期间，入选48名受试者并将其1:1随机分配至cenegermin 20
液指标显示溴珠单抗优于阿柏西普，总体安全性而言溴珠单 mg/ml或空白滴眼液用药组。Cenegermin治疗角膜愈合率高，
抗和阿柏西普相当。作者总结溴珠单抗治疗48周时视功能非 治疗8周时角膜愈合情况以传统方法或保守评估法在用药组间
劣于阿柏西普，可作为阿柏西普的替代治疗。 均显示出统计学显著差异。治疗4周时角膜愈合情况以保守方
法评估亦显示出统计学显著性。与空白用药组患者相比，
探究视网膜厚度和意识功能的相关性 Cenegermin治疗组患者显示病灶范围明显减小；8周治疗过程
中，较少cenegermin治疗患者出现病情进展。研究者发现cen-
在一项横断面、人群为基础的调查中，Ito (p. 107) 等就视网
egermin耐受度良好，不良反应多为局部、轻微且一过性。他
膜厚度和意识功能的相关性进行研究。对1293名65至86岁的
们总结cenegermin治疗神经营养性角膜炎角膜愈合率高，有望
日本人进行全面眼科评估和详细精神健康检查，以将其分为
作为安全且非侵入的治疗选择。
以下三组：正常、轻度意识障碍、痴呆。显著差异见于三组
中视网膜各层和节细胞复合体（ganglion cell complex Sandeep Ravindran, PhD
 Correspondence

Re: Heindl et al.: The eyes of Oetzi: FRANCESCO M. GALASSI, MD1,2

the Tyrolean iceman mummy ELENA VAROTTO, MA, PGD2,3
(Ophthalmology. 2019;126:530) ANDREAS G. NERLICH, MD, PHD4
TO THE EDITOR: We read the Pictures & Perspectives by Heindl DONG HOON SHIN, MD, PHD5,6
et al1 regarding the eyes of Oetzi, the celebrated Tyrolean Iceman RAFFAELLA BIANUCCI, PHD7,8,9
dating back to the 4th millennium BCE. Building on previous 1
Archaeology, College of Humanities, Arts and Social Sciences,
studies, the authors mention a “unique view” onto this ancient Flinders University, Adelaide, Australia; 2FAPAB Research Center,
corpse’s anatomy, thanks to the “natural mummification in the Sicily, Italy; 3Department of Humanities (DISUM), Catania CT, Sicily,
glacier ice,”1 which has allowed many important discoveries, Italy; 4Institute of Pathology, Academic Clinic Munich-Bogenhausen,
including the determination of this prehistoric individual’s violent Munich, Germany; 5Laboratory of Bioanthropology, Paleopathology
and History of Diseases, Seoul National University College of Medicine,
cause of death. Specifically, addressing the issue of the preservation
Seoul, South Korea; 6Institute of Forensic Science, Seoul National
of ocular structures,2,3 it is interesting to examine the actual role University College of Medicine, Seoul, Korea; 7Legal Medicine Section,
played by environmental and climate conditions as opposed to purely Department of Public Health and Paediatric Sciences, University of
anatomic characteristics of the tissues themselves. Turin, Turin, Piedmont, Italy; 8Warwick Medical School, Biomedical
The available literature shows a consistent presence of ocular Sciences, University of Warwick, Coventry, West Midlands, United Kingdom;
structures in mummified remains,2 either natural or artificial.3-6 9
UMR 7268, Laboratoire d’Anthropologie bio-culturelle, Droit, Etique &
Twenty-one of 52 Third Intermediate Period (1064e656 BCE) to Santé (Adés), Faculté de Médecine, Provence-Alpes-Côte d’Azur, France
Graeco-Roman (332 BCEe395 CE) period in Egyptian mummies7
revealed preservation of ocular structures exclusively through Financial Disclosures: The authors have no proprietary or commercial
desiccation in 40.4%.4 Similarly, the globes of the eyes with interest in any materials discussed in this article.
optic nerves and orbital muscles could be identified in the
Available online: November 4, 2019.
mummified remains of 2 remarkable 18th Dynasty Egyptian
dignitaries: Nebiri and the Royal Architect Kha.8 Correspondence:
The Korean Mummy Project, assessing 7 computed tomography Francesco M. Galassi, MD, Archaeology, College of Humanities, Arts
scanned mummies, indicates eye structures were likely to have and Social Sciences, Flinders University, Humanities Building, GPO
preserved in 4 of 7 (57.1 %) of naturally mummified Joseon Dy- Box 2100, Adelaide, SA 5001, Australia. E-mail: francescom.galassi@
nasty individuals from the 17the19th centuries (Shin DH, unpub- flinders.edu.au.
lished data), which approximately equals the number of preserved
eye structures in 18th/19th century German crypt mummies (5/10 References
[50%] individuals) [Nerlich AG, personal communication].
In a series of 23 naturally and 23 artificially mummified 1. Heindl LM, Pernter P, Zink AR, Panzer S. The eyes of Oetzi: the
Capuchin mummies from Palermo, Sicily, Panzer et al5 reported a Tyrolean Iceman mummy. Ophthalmology. 2019;126:530.
higher prevalence for the preservation of the bulbs and optic nerves 2. Holz HA, Lloyd III WC, Mannis MJ, Aufderheide AC. Histo-
bilaterally, namely in 21 of 23 mummies (6 natural and 17 artificial pathologic findings in naturally preserved mummified human
mummies), yielding a 91.3% figure. eyes. Arch Ophthalmol. 2007;125:978e981.
Apparently, this preservative phenomenon has little to do with 3. Esteban J, Cases-Mérida S, Tortosa M, et al. Histopathological
climate since ocular structures are a common feature of both cold study of a mummified eye and optic nerve from a strangled
(Alpine glaciers1), hot and sultry (Atacama desert of Northern Peruvian mummy Lamabyeque (900e1,200 AD). Pathobiology.
2015;82:90e93.
Chile,2 Peru,3 Northern African desert4), hot summer subtropical
4. Loynes RD. Prepared for eternity. A study of human embalming
Mediterranean climate (Sicily5), and humid continental regions of techniques in Ancient Egypt using computerized tomography
the world (South Korea, Germany). scans of mummies. Oxford: Archeopress; 2015:66e67.
In Oetzi’s case, as in all others, the most likely reason for the 5. Panzer S, Augat P, Zink AR, Piombino-Mascali D. CT checklist
preservation of intact ocular globes is to be searched in the eye’s and scoring system for the assessment of soft tissue preservation
anatomy. The sclera is a tough connective tissue extremely resistant to in human mummies: application to catacomb mummies from
destruction. Scleral collagen is, in composition and arrangement, Palermo, Sicily. Int J Paleopathol. 2018;20:50e59.
more similar to that seen in skin, with wider fibrils and a much more 6. Meek K. The cornea and sclera. In: Fratzl P, ed. Collagen.
interwoven structure than the cornea.6 Dehydration leads to a collapse Boston: Springer; 2008:359e396.
of the vitreous body and a retraction of the sclera, resulting in a 7. Ikram S. Death and Burial in Ancient Egypt. Longman; 2002:
216e221.
shrunken but compact eye globe. As to the optic nerve, the narrow
8. Loynes RD, Charlier P, Froesch P, et al. Virtopsy shows a high status
and compact optic canal and the myelin sheath, whose dry mass is funerary treatment in an early 18th Dynasty non-royal individual.
characterized by a high proportion of lipids9 act as a protective shield. Forensic Science, Medicine and Pathology. 2017;3:302e311.
In summary, the preservation of eye structures in mummies is 9. Morelland P, Quarles H. Characteristic composition of myelin.
independent of region, time period and cause of mummification, but In: Siegel GJ, Agranoff BW, Albers RW, et al., eds. Philadel-
depends on the overall preservation of a mummified human body. phia: Lippincott-Raven; 1999.

e4
 Correspondence

REPLY: We thank Galassi et al for their comments on University Hospital Cologne, Cologne, Germany; 4European Academy
our study.1 We are aware of the fact that the preservation of Bolzano (EURAC)-Institute for Mummy Studies, Bolzano, Italy
of ocular structures in spontaneous and anthropogenic
mummies is well known in the paleopathological literature. From Financial Disclosures: The authors have no proprietary or commercial
interest in any materials discussed in this article.
the radiologist’s point of view, the first report of the application
of computed tomography (CT) to the study of ancient mummies Available online: November 4, 2019.
by Harwood-Nash2 in 1979 has to be mentioned. Already in this
article, the optic nerve and strap muscles were described to be Correspondence:
extraordinarily well preserved in an Ancient Egyptian Theban Ludwig M. Heindl, MD, Department of Ophthalmology, University of
mummy of the mid-9th century BCE. Cologne, Kerpener Strasse 62, 50937 Cologne, Germany. E-mail:
We fully agree that the preservation of eye structures occurs in [email protected].
mummies from various regions, time periods, and differing mummi-
fication types. In our findings, however, we encountered differences in
the morphology of preserved intraorbital structures, such as the eye References
globe, optic nerve, eye muscles, and the fat pads on CT images be-
tween mummies from hot regions (e.g., Egypt, Sicily) and the glacier 1. Heindl LM, Pernter P, Zink AR, Panzer S. The eyes of Oetzi: the
mummy Oetzi. In ancient Egyptian mummies and the mummies from Tyrolean iceman mummy. Ophthalmology. 2019;126:530.
Palermo, Sicily, the intraorbital structures are usually visible as clearly 2. Harwood-Nash DC. Computed tomography of ancient Egyptian
shrunken relatively dense structures whereby the rest of the intra- mummies. J Comput Assist Tomogr. 1979;3:768e773.
orbital space is filled with air. In the Tyrolean Iceman, the eye globes 3. Previgliano CH, Ceruti C, Reinhard J, et al. Radiologic evalu-
are shrunken and retracted.1 The majority of the orbit, however, is full ation of the Llullaillaco mummies. AJR Am J Roentgenol.
of soft tissue containing eye muscles, optic nerve, and preserved fat 2003;181:1473e1479.
4. Panzer S, Augat P, Zink AR, Piombino-Mascali D. CT checklist
pads between these structures.1 Only a thin layer of air surrounded
and scoring system for the assessment of soft tissue preservation
the mass of preserved soft tissues and the optic nerves.1 A similar in human mummies: application to catacomb mummies from
CT appearance of preserved eyes, optic nerve, and orbital muscles Palermo, Sicily. Int J Paleopathol. 2018;20:50e589.
was shown in 1 of the 3 Incan children who were mummified by
freezing on the summit of Mount Llullaillaco in the Argentinean
Andes.3 These children were sacrificed 500 years ago.3
Moreover, we highlighted the preservation of the optic chiasm Re: Lambert et al.: Intraocular lens
in the Tyrolean Iceman.1 It is clearly discernible in its anatomic implantation during early
location at the suprasellar region.1 Analysis of the Sicilian
childhood: A Report by the
mummies from Palermo revealed no preservation of the optic
chiasm in the 6 spontaneous mummies and only one case with American Academy of
clear preservation in the anthropogenic mummies (own data of Ophthalmology
the collections published4). Because the intraorbital structures (Ophthalmology. 2019;126:1454-1461)
were present in almost all of the Palermo mummies, but the optic
chiasm only in one, there seem to be varying conditions for the TO THE EDITOR: The article by Lambert et al1 published as a report
preservation of intraorbital and intracranial soft tissues. In Oetzi, from the American Academy of Ophthalmology attracted our
the intracranial microenvironment led to remarkable preservation attention. To explore outcomes of intraocular lens (IOL)
of the brain and optic chiasm.1 Thus, we agree with the authors implantation in infants, the best studies in quality have been
that the preservation of eye structures occurs in other mummies, collected and well-reviewed among 222 potential papers, which is
but the way of preservation clearly depends on the region, time a daunting task. Because the academy represents an elite class of
period, and type of mummification. surgeons from very highly ranked centers in the developed world,
Finally, in our contribution to Pictures & Perspectives, we we believe this article misses some important issues before gener-
wanted to illustrate insights into the eyes of Oetzi by means of CT alizing it for the pediatric ophthalmology community across the
images.1 We did not present it as a unique finding, but we referred globe.
it to the mummy in general, as the Iceman is the world oldest The authors selected 14 studies to arrive at their conclusions.
glacier mummy and probably the most heavily studied human We must point out that these studies had a lot of age variability and/
corpse. Despite that, little attention has been given to the or were retrospective. Of the 14 studies, 8 were publications of the
preservation of his eyes. Infant Aphakia Treatment Study (IATS) group only, and hence
were not isolated in the true sense. It is not justifiable to compare
STEPHANIE PANZER, MD1 the visual acuity and opacification rates of cataract surgery in a 3-
PATRIZIA PERNTER, MD2 month-old child and a 2-year-old child. When the threshold for IOL
LUDWIG M. HEINDL, MD3 implantation is becoming as early as 6 months, we need to critically
ALBERT R. ZINK, PHD4 review the relevant literature only.
It is to be emphasized that, from all studies reviewed by the
1
Department of Radiology, Trauma Center Murnau and Institute of
Biomechanics, Trauma Center Murnau and Paracelsus Medical authors, the most unbiased conclusion is that “timing of surgery”
University Salzburg, Murnau, Germany; 2Department of and not the IOL implantation that affects the incidence of glaucoma.
Radiodiagnostics, Central Hospital, Bolzano, Italy; 3Department of Moreover, all these studies keep the adult cutoff for intraocular
Ophthalmology, University of Cologne, Faculty of Medicine and pressure (21 or 25 mmHg). Until we have sufficiently powered and

e5
 Ophthalmology Volume 127, Number 1, January 2020

tailored trials to study glaucoma in infant cataract surgery, our Correspondence:

decision to implant or not implant an IOL should not be based on a Jagat Ram, MS, Professor, Advanced Eye Centre, Post Graduate Insti-
concern for glaucoma alone. tute of Medical Education and Research, Chandigarh, India. E-mail:
[email protected].
Another important point missed in the discussion is the vari-
ability of postoperative target refraction. There is no consensus on
IOL power calculation formula used, biometry methods, or rate of References
undercorrection in IOL power (ranging from 20% to 35% under-
correction or constant þ8 diopters by the IATS2). It is imperative to
1. Lambert SR, Aakalu VK, Hutchinson AK, et al. Intraocular lens
discuss this issue because the visual outcomes, choice of implantation during early childhood: a report by the American
postoperative rehabilitation, and myopic shifts are all interlinked Academy of Ophthalmology. Ophthalmology. 2019;126:
with the target refraction. 1454e1461.
It is curious that, despite numerous publications on IOL im- 2. Plager DA, Lynn MJ, Buckley EG, et al. Infant Aphakia
plantation in infants, studies from underdeveloped countries (where Treatment Study Group. Complications in the first 5 years
the incidence of congenital cataract is maximal) failed to reach the following cataract surgery in infants with and without intraoc-
quality of evidence required to make it to this report. The rates of ular lens implantation in the Infant Aphakia Treatment Study.
visual axis opacification reported in some studies are much less than Am J Ophthalmol. 2014;158:892.
that of the IATS (17% vs 40%).2e4 Even after the IATS published 3. Sukhija J, Ram J, Gupta N, et al. Long-term results after primary
intraocular lens implantation in children operated less than 2
in favor of contact lenses,2 the developing world continues to
years of age for congenital cataract. Indian J Ophthalmol.
support IOL implantation.3e5 There could be additional under- 2014;62:1132e1135.
studied factors responsible and we tried to review the same studies 4. Sukhija J, Kaur S, Ram J. Outcome of primary intraocular
as the authors. We observed that those children who had anterior lens implantation in infants: complications and rates of
vitrectomy after IOL implantation2 had higher opacification rates additional surgery. J Cataract Refract Surg. 2016;42:
than those who had vitrectomy before placing the IOL.3e5 We 1060e1065.
believe the adequacy of vitrectomy cannot be judged efficiently 5. Vasavada AR, Vasavada V, Shah SK, et al. Five-year post-
after IOL implantation and hence the higher rates of opacification. operative outcomes of bilateral aphakia and pseudophakia in
It is not the merits of IOL implantation, but rather the de- children up to 2 years of age: a randomized clinical trial. Am J
merits of the alternatives in developing countries that is affecting Ophthalmol. 2018;193:33e44.
the decision making. Belonging to a very poor uneducated strata
with no insurance, bearing the initial cost of contact lenses, along
with their maintenance and hygiene is out of the question. Un- REPLY: We appreciate the interest of Kaur et al in our
corrected aphakia thus leads to the unintended consequence of Ophthalmic Technology Assessment article. The pur-
poor visual outcome. Placing an IOL gives a partial rehabilitation pose of this assessment was to compare visual outcomes
in a scenario where the patient cannot even afford spectacles. and adverse events for children 2 years of age undergoing cataract
Amblyopia can be averted by constant optical correction and surgery with or without intraocular lens (IOL) implantation. All but
faster visual rehabilitation during the early years of visual 2 of the 14 studies reviewed were randomized clinical trials or
development ensured by an IOL placement in the capsular bag prospective cohort studies. One of the randomized clinical trials we
and “moderate hyperopia” as the postoperative target. The visual reviewed was performed in India (reference 12 in the original
axis opacification can be dealt with through an efficient second article). The 2 retrospective studies cited by Kaur et al were not
operation, which is not a complicated procedure in the hands of appropriate for our review because one only included 13 patients
an experienced surgeon. (our inclusion criteria required 40 patients) and the second did not
Perhaps a concurrent randomized trial from the developing and have a control group (e.g., all of the patients underwent IOL
the developed world would provide more level 1 evidence. The implantation).1,2
Academy members have a crucial responsibility in laying down the We agree with many of the points raised by the authors. First,
practice patterns that the rest of the world follows. We urge them to we agree that age at cataract surgery is the most important risk
consider the involvement of centers across the world into their trials factor for glaucoma following cataract surgery (reference 19 in the
henceforth. Until then, IOL implantation could be a good practical original article). Second, we agree that there are significant barriers
option (even if not the best) in the developing world for very young to contact lens wear for aphakic children in developing countries.
children. One of the randomized clinical trials we cited was performed in
SAVLEEN KAUR, MS India (reference 12 in the original article). In that study, 60 children
underwent bilateral cataract surgery and were randomized to either
VIJAY SHARMA, MS
aphakia or primary IOL implantation. They reported that the parents
JASPREET SUKHIJA, MS
of 13 of the 30 children randomized to aphakia agreed to have their
JAGAT RAM, MS child wear contact lenses, but only 2 continued to wear them for 1
Department of Ophthalmology, Post Graduate Institute of Medical year or longer. Although contact lens adherence is much better in
Education and Research, Chandigarh, India the United States, there remain significant cost barriers that limit
their use, just as there may be cost barriers in other countries. For
Financial Disclosures: The authors have no proprietary or commercial
this reason, the Infant Aphakia Treatment Study (IATS) in-
interest in any materials discussed in this article.
vestigators have advocated that aphakic contact lenses should be a
Available online: November 4, 2019. mandated benefit for medical insurance plans in the United States.3

e6
 Correspondence

The 14 studies reviewed for this Ophthalmic Technology Available online: November 4, 2019.
Assessment did not report improved visual outcomes with IOL
implantation in young children. IATS was designed to test the Correspondence:
Scott R. Lambert, MD, Byers Eye Institute, 2452 Watson Ct, Palo Alto,
hypothesis that the partial optical correction provided by IOLs
CA 94303. E-mail: [email protected].
would result in better visual outcomes in infants undergoing uni-
lateral cataract surgery. However, visual outcomes were similar
regardless of whether an IOL was implanted, and the children who References
underwent IOL implantation had a 3 times higher risk of needing
additional intraocular surgeries (reference 13 in the original article). 1. Sukhija J, Ram J, Gupta N, et al. Long-term results after primary
For this reason, the IATS investigators concluded that, “When intraocular lens implantation in children operated less than 2
operating on an infant younger than 7 months of age with a uni- years of age for congenital cataract. Indian J Ophthalmol.
lateral cataract, we recommend leaving the eye aphakic and 2014;62:1132e1135.
focusing the eye with a contact lens. Primary IOL implantation 2. Sukhija J, Kaur S, Ram J. Outcome of primary intraocular lens
should be reserved for those infants where, in the opinion of the implantation in infants: Complications and rates of additional
surgeon, the cost and handling of a contact lens would be so surgery. J Cataract Refract Surg. 2016;42:1060e1065.
burdensome as to result in significant periods of uncorrected 3. Kruger SJ, Vanderveen DK, Freedman SF, et al. Third-party
aphakia.” Some clinicians have questioned whether these findings coverage for aphakic contact lenses for children. Transl Vis Sci
Technol. 2019;8:41.
are relevant for children living in developing countries. We
acknowledge that there are many differences between pediatric
cataract surgery and the subsequent management of these eyes in
developed and developing countries. First, children undergoing Re: Lambert et al.: Intraocular lens
congenital cataract surgery in developed countries are usually much implantation during early
younger. In the IATS, children underwent congenital cataract sur- childhood: a report by the
gery at a median age of 1.8 months, whereas Vasavada et al re- American Academy of
ported performing congenital cataract surgery and IOL implantation
at a median age of 6.0 months, and Sukhija et al at a mean age of
Ophthalmology
7.1 months (reference 12 in the original article).2 Second, greater (Ophthalmology. 2019;126:1454-1461)
resources are available in developed countries to support TO THE EDITOR: We read the American Academy of Ophthalmol-
successful contact lenses and spectacle wear. Although we agree ogy’s (AAO) report on intraocular lens (IOL) implantation in
with Kaur et al that primary IOL implantation may be an children 2 years of age, which is based on a systematic literature,
excellent choice in older infants, particularly those with unilateral without a meta-analysis, by Lambert et al.1 Based on this review,
cataracts, evidence is lacking to support the superiority of the AAO states that although IOL implantation in children aged
primary IOL implantation in children <6 months of age. The under 6 months is not recommended, the “best available evidence
conclusion of our review states that, “the best available evidence suggests that IOL implantation can be done safely with
suggests that IOL implantation can be done safely with acceptable side effects in children older than 6 months of age.”1
acceptable side effects in children older than 6 months of age.” We disagree with this statement because the best available ev-
This is supported by a recent retrospective study performed in the idence does not support the absence of important adverse outcomes
United States that reported a low adverse event rate (including a in children >6 months of age.
very low rate of glaucoma) in children undergoing cataract The review adopts the internationally accepted framework for
surgery and primary IOL implantation between 7 and 24 months evidence appraisal developed by the Oxford Centre for Evidence-
of age (reference 33 in the original article). based Medicine. However, the authors have applied the 20092
Finally, it is also important to emphasize that congenital cataract version of this guidance rather than the version updated in 2011.
surgery needs to be performed in the first 2 months of life for The authors do not justify this. Furthermore, they have
unilateral cataracts and 3 months for bilateral cataracts in order to misinterpreted the 2009 guidance from Oxford Centre for
obtain the best visual results. If cataract surgery is delayed until Evidence-based Medicine, which states that an inception cohort
after 6 months of age, when IOLs can be implanted with a lower study with >80% follow-up is at an equivalent level to individual
risk of adverse events, the visual results will generally be subop- (i.e., single) randomized controlled trials.2 IoLunder2, the national
timal. study of primary IOL implantation in children aged 2 years, is
an inception cohort study with 96% follow-up.3 Findings from
SCOTT R. LAMBERT, MD1 IoLunder2 and the Infant Aphakia Treatment Study randomized
DEBORAH K. VANDERVEEN, MD2 controlled trial on outcomes for children undergoing surgery for
unilateral cataract at <7 months of age are strikingly similar.3,4
STEPHEN J. KIM, MD3
The updated 2011 guidance Oxford Centre for Evidence-based
1
Department of Ophthalmology, Stanford University School of Medicine restates the equivalence between single trials and incep-
Medicine, Palo Alto, California; 2Division of Ophthalmology, Boston
tion cohort studies with good follow-up, and also states that indi-
Children’s Hospital, Boston, Massachusetts; 3Vanderbilt Eye Institute,
Vanderbilt University School of Medicine, Nashville, Tennessee
vidual inception cohort studies and randomized controlled trials
provide level 2 evidence on treatment harms or benefits.5 Thus, the
Financial Disclosures: The authors made the following disclosures: review by Lambert et al1 has not appropriately weighed the
D.K.V.: Personal fees e Ophtec, during the conduct of the study. literature, undermining the accuracy of the AAO report.

e7
 Ophthalmology Volume 127, Number 1, January 2020

Lambert et al1 also suggest that findings from IoLunder2 are evidence-based-medicine-levels-evidence-march-2009/; 2009.
limited by differences within the cohort of age at surgery, Accessed August 21, 2019.
operative techniques, perioperative medications, or level of surgeon 3. Solebo AL, Cumberland P, Rahi JS. 5-year outcomes after pri-
experience. These and other possible confounders were considered mary intraocular lens implantation in children aged 2 years or
in the gold standard approach of using multivariable analysis to younger with congenital or infantile cataract: findings from the
IoLunder2 prospective inception cohort study. Lancet Child
adjust for confounding. The increased risk of visual axis opacity
Adolesc Health. 2018;2:863e871.
with IOL implantation in children aged <2 years of age is 4. Infant Aphakia Treatment Study Group, Lambert SR,
independent of these factors.3 Thus, taken together, the evidence Lynn MJ, et al. Comparison of contact lens and intraocular
from the work cited within the review,1 specifically IoLunder2, Li lens correction of monocular aphakia during infancy: a ran-
et al and Vasavada et al, suggests that for children aged between 6 domized clinical trial of HOTV optotype acuity at age 4.5
months and 2 years of age, primary pseudophakic carries an years and clinical findings at age 5 years. JAMA Ophthalmol.
increased risk of visual axis opacification (VAO) over aphakia. 2014;132:676e682.
Although Vasavada et al reported similar rates of VAO in the 5. OCEBM Levels of Evidence Working Group. The Oxford 2011
pseudophakia and aphakia group, there was a significantly higher levels of evidence. Available: www.cebm.net/wp-content/uploads/
rate of structural inflammatory sequelae following IOL implantation. 2014/06/CEBM-Levels-of-Evidence-2.1.pdf; 2011. Accessed
August 21, 2019.
We think that it is important that recommendations about clin-
ical practice accurately assess and draw together the best available
evidence. We hope that drawing attention to the flaws in the review
underlying the AAO recommendations will enable clinicians to REPLY: We appreciate the interest of Solebo et al in our
counsel families of affected children appropriately, before surgery, American Academy of Ophthalmology (AAO)
on the increased risk of reoperation with primary IOL implantation, Ophthalmic Technology Assessment (OTA) article. Sol-
independent of age, for all children aged <2 years old at cataract ebo et al are correct that OTAs are not meta-analyses; rather they
surgery. are systematic reviews. A systematic review, as described by the
AMEENAT LOLA SOLEBO, PHD, FRCOPHTH Cochrane Handbook, “summarize[s] the results of available
carefully designed healthcare studies and provides a high level of
IAN CHRISTOPHER LLOYD, FRCOPHTH
evidence on the effectiveness of healthcare interventions.”1
JUGNOO RAHI, PHD, FRCOPHTH Furthermore, systematic reviews minimize publication bias which
UCL Great Ormond Street Institute of Child Health, UCL Great can confound the interpretation of meta-analyses.
Ormond Street Institute, London, UK All studies included in AAO OTAs are reviewed by one or more
methodologists and then graded using a scale modified from the
Financial Disclosures: The authors made the following disclosures:
A.L.S.: Fellowship e Ulverscroft Vision Research Group fellowship;
Oxford Center for Evidence-based Medicine 2009 Levels of Evi-
Lecturer award e Academy of Medical Sciences; Clinician Scientist dence, as mandated by the AAO, and given an AAO Grade of I, II,
award e National Institute for Health Research (NIHR). or III. There are multiple levels of internal and external review for
J.R.: Senior investigator award e NIHR. each OTA by experienced epidemiologists, subject matter, and
A.L.S. and J.R.: Supported in part by the NIHR Biomedical Research methodologic panel chairs.
Centre (BRC) based at Moorfields Eye Hospital NHS Foundation Trust We do agree with Solebo et al that the 2011 Oxford Center
and UCL Institute of Ophthalmology. This work was undertaken at UCL for Evidence-based Medicine guidelines provide some enhance-
Institute of Child Health/Great Ormond Street Hospital, which received ments to the 2009 guidelines. However, in the opinion of our
a proportion of funding from the NIHR BRC scheme. The funding or- methodologist and panel chair methodologist, the Solebo et al
ganizations had no role in the design or conduct of this research. The
(references 11 and 21 in the original article) study represents
views expressed are those of the authors and not necessarily those of the
NHS, the NIHR or the Department of Health and Social Care. The au-
AAO Grade II evidence. Solebo et al are correct that the 2011
thors have no other financial disclosures to declare. guidelines specify that single observational studies can qualify as
Oxford Center for Evidence-based Medicine Level 2 evidence,
Available online: November 4, 2019. but the guidelines also state that studies may be downgraded
based on study quality. The Solebo study is a large prospective
Correspondence: cohort study conducted by a network of 69 ophthalmologists in
Jugnoo Rahi, PhD, FRCOphth, UCL Great Ormond Street Institute of the UK and the Republic of Ireland. Although their 5-year
Child Health, UCL Great Ormond Street Institute, 30 Guilford Street, follow-up rate of 92% is admirable, there are shortcomings in
London, WC1N 1EH United Kingdom. E-mail: [email protected]. the design of the study that led our methodologist to downgrade
it to AAO Level II evidence. These limitations include the
absence of a standardized surgical protocol, absence of a stan-
References dardized visual acuity testing protocol, absence of masked ex-
aminers, no power analysis, and the small size of the unilateral
1. Lambert SR, Aakalu VK, Hutchinson AK, et al. Intraocular lens cataract group. The absence of a surgical protocol is particularly
implantation during early childhood: a report by the American problematic and may account for the 5 times higher risk of
Academy of Ophthalmology. Ophthalmology. 2019;126: reoperation for visual axis opacity in the bilateral group and 20
1454e1461. times higher risk in the unilateral cataract group. For example, 8
2. Oxford Centre for Evidence-based Medicine. Levels of evi- children in their study who underwent intraocular lens (IOL)
dence. Available: www.cebm.net/2009/06/oxford-centre- implantation did not receive a posterior capsulotomy, whereas all

e8
 Correspondence

of the aphakic patients did. Because posterior capsule opacifi- SCOTT R. LAMBERT, MD1
cation always occurs in young children after cataract surgery if DEBORAH K. VANDERVEEN, MD2
the posterior capsule is left intact, it is not surprising that there VINAY K. AAKALU, MD, MPH3
was a higher rate of visual axis opacification in the pseudophakic STEPHEN J. KIM, MD4
group. Furthermore, Solebo et al (reference 21 in the original 1
Department of Ophthalmology, Stanford University School of
article) report that 67% of the patients in their study had a
Medicine, Palo Alto, California; 2Division of Ophthalmology, Boston
3-piece IOL implanted, suggesting that many of these IOLs were Children’s Hospital, Boston, Massachusetts; 3Illinois Eye and Ear
implanted in the sulcus, which can be more inflammatory. In Infirmary, University of Illinois College of Medicine at Chicago,
contrast, Vasavada et al (reference 12 in the original article) Chicago, Illinois; 4Vanderbilt Eye Institute, Vanderbilt University
reported that all 60 eyes randomized to IOL implantation at a School of Medicine, Nashville, Tennessee
median age of 6.0 months received a posterior capsulorrhexis
and anterior vitrectomy and all had a 1-piece IOL implanted into Financial Disclosures: The authors have no proprietary or commercial
the capsular bag. This difference may explain why only 10% of interests in any materials discussed in this letter.
the pseudophakic eyes in their clinical trial required surgery to D.K.V.: Lecture Fees e Ophtec
remove a visual axis opacity after a 5-year follow-up. The
Available online: November 4, 2019.
Toddler Aphakia and Pseudophakia Study (TAPS) also reported
a low rate of visual axis opacities in children undergoing uni- Correspondence:
lateral cataract surgery at a median age of 13.9 months (range, Scott R. Lambert, MD, Byers Eye Institute, 2452 Watson Ct, Palo Alto,
7e24 months) after a 5-year follow-up (reference 33 in the CA 94303. E-mail: [email protected].
original article). All but one of the eyes underwent a primary
posterior capsulotomy and anterior vitrectomy at the time of
cataract surgery and 44 of 51 (86%) eyes had the IOL implanted
Reference
in the capsular bag. An additional unplanned intraocular surgery
was performed on only 14% of these eyes compared with 48% 1. Higgins JPT, Green S, eds. Cochrane Handbook for Systematic
of the pseudophakic eyes in the Solebo et al study. Reviews of Interventions Version 5.1.0 [updated March 2011].
Solebo et al correctly quote our article as stating, “The best The Cochrane Collaboration; 2011. Available: www.handbook.
available evidence suggests that IOL implantation can be done cochrane.org.
safely with acceptable side effects in children older than 6
months of age.” However, they fail to acknowledge that we Re: Lee et al.: Longitudinal changes
subsequently state, “The training and experience of the surgeon in peripapillary retinal nerve fiber
as well as ocular and systemic comorbidities should be taken layer thickness in high myopia: a
into consideration when deciding when IOL implantation would prospective, observational study
be appropriate.” The 10 surgeons who participated in the TAPS
(Ophthalmology. 2019;126:522-528)
study also participated in the Infant Aphakia Treatment Study
(reference 13 in the original article). They were experienced TO THE EDITOR: We were interested to read this study, with its
surgeons who had undergone special training to perform this finding that “highly myopic eyes had a significantly greater
type of surgery. Although Solebo et al (reference 21 in the decrease in [the peripapillary retinal nerve fiber layer] pRNFL over
original article) state that the 69 surgeons participating in their 2 years than normal eyes.”1 We have found the pRNFL to be cross-
cohort study specialized in congenital and infantile cataracts, it sectionally associated with myopia (and intraocular pressure, age,
is unclear whether they had received special training to perform and smoking, among others)2 as well as optic disc area.3 The
this difficult surgery. In our opinion, IOL implantation in young possible reasons for the greater thinning are of interest,
children should not be performed by inexperienced surgeons. particularly with the increasing prevalence of high myopia in the
The age that IOLs can be safely implanted in the eyes of young world. The authors speculate that preperimetric glaucoma might
children likely depends on the training and skill of the surgeon. be a cause; while high myopia is a risk factor for glaucoma,4 the
Similarly, it should be recognized that there are significant authors have taken care to exclude individuals with glaucoma or
differences between IOL implantation in, for example, a 2- with ocular hypertension. There are 2 possible other reasons: one,
month-old compared with an 18-month-old, and there are with potentially important implications given the interest in
distinct reasons why one might opt for primary IOL implanta- pRNFL as a biomarker of brain health, is that myopes are more
tion in an older infant or toddler. IOL implantation may be likely to undergo neurodegeneration, although this seems
particularly helpful for children who are not likely to wear unlikely. The other possibility is that retinal stretching and
contact lenses for economic or social reasons. thinning is continuing in high myopes, particularly at older ages
We thank Solebo et al for their commentary and their important (either through progressive axial growth or staphyloma formation
contributions to the literature, and stress that there is room for at the disc). The authors mention that axial length was measured,
reasonable differences of opinion in this area. The reader of this but other than baseline measures this parameter was not analyzed.
study, and all OTAs, should understand the methods by which We wonder if the authors would consider examining these data,
studies are graded and that every effort has been made to equitably to see if there was progressive axial elongation in this group, or
and critically evaluate the literature in this, and all, AAO OTAs. It if the optic disc area and peripapillary atrophy increased during
is also important that the reader be exposed to dissenting opinions the 3-year follow-up, as a possible explanation for the greater
such as those of Solebo et al. progression of pRNFL thinning?

e9
 Ophthalmology Volume 127, Number 1, January 2020

As a final comment, we note a discrepancy between the inclu- Although the subjects did not overlap completely because of the
sion criteria for the cases and control groups. The stated inclusion different starting point and study period between 2 studies, the
criteria for high myopia eyes was based on an axial length of >26.0 inclusion criteria of 2 studies were almost same. We identified
mm, whereas that for the control group eyes was based on spherical that myopic eyes showed a consistent increase in axial length
equivalent between þ3.0 and -6.0 diopters. Can the authors confirm over 4 years, in which the rate of change was 0.047 mm/y, and
there was no overlap in either criterion between groups? Similarly, eyes with a longer baseline axial length showed a greater increase
the analysis consisted of 80 eyes in 79 subjects in the high myopia in axial length than eyes with a shorter axial length. Age was
group and 80 eyes in 77 subjects in the control group and the also one of the significant factors which might be associated with
criteria for which subjects had both eyes included (and the choice of changes of axial length in high myopia subjects, although it did
which eyes in the others) would be helpful to know. not show a significant result in multivariate analyses. These
findings would be helpful to explain the mechanism of the
PHILIP H. WRIGHT, BMBS, BMEDSCI1,2 reduction in pRNFL thickness in high myopia, which might be
OMAR A. MAHROO, PHD, FRCOPHTH1,2,3 related to mechanical stretching.
CHRISTOPHER J. HAMMOND, MD, FRCOPHTH1,3 We did not exclude patients with mild peripapillary cho-
1
Department of Ophthalmology, King’s College London, St Thomas’ rioretinal atrophy, which may not influence the measurements
Hospital Campus, London, UK; 2Institute of Ophthalmology, University of pRNFL thickness owing to its small size, and the atrophy
College London, London, UK; 3Department of Twin Research and did not show any changes until the last follow-up. Additionally,
Genetic Epidemiology, King’s College London, St Thomas’ Hospital glaucomatous changes such as an enlarged cup or increased
Campus, London, UK cup/disc ratio were not found during the study period. Thus,
these factors may be less relevant to the reduction of pRNFL
Financial Disclosures: The authors have no proprietary or commercial
interest in any materials discussed in this article.
thickness.
We enrolled the high myopia group based on axial length
Available online: November 4, 2019. rather than SE because SE could be affected by the cataract or
refractive surgeries. So subjects with an axial length of 26.00
Correspondence: mm were classified into the high myopia group regardless of the
Christopher Hammond, MD, FRCOphth, Department of Ophthal- SE. We defined the control group with SE to show that there was
mology, King’s College London, St Thomas’ Hospital Campus, West- no high myopia in terms of both axial length and SE in the
minster Road, London, UK. E-mail: [email protected]. control group. Actually, the longest axial length in the control
group was 25.88 mm, so there was no overlapping patients in the
References 2 groups.
The patients were enrolled with unilateral rhegmatogenous
retinal detachment, epiretinal membrane, macular hole, or intraoc-
1. Lee MW, Kim JM, Shin YI, et al. Longitudinal changes in peri-
ular lens dislocation as well as patients without any ophthalmic
papillary retinal nerve fiber layer thickness in high myopia: a pro-
spective, observational study. Ophthalmology. 2019;126:522e528. disease in both eyes. One eye was selected randomly in subjects
2. Mauschitz MM, Bonnemaiher PWM, Diers K, et al. systemic without any disease in both eyes, except for a few cases; 1 in the
and ocular determinants of peripapillary retinal nerve fiber layer 40- to 49-year age subgroup of the high myopia group, 1 in the 40-
thickness measurements in the European Eye Epidemiology (E3) to 49-year age subgroup of the control group, and 2 in 30- to 39-
population. Ophthalmology. 2018;125:1526e1536. year age subgroup of the control group. Although there might be
3. Jones-Odeh E, Yonova-Doing E, Bloch E, et al. The correlation some bias because patients enrolled both eyes, the number was not
between cognitive performance and retinal nerve fibre layer critical to the results of our report.
thickness is largely explained by genetic factors. Sci Rep.
2016;6, 34116-34116. MIN-WOO LEE, MD1,2
4. Marcus MW, de Vries MM, Junoy Montolio FG, et al. Myopia JU-MI KIM, MD1
as a risk factor for open-angle glaucoma: a systematic review YONG-IL SHIN, MD1
and meta-analysis. Ophthalmology. 2011;118:1989e1994.e2.
YOUNG-JOON JO, MD, PHD1
JUNG YEUL KIM, MD, PHD1
1
Department of Ophthalmology, Chungnam National University College
REPLY: We thank Wright et al for their interest and of Medicine, Daejeon, Republic of Korea; 2Department of
comments regarding our article, which evaluated lon- Ophthalmology, Konyang University College of Medicine, Daejeon,
gitudinal changes in peripapillary retinal nerve fiber Republic of Korea
layer (pRNFL) thickness in patients with high myopia by
comparing their eyes with the eyes of normal individuals.1 Financial Disclosures: The authors have no proprietary or commercial
interest in any materials discussed in this article.
We hypothesized the mechanism of the decrease in pRNFL
thickness as the globe elongation that leads to mechanical stretching
Available online: November 4, 2019.
and thinning of the retina as the previous study, and Wright et al
recommended to identify if there was progressive axial elongation Correspondence:
in the subjects who were enrolled in the study.2 Recently, we Jung-Yeul Kim, MD, PhD, Department of Ophthalmology, Chungnam
reported the longitudinal changes in the axial length in patients National University College of Medicine, #640 Daesa-dong, Jung-gu,
with high myopia without any other ophthalmic disease.3 Daejeon, 301-721, Korea. E-mail: [email protected].

e10
 Correspondence

References 2. Leung CK-S, Mohamed S, Leung KS, et al. Retinal nerve fiber
layer measurements in myopia: an optical coherence tomogra-
phy study. Invest Ophthalmol Vis Sci. 2006;47:5171e5176.
1. Lee M-W, Kim J-m, Shin Y-I, et al. Longitudinal changes in peri- 3. Lee M-W, Lee S-E, Lim H-B, et al. Longitudinal changes in
papillary retinal nerve fiber layer thickness in high myopia: a pro- axial length in high myopia: a 4-year prospective study. Br J
spective, observational study. Ophthalmology. 2019;126:522e528. Ophthalmol. 2019 Aug 13 [Epub ahead of print].

e11
over

Age-Related Macular
Degeneration Preferred
Practice Pattern®

© 2019 by the American Academy of Ophthalmology http://dx.doi.org/10.1016/j.ophtha.2019.09.024

Published by Elsevier Inc. ISSN 0161-6420/19
 Age-Related Macular Degeneration PPP

Secretary for Quality of Care:

Timothy W. Olsen, MD

Academy Staff:
Ali Al-Rajhi, PhD, MPH
Andre Ambrus, MLIS
Meghan Daly
Flora C. Lum, MD

Medical Editor: Susan Garratt

Approved by: Board of Trustees

September 7, 2019

© 2019 American Academy of Ophthalmology®

All rights reserved

AMERICAN ACADEMY OF OPHTHALMOLOGY and PREFERRED PRACTICE PATTERN are

registered trademarks of the American Academy of Ophthalmology. All other trademarks are the property of
their respective owners.

Preferred Practice Pattern® guidelines are developed by the Academy’s H. Dunbar Hoskins Jr., MD Center
for Quality Eye Care without any external financial support. Authors and reviewers of the guidelines are
volunteers and do not receive any financial compensation for their contributions to the documents. The
guidelines are externally reviewed by experts and stakeholders before publication.

Correspondence:
Ali A. Al-Rajhi, PhD, MPH, American Academy of Ophthalmology, P. O. Box 7424, San Francisco, CA
94120-7424. E-mail: [email protected].

2
P2
 Age-Related Macular Degeneration PPP

The Retina/Vitreous Preferred Practice Pattern® Panel members wrote the Age-Related Macular
Degeneration Preferred Practice Pattern® (PPP) guidelines. The PPP Panel members discussed and reviewed
successive drafts of the document, meeting in person twice and conducting other review by e-mail discussion, to
develop a consensus over the final version of the document.

Retina/Vitreous Preferred Practice Pattern Panel 2018–2019

Ron A. Adelman, MD, MPH, MBA, FACS
Gurunadh A. Vemulakonda, MD, American Society of Retina Specialists Representative
Steven T. Bailey, MD, Retina Society Representative
Amani Fawzi, MD, Macula Society Representative
Jennifer I. Lim, MD
Gui-shang Ying, MD, PhD, Methodologist
Christina J. Flaxel, MD, Chair

We thank our partners, the Cochrane Eyes and Vision US Satellite (CEV@US), for identifying reliable systematic
reviews that we cite and discuss in support of the PPP recommendations.

The Preferred Practice Patterns Committee members reviewed and discussed the document during a meeting
in June 2019. The document was edited in response to the discussion and comments.

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair
Roy S. Chuck, MD, PhD
Steven P. Dunn, MD
Christina J. Flaxel, MD
Steven J. Gedde, MD
Francis S. Mah, MD
Randall J. Olson, MD
David K. Wallace, MD, MPH
David C. Musch, PhD, MPH, Methodologist

The Age-Related Macular Degeneration PPP was then sent for review to additional internal and external groups and
individuals in July 2019. All those returning comments were required to provide disclosure of relevant relationships
with industry to have their comments considered (indicated with an asterisk below). Members of the Retina/Vitreous
Preferred Practice Pattern Panel reviewed and discussed these comments and determined revisions to the document.

P3
 Age-Related Macular Degeneration PPP

In compliance with the Council of Medical Specialty Societies’ Code for Interactions with Companies
(available at www.cmss.org/codeforinteractions.aspx), relevant relationships with industry are listed. The
Academy has Relationship with Industry Procedures to comply with the Code (available at
http://one.aao.org/CE/PracticeGuidelines/PPP.aspx). A majority (88%) of the members of the Retina/Vitreous
Preferred Practice Pattern Panel 2018–2019 had no financial relationship to disclose.

Retina/Vitreous Preferred Practice Pattern Panel 2018–2019

Christina J. Flaxel, MD: No financial relationships to disclose
Ron A. Adelman, MD, MPH, MBA, FACS: No financial relationships to disclose
Steven T. Bailey, MD: No financial relationships to disclose
Amani Fawzi, MD: No financial relationships to disclose
Jennifer I. Lim, MD: Genentech, Alcon Laboratories Inc., Kodiak Sciences, Opthea, Novartis—Consultant/Advisor;
Genentech, Novartis Alcon Pharmaceuticals—Lecture Fees
Gurunadh A. Vemulakonda, MD: No financial relationships to disclose
Gui-shang Ying, MD, PhD: No financial relationships to disclose

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair: No financial relationships to disclose
Roy S. Chuck, MD, PhD: No financial relationships to disclose
Steven P. Dunn, MD: No financial relationships to disclose
Christina J. Flaxel, MD: No financial relationships to disclose
Steven J. Gedde, MD: No financial relationships to disclose
Francis S. Mah, MD: Alcon Laboratories, Abbott Medical Optics, Bausch + Lomb—Consultant/Advisor; Abbott
Medical Optics, Bausch + Lomb—Lecture Fees
Randall J. Olson, MD: No financial relationships to disclose
David K. Wallace, MD, MPH: No financial relationships to disclose
David C. Musch, PhD, MPH, Methodologist: Chengdu Kanghong Biotechnology Opthea, IRIDEX, Notal Vision—
Consultant/Advisor

Secretary for Quality of Care

Timothy W. Olsen, MD: No financial relationships to disclose

Academy Staff
Ali Al-Rajhi, PhD, MPH: No financial relationships to disclose
Andre Ambrus, MLIS: No financial relationships to disclose
Meghan Daly: No financial relationships to disclose
Flora C. Lum, MD: No financial relationships to disclose

The disclosures of relevant relationships to industry of other reviewers of the document from January to
October 2019 are available online at www.aao.org/ppp.

P4
 Age-Related
Age-Related Macular
MacularDegeneration
DegenerationPPP
PPP

TABLE OF CONTENTS
OBJECTIVES OF PREFERRED PRACTICE PATTERN GUIDELINES ............................................. P7
METHODS AND KEY TO RATINGS ......................................................................................................... P8
HIGHLIGHTED FINDINGS AND RECOMMENDATIONS FOR CARE ............................................. P9
INTRODUCTION ........................................................................................................................................ P10
Disease Definition .......................................................................................................................................... P10
Patient Population ........................................................................................................................................... P11
Clinical Objectives ......................................................................................................................................... P11
BACKGROUND ........................................................................................................................................... P11
Incidence and Prevalence ............................................................................................................................... P11
Risk Factors .................................................................................................................................................... P13
Smoking, Hypertension, and Cardiovascular Disease ........................................................................... P13
Levels of Antioxidants ........................................................................................................................... P13
Diet ........................................................................................................................................................ P14
Aspirin ................................................................................................................................................... P14
Genetic Factors ...................................................................................................................................... P14
Other Risk Factors ................................................................................................................................. P16
Natural History ............................................................................................................................................... P16
Early Age-Related Macular Degeneration ............................................................................................. P16
Intermediate Age-Related Macular Degeneration ................................................................................. P16
Advanced Age-Related Macular Degeneration ..................................................................................... P17
Rationale for Treatment .................................................................................................................................. P18
Treatment Modalities ...................................................................................................................................... P19
Early Age-Related Macular Degeneration ............................................................................................. P19
Intermediate Age-Related Macular Degeneration ................................................................................. P19
Neovascular Age-Related Macular Degeneration .................................................................................. P22
CARE PROCESS ......................................................................................................................................... P29
Patient Outcome Criteria ................................................................................................................................ P29
Diagnosis ........................................................................................................................................................ P29
History ................................................................................................................................................... P29
Examination ........................................................................................................................................... P29
Diagnostic Tests..................................................................................................................................... P29
Management ................................................................................................................................................... P31
Monitoring and Early Detection ............................................................................................................ P32
Indications for Treatment for Choroidal Neovascularization ................................................................ P33
Complications of Treatment .................................................................................................................. P36
Follow-up Evaluation ............................................................................................................................ P39
Provider and Setting ....................................................................................................................................... P40
Counseling and Referral ................................................................................................................................. P40
Socioeconomic Considerations ....................................................................................................................... P41
APPENDIX 1. QUALITY OF OPHTHALMIC CARE CORE CRITERIA .......... ............................... P43
APPENDIX 2. INTERNATIONAL STATISTICAL CLASSIFICATION OF DISEASES AND
RELATED HEALTH PROBLEMS (ICD) CODES .................................................................................. P45
GLOSSARY .................................................................................................................................................. P46
LITERATURE SEARCHES FOR THIS PPP ........................................................................................... P50
RELATED ACADEMY MATERIALS ...................................................................................................... P51
REFERENCES ............................................................................................................................................. P52

P5
P5
 Age-Related Macular Degeneration PPP

Age-Related Macular Degeneration PPP

Background:
Age-related macular degeneration is a leading cause of severe, irreversible vision impairment in
developed countries. The primary risk factors for the development of advanced AMD include
increasing age, northern European ancestry, and genetic factors. Smoking has been shown by
numerous studies to be the main modifiable risk factor.
This Preferred Practice Pattern (PPP) uses the classification of the Age-Related Eye Disease
Study (AREDS) and a more recent clinical classification to define the early and intermediate
stages of AMD since current treatment recommendations are based on these classifications. The
PPP recommendations are based on Cochrane-identified reliable systematic reviews.
Rationale for treatment:
Prospective randomized controlled clinical trials support the use of antioxidant vitamins and
minerals for slowing the progression to later stages of AMD, intravitreal injection of anti-VEGF
agents, photodynamic therapy (PDT), and laser photocoagulation surgery to treat neovascular
AMD. It should be noted that intravitreal injection therapy using anti-vascular endothelial
growth factor (VEGF) agents (e.g., aflibercept, bevacizumab, and ranibizumab) is the most
effective way to manage neovascular AMD and represents the first line of treatment.
Care Process:
Patient outcome criteria are to reverse or minimize visual loss and improve visual function. The
initial evaluation of a patient with signs and symptoms suggestive of AMD includes all features
of the comprehensive adult medical eye evaluation, with particular attention to those aspects
relevant to AMD. In patients with neovascular AMD, early detection and prompt treatment
improves the visual outcome. Symptoms suggestive of postinjection endophthalmitis or retinal
detachment require prompt evaluation.
Fundus fluorescein angiography and optical coherence tomography (OCT) are useful diagnostic
tests in clinical practice to detect new or recurrent neovascular disease activity and guide
therapy.
Management options for AMD include observation and early detection, antioxidant vitamin and
mineral supplements, intravitreal injection of anti-VEGF agents, PDT, laser photocoagulation
surgery, and the encouragement of smoking cessation for patients who currently smoke. All
patients with AMD should be educated about the prognosis of the disease and the potential value
of treatment as appropriate for their visual and functional status.

P6

P6
 Age-Related Macular Degeneration PPP

As a service to its members and the public, the American Academy of Ophthalmology has developed a series
of Preferred Practice Pattern® guidelines that identify characteristics and components of quality eye care.
Appendix 1 describes the core criteria of quality eye care.

The Preferred Practice Pattern® guidelines are based on the best available scientific data as interpreted by
panels of knowledgeable health professionals. In some instances, such as when results of carefully conducted
clinical trials are available, the data are particularly persuasive and provide clear guidance. In other instances,
the panels have to rely on their collective judgment and evaluation of available evidence.

These documents provide guidance for the pattern of practice, not for the care of a particular
individual. While they should generally meet the needs of most patients, they cannot possibly best meet the
needs of all patients. Adherence to these PPPs will not ensure a successful outcome in every situation. These
practice patterns should not be deemed inclusive of all proper methods of care or exclusive of other methods
of care reasonably directed at obtaining the best results. It may be necessary to approach different patients’
needs in different ways. The physician must make the ultimate judgment about the propriety of the care of a
particular patient in light of all of the circumstances presented by that patient. The American Academy of
Ophthalmology is available to assist members in resolving ethical dilemmas that arise in the course of
ophthalmic practice.

Preferred Practice Pattern® guidelines are not medical standards to be adhered to in all individual
situations. The Academy specifically disclaims any and all liability for injury or other damages of any kind,
from negligence or otherwise, for any and all claims that may arise out of the use of any recommendations or
other information contained herein.

References to certain drugs, instruments, and other products are made for illustrative purposes only and are
not intended to constitute an endorsement of such. Such material may include information on applications
that are not considered community standard, that reflect indications not included in approved U.S. Food and
Drug Administration (FDA) labeling, or that are approved for use only in restricted research settings. The
FDA has stated that it is the responsibility of the physician to determine the FDA status of each drug or
device he or she wishes to use, and to use them with appropriate patient consent in compliance with
applicable law.

Innovation in medicine is essential to ensure the future health of the American public, and the Academy
encourages the development of new diagnostic and therapeutic methods that will improve eye care. It is
essential to recognize that true medical excellence is achieved only when the patients’ needs are the foremost
consideration.

All Preferred Practice Pattern® guidelines are reviewed by their parent panel annually or earlier if
developments warrant and updated accordingly. To ensure that all PPPs are current, each is valid for 5 years
from the approved by date unless superseded by a revision. Preferred Practice Pattern guidelines are funded
by the Academy without commercial support. Authors and reviewers of PPPs are volunteers and do not
receive any financial compensation for their contributions to the documents. The PPPs are externally
reviewed by experts and stakeholders, including consumer representatives, before publication. The PPPs are
developed in compliance with the Council of Medical Specialty Societies’ Code for Interactions with
Companies. The Academy has Relationship with Industry Procedures (available at www.aao.org/about-
preferred-practice-patterns) to comply with the Code.

Appendix 2 contains the International Statistical Classification of Diseases and Related Health Problems
(ICD) codes for the disease entities that this PPP covers. The intended users of the Age-Related Macular
Degeneration PPP are ophthalmologists.

P7
6
 Age-Related Macular Degeneration PPP

Preferred Practice Pattern® guidelines should be clinically relevant and specific enough to provide useful
information to practitioners. Where evidence exists to support a recommendation for care, the
recommendation should be given an explicit rating that shows the strength of evidence. To accomplish these
aims, methods from the Scottish Intercollegiate Guideline Network1 (SIGN) and the Grading of
Recommendations Assessment, Development and Evaluation 2 (GRADE) group are used. GRADE is a
systematic approach to grading the strength of the total body of evidence that is available to support
recommendations on a specific clinical management issue. Organizations that have adopted GRADE include
SIGN, the World Health Organization, the Agency for Healthcare Research and Policy, and the American
College of Physicians.3
◆ All studies used to form a recommendation for care are graded for strength of evidence individually, and
that grade is listed with the study citation.
◆ To rate individual studies, a scale based on SIGN1 is used. The definitions and levels of evidence to rate
individual studies are as follows:
I++ High-quality meta-analyses, systematic reviews of randomized controlled trials (RCTs), or
RCTs with a very low risk of bias
I+ Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias
I- Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias
II++ High-quality systematic reviews of case-control or cohort studies
High-quality case-control or cohort studies with a very low risk of confounding or bias and a
high probability that the relationship is causal
II+ Well-conducted case-control or cohort studies with a low risk of confounding or bias and a
moderate probability that the relationship is causal
II- Case-control or cohort studies with a high risk of confounding or bias and a significant risk that
the relationship is not causal
III Nonanalytic studies (e.g., case reports, case series)
◆ Recommendations for care are formed based on the body of the evidence. The body of evidence quality
ratings are defined by GRADE2 as follows:
Good quality Further research is very unlikely to change our confidence in the estimate of
effect
Moderate quality Further research is likely to have an important impact on our confidence in the
estimate of effect and may change the estimate
Insufficient quality Further research is very likely to have an important impact on our confidence in
the estimate of effect and is likely to change the estimate
Any estimate of effect is very uncertain
◆ Key recommendations for care are defined by GRADE 2 as follows:
Strong Used when the desirable effects of an intervention clearly outweigh the
recommendation undesirable effects or clearly do not
Discretionary Used when the trade-offs are less certain—either because of low-quality evidence
recommendation or because evidence suggests that desirable and undesirable effects are closely
balanced

◆ The Highlighted Findings and Recommendations for Care section lists points determined by the PPP
Panel to be of particular importance to vision and quality of life outcomes.
◆ All recommendations for care in this PPP were rated using the system described above. Ratings are embedded
throughout the PPP main text in italics.
◆ Literature searches to update the PPP were undertaken in March 2018 and June 2019 in PubMed and the
Cochrane Library. Complete details of the literature searches are available online at www.aao.org/ppp.

P8
7
 Age-Related Macular Degeneration PPP

Although an estimated 80% of age-related macular degeneration (AMD) patients have non-neovascular or
atrophic AMD, the neovascular form is responsible for the majority of the severe central visual acuity (VA)
loss associated with AMD.

The primary risk factors for the development of advanced AMD include increasing age, northern European
ancestry, and genetic factors. Cigarette smoking is the main modifiable risk factor that has been consistently
identified in numerous studies. Smoking cessation is strongly recommended when advising patients who
have AMD or are at risk for AMD. The routine use of genetic testing is not recommended at this time.

A meta-analysis of 10 studies found that the use of aspirin was not associated with an increased risk of AMD.
Therefore, patients who have been instructed by a physician to use aspirin should continue to use it as
prescribed.

Antioxidant vitamin and mineral supplementation as per the Age-Related Eye Disease Study (AREDS2)
should be considered in patients with intermediate or advanced AMD. There is no evidence to support the
use of these supplements for patients who have less than intermediate AMD and no evidence of any
prophylactic value for family members without signs of AMD.

Fluorescein angiography, optical coherence tomography (OCT), and optical coherence tomography
angiography (OCTA) are useful diagnostic tests in clinical practice to detect new or recurrent neovascular
disease activity and guide therapy.

In patients with neovascular AMD, early detection and prompt treatment improves the visual outcome.
Intravitreal injection therapy using anti-vascular endothelial growth factor (VEGF) agents (e.g., aflibercept,
bevacizumab, and ranibizumab) is the most effective way to manage neovascular AMD and represents the
first line of treatment. Symptoms suggestive of postinjection endophthalmitis or retinal detachment require
prompt evaluation.

P9
8
 Age-Related Macular Degeneration PPP

Age-related macular degeneration (AMD) is a disorder of the macula characterized by one or more of
the following (for specific terms, see Glossary):

◆ Presence of at least intermediate-size drusen (>63 µm in diameter)

◆ Retinal pigment epithelium (RPE) abnormalities such as hypopigmentation or hyperpigmentation

◆ Presence of any of the following features: geographic atrophy of the RPE, choroidal
neovascularization ([CNV] exudative, wet), polypoidal choroidal vasculopathy (PCV), reticular
pseudodrusen, or retinal angiomatous proliferation

This Preferred Practice Pattern uses the classification of the Age-Related Eye Disease Study
(AREDS) and a more recent clinical classification4 to define the early and intermediate stages of
AMD because current treatment recommendations are based on these classifications. The AREDS
was a prospective multicenter randomized clinical trial conducted between 1992 and 2006 designed to
assess the natural course and risk factors for age-related cataract and AMD. The effects of antioxidant
vitamins and minerals on these two ocular conditions were studied.

The classification of AMD from the AREDS is as follows:5

◆ No AMD (AREDS category 1) represented the control group; it is characterized by no or few small
drusen (<63 µm in diameter).

◆ Early AMD (AREDS category 2) is characterized by a combination of multiple small drusen, few
intermediate drusen (63–124 µm in diameter), or mild RPE abnormalities.

◆ Intermediate AMD (AREDS category 3) is characterized by any of the following features:

◆ Numerous intermediate drusen

◆ At least one large druse (125 µm in diameter)

◆ Geographic atrophy (a sharply demarcated, usually round or oval, area of atrophy of the RPE not
involving the center of the fovea)
◆ Advanced AMD (AREDS category 4) is characterized by one or more of the following (in the absence
of other causes) in one eye:
◆ Geographic atrophy of the RPE involving the foveal center
◆ Neovascular maculopathy that includes the following:
o CNV defined as pathologic angiogenesis originating from the choroidal vasculature that
extends through a defect in Bruch’s membrane
o Serous and/or hemorrhagic detachment of the neurosensory retina or RPE
o Retinal hard exudates (a secondary phenomenon resulting from chronic vascular leakage)
o Subretinal and sub-RPE fibrovascular proliferation

P10
9
 Age-Related Macular Degeneration PPP

o Disciform scar (subretinal fibrosis)

See Glossary for definitions of important terms. Clinical details are available in standard texts. 6,7

Patients are typically aged 50 years or older, with or without visual symptoms. Clinicians should
consider the possibility of hereditary macular dystrophies in patients under 50 years of age who have
clinical features that resemble AMD.

◆ Identify patients at risk of visual loss related to AMD

◆ Educate patients and their families about the disease, risk factors, and preventive measures
◆ Minimize or reverse visual loss and functional impairment in these patients through appropriate
detection, self-assessment, treatment, and follow-up examinations
◆ Help patients identify expert physicians and resources needed to facilitate improvement in vision

Age-related macular degeneration is a leading cause of severe, irreversible vision impairment in

developed countries.8-13 In 2004, it was estimated that approximately 1.75 million people aged 40
years or older in the United States have advanced AMD, either neovascular AMD or geographic
atrophy in at least one eye; and 7.3 million were considered to have high-risk features, such as large
drusen (≥125 µm in diameter) in one or both eyes.12 The authors projected that the number of
individuals affected by advanced AMD in at least one eye will increase to nearly 3 million by year
2020,12 based on the aging population demographics in the United States. 14 Aging is the greatest risk
factor; therefore, the prevalence of AMD in the United States is anticipated to increase to 22 million
by the year 2050, while the global prevalence is expected to increase to 288 million by the year
2040.15 These predictions are likely to be affected by both more effective treatments for the
neovascular forms of AMD using anti-vascular endothelial growth factor (VEGF) agents as well as
the slowing of the disease progression using antioxidant vitamins with zinc. The use of anti-VEGF
agents will likely reduce the odds of legal blindness from neovascular AMD and could theoretically
reduce the rate of legal blindness by up to 70% over 2 years.16 However, longer-term follow-up
studies from the population originally treated with regular anti-VEGF agents suggest that these gains
in visual acuity (VA) are largely lost in two-thirds of patients followed for over 7 years.17 The use of
antioxidant vitamins (i.e., vitamin C, vitamin E), lutein, zeaxanthin, and zinc in an otherwise well-
nourished population with intermediate AMD has been demonstrated to reduce the progression

10 P11
 Age-Related Macular Degeneration PPP

toward more advanced stages of AMD by approximately 25% at 5 years.5,18 A study forecasting the
potential impact of treatments in AMD concluded that though the prevalence of AMD will increase
substantially by 2050 in the United States, the use of anti-VEGF therapies and vitamin therapies will
mitigate these effects.19

Overall, AMD is responsible for an estimated 46% of cases of severe visual loss (VA 20/200 or
worse) in persons over age 40 in the United States.13 While most consider the onset of AMD as
occurring in individuals over the age of 50, there are variations in the epidemiologic literature. While
relatively few cases of advanced AMD occur between ages 40 and 50, detection of earlier AMD
stages, which are precursors of more advanced AMD, are not uncommon occurrences during this
decade. Therefore, the reader must keep in mind that AMD is a disease spectrum that has early and
later stages. Although an estimated 80% of AMD patients have non-neovascular or atrophic AMD,9
the neovascular form with its natural history is responsible for nearly 90% of the severe VA loss
(20/200 or worse) from AMD.20,21

The prevalence, incidence, and progression of AMD and most associated features (e.g., large drusen)
increase with age. The prevalence of AMD also varies by ethnicity.13,22-24 In the Beaver Dam Eye
Study, consisting of primarily a Caucasian population base, the prevalence of any AMD (referred to
as age-related maculopathy) was less than 10% in persons aged 43 to 54 years yet more than tripled
for persons aged 75 to 85 years of age.8 The Beaver Dam Eye Study demonstrated that the
development of any AMD over a 10-year period was 4.2% for persons 43 to 54 years old and 46% for
those 75 and older.25 The Beaver Dam Eye Study has identified that soft, indistinct drusen and
pigmentary abnormalities also increase in frequency with increasing age and are strongly predictive of
progression to more advanced AMD. In the Los Angeles Latino Eye Study, prevalence of advanced
AMD increased from 0% in individuals 40 to 49 years old to 8.5% in those 80 years old and older.26
The Proyecto Vision Evaluation and Research study of Hispanic participants in Arizona found that the
prevalence of advanced AMD increased from 0.1% in persons 50 to 59 years old to 4.3% in those 80
and older.27

Observations from the Barbados Eye Study, 28 the Baltimore Eye Study,29 and the Macular
Photocoagulation Study (MPS)30 suggest that late stages of AMD are more common among
Caucasians. Findings from the Multi-ethnic Study of Atherosclerosis also suggest that neovascular
AMD may be more common in Caucasians than in African Americans.23 In Asian populations, there
are racial variations in the prevalence of early and late AMD, and Caucasian and Asian populations
are at higher risk than Hispanic and African individuals.31-36 A recent meta-analysis and systematic
review reported a higher prevalence of AMD in Europeans than in Asians or Africans, with no
difference in prevalence between Asians and Africans. The global number of people with AMD was
projected to be 196 million in 2020, increasing to 288 million in 2040. 15

P12
11
 Age-Related Macular Degeneration PPP

The main risk factors for the development of advanced AMD are increasing age, ethnicity (i.e.,
Caucasian and family history). Although a number of modifiable risk factors have been investigated,
cigarette smoking is the main modifiable risk factor that has been consistently identified in numerous
studies.37-46 Importantly, it is essential to recognize that the associations found in observational studies
that analyze risk factors should not be interpreted as cause and effect. Such associations may not
necessarily translate into treatment recommendations, as there may be multiple confounding variables
that are not accounted for in the studies.

Smoking significantly increases the risk of AMD and there appears to be a dose response
relationship, because the odds ratio increases with an increased number of pack-year
exposure.39,47 Smoking cessation is associated with a reduced risk of AMD progression; the risk
of developing AMD in individuals who have not smoked for more than 20 years is comparable
to the risk in nonsmokers.39 Thus, smoking cessation is strongly recommended when advising
patients, as it represents a key and important modifiable risk factor. A number of case-control
and population-based studies have examined the relationship between AMD, hypertension, and
other cardiovascular diseases. These studies have shown conflicting results.22,48-54 Passive
smoking exposure was associated with an increased risk of AMD (odds ratio 1.87%; 95%
confidence interval [CI] 1.03 – 3.40) in non-smokers.39

Additional risk factors may include low systemic levels of antioxidants. Data from
observational studies have been inconsistent in identifying low levels of plasma and dietary
antioxidants of vitamins C and E, carotenoids (e.g., lutein, zeaxanthin), and zinc as risk factors
for AMD.55-61 The original AREDS results demonstrated a beneficial effect for the use of high-
dose oral antioxidant vitamins (vitamins C, E, beta-carotene) and zinc supplementation in
reducing progression of intermediate AMD or advanced AMD in the fellow eye to advanced
AMD by 25%.62 However, additional vitamin E supplementation above the AREDS levels
should be avoided.63 Results of AREDS2 support the removal of beta-carotene (found in the
original AREDS supplements) and the addition of lutein/zeaxanthin in the AREDS2
supplements.18 Furthermore, elimination of the beta-carotene component may reduce the
competitive absorption of the lutein/zeaxanthin. Importantly, removal of beta-carotene may also
decrease higher incidence of lung cancer associated with the use of supplemental beta-
carotene.64 Finally, AREDS2 demonstrated that there was no effect on the progression of AMD
by either reducing the zinc dose (from 80 mg to 25 mg) or adding an omega-3 polyunsaturated
fatty acid supplement (docosahexaenoic acid [DHA] and eicosapentaenoic acid [EPA]). 64 A
recent Cochrane systematic review concluded that taking antioxidant vitamins plus zinc

P13
12
 Age-Related Macular Degeneration PPP

probably slows the progression to late AMD and vision loss (moderate-certainty evidence).
They also concluded that supplements containing only lutein and zeathanthin may have little or
no effect on the progression of AMD.65

Several studies have also identified an association between dietary fat and advanced AMD. 40,66-71
Similar to the reports on risk factors for cardiovascular disease, a number of reports from
population-based studies have demonstrated that a reduced risk of AMD is associated with higher
dietary intake of foods rich in omega-3 long-chain polyunsaturated fatty acids, such as fish.40,70-73
In a nested cohort study from the original AREDS population of 1837 patients who were at
moderate risk for progression, participants who reported the highest omega-3 intake (note that
this was not in the form of a supplement) were 30% less likely to develop advanced AMD after
12 years.71 These dietary long-chain fatty acids are felt to decrease inflammatory mediators via
immunomodulation, thus decreasing disease progression to advanced AMD. 71 An increased risk
of AMD was found in individuals who had a higher intake of saturated fats and cholesterol and in
those with a higher body mass index.44 Despite this dietary association, AREDS2 failed to
demonstrate a benefit from the use of DHA and EPA as oral supplements at the doses tested; both
are omega-3 poly-unsaturated fatty acids.64 The EYE-RISK consortium recently published their
evaluation of the pooled data from the Rotterdam Study-1 and the Alienor Study populations,
which included over 4000 participants with mean follow-up of 9.9 years and 4.1 years,
respectively, and adherence to Mediterranean diet and found this diet was associated with 41%
reduced risk of advanced AMD. The Mediterranean diet includes a diet rich in fruits, vegetables,
legumes and fish.74,75

Recent observational studies have indicated a possible link between aspirin use and AMD. The
Beaver Dam Eye Study reported two times the incidence of late macular degeneration in patients
who used aspirin at least twice weekly for 10 years compared with those who used no aspirin.76,77
Other studies have shown a potential protective effect of aspirin against the development of
AMD.78 In a meta-analysis of 10 studies including over 171,000 patients, the use of aspirin was
not associated with an increased risk of AMD. 79 In light of all of the available information on the
subject of aspirin use and AMD, the current preferred practice is for patients who have been
instructed to use aspirin by a physician to continue their aspirin therapy as prescribed.80 81

Molecular genetic studies and epidemiologic studies have determined some of the genetic factors
in AMD.82-88 Several studies published in 2005 identified a strong association of the complement
factor H (CFH) Y402H polymorphism with a higher risk of AMD.89-94

P14
13
 Age-Related Macular Degeneration PPP

The CFH gene product is involved in regulation of the complement system through binding to
factor C3b. This specific complement factor represents a key regulator of the innate rather than the
adaptive immune system. An alteration of regulation that occurs as a result of modification at the
C3b site leads to a defective regulation of the alternative complement pathway and results in an
up-regulation of inflammation to host cells that are mediated by the membrane attack complex.
Patients homozygous for the Y402H risk allele of CFH possess a 7.4-fold increased risk of AMD.
The CFH gene is located on chromosome 1, in a region linked to AMD in multiple family
studies.89 Studies report an association of a CFH variant (homozygous individuals) with other
factors for the risk of progression to advanced AMD compared with noncarriers who lack these
determinants.95,96 Other factors associated with abnormal complement variants and AMD
progression include an elevated erythrocyte sedimentation rate, an elevated serum C-reactive
protein, and smoking. Such findings support the combined pathogenic mechanisms for AMD
progression that include an interplay of environmental factors, heredity, and inflammation.

Strong linkage disequilibrium has been shown across the ARMS2-HTRA1 region, and these two
genes are also strongly associated with AMD.97-99 The exact mechanism that explains this
association has not been clearly determined.100 Other proposed genetic variants associated with
AMD include a variant in the hepatic lipase (LIPC) gene101 and the rs3775291 variant in the toll-
like receptor 3 (TLR3) gene.102,103 A number of other genes have also been identified as well as
several other rare variants of genes.104 A combination of genes and other risk factors may dispose
an individual to varying AMD risks more than any one variant taken in isolation. 105 A recent
genome-wide association study has identified 19 loci (P<5x10-8), seven of which are newly
described.106

Age-related macular degeneration has a complex genetic background with similar phenotypes.
Many genetic associations have been identified—some are protective,107 some are associated with
disease progression, and others have been reported yet not confirmed and require further
investigation.

In 2013, several authors proposed that genetic selection of subjects who would most benefit from
nutritional supplementation should be used to guide therapy based on a post hoc analysis of a
subset of the AREDS population. Thus, the authors recommend using a personalized genetic
testing approach to guide therapy in AMD.108,109 However, an analysis of the AREDS population
that included an additional 526 AREDS subjects concluded that genetic testing does not provide
benefits in managing nutritional supplements in this population. 110-112

Statistical experts found errors in the data used to support an association, and bias in the analyses
used to support genetic testing. They concluded that there was no evidence to support the need for
genotyping to guide recommendations for use of supplements containing antioxidants and zinc in
AMD.113

P15
14
 Age-Related Macular Degeneration PPP

A recent prospective, multicenter study looked at genome-wide associations with treatment

outcomes in a cohort of 465 patients with exudative AMD who were initiating ranibizumab
therapy.114 Although there was no association of any single-nucleotide polymorphism with 12-
month treatment outcomes (i.e., achieving a dry macula, requiring additional treatment, and visual
acuity change), the authors found preliminary evidence of a predictive association of the
ARMS/HTRA1 polymorphism with the need for additional treatment. They postulated that testing
for this polymorphism might be able to predict the frequency of injection after initial ranibizumab
therapy. However, a systematic review published in 2015 looked at the association between anti-
VEGF response and variations in AMD-associated genes and concluded that genetic background
may influence an individual’s response to treatment, however further studies are needed to better
understand the contribution of various genes to treatment response. 115

Currently, only post hoc analysis data is available and results are conflicting.116 One or more
prospective clinical trials will need to demonstrate the value of genetic testing in AMD. That is,
randomization based on genetic type has not been done for neovascular AMD treatment response
to date. Thus, the routine use of genetic testing is not supported by the existing literature and is not
recommended at this time.

An increased waist/hip ratio for men has been associated with an increase in the risk of both early
and late AMD.117 Markers of inflammation, such as C-reactive protein, may be associated with a
higher risk of AMD progression.118-120 Other possible factors that have been considered in various
studies, with inconclusive findings, include hormonal status, 121-125 sunlight exposure,126-128 alcohol
use,129-131 and vitamins B and D status.132,133 A Cochrane systematic review in 2016 concluded
that there was insufficient evidence to define a role of statins in the onset or progression of
AMD.134

As defined by the AREDS, early AMD (category 2) is characterized by small drusen (<63 µm
in diameter), few medium drusen (63–125 µm in diameter), and/or minimally detected or no
pigment epithelial abnormalities in the macula. Patients in this category have a low risk of
progressing to advanced AMD after 5 years in either eye.5 More recently, the AREDS study
group published a report based on 10-year follow-up data obtained from approximately 85% of
the originally enrolled patients.135 In the group with a combination of small drusen or no drusen
at baseline, approximately 15% developed large drusen at 10 years.

P16
15
 Age-Related Macular Degeneration PPP

Intermediate AMD (category 3) is a more critical distinction clinically because it places the
individual at risk for progression to more advanced AMD. It has been defined by the AREDS as
having extensive medium drusen (63–124 µm in diameter) or one or more large drusen (125
µm in diameter) in one or both eyes. The progression to advanced AMD at 5 years in this group
is approximately 18% according to the original AREDS. However, for patients with large
drusen in one eye, the rate of development of advanced AMD at 5 years is 6.3%, whereas the
rate for patients with multiple bilateral large drusen increases to 26% at 5 years.5,136 In the 10-
year follow-up study of the AREDS, 37% of patients developed large drusen when medium
drusen were present at baseline in one eye, and 71% developed large drusen when medium
drusen were present in both eyes at baseline.135 When medium drusen were present at baseline,
14% progressed to advanced AMD at 10 years.

In 2005, a simplified severity scale was developed for assessing AMD risk progression that is
based on two primary ophthalmoscopic features: one or more large drusen (≥125 µm in
diameter) and the presence of pigmentary changes.137 Individuals with two affected eyes could
then be given a five-step grading score of 0–4 (based on one point for each factor being present
in each eye). The following scores enable the clinician to communicate with the patient about
the approximate 5-year risk for developing advanced AMD: four factors, 45%; three factors,
26%; two factors, 9%; one factor, 4%; and zero factors, 0.5%. The approximate 10-year risks
were 71%, 53%, 28%, 8%, and 1.5%, respectively.135

For patients without large drusen, the presence of intermediate drusen in both eyes is considered
to represent one risk factor using this severity scale. Advanced AMD in one eye is counted as
two risk factors. Often, the eye contralateral to the eye with Advanced AMD has large drusen
and RPE pigmentary disturbances and therefore has four risk factors, the highest risk-level for
progression of all patients with AMD (50% by 5 years and 71% by 10 years). Interestingly, an
online AMD risk calculator that includes phenotype (simplified severity scale score described
above) and demographic information (age, smoking, and family history of AMD) had excellent
calibration and overall performance, whereas the addition of specific genetic analysis added
little to the 9- to 10-year trend for the development of advanced AMD. 138

Reticular pseudodrusen (also referred to as subretinal drusenoid deposits) may be under-

recognized. They are best imaged using fundus autofluorescence, infrared reflectance, and/or
spectral-domain optical coherence tomography (SD-OCT), and they appear to represent a
meaningful risk factor associated with progression to the geographic atrophy.139-144 (See
Glossary.)

P17
16
 Age-Related Macular Degeneration PPP

Advanced AMD (category 4) as defined in the AREDS refers to either neovascular AMD or
geographic atrophy involving the center of the macula. Visual acuity in one eye is affected in all
category 4 patients. In the Beaver Dam Eye Study, approximately 22% of the fellow eyes of
such patients developed neovascular changes or geographic atrophy involving the fovea over 5
years.145 In the AREDS, for patients with advanced AMD in one eye, the risk of progression to
an advanced stage in the fellow eye ranged from 35% to 50% at 5 years, depending largely on
the phenotype in the better eye.137 In the Submacular Surgery Trial (SST), these findings were
also confirmed and further emphasize the value of the simple risk scale. 146

The phenotype of central geographic atrophy, the advanced form of non-neovascular AMD,
will have one or more zones of well-demarcated RPE and/or choriocapillaris atrophy. Drusen
and other pigmentary abnormalities may surround the atrophic areas. Severe VA loss occurs
less commonly and more slowly in patients with geographic atrophy than in patients with
neovascular AMD. Geographic atrophy involving the foveal center causes approximately 10%
of all AMD-related visual loss of 20/200 or worse.147 Patients with geographic atrophy not
necessarily involving the central fovea may have relatively good distance VA yet manifest a
substantially decreased ability to perform near visual tasks such as reading.147 Doubling of the
visual angle in patients with geographic atrophy has been reported to occur in as many as 50%
of patients over a 2-year period.147 Choroidal neovascularization also may occur.

Neovascular AMD is characterized angiographically as either classic, occult, predominantly

classic, minimally classic, or mixed lesions. (See Glossary.) Serous and/or hemorrhagic
detachment of the neurosensory retina or the RPE, and/or various stages of an elevated,
fibrovascular disciform scar, may also occur.

In the Macular Photocoagulation Study (MPS), classification of neovascular AMD with CNV
was based on fluorescein angiography. Classic CNV (Gass Type 2 membrane)148-151 is defined
as a well-demarcated hyperfluorescence in the early phase of the angiogram, with progressive
leakage of dye into the overlying subneurosensory retinal space during the late phases of the
angiogram. Occult CNV (Gass Type 1 membrane)148-151 is characterized by either a
fibrovascular pigment epithelial detachment (PED) or late leakage of undetermined source. A
fibrovascular PED is an irregular elevation of the RPE that has accompanying stippled
heterofluorescence or even hypofluorescence early in the angiogram, with progressive late
leakage in the later stages of the angiogram. An occult lesion with late leakage of undetermined
source is not elevated yet shows a similar pattern of late leakage (usually after 1 minute). Other
clinical subtypes or features of neovascular AMD may include the following:

◆ Retinal PED
◆ Idiopathic PCV,152,153 which should be suspected in patients with orange polypoid lesions and
especially in patients of African or Asian descent. The lesions are often located in the

P18
17
 Age-Related Macular Degeneration PPP

peripapillary region, but may also present in the central macula or the macular arcades initially
as large hemorrhagic retinal PED, lipid exudation, and subretinal fluid. An indocyanine green
(ICG) angiogram is often useful in confirming the diagnosis.
◆ Retinal angiomatous proliferation 154

Prospective, randomized, controlled clinical trials support the use of antioxidant vitamins and
minerals for slowing the progression to later stages of AMD, intravitreal injection of anti-VEGF
agents, photodynamic therapy (PDT), and laser photocoagulation surgery to treat neovascular AMD.
However, thermal laser photocoagulation surgery is no longer recommended for subfoveal CNV
treatment (See Glossary.) At present, there is no proven therapy to prevent or treat geographic
atrophy.155

The use of the combination of antioxidant vitamins and minerals did not reduce the progression
of early AMD to the intermediate stage of AMD, and there was insufficient power to determine
the effects of the combination treatment on the progression to more advanced AMD. Therefore,
there is no evidence to support the use of these supplements for patients who have less than
intermediate AMD. In early AMD (AREDS category 2), only 1.3% of participants progressed
to advanced AMD in 5 years. A meta-analysis by Evans in 2012 that looked at the evidence
about whether to take an antioxidant vitamin or mineral supplement prevents the development
of AMD concluded that there was accumulating evidence that taking vitamin E or beta-carotene
supplements will not prevent or delay the onset of AMD. 156

The original AREDS used a factorial design whereby 4757 participants were randomized to
antioxidant vitamins, zinc, a combination of antioxidant vitamins and minerals (zinc and
copper), or a placebo, and they were followed for a mean of 6 years. 5 Of these, 3640
participants were enrolled in the study for AMD. In the AREDS, daily doses of vitamin C (500
mg), vitamin E (400 IU), beta-carotene (15 mg), zinc (80 mg as zinc oxide), and copper (2 mg
as cupric oxide, to reduce the risk of zinc-induced copper deficiency anemia) were evaluated. In
the AREDS2, the replacement of beta-carotene with lutein (10 mg) and zeaxanthin (2 mg) was
explored, along with a lower dose (25 mg) of zinc oxide (see Table 1).

P19
18
 Age-Related Macular Degeneration PPP

The AREDS2 study was a multicenter, randomized, double-masked, placebo-controlled phase

III study that used a 2 x 2 factorial study design. 64 The study enrolled 4203 participants with
either bilateral large drusen or large drusen in one eye and advanced disease in the fellow eye.
This population represented a high-risk group for progression to more advanced stages as
identified in the original AREDS.157 Participants were randomized to receive either
supplemental lutein and zeaxanthin, supplemental omega-3, or the original formulation. A
secondary randomization to four variations included elimination of beta-carotene, lower zinc
levels (25 mg), or both. The final results of the AREDS2 support the recommendation for
substitution of beta-carotene with lutein (10 mg) and zeaxanthin (2 mg).

In the original AREDS and in AREDS2, participants who benefited from antioxidant vitamin
and mineral supplementation were those who had either intermediate AMD or advanced AMD
in one eye. For participants with extensive intermediate (i.e., medium-sized) drusen in one or
both eyes, one or more large drusen in at least one eye, nonsubfoveal geographic atrophy in one
eye, or advanced AMD (i.e., subfoveal geographic atrophy or CNV) in one eye, the rate of
development of advanced AMD at 5 years was reduced by 25% in the participants using the
combination treatment of antioxidant vitamins with zinc and copper. The risk of losing vision of
3 or more lines (doubling of the visual angle) was reduced by 19% with this combination
treatment. Although zinc alone or antioxidants alone reduced progression, the therapy that
resulted in a statistically significant reduction in both the development of advanced AMD and
vision loss was the combination treatment of antioxidant vitamins and minerals (Table 2).

19

P20
 Age-Related Macular Degeneration PPP

A meta-analysis by Evans in 2017 concluded that individuals with AMD may experience
delay in progression of the disease with antioxidant vitamin and mineral supplementation.158
This finding is drawn from one large trial conducted in a relatively well-nourished American
population. The generalizability of these findings to other populations is not known. Although
generally regarded as safe, vitamin supplements may have side effects.65 Evans also published
a second meta-analysis concluding that taking vitamin E or beta-carotene supplements will
not prevent or delay the onset of AMD. The same probably applies to vitamin C and the
multivitamin (Centrum Silver) investigated in the one trial reported to date. There is no
evidence with respect to other antioxidant supplements, such as lutein and zeaxanthin.158 A
meta-analysis of the adverse effects of nutritional supplementation reported that there is an
increased risk of death from vitamin A, beta-carotene, and vitamin E supplements (16%, 7%,
4%, respectively), but not from vitamin C supplements. 159 Other investigators have raised
concerns about the methodology for this meta-analysis. There is potential bias in the analyses
owing to the omission of clinical trials that had no deaths and the lack of biological
plausibility in the authors’ interpretation of the results of the subgroup analyses. 160-162 Also a
number of studies in the meta-analysis used antioxidant dosages much higher than those used
in the AREDS and did not find an adverse association of high-dose antioxidant
supplementation.163 Of great concern, two studies reported an increased mortality among
patients who were heavy smokers and were also taking beta-carotene supplements to prevent
lung cancer.164,165

The AREDS2 study results demonstrated that in patients at high risk for progression, there
was no statistically significant difference associated with supplementation with the original
AREDS formula versus each of the other modifications on AMD progression. As mentioned
earlier, the addition of omega-3 supplementation (DHA and EPA) had no further benefit. This
20

P21
 Age-Related Macular Degeneration PPP

result was also suggested by a meta-analysis by Chong et al in 2008. 73 Subgroup analysis

indicated that for those in the lowest quartile for lutein and zeaxanthin intake, supplemental
lutein and zeaxanthin was protective (95% CI, 0.59–0.94; P=0.01). The authors concluded
from all available evidence that lutein and zeaxanthin represent an appropriate substitute for
beta-carotene in the supplement.64 Finally, there was no demonstrated detrimental effect of
lowering the zinc levels (25 mg) on progression to advanced disease. 64 A meta-analysis by
Vishwanathan in 2013 did show that zinc supplementation alone may not be sufficient to
produce clinically meaningful changes in VA.166

With the introduction of the VEGF inhibitors pegaptanib sodium (Macugen®, Eyetech, Inc.,
Cedar Knolls, NJ) in 2004, off-label bevacizumab (Avastin®, Genentech, Inc., South San
Francisco, CA) in 2005, ranibizumab (Lucentis®, Genentech, Inc., South San Francisco, CA)
in 2006, and aflibercept (Eylea™, Regeneron Pharmaceuticals, Inc., Tarrytown, NY) in 2011,
more effective treatments for neovascular AMD exist. The VEGF inhibitors have
demonstrated improved visual and anatomic outcomes compared with other therapies. Anti-
VEGF therapies have become first-line therapy for treating and stabilizing most cases of
neovascular AMD and a Cochrane systematic review demonstrates the effectiveness of these
agents to maintain visual acuity.167 (I+, Good quality, Strong recommendation)

Aflibercept is a pan–VEGF-A and placental growth factor (PGF) blocker approved by the US
Food and Drug Administration (FDA) that has been documented to be of similar efficacy to
ranibizumab in the head-to-head phase III VEGF Trap-Eye: Investigation of Efficacy and
Safety in Wet AMD (VIEW) trials.168 In these pivotal studies, the currently approved 2-mg
dose of aflibercept was administered by intravitreal injection every 4 weeks and every 8
weeks after three monthly loading doses. In the first year, both study arms were similar to 0.5-
mg ranibizumab dosed every 4 weeks.

Bevacizumab is a full-length monoclonal antibody that binds all isoforms of VEGF. It is FDA
approved for intravenous use in the treatment of metastatic colorectal, metastatic breast, and
non-small cell lung cancer. Bevacizumab was investigated first as a systemic intravenous
treatment for AMD and then as an intravitreal injection (1.25 mg) before the FDA approved
ranibizumab.169,170 Because preliminary reports appeared favorable, ophthalmologists began
to use intravitreal bevacizumab off-label to treat CNV. Comparative trials and uncontrolled
case series reported improvements in VA and decreased retinal thickness by optical coherence
tomography (OCT) following intravitreal bevacizumab treatment.171-177 Informed consent
information is available on the benefits and risks of intravitreal bevacizumab and its off-label
status.178

21

P22
 Age-Related Macular Degeneration PPP

Intravitreal ranibizumab (0.5 mg) is FDA approved for the treatment of all subtypes of
neovascular AMD, based on results from three double-masked, randomized controlled
trials.179,180 (See Table 3.) Ranibizumab is a recombinant, humanized immunoglobulin G1
kappa isotype therapeutic antibody fragment developed for intraocular use. Ranibizumab
binds to and inhibits the biologic activity of all isoforms of human VEGF-A.

22

P23
 Age-Related Macular Degeneration PPP

23
≥
≤

P24
 Age-Related Macular Degeneration PPP

24
≤
≥

P25
 Age-Related Macular Degeneration PPP

The Comparison of AMD Treatment Trials (CATT) was a multicenter clinical trial that
compared the safety and effectiveness of bevacizumab with ranibizumab and an
individualized dosing regimen (as needed, or PRN) with monthly injections. At 1 year, the
CATT study found that ranibizumab and bevacizumab had comparable equivalence VA
improvements for monthly dosing.174 Ranibizumab PRN had similar VA improvements
compared with a fixed schedule of monthly injections. Further follow-up at 2 years showed
that the two drugs remained comparable in both efficacy and safety, but the PRN arms
together did not perform as well in terms of maintaining the visual gains at the end of year 1
compared with the two monthly arms, especially in the bevacizumab PRN group.183 The
CATT 5-year follow-up study demonstrated vision gains during the first 2 years that were not
maintained at 5 years. However, 50% of eyes had VA of 20/40 or better, confirming anti-
VEGF therapy as a major long-term therapeutic advance for neovascular AMD. 184 Similar
results were seen in the 2-year Inhibition of VEGF in Age-related choroidal
Neovascularization (IVAN) trial conducted in the United Kingdom.185,186 (See Glossary.)
Presently, there does not appear to be a significant difference in efficacy between ranibizumab
and bevacizumab.184 A meta-analysis by Nguyen in 2018 of over 8,000 eyes comparing all
three drugs concluded that bevacizumab and ranibizumab had equivalent efficacy for best-
corrected visual acuity (BCVA), whereas ranibizumab had greater reduction in central
macular thickness, and aflibercept and ranibizumab had comparable efficacy for BCVA and
central macular thickness.187 The review by Chen in 2015 also elicited similar results. 188 The
systemic safety data in the CATT and IVAN studies are inconclusive and two Cochrane
systematic reviews have also concluded that if a difference in safety between these anti-VEGF
drugs exists, it is minimal.189,190 (I+, Good quality, Strong recommendation) A real world
analysis of 13,859 patients found that all three agents improved visual acuity similarly over 1
year.191

Pegaptanib sodium is a selective VEGF antagonist that binds to the 165 isoform of VEGF-A.
It was the first anti-VEGF agent available for treating neovascular AMD. Pegaptanib sodium
injection is FDA approved for the treatment of all subtypes of neovascular AMD, with a
recommended dosage of 0.3 mg injected every 6 weeks into the vitreous. These
recommendations were based on results from two double-masked, randomized controlled
trials.181 (See Table 3.) Unlike the other anti-VEGF agents that are currently available
(ranibizumab, aflibercept, and bevacizumab), pegaptanib treatment does not improve VA on
average in patients with new-onset neovascular AMD and is rarely used in current clinical
practice.

Randomized clinical trials have been performed to study the adjunct use of intravitreal
corticosteroids and/or anti-VEGF agents in various drug combinations or with verteporfin

25

P26
 Age-Related Macular Degeneration PPP

PDT, following the publication of results from uncontrolled case series.192-195 However, the
data do not currently support the use of combination therapy with steroids, especially given
the long-term side effects of glaucoma and cataract that are associated with corticosteroid use.

The DENALI and MONT BLANC studies (ranibizumab and verteporfin PDT compared with
ranibizumab alone) did not show a significant benefit of adding PDT to anti-VEGF therapy in
new-onset neovascular AMD.196,197 (See Glossary.) However, the EVEREST study
demonstrated that fewer anti-VEGF injections were needed in combination therapy compared
with anti-VEGF monotherapy in eyes with the PCV variant of neovascular AMD.198 A 2017
meta-analysis and systematic review also concluded that treatment of PCV by PDT combined
with ranibizumab is valuable in improving VA and maintaining long-term effectiveness but
recommended further study.199,200 A randomized trial of 310 subjects has shown aflibercept to
effectively treat PCV in 85% of patients; 15% required PDT for control. 200 A 2018 meta-
analysis of 16 studies by Gao et al compared 587 patients in the monotherapy group with
various anti-VEGF agents against 673 patients in the combination group and found no
statistically significant difference between groups in mean BCVA, the proportion of patients
who gained 15 or more letters, or central retinal thickness at the end of the study. 201 However,
combination therapy did require fewer anti-VEGF injections, as noted in other studies with
reduced-fluence PDT demonstrating this reduction in number of injections at a statistically
significant level as opposed to the standard fluence group. 201

In addition to intravitreal injections of VEGF inhibitors, verteporfin PDT and thermal

laser photocoagulation surgery remain approved options for the treatment of subfoveal
lesions. Current practice patterns support the use of anti-VEGF monotherapy for patients
with newly diagnosed neovascular AMD and suggest that these other therapies are rarely
needed. Photodynamic therapy with verteporfin has FDA approval for the treatment of
AMD-related, predominantly classic, subfoveal CNV; treatment trial results are described
in Table 3. The efficacy of thermal laser photocoagulation surgery for CNV was studied
in the MPS (early 1990s) in a randomized, controlled, multicenter trial.148-151 The MPS
directly treated eyes that had subfoveal lesions using thermal laser surgery,150 but the
outcomes were poor and do not compare with the positive VA benefits found with current
anti-VEGF therapy. Thus, thermal laser photocoagulation surgery is no longer
recommended for subfoveal CNV treatment.

Table 3 (at the end of this section) summarizes the findings from randomized controlled
trials of verteporfin PDT and VEGF inhibitors for the treatment of subfoveal CNV. The

26

P27
 Age-Related Macular Degeneration PPP

entry criteria varied among these studies and may have contributed to the differences
among treatment cohorts.

Although randomized, controlled clinical trials have not routinely included patients with
juxtafoveal CNV, many clinicians extrapolated the data from current trials to consider
intravitreal injections of anti-VEGF agent as the primary therapy for juxtafoveal lesions.

In the MPS, treatment of well-demarcated juxtafoveal CNV lesions resulted in a small

overall treatment benefit.151 The rates of “persistence” (CNV leakage within 6 weeks of
laser photocoagulation surgery) and “recurrence” (CNV leakage more than 6 weeks after
laser photocoagulation surgery) were high (80%) at 5 years. After 5 years of follow-up,
52% of eyes treated for juxtafoveal lesions progressed to visual loss of 30 or more letters
(quadrupling of the visual angle) compared with 61% of untreated eyes. 151

There still remains a possible role for thermal laser surgery treatment in eyes with
extrafoveal and peripapillary CNV lesions as defined by the MPS.148,202 Although
photocoagulation of well-demarcated extrafoveal CNV lesions resulted in a substantial
reduction in the risk of severe visual loss for the first 2 years, recurrence or persistence
occurs in approximately 50% of cases, thus reducing this benefit over the subsequent 3
years of follow-up.148 After 5 years of follow-up, 48% of eyes treated for extrafoveal
lesions progressed to VA loss of 30 or more letters when compared with 62% of
untreated eyes.148 The historical data are important to recognize in current practice
patterns, as none of the anti-VEGF or PDT trials included extrafoveal lesions.
Practitioners have extrapolated and applied data from the dramatic improvements seen in
the treatment of subfoveal lesions to extrafoveal lesions. The current trend is to use anti-
VEGF agents in preference to laser photocoagulation surgery. Laser surgery for
extrafoveal lesions remains a less commonly used, yet reasonable, therapy. Current
therapies that have insufficient data to demonstrate clinical efficacy include radiation
therapy, acupuncture, electrical stimulation, macular translocation surgery, and
adjunctive use of intravitreal corticosteroids with verteporfin PDT. Therefore, at this
time, these therapies are not recommended.

27

P28
 Age-Related Macular Degeneration PPP

Patient outcome criteria are to reverse or minimize visual loss and improve visual function.

The initial evaluation of a patient with signs and symptoms suggestive of AMD includes all features
of the comprehensive adult medical eye evaluation,203 with particular attention to those aspects
relevant to AMD.

An initial history should consider the following elements:

◆ Symptoms204
◆ Metamorphopsia
◆ Decreased vision
◆ Scotoma
◆ Photopsia
◆ Difficulties in dark adaptation
◆ Medication and nutritional supplement use
◆ Ocular history11,205,206
◆ Medical history11,205,206 (including any hypersensitivity reactions181,207)
◆ Family history, especially family history of AMD 85,208
◆ Social history, especially a quantitative smoking history 39-43

◆ Comprehensive eye examination

◆ Amsler grid
◆ Stereoscopic biomicroscopic examination of the macula

Binocular slit-lamp biomicroscopy of the ocular fundus is often necessary to detect subtle
clinical signs of CNV. These include small areas of hemorrhage, hard exudates, subretinal fluid,
macular edema, subretinal fibrosis, or pigment epithelial elevation.

P29
28
 Age-Related Macular Degeneration PPP

Optical coherence tomography is important in diagnosing and managing AMD, particularly

with respect to determining the presence of subretinal and intraretinal fluid and in
documenting the degree of retinal thickening.209 Optical coherence tomography defines the
cross-sectional architecture of the retina, which is not possible with any other imaging
technology. It may reveal the presence of fluid that is not apparent on biomicroscopy alone.
It also helps in evaluating the response of the retina and RPE to therapy by allowing
structural changes to be followed accurately.210-213 Newer-generation OCT modalities,
including SD-OCT, are preferred technologies. Advances in OCT have increased the image
resolution and enhanced our ability to detect structural changes of the retina and
choroid.214-217 The implementation of newer technologies, such as swept-source OCT (that
is FDA approved), is evolving at this time.215-217

Optical coherence tomography angiography (OCTA) is a newer imaging modality that

provides noninvasive evaluation of the retinal and choroidal vasculature and is becoming
more commonly applied in the evaluation and management of AMD, but it has not
replaced other angiographic methods.218

Intravenous fundus fluorescein angiography is indicated 148,150,151 when the patient

complains of new metamorphopsia or has unexplained blurred vision, and/or when clinical
examination reveals elevation of the RPE or retina, macular edema, subretinal blood, hard
exudates, or subretinal fibrosis, or the OCT shows evidence of fluid. Fluorescein
angiography is also warranted as follows:

◆ To detect the presence of and determine the extent, type, size, and location of CNV. If
verteporfin PDT or laser photocoagulation surgery is being considered, the angiogram is
used as a guide to direct treatment. The role and indications for fluorescein angiography are
evolving as continued advances in OCT occur.
◆ To detect persistent or recurrent CNV or other retinal diseases following treatment. (See
Glossary.)
◆ To assist in determining the cause of visual loss that is not explained by the clinical
examination.

P30
29
 Age-Related Macular Degeneration PPP

If CNV is suspected on the basis of new symptoms or ocular findings, fluorescein

angiography should be performed and interpreted expeditiously by an individual
experienced in managing patients with neovascular AMD. 148,150,151

When fluorescein angiography is performed, the physician must be aware of potential risks
associated with this procedure:219,220 tissue infiltration (if the drug extravasates the vein),
pain, and allergic reactions. Even death from anaphylaxis has been reported (approximately
1 in 200,000 patients). Each angiographic facility should have a care plan in place for an
emergency situation as well as a clear protocol to minimize the risks and to manage
complications.

Color fundus photographs may be obtained when angiography is performed, because they
are useful in finding landmarks, evaluating serous detachments of the neurosensory retina
and RPE, and determining the etiology of blocked fluorescence. Fundus photographs may
also be used as a baseline reference for selected patients with advanced non-neovascular
AMD and for follow-up of treated patients.

Fundus autofluorescence is helpful to demonstrate areas of geographic atrophy and monitor

their progression. Some patterns of autofluorescence may predict faster rates of geographic
atrophy.221 Also, fundus autofluorescence may be used to quantify lipofuscin in the RPE. 221

Indocyanine green angiography is a technique that allows visualization of the choroidal

circulation. The value of this test in evaluating and treating AMD has been debated.222
Indocyanine green angiography has been shown to be useful in evaluating specific forms of
AMD, such as PED, poorly defined CNV, occult CNV, and lesions including retinal
angiomatous proliferation or idiopathic PCV.154,223 The PCV form of neovascular AMD
may be more easily identified when ICG is used, particularly in patients of African or
Asian descent.12,224 When ICG angiography is performed, the physician must be aware of
potential risks associated with this procedure: severe medical complications, allergic
reactions, and even death.225

Several other tests including microperimetry226 (to measure macular sensitivity), and
adaptive optics (to identify individual rods and cones)227 have been used to evaluate
patients with AMD; however, their specific role in clinical practice has yet to be
specifically defined.

P31
30
 Age-Related Macular Degeneration PPP

Consequences of untreated neovascular macular degeneration include a substantial economic burden

on patients, their family and society. Anti-VEGF agents are cost-effective for management of
neovascular macular degeneration. Choice of the anti-VEGF agent to use should be individually
tailored based on discussion between the patient and physician. Early detection and treatment of AMD
to arrest the deterioration in vision may help preserve patients’ quality of life and independence.
Management options for AMD include observation, antioxidant vitamin and mineral supplements,
intravitreal injection of anti-VEGF agents, PDT, and laser photocoagulation surgery. Several new
treatments such as stem cells and gene therapy are currently under investigation.228-230

Patients who are currently smoking should be advised to stop.231,232 Studies have found that the
physician’s advice to stop smoking is a helpful motivator for patients who are attempting to quit 231
and is associated with increased long-term smoking abstinence rates.232 An important component of
care for an AMD patient is referral for vision rehabilitation as well as continued follow-up for general
eye care.

Patients with early AMD and/or a family history of AMD should be encouraged to assess their
own VA using monocular vision testing (i.e., Amsler grid or electronic home monitoring233,234)
and have scheduled dilated eye examinations for detecting the intermediate stage of AMD. (See
Glossary.) Treatment with antioxidants and minerals as described previously in the original
AREDS and AREDS2 trials should be considered for patients who have progressed to
intermediate or advanced AMD in at least one eye.

Patients with a high-risk AMD phenotype are at increased risk of progression to advanced
AMD and should be educated about methods of detecting new symptoms of CNV, including
self-monitoring. They should also be educated about the need for promptly reporting new
symptoms to an ophthalmologist who can confirm if the new symptoms are from CNV and who
can begin any necessary treatment.

Follow-up examinations of patients at increased risk of progression to advanced AMD may

enable (1) early detection of asymptomatic and treatable neovascular lesions that could improve
or preserve VA, (2) education about the possible benefit of AREDS2-based nutritional
supplements, and (3) reinforcement of the need for self-monitoring and prompt evaluation with
the onset of new symptoms. Patients who check monocular near vision (reading/Amsler
grid/Amsler-grid equivalent) may be more likely to become aware of subtle visual symptoms
due to CNV, increasing the likelihood of detecting CNV at an early stage which, on average,

31

P32
 Age-Related Macular Degeneration PPP

yields better long-term visual outcomes with treatment compared with neovascular disease
detected at a more advanced stage.

Sensitivity and specificity for CNV detection with en face OCTA combined with cross-
sectional OCTA approaches that of the gold standard of fluorescein angiography with OCT, and
it is better than en face OCTA alone.235 Structural OCT alone has excellent sensitivity for CNV
detection. False positives from the structural OCT can be mitigated with the addition of flow
information with OCTA.235 Optical coherence tomography angiography may detect subclinical
CNV, which needs close monitoring and not treatment.179,218,236,237 Electronic monitoring
devices are now available to aid in the detection of neovascularization at an early stage. Such
devices use hyperacuity perimetry (or vernier acuity) to create a quantified central visual map
of metamorphopsia.238 Further studies of a variety of such devices are ongoing.

Assessment and treatment plans for non-neovascular and neovascular AMD are listed in Table
4. The criteria for treatment of AMD and the techniques of therapy are described in the
aflibercept, bevacizumab, ranibizumab, pegaptanib, MPS, and AREDS literature. Aflibercept,
ranibizumab, and pegaptanib-injection product labeling and other literature discuss techniques
of intravitreal injection.181,207,239-241 Recently, conbercept has shown promising results in the
management of wet AMD,242 although it has yet to receive FDA approval for its use. Similarly,
abicipar has completed phase II clinical trials and has shown an extended duration of effect with
a good safety profile; however, it has not received FDA approval. 243,244 Recently reported
results from the HAWK and HARRIER phase III clinical trials showed that brolucizumab
achieved its primary endpoint of noninferiority of BCVA change compared with aflibercept at
week 48. Patients treated with brolucizumab achieved superior reductions in central subfield
thickness compared with aflibercept. Fewer patients treated with brolucizumab had sub-retinal
fluid, inter-retinal fluid, and sub-RPE fluid. Brolucizumab received FDA approval in October
2019.245

As is the case with most clinical trials, these treatment trials do not provide clear guidance for
the management of all patients encountered in clinical practice. To date, the major prospective
randomized anti-VEGF treatment trials (Anti-VEGF Antibody for the Treatment of
Predominantly Classic CNV in AMD [ANCHOR], Minimally Classic/Occult Trial of the Anti-
VEGF Antibody Ranibizumab in the Treatment of Neovascular AMD [MARINA], VIEW,
CATT, IVAN, HARBOR) used either a fixed continuous treatment regimen (approximately
every 4 or 8 weeks) or an individualized discontinuous treatment regimen
(PRN).168,174,179,180,183,185,186,246

P33
 Age-Related Macular Degeneration PPP

•
•

33
•

P34
 Age-Related Macular Degeneration PPP

•
•

•
•

34
≤
•

P35
 Age-Related Macular Degeneration PPP

The PRN regimens using ranibizumab appear to have efficacy and safety comparable to fixed
monthly regimens over 1 year of treatment, but they do not maintain the initial visual gains with
longer follow-up.183,255 Caution should be used when dosing PRN bevacizumab, as it may be
slightly less effective than other monthly anti-VEGF regimens and other PRN anti-VEGF
regimens.183 Vision gains during the first 2 years of the CATT clinical trials were not
maintained at the 5-year follow-up visit, but 50% of the patients maintained a VA of 20/40.184
A continuous, variable dosing regimen that attempts to individualize therapy, commonly
referred to as “treat and extend,” is frequently used in clinical practice as an alternative to the
two treatment approaches above.248-251 Prospective studies such as Lucentis Compared to
Avastin Study (LUCAS) have shown similar efficacy between monthly and treat-and-extend for
bevacizumab and ranibizumab.256

Subretinal hemorrhages are relatively common in neovascular AMD. Small subretinal

hemorrhages are a sign of active CNV or PCV and may be managed with anti-VEGF therapy.
For the management of larger submacular hemorrhages, the SST study was inconclusive.
Pneumatic displacement procedures, the use of tPA, and/or pars plana vitrectomy have been
proposed. The data on management of these larger hemorrhages are inadequate to make a
recommendation at this time.257

The risks, benefits, and complications of the treatment and the alternatives to it should be
discussed with the patient and informed should be consent obtained.146,258

Possible complications of the four main modalities of treatment for AMD are listed below.
Retinal pigment epithelium rips (tears) may occur with or without these treatment modalities,
yet this is not a contraindication to continued anti-VEGF therapy.

All anti-VEGF treatments may carry theoretical risks for systemic arterial thromboembolic
events and increased intraocular pressure, although the results of clinical trials studying
these risks remain inconclusive.259-262 A recent review of the literature concluded that anti-
VEGF therapy is safe and effective for neovascular AMD. 263 The risks of intravitreal anti-
VEGF agents in pregnant or lactating women have not been studied.264,265 Intravitreal
pharmacotherapy can result in endophthalmitis, noninfectious inflammation, retinal tear, or
detachment.

◆ Aflibercept injection
 Endophthalmitis (cumulative ≤1.0% over 1 year in VIEW studies) 168

P36
35
 Age-Related Macular Degeneration PPP

At 1 year, there were no statistically significant differences in rates of serious systemic

adverse events such as death, arteriothrombotic events, or venous thrombotic events
between ranibizumab and aflibercept.168,266

◆ Bevacizumab injection
 Reported safety data are limited by relatively short and variable follow-up periods and
by differences in reporting criteria.267,268
 Reported ocular adverse events include bacterial endophthalmitis per injection
(0.16%), tractional retinal detachments (0.16%), uveitis (0.09%), rhegmatogenous
retinal detachment (0.02%), and vitreous hemorrhage (0.16%).240,269

The CATT study had limited statistical power to identify any differences in treatment-
related adverse events between bevacizumab and ranibizumab. At 1 year, there were no
statistically significant differences in rates of death, arteriothrombotic events, or venous
thrombotic events for the two drugs. There was a higher rate of serious systemic events
(e.g., arteriothrombotic events, venous thrombosis, or gastrointestinal disorders such as
hemorrhage) among patients treated with bevacizumab compared with ranibizumab (24%
vs. 19%; P=0.04), and this statistically significant difference was persistent at 2 years of
follow-up.174,183 The IVAN trial showed greater serum VEGF suppression with
bevacizumab but did not show any statistically significant difference in serious systemic
adverse events.185

◆ Ranibizumab injection
 Endophthalmitis (cumulative ≤1.0% over 2 years in MARINA study; <1.0% over 1
year in ANCHOR study)
 Retinal detachment or traumatic injury to the lens (<0.1% of treated cases during the
first year of treatment)179,180

◆ Pegaptanib sodium injection270

 Endophthalmitis (1.3% of treated cases during the first year of treatment)
 Traumatic injury to the lens (0.6% of treated cases during the first year of treatment)
 Retinal detachment (0.7% of treated cases during the first year of treatment)
 Anaphylaxis/anaphylactoid reactions including angioedema (rare; these were reported
following FDA approval)

◆ A severe decrease in central vision occurred within 1 week following treatment in 1% to

4% of patients, and may be permanent182,252,253

◆ Infusion site extravasation

◆ Idiosyncratic back pain during infusion of the drug (1%–2% of patients)182,252,253

P37
36
 Age-Related Macular Degeneration PPP

◆ Photosensitivity reaction (<3% of patients).182,252,253 The stated, current recommendations

are to avoid direct sunlight for the first 5 days after a treatment.

Verteporfin is contraindicated in patients with porphyria or a known allergy or sensitivity

to the drug. Careful consideration should be given to patients with liver dysfunction and to
patients who are pregnant, breastfeeding, or of pediatric age, because these patients were
not studied in published reports.

◆ Severe vision loss following treatment, which may be permanent

◆ Rupture of Bruch’s membrane with subretinal or vitreous hemorrhage

◆ Effects on the fovea in subfoveal or juxtafoveal CNV

Thermal laser is no longer recommended for subfoveal CNV. Introduction or enlargement

of a pre-existing scotoma, with or without VA loss, is not a complication of thermal laser
photocoagulation surgery; rather, it is an anticipated side effect of the treatment. Similarly,
recurrence or persistence of CNV, or the development of new CNV and further visual
deterioration after adequate thermal laser surgery, is usually a result of the disease process
and is not a complication. These realities must be emphasized to the patient and family
before treatment.

◆ Beta-carotene
 Self-reported yellowing of the skin (8.3% in the antioxidant arm compared with 6.0%
in the no antioxidant arm; P=0.008)5
 Increased risk of developing lung cancer in current smokers (an excess cumulative
incidence of lung cancer was observed after18 months and increased progressively
thereafter, resulting in an 18% difference in incidence by the end of the study (95% CI,
3%–36%; P=0.01) between the patients who received beta-carotene and those who did
not).164 The active treatment group had a relative risk of lung cancer of 1.28 (95% CI,
1.04–1.57; P=0.02), as compared with the placebo group.165

◆ Zinc
 Increased risk of hospitalizations for genitourinary causes, i.e., unspecified urinary
tract infection and prostatic hyperplasia in men and stress incontinence in women
(7.5% in those treated with zinc compared with 4.9% in those not treated with 80 mg
of zinc; P=0.001).5 In the AREDS2, there was no significant difference in AMD
progression between 80 mg and 25 mg of zinc.
 Copper-deficiency anemia (concomitant administration of copper is necessary;
included in the AREDS and AREDS2)

P38
37
 Age-Related Macular Degeneration PPP

When considering long-term supplementation, some people may have reason to avoid one
or more of the supplements evaluated in the original AREDS or AREDS2. Because of the
potential adverse effects, such as increased rate of genitourinary conditions that may
require hospitalizations, the high doses of antioxidant vitamins and minerals recommended
by the original AREDS and AREDS2 should be reviewed by the patient’s primary care
physician.

A history and examination are the recommended elements of the follow-up visits.
Recommended follow-up intervals are listed in Table 4.

The follow-up history should take into account the following:

◆ Symptoms, including decreased vision and metamorphopsia 204

◆ Changes in medications and nutritional supplements

◆ Changes in medical and ocular history11,205,206

◆ Changes in social history (smoking)39-43

The examination on the follow-up visit should include the following:

◆ VA at distance with correction

◆ Amsler grid

◆ Stereoscopic biomicroscopic examination of the fundus

In addition to the above recommendations, patients who have been treated with aflibercept,
bevacizumab, ranibizumab, or pegaptanib sodium injection; verteporfin PDT; or thermal
laser photocoagulation surgery should be examined at regular intervals by means of
biomicroscopy of the fundus. Optical coherence tomography,209 OCTA,271-274 fluorescein
angiography,148,150,151 and fundus photography may be helpful to detect signs of active
exudation or disease progression and should be used when clinically indicated. In common
clinical practice, OCT is a simple, noninvasive procedure that is well accepted by the
patient and provides important information for the provider to manage AMD.

Initial treatment and follow-up with intravitreal anti-VEGF therapy (aflibercept,

bevacizumab and ranibizumab) should be at approximately 4-week intervals.168,179,183

P39
38
 Age-Related Macular Degeneration PPP

Subsequent follow-up and treatment intervals vary depending on the clinical findings and
judgment of the treating ophthalmologist. After three loading doses administered at 4 week
intervals, a maintenance treatment regimen every 8 weeks with aflibercept has been shown
to have comparable efficacy to every 4 weeks of either ranibizumab and aflibercept in the
first year of therapy.168 There is no consensus about the ideal treatment intervals with anti-
VEGF agents. There are three protocols: monthly or bimonthly injections, treat-and-extend,
or PRN. A minority of retina specialists will treat patients monthly. Treat-and-extend is
based on anti-VEGF injection following an interval based on treatment response. As-
needed treatment is based on the presence or absence of subretinal or intraretinal fluid. The
few patients currently being treated with pegaptanib sodium injection should have follow-
up examinations approximately 6 weeks after each injection.

Subsequent examinations, OCT, OCTA, and fluorescein angiography should be performed

as indicated depending on the clinical findings and the judgment of the treating
ophthalmologist. Treated patients should be instructed to report symptoms of
endophthalmitis, retinal detachment, or decreased vision, and they should be re-examined
promptly.

For patients with unilateral disease, the fellow eye without CNV remains at high risk of
developing advanced AMD.275 The risk can be lowered by as much as 36% over a 10-year
period by taking the AREDS/AREDS2 supplements.5 Patients should be instructed to
monitor their vision and to return to the ophthalmologist periodically, even in the absence
of symptoms, but promptly after the onset of any new or significant visual symptoms.
Patients at exceptionally high risk (e.g., the presence of advanced AMD in one eye and
large drusen with RPE changes in the fellow eye) may be examined more frequently (i.e.,
every 6–12 months) in an effort to detect asymptomatic CNV at a treatable stage. Since
some patients with AMD also have cognitive impairment, a family member or care
assistant should prompt the patient to self-test. Optical coherence tomography is useful and
OCTA may be useful for evaluating the status of high-risk fellow eyes.

Ancillary clinical personnel should be aware that patients with the onset of new symptoms suggestive
of AMD (e.g., new visual loss, metamorphopsia, or scotoma) should be examined promptly. The
ophthalmologist will perform most of the examination and all treatment, and certain aspects of the
testing may be conducted by other trained individuals under the ophthalmologist's supervision.

P40
39
 Age-Related Macular Degeneration PPP

All patients with AMD should be educated about the prognosis of the disease and the potential value
of treatment as appropriate for their visual and functional status. Patients can be informed that while
central visual loss is common, total visual loss is extremely rare. Patients with AMD can be reassured
that there is no harm in using their eyes for normal visual tasks, and they may be told that the effect of
total sunlight exposure remains uncertain. Insofar as cigarette smoking is a key modifiable risk factor,
smoking cessation is strongly recommended when advising patients with AMD or at risk for AMD.

The informed consent process should include a discussion of the risks and benefits of treatment and
treatment alternatives. The off-label status of bevacizumab for neovascular AMD should be included
in the discussion; information and a consent form are available from the Ophthalmic Mutual
Insurance Company.178

Vision rehabilitation optimizes the patient’s functional ability,276 and patients with reduced visual
function should be referred for vision rehabilitation and social services.277 Patients with severe visual
loss related to AMD who are referred for vision rehabilitation services often have unrealistic
expectations. Educating patients that the visual rehabilitation specialist helps to optimize their existing
visual function, rather than “helping them see better” will establish more appropriate expectations
around such services. Special optical or electronic magnifying lenses, bright lights, and electronic
reading aids may help patients to read more effectively, but not as well as they did before the onset of
AMD. An Implantable Miniature Telescope (IMT) is an FDA-approved device that may be effective
for screened, phakic, motivated patients with end-stage AMD, and it appears to be cost-effective.278,279
A systematic review in 2018 found insufficient evidence on the IMT’s safety and effectiveness in
patients with late or advanced AMD.267 More information on vision rehabilitation, including materials
for patients, is available at www.aao.org/low-vision-and-vision-rehab.

Loss of VA increases the risk of frequent falls.280,281 Depression and visual hallucinations (Charles
Bonnet syndrome) frequently accompany severe central vision loss. Patients who have Charles
Bonnet syndrome and their family members should be informed that visual symptoms are not unusual
and do not represent a sign of psychosis or mental deterioration. The ophthalmologist may inquire
about symptoms of clinical depression and, when appropriate, suggest that the patient seek
professional advice, as depression may exacerbate the effects of AMD. 282

Direct medical costs (taken from private insurance and Medicare claims data) related to treatment for
AMD in the United States were estimated to be approximately $574 million in 2004.283 However,
these studies were conducted prior to the use of anti-VEGF agents.

The considerable burden of disease associated with AMD, as well as the public health benefits of
prevention, are highlighted in analyses conducted by the AREDS authors. This research, published in
2003, estimated that 8 million Americans aged 55 and older are at high risk for developing advanced

P41
40
 Age-Related Macular Degeneration PPP

AMD. If these persons received AREDS-formulation supplements, it was estimated that

approximately 300,000 would avoid advanced AMD and any associated vision loss over a 5-year
period.284 In the Salisbury Eye Study, Christ et al reported that VA loss adversely affected activities of
daily living levels which subsequently increased mortality risk in older adults, further calculations
estimated that treating AMD with anti-VEGF agents saves 1 to 2 years of life.285

More recent cost-effectiveness studies on the use of anti-VEGF therapies have demonstrated this
newer therapy to be highly cost-effective over prior therapies such as PDT.286 287-290 The off-label use
of intravitreal bevacizumab was suggested to represent a highly cost-effective, off-label option for
management of neovascular AMD compared with the higher cost of ranibizumab.289 Others have
investigated the cost utility of various treatments for AMD. One analysis using CATT trial data found
that bevacizumab with PRN dosing offered considerably greater value than ranibizumab in the
treatment of neovascular AMD among patients 80 and older. 290 Another analysis using CATT and
MARINA data evaluated the relative 10-year cost-effectiveness of bevacizumab and ranibizumab in
65-year-old patients with neovascular AMD. This study estimated the cost utility of bevacizumab
treatment (relative to no treatment) at approximately $2,700 per quality-adjusted life year ([QALY]
for monthly dosing) and $3,300 per QALY (for PRN dosing). In contrast, the cost-effectiveness of
ranibizumab was estimated as $63,300/QALY for monthly dosing and $18,600 per QALY for PRN
dosing.287 Wholesale prices of anti-VEGF medications range from $50 to $1,950 per dose, depending
on the medication.291,292 The use of personalized anti-VEGF treatment guided by OCT has resulted in
savings for the US government in neovascular AMD patients of $9 billion and $22 billion,
respectively.293

P42
41
 Age-Related Macular Degeneration PPP

Providing quality care

is the physician's foremost ethical obligation, and is
the basis of public trust in physicians.
AMA Board of Trustees, 1986

Quality ophthalmic care is provided in a manner and with the skill that is consistent with the best interests of
the patient. The discussion that follows characterizes the core elements of such care.
The ophthalmologist is first and foremost a physician. As such, the ophthalmologist demonstrates
compassion and concern for the individual, and utilizes the science and art of medicine to help alleviate
patient fear and suffering. The ophthalmologist strives to develop and maintain clinical skills at the highest
feasible level, consistent with the needs of patients, through training and continuing education. The
ophthalmologist evaluates those skills and medical knowledge in relation to the needs of the patient and
responds accordingly. The ophthalmologist also ensures that needy patients receive necessary care directly or
through referral to appropriate persons and facilities that will provide such care, and he or she supports
activities that promote health and prevent disease and disability.
The ophthalmologist recognizes that disease places patients in a disadvantaged, dependent state. The
ophthalmologist respects the dignity and integrity of his or her patients and does not exploit their
vulnerability.
Quality ophthalmic care has the following optimal attributes, among others.
◆ The essence of quality care is a meaningful partnership relationship between patient and physician. The
ophthalmologist strives to communicate effectively with his or her patients, listening carefully to their
needs and concerns. In turn, the ophthalmologist educates his or her patients about the nature and
prognosis of their condition and about proper and appropriate therapeutic modalities. This is to ensure
their meaningful participation (appropriate to their unique physical, intellectual, and emotional state) in
decisions affecting their management and care, to improve their motivation and compliance with the
agreed plan of treatment, and to help alleviate their fears and concerns.
◆ The ophthalmologist uses his or her best judgment in choosing and timing appropriate diagnostic and
therapeutic modalities as well as the frequency of evaluation and follow-up, with due regard to the
urgency and nature of the patient's condition and unique needs and desires.
◆ The ophthalmologist carries out only those procedures for which he or she is adequately trained,
experienced, and competent, or, when necessary, is assisted by someone who is, depending on the
urgency of the problem and availability and accessibility of alternative providers.
◆ Patients are assured access to, and continuity of, needed and appropriate ophthalmic care, which can be
described as follows.
 The ophthalmologist treats patients with due regard to timeliness, appropriateness, and his or her own
ability to provide such care.
 The operating ophthalmologist makes adequate provision for appropriate pre- and postoperative
patient care.
 When the ophthalmologist is unavailable for his or her patient, he or she provides appropriate alternate
ophthalmic care, with adequate mechanisms for informing patients of the existence of such care and
procedures for obtaining it.
 The ophthalmologist refers patients to other ophthalmologists and eye care providers based on the
timeliness and appropriateness of such referral, the patient's needs, the competence and qualifications
of the person to whom the referral is made, and access and availability.
 The ophthalmologist seeks appropriate consultation with due regard to the nature of the ocular or other
medical or surgical problem. Consultants are suggested for their skill, competence, and accessibility.
P43
42
 Age-Related Macular Degeneration PPP

They receive as complete and accurate an accounting of the problem as necessary to provide efficient
and effective advice or intervention, and in turn they respond in an adequate and timely manner. The
ophthalmologist maintains complete and accurate medical records.
 On appropriate request, the ophthalmologist provides a full and accurate rendering of the patient's
records in his or her possession.
 The ophthalmologist reviews the results of consultations and laboratory tests in a timely and effective
manner and takes appropriate actions.
 The ophthalmologist and those who assist in providing care identify themselves and their profession.
 For patients whose conditions fail to respond to treatment and for whom further treatment is
unavailable, the ophthalmologist provides proper professional support, counseling, rehabilitative and
social services, and referral as appropriate and accessible.
◆ Prior to therapeutic or invasive diagnostic procedures, the ophthalmologist becomes appropriately
conversant with the patient's condition by collecting pertinent historical information and performing
relevant preoperative examinations. Additionally, he or she enables the patient to reach a fully informed
decision by providing an accurate and truthful explanation of the diagnosis; the nature, purpose, risks,
benefits, and probability of success of the proposed treatment and of alternative treatment; and the risks
and benefits of no treatment.
◆ The ophthalmologist adopts new technology (e.g., drugs, devices, surgical techniques) in judicious
fashion, appropriate to the cost and potential benefit relative to existing alternatives and to its
demonstrated safety and efficacy.
◆ The ophthalmologist enhances the quality of care he or she provides by periodically reviewing and
assessing his or her personal performance in relation to established standards, and by revising or altering
his or her practices and techniques appropriately.
◆ The ophthalmologist improves ophthalmic care by communicating to colleagues, through appropriate
professional channels, knowledge gained through clinical research and practice. This includes alerting
colleagues of instances of unusual or unexpected rates of complications and problems related to new
drugs, devices, or procedures.
◆ The ophthalmologist provides care in suitably staffed and equipped facilities adequate to deal with
potential ocular and systemic complications requiring immediate attention.
◆ The ophthalmologist also provides ophthalmic care in a manner that is cost effective without
unacceptably compromising accepted standards of quality.

Reviewed by: Council

Approved by: Board of Trustees
October 12, 1988
2nd Printing: January 1991
3rd Printing: August 2001
4th Printing: July 2005

P44
43
 Age-Related Macular Degeneration PPP

Age-related macular degeneration, which includes entities with the following ICD-9 and ICD-10
classifications (see Glossary):

•
•

•
•
•

P45
44
 Age-Related Macular Degeneration PPP

Advanced age-related macular degeneration (advanced AMD): This is the most severe form of AMD,
defined as geographic atrophy involving the center of the macula (fovea) or features of CNV.

Age-Related Eye Disease Study (AREDS): A prospective, multicenter, randomized clinical trial designed to
assess the natural course and risk factors of age-related cataract and AMD and the effects of antioxidants and
minerals on these two conditions.

Age-Related Eye Disease Study (AREDS2): A prospective, multicenter, randomized clinical trial of 4000
participants designed to assess the effects of oral supplementation of high doses of macular xanthophylls
(lutein and zeaxanthin) and/or omega-3 long-chain polyunsaturated fatty acids (docosahexaenoic acid and
eicosapentaenoic acid) for the treatment of AMD and cataract. All participants were offered the AREDS
supplements. A secondary randomization evaluated the possibility of deleting beta-carotene and decreasing
the original levels of zinc in the AREDS formulation. Follow-up occurs over 5 years.

Age-related macular degeneration (AMD): There is no universally accepted definition of this term. The
condition is characterized by the presence of drusen and alterations of the RPE as well as by the fundus
abnormalities associated with CNV, and it generally occurs in persons over age 65. The VA may vary from
normal to severe impairment.

AMD: See Age-related macular degeneration.

Amsler grid: This is a graph paper with a central dot for fixation. While viewing this central spot, the patient
is asked to evaluate vision for the early signs of metamorphopsia by looking for any changes in the grid.

ANCHOR Study: Anti-VEGF antibody (ranibizumab) for the treatment of predominantly classic CNV in
AMD study.

Anti-VEGF: See Anti-vascular endothelial growth factor.

Anti-vascular endothelial growth factor (VEGF): Substances that inhibit the action of vascular endothelial
growth factor protein.

AREDS: See Age-Related Eye Disease Study (AREDS).

AREDS2: See Age-Related Eye Disease Study (AREDS2).

Bevacizumab (Avastin): Bevacizumab is a full-length monoclonal antibody that binds all isoforms of VEGF
and has FDA approval for intravenous use in the treatment of metastatic colorectal, metastatic breast, and
non-small cell lung cancer.

CATT: See Comparison of AMD Treatment Trials.

Choroidal neovascularization (CNV): Synonymous with subretinal or choroidal neovascular membrane.

These are vessels from the choriocapillaris that perforate and grow through Bruch’s membrane and enter the
subretinal pigment epithelial and/or subretinal spaces.

Classic choroidal neovascularization: The angiographic findings in which the CNV is recognized in the
early phase of the fluorescein angiogram as an area of bright, well-demarcated hyperfluorescence and during
the late phases of the angiogram as progressive pooling of dye in the overlying subsensory retinal space.
Usually considered a Gass Type 2 membrane.

CNV: See Choroidal neovascularization.

P46
45
 Age-Related Macular Degeneration PPP

Comparison of AMD Treatment Trials (CATT): A multicenter clinical trial that compared the safety and
efficacy of bevacizumab and ranibizumab and an individualized dosing regimen (PRN) to monthly
injections.

DENALI study: Part of the SUMMIT studies, this trial compares ranibizumab and verteporfin PDT
combination therapy with ranibizumab alone.

Disc area: As defined by the Macular Photocoagulation Study, the area of a circle with a diameter of 1.5
millimeters (1500 µm) equal to 1.77 square millimeters. The area on a photograph will vary with the type of
fundus camera used.

Disciform scar: Subretinal fibrovascular tissue that usually becomes more fibrous within a few years and that
is often the end result of CNV.

Drusen: Yellow lesions at the level of the basement membrane of the RPE. They are the ophthalmoscopic
and histologic hallmark of AMD. They are considered to be small if they are less than 63 µm in diameter,
intermediate if they are greater than or equal to 63 and less than or equal to 125 µm, and large when the
diameter is greater than 125 µm, and they may be considered soft if they have ill-defined edges.

EVEREST study: A study conducted in Asia that investigated combination PDT and anti-VEGF therapy.

Extrafoveal choroidal neovascularization: A choroidal neovascular membrane that comes no closer than
200 µm from the center of the foveal avascular zone, as defined by the Macular Photocoagulation Study.

Foveal avascular zone: An area usually 300 to 500 millimeters in diameter centered on the foveola and
lacking retinal blood vessels, also known as the capillary-free zone.

Geographic atrophy: One or several well-demarcated zones of RPE atrophy (and sometimes choriocapillaris
atrophy). Drusen are usually present surrounding these zones and there may be surrounding pigment
clumping. This is an advanced form of AMD when the center of the fovea is involved.

HARBOR study: A 12-month dose-comparison study of 0.5 mg and 2 mg of ranibizumab. It also compared
monthly to PRN treatment over 2 years.

ICD-9: International Statistical Classification of Diseases and Related Health Problems, Ninth Edition.

ICD-10: International Statistical Classification of Diseases and Related Health Problems, Tenth Edition.

ICG: See Indocyanine green.

Indocyanine green (ICG): A cyanine dye that fluoresces in the near-infrared spectrum and is used in
diagnostic evaluation to visualize CNV.

Inhibition of VEGF in Age-related choroidal Neovascularization (IVAN trial): This study compared
intravitreal bevacizumab with ranibizumab dosed either on a continuous (monthly) or discontinuous (PRN)
basis. It was a 2-year study conducted in the United Kingdom.

IVAN trial: See Inhibition of VEGF in Age-related choroidal Neovascularization.

Juxtafoveal choroidal neovascularization: Well-demarcated CNV that is between 1 µm and 199 µm from
the center of the foveal avascular zone but that does not reach its center, as defined by the Macular
Photocoagulation Study.

LUCAS: Lucentis Compared to Avastin Study

Macular Photocoagulation Study (MPS): A series of prospective randomized multicenter clinical trials
designed to determine the efficacy of laser photocoagulation surgery in CNV caused by AMD, ocular
histoplasmosis, and idiopathic causes.

P47
46
 Age-Related Macular Degeneration PPP

Macular translocation: An operation designed to move the sensory retina from an area of damaged RPE to
another area of more intact RPE.

MARINA study: Study of minimally classic/occult trial of the anti-VEGF antibody, ranibizumab, in the
treatment of neovascular AMD.

MONT BLANC study: Part of the SUMMIT study, this European trial compares ranibizumab and
verteporfin PDT combination treatment with ranibizumab alone.

MPS: See Macular Photocoagulation Study.

Neovascular macular degeneration: Manifestations of CNV and/or RPE detachment associated with
subretinal serous fluid, exudates, and/or blood.

Occult choroidal neovascularization: Angiographic findings characterized by a fibrovascular RPE

detachment and/or late leakage of an undetermined source. This is also referred to as poorly defined CNV
that has indistinct or poorly demarcated boundaries on fluorescein angiography. Usually considered a Gass
Type 1 membrane.

Optical coherence tomography (OCT): A noninvasive technique to image intraocular tissues by measuring
the echo time delay and intensity of back-reflected light. The resulting image provides high-resolution, cross-
sectional representation of structure with near-histological detail.

Optical coherence tomography angiography (OCTA): A non-invasive imaging technique for the
microvasculature of the retina and choroid.

PDT: See Photodynamic therapy.

PED: See Pigment epithelial detachment.

Pegaptanib sodium (Macugen): A compound that binds to a specific isoform of vascular endothelial growth
factor (VEGF165) and thus blocks its activity. It is administered by intravitreal injection.

Persistent choroidal neovascularization: Angiographically documented CNV found within 6 weeks of laser
surgery, typically but not always at the site of the previously treated CNV, according to the Macular
Photocoagulation Study definition.

Photodynamic therapy (PDT): A method of treating CNV with a two-part process involving systemic
administration of a photosensitizing drug followed by nonthermal light application to the macular pathology.

Pigment epithelial detachment (PED): Accumulation of fluid (serous RPE detachment) or blood
(hemorrhagic RPE detachment) beneath the RPE. Associated CNV is usually present in older patients and/or
patients with drusen. Another form is the fibrovascular pigment epithelial detachment, which is a form of
occult CNV.

PGF: See Placental growth factor.

Placental growth factor (PGF): A growth factor related to VEGF that may play a role in ocular
angiogenesis.

Polypoidal choroidopathy: Characterized by multiple and recurrent serosanguineous RPE detachments,

which often resemble hemorrhagic detachment in AMD. A fluorescein angiogram and ICG may be helpful in
distinguishing these conditions.

Predominantly classic lesion: CNV in which classic CNV occupies more than 50% of the entire lesion area.

Ranibizumab (Lucentis): A recombinant humanized immunoglobulin G1 kappa isotype therapeutic antibody

fragment that binds to and inhibits the biologic activity of a form of VEGF-A.

P48
47
 Age-Related Macular Degeneration PPP

Recurrent choroidal neovascularization: Angiographically documented CNV found more than 6 weeks after
laser surgery and typically occurring on the perimeter of the previous treatment scar, as defined by the
Macular Photocoagulation Study.

Reticular pseudodrusen: Also referred to as subretinal drusenoid deposits.

Retinal angiomatous proliferation: Characterized by proliferation of retinal capillaries in the paramacular

area that may present as intraretinal, subretinal, or CNV.

Retinal pigment epithelial (RPE) abnormalities: Alterations of the retinal pigment epithelium-Bruch’s
membrane complex that lead to an appearance of hypopigmentation and/or hyperpigmentation. Its extreme
form is geographic atrophy.

RPE: See Retinal pigment epithelium (RPE) abnormalities.

Severe visual loss: In this document, severe visual loss means quadrupling or more of the visual angle (e.g.,
20/20 to 20/80 or worse, or 20/50 to 20/200 or worse).

Subfoveal choroidal neovascularization: CNV that underlies the center of the foveal avascular zone.

SST: See Submacular Surgery Trial.

Submacular Surgery Trial (SST): A trial conducted in the mid-1990s, prior to the emergence of currently
used therapies, that evaluated the efficacy of submacular surgery for treating complications of CNV and
subretinal hemorrhage.

Subretinal drusenoid deposits: See Reticular pseudodrusen.

SUMMIT: Two studies, called DENALI in North America and MONT BLANC in Europe, that compare
ranibizumab and verteporfin PDT combination therapy with ranibizumab alone.

Vascular endothelial growth factor (VEGF): A significant mediator in the process of angiogenesis and
increased vascular permeability and inflammation. It has been identified in neovascularization related to both
diabetic retinopathy and AMD. In animal models, the introduction of VEGF has initiated the cascade of
neovascularization seen in AMD. Thus, the inhibition or antagonism of the action of VEGF is a targeted area
of research, with several novel therapeutic agents being developed, and in various stages of investigation and
FDA approval.

VEGF: See Vascular endothelial growth factor.

Verteporfin (Visudyne): A drug used as a photosensitizer in conjunction with a nonthermal PDT laser.

VIEW Study: VEGF Trap-Eye: Investigation of Efficacy and Safety in Wet AMD

48
P49
 Age-Related Macular Degeneration PPP

Literature searches of the PubMed and Cochrane databases were conducted in March 2018; the search
strategies are provided at www.aao.org/ppp. Specific limited update searches were conducted after June 2019.
"Macular Degeneration/diagnosis"[Mesh]

("Macular Degeneration/epidemiology"[Mesh] OR "Macular

Degeneration/ethnology"[Mesh])

"Macular Degeneration/genetics"[Mesh]

"Macular Degeneration"[Mesh] AND "Risk Factors"[Mesh]

("Macular Degeneration/therapy"[Mesh] AND "Quality of Life"[Mesh]) OR ("Macular

Degeneration"[Mesh]) AND ("Quality of Life"[Mesh] not treatment) OR ("Macular
Degeneration"[Mesh] AND "Cost of Illness"[Mesh])

("Macular Degeneration/economics"[Mesh] OR ("Macular Degeneration"[Mesh] AND

"Cost-Benefit Analysis"[Mesh])) NOT "Cost of Illness"[Mesh]

("Macular Degeneration/therapy"[Mesh] OR ("Macular Degeneration"[Mesh] AND (combinations[tiab] OR

combined[tiab])) OR (("Drug Therapy, Combination"[Mesh] OR "Drug Combinations"[Mesh]) OR
"Combined Modality Therapy"[Mesh])

49
P50
 Age-Related Macular Degeneration PPP

Basic and Clinical Science Course

Retina and Vitreous (Section 12, 2019–2020)

Focal Points
Neovascular Age-Related Macular Degeneration (2016)
Masquerades of Age-related Macular Degeneration (2018)

Ophthalmic Technology Assessment –

Published in Ophthalmology, which is distributed free to Academy members; links to full text available
at www.aao.org/ota.
Safety and Efficacy of Anti-Vascular Endothelial Growth Factor Therapies for Neovascular Age-Related
Macular Degeneration (2018)

Patient Education
Age-Related Macular Degeneration Brochure (AMD) (2014)
AMD and Nutritional Supplements Brochure (2014)
Anti-VEGF Treatment for AMD Brochure (2014)

Preferred Practice Pattern® Guidelines – Free download available at www.aao.org/ppp.

Comprehensive Adult Medical Eye Evaluation (2015)

To order any of these products, except for the free materials, please contact the Academy’s Customer Service
at 866.561.8558 (U.S. only) or 415.561.8540 or www.aao.org/store

50

P51
 Age-Related Macular Degeneration PPP

1. Bourne RR, Sorensen KE, Klauber A, Foster PJ, Johnson GJ, Alsbirk PH. Glaucoma in East
Greenlandic Inuit--a population survey in Ittoqqortoormiit (Scoresbysund). Acta Ophthalmol Scand.
2001;79(5):462-467.
2. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of
evidence and strength of recommendations. Bmj. 2008;336(7650):924-926.
3. Bourne RR, Sukudom P, Foster PJ, et al. Prevalence of glaucoma in Thailand: a population based
survey in Rom Klao District, Bangkok. Br J Ophthalmol. 2003;87(9):1069-1074.
4. Ferris FL, III, Wilkinson CP, Bird A, et al. Clinical classification of age-related macular
degeneration. Ophthalmology. 2013;120(4):844-851.
5. Age-Related Eye Disease Study Research Group. A randomized, placebo-controlled, clinical trial of
high-dose supplementation with vitamins C and E, beta carotene, and zinc for age-related macular
degeneration and vision loss: AREDS report number 8. Arch Ophthalmol. 2001;119(10):1417-1436.
6. Gass JDM. Stereoscopic Atlas of Macular Diseases: Diagnosis and Treatment. 4th ed. St. Louis,
MO: CV Mosby; 1997.
7. Ryan SJ, Hinton DR, Schachat AP, Wilkinson CP, eds. Retina. 4th ed. St. Louis, MO: CV Mosby;
2005.
8. Klein R, Klein BE, Linton KL. Prevalence of age-related maculopathy. The Beaver Dam Eye Study.
Ophthalmology. 1992;99(6):933-943.
9. Kahn HA, Leibowitz HM, Ganley JP, et al. The Framingham Eye Study. I. Outline and major
prevalence findings. Am J Epidemiol. 1977;106(1):17-32.
10. Sommer A, Tielsch JM, Katz J, et al. Racial differences in the cause-specific prevalence of
blindness in east Baltimore. N Engl J Med. 1991;325(20):1412-1417.
11. Klein BE, Klein R. Cataracts and macular degeneration in older Americans. Arch Ophthalmol.
1982;100(4):571-573.
12. Friedman DS, O'Colmain BJ, Munoz B, et al. Prevalence of age-related macular degeneration in the
United States. Arch Ophthalmol. 2004;122(4):564-572.
13. Congdon N, O'Colmain B, Klaver CC, et al. Causes and prevalence of visual impairment among
adults in the United States. Arch Ophthalmol. 2004;122(4):477-485.
14. Vincent GK, Velkoff VA. The next four decades, the older population in the United States: 2010 to
2050. 2010; P25-1138. Available at: www.census.gov/prod/2010pubs/p25-1138.pdf. Accessed
September 2019.
15. Wong WL, Su X, Li X, et al. Global prevalence of age-related macular degeneration and disease
burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet Glob Health.
2014;2(2):e106-116.
16. Bressler NM, Doan QV, Varma R, et al. Estimated cases of legal blindness and visual impairment
avoided using ranibizumab for choroidal neovascularization: non-Hispanic white population in the
United States with age-related macular degeneration. Arch Ophthalmol. 2011;129(6):709-717.
17. Rofagha S, Bhisitkul RB, Boyer DS, Sadda SR, Zhang K. Seven-year outcomes in ranibizumab-
treated patients in ANCHOR, MARINA, and HORIZON: a multicenter cohort study (SEVEN-UP).
Ophthalmology. 2013;120(11):2292-2299.
18. Age-Related Eye Disease Study 2 (AREDS2) Research Group, Chew EY, SanGiovanni JP, Ferris
FL, et al. Lutein/zeaxanthin for the treatment of age-related cataract: AREDS2 randomized trial
report number 4. JAMA Ophthalmol. 2013;131(7):843-850.
19. Rein DB, Wittenborn JS, Zhang X, et al. Forecasting age-related macular degeneration through the
year 2050: the potential impact of new treatments. Arch Ophthalmol. 2009;127(4):533-540.
20. Ambati J, Ambati BK, Yoo SH, Ianchulev S, Adamis AP. Age-related macular degeneration:
etiology, pathogenesis, and therapeutic strategies. Surv Ophthalmol. 2003;48(3):257-293.
21. Ferris FL, III, Fine SL, Hyman L. Age-related macular degeneration and blindness due to
neovascular maculopathy. Arch Ophthalmol. 1984;102(11):1640-1642.

P52
51
 Age-Related Macular Degeneration PPP

22. Age-Related Eye Disease Study Research Group. Risk factors associated with age-related macular
degeneration. A case-control study in the age-related eye disease study: Age-Related Eye Disease
Study report number 3. Ophthalmology. 2000;107(12):2224-2232.
23. Klein R, Klein BE, Knudtson MD, et al. Prevalence of age-related macular degeneration in 4
racial/ethnic groups in the multi-ethnic study of atherosclerosis. Ophthalmology. 2006;113(3):373-
380.
24. Cotter SA, Varma R, Ying-Lai M, Azen SP, Klein R. Causes of low vision and blindness in adult
Latinos: the Los Angeles Latino Eye Study. Ophthalmology. 2006;113(9):1574-1582.
25. Klein R, Klein BE, Tomany SC, Meuer SM, Huang GH. Ten-year incidence and progression of age-
related maculopathy: The Beaver Dam Eye Study. Ophthalmology. 2002;109(10):1767-1779.
26. Varma R, Fraser-Bell S, Tan S, Klein R, Azen SP. Los Angeles Latino Eye Study Group.
Prevalence of age-related macular degeneration in Latinos: the Los Angeles Latino Eye Study.
Ophthalmology. 2004;111(7):1288-1297.
27. Munoz B, Klein R, Rodriguez J, Snyder R, West SK. Prevalence of age-related macular
degeneration in a population-based sample of Hispanic people in Arizona: Proyecto VER. Arch
Ophthalmol. 2005;123(11):1575-1580.
28. Schachat AP, Hyman L, Leske MC, Connell AM, Wu SY. The Barbados Eye Study Group. Features
of age-related macular degeneration in a black population. Arch Ophthalmol. 1995;113(6):728-735.
29. Friedman DS, Katz J, Bressler NM, Rahmani B, Tielsch JM. Racial differences in the prevalence of
age-related macular degeneration: the Baltimore Eye Survey. Ophthalmology. 1999;106(6):1049-
1055.
30. Jampol LM, Tielsch J. Race, macular degeneration, and the Macular Photocoagulation Study. Arch
Ophthalmol. 1992;110(12):1699-1700.
31. Yang K, Liang YB, Gao LQ, et al. Prevalence of age-related macular degeneration in a rural
Chinese population: the Handan Eye Study. Ophthalmology. 2011;118(7):1395-1401.
32. You QS, Xu L, Yang H, et al. Five-year incidence of age-related macular degeneration: the Beijing
Eye Study. Ophthalmology. 2012;119(12):2519-2525.
33. Kawasaki R, Yasuda M, Song SJ, et al. The prevalence of age-related macular degeneration in
Asians: a systematic review and meta-analysis. Ophthalmology. 2010;117(5):921-927.
34. Klein R, Chou CF, Klein BE, Zhang X, Meuer SM, Saaddine JB. Prevalence of age-related macular
degeneration in the US population. Arch Ophthalmol. 2011;129(1):75-80.
35. Nakata I, Yamashiro K, Nakanishi H, et al. Prevalence and characteristics of age-related macular
degeneration in the Japanese population: the Nagahama study. Am J Ophthalmol. 2013;156(5):1002-
1009.
36. Delcourt C. Epidemiology of AMD. In: Bandello F, ed. AMD: Age-Related Macular Degeneration.
Loures, Portugal: Théa Portugal, SA; 2010:13-21.
37. Tomany SC, Wang JJ, Van Leeuwen R, et al. Risk factors for incident age-related macular
degeneration: pooled findings from 3 continents. Ophthalmology. 2004;111(7):1280-1287.
38. Thornton J, Edwards R, Mitchell P, Harrison RA, Buchan I, Kelly SP. Smoking and age-related
macular degeneration: a review of association. Eye. 2005;19:935-944.
39. Khan JC, Thurlby DA, Shahid H, et al. Smoking and age related macular degeneration: the number
of pack years of cigarette smoking is a major determinant of risk for both geographic atrophy and
choroidal neovascularisation. Br J Ophthalmol. 2006;90(1):75-80.
40. Seddon JM, George S, Rosner B. Cigarette smoking, fish consumption, omega-3 fatty acid intake,
and associations with age-related macular degeneration: the US Twin Study of Age-Related Macular
Degeneration. Arch Ophthalmol. 2006;124(7):995-1001.
41. Fraser-Bell S, Wu J, Klein R, Azen SP, Varma R. Smoking, alcohol intake, estrogen use, and age-
related macular degeneration in Latinos: the Los Angeles Latino Eye Study. Am J Ophthalmol.
2006;141(1):79-87.
42. Tan JS, Mitchell P, Kifley A, Flood V, Smith W, Wang JJ. Smoking and the long-term incidence of
age-related macular degeneration: the Blue Mountains Eye Study. Arch Ophthalmol.
2007;125(8):1089-1095.
43. Klein R, Knudtson MD, Cruickshanks KJ, Klein BE. Further observations on the association
between smoking and the long-term incidence and progression of age-related macular degeneration:
the Beaver Dam Eye Study. Arch Ophthalmol. 2008;126(1):115-121.
44. Clemons TE, Milton RC, Klein R, Seddon JM, Ferris FL, III. Age-Related Eye Disease Study
Research Group. Risk factors for the incidence of advanced age-related macular degeneration in the

P53
52
 Age-Related Macular Degeneration PPP

Age-Related Eye Disease Study (AREDS): AREDS report number 19. Ophthalmology.
2005;112(4):533-539.
45. Complications of Age-related Macular Degeneration Prevention Trial (CAPT) Research Group.
Risk factors for choroidal neovascularization and geographic atrophy in the complications of age-
related macular degeneration prevention trial. Ophthalmology. 2008;115(9):1474-1479.
46. Hurley SF, Matthews JP, Guymer RH. Cost-effectiveness of smoking cessation to prevent age-
related macular degeneration. Cost Eff Resour Alloc. 2008;6:18.
47. Christen WG, Glynn RJ, Manson JE, Ajani UA, Buring JE. A prospective study of cigarette
smoking and risk of age-related macular degeneration in men. JAMA. 1996;276(14):1147-1151.
48. Delcourt C, Michel F, Colvez A, Lacroux A, Delage M, Vernet MH. Associations of cardiovascular
disease and its risk factors with age-related macular degeneration: the POLA Study. Ophthalmic
Epidemiol. 2001;8(4):237-249.
49. Fernandez AB, Wong TY, Klein R, et al. Age-related macular degeneration and incident
cardiovascular disease: the Multi-Ethnic Study of Atherosclerosis. Ophthalmology.
2012;119(4):765-770.
50. Hyman L, Schachat AP, He Q, Leske MC. Age-Related Macular Degeneration Risk Factors Study
Group. Hypertension, cardiovascular disease, and age-related macular degeneration. Arch
Ophthalmol. 2000;118(3):351-358.
51. Keilhauer CN, Fritsche LG, Guthoff R, Haubitz I, Weber BH. Age-related macular degeneration and
coronary heart disease: evaluation of genetic and environmental associations. Eur J Med Genet.
2013;56(2):72-79.
52. Klein R, Deng Y, Klein BE, et al. Cardiovascular disease, its risk factors and treatment, and age-
related macular degeneration: Women's Health Initiative Sight Exam ancillary study. Am J
Ophthalmol. 2007;143(3):473-483.
53. McCarty CA, Mukesh BN, Fu CL, Mitchell P, Wang JJ, Taylor HR. Risk factors for age-related
maculopathy: the Visual Impairment Project. Arch Ophthalmol. 2001;119(10):1455-1462.
54. Olea JL, Tunon J. Patients with neovascular age-related macular degeneration in Spain display a
high cardiovascular risk. Eur J Ophthalmol. 2012;22(3):404-411.
55. Mares-Perlman JA, Fisher AI, Klein R, et al. Lutein and zeaxanthin in the diet and serum and their
relation to age-related maculopathy in the third national health and nutrition examination survey.
Am J Epidemiol. 2001;153(5):424-432.
56. Delcourt C, Cristol JP, Tessier F, Leger CL, Descomps B, Papoz L. POLA Study Group. Age-
related macular degeneration and antioxidant status in the POLA Study, Pathologies Oculaires Liees
a l'Age. Arch Ophthalmol. 1999;117(10):1384-1390.
57. Cho E, Stampfer MJ, Seddon JM, et al. Prospective study of zinc intake and the risk of age-related
macular degeneration. Ann Epidemiol. 2001;11(5):328-336.
58. van Leeuwen R, Boekhoorn S, Vingerling JR, et al. Dietary intake of antioxidants and risk of age-
related macular degeneration. JAMA. 2005;294(24):3101-3107.
59. Age-Related Eye Disease Study Research Group, SanGiovanni JP, Chew EY, Clemons TE, et al.
The relationship of dietary carotenoid and vitamin A, E, and C intake with age-related macular
degeneration in a case-control study: AREDS report number 22. Arch Ophthalmol.
2007;125(9):1225-1232.
60. Chong EW, Wong TY, Kreis AJ, Simpson JA, Guymer RH. Dietary antioxidants and primary
prevention of age related macular degeneration: systematic review and meta-analysis. BMJ.
2007;335(7623):755.
61. Moriarty-Craige SE, Adkison J, Lynn M, et al. Antioxidant supplements prevent oxidation of
cysteine/cystine redox in patients with age-related macular degeneration. Am J Ophthalmol.
2005;140(6):1020-1026.
62. Age-Related Eye Disease Study Research Group. A randomized, placebo-controlled, clinical trial of
high-dose supplementation with vitamins C and E and beta carotene for age-related cataract and
vision loss: AREDS report number 9. Arch Ophthalmol. 2001;119(10):1439-1452.
63. Miller ER, III, Pastor-Barriuso R, Dalal D, Riemersma RA, Appel LJ, Guallar E. Meta-analysis:
high-dosage vitamin E supplementation may increase all-cause mortality. Ann Intern Med.
2005;142(1):37-46.
64. Age-Related Eye Disease Study 2 Research Group. Lutein + zeaxanthin and omega-3 fatty acids for
age-related macular degeneration: the Age-Related Eye Disease Study 2 (AREDS2) randomized
clinical trial. JAMA. 2013;309(19):2005-2015.

P54
53
 Age-Related Macular Degeneration PPP

65. Evans JR, Lawrenson JG. Antioxidant vitamin and mineral supplements for slowing the progression
of age-related macular degeneration. Cochrane Database Syst Rev. 2017;7:CD000254.
66. Seddon JM, Rosner B, Sperduto RD, et al. Dietary fat and risk for advanced age-related macular
degeneration. Arch Ophthalmol. 2001;119(8):1191-1199.
67. Mares-Perlman JA, Brady WE, Klein R, VandenLangenberg GM, Klein BE, Palta M. Dietary fat
and age-related maculopathy. Arch Ophthalmol. 1995;113(6):743-748.
68. Smith W, Mitchell P, Leeder SR. Dietary fat and fish intake and age-related maculopathy. Arch
Ophthalmol. 2000;118(3):401-404.
69. Cho E, Hung S, Willett WC, et al. Prospective study of dietary fat and the risk of age-related
macular degeneration. Am J Clin Nutr. 2001;73(2):209-218.
70. Chua B, Flood V, Rochtchina E, Wang JJ, Smith W, Mitchell P. Dietary fatty acids and the 5-year
incidence of age-related maculopathy. Arch Ophthalmol. 2006;124(7):981-986.
71. SanGiovanni JP, Agron E, Meleth AD, et al. Age-Related Eye Disease Study Research Group.
{omega}-3 Long-chain polyunsaturated fatty acid intake and 12-y incidence of neovascular age-
related macular degeneration and central geographic atrophy: AREDS report 30, a prospective
cohort study from the Age-Related Eye Disease Study. Am J Clin Nutr. 2009;90(6):1601-1607.
72. SanGiovanni JP, Chew EY, Clemons TE, et al. Age-Related Eye Disease Study Research Group.
The relationship of dietary lipid intake and age-related macular degeneration in a case-control study:
AREDS report number 20. Arch Ophthalmol. 2007;125(5):671-679.
73. Chong EW, Kreis AJ, Wong TY, Simpson JA, Guymer RH. Dietary omega-3 fatty acid and fish
intake in the primary prevention of age-related macular degeneration: a systematic review and meta-
analysis. Arch Ophthalmol. 2008;126(6):826-833.
74. Hogg RE, Woodside JV, McGrath A, et al. Mediterranean Diet Score and Its Association with Age-
Related Macular Degeneration: The European Eye Study. Ophthalmology. 2017;124(1):82-89.
75. Merle BMJ, Colijn JM, Cougnard-Gregoire A, et al. Mediterranean Diet and Incidence of Advanced
Age-Related Macular Degeneration: The EYE-RISK Consortium. Ophthalmology.
2019;126(3):381-390.
76. Klein BE, Howard KP, Gangnon RE, Dreyer JO, Lee KE, Klein R. Long-term use of aspirin and
age-related macular degeneration. JAMA. 2012;308(23):2469-2478.
77. de Jong PT, Wang JJ, Fletcher AE. Aspirin use and aging macula disorder. JAMA Ophthalmol.
2014;132(1):9-10.
78. Christen WG, Glynn RJ, Ajani UA, et al. Age-related maculopathy in a randomized trial of low-
dose aspirin among US physicians. Arch Ophthalmol. 2001;119(8):1143-1149.
79. Zhu W, Wu Y, Xu D, et al. Aspirin use and risk of age-related macular degeneration: a meta-
analysis. PLoS One. 2013;8(3):e58821.
80. Rim TH, Yoo TK, Kwak J, et al. Long-Term Regular Use of Low-Dose Aspirin and Neovascular
Age-Related Macular Degeneration: National Sample Cohort 2010-2015. Ophthalmology.
2019;126(2):274-282.
81. Modjtahedi BS, Fong DS, Jorgenson E, Van Den Eeden SK, Quinn V, Slezak JM. The Relationship
Between Nonsteroidal Anti-inflammatory Drug Use and Age-related Macular Degeneration. Am J
Ophthalmol. 2018;188:111-122.
82. Hyman LG, Lilienfeld AM, Ferris FL, III, Fine SL. Senile macular degeneration: a case-control
study. Am J Epidemiol. 1983;118(2):213-227.
83. Piguet B, Wells JA, Palmvang IB, Wormald R, Chisholm IH, Bird AC. Age-related Bruch's
membrane change: a clinical study of the relative role of heredity and environment. Br J
Ophthalmol. 1993;77(7):400-403.
84. Silvestri G, Johnston PB, Hughes AE. Is genetic predisposition an important risk factor in age-
related macular degeneration? Eye. 1994;8 (Pt 5):564-568.
85. Seddon JM, Ajani UA, Mitchell BD. Familial aggregation of age-related maculopathy. Am J
Ophthalmol. 1997;123(2):199-206.
86. Meyers SM. A twin study on age-related macular degeneration. Trans Am Ophthalmol Soc.
1994;92:775-843.
87. Hammond CJ, Webster AR, Snieder H, Bird AC, Gilbert CE, Spector TD. Genetic influence on
early age-related maculopathy: a twin study. Ophthalmology. 2002;109(4):730-736.
88. Seddon JM, Cote J, Page WF, Aggen SH, Neale MC. The US twin study of age-related macular
degeneration: relative roles of genetic and environmental influences. Arch Ophthalmol.
2005;123(3):321-327.

P55
54
 Age-Related Macular Degeneration PPP

89. Klein RJ, Zeiss C, Chew EY, et al. Complement factor H polymorphism in age-related macular
degeneration. Science. 2005;308(5720):385-389.
90. Edwards AO, Ritter R, III, Abel KJ, Manning A, Panhuysen C, Farrer LA. Complement factor H
polymorphism and age-related macular degeneration. Science. 2005;308(5720):421-424.
91. Haines JL, Hauser MA, Schmidt S, et al. Complement factor H variant increases the risk of age-
related macular degeneration. Science. 2005;308(5720):419-421.
92. Hageman GS, Anderson DH, Johnson LV, et al. A common haplotype in the complement regulatory
gene factor H (HF1/CFH) predisposes individuals to age-related macular degeneration. Proc Natl
Acad Sci U S A. 2005;102(20):7227-7232.
93. Zareparsi S, Branham KE, Li M, et al. Strong association of the Y402H variant in complement
factor H at 1q32 with susceptibility to age-related macular degeneration. Am J Hum Genet.
2005;77(1):149-153.
94. Gold B, Merriam JE, Zernant J, et al. Variation in factor B (BF) and complement component 2 (C2)
genes is associated with age-related macular degeneration. Nat Genet. 2006;38(4):458-462.
95. Despriet DD, Klaver CC, Witteman JC, et al. Complement factor H polymorphism, complement
activators, and risk of age-related macular degeneration. JAMA. 2006;296(3):301-309.
96. Yates JR, Sepp T, Matharu BK, et al. Complement C3 variant and the risk of age-related macular
degeneration. N Engl J Med. 2007;357(6):553-561.
97. Yang Z, Camp NJ, Sun H, et al. A variant of the HTRA1 gene increases susceptibility to age-related
macular degeneration. Science. 2006;314(5801):992-993.
98. Dewan A, Liu M, Hartman S, et al. HTRA1 promoter polymorphism in wet age-related macular
degeneration. Science. 2006;314(5801):989-992.
99. Kanda A, Chen W, Othman M, et al. A variant of mitochondrial protein LOC387715/ARMS2, not
HTRA1, is strongly associated with age-related macular degeneration. Proc Natl Acad Sci U S A.
2007;104(41):16227-16232.
100. Wang G, Spencer KL, Court BL, et al. Localization of age-related macular degeneration-associated
ARMS2 in cytosol, not mitochondria. Invest Ophthalmol Vis Sci. 2009;50(7):3084-3090.
101. Neale BM, Fagerness J, Reynolds R, et al. Genome-wide association study of advanced age-related
macular degeneration identifies a role of the hepatic lipase gene (LIPC). Proc Natl Acad Sci U S A.
2010;107(16):7395-7400.
102. Yang Z, Stratton C, Francis PJ, et al. Toll-like receptor 3 and geographic atrophy in age-related
macular degeneration. N Engl J Med. 2008;359(14):1456-1463.
103. Cho Y, Wang JJ, Chew EY, et al. Toll-like receptor polymorphisms and age-related macular
degeneration: replication in three case-control samples. Invest Ophthalmol Vis Sci.
2009;50(12):5614-5618.
104. Fritsche LG, Igl W, Bailey JN, et al. A large genome-wide association study of age-related macular
degeneration highlights contributions of rare and common variants. Nat Genet. 2016;48(2):134-143.
105. Francis PJ, Klein ML. Update on the role of genetics in the onset of age-related macular
degeneration. Clin Ophthalmol. 2011;5:1127-1133.
106. Fritsche LG, Chen W, Schu M, et al. AMD Gene Consortium. Seven new loci associated with age-
related macular degeneration. Nat Genet. 2013;45(4):433-439.
107. Seddon JM. Macular Degeneration Epidemiology: Nature-Nurture, Lifestyle Factors, Genetic Risk,
and Gene-Environment Interactions - The Weisenfeld Award Lecture. Invest Ophthalmol Vis Sci.
2017;58(14):6513-6528.
108. Awh CC, Lane AM, Hawken S, Zanke B, Kim IK. CFH and ARMS2 genetic polymorphisms
predict response to antioxidants and zinc in patients with age-related macular degeneration.
Ophthalmology. 2013;120(11):2317-2323.
109. Awh CC, Hawken S, Zanke BW. Treatment response to antioxidants and zinc based on CFH and
ARMS2 genetic risk allele number in the Age-Related Eye Disease Study. Ophthalmology.
2015;122(1):162-169.
110. Chew EY, Klein ML, Clemons TE, Agron E, Abecasis GR. Genetic testing in persons with age-
related macular degeneration and the use of the AREDS supplements: to test or not to test?
Ophthalmology. 2015;122(1):212-215.
111. Chew EY, Klein ML, Clemons TE, et al. Age-Related Eye Disease Study Research Group. No
clinically significant association between CFH and ARMS2 genotypes and response to nutritional
supplements: AREDS report number 38. Ophthalmology. 2014;121(11):2173-2180.

P56
55
 Age-Related Macular Degeneration PPP

112. Klein ML, Francis PJ, Rosner B, et al. CFH and LOC387715/ARMS2 genotypes and treatment with
antioxidants and zinc for age-related macular degeneration. Ophthalmology. 2008;115(6):1019-
1025.
113. Assel MJ, Li F, Wang Y, Allen AS, Baggerly KA, Vickers AJ. Genetic Polymorphisms of CFH and
ARMS2 Do Not Predict Response to Antioxidants and Zinc in Patients with Age-Related Macular
Degeneration: Independent Statistical Evaluations of Data from the Age-Related Eye Disease Study.
Ophthalmology. 2018;125(3):391-397.
114. Yamashiro K, Mori K, Honda S, et al. A prospective multicenter study on genome wide associations
to ranibizumab treatment outcome for age-related macular degeneration. Sci Rep. 2017;7(1):9196.
115. Dedania VS, Grob S, Zhang K, Bakri SJ. Pharmacogenomics of response to anti-VEGF therapy in
exudative age-related macular degeneration. Retina. 2015;35(3):381-391.
116. Vavvas DG, Small KW, Awh CC, Zanke BW, Tibshirani RJ, Kustra R. CFH and ARMS2 genetic
risk determines progression to neovascular age-related macular degeneration after antioxidant and
zinc supplementation. Proc Natl Acad Sci U S A. 2018;115(4):E696-E704.
117. Adams MK, Simpson JA, Aung KZ, et al. Abdominal obesity and age-related macular degeneration.
Am J Epidemiol. 2011;173(11):1246-1255.
118. Seddon JM, George S, Rosner B, Rifai N. Progression of age-related macular degeneration:
prospective assessment of C-reactive protein, interleukin 6, and other cardiovascular biomarkers.
Arch Ophthalmol. 2005;123(6):774-782.
119. Schaumberg DA, Christen WG, Buring JE, Glynn RJ, Rifai N, Ridker PM. High-sensitivity C-
reactive protein, other markers of inflammation, and the incidence of macular degeneration in
women. Arch Ophthalmol. 2007;125(3):300-305.
120. Laine M, Jarva H, Seitsonen S, et al. Y402H polymorphism of complement factor H affects binding
affinity to C-reactive protein. J Immunol. 2007;178(6):3831-3836.
121. Klein BE, Klein R, Jensen SC, Ritter LL. Are sex hormones associated with age-related
maculopathy in women? The Beaver Dam Eye Study. Trans Am Ophthalmol Soc. 1994;92:289-297.
122. Smith W, Mitchell P, Wang JJ. Gender, oestrogen, hormone replacement and age-related macular
degeneration: results from the Blue Mountains Eye Study. Aust N Z J Ophthalmol. 1997;25
(suppl):13-15.
123. Snow KK, Cote J, Yang W, Davis NJ, Seddon JM. Association between reproductive and hormonal
factors and age-related maculopathy in postmenopausal women. Am J Ophthalmol.
2002;134(6):842-848.
124. Vingerling JR, Dielemans I, Witteman JC, Hofman A, Grobbee DE, de Jong PT. Macular
degeneration and early menopause: a case-control study. BMJ. 1995;310(6994):1570-1571.
125. Feskanich D, Cho E, Schaumberg DA, Colditz GA, Hankinson SE. Menopausal and reproductive
factors and risk of age-related macular degeneration. Arch Ophthalmol. 2008;126(4):519-524.
126. Delcourt C, Carriere I, Ponton-Sanchez A, Fourrey S, Lacroux A, Papoz L. POLA Study Group.
Light exposure and the risk of age-related macular degeneration: the Pathologies Oculaires Liees a
l'Age (POLA) Study. Arch Ophthalmol. 2001;119(10):1463-1468.
127. Cruickshanks KJ, Klein R, Klein BE, Nondahl DM. Sunlight and the 5-year incidence of early age-
related maculopathy: the Beaver Dam Eye Study. Arch Ophthalmol. 2001;119(2):246-250.
128. Khan JC, Shahid H, Thurlby DA, et al. Age related macular degeneration and sun exposure, iris
colour, and skin sensitivity to sunlight. Br J Ophthalmol. 2006;90(1):29-32.
129. Cho E, Hankinson SE, Willett WC, et al. Prospective study of alcohol consumption and the risk of
age-related macular degeneration. Arch Ophthalmol. 2000;118(5):681-688.
130. Moss SE, Klein R, Klein BE, Jensen SC, Meuer SM. Alcohol consumption and the 5-year incidence
of age-related maculopathy: the Beaver Dam Eye Study. Ophthalmology. 1998;105(5):789-794.
131. Chong EW, Kreis AJ, Wong TY, Simpson JA, Guymer RH. Alcohol consumption and the risk of
age-related macular degeneration: a systematic review and meta-analysis. Am J Ophthalmol.
2008;145(4):707-715.
132. Gopinath B, Flood VM, Rochtchina E, Wang JJ, Mitchell P. Homocysteine, folate, vitamin B-12,
and 10-y incidence of age-related macular degeneration. Am J Clin Nutr. 2013;98(1):129-135.
133. Millen AE, Voland R, Sondel SA, et al. Vitamin D status and early age-related macular
degeneration in postmenopausal women. Arch Ophthalmol. 2011;129(4):481-489.
134. Gehlbach P, Li T, Hatef E. Statins for age-related macular degeneration. Cochrane Database Syst
Rev. 2016(8):CD006927.

P57
56
 Age-Related Macular Degeneration PPP

135. Chew EY, Clemons TE, Agron E, et al. Age-Related Eye Disease Study Research Group. Ten-year
follow-up of age-related macular degeneration in the age-related eye disease study: AREDS report
number 36. JAMA Ophthalmol. 2014;132(3):272-277.
136. Chew EY, Davis MD, Seddon JM, Clemons TE, Hubbard LD. Age-Related Eye Disease Study
Research Group. The effect of antioxidant and zinc supplements on change in drusen size/area in the
Age-Related Eye Disease Study (AREDS). Invest Ophthalmol Vis Sci. 2002;43:E-Abstract 1903.
137. Ferris FL, Davis MD, Clemons TE, et al. Age-Related Eye Disease Study (AREDS) Research
Group. A simplified severity scale for age-related macular degeneration: AREDS report number 18.
Arch Ophthalmol. 2005;123(11):1570-1574.
138. Klein ML, Francis PJ, Ferris FL, III, Hamon SC, Clemons TE. Risk assessment model for
development of advanced age-related macular degeneration. Arch Ophthalmol. 2011;129(12):1543-
1550.
139. Curcio CA, Messinger JD, Sloan KR, McGwin G, Medeiros NE, Spaide RF. Subretinal drusenoid
deposits in non-neovascular age-related macular degeneration: morphology, prevalence, topography,
and biogenesis model. Retina. 2013;33(2):265-276.
140. Sarks J, Arnold J, Ho IV, Sarks S, Killingsworth M. Evolution of reticular pseudodrusen. Br J
Ophthalmol. 2011;95(7):979-985.
141. Zweifel SA, Imamura Y, Spaide TC, Fujiwara T, Spaide RF. Prevalence and significance of
subretinal drusenoid deposits (reticular pseudodrusen) in age-related macular degeneration.
Ophthalmology. 2010;117(9):1775-1781.
142. Arnold JJ, Sarks SH, Killingsworth MC, Sarks JP. Reticular pseudodrusen. A risk factor in age-
related maculopathy. Retina. 1995;15(3):183-191.
143. Mimoun G, Soubrane G, Coscas G. Macular drusen [in French]. J Fr Ophtalmol. 1990;13(10):511-
530.
144. Ueda-Arakawa N, Ooto S, Tsujikawa A, Yamashiro K, Oishi A, Yoshimura N. Sensitivity and
specificity of detecting reticular pseudodrusen in multimodal imaging in Japanese patients. Retina.
2013;33(3):490-497.
145. Klein R, Klein BE, Jensen SC, Meuer SM. The five-year incidence and progression of age-related
maculopathy: the Beaver Dam Eye Study. Ophthalmology. 1997;104(1):7-21.
146. Solomon SD, Jefferys JL, Hawkins BS, Bressler NM, Bressler SB. Submacular Surgery Trials
Research Group. Risk factors for second eye progression to advanced age-related macular
degeneration: SST report number 21. Retina. 2009;29(8):1080-1090.
147. Sunness JS, Rubin GS, Applegate CA, et al. Visual function abnormalities and prognosis in eyes
with age-related geographic atrophy of the macula and good visual acuity. Ophthalmology.
1997;104(10):1677-1691.
148. Macular Photocoagulation Study Group. Argon laser photocoagulation for neovascular
maculopathy: five-year results from randomized clinical trials. Arch Ophthalmol. 1991;109(8):1109-
1114.
149. Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions in
age-related macular degeneration: results of a randomized clinical trial. Arch Ophthalmol.
1991;109(9):1220-1231.
150. Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions of
age-related macular degeneration: updated findings from two clinical trials. Arch Ophthalmol.
1993;111(9):1200-1209.
151. Macular Photocoagulation Study Group. Laser photocoagulation for juxtafoveal choroidal
neovascularization: five-year results from randomized clinical trials. Arch Ophthalmol.
1994;112(4):500-509.
152. Thomas R, Parikh R, Muliyil J, Kumar RS. Five-year risk of progression of primary angle closure to
primary angle closure glaucoma: a population-based study. Acta Ophthalmol Scand.
2003;81(5):480-485.
153. Yannuzzi LA, Sorenson J, Spaide RF, Lipson B. Idiopathic polypoidal choroidal vasculopathy
(IPCV). Retina. 1990;10(1):1-8.
154. Yannuzzi LA, Negrao S, Iida T, et al. Retinal angiomatous proliferation in age-related macular
degeneration. Retina. 2001;21(5):416-434.
155. Holz FG, Sadda SR, Busbee B, et al. Efficacy and Safety of Lampalizumab for Geographic Atrophy
Due to Age-Related Macular Degeneration: Chroma and Spectri Phase 3 Randomized Clinical
Trials. JAMA Ophthalmol. 2018;136(6):666-677.

P58
57
 Age-Related Macular Degeneration PPP

156. Evans JR, Lawrenson JG. Antioxidant vitamin and mineral supplements for preventing age-related
macular degeneration. Cochrane Database Syst Rev. 2012(6):CD000253.
157. Davis MD, Gangnon RE, Lee LY, et al. Age-Related Eye Disease Study Group. The Age-Related
Eye Disease Study severity scale for age-related macular degeneration: AREDS report number 17.
Arch Ophthalmol. 2005;123(11):1484-1498.
158. Evans JR, Lawrenson JG. Antioxidant vitamin and mineral supplements for preventing age-related
macular degeneration. Cochrane Database Syst Rev. 2017;7:CD000253.
159. Bjelakovic G, Nikolova D, Gluud LL, Simonetti RG, Gluud C. Mortality in randomized trials of
antioxidant supplements for primary and secondary prevention: systematic review and meta-
analysis. JAMA. 2007;297(8):842-857.
160. Albanes D. Antioxidant supplements and mortality. JAMA. 2007;298(4):400; author reply 402-403.
161. Hemila H. Antioxidant supplements and mortality. JAMA. 2007;298(4):401; author reply 402-403.
162. Taylor PR, Dawsey S. Antioxidant supplements and mortality. JAMA. 2007;298(4):401-402; author
reply 402-403.
163. Clemons TE, Kurinij N, Sperduto RD. AREDS Research Group. Associations of mortality with
ocular disorders and an intervention of high-dose antioxidants and zinc in the Age-Related Eye
Disease Study: AREDS report number 13. Arch Ophthalmol. 2004;122(5):716-726.
164. Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. The effect of vitamin E and beta
carotene on the incidence of lung cancer and other cancers in male smokers. N Engl J Med.
1994;330(15):1029-1035.
165. Omenn GS, Goodman GE, Thornquist MD, et al. Effects of a combination of beta carotene and
vitamin A on lung cancer and cardiovascular disease. N Engl J Med. 1996;334(18):1150-1155.
166. Vishwanathan R, Chung M, Johnson EJ. A systematic review on zinc for the prevention and
treatment of age-related macular degeneration. Invest Ophthalmol Vis Sci. 2013;54(6):3985-3998.
167. Solomon SD, Lindsley K, Vedula SS, Krzystolik MG, Hawkins BS. Anti-vascular endothelial
growth factor for neovascular age-related macular degeneration. Cochrane Database Syst Rev.
2019;3:CD005139.
168. Heier JS, Brown DM, Chong V, et al. VIEW 1 and VIEW 2 Study Groups. Intravitreal aflibercept
(VEGF trap-eye) in wet age-related macular degeneration. Ophthalmology. 2012;119(12):2537-
2548.
169. Michels S, Rosenfeld PJ, Puliafito CA, Marcus EN, Venkatraman AS. Systemic bevacizumab
(Avastin) therapy for neovascular age-related macular degeneration twelve-week results of an
uncontrolled open-label clinical study. Ophthalmology. 2005;112(6):1035-1047.
170. Avery RL, Pieramici DJ, Rabena MD, Castellarin AA, Nasir MA, Giust MJ. Intravitreal
bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology.
2006;113(3):363-372.
171. Bashshur ZF, Haddad ZA, Schakal A, Jaafar RF, Saab M, Noureddin BN. Intravitreal bevacizumab
for treatment of neovascular age-related macular degeneration: a one-year prospective study. Am J
Ophthalmol. 2008;145(2):249-256.
172. Arevalo JF, Sanchez JG, Wu L, et al. Pan-American Collaborative Retina Study Group. Intravitreal
bevacizumab for subfoveal choroidal neovascularization in age-related macular degeneration at
twenty-four months: the Pan-American Collaborative Retina Study. Ophthalmology.
2010;117(10):1974-1981.
173. Lai TY, Liu DT, Chan KP, Luk FO, Pang CP, Lam DS. Visual outcomes and growth factor changes
of two dosages of intravitreal bevacizumab for neovascular age-related macular degeneration: a
randomized, controlled trial. Retina. 2009;29(9):1218-1226.
174. Martin DF, Maguire MG, Ying GS, Grunwald JE, Fine SL, Jaffe GJ. Comparison of Age-related
Macular Degeneration Treatments Trials (CATT) Research Group. Ranibizumab and bevacizumab
for neovascular age-related macular degeneration. N Engl J Med. 2011;364(20):1897-1908.
175. Modarres M, Naseripour M, Falavarjani KG, Nikeghbali A, Hashemi M, Parvaresh MM. Intravitreal
injection of 2.5 mg versus 1.25 mg bevacizumab (Avastin) for treatment of CNV associated with
AMD. Retina. 2009;29(3):319-324.
176. Subramanian ML, Abedi G, Ness S, et al. Bevacizumab vs ranibizumab for age-related macular
degeneration: 1-year outcomes of a prospective, double-masked randomised clinical trial. Eye
(Lond). 2010;24(11):1708-1715.

P59
58
 Age-Related Macular Degeneration PPP

177. Tufail A, Patel PJ, Egan C, et al. ABC Trial Investigators. Bevacizumab for neovascular age related
macular degeneration (ABC Trial): multicentre randomised double masked study. BMJ.
2010;340:c2459.
178. Friedman DS, Gazzard G, Foster P, et al. Ultrasonographic biomicroscopy, Scheimpflug
photography, and novel provocative tests in contralateral eyes of Chinese patients initially seen with
acute angle closure. Arch Ophthalmol. 2003;121(5):633-642.
179. Rosenfeld PJ, Brown DM, Heier JS, et al. MARINA Study Group. Ranibizumab for neovascular
age-related macular degeneration. N Engl J Med. 2006;355(14):1419-1431.
180. Brown DM, Kaiser PK, Michels M, et al. ANCHOR Study Group. Ranibizumab versus verteporfin
for neovascular age-related macular degeneration. N Engl J Med. 2006;355(14):1432-1444.
181. Stein JD, Talwar N, Laverne AM, Nan B, Lichter PR. Racial disparities in the use of ancillary
testing to evaluate individuals with open-angle glaucoma. Arch Ophthalmol. 2012;130(12):1579-
1588.
182. Bressler NM. Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP)
Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in age-related
macular degeneration with verteporfin: two-year results of 2 randomized clinical trials-TAP report
2. Arch Ophthalmol. 2001;119(2):198-207.
183. Martin DF, Maguire MG, Fine SL, et al. Comparison of Age-related Macular Degeneration
Treatments Trials (CATT) Research Group. Ranibizumab and bevacizumab for treatment of
neovascular age-related macular degeneration: two-year results. Ophthalmology. 2012;119(7):1388-
1398.
184. Comparison of Age-related Macular Degeneration Treatments Trials Research G, Maguire MG,
Martin DF, et al. Five-Year Outcomes with Anti-Vascular Endothelial Growth Factor Treatment of
Neovascular Age-Related Macular Degeneration: The Comparison of Age-Related Macular
Degeneration Treatments Trials. Ophthalmology. 2016;123(8):1751-1761.
185. Chakravarthy U, Harding SP, Rogers CA, et al. IVAN Study Investigators. Ranibizumab versus
bevacizumab to treat neovascular age-related macular degeneration: one-year findings from the
IVAN randomized trial. Ophthalmology. 2012;119(7):1399-1411.
186. Chakravarthy U, Harding SP, Rogers CA, et al. IVAN Study Investigators. Alternative treatments to
inhibit VEGF in age-related choroidal neovascularisation: 2-year findings of the IVAN randomised
controlled trial. Lancet. 2013;382(9900):1258-1267.
187. Nguyen CL, Oh LJ, Wong E, Wei J, Chilov M. Anti-vascular endothelial growth factor for
neovascular age-related macular degeneration: a meta-analysis of randomized controlled trials. BMC
Ophthalmol. 2018;18(1):130.
188. Chen G, Li W, Tzekov R, Jiang F, Mao S, Tong Y. Bevacizumab versus ranibizumab for
neovascular age-related macular degeneration: a meta-analysis of randomized controlled trials.
Retina. 2015;35(2):187-193.
189. Sarwar S, Clearfield E, Soliman MK, et al. Aflibercept for neovascular age-related macular
degeneration. Cochrane Database Syst Rev. 2016;2:CD011346.
190. Moja L, Lucenteforte E, Kwag KH, et al. Systemic safety of bevacizumab versus ranibizumab for
neovascular age-related macular degeneration. Cochrane Database Syst Rev. 2014(9):CD011230.
191. Rao P, Lum F, Wood K, et al. Real-World Vision in Age-Related Macular Degeneration Patients
Treated with Single Anti-VEGF Drug Type for 1 Year in the IRIS Registry. Ophthalmology.
2018;125(4):522-528.
192. Schmidt-Erfurth U, Michels S, Augustin A. Perspectives on verteporfin therapy combined with
intravitreal corticosteroids. Arch Ophthalmol. 2006;124(4):561-563.
193. Zarbin M. Should corticosteroids be considered as part of the standard care with photodynamic
therapy? Arch Ophthalmol. 2006;124(4):563-571.
194. Sacu S, Varga A, Michels S, et al. Reduced fluence versus standard photodynamic therapy in
combination with intravitreal triamcinolone: short-term results of a randomised study. Br J
Ophthalmol. 2008;92(10):1347-1351.
195. Su Y, Wu J, Gu Y. Photodynamic therapy in combination with ranibizumab versus ranibizumab
monotherapy for wet age-related macular degeneration: A systematic review and meta-analysis.
Photodiagnosis Photodyn Ther. 2018;22:263-273.
196. Kaiser PK, Boyer DS, Cruess AF, Slakter JS, Pilz S, Weisberger A. DENALI Study Group.
Verteporfin plus ranibizumab for choroidal neovascularization in age-related macular degeneration:
twelve-month results of the DENALI study. Ophthalmology. 2012;119(5):1001-1010.

P60
59
 Age-Related Macular Degeneration PPP

197. Larsen M, Schmidt-Erfurth U, Lanzetta P, et al. MONT BLANC Study Group. Verteporfin plus
ranibizumab for choroidal neovascularization in age-related macular degeneration: twelve-month
MONT BLANC study results. Ophthalmology. 2012;119(5):992-1000.
198. Koh A, Lee WK, Chen LJ, et al. EVEREST study: efficacy and safety of verteporfin photodynamic
therapy in combination with ranibizumab or alone versus ranibizumab monotherapy in patients with
symptomatic macular polypoidal choroidal vasculopathy. Retina. 2012;32(8):1453-1464.
199. Liu L, Tham YC, Wu J, Yue S, Cheng CY. Photodynamic therapy in combination with ranibizumab
versus ranibizumab monotherapy for polypoidal choroidal vasculopathy: A systematic review and
meta-analysis. Photodiagnosis Photodyn Ther. 2017;20:215-220.
200. Lee WK, Iida T, Ogura Y, et al. Efficacy and Safety of Intravitreal Aflibercept for Polypoidal
Choroidal Vasculopathy in the PLANET Study: A Randomized Clinical Trial. JAMA Ophthalmol.
2018;136(7):786-793.
201. Gao Y, Yu T, Zhang Y, Dang G. Anti-VEGF Monotherapy Versus Photodynamic Therapy and
Anti-VEGF Combination Treatment for Neovascular Age-Related Macular Degeneration: A Meta-
Analysis. Invest Ophthalmol Vis Sci. 2018;59(10):4307-4317.
202. Macular Photocoagulation Study Group. Laser photocoagulation for neovascular lesions nasal to the
fovea: results from clinical trials for lesions secondary to ocular histoplasmosis or idiopathic causes.
Arch Ophthalmol. 1995;113(1):56-61.
203. Feder RS, Olsen TW, Prum BE, Jr., et al. Comprehensive adult medical eye evaluation Preferred
Practice Pattern® Guidelines. Ophthalmology. 2016;123(1):P209-236.
204. Fine AM, Elman MJ, Ebert JE, Prestia PA, Starr JS, Fine SL. Earliest symptoms caused by
neovascular membranes in the macula. Arch Ophthalmol. 1986;104(4):513-514.
205. Eye Disease Case-Control Study Group. Risk factors for neovascular age-related macular
degeneration. Arch Ophthalmol. 1992;110(12):1701-1708.
206. Macular Photocoagulation Study Group. Risk factors for choroidal neovascularization in the second
eye of patients with juxtafoveal or subfoveal choroidal neovascularization secondary to age-related
macular degeneration. Arch Ophthalmol. 1997;115(6):741-747.
207. US Food and Drug Administration, Center for Drug Evaluation and Research. Lucentis
(ranibizumab injection). BLA 25156. Available at:
www.accessdata.fda.gov/drugsatfda_docs/label/2012/125156s0069s0076lbl.pdf. Accessed
September 2019.
208. Meyers SM, Greene T, Gutman FA. A twin study of age-related macular degeneration. Am J
Ophthalmol. 1995;120(6):757-766.
209. McDonald HR, Williams GA, Scott IU, Haller JA, Maguire AM, Marcus DM. Laser scanning
imaging for macular disease: a report by the American Academy of Ophthalmology.
Ophthalmology. 2007;114(6):1221-1228.
210. Fung AE, Lalwani GA, Rosenfeld PJ, et al. An optical coherence tomography-guided, variable
dosing regimen with intravitreal ranibizumab (Lucentis) for neovascular age-related macular
degeneration. Am J Ophthalmol. 2007;143(4):566-583.
211. Kaiser PK, Blodi BA, Shapiro H, Acharya NR. Angiographic and optical coherence tomographic
results of the MARINA study of ranibizumab in neovascular age-related macular degeneration.
Ophthalmology. 2007;114(10):1868-1875.
212. Krebs I, Binder S, Stolba U, et al. Optical coherence tomography guided retreatment of
photodynamic therapy. Br J Ophthalmol. 2005;89(9):1184-1187.
213. Ahlers C, Golbaz I, Stock G, et al. Time course of morphologic effects on different retinal
compartments after ranibizumab therapy in age-related macular degeneration. Ophthalmology.
2008;115(8):e39-46.
214. Spaide RF, Koizumi H, Pozzoni MC. Enhanced depth imaging spectral-domain optical coherence
tomography. Am J Ophthalmol. 2008;146(4):496-500.
215. Keane PA, Patel PJ, Liakopoulos S, Heussen FM, Sadda SR, Tufail A. Evaluation of age-related
macular degeneration with optical coherence tomography. Surv Ophthalmol. 2012;57(5):389-414.
216. Hu Z, Wu X, Ouyang Y, Sadda SR. Semiautomated segmentation of the choroid in spectral-domain
optical coherence tomography volume scans. Invest Ophthalmol Vis Sci. 2013;54(3):1722-1729.
217. Karampelas M, Sim DA, Keane PA, et al. Evaluation of retinal pigment epithelium-Bruch's
membrane complex thickness in dry age-related macular degeneration using optical coherence
tomography. Br J Ophthalmol. 2013;97(10):1256-1261.

P61
60
 Age-Related Macular Degeneration PPP

218. Wang R, Liang Z, Liu X. Diagnostic accuracy of optical coherence tomography angiography for
choroidal neovascularization: a systematic review and meta-analysis. BMC Ophthalmol.
2019;19(1):162.
219. Yannuzzi LA, Rohrer KT, Tindel LJ, et al. Fluorescein angiography complication survey.
Ophthalmology. 1986;93(5):611-617.
220. Kwiterovich KA, Maguire MG, Murphy RP, et al. Frequency of adverse systemic reactions after
fluorescein angiography. Results of a prospective study. Ophthalmology. 1991;98(7):1139-1142.
221. Holz FG, Bindewald-Wittich A, Fleckenstein M, et al. Progression of geographic atrophy and
impact of fundus autofluorescence patterns in age-related macular degeneration. Am J Ophthalmol.
2007;143(3):463-472.
222. American Academy of Ophthalmology. Indocyanine green angiography. Ophthalmology.
1998;105(8):1564-1569.
223. Spaide RF, Yannuzzi LA, Slakter JS, Sorenson J, Orlach DA. Indocyanine green videoangiography
of idiopathic polypoidal choroidal vasculopathy. Retina. 1995;15(2):100-110.
224. Sho K, Takahashi K, Yamada H, et al. Polypoidal choroidal vasculopathy: incidence, demographic
features, and clinical characteristics. Arch Ophthalmol. 2003;121(10):1392-1396.
225. Su Z, Ye P, Teng Y, Zhang L, Shu X. Adverse reaction in patients with drug allergy history after
simultaneous intravenous fundus fluorescein angiography and indocyanine green angiography. J
Ocul Pharmacol Ther. 2012;28(4):410-413.
226. Midena E, Vujosevic S, Convento E, Manfre A, Cavarzeran F, Pilotto E. Microperimetry and
fundus autofluorescence in patients with early age-related macular degeneration. Br J Ophthalmol.
2007;91(11):1499-1503.
227. Paques M, Meimon S, Rossant F, et al. Adaptive optics ophthalmoscopy: Application to age-related
macular degeneration and vascular diseases. Prog Retin Eye Res. 2018;66:1-16.
228. Constable IJ, Lai CM, Magno AL, et al. Gene therapy in neovascular age-related macular
degeneration: three-year follow-up of a phase 1 randomized dose escalation trial. Am J Ophthalmol.
2017;177:150-158.
229. Campochiaro PA, Lauer AK, Sohn EH, et al. Lentiviral vector gene transfer of
endostatin/angiostatin for macular degeneration (GEM) study. Hum Gene Ther. 2017;28(1):99-111.
230. Ho AC, Chang TS, Samuel M, Williamson P, Willenbucher RF, Malone T. Experience with a
subretinal cell-based therapy in patients with geographic atrophy secondary to age-related macular
degeneration. Am J Ophthalmol. 2017;179:67-80.
231. Stead LF, Bergson G, Lancaster T. Physician advice for smoking cessation. Cochrane Database Syst
Rev 2008, Issue 2. Art. No.: CD000165. DOI: 10.1002/14651858.CD000165.pub3.
232. Tiedeman JS. A physical analysis of the factors that determine the contour of the iris. Am J
Ophthalmol. 1991;111(3):338-343.
233. AREDS2-HOME Study Research Group, Chew EY, Clemons TE, et al. Randomized trial of a home
monitoring system for early detection of choroidal neovascularization home monitoring of the Eye
(HOME) study. Ophthalmology. 2014;121(2):535-544.
234. Chew EY, Clemons TE, Bressler SB, et al. Randomized trial of the ForeseeHome monitoring device
for early detection of neovascular age-related macular degeneration. The HOme Monitoring of the
Eye (HOME) study design - HOME Study report number 1. Contemp Clin Trials. 2014;37(2):294-
300.
235. Faridi A, Jia Y, Gao SS, et al. Sensitivity and specificity of OCT angiography to detect choroidal
neovascularization. Ophthalmol Retina. 2017;1(4):294-303.
236. de Oliveira Dias JR, Zhang Q, Garcia JMB, et al. Natural History of Subclinical Neovascularization
in Nonexudative Age-Related Macular Degeneration Using Swept-Source OCT Angiography.
Ophthalmology. 2018;125(2):255-266.
237. Carnevali A, Sacconi R, Querques L, et al. Natural History of Treatment-Naive Quiescent Choroidal
Neovascularization in Age-Related Macular Degeneration Using OCT Angiography. Ophthalmol
Retina. 2018;2(9):922-930.
238. AREDS2-HOME Study Research Group, Chew EY, Clemons TE, Bressler SB, et al. Randomized
trial of a home monitoring system for early detection of choroidal neovascularization Home
Monitoring of the Eye (HOME) Study. Ophthalmology. 2014;121(2):535-544.
239. Fong DS, Ferris FL, III, Davis MD, Chew EY. Early Treatment Diabetic Retinopathy Study
Research Group. Causes of severe visual loss in the early treatment diabetic retinopathy study:
ETDRS report no. 24. Am J Ophthalmol. 1999;127(2):137-141.

P62
61
 Age-Related Macular Degeneration PPP

240. American Academy of Ophthalmology. Verifying the source of compounded bevacizumab for
intravitreal injections - 2014. Available at: https://www.aao.org/clinical-statement/verifying-source-
of-compounded-bevacizumab-intravi-2. Accessed September 2019.
241. Aiello LP, Brucker AJ, Chang S, et al. Evolving guidelines for intravitreous injections. Retina.
2004;24(5 Suppl):S3-19.
242. Zhang J, Liang Y, Xie J, et al. Conbercept for patients with age-related macular degeneration: a
systematic review. BMC Ophthalmol. 2018;18(1):142.
243. Callanan D, Kunimoto D, Maturi RK, et al. Double-Masked, Randomized, Phase 2 Evaluation of
Abicipar Pegol (an Anti-VEGF DARPin Therapeutic) in Neovascular Age-Related Macular
Degeneration. J Ocul Pharmacol Ther. 2018.
244. Kunimoto D, Ohji M, Maturi RK, et al. Evaluation of Abicipar Pegol (an Anti-VEGF DARPin
Therapeutic) in Patients With Neovascular Age-Related Macular Degeneration: Studies in Japan and
the United States. Ophthalmic Surg Lasers Imaging Retina. 2019;50(2):e10-e22.
245. Dugel PU, Koh A, Ogura Y, et al. HAWK and HARRIER: Phase 3, Multicenter, Randomized,
Double-Masked Trials of Brolucizumab for Neovascular Age-Related Macular Degeneration.
Ophthalmology. 2019.
246. Busbee BG, Ho AC, Brown DM, et al. HARBOR Study Group. Twelve-month efficacy and safety
of 0.5 mg or 2.0 mg ranibizumab in patients with subfoveal neovascular age-related macular
degeneration. Ophthalmology. 2013;120(5):1046-1056.
247. Macular Photocoagulation Study Group. Five-year follow-up of fellow eyes of patients with age-
related macular degeneration and unilateral extrafoveal choroidal neovascularization. Arch
Ophthalmol. 1993;111(9):1189-1199.
248. Shienbaum G, Gupta OP, Fecarotta C, Patel AH, Kaiser RS, Regillo CD. Bevacizumab for
neovascular age-related macular degeneration using a treat-and-extend regimen: clinical and
economic impact. Am J Ophthalmol. 2012;153(3):468-473.
249. Gupta OP, Shienbaum G, Patel AH, Fecarotta C, Kaiser RS, Regillo CD. A treat and extend
regimen using ranibizumab for neovascular age-related macular degeneration clinical and economic
impact. Ophthalmology. 2010;117(11):2134-2140.
250. Oubraham H, Cohen SY, Samimi S, et al. Inject and extend dosing versus dosing as needed: a
comparative retrospective study of ranibizumab in exudative age-related macular degeneration.
Retina. 2011;31(1):26-30.
251. Toalster N, Russell M, Ng P. A 12-month prospective trial of inject and extend regimen for
ranibizumab treatment of age-related macular degeneration. Retina. 2013;33(7):1351-1358.
252. Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study Group.
Photodynamic therapy of subfoveal choroidal neovascularization in age-related macular
degeneration with verteporfin: one-year results of 2 randomized clinical trials--TAP report. Arch
Ophthalmol. 1999;117(10):1329-1345.
253. Verteporfin in Photodynamic Therapy Study Group. Verteporfin therapy of subfoveal choroidal
neovascularization in age-related macular degeneration: two-year results of a randomized clinical
trial including lesions with occult with no classic choroidal neovascularization--verteporfin in
photodynamic therapy report 2. Am J Ophthalmol. 2001;131(5):541-560.
254. Barbazetto I, Burdan A, Bressler NM, et al. Photodynamic therapy of subfoveal choroidal
neovascularization with verteporfin: fluorescein angiographic guidelines for evaluation and
treatment--TAP and VIP report number 2. Arch Ophthalmol. 2003;121(9):1253-1268.
255. Comparison of Age-related Macular Degeneration Treatments Trials (CATT) Research Group,
Maguire MG, Martin DF, et al. Five-year outcomes with anti-vascular endothelial growth factor
treatment of neovascular age-related macular degeneration: the comparison of age-related macular
degeneration treatments trials. Ophthalmology. 2016;123(8):1751-1761.
256. Berg K, Pedersen TR, Sandvik L, Bragadottir R. Comparison of ranibizumab and bevacizumab for
neovascular age-related macular degeneration according to LUCAS treat-and-extend protocol.
Ophthalmology. 2015;122(1):146-152.
257. Giansanti F, Eandi CM, Virgili G. Submacular surgery for choroidal neovascularisation secondary
to age-related macular degeneration. Cochrane Database Syst Rev. 2009(2):CD006931.
258. American Academy of Ophthalmology. An ophthalmologist's duties concerning postoperative care.
Available at: https://www.aao.org/clinical-statement/ophthalmologists-duties-concerning-
postoperative-c. Accessed September 2019.

P63
62
 Age-Related Macular Degeneration PPP

259. Virgili G, Parravano M, Menchini F, Brunetti M. Antiangiogenic therapy with anti-vascular

endothelial growth factor modalities for diabetic macular oedema. Cochrane Database Syst Rev
2012, Issue 12. Art. No.: CD007419. DOI:10.1002/14651858.CD007419.pub3.
260. Hoang QV, Mendonca LS, Della Torre KE, Jung JJ, Tsuang AJ, Freund KB. Effect on intraocular
pressure in patients receiving unilateral intravitreal anti-vascular endothelial growth factor
injections. Ophthalmology. 2012;119(2):321-326.
261. Wehrli SJ, Tawse K, Levin MH, Zaidi A, Pistilli M, Brucker AJ. A lack of delayed intraocular
pressure elevation in patients treated with intravitreal injection of bevacizumab and ranibizumab.
Retina. 2012;32(7):1295-1301.
262. Pielen A, Feltgen N, Isserstedt C, Callizo J, Junker B, Schmucker C. Efficacy and safety of
intravitreal therapy in macular edema due to branch and central retinal vein occlusion: a systematic
review. PLoS One. 2013;8(10):e78538.
263. Bakri SJ, Thorne JE, Ho AC, et al. Safety and Efficacy of Anti-Vascular Endothelial Growth Factor
Therapies for Neovascular Age-Related Macular Degeneration: A Report by the American
Academy of Ophthalmology. Ophthalmology. 2019;126(1):55-63.
264. Tarantola RM, Folk JC, Boldt HC, Mahajan VB. Intravitreal bevacizumab during pregnancy.
Retina. 2010;30(9):1405-1411.
265. Ehlken C, Martin G, Stahl A, Agostini HT. Reduction of vascular endothelial growth factor a in
human breast milk after intravitreal injection of bevacizumab but not ranibizumab. Arch
Ophthalmol. 2012;130(9):1226-1227.
266. Goldberg RA, Shah CP, Wiegand TW, Heier JS. Noninfectious inflammation after intravitreal
injection of aflibercept: clinical characteristics and visual outcomes. Am J Ophthalmol.
2014;158(4):733-737 e731.
267. Gupta A, Lam J, Custis P, Munz S, Fong D, Koster M. Implantable miniature telescope (IMT) for
vision loss due to end-stage age-related macular degeneration. Cochrane Database Syst Rev.
2018;5:CD011140.
268. Wei Y, Liao H, Ye J. Therapeutic effects of various therapeutic strategies on non-exudative age-
related macular degeneration: A PRISMA-compliant network meta-analysis of randomized
controlled trials. Medicine (Baltimore). 2018;97(21):e10422.
269. Wu L, Martinez-Castellanos MA, Quiroz-Mercado H, et al. Pan American Collaborative Retina
Group (PACORES). Twelve-month safety of intravitreal injections of bevacizumab (Avastin):
results of the Pan-American Collaborative Retina Study Group (PACORES). Graefes Arch Clin Exp
Ophthalmol. 2008;246(1):81-87.
270. Gragoudas ES, Adamis AP, Cunningham ET, Jr., Feinsod M, Guyer DR. Pegaptanib for
neovascular age-related macular degeneration. N Engl J Med. 2004;351(27):2805-2816.
271. McClintic SM, Gao S, Wang J, et al. Quantitative Evaluation of Choroidal Neovascularization under
Pro Re Nata Anti-Vascular Endothelial Growth Factor Therapy with OCT Angiography.
Ophthalmol Retina. 2018;2(9):931-941.
272. Pilotto E, Frizziero L, Daniele AR, et al. Early OCT angiography changes of type 1 CNV in
exudative AMD treated with anti-VEGF. Br J Ophthalmol. 2018:[Epub ahead of print].
273. Schneider EW, Fowler SC. Optical coherence tomography angiography in the management of age-
related macular degeneration. Curr Opin Ophthalmol. 2018;29(3):217-225.
274. Told R, Sacu S, Hecht A, et al. Comparison of SD-optical coherence tomography angiography and
indocyanine green angiography in type 1 and 2 neovascular age-related macular degeneration. Invest
Ophthalmol Vis Sci. 2018;59(6):2393-2400.
275. Wong T, Chakravarthy U, Klein R, et al. The natural history and prognosis of neovascular age-
related macular degeneration: a systematic review of the literature and meta-analysis.
Ophthalmology. 2008;115(1):116-126.
276. Stelmack JA, Tang XC, Reda DJ, Rinne S, Mancil RM, Massof RW. LOVIT Study Group.
Outcomes of the Veterans Affairs Low Vision Intervention Trial (LOVIT). Arch Ophthalmol.
2008;126(5):608-617.
277. Fontenot JL, Bona MD, Kaleem MA, et al. Vision Rehabilitation Preferred Practice Pattern®.
Ophthalmology. 2018;125(1):P228-P278.
278. Hudson HL, Lane SS, Heier JS, et al. Implantable miniature telescope for the treatment of visual
acuity loss resulting from end-stage age-related macular degeneration: 1-year results.
Ophthalmology. 2006;113(11):1987-2001.

P64
63
 Age-Related Macular Degeneration PPP

279. Brown GC, Brown MM, Lieske HB, Lieske PA, Brown KS, Lane SS. Comparative effectiveness
and cost-effectiveness of the implantable miniature telescope. Ophthalmology. 2011;118(9):1834-
1843.
280. Coleman AL, Stone K, Ewing SK, et al. Higher risk of multiple falls among elderly women who
lose visual acuity. Ophthalmology. 2004;111(5):857-862.
281. Soubrane G, Cruess A, Lotery A, et al. Burden and health care resource utilization in neovascular
age-related macular degeneration: findings of a multicountry study. Arch Ophthalmol.
2007;125(9):1249-1254.
282. Rovner BW, Casten RJ, Tasman WS. Effect of depression on vision function in age-related macular
degeneration. Arch Ophthalmol. 2002;120(8):1041-1044.
283. Rein DB, Zhang P, Wirth KE, et al. The economic burden of major adult visual disorders in the
United States. Arch Ophthalmol. 2006;124(12):1754-1760.
284. Bressler NM, Bressler SB, Congdon NG, et al. Age-Related Eye Disease Study Research Group.
Potential public health impact of Age-Related Eye Disease Study results: AREDS report no. 11.
Arch Ophthalmol. 2003;121(11):1621-1624.
285. Christ SL, Zheng DD, Swenor BK, et al. Longitudinal relationships among visual acuity, daily
functional status, and mortality: the Salisbury Eye Evaluation Study. JAMA Ophthalmol.
2014;132(12):1400-1406.
286. van Asten F, Michels CTJ, Hoyng CB, et al. The cost-effectiveness of bevacizumab, ranibizumab
and aflibercept for the treatment of age-related macular degeneration-A cost-effectiveness analysis
from a societal perspective. PLoS One. 2018;13(5):e0197670.
287. Nwanze CC, Akinwale A, Adelman RA. Bevacizumab vs. ranibizumab in preserving or improving
vision in patients with wet, age-related macular degeneration: a cost-effectiveness review. Clin Med
Insights Ther. 2012;4:29-38.
288. Mitchell P, Annemans L, White R, Gallagher M, Thomas S. Cost effectiveness of treatments for wet
age-related macular degeneration. Pharmacoeconomics. 2011;29(2):107-131.
289. Patel JJ, Mendes MA, Bounthavong M, Christopher ML, Boggie D, Morreale AP. Cost-utility
analysis of bevacizumab versus ranibizumab in neovascular age-related macular degeneration using
a Markov model. J Eval Clin Pract. 2012;18(2):247-255.
290. Stein JD, Newman-Casey PA, Mrinalini T, Lee PP, Hutton DW. Cost-effectiveness of bevacizumab
and ranibizumab for newly diagnosed neovascular macular degeneration. Ophthalmology.
2014;121(4):936-945.
291. Chapman JA, Beckey C. Pegaptanib: a novel approach to ocular neovascularization. Ann
Pharmacother. 2006;40(7-8):1322-1326.
292. Web JA. Genentech decision expands access to bevacizumab. Ophthalmol Times. January 15, 2008.
293. Windsor MA, Sun SJJ, Frick KD, Swanson EA, Rosenfeld PJ, Huang D. Estimating Public and
Patient Savings From Basic Research-A Study of Optical Coherence Tomography in Managing
Antiangiogenic Therapy. Am J Ophthalmol. 2018;185:115-122.

64
P65
over

Diabetic Retinopathy
Preferred Practice
Pattern®

© 2019 by the American Academy of Ophthalmology http://dx.doi.org/10.1016/j.ophtha.2019.09.025

Published by Elsevier Inc. ISSN 0161-6420/19
 Diabetic Retinopathy PPP

Secretary for Quality of Care:

Timothy W. Olsen, MD

Academy Staff:
Ali Al-Rajhi, PhD, MPH
Andre Ambrus, MLIS
Meghan Daly
Flora C. Lum, MD

Medical Editor: Susan Garratt

Approved by: Board of Trustees

September 7, 2019

© 2019 American Academy of Ophthalmology®

All rights reserved

AMERICAN ACADEMY OF OPHTHALMOLOGY and PREFERRED PRACTICE PATTERN are

registered trademarks of the American Academy of Ophthalmology. All other trademarks are the property of
their respective owners.

Preferred Practice Pattern® guidelines are developed by the Academy’s H. Dunbar Hoskins Jr., MD Center
for Quality Eye Care without any external financial support. Authors and reviewers of the guidelines are
volunteers and do not receive any financial compensation for their contributions to the documents. The
guidelines are externally reviewed by experts and stakeholders before publication.

Correspondence:
Ali A. Al-Rajhi, PhD, MPH American Academy of Ophthalmology, P. O. Box 7424, San Francisco, CA
94120-7424. E-mail: [email protected].

P67
 Diabetic Retinopathy PPP

The Retina/Vitreous Preferred Practice Pattern® Panel members wrote the Diabetic Retinopathy Preferred
Practice Pattern® (PPP) guidelines. The PPP Panel members discussed and reviewed successive drafts of the
document, meeting in person twice and conducting other review by e-mail discussion, to develop a consensus
over the final version of the document.

Retina/Vitreous Preferred Practice Pattern Panel 2018–2019

Steven T. Bailey, MD, Retina Society Representative
Amani Fawzi, MD, Macula Society Representative
Jennifer I. Lim, MD
Ron A. Adelman, MD, MPH, MBA, FACS
Gurunadh A. Vemulakonda, MD, American Society of Retina Specialists Representative
Gui-shang Ying, MD, PhD, Methodologist
Christina J. Flaxel, MD, Chair

We thank our partners, the Cochrane Eyes and Vision US Satellite (CEV@US), for identifying reliable systematic
reviews that we cite and discuss in support of the PPP recommendations.

The Preferred Practice Patterns Committee members reviewed and discussed the document during a meeting
in June 2019. The document was edited in response to the discussion and comments.

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair
Roy S. Chuck, MD, PhD
Steven P. Dunn, MD
Christina J. Flaxel, MD
Steven J. Gedde, MD
Francis S. Mah, MD
Randall J. Olson, MD
David K. Wallace, MD, MPH
David C. Musch, PhD, MPH, Methodologist

The Diabetic Retinopathy PPP was then sent for review to additional internal and external groups and individuals in
July 2019. All those returning comments were required to provide disclosure of relevant relationships with industry
to have their comments considered (indicated with an asterisk below). Members of the Retina/Vitreous Preferred
Practice Pattern Panel reviewed and discussed these comments and determined revisions to the document.

P68
 Diabetic Retinopathy PPP

In compliance with the Council of Medical Specialty Societies’ Code for Interactions with Companies
(available at www.cmss.org/codeforinteractions.aspx), relevant relationships with industry are listed. The
Academy has Relationship with Industry Procedures to comply with the Code (available at
http://one.aao.org/CE/PracticeGuidelines/PPP.aspx). A majority (88%) of the members of the Retina/Vitreous
Preferred Practice Pattern Panel 2018–2019 had no financial relationship to disclose.

Retina/Vitreous Preferred Practice Pattern Panel 2018–2019

Christina J. Flaxel, MD: No financial relationships to disclose
Ron A. Adelman, MD, MPH, MBA, FACS: No financial relationships to disclose
Steven T. Bailey, MD: No financial relationships to disclose
Amani Fawzi, MD: No financial relationships to disclose
Jennifer I. Lim, MD: Alcon Laboratories, Genentech, Kodiak Sciences, EyePoint Pharmaceuticals—
Consultant/Advisor; Genentech—Lecture Fees
Gurunadh A. Vemulakonda, MD: No financial relationships to disclose
Gui-shang Ying, MD, PhD: No financial relationships to disclose

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair: No financial relationships to disclose
Roy S. Chuck, MD, PhD: Novartis, Shire—Consultant/Advisor
Steven P. Dunn, MD: No financial relationships to disclose
Christina J. Flaxel, MD: No financial relationships to disclose
Steven J. Gedde, MD: No financial relationships to disclose
Francis S. Mah, MD: Aerie Pharmaceuticals, Bausch + Lomb, EyePoint Pharmaceuticals, Novartis, Ocular
Therapeutix, Shire, Sun Pharma—Consultant/Advisor; Bausch + Lomb, Novartis, Shire, Sun Pharma—Lecture Fees
Randall J. Olson, MD: No financial disclosures
David K. Wallace, MD, MPH: No financial disclosures
David C. Musch, PhD, MPH, Methodologist:, IRIDEX, Notal Vision—Consultant/Advisor

Secretary for Quality of Care

Timothy W. Olsen, MD: No financial relationships to disclose

Academy Staff
Ali Al-Rajhi, PhD, MPH: No financial relationships to disclose
Andre Ambrus, MLIS: No financial relationships to disclose
Meghan Daly: No financial relationships to disclose
Flora C. Lum, MD: No financial relationships to disclose

The disclosures of relevant relationships to industry of other reviewers of the document from January to
October 2019 are available online at www.aao.org/ppp.

ii

P69
 Abs

DiabeticRetinopathy
Diabetic RetinopathyPPP
PPP

TABLE OF CONTENTS

OBJECTIVES OF PREFERRED PRACTICE PATTERN GUIDELINES ........................................... P72

METHODS AND KEY TO RATINGS ....................................................................................................... P73
HIGHLIGHTED FINDINGS AND RECOMMENDATIONS FOR CARE ........................................... P74
INTRODUCTION ........................................................................................................................................ P75
Disease Definition .......................................................................................................................................... P75
Patient Population ........................................................................................................................................... P75
Clinical Objectives ......................................................................................................................................... P75
BACKGROUND ........................................................................................................................................... P76
Introduction .................................................................................................................................................... P76
Prevalence of Diabetes ........................................................................................................................... P76
Prevalence of Diabetic Retinopathy ....................................................................................................... P77
Risk Factors .................................................................................................................................................... P78
Natural History ............................................................................................................................................... P80
CARE PROCESS ......................................................................................................................................... P83
Patient Outcome Criteria ................................................................................................................................ P84
Diagnosis ........................................................................................................................................................ P84
History .................................................................................................................................................... P84
Examination ............................................................................................................................................ P84
Examination Schedule ............................................................................................................................ P85
Ancillary Tests........................................................................................................................................ P86
Management ................................................................................................................................................... P89
Prevention of Diabetic Retinopathy ........................................................................................................ P89
Early Detection of Diabetic Retinopathy ................................................................................................ P90
Medical and Surgical Management ........................................................................................................ P91
Follow-up Evaluation ........................................................................................................................... P104
Provider and Setting ..................................................................................................................................... P104
Counseling and Referral ............................................................................................................................... P104
Socioeconomic Considerations ..................................................................................................................... P105
APPENDIX 1. QUALITY OF OPHTHALMIC CARE CORE CRITERIA ......................................... P106
APPENDIX 2. INTERNATIONAL STATISTICAL CLASSIFICATION OF DISEASES
AND RELATED HEALTH PROBLEMS (ICD) CODES ...................................................................... P108
APPENDIX 3. MAJOR STUDY RESULTS ........................................................................................... P110
APPENDIX 4. GLYCEMIC CONTROL ................................................................................................ P119
APPENDIX 5. CLASSIFICATION OF DIABETIC RETINOPATHY IN THE EARLY
TREATMENT OF DIABETIC RETINOPATHY STUDY .............................................................. …..P123
GLOSSARY ................................................................................................................................................ P124
LITERATURE SEARCHES FOR THIS PPP ......................................................................................... P129
RELATED ACADEMY MATERIALS .................................................................................................... P130
REFERENCES ........................................................................................................................................... P131

P70
P70
 Diabetic Retinopathy PPP
Diabetic Retinopathy PPP

Background:
Diabetic retinopathy is a leading cause of visual impairment in working-age adults worldwide.
Duration of diabetes is a major risk factor associated with the development of diabetic
retinopathy. Due to the disproportionately large number of patients with type 2 diabetes, this
group comprises a larger proportion of the disease burden in patients with visual impairment
from diabetic retinopathy compared to patients with type 1 diabetes. The recommendations of
this Preferred Practice Pattern are based on Cochrane-identified reliable systematic reviews.
Rationale for treatment:
Both clinical trials and epidemiological studies have shown that the two key modifiable risk
factors associated with developing diabetic retinopathy are blood sugar and blood pressure
control. Maintaining near-normal glucose levels and near-normal blood pressure lowers the risk
of retinopathy developing and/or progressing.
Care Process:
The care process for diabetic retinopathy includes a medical history, a regular ophthalmologic
examination or screening of high-quality retinal photographs of patients who have not had
previous treatment for diabetic retinopathy or other eye disease, and regular follow-up. The goal
of treatment is to improve or stabilize visual function, improve vision-related quality of life; and,
through close communication with the patient’s primary care physician achieve optimal control
of blood glucose, blood pressure and other metabolic risk factors.

The initial examination for a patient with diabetes mellitus includes all features of the
comprehensive adult medical eye evaluation, with particular attention to those aspects relevant to
diabetic retinopathy. The examination schedule is detailed in this Preferred Practice Pattern for
patients diagnosed with type 1 or type 2 diabetes. Additionally, ancillary tests (e.g., fundus
photography, OCT, and FA) to clinical examinations may enhance patient care.

Management options for diabetic retinopathy includes following a healthy diet and lifestyle,
medical management, timely ophthalmic evaluation, and treatment under the care of an
ophthalmologist. Cost-effective treatments with laser, anti-VEGF agents, or intravitreal
corticosteroids may also be considered. Because patients with diabetes may be under the care of
multiple practitioners, effective communication and care coordination is necessary to optimize
care.

P71

P71
 Diabetic Retinopathy PPP

As a service to its members and the public, the American Academy of Ophthalmology has developed a series
of Preferred Practice Pattern® guidelines that identify characteristics and components of quality eye care.
Appendix 1 describes the core criteria of quality eye care.

The Preferred Practice Pattern® guidelines are based on the best available scientific data as interpreted by
panels of knowledgeable health professionals. In some instances, such as when results of carefully conducted
clinical trials are available, the data are particularly persuasive and provide clear guidance. In other instances,
the panels have to rely on their collective judgment and evaluation of available evidence.

These documents provide guidance for the pattern of practice, not for the care of a particular
individual. While they should generally meet the needs of most patients, they cannot possibly best meet the
needs of all patients. Adherence to these PPPs will not ensure a successful outcome in every situation. These
practice patterns should not be deemed inclusive of all proper methods of care or exclusive of other methods
of care reasonably directed at obtaining the best results. It may be necessary to approach different patients’
needs in different ways. The physician must make the ultimate judgment about the propriety of the care of a
particular patient in light of all of the circumstances presented by that patient. The American Academy of
Ophthalmology is available to assist members in resolving ethical dilemmas that arise in the course of
ophthalmic practice.

Preferred Practice Pattern® guidelines are not medical standards to be adhered to in all individual
situations. The Academy specifically disclaims any and all liability for injury or other damages of any kind,
from negligence or otherwise, for any and all claims that may arise out of the use of any recommendations or
other information contained herein.

References to certain drugs, instruments, and other products are made for illustrative purposes only and are
not intended to constitute an endorsement of such. Such material may include information on applications
that are not considered community standard, that reflect indications not included in approved U.S. Food and
Drug Administration (FDA) labeling, or that are approved for use only in restricted research settings. The
FDA has stated that it is the responsibility of the physician to determine the FDA status of each drug or
device he or she wishes to use, and to use them with appropriate patient consent in compliance with
applicable law.

Innovation in medicine is essential to ensure the future health of the American public, and the Academy
encourages the development of new diagnostic and therapeutic methods that will improve eye care. It is
essential to recognize that true medical excellence is achieved only when the patients’ needs are the foremost
consideration.

All Preferred Practice Pattern® guidelines are reviewed by their parent panel annually or earlier if
developments warrant and updated accordingly. To ensure that all PPPs are current, each is valid for 5 years
from the approved by date unless superseded by a revision. Preferred Practice Pattern guidelines are funded
by the Academy without commercial support. Authors and reviewers of PPPs are volunteers and do not
receive any financial compensation for their contributions to the documents. The PPPs are externally
reviewed by experts and stakeholders, including consumer representatives, before publication. The PPPs are
developed in compliance with the Council of Medical Specialty Societies’ Code for Interactions with
Companies. The Academy has Relationship with Industry Procedures (available at www.aao.org/about-
preferred-practice-patterns) to comply with the Code.

Appendix 2 contains the International Statistical Classification of Diseases and Related Health Problems
(ICD) codes for the disease entities that this PPP covers. The intended users of the Diabetic Retinopathy PPP
are ophthalmologists.

P72
2
 Diabetic Retinopathy PPP

Preferred Practice Pattern® guidelines should be clinically relevant and specific enough to provide
useful information to practitioners. Where evidence exists to support a recommendation for care, the
recommendation should be given an explicit rating that shows the strength of evidence. To accomplish
these aims, methods from the Scottish Intercollegiate Guideline Network1 (SIGN) and the Grading of
Recommendations Assessment, Development and Evaluation2 (GRADE) group are used. GRADE is a
systematic approach to grading the strength of the total body of evidence that is available to support
recommendations on a specific clinical management issue. Organizations that have adopted GRADE
include SIGN, the World Health Organization, the Agency for Healthcare Research and Policy, and the
American College of Physicians.3
◆ All studies used to form a recommendation for care are graded for strength of evidence individually, and
that grade is listed with the study citation.
◆ To rate individual studies, a scale based on SIGN1 is used. The definitions and levels of evidence to rate
individual studies are as follows:
I++ High-quality meta-analyses, systematic reviews of randomized controlled trials (RCTs), or
RCTs with a very low risk of bias
I+ Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias
I- Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias
II++ High-quality systematic reviews of case-control or cohort studies
High-quality case-control or cohort studies with a very low risk of confounding or bias and a
high probability that the relationship is causal
II+ Well-conducted case-control or cohort studies with a low risk of confounding or bias and a
moderate probability that the relationship is causal
II- Case-control or cohort studies with a high risk of confounding or bias and a significant risk that
the relationship is not causal
III Nonanalytic studies (e.g., case reports, case series)

◆ Recommendations for care are formed based on the body of the evidence. The body of evidence quality
ratings are defined by GRADE2 as follows:
Good quality Further research is very unlikely to change our confidence in the estimate of
effect
Moderate quality Further research is likely to have an important impact on our confidence in the
estimate of effect and may change the estimate
Insufficient quality Further research is very likely to have an important impact on our confidence in
the estimate of effect and is likely to change the estimate
Any estimate of effect is very uncertain

◆ Key recommendations for care are defined by GRADE 2 as follows:

Strong Used when the desirable effects of an intervention clearly outweigh the
recommendation undesirable effects or clearly do not
Discretionary Used when the trade-offs are less certain—either because of low-quality evidence
recommendation or because evidence suggests that desirable and undesirable effects are closely
balanced

◆ The Highlighted Findings and Recommendations for Care section lists points determined by the PPP
Panel to be of particular importance to vision and quality of life outcomes.
◆ All recommendations for care in this PPP were rated using the system described above. Ratings are embedded
throughout the PPP main text in italics.
◆ Literature searches to update the PPP were undertaken in April 2018 and June 2019 in PubMed and the
Cochrane Library. Complete details of the literature searches are available online at www.aao.org/ppp.

P73
3
 Diabetic Retinopathy PPP

The prevalence of diabetes is increasing with increasing industrialization and globalization. Consequently,
the prevalence of diabetic retinopathy and vision-threatening diabetic retinopathy is also expected to
increase. Only about 60% of people with diabetes have recommended yearly screenings for diabetic
retinopathy. Referral to an ophthalmologist is required when there is any evidence of diabetic retinopathy.

People with type 1 diabetes should have annual screenings for diabetic retinopathy beginning 5 years after
the onset of their disease, whereas those with type 2 diabetes should have a prompt screening at the time of
diagnosis and at least yearly screenings thereafter.

Maintaining control of glucose and blood pressure lowers the risk of retinopathy developing and/or
progressing, so patients should be informed of the importance of maintaining good levels of glycosylated
hemoglobin, and blood pressure.

Patients with diabetes may use aspirin for other medical indications (as antiplatelet therapy) without an
adverse effect on their risk of diabetic retinopathy.

Women with diabetes who become pregnant should be examined early and closely in the course of the
pregnancy because the disease can progress rapidly. However, an eye examination is not required when
gestational diabetes occurs during pregnancy. Patients with diabetes have an accelerated rate of diabetic
retinopathy progression during puberty and should be followed more closely.

Intravitreal anti-vascular endothelial growth factor (anti-VEGF) agents are effective in the treatment of
center-involved diabetic macular edema with vision loss. At this time, laser photocoagulation surgery
remains the preferred treatment for non-center-involved diabetic macular edema and pan-retinal
photocoagulation (PRP) surgery remains the mainstay treatment for proliferative diabetic retinopathy (PDR).

P74
4
 Diabetic Retinopathy PPP

Diabetic retinopathy is a common complication in type 1 and type 2 diabetes. Diabetic retinopathy is
the ocular manifestation of end-organ damage in diabetes mellitus.4 Diabetic retinopathy has been
classically considered as a microvascular disease of the retina. However, growing evidence suggests
that retinal neurodegeneration is an early event in the pathogenesis of diabetic retinopathy, which
could contribute to the development of microvascular abnormalities. 5 Although defects in
neurosensory function have been demonstrated in patients with diabetes mellitus prior to the onset of
vascular lesions, the most common early clinically visible manifestations of diabetic retinopathy
include microaneurysm formation and intraretinal hemorrhages. Microvascular damage leads to
retinal capillary nonperfusion, cotton wool spots, an increased number of hemorrhages, venous
abnormalities, and intraretinal microvascular abnormalities (IRMA). During this stage, increased
vasopermeability can result in retinal thickening (edema) and/or exudates that may lead to a loss in
central visual acuity. The proliferative stage results in proliferation of new vessels on the disc, retina,
and iris, and in the filtration angle. These new vessels then lead to traction retinal detachments and
neovascular glaucoma, respectively. Vision can be substantially impaired in this stage as a result of
capillary nonperfusion or edema in the macula, vitreous hemorrhage, and distortion or traction retinal
detachment.

A description of the fundus findings in various stages of diabetic retinopathy is included in the
Natural History section, and important terms are defined in the Glossary.

The patient population includes all patients with diabetes mellitus.

◆ Identify patients at risk of developing diabetic retinopathy

◆ Encourage a collaborative approach between the patient, the primary care physician, and
subspecialists in the management of the patient’s systemic disorder, with specific attention to control
of blood sugar (hemoglobin A1c [HbA1c]), blood pressure, serum lipids, body weight, and the
management of renal disease, coronary artery disease, 6 and neuropathy
◆ Encourage and provide lifelong monitoring of retinopathy progression
◆ Treat patients with visual loss or those at risk for visual loss from diabetic retinopathy
◆ Minimize the side effects of treatment that might adversely affect the patient’s vision and/or vision-
related quality of life
◆ Provide or refer for visual rehabilitation services when a patient has visual impairment from the
disease

P75
5
 Diabetic Retinopathy PPP

◆ Refer for ophthalmological follow-up for potentially reversable vision loss such as cataracts,
glaucoma, or refractive changes
◆ Develop new technologies for telemedicine improvement

In the United States, an estimated three out of five people with diabetes have one or more of the
complications associated with the disease.7 Two main forms of diabetes mellitus are recognized. Type
1, previously called juvenile-onset or insulin-dependent diabetes, is characterized by cellular-
mediated autoimmune destruction of the beta cells in the pancreas and usually leads to severe insulin
deficiency. Type 2 diabetes was previously referred to as adult-onset or noninsulin-dependent
diabetes. Type 2 is characterized by a range of disease from insulin resistance with relative insulin
deficiency to predominately an insulin secretory defect combined with insulin resistance. Type 2
patients usually have a relative rather than an absolute insulin deficiency; they may take insulin, yet
typically do not need insulin for survival. Many patients with type 2 diabetes are obese, and obesity
itself causes relative insulin resistance. Between 90% and 95% of all patients with diabetes have type
2 diabetes.8 Because of the disproportionately large number of patients with type 2 diabetes, this
group comprises a larger proportion of the disease burden in patients with visual impairment from
diabetic retinopathy, even though type 1 diabetes is associated with more frequent and more severe
ocular complications.9,10

An estimated 100 million Americans aged 18 years and older have either been diagnosed with
diabetes or are prediabetic, according to a 2015 report by the Centers for Disease Control and
Prevention (CDC). As reported by the CDC, 30.3 million Americans 18 or older have diabetes
(9.4% of people in this age group),11 and about one-quarter are not aware that they have the
disease.12 An additional 79 million persons have impaired fasting blood glucose levels (based
on both fasting blood glucose levels and HbA1c levels).12 In 2015, an estimated 1.5 million new
cases of diabetes were diagnosed among people aged 18 and older.11

Rates of diagnosed diabetes increased with age: among individuals 18 to 44 years old, 4% had
diabetes; among those 45 to 64 years old, 17% had diabetes; and among those 65 and older,
25% had diabetes. Rates of diagnosed diabetes were higher among Native Americans and
Alaskan Natives (15.1%), non-Hispanic blacks (12.7%), and Hispanics (12.1%) compared with
Asians (8.0%) and non-Hispanic whites (7.4%).11

Rates of prediabetes (HbA1c levels between 5.7% and 6.4%) are also increasing.13 It is
estimated that 33.9% of US adults 18 or older (84.1 million people) have prediabetes based on
their fasting glucose or HbA1c level. Nearly half (48.3%) of adults 65 or older had prediabetes.11
6

P76
 Diabetic Retinopathy PPP

Age-adjusted incidence of diabetes was two times higher for people with less than a high school
education (10.4/1000 persons) compared with those with more than a high school education
(5.3/1000 persons) from 2013 to 2015. Rates of diabetes and prediabetes are similarly high
among children and adolescents (younger than 20).7 Compared with members of other US
racial and ethnic groups, non-Hispanic whites had the highest rate of new cases of type 1
diabetes. Among children and adolescents aged 10 to 19, U.S. minority populations had higher
rates of new cases of type 2 diabetes compared with non-Hispanic whites.

The 2015 CDC report notes a higher prevalence of diabetes among American Indians/Alaska
Natives (15.1%), non-Hispanic blacks (12.7%), and people of Hispanic ethnicity (12.1%) than
among non-Hispanic whites (7.4%) and Asians (8.0%) among adults aged 18 years or older.11
Americans of African descent or Hispanic ethnicity have a disproportionately high prevalence
of diabetes compared with Americans of European descent (12.6%, 11.8%, 7.0%, respectively),
whereas Asian Americans have only a slightly higher prevalence (8.4%). 12 Native Americans
and Alaskan Natives had an approximate diabetes prevalence of 6.4 per 1000 in 1990 and
increased to 9.3 per 1000 in 1998 (approximately 45% increase) in children and young adults
under the age of 35 years.14,15 Other research suggests a high prevalence of diabetes in Asia. 16,17

According to estimates based on data from the U.S. Census Bureau, approximately one-third of
Americans are at risk of developing diabetes mellitus during their lifetime. 18 With increasing
industrialization and globalization, there is a concomitant increasing prevalence of diabetes that
is leading to a worldwide epidemic.19 An alarming increase in the frequency of type 2 diabetes
in the pediatric age group has been noted in several countries, 10,20-24 including in the United
States, and has been associated with the increased frequency of childhood obesity. 25 Diabetes is
one of the most common diseases in school-aged children. Clearly, these trends predict an
increase in the number of individuals with diabetes as well as the associated increased costs for
health care and the burdens of disability associated with diabetes and its complications. In
addition, there is evidence suggesting that diabetes develops at earlier ages and carries a higher
incidence of complications among ethnic minorities.26-28

Diabetic retinopathy is a leading cause of new cases of legal blindness among working-age
Americans and represents a leading cause of blindness in this age group worldwide. 29 The
prevalence rate for retinopathy for all adults with diabetes aged 40 and older in the United
States is 28.5% (4.2 million people); worldwide, the prevalence rate has been estimated at
34.6% (93 million people).30,31 An estimate of the prevalence rate for vision-threatening
diabetic retinopathy in the United States is 4.4% (0.7 million people). Worldwide, this
prevalence rate has been estimated at 10.2% (28 million people). 30,31 Assuming a similar
prevalence of diabetes mellitus, the projected prevalence of individuals with any diabetic

P77
7
 Diabetic Retinopathy PPP

retinopathy in the United States by the year 2020 is 6 million persons, and 1.34 million persons
will have vision-threatening diabetic retinopathy.

The prevalence of diabetic retinopathy increases with increasing duration of disease. In the
United States, the prevalence is predicted to increase as the incidence and duration of diabetes
in the population increases. More than 50% of worldwide visual impairment or blindness from
diabetic retinopathy is estimated to exist in the Asia-Pacific region (51% of all those with
blindness due to diabetic retinopathy globally [n=424,400], and 56% of those with visual
impairment). Prevalence rates of diabetic retinopathy in patients with diabetes range from 10%
in India to 43% in Indonesia.32

The Chinese American Study has found slightly lower prevalence rates of diabetic retinopathy
in Chinese American than in Latino type 2 patients (35.8% in Chinese Americans vs. 42.0% in
Latinos). Increasing duration of diabetes was associated with higher probability of diabetic
retinopathy in Latinos than Chinese Americans, even after controlling for other known
predictors.33

Duration of diabetes is a major risk factor associated with the development of diabetic retinopathy.
After 5 years, approximately 25% of type 1 patients will have retinopathy. After 10 years, almost 60%
will have retinopathy, and after 15 years, 80% will have retinopathy.34,35 In the Wisconsin
Epidemiologic Study of Diabetic Retinopathy (WESDR) for patients 30 and younger, proliferative
diabetic retinopathy (PDR), the most vision-threatening form of the disease, was present in
approximately 50% of type 1 patients who had the disease for 20 years.36 In the Los Angeles Latino
Eye Study (LALES) and in Proyecto VER (Vision, Evaluation and Research), 18% of participants
with diabetes of more than 15 years’ duration had PDR, and there was no difference in the percentage
with PDR between those with type 1 and type 2 diabetes.35,37 In the recent Singapore Eye Disease
Study, independent risk factors for any diabetic retinopathy included Indian ethnicity, diabetes
duration , HbA1c , serum glucose , and systolic blood pressure 38 Diastolic blood pressure , total
cholesterol , and low-density lipoprotein cholesterol were associated with lower odds of any diabetic
retinopathy. In a study of First Nations people in Canada, HbA1c and systolic blood pressure were
found to be independent predictors of 2-step progression of diabetic retinopathy (hazard ratio, 1.42;
P<0.0001) and systolic blood pressure (hazard ratio, 1.24 per 10 mm Hg; P=0.009).39

Of type 2 patients over the age of 30 who have a known duration of diabetes of less than 5 years, 40%
of those patients taking insulin and 24% of those not taking insulin have retinopathy. These rates
increase to 84% and 53%, respectively, when the duration of diabetes has been documented for up to
19 years.40 Proliferative diabetic retinopathy develops in 2% of type 2 patients who have diabetes for
less than 5 years and in 25% of patients who have diabetes for 25 years or more. 40 Comparisons of
information from WESDR and more recent population-based studies such as Proyecto VER and

P78
8
 Diabetic Retinopathy PPP

LALES may reflect differences in blood glucose and hypertension management that have occurred
over time.

Blood sugar and blood pressure control are the key modifiable risk factors associated with the
development of diabetic retinopathy.41 Support for this association is based on both clinical trials and
epidemiologic studies.42-49 There is general agreement that duration of diabetes and severity of
hyperglycemia are the major risk factors for developing retinopathy. 41,50-53 Once retinopathy is
present, duration of diabetes appears to be a less important factor than glycemic control in forecasting
progression from earlier to later stages of retinopathy.54,55 It is recommended that a HbA1c of 7% or
lower is the target for glycemic control in most patients, whereas in selected patients, there may be
some benefit to setting a lower target of 6.5%.56 In fact, an increase in HbA1c corresponds to an
increased risk of diabetic macular edema (DME).57

Treatment of hypertension remains important, although the benefits of intensive management of

hypertension is inconclusive.58,59 Large studies have suggested that management of serum lipids may
reduce retinopathy progression and the need for treatment. 60-64 There is less agreement among studies
concerning the importance of other factors such as age, type of diabetes, clotting factors, renal
disease, physical inactivity, inflammatory biomarkers, and use of angiotensin-converting enzyme
inhibitors.54,61,65-69 Many of these factors are associated with substantial cardiovascular morbidity and
mortality and other complications associated with diabetes. Thus, ophthalmologists should encourage
patients with diabetes to be as compliant as possible with therapy of all medical aspects of their
disease.70,71

More recently, lipid-lowering agents have shown a positive effect on slowing progression of diabetic
retinopathy. In a recent meta-analysis, lipid-lowering agents showed a protective effect on diabetic
retinopathy progression and suggest a possible reduced risk of developing DME. Despite this, there
was no effect on visual acuity or on the presence of hard exudates. 72

There is conflicting evidence that genetics and epigenetic factors may explain differences in
progression rates of diabetic retinopathy between groups of individuals with similar duration of
diabetes or HbA1c levels. A study found that mitochondrial genetic haplogroups modify the risk for
progression of disease despite similar HbA1c level and duration of diabetes. For patients with
haplogroup H, longer diabetes duration and increasing HbA 1c level were significant risk factors for
PDR (P=0.0001 and P=0.011, respectively). However, for patients with haplogroup UK, neither
diabetes duration nor HbA1c level was a significant risk factor for PDR.73 A larger more recent study
looking at the same haplotypes failed to show that association. 74

Another genetic study evaluated patients with type 2 diabetes who were carriers of the HMGA1
rs139876191 variant. Patients with this variant had a significantly lower risk of developing PDR
compared with noncarrier diabetic patients.75 It is believed that the HMGA1 rs139876191 variant
confers protection by downregulating the expression of vascular endothelial growth factor A in
diabetic retinopathy.

P79
9
 Diabetic Retinopathy PPP

Metabolic syndrome refers to a group of risk factors that increases the risk for developing heart
disease, diabetes, and stroke.76 Metabolic syndrome has also been found to be associated with
microvascular and macrovascular disease in a study of patients with type 2 diabetes. More patients
with metabolic syndrome had higher rates of albuminuria (40.8% vs 21.8%; P<0.001), retinopathy
(37.9% vs 28.6%; P<0.001), coronary artery disease (19.4% vs 11.6%; P<0.001), cerebrovascular
disease (5.8% vs 3.2%; P=0.014), and an ankle-brachial index of less than 0.9 or of 1.3 or higher
(6.1% vs 3.0%; P=0.015).77 There was also a trend for stepwise increases in albuminuria, retinopathy,
coronary artery disease, cerebrovascular disease, and peripheral artery disease corresponding to the
number of metabolic syndrome components (all P for trend <0.05). Screening programs for metabolic
syndrome may therefore be helpful in finding patients at higher risk for progression.

In a recent study of 50,254 eyes, baseline features and level of nonproliferative diabetic retinopathy
(NPDR) were associated with 5-year progression to PDR.78 Eyes with IRMA had an increased hazard
ratio of developing PDR (hazard ratio, 1.77; P=0.0013) compared with eyes with venous beading, and
eyes with 4-quadrant dot-blot hemorrhages had higher risk for developing vitreous hemorrhage
(hazard ratio, 3.84; P=0.0095).78,79 For eyes with PDR, the Diabetic Retinopathy Clinical Research
Network (DRCR.net) Protocol S study found that worse baseline levels of PDR were associated with
an increased risk of PDR progressing, regardless of treatment with anti-vascular endothelial growth
factor (anti-VEGF) or panretinal photocoagulation (PRP) (64% [high-risk PDR or worse] vs 23%
[moderate PDR or better]; hazard ratio, 3.97; P<0.001). In the PRP group, eyes receiving pattern scan
versus conventional single-spot PRP were at higher risk for worsening PDR (60% vs 39%; hazard
ratio, 2.04; P=0.008), regardless of the number of spots placed. 79

Diabetic retinopathy progresses in an orderly fashion from mild to more severe stages when there is
not appropriate intervention. It is important to recognize the stages when treatment may be most
beneficial. Several decades of clinical research have provided excellent data on the natural course of
the disease and on treatment strategies that are 90% effective in preventing the occurrence of severe
vision loss.80 The outcomes of key clinical trials form a solid foundation in support of treating diabetic
retinopathy. The results of these studies are summarized in Appendices 3 and 4. Major studies include
the following (see Glossary):

◆ Diabetes Control and Complications Trial (DCCT)44,81,82

◆ Follow-up study to the DCCT titled Epidemiology of Diabetes Interventions and Complications
(EDIC)43,45,62,83,84
◆ Diabetic Retinopathy Study (DRS)85,86
◆ Early Treatment Diabetic Retinopathy Study (ETDRS) 87-89
◆ Diabetic Retinopathy Vitrectomy Study (DRVS)90
◆ Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) 91
◆ Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study 92

P80
10
 Diabetic Retinopathy PPP

◆ Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial93

◆ Diabetic Retinopathy Clinical Research Network (DRCR.net) Protocol I, S, and T studies94-96
◆ United Kingdom Prospective Diabetes Study (UKPDS)46,58,97

The nonproliferative stages of diabetic retinopathy are characterized by retinal vascular related
abnormalities such as microaneurysms, intraretinal hemorrhages, venous dilation, and cotton wool
spots. Increased retinal vascular permeability that occurs at these or later stages of retinopathy may
result in retinal thickening (edema) and lipid deposits (hard exudates). Clinically significant macular
edema (CSME) is a term commonly used to describe retinal thickening and/or adjacent hard exudates
that either involve the center of the macula or threaten to involve it. Patients with CSME should be
considered for prompt treatment, particularly when the center of the macula is already involved or if
retinal thickening and/or hard exudates are very close to the center (see Care Process). Clinically
significant macular edema can be divided into center-involved and non-center-involved macular
edema. (See Glossary.)

As diabetic retinopathy progresses, there is a gradual closure of retinal vessels that results in impaired
perfusion and retinal ischemia. Signs of increasing ischemia include venous abnormalities (e.g.,
dilation, beading, loops), IRMA, and more severe and extensive vascular leakage characterized by
increasing retinal hemorrhages and exudation. When these signs progress beyond certain defined
thresholds, severe NPDR is diagnosed (see Table 1). Such patients should be considered candidates
for treatment with panretinal (scatter) photocoagulation (see Care Process).

•
•

•
•
•

•
•

P81
11
 Diabetic Retinopathy PPP

The more advanced stage, PDR, is characterized by the onset of neovascularization at the inner
surface of the retina and into the vitreous induced by more global retinal ischemia. New vessels on or
near the optic disc (NVD) and new vessels elsewhere in the retina are prone to bleed, resulting in
vitreous hemorrhage. These new vessels may undergo fibrosis and contraction; this and other fibrous
proliferation may result in epiretinal membrane formation, vitreoretinal traction, retinal tears, and
retinal detachments. When new vessels are accompanied by vitreous hemorrhage, or when NVD
occupy greater than or equal to about one-quarter to one-third disc area, even in the absence of
vitreous hemorrhage, PDR is considered high-risk. (See Glossary.) Neovascular glaucoma can result
from new vessels growing on the iris and anterior chamber angle structures. Patients with neovascular
glaucoma or high-risk PDR should receive prompt PRP, and their treating ophthalmologist could also
consider initiating anti-VEGF therapy (see Care Process and Glossary).

Table 1 classifies diabetic retinopathy by severity based on clinical findings. The ETDRS
classification (Appendix 5) has clinical significance because risk of diabetic retinopathy progression
is associated with increasing severity level.87,88,98,99 A higher risk of incident DME in eyes with more
severe levels of baseline NPDR has been reported.100,101

A study of 2240 youths (21 or younger) with type 1 diabetes and 1768 youths with type 2 diabetes
evaluated the rates of diabetic retinopathy development between type 1 and type 2 diabetics.102 Rates
of developing diabetic retinopathy were 20.1% for type 1 and 7.2% for type 2 over a median follow-
up time of 3.2 and 3.1 years, respectively. Survival curves demonstrated that type 1 patients
developed diabetic retinopathy faster than type 2 youths (P<0.0001). The hazard for diabetic
retinopathy increases with increasing HbAlc.

P82
12
 Diabetic Retinopathy PPP

The care process for diabetic retinopathy includes a medical history, a regular ophthalmologic examination or
screening of high-quality retinal photographs of patients who have not had previous treatment for diabetic
retinopathy or other eye disease, and regular follow-up. The purpose of an effective screening program is to
determine who needs to be referred to an ophthalmologist for close follow-up and possible treatment, and
who may simply be screened annually. Early detection of retinopathy depends on educating patients who
have diabetes, as well as their family, friends, and health care providers, about the importance of regular eye
examination even though the patient may be asymptomatic. In lay terms, patients must be informed that they
may have good vision and no ocular symptoms but that they may still have significant disease that needs
treatment. They should be educated that early treatment works best and that is why they need to return for an
annual eye examination, even when their vision is good. Individuals with type 2 diabetes mellitus without
diabetic retinopathy should be encouraged to have an annual dilated eye examination to detect the onset of
diabetic retinopathy.34,40,103-120 Individuals with type 1 diabetes mellitus without diabetic retinopathy should
have annual dilated eye examinations beginning 5 years after the onset of diabetes.34,121 The recommended
timing of the first ophthalmic examination and subsequent follow-up examinations for patients with diabetes
is listed in Table 2 and described in the Management section.

Maintaining near-normal glucose levels and near-normal blood pressure lowers the risk of retinopathy
developing and/or progressing,43,44,46,58,126 so patients should be informed of the importance of maintaining
good glycosylated hemoglobin levels, serum lipids, and blood pressure. Aspirin may be used by diabetic
patients for other medical indications without concern that the aspirin therapy will worsen diabetic
retinopathy or worsen a vitreous hemorrhage should it occur.127,128

13

P83
 Diabetic Retinopathy PPP

Patient outcome criteria include the following:

◆ Improvement or stabilization of visual function

◆ Improvement or stabilization of vision-related quality of life
◆ Optimal control of blood glucose, blood pressure, and other risk factors through close communication
with the patient’s primary care physician on the status of the diabetic retinopathy and the need for
optimal metabolic control

The initial examination for a patient with diabetes mellitus includes all features of the comprehensive
adult medical eye evaluation,129 with particular attention to those aspects relevant to diabetic
retinopathy.

An initial history should consider the following elements:

◆ Duration of diabetes34,54,130
◆ Past glycemic control (HbA1c)54,82,130
◆ Medications
◆ Medical history (e.g., obesity, renal disease,34,40 systemic hypertension,34,40 serum lipid levels,131
pregnancy,123,124 neuropathy)
◆ Ocular history (e.g., trauma, other eye diseases, ocular injections, surgery, including retinal laser
treatment and refractive surgery)

The initial examination should include the following elements:

◆ Visual acuity132
◆ Slit-lamp biomicroscopy
◆ Intraocular pressure (IOP)
◆ Gonioscopy before dilation, when indicated. Iris neovascularization is best recognized prior to
dilation. When neovascularization of the iris is present or suspected, or if the IOP is elevated,
undilated gonioscopy can be used to detect neovascularization in the anterior chamber angle.
◆ Pupillary assessment for optic nerve dysfunction
◆ Thorough fundoscopy, including stereoscopic examination of the posterior pole 89
◆ Examination of the peripheral retina and vitreous

A dilated pupil is preferred to ensure optimal examination of the retina, because only 50% of
eyes are correctly classified for the presence and severity of retinopathy through undilated
pupils.133 Slit-lamp biomicroscopy is the recommended method to evaluate retinopathy in the

P84
14
 Diabetic Retinopathy PPP

posterior pole and midperipheral retina with a 90D or 78D lens.89 Examination of the peripheral
retina is best performed using indirect ophthalmoscopy or slit-lamp biomicroscopy.

Because treatment is effective in reducing the risk of visual loss, a detailed examination is
indicated to assess for the following features that often lead to visual impairment:

◆ Macular edema
◆ Signs of severe NPDR (extensive retinal hemorrhages/microaneurysms, venous beading, and
IRMA)
◆ Optic nerve head neovascularization and/or neovascularization elsewhere
◆ Vitreous or preretinal hemorrhage

Many studies of patients with type 1 diabetes have reported a direct relationship between
the prevalence and severity of retinopathy and the duration of diabetes. 40,134,135 The
development of vision-threatening retinopathy is rare in children prior to puberty.134,136
Among patients with type 1 diabetes, substantial retinopathy may become apparent as early
as 6 to 7 years after onset of the disease.34 Ophthalmic examinations are recommended
beginning 5 years after the diagnosis of type 1 diabetes and annually thereafter, which will
detect the vast majority of type 1 patients who require therapy.34,121 Patient education about
the visual impact of early glucose control is important and should begin with the onset of
disease.

The time of onset of type 2 diabetes is often difficult to determine and may precede the
diagnosis by a number of years.137 Up to 3% of patients whose diabetes is first diagnosed at
age 30 or later will have CSME or high-risk features at the time of the initial diagnosis of
diabetes.34 About 30% of patients will have some manifestation of diabetic retinopathy at
diagnosis. Therefore, the patient should be referred for ophthalmologic evaluation at the
time of diagnosis.40,122

Diabetic retinopathy can worsen during pregnancy due to the physiologic changes of
pregnancy itself or changes in overall metabolic control. 123-125 Patients with diabetes who
plan to become pregnant should have an ophthalmologic examination prior to pregnancy
and counseled about the risk of development and/or progression of diabetic retinopathy.
The obstetrician or primary care physician should carefully guide the management of the
pregnant patient with diabetes’ blood glucose, blood pressure, as well as other issues
related to pregnancy.123-125 During the first trimester, an eye examination should be
performed with repeat and follow-up visits scheduled, depending on the severity of

P85
15
 Diabetic Retinopathy PPP

retinopathy. (See Table 2.) Women who develop gestational diabetes138 do not require an
eye examination during pregnancy and do not appear to be at increased risk for diabetic
retinopathy during pregnancy.

After the examination, the ophthalmologist should discuss the results and their implications
with the patient. Both eyes should be classified according to the categories of diabetic
retinopathy and macular edema discussed in the Natural History and Treatment sections.
Each category has an inherent risk for progression and is dependent upon adherence to
overall diabetes control. Thus, the diagnostic category, combined with the level of diabetes
control, determines the timing for both the intervention and follow-up examination.

Patients with diabetes have an accelerated rate of diabetic retinopathy progression during
puberty. This relative risk has been reported to be 4.8 in pubescent patients compared with
their prepubescent counterparts despite similar durations of diabetes mellitus.121

If used appropriately, a number of tests ancillary to the clinical examination may enhance
patient care. The most common tests include the following:

◆ Color and red-free fundus photography

◆ Optical coherence tomography (OCT)
◆ Fluorescein angiography (FA)
◆ OCT angiography139-142
◆ B-scan ultrasonography

Fundus photography (with or without pupillary dilation) is a reproducible technique for

detecting diabetic retinopathy and has been used in large clinical research studies. Fundus
photography is also useful for documenting the severity of the diabetes, the presence of
new vessels elsewhere in the retina and NVD, the response to treatment, and the need for
additional treatment at future visits.

Optical coherence tomography provides high-resolution imaging of the vitreoretinal

interface, neurosensory retina, and subretinal space. It can be used to quantify retinal
thickness, monitor macular edema, identify vitreomacular traction, and detect other forms
of macular disease in patients with DME.143-148 (See Table 3.) Large clinical trials testing
anti-VEGF treatment have utilized OCT rather than stereoscopic photographs or clinical
examination to evaluate and follow macular edema status because it allows an objective,
accurate assessment of the amount and location of retinal thickening.94,149-153 In clinical
practice, decisions are often based on OCT findings. For example, the decision to treat with
P86
16
 Diabetic Retinopathy PPP

anti-VEGF injections, change therapeutic agents (e.g., intraocular corticosteroids), initiate

laser treatment, or even consider vitrectomy surgery is often based in part on OCT findings.
Nevertheless, retinal thickness, even when measured by OCT, is not always consistently
correlated with visual acuity.154,155 Optical coherence tomography can demonstrate the
microstructural changes secondary to ischemia. Loss of inner retinal layers at the fovea with
high-resolution spectral-domain OCT has been shown to correlate with vision loss in eyes
with diabetic macular ischemia.

⚫
⚫
⚫

⚫
⚫

Routine FA is not indicated as a part of the regular examination of patients with diabetes.
Clinical examination, OCT, and/or FA are used in the diagnosis of macular edema and
PDR. As the use of anti-VEGF agents and intraocular corticosteroids has increased for the
treatment of macular edema, the use of focal laser surgery has decreased. Therefore, the
need for angiography that localizes leaking microaneurysms or areas of capillary dropout
has also declined.

Nevertheless, FA is useful to differentiate diabetic macular swelling from other macular

disease or for a patient with unexplained vision loss. (See Table 4.) Angiography can
identify macular capillary nonperfusion 156 appearing as enlargement of the foveal
avascular zone or anywhere in the macular region as an explanation for vision loss that is
unresponsive to therapy. Fluorescein angiography may also detect areas of untreated retinal
capillary nonperfusion that could explain persistent retinal or disc neovascularization after
previous scatter laser surgery. Advances in widefield FA have shown improved detection
of peripheral ischemia and peripheral lesions, including neovascularization that may not be
clinically apparent. 157 Thus, FA remains a valuable tool, and facilities for conducting FA
should be available to physicians who diagnose and treat patients with diabetic retinopathy.

P87
17
 Diabetic Retinopathy PPP

An ophthalmologist who orders FA must be aware of the potential risks associated with the
procedure, because severe medical complications may occur, including death in about
1/200,000 patients.158 Each angiography facility should have in place an emergency care
plan and a clear protocol to minimize the risks and to manage complications. Fluorescein
dye crosses the placenta into the fetal circulation, 159 but detrimental effects of fluorescein
dye on a fetus have not been documented.

The use of OCT angiography (OCTA) has added a new perspective on our understanding of
diabetic retinopathy. Although the technology is FDA approved, the guidelines and
indications for use during screening and management of diabetic retinopathy are currently
evolving. The major advances offered by OCTA have been its noninvasive nature and the
ability to visualize depth-resolved, capillary-level abnormalities in the three retinal
plexuses, offering a much more quantitative assessment of macular ischemia. 139-142,160-163
Even though the technology is very effective at revealing vascular abnormalities, including
neovascularization on the surface of the retina and optic nerve, it is not capable of
visualizing leakage, which could be construed as possible limitation, though it permits a
much better unperturbed view of the underlying ischemia. 164,165 Using this technique
preclinical microvascular changes can be detected,166 regions of macular nonperfusion can
be quantified, where studies have shown that nonperfusion correlates to severity of diabetic
retinopathy,140,162 and retinal neovascular tissue can be identified.167,168 The current
limitations include projection artifacts and the lack of consensus on segmentation
algorithms.169,170 They should also include a reduced field of view, which limits the view of
peripheral retinal ischemia and neovascularization unless the clinicians use image
montages.171-173

18
P88
 Diabetic Retinopathy PPP

Ultrasonography is an extremely valuable diagnostic tool that enables assessment of the

status of the retina in the presence of a vitreous hemorrhage or other media opacity. It can
be used to assess the amount of vitreous hemorrhage and to define the extent and severity of
vitreoretinal traction and diagnose diabetic retinal detachments in the setting of media
opacity.

Untreated diabetic retinopathy and its accompanying visual loss result in a substantial economic
burden on patients, their family and society. Treatment with laser, anti-VEGF agents, or intravitreal
corticosteroids is cost-effective for managing diabetic retinopathy to varying degrees.174,175 Choice of
laser, individual anti-VEGF agents, or approved intravitreal corticosteroids should be individually
tailored based on discussion between the patient and physician.

Management of diabetic retinopathy includes following a healthy diet and lifestyle, medical
management, timely ophthalmic evaluation, and treatment under the care of an ophthalmologist.
Because patients with diabetes may be under the care of multiple practitioners, effective
communication and care coordination is necessary to optimize care. 176 Physicians and patients need to
be educated and informed of the need for ophthalmic referral and routine surveillance. Finally,
patients need to understand that current treatments often require multiple visits and evaluations over
time for adequate delivery of therapeutic effect.

A healthy diet and lifestyle that includes exercise and weight control may decrease the risk of
developing diabetes in some patients.177,178 The visual complications of diabetes mellitus can at
least be moderated by a healthy lifestyle; however, diabetes complications simply cannot be
prevented in all cases.

The DCCT showed that the development and progression of diabetic retinopathy in patients
with type 1 diabetes can be delayed when the HbA1c is optimized.44 (See Appendix 4.)
Establishing a close partnership between the ophthalmologist and the primary care physician is
an important step to ensure optimal patient care. Furthermore, it is important to help educate
patients with diabetes as well as their primary care physician about the ophthalmologic
implications of controlling blood glucose (as monitored by HbA 1c) to as near normal as is safely
possible. Results from multiple studies have demonstrated the value of controlling blood
glucose, serum lipid levels, and blood pressure in patients with type 2 diabetes. (See Appendix
4 for further information.)

The ETDRS found that aspirin therapy at a dose of 650 mg per day does not slow the
progression of diabetic retinopathy.127 Also, any aspirin therapy did not cause more severe,
more frequent, or longer-lasting vitreous hemorrhages in patients with PDR.128 As such, aspirin
appears to be neither helpful nor harmful in the management of diabetic retinopathy. Therefore,

P89
19
 Diabetic Retinopathy PPP

no recommended changes in medically administered aspirin therapy are indicated in the setting
of diabetic retinal disease.

Diabetic retinopathy may be asymptomatic for years, even at an advanced stage, so screening,
using new technologies such as telemedicine, is essential to identify, monitor, and guide the
treatment of disease. When visual complications occur, treatment preserves visual function and
is believed to yield a substantial cost savings when compared with the direct costs for
individuals disabled by vision loss (see Socioeconomic Considerations section). According to
the National Committee for Quality Assurance’s Health Plan Employers Data Information Set
System, national monitoring of quality data has shown a slow but definite trend toward
improving rates of screening examinations.179 Still, screening rates remain lower than ideal in
spite of evidence supporting the effectiveness of treatment. Physicians who care for patients
with diabetes, and patients themselves, need to be educated about indications for
ophthalmologic referral. (See Table 5.)

P90
20
 Diabetic Retinopathy PPP

Analyses from two clinical trials show that treatment for diabetic retinopathy may be 90%
effective in preventing severe vision loss (visual acuity <5/200) using current therapeutic
treatment strategies.80 Although effective treatment is available, fewer patients with diabetes are
referred by their primary care physicians for ophthalmic care than would be expected according
to guidelines by the American Diabetes Association and the American Academy of
Ophthalmology.180 In two community-based studies, 43% to 65% of participants had not
received a dilated eye examination at the time of enrollment. 181,182

The purpose of an effective screening program for diabetic retinopathy is to determine who
needs to be referred to an ophthalmologist for close follow-up and possible treatment and who
may simply be screened annually. Some studies have shown that screening programs using
digital retinal images taken with or without dilation may enable early detection of diabetic
retinopathy along with an appropriate referral. 103-113 Optical coherence tomography appears to
be an effective and sensitive imaging tool for detecting DME as long as there are no other
causes for cystoid macular edema.147,183 (I+, Good quality, Strong recommendation)

Studies have found a positive association between participating in a photographic screening

program and subsequent adherence to receiving recommended comprehensive dilated eye
examinations by a clinician.114,115 Of course, such screening programs are more relevant when
access to ophthalmic care is limited.116-119 Screening programs should follow established
guidelines.120 Given the known gap in accessibility of direct ophthalmologic screening, retinal
imagining screening programs may help increase the chances that at-risk individuals will be
promptly referred for more detailed evaluation and management.

Management recommendations for patients with diabetes are described according to severity of
the retinopathy as well as the presence and type of DME. Diabetic macular edema should be
classified as either center-involved (CI-DME) or noncenter-involved DME (NCI-DME).
Follow-up recommendations and treatment options based on severity of disease are summarized
in Table 5. Diabetic macular edema can be present in all stages of diabetic retinopathy.
Clinicians need to consider certain treatment interactions when deciding treatment options. For
example, DME can worsen following PRP for PDR.184 There have been case reports of
idiosyncratic macular edema that is temporally associated with use of the glitazone class of oral
antihyperglycemic agents.185,186 Alternatively, the severity of diabetic retinopathy can improve
in eyes receiving treatment with anti-VEGF treatment for DME.95,187 Table 5 provides guidance

P91
21
 Diabetic Retinopathy PPP

for managing of patients with diabetes; however, individual patient needs may vary. Table 6
summarizes the side effects and complications associated with currently available treatments.

Historically, CSME is defined by the ETDRS to include any of the following features:

◆ Thickening of the retina at or within 500 µm of the center of the macula

◆ Hard exudates at or within 500 µm of the center of the macula, when associated with
adjacent retinal thickening. (This criteria does not apply to residual hard exudates that
remain after successful treatment of prior retinal thickening.)
◆ A zone or zones of retinal thickening 1 disc area or larger, where any portion of the
thickening is within 1 disc diameter of the center of the macula
Because the risk of visual loss is greatest if macular edema is at the center of the macula
DME is now subdivided as either center involved (CI-DME) or noncenter-involved (NCI-
DME) . OCT is the best way to detect and quantitate CI-DME and recent clinical trials
have required CI-DME as inclusion criteria. A Diabetic Retinopathy Clinical Research
Newtwork (DRCR.net) study determined a reasonable clinical threshold for CI-DME was a
central macular thickness 2 standard deviations above the normative study population of
diabetics without macular edema.192 Changes in central macular thickness measurements
based on OCT is a useful marker for assessing response to treatment. Treating
ophthalmologists should be familiar with relevant studies and techniques as described in
the ETDRS, trials under the guidance of the DRCR.net Protocol trial,94 and other studies
involving anti-VEGF treatment.89,156

Patients commonly present with good visual acuity despite the presence of CI-DME. An
estimated 40% of eyes with DME in the ETDRS had visual acuity of 20/20 or better. 193
Studies that have demonstrated the benefit of anti-VEGF therapy for CI-DME required
visual acuity loss (20/32 or worse).151,194,195 DRCR Protocol V found that in eyes with
CI-DME and visual acuity of 20/25 or better, there was no difference in visual acuity
loss in eyes treated with aflibercept, focal laser photocoagulation with aflibercept if
visual acuity decreased per criteria, or observation with aflibercept if visual acuity
decreased per criteria.190 The visual criteria for adding aflibercept to the laser or
observation strategy were a decrease from baseline by at least 10 letters (>2 lines on an
eye chart) at any one visit or by 5 to 9 letters (1 to 2 lines) at two consecutive visits.
After 2 years, all three strategies resulted in mean visual acuity of 20/20 and the central
subfield thickness on OCT did not significantly change compared with baseline. In eyes
with good visual acuity and CI-DME, treatment is reasonably deferred until the visual
acuity is affected (20/30 or worse). These patients should be examined every 2 to 4
months.89

P92
22
 Diabetic Retinopathy PPP

Multiple, high quality clinical trials have demonstrated that anti-VEGF therapy is more
effective in improving vision in CI-DME than monotherapy with focal laser treatment,
supplanting it as the first-line therapy.89,94,150,156,187,189,196-201 With a monthly or a
protocol-driven strategy such as DRCR.net studies with anti-VEGF, eyes with vision
worse than 20/32 or worse due to CI-DME gained around 2 lines of vision at 2 years
compared with stabilization of vision with focal treatment alone. This was demonstrated
with ranibizumab, bevacizumab, and aflibercept. A significant portion of patients in
these trials (30%-46%) underwent focal laser treatment. The timing of the laser—
deferred or prompt—did not affect the outcome. DRCR Protocol T, a head-to-head trial
comparing bevacizumab, ranibizumab, and aflibercept, demonstrated effectiveness for
all three agents with comparable safety profile in eyes with CI-DME. For eyes with
visual acuity of 20/40 or better, the visual gains were similar between the three groups.
In eyes with visual acuity of 20/50 or worse, mean visual acuity gains were 18.3, 13.3
and 16.1 letters for aflibercept, bevacizumab, ranibizumab, respectively at 2 years, with
a statistically significant difference only found between aflibercept and bevacizumab
groups. In the second year, the average number of injections decreased to about half of
the number in the first year. Over 2 years, the percentage of eyes undergoing focal laser
for persistent edema was 41%, 64%, and 52% for aflibercept, bevacizumab, and
ranibizumab groups, respectively (all pairwise comparisons were P <0.05).

The DRCR protocol for CI-DME starts with monthly injections for 4 to 6 months
initially, then allows for holding treatment if there is no improvement in vision or
central macular thickness, or if 20/20 vision and/or resolution of macular edema has
been achieved. If there is worsening vision or central macular thickness on subsequent
visits, injection is resumed. If consecutive visits do not require treatment, the follow up
interval is doubled up to 4 months. This approach has been demonstrated to reduce the
number of injections while delivering excellent visual acuity gains.

An alternative approach to reducing the injection burden is treat-and-extend, whereby

the interval between visits is adjusted based on the treatment response. A recent
prospective trial showed that this approach is comparable in visual and anatomic results
at 2 years to monthly dosing with fewer injections. 202

The DRCR.net Protocol T demonstrated that anti-VEGF therapy using either

bevacizumab, ranibizumab, or aflibercept is effective treatment for CI-DME.96 The 2-
year results did not reveal a statistical difference among the three drugs in serious
adverse events and all three treatments provided substantial visual acuity improvement.
In eyes with visual acuity of 20/40 or better, there were no visual acuity differences
between treatment regimens. In eyes 20/50 or worse, aflibercept was superior to

23
P93
 Diabetic Retinopathy PPP

ranibizumab and bevacizumab at year 1. However, at year 2, the mean visual acuity in
the aflibercept group was superior only to the bevacizumab group.155

The visual acuity gain and reduction in macular thickness following administration of
combined intravitreal ranibizumab, with prompt or deferred laser surgery, had better
outcomes than laser alone after 2 years of follow-up.189 Appendix 3 summarizes the
results of several studies that have demonstrated the benefit of different anti-VEGF
agents for CI-DME. Based on these studies, anti-VEGF therapy is the initial treatment
choice for CI-DME, with possible subsequent focal laser treatment for persistent edema.
The Ranibizumab for Edema of the Macula in Diabetes-2 (READ-2) study involved
126 patients randomized to either anti-VEGF therapy (in this case ranibizumab alone),
laser alone, or focal/grid laser combined with anti-VEGF therapy. (See Glossary.) The
group that received anti-VEGF therapy alone or with laser treatment did better than the
group treated with laser alone.203 The DRCR.net Protocol I also showed that anti-VEGF
with either prompt or deferred laser photocoagulation surgery was better than either
laser alone or laser combined with triamcinolone acetonide. 94 (See Glossary.) Prompt
laser demonstrated no additional benefit. During the 2 years of the RISE and RIDE
trials, approximately 30% of patients were treated with focal laser.187 In the DRCR.net
Protocol I, 46% of patients were treated with laser for persistent CI-DME prior to the 3-
year visit.94 In this study, after 6 months of treatment, as-needed protocol was followed,
and the number of injections decreased in years 2 and 3 while visual acuity remained
stable. It is possible that focal laser for persistent macular edema despite anti-VEGF
treatment may reduce the number of injections. The studies above used ranibizumab,
whereas the Bevacizumab or Laser Treatment in the Management of Diabetic Macular
Edema (BOLT) study showed favorable outcomes for bevacizumab over macular laser
treatment in eyes with CI-DME.204 (See Glossary.) The DME and VEGF Trap-Eye:
Investigation of Clinical Impact (DA VINCI) study demonstrated better outcomes using
aflibercept over laser treatment for CI-DME.195 (See Glossary.) A meta-analysis
provided additional evidence that both ranibizumab and aflibercept have superior
efficacy for DME treatment compared with conventional laser. 205 (I++, Good Quality,
Strong Recommendation)

The most serious complication of anti-VEGF injections is infectious endophthalmitis with

rates between 0.019% and 0.09% in clinical trial settings.206 The use of topical povidone
iodine is recommended for intravitreal injections as its non-use has been reported to have an
unacceptably high risk of endophthalmitis. The use of routine antibiotic eye drops is not
recommended before or following intravitreal injection procedures, because it does not
decrease the risk of endophthalmitis.207 Other complications, such as retinal detachment,
cataract formation, and sustained elevated IOP are rare. 208-210 Individuals receiving the
intravitreal injections of anti-VEGF agents may be examined at 1 month following therapy.

24
P94
 Diabetic Retinopathy PPP

(See Table 5.) Systematic adverse events, particularly thromboembolic events, have been
considered a potential side effect of anti-VEGF treatment. An additional meta-analysis
suggests there may be a modest increased risk of death and cerebrovascular events in
patients receiving monthly therapy for two years.211 (See Table 7.). However, a 2018
Cochrane systematic review has reported that there is “moderate certainty evidence” of
safety of anti-VEGF injections and as of 2019 no studies have shown a definite increased
risk.212 (I+, Moderate quality, Strong recommendation)

The ETDRS demonstrated that focal laser photocoagulation reduces the risk of
moderate vision loss in eyes with CSME.85,213,214 The DRCR.net Protocols B and I
demonstrated a beneficial treatment effect of focal laser treatment for CI-DME . The
role of anti-VEGF in NCI-DME has not been studied, and the focal/grid laser treatment
option is recommended in this scenario. A modified ETDRS laser treatment is currently
recommended; it includes a less intense laser treatment, has greater spacing than for a
grid, directly targets microaneurysms, and avoids foveal vasculature within at least 500
µm of the center of the macula.215 A recent Cochrane systematic review concluded that
laser photocoagulation reduces the changes of visual loss and increases those of partial
to complete resolution of DME compared to no intervention at 1-3 years.216 (I,
Moderate quality, Strong recommendation) Preoperatively, the ophthalmologist should
discuss with the patient the side effects and risks of treatment.89,156 Fluorescein
angiography prior to laser surgery for CSME can be helpful for identifying leaking
microaneurysms in areas of thickened retina. Fluorescein angiography is also useful for
detecting capillary dropout and pathologic enlargement of the foveal avascular zone,
information that may be useful when planning focal laser treatment.89 Optical coherence
tomography angiography can detect capillary drop out and enlarged an foveal avascular
zone; however, it does not reveal leakage. A posttreatment evaluation should be
scheduled within 3 to 4 months of laser surgery.89 Rarely, focal laser photocoagulation
surgery may induce subretinal fibrosis with choroidal neovascularization, a
complication that may be associated with permanent central vision loss. 217-219 Other
than choroidal neovascularization, the most important factor associated with the
development of subretinal fibrosis includes both the more severe levels of subretinal
hard exudates and elevated serum lipids prior to laser photocoagulation surgery.220

Several studies have evaluated the use of intravitreal administration of short- and long-
acting corticosteroids for the treatment of DME. Topical corticosteroids and periocular
steroid injection demonstrated no significant benefit.221 The role of intravitreal
triamcinolone acetonide was compared with focal laser photocoagulation surgery.
Treatment with intravitreal triamcinolone
P95 acetonide resulted in an early decrease in
25
 Diabetic Retinopathy PPP

retinal thickness at 4 months, yet by 24 months those patients randomized to focal/grid

laser photocoagulation surgery had better mean visual acuity. Of the triamcinolone
group, half of phakic eyes underwent cataract surgery within 2 years and about 30% of
eyes developed elevated IOP above 10 mm Hg compared with baseline.222 At 3 years,
these results were largely unchanged.223 A subsequent study showed that pseudophakic
eyes treated with the combination of the intravitreal triamcinolone acetonide and focal
laser had visual gains similar to eyes treated with anti-VEGF agents.224 The sustained-
release dexamethasone implant for treatment naïve CI-DME improved visual acuity
compared with sham treatment. In this study, the mean number of treatments was four
to five injections over 3 years’ time.225 The fluocinolone acetonide implant for DME
treatment study revealed improved visual acuity relative to sham at 3 years. At three
years, 75% of patients were treated with only one implant. Rates of cataract extraction
of phakic eyes was 74.9% with an implant versus 23.1% for sham. Rates of incisional
glaucoma surgery were 3.7% versus 0.5% for sham at 2 years.226

The DRCR.net phase II, randomized clinical trial evaluated the role of combination
anti-VEGF treatment with intravitreous dexamethasone in a sustained-release drug
delivery system to eyes with persistent CI-DME after at least three anti-VEGF
injections in the previous 20 weeks.227 The addition of the dexamethasone implant
reduced central macular thickness; however there was no benefit in visual acuity.
Pseudophakic eyes improved by 3 letters, but there was insufficient power to be
confident of this effect.

A Cochrane systematic review concluded that a combination of steroid with anti-VEGF

did not provide additional benefit to anti-VEGF monotherapy.228 (I, Moderate quality,
Strong recommendation) However, the evidence base for this conclusion was rated as
low-certainty given the relative paucity of studies with long-term follow-up.228 Multiple
studies consistently found that corticosteroids carry higher risk for cataract and elevated
IOP compared with anti-VEGF therapy (See Table 6).189,197

Studies of intravitreal corticosteroids for DME have evaluated them as first-line agents
only. Because of their side-effect profile, including cataract progression and elevated
IOP, they are generally used as second-line agents for DME, especially for phakic
patients. To date, no large randomized clinical trial has evaluated the use of intravitreal
corticosteroid injection as a rescue treatment for eyes with persistent DME after anti-
VEGF injection therapy.

When substantial vitreomacular traction is present, pars plana vitrectomy may improve
visual acuity in selected patients who have diffuse CSME that is unresponsive to
previous macular laser photocoagulation surgery and/or anti-VEGF therapy.229-231 The

P96
26
 Diabetic Retinopathy PPP

DRCR.net Protocol D found that 38% of eyes with DME and vitreomacular traction had
improved visual acuity, whereas 22% of eyes experienced visual acuity loss. However,
the value of vitrectomy in CSME is difficult to study in a randomized clinical trial, as
there are many variables that affect visual acuity. (See DRCR.net Protocol D.232)
Because the majority of studies evaluating vitrectomy for DME precede the use of anti-
VEGF treatment, it is difficult to determine the role of vitrectomy with concomitant
anti-VEGF treatment.

Some authors have suggested that micropulse laser induces less damage to the
macula.233 A recent meta-analysis found no difference in visual acuity with
conventional laser photocoagulation surgery compared with subthreshold diode
micropulse laser photocoagulation surgery.234

A Cochrane systematic review did not find any randomized controlled clinical trials
evaluating use of NSAIDS for DME.235

P97
27
 Diabetic Retinopathy PPP

•
•

•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

The patient with a normal retinal examination or with rare microaneurysms should be
re-examined annually,34 because within 1 year 5% to 10% of patients without
retinopathy will develop diabetic retinopathy. Existing retinopathy will worsen by a
similar percentage.65,66,81

Patients with retinal microaneurysms and occasional blot hemorrhages or hard exudates
should be re-examined within 6 to 12 months, because disease progression is
common.65 In The Wisconsin Epidemiologic Study of Diabetic Retinopathy, the natural
history of type 1 diabetic patients suggests that approximately 16% of patients with
mild retinopathy (hard exudates and microaneurysms only) will progress to proliferative
stages within 4 years.65

P98
28
 Diabetic Retinopathy PPP

For patients with mild NPDR, the 4-year incidence of either CSME or macular edema
that is not clinically significant is approximately 12%. For moderate NPDR, the risk
increases to 23% for patients with either type 1 or 2 diabetes.132 For patients undergoing
anti-VEGF treatment, the clinically observed level of retinopathy may become
consistent with mild to moderate retinopathy. Especially when anti-VEGF treatment is
stopped because edema is well controlled, and the patient had previously been noted to
have a higher level of retinopathy, a closer follow-up of retinopathy may be necessary,
as the progression of disease may be higher in those patients.

The DRS demonstrated that eyes with severe NPDR and non-high-risk PDR had a
reduced risk of severe vision loss with PRP but suggested that a deferral of
photocoagulation is reasonable until high-risk characteristics develop.246 The ETDRS
showed that although deferral of PRP until high-risk characteristics develop, especially
in eyes with DME, early PRP could be considered, especially for eyes with very severe
NPDR and non-high risk PDR, who have close to a 50% risk of progressing to high-risk
PDR within 1 year. Very severe NPDR is defined as an eye with 2 or more of the 4-2-1
characteristics summarized in Table 1.

Severe NPDR and non–high-risk PDR are discussed together because ETDRS data
showed that they have a similar clinical course and subsequent recommendations for
treatment are similar. The study demonstrated that the risk of progression to
proliferative disease was high, with 45% of patients with very severe NPDR, 46% of
patients with moderate PDR, 22% of patients with mild PDR, and 15% of patients with
severe NPDR developing PDR within 1 year.132 Therefore, these patients should be re-
examined within 2 to 4 months.1,132 Refer to Table 1 for the definition of severe NPDR
and very severe NPDR.

The presence of any three of the following four features characterizes DRS high-risk
PDR:85,86

◆ Neovascularization (at any location)

◆ Neovascularization at or near the optic disc (see standard photograph 10A in

Glossary)

◆ At least moderate neovascularization, defined as:

o New vessels within 1 disc diameter of the optic nerve head that are
larger than one-quarter to one-third disc area in size

o New vessels elsewhere that are at least one-half disc area in size

P99
29
 Diabetic Retinopathy PPP

o Vitreous or preretinal hemorrhage

The DRS showed that the risk of severe visual loss among patients with high-risk PDR is
high and is reduced substantially by PRP. (See Glossary) Most patients with high-risk PDR
should receive PRP expeditiously, as it usually induces regression of retinal
neovascularization.85,247

The DRCR.net study Protocol S that examined patients with PDR primarily has
demonstrated that a series of anti-VEGF injections (ranibizumab was used in this protocol)
is noninferior to PRP at 2 years.95 The patients undergoing anti-VEGF injections were less
likely to have worsening macular edema or to have peripheral vision loss as measured by
automated visual field testing compared with the PRP group. However, when patients with
PDR undergoing anti-VEGF injections are lost to follow up, their visual and anatomic
outcomes are inferior to those who received PRP. 248 Therefore, the decision to choose anti-
VEGF over PRP must be made cautiously with a careful consideration of patient-related
factors. The anti-VEGF injection alone could be considered for patients with reliable
follow-up.

Additional PRP or anti-VEGF therapy should be considered in situations involving the

following:

◆ Failure of the neovascularization to regress

◆ Increasing neovascularization of the retina or iris

◆ New vitreous hemorrhage

◆ New areas of neovascularization

In cases of involutional PDR, vitreous hemorrhage may occur due to vitreous traction on
involuted neovascularization. These eyes may not necessarily require additional PRP,
especially in the absence of venous dilation. Pars plana vitrectomy should be considered for
patients with PDR and vitreous opacities interfering with vision or treatment, severe
fibrovascular proliferation, and traction retinal detachment that is threatening or involving
the macula.90,249-251 The value of early pars plana vitrectomy increases with the increasing
severity of neovascularization. (See Appendix 3.) The role of anti-VEGFs in these later
stages of proliferative retinopathy is under investigation.

Panretinal photocoagulation has been demonstrated to reduce the risk of severe vision
loss in PDR and severe NPDR. The ETDRS protocol for full PRP included 1200 to1600
spots of moderate burns of 0.1 second duration that is a one-half burn width apart and at
least 2 disc diameters from the fovea out to the equator.132 If laser surgery is elected,
full PRP is a proven treatment approach. Partial or limited PRP treatment is not

P100
30
 Diabetic Retinopathy PPP

recommended.85 Fluorescein angiography does not usually need to be performed to

apply the PRP effectively.

Additional analyses of visual outcome in ETDRS patients with severe NPDR to non–
high-risk PDR suggest that the recommendation to consider PRP before the
development of high-risk PDR is particularly appropriate for patients with type 2
diabetes. The risk of severe vision loss or vitrectomy was reduced by 50% (2.5% vs.
5%; P=0.0001) in patients with type 2 diabetes who were treated early when compared
with deferral PRP until high-risk PDR developed.1 For patients with type 1 diabetes, the
timing of the PRP depends on the patient’s compliance with follow-up and the status
and response to treatment of the fellow eye. For both patients with type 1 and type 2
diabetes, impending or recent cataract surgery or pregnancy may increase the risk of
progression and may influence the decision to perform PRP.

The goal of PRP is to reduce the risk of vision loss. Preoperatively, the ophthalmologist
should assess for the presence of macular edema, discuss side effects of treatment and
risks of visual loss with the patient, and obtain informed consent.213,214 This technique
has been fully described85,213 and the results are summarized in Appendix 3.

The results of clinical trials suggest that PRP is to be performed on eyes with CSME;
focal photocoagulation and/or anti-VEGF therapy prior to or concomitant with PRP
should be performed when there is evidence that PRPmay exacerbate macular edema
and increase the rate of moderate visual loss (i.e., doubling of the visual angle)
compared with untreated control eyes.132 (See Glossary.) However, PRP should not be
delayed when PDR is at the high-risk stage (i.e., if NVD is extensive or
vitreous/preretinal hemorrhage has occurred recently). In such cases, anti-VEGF
therapy and PRP may be performed concomitantly. For patients who have concurrent
CI-DME, combined anti-VEGF therapy and PRP at the first treatment session should be
considered (Table 6).

The DRCR.net Protocol S was a randomized controlled trial that compared PRP with
ranibizumab in patients primarily with PDR with and without DME, and approximately
11% had mild to severe NPDR.252 The patients received ranibizumab monthly for 6
months, unless complete neovascular regression was obtained at 4 months, followed by
treatment as needed based on a specific protocol for evaluating the presence and/or
activity of retinal neovascularization.253 The study concluded that ranibizumab resulted
in not more than 5 letters worse visual acuity than PRP at 2 years. The ranibizumab
group seemed to have better average visual acuity, less visual field loss, fewer
vitrectomies, and fewer new developments of DME-related vision loss. However, the
ranibizumab group had a higher number of treatments and visits than the group

P101
31
 Diabetic Retinopathy PPP

receiving PRP.254 Patient compliance is a major concern for management of patients

with PDR. An additional study demonstrated that aflibercept is similar to PRP for
treatment of PDR and may have superior visual acuity outcomes in eyes without CI-
DME at 1 year.255 A follow-up of patients from the RIDE and RISE studies found that
more patients receiving ranibizumab treatment had a 2-step or 3-step or more
improvement in diabetic retinopathy compared with the sham crossover group at a
median level of moderate NPDR. (See Glossary.) 256 It is not yet known whether anti-
VEGF treatment would benefit patients with severe NPDR for whom PRP is
considered.

A key clinical consideration for determining the use of anti-VEGF versus PRP is the
reliability of patient follow-up. A recent analysis found that over a 4-year period, 22%
of patients with PDR under treatment with anti-VEGF injections were lost to follow-
up.248 Further studies are required to determine the long-term implications of using anti-
VEGF agents alone.188 Recent reports raise into question the implications of using anti-
VEGF therapy in PDR patients and the severe consequences of such a decision and a
higher rate of NVG.257 The clinical indications for use in patients with moderate or mild
NPDR are unknown and also depend on other factors such as systemic blood glucose
control and compliance with follow-up examinations. Clinical judgment is important for
guiding therapy.

Although some studies have reported evidence for the beneficial use of anti-VEGF for
treating vitreous hemorrhage,258 a DRCR trial found no difference between anti-VEGF
and intravitreal saline injection.259 Following anti-VEGF injection, cases with severe
PDR may develop traction or pre-existing traction may progress.260 However, Protocol
S showed that there was no statistically significant difference between rates of tractional
retinal detachment in PRP compared with anti-VEGF.254

Several anti-VEGF studies have also found a significant difference in the rates of 2-step
and 3-step improvements in severity of diabetic retinopathy between eyes receiving anti-
VEGF and control eyes. The DRCR.net has shown that in the short-term, anti-VEGF
treatment lowers the risk of progression to PDR.261,262 In the DRCR.net Protocol T year
1, of the 423 NPDR eyes, 44 of 141 (31.2%) treated with aflibercept, 29 of 131 (22.1%)
with bevacizumab, and 57 of 151 (37.7%) with ranibizumab had improvement of
diabetic retinopathy severity. The adjusted difference for aflibercept versus
bevacizumab was 11.7% (95% CI, 2.9%–20.6%; P=0.004), for ranibizumab versus
bevacizumab was 8.9% (95% CI, 1.7%–16.1%; P=0.01), and for aflibercept versus
ranibizumab was 2.9% (95% CI, -5.7% to 11.4%; P=0.51). At year 2, despite fewer
injections of an anti-VEGF drug given to these eyes, 25% of the aflibercept group, 22%
of the bevacizumab group, and 21% of the ranibizumab group showed diabetic

P102
32
 Diabetic Retinopathy PPP

retinopathy improvement. Rates of worsening retinopathy were uniformly low for all
three drugs.

In the RIDE and RISE trials, approximately 11% of ranibizumab-treated eyes showed
progression of diabetic retinopathy compared with 34% of sham-treated eyes at 2
years.263 The percentage of eyes with worsening diabetic retinopathy by 2 steps or more
(Table 5) was significantly greater for the sham-treated eyes than the ranibizumab-
treated eyes. Post hoc analysis of RIDE and RISE trials revealed that ranibizumab
treatment improved diabetic retinopathy severity in all subsets. The greatest
improvement occurred in eyes with a baseline of moderately severe to severe NPDR.264

In the VIVID and VISTA trials, eyes treated with aflibercept (every 4 or 8 weeks) for
DME had a significantly higher chance of a 2-step (Table 5) improvement in the
Diabetic Retinopathy Severity Scale score compared with eyes treated with laser
control. (See Glossary). In the VIVID trial, the improvement was 29.3% and 32.6%,
respectively, versus 8.2% ( P<0.0004 for every 4 weeks and P<0.0001 for every 8
weeks), and in the VISTA trial, the improvement was 37.0% and 37.1%, respectively,
versus 15.6% (P<0.0001 for both aflibercept vs control comparisons). 265

Vitrectomy surgery typically is reserved for cases with persistent disease activity despite
medical management with anti-VEGF or PRP, or if disease is unamenable to medical
management alone. Typical indications for vitrectomy include:
• Nonclearing vitreous hemorrhage
• Tractional retinal detachment threatening the macula
• Combined rhegmatogenous and tractional retinal detachment
• Dense pre-macular subhyaloid hemorrhage
The DRVS demonstrated improved outcomes if vitrectomy for vitreous hemorrhage is
done within 1 to 6 months of onset compared with later vitrectomy at 1 year.266,267 Vitreous
hemorrhage should be followed with serial ultrasounds to evaluate for possible retinal tear,
tractional retinal detachment that threatens the macula, or rhegmatogenous retinal
detachment. Recent advances, including endolaser and small-gauge instruments have
improved outcomes and decreased adverse events. 268 One meta-analysis suggested that pre-
operative anti-VEGF treatment reduces the duration of surgery, the number of retinal
breaks, and the amount of intra-operative bleeding.269(I+, Moderate quality, Strong
recommendation) A Cochrane systematic review suggested pre-operative or intra-operative
bevacizumab may reduce the incidence of post-operative vitreous hemorrhage.270,271(I+,
Moderate quality, Strong recommendation)

P103
33
 Diabetic Retinopathy PPP

The follow-up evaluation includes a history and examination.

A follow-up history should include changes in the following:

◆ Symptoms
◆ Systemic status (pregnancy, blood pressure, serum lipids, renal status)
◆ Glycemic status (HbA1c)54,82,130
◆ Other treatments such as dialysis and fenofibrates

A follow-up examination should include the following elements:

◆ Visual acuity132
◆ Slit-lamp biomicroscopy with iris examination272
◆ IOP
◆ Gonioscopy (preferably before dilation when iris neovascularization is suspected or if IOP
is elevated)272
◆ Stereoscopic examination of the posterior pole after dilation of the pupils89
◆ OCT imaging, when appropriate
◆ Peripheral retina and vitreous examination, when indicated 88

Recommended intervals for follow-up are given in Table 5.

Although the ophthalmologist will perform most of the examination and all surgery, certain aspects of
examination may be performed by trained individuals under the ophthalmologist’s supervision and
review. Because of the complexities of the diagnosis and treatment for diabetic retinopathy, the
ophthalmologist caring for patients with this condition should be familiar with the specific
recommendations of relevant clinical trials.45,94,131,132,151,195,203,204,214,246,273-279

The ophthalmologist should refer patients with diabetes to a primary care physician for appropriate
management of their systemic condition and should communicate examination results to the physician
managing the patient’s ongoing diabetes care. An Eye MD Examination Report Form is available
from the American Academy of Ophthalmology. 280

Some patients with diabetic retinopathy will lose substantial vision despite being treated according to
the recommendations in this document.1 Patients whose conditions fail to respond to surgery and
those for whom further treatment is unavailable should be provided with professional support and
offered referral for counseling, vision rehabilitation, or social services as appropriate. 281 Vision

P104
34
 Diabetic Retinopathy PPP

rehabilitation improves functional ability,282 and so patients with functionally limiting postoperative
visual impairment should be referred for vision rehabilitation and social services. 281 More information
on vision rehabilitation, including materials for patients, is available at www.aao.org/low-vision-and-
vision-rehab.

In the era before anti-VEGF treatment, an analysis of medical and economic effects of diabetic
retinopathy control predicted that over their lifetime, 72% of patients with type 1 diabetes would
eventually develop PDR requiring PRP and that 42% would develop macular edema.283 If treatments
are delivered as recommended in the clinical trials, the model predicted a cost of $966 per person-year
of vision saved for patients with PDR and $1,120 per person-year of central visual acuity saved for
patients with macular edema. These costs are less than the cost of a year of Social Security disability
payments for patients disabled by vision loss. Therefore, treatment yields a substantial savings
compared with the direct cost to society of untreated PDR in a type 1 diabetic patient.284 The indirect
costs in lost productivity and human suffering are even greater.

Another analysis estimated that screening and treatment of eye disease in patients with diabetes costs,
on average, $3,190 per quality-adjusted life year (QALY) saved.285 For patients with type 1 diabetes,
it costs $1996 per QALY saved; for patients with type 2 diabetes who use insulin, it costs $2,933 per
QALY saved; and for patients with type 2 diabetes who do not use insulin, it costs $3,530 per QALY
saved. Insofar as patients with type 2 diabetes not using insulin represent the largest subset of the
patient population, most of the economic benefits of screening and treatment are realized among these
patients.

A 2013 cost-effectiveness analysis of various interventions for DME evaluated the cost-effectiveness
of anti-VEGF therapies for CSME. Compared with laser alone, the incremental cost-effectiveness of
laser plus bevacizumab is $11,138 per QALY and thus seems to confer the greatest value among the
various treatment options for CSME.286 By comparison, the cost-utility of laser photocoagulation
surgery for DME is $3,101 per QALY,287 whereas laser photocoagulation surgery for extrafoveal
choroidal neovascularization is $23,640 per QALY.288 174 Finally, a cost-utility analysis of detection
and treatment of diabetic retinopathy in patients with type 1 and type 2 diabetes demonstrates that
provision of recommended ophthalmic care would reduce the prevalence of blindness by 52% and
that the direct costs of care would be less than the losses in productivity and the costs of facilities
provided for disability.289

P105
35
 Diabetic Retinopathy PPP

Providing quality care

is the physician's foremost ethical obligation, and is
the basis of public trust in physicians.
AMA Board of Trustees, 1986

Quality ophthalmic care is provided in a manner and with the skill that is consistent with the best interests of
the patient. The discussion that follows characterizes the core elements of such care.
The ophthalmologist is first and foremost a physician. As such, the ophthalmologist demonstrates
compassion and concern for the individual, and utilizes the science and art of medicine to help alleviate
patient fear and suffering. The ophthalmologist strives to develop and maintain clinical skills at the highest
feasible level, consistent with the needs of patients, through training and continuing education. The
ophthalmologist evaluates those skills and medical knowledge in relation to the needs of the patient and
responds accordingly. The ophthalmologist also ensures that needy patients receive necessary care directly or
through referral to appropriate persons and facilities that will provide such care, and he or she supports
activities that promote health and prevent disease and disability.
The ophthalmologist recognizes that disease places patients in a disadvantaged, dependent state. The
ophthalmologist respects the dignity and integrity of his or her patients, and does not exploit their
vulnerability.
Quality ophthalmic care has the following optimal attributes, among others.
◆ The essence of quality care is a meaningful partnership relationship between patient and physician. The
ophthalmologist strives to communicate effectively with his or her patients, listening carefully to their
needs and concerns. In turn, the ophthalmologist educates his or her patients about the nature and
prognosis of their condition and about proper and appropriate therapeutic modalities. This is to ensure
their meaningful participation (appropriate to their unique physical, intellectual, and emotional state) in
decisions affecting their management and care, to improve their motivation and compliance with the
agreed plan of treatment, and to help alleviate their fears and concerns.
◆ The ophthalmologist uses his or her best judgment in choosing and timing appropriate diagnostic and
therapeutic modalities as well as the frequency of evaluation and follow-up, with due regard to the
urgency and nature of the patient's condition and unique needs and desires.
◆ The ophthalmologist carries out only those procedures for which he or she is adequately trained,
experienced, and competent, or, when necessary, is assisted by someone who is, depending on the
urgency of the problem and availability and accessibility of alternative providers.
◆ Patients are assured access to, and continuity of, needed and appropriate ophthalmic care, which can be
described as follows.
 The ophthalmologist treats patients with due regard to timeliness, appropriateness, and his or her own
ability to provide such care.
 The operating ophthalmologist makes adequate provision for appropriate pre- and postoperative
patient care.
 When the ophthalmologist is unavailable for his or her patient, he or she provides appropriate alternate
ophthalmic care, with adequate mechanisms for informing patients of the existence of such care and
procedures for obtaining it.
 The ophthalmologist refers patients to other ophthalmologists and eye care providers based on the
timeliness and appropriateness of such referral, the patient's needs, the competence and qualifications
of the person to whom the referral is made, and access and availability.

P106
36
 Diabetic Retinopathy PPP

 The ophthalmologist seeks appropriate consultation with due regard to the nature of the ocular or other
medical or surgical problem. Consultants are suggested for their skill, competence, and accessibility.
They receive as complete and accurate an accounting of the problem as necessary to provide efficient
and effective advice or intervention, and in turn they respond in an adequate and timely manner. The
ophthalmologist maintains complete and accurate medical records.
 On appropriate request, the ophthalmologist provides a full and accurate rendering of the patient's
records in his or her possession.
 The ophthalmologist reviews the results of consultations and laboratory tests in a timely and effective
manner and takes appropriate actions.
 The ophthalmologist and those who assist in providing care identify themselves and their profession.
 For patients whose conditions fail to respond to treatment and for whom further treatment is
unavailable, the ophthalmologist provides proper professional support, counseling, rehabilitative and
social services, and referral as appropriate and accessible.
◆ Prior to therapeutic or invasive diagnostic procedures, the ophthalmologist becomes appropriately
conversant with the patient's condition by collecting pertinent historical information and performing
relevant preoperative examinations. Additionally, he or she enables the patient to reach a fully informed
decision by providing an accurate and truthful explanation of the diagnosis; the nature, purpose, risks,
benefits, and probability of success of the proposed treatment and of alternative treatment; and the risks
and benefits of no treatment.
◆ The ophthalmologist adopts new technology (e.g., drugs, devices, surgical techniques) in judicious
fashion, appropriate to the cost and potential benefit relative to existing alternatives and to its
demonstrated safety and efficacy.
◆ The ophthalmologist enhances the quality of care he or she provides by periodically reviewing and
assessing his or her personal performance in relation to established standards, and by revising or altering
his or her practices and techniques appropriately.
◆ The ophthalmologist improves ophthalmic care by communicating to colleagues, through appropriate
professional channels, knowledge gained through clinical research and practice. This includes alerting
colleagues of instances of unusual or unexpected rates of complications and problems related to new
drugs, devices, or procedures.
◆ The ophthalmologist provides care in suitably staffed and equipped facilities adequate to deal with
potential ocular and systemic complications requiring immediate attention.
◆ The ophthalmologist also provides ophthalmic care in a manner that is cost effective without
unacceptably compromising accepted standards of quality.

Reviewed by: Council

Approved by: Board of Trustees
October 12, 1988
2nd Printing: January 1991
3rd Printing: August 2001
4th Printing: July 2005

P107
37
 Diabetic Retinopathy PPP

Diabetic retinopathy, which includes entities with the following ICD-9 and ICD-10 classifications (see
Glossary):

•
•
•
•
•

•
•
•
•
•

•
•
•
•
•

•
•
•
•
•

P108
38
 Diabetic Retinopathy PPP

•
•
•
•
•

•
•

•
•

•
•
•

P109
39
 Diabetic Retinopathy PPP

The Diabetic Retinopathy Study (DRS) was designed to investigate the value of laser
photocoagulation surgery for patients with severe nonproliferative diabetic retinopathy (NPDR) and
proliferative diabetic retinopathy (PDR).85 The results are shown in Table A4-1.

The Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) began in 1979. It was
initially funded by the National Eye Institute, which is part of the National Institutes of Health. The
purpose of the WESDR is to describe the frequency and incidence of complications associated with
diabetes (eye complications such as diabetic retinopathy and visual loss, kidney complications such as
diabetic nephropathy, and amputations), and to identify risk factors (such as poor glycemic control,
smoking, and high blood pressure) that may contribute to the development of these complications. 91

The Early Treatment Diabetic Retinopathy Study (ETDRS) investigated the value of photocoagulation
surgery for patients with NPDR or PDR without high-risk characteristics.89,132 The results for eyes
with macular edema are shown in Table A4-2. Visual loss was defined as at least doubling of the
visual angle (e.g., 20/20 to 20/40, or 20/50 to 20/100).

P110
40
 Diabetic Retinopathy PPP

In eyes with NPDR or non-high-risk PDR, early panretinal photocoagulation (PRP) was
compared with deferral of photocoagulation, and although there was a beneficial treatment
effect, the outlook for maintaining vision was good in both groups. The 5-year rates of severe
visual loss or vitrectomy ranged from 2% to 6% in eyes assigned to early photocoagulation and
from 4% to 10% in eyes assigned to deferral. Early PRP was associated with side effects (small
decreases in visual acuity and visual field) in some eyes, and the ETDRS concluded that
deferral of photocoagulation was preferable at least until retinopathy was approaching the high-
risk stage. Eyes approaching that stage had a 50% risk of reaching it within 12 to 18 months.
Eyes in this category had very severe NPDR or non-high-risk PDR characterized by NVD less
than one-quarter to one-third disc area and/or NVE, without vitreous or preretinal hemorrhage.

Recent additional analyses of visual outcome in ETDRS patients with severe NPDR to non-
high-risk PDR suggest that the recommendation to consider PRP before the development of
high-risk PDR is particularly appropriate for patients with type 2 diabetes.1 The risk of severe
vision loss or vitrectomy was reduced by 50% in patients who were treated early compared with
those who deferred treatment until high-risk PDR developed.

For patients with type 1 diabetes, the timing of the PRP will depend on the compliance with
follow-up, status and response to treatment of the fellow eye, impending cataract surgery,
and/or pregnancy status.

The Diabetic Retinopathy Vitrectomy Study (DRVS) investigated the role of vitrectomy in managing
eyes with very severe PDR.90,249-251 The benefit of early vitrectomy for severe vitreous hemorrhage
(defined as hemorrhage obscuring the macula or major retinal vessels for 3 disc diameters from the
macular center) was seen in type 1 patients, but no such advantage was found in type 2 patients, who
did not benefit from earlier surgery. Early vitrectomy was beneficial among patients with visual acuity
of 5/200 or worse and severe vitreous hemorrhage with reduced vision for at least 1 month and

P111
41
 Diabetic Retinopathy PPP

without previous treatment or complications such as retinal detachment or neovascularization of the

iris. Overall, at 2 years after surgery, 25% of the early vitrectomy group and 15% of the deferral group
had visual acuity of 20/40 or better. The advantage was most pronounced in patients with type 1
diabetes (36% vs. 12% for early vitrectomy versus deferral of vitrectomy, respectively) and was not
statistically significant for patients with type 2 diabetes.

The DRVS showed that early vitrectomy was beneficial for patients with visual acuity of 20/400 or
better plus 1 of the following: (1) severe neovascularization and fibrous proliferation; (2) fibrous
proliferation and moderate vitreous hemorrhage; or (3) moderate neovascularization, severe fibrous
proliferation, and moderate vitreous hemorrhage. Among such patients, 44% with early vitrectomy
and 28% in the observation group had visual acuity of 20/40 or better at 4 years of follow-up.

The results of the DRVS should be interpreted in light of subsequent advances in vitreoretinal
surgery, such as the introduction of small-gauge vitrectomy technology, endoscopic and indirect
ophthalmoscopic laser photocoagulation surgery, and advanced instrumentation. The use of long-
acting intraocular gases such as sulfur hexafluoride (SF6) and perfluoropropane (C3F8), the use of
viscodissection, and the use of heavier-than-water liquids such as perfluoro-octane are advances in
vitreoretinal surgery that developed after the DRVS. Thus, the results may actually be better than
those reported in the DRVS.239,290 Early vitrectomy should be considered for selected patients with
type 2 diabetes, particularly those in whom severe vitreous hemorrhage prohibits laser therapy
photocoagulation of active neovascularization.

The FIELD study was a randomized controlled trial that evaluated long-term fenofibrate therapy for
the reduction of cardiovascular events in 9795 patients with type 2 diabetes mellitus. Fenofibrate did
not significantly reduce the risk of the primary outcome of coronary events. It did reduce total
cardiovascular events, mainly due to fewer nonfatal myocardial infarctions and revascularizations.
The higher rate of starting statin therapy in patients allocated to receive placebo might have masked a
moderately larger treatment benefit.

The Diabetic Retinopathy Clinical Research Network (DRCR.net) is a collaborative network

dedicated to facilitating multicenter clinical research of diabetic retinopathy, diabetic macular edema
(DME), and associated conditions. The DRCR.net supports the identification, design, and
implementation of multicenter clinical research initiatives focused on diabetes-induced retinal
disorders. Principal emphasis is placed on clinical trials, but epidemiologic outcomes and other
research may be supported as well.

P112
42
 Diabetic Retinopathy PPP

The DRCR.net was formed in 2002 and currently includes over 115 participating sites (offices) with
over 400 physicians throughout the United States. The DRCR.net is funded by the National Eye
Institute (NEI), which is a part of the National Institutes of Health, the branch of government that
funds medical research.

The DRCR.net has completed multiple clinical trials evaluating the role of anti-vascular endothelial
growth factor (anti-VEGF), laser treatment, and corticosteroids in DME, anti-VEGF efficacy in PDR
and vitreous hemorrhage, and even diabetes education effectiveness on DME (See Table A3-3). Most
importantly, DRCR.net Protocol T (Comparative Effectiveness Study of Intravitreal Aflibercept,
Bevacizumab, and Ranibizumab for Diabetic Macular Edema) compared the effectiveness of
ranibizumab, aflibercept, and bevacizumab in the treatment of DME.96 This study found that all three
drugs resulted in improvement in visual acuity at 1 year with similar safety profiles. However, the
mean visual acuity using aflibercept was better for eyes with visual acuity of 20/50 or worse at 1 year.
At 2 years, the mean visual acuity in the aflibercept was no longer superior to ranibizumab, although
it remained superior to bevacizumab.

Another important treatment comparison was done in Protocol I: Intravitreal Ranibizumab for
Diabetic Macular Edema with Prompt vs. Deferred Laser Treatment. Three-year results were reported
in 2012. The study utilized ranibizumab monthly until improvement no longer occurred (with
resumption if the condition worsened) and random assignment to focal/grid laser treatment promptly
or deferred (≥24 weeks). The 3-year results suggest that focal/grid laser treatment at the initiation of
intravitreal ranibizumab is no better, and possibly worse for vision outcomes, than deferring laser
treatment for ≥24 weeks in eyes with DME involving the fovea and with vision impairment. 94

A previous publication from Protocol I results confirmed the 1-year results that intravitreal
ranibizumab with prompt or deferred laser was more effective through 2 years compared with prompt
laser alone for the treatment of DME involving the central macula. Laser was not associated with
endophthalmitis, the rare but potentially devastating complication of injecting ranibizumab. In
pseudophakic eyes, results with intravitreal triamcinolone plus prompt laser appeared similar to
results in the ranibizumab arms and were more effective than laser alone, but the triamcinolone plus
prompt laser arm had an increased risk of IOP elevation.189

Most recently, the DRCR.net Protocol S evaluated the effects of anti-VEGF versus PRP.291 In a
randomized, multicenter, noninferiority trial, 394 eyes of 305 adults with PDR were randomized to
receive either PRP or anti-VEGF therapy. Ranibizumab 0.5 mg was given at baseline and as
frequently as every 4 weeks based on a structured retreatment design. Eyes in both groups were
allowed ranibizumab if DME was present. In eyes with PDR, ranibizumab was not inferior to PRP in
terms of visual acuity outcomes at 2 years. Mean visual acuity improvement was +2.8 letters for
ranibizumab and +0.2 letters for PRP-treated eyes (P<0.001). When the totality of the visual acuity
data was included (area under the curve analysis), eyes given ranibizumab had overall better visual
acuity outcomes than eyes treated with PRP. There was less mean reduction in peripheral visual field

43
P113
 Diabetic Retinopathy PPP

(-23 dB vs. -422 dB; P<0.001) with ranibizumab than with PRP treatment. The rates for vitrectomy
were more frequent (15% vs. 4%; P<0.001), and DME development was more frequent (28% vs. 9%;
P<0.001) in the PRP group than in the ranibizumab group. Moreover, rates of active
neovascularization or rates of regression of neovascularization were similar between the two groups.

P114
44
 Diabetic Retinopathy PPP

P115
45
 Diabetic Retinopathy PPP

P116
46
 Diabetic Retinopathy PPP

The RISE and RIDE trials were parallel phase III multicenter double-masked sham injection–
controlled randomized studies conducted at private and university-based retina specialty clinics in the
United States and South America. (See Glossary.)

The phase III results for both studies were published in 2012. The studies utilized monthly intravitreal
ranibizumab (0.5 or 0.3 mg) or sham injections, with macular laser available if needed. The study
concluded that ranibizumab rapidly and sustainably improved vision, reduced the risk of further
vision loss, and improved macular edema in patients with DME, with low rates of ocular and
nonocular side effects.187

READ-2 was a phase II multicenter randomized controlled trial that compared 0.5 mg injections of
ranibizumab versus focal laser treatment over 2 years in patients with type 1 or type 2 diabetes
mellitus and DME. Patients randomized to one arm of the trial received ranibizumab at baseline, and
at 1, 3 and 5 months after baseline; a second arm received laser treatment at baseline and at 3 months
(if needed); the third arm received both ranibizumab and laser treatment at baseline and 3 months.
From month 5, all subjects received ranibizumab every 2 months and/or maintenance laser treatment
every 3 months.

At 24 months, differences between the groups were not statistically significant, and all groups
experienced improved visual acuity. Patients receiving combined ranibizumab and laser treatment
required fewer injections than patients receiving ranibizumab alone. 203

BOLT was a phase II 2-year randomized controlled trial that compared intravitreal 1.25 mg
bevacizumab injections and focal laser treatment in patients with persistent DME and visual
impairment. Bevacizumab patients received an injection every 6 weeks, whereas laser patients were
treated every 4 weeks.

At 2 years, visual acuity results were substantially better in the bevacizumab group compared with the
laser group, with significant differences in the proportions of patients gaining 10 letters and 15 letters.
No patients lost 10 or more letters in the bevacizumab group, compared with 14% of patients treated
with laser.204

P117
47
 Diabetic Retinopathy PPP

These studies compared the efficacy and safety of intravitreal aflibercept injection (IAI) with macular
laser photocoagulation surgery for DME. Visual improvement were observed in the IAI treatment
regimens over laser control at 52, 100 and 148 weeks. Incidence of adverse events was consistent with
the known safety profile of IAI.292

The DRCR.net compared the efficacy and safety of bevacizumab, ranibizumab, and aflibercept in a
multicentered, randomized clinical trial.152 At the primary endpoint at 1 year, the mean change in
vision was greater for aflibercept than for either of the other two drugs. However, the mean visual
acuity changes were dependent on the baseline visual acuity. For eyes with milder visual acuity loss,
the drugs resulted in similar visual outcomes (8.0 with aflibercept, 7.5 with bevacizumab, and 8.3
with ranibizumab; P>0.50 for each pairwise comparison). However, for eyes with 20/50 or worse
vision, the mean visual acuity in eyes treated with aflibercept had greater improvements in vision
(18.9 with aflibercept, 11.8 with bevacizumab, and 14.2 with ranibizumab; P<0.001 for aflibercept vs.
bevacizumab, P=0.003 for aflibercept vs. ranibizumab, and P=0.21 for ranibizumab vs.
bevacizumab). There were no significant differences in rates of adverse events. However, at 2 years,
the mean visual acuity results were similar for ranibizumab and aflibercept, although aflibercept
results remained significantly better than bevacizumab results. There was a slightly higher rate of
Antiplatelet Trialists' Collaboration (APTC) events with ranibizumab compared with the other two
drugs at the 2-year endpoint. All three drugs improved visual acuity at 2 years, and the number of
injections decreased in year 2 compared to year 1.

P118
48
 Diabetic Retinopathy PPP

The Diabetes Control and Complications Trial (DCCT) was a multicenter, randomized controlled trial
designed to study the connection between glycemic control and retinal, renal, and neurologic complications
of type 1 diabetes mellitus. Published results from this trial demonstrated that improved blood sugar control
can delay the onset and slow the progression of diabetic retinopathy, nephropathy, and neuropathy in type 1
patients.81 The DCCT showed a strong exponential relationship between the risk of diabetic retinopathy and
the mean HbA1c level. For each 10% decrease in the HbA1c (e.g., from 9% to 8.1%), there was a 39%
decrease in the risk of progression of retinopathy over the range of HbA1c values. There was no glycemic
threshold when the risk of retinopathy was eliminated above the nondiabetic range of HbA1c (4% to 6.05%).

After 6.5 years of follow-up, the DCCT ended, and all patients were encouraged to pursue strict control of
blood sugar. Most of these patients are being followed in the Epidemiology of Diabetes Interventions and
Complications (EDIC) study, which includes 95% of the DCCT subjects. A total of 1294 to 1335 patients
have been examined annually in the EDIC study. Further progression of diabetic retinopathy during the first
4 years of the EDIC study was 66% to 77% less in the former intensive treatment group than in the former
conventional treatment group.43 The benefit persisted even at 7 years. This benefit included an effect on
severe diabetic retinopathy, including severe nonproliferative diabetic retinopathy (NPDR), proliferative
diabetic retinopathy (PDR), clinically significant macular edema, and the need for focal/grid or panretinal
laser photocoagulation surgery.45 The decrease in HbA1c from 9% to approximately 8% did not drastically
reduce the progression of diabetic retinopathy in the former conventional treatment group, nor did the
increase in HbA1c from approximately 7% to approximately 8% drastically accelerate diabetic retinopathy in
the former intensive treatment group.43 Thus, it takes time for improvements in control to negate the long-
lasting effects of prior prolonged hyperglycemia, and once the biological effects of prolonged improved
control are manifest, the benefits are long-lasting. Furthermore, the total glycemic exposure of the patient
(i.e., degree and duration) determines the degree of retinopathy observed at any one time.

A positive relationship between the 4-year incidence and progression of retinopathy and glycosylated
hemoglobin remains after controlling for other risk factors, such as duration of diabetes and severity of
retinopathy at a baseline examination.65,66,130 Extrapolation of pathologic and clinical experience strongly
suggests that poor levels of control contribute to microangiopathy, including retinopathy. 293 The development
of PDR parallels an increased risk of nephropathy, myocardial infarction, and/or cerebral vascular accidents.

Although good glycemic control is advised, there is some evidence that rapid improvement of long-standing
poor control may increase the risk of retinopathy progression over the first year for some patients. About
10% of type 1 patients who had initial retinopathy at the beginning of the DCCT had increased retinopathy
progression.294 Specifically, there may be a transient increase in the number of cotton wool spots seen on
retinal examination. Frequent ophthalmologic monitoring is important when diabetic patients are being
brought under better metabolic control.294

P119
49
 Diabetic Retinopathy PPP

In the DCCT there was a threefold increase in severe hypoglycemic events and excess weight gain among
patients using intensive treatment regimens. Increased risk of hypoglycemia is a consequence of strict blood
glucose control. Irregular food intake, failure to check blood glucose before planned or unplanned vigorous
exercise or before operating a motor vehicle, and excess alcohol are risk factors for hypoglycemia. Diabetes
mellitus education and regular reinforcement should be provided by diabetes nurses and dietitian educators
and may help minimize the risk of hypoglycemia.

The United Kingdom Prospective Diabetes Study (UKPDS), 46,126 a randomized controlled clinical trial of
blood glucose control, enrolled 3867 patients with newly diagnosed type 2 diabetes. Intensive blood glucose
control by either the sulfonylureas or insulin decreased the risk of microvascular complications but not the
risk of macrovascular disease. There were no adverse effects of the individual drugs on the cardiovascular
outcome. In this study, there was a 29% reduction in the need for retinal photocoagulation in the group that
had intensive glucose therapy compared with those that had conventional treatment (relative risk, 0.71; 95%
confidence interval, 0.53–0.96; P=0.003).

The ACCORD (Action to Control Cardiovascular Risk in Diabetes) study (www.accordtrial.org) was a large
clinical trial of adults with established type 2 diabetes who are at especially high risk of cardiovascular
disease (CVD). Type 2 diabetes increases the risk of a number of complications, especially CVD, which is
the leading cause of early death in people with diabetes.

The ACCORD study consisted primarily of three clinical trials that tested treatment approaches to determine
the best ways to decrease the high rate of major CVD events—heart attack, stroke, or death from CVD—
among people with type 2 diabetes who are at especially high risk of having such a CVD event. These three
treatment approaches were intensive lowering of blood sugar levels compared with a more standard blood
sugar treatment; intensive lowering of blood pressure compared with standard blood pressure treatment; and
treatment of multiple blood lipids with two drugs—a fibrate plus a statin—compared with one drug, a statin
alone.295

The study began enrolling participants in 2001 and took place in 77 clinical sites across the United States and
Canada. A total of 10,251 adults with established type 2 diabetes participated in ACCORD. At enrollment,
study participants were between age 40 and 79 (average age 62), had diabetes for an average of 10 years, and
were at especially high risk for CVD events because they already had pre-existing CVD, evidence of
subclinical CVD, or at least two CVD risk factors in addition to type 2 diabetes. The other CVD risk factors
could be high low-density lipoprotein cholesterol, high blood pressure, smoking, or obesity.

The primary outcome measure for all three trials was the first occurrence after randomization of a major
CVD event, specifically nonfatal heart attack, nonfatal stroke, or CVD death. Secondary outcomes include
total mortality (death), microvascular outcomes (e.g., eye, kidney, and nerve complications), health-related
quality of life, and cost-effectiveness.

All three ACCORD clinical trials have ended. The National Heart, Lung, and Blood Institute (NHLBI)
stopped the intensive blood sugar lowering strategy in 2008 due to safety concerns. Participants in the
intensive blood sugar treatment strategy group were transitioned to the standard treatment strategy. The blood

P120
50
 Diabetic Retinopathy PPP

pressure and lipid treatment trials continued until the planned end of the study in 2009. In its regular review
of the available study data, the ACCORD Data and Safety Monitoring Board (DSMB) noticed an unexpected
increase in total deaths from any cause among participants who had been randomly (by chance) assigned to
the intensive lowering of blood sugar levels group compared with those assigned to the standard blood sugar
treatment group. The data analyses showed that over an average of 3.5 years of treatment (ranging from
about 2 years to about 7 years), 257 participants in the intensive group died compared with 203 in the
standard group—a difference of 54 deaths, or an excess of about 3 deaths per 1,000 participants treated for a
year. This translates to a statistically significant 22% higher rate of death in the intensive group than in the
standard group.

There was a trend toward lower (10% lower) rate of primary outcome events, primarily nonfatal heart
attacks, in the intensive group compared with the standard treatment group. However, the DSMB
recommended discontinuing intensive blood sugar treatment because the harm of the intensive strategy
outweighed the potential benefit. The NHLBI accepted the DSMB’s recommendation and decided to
transition all participants to the standard blood sugar strategy.

The results of the blood sugar trial were published in 2008. 296 There was no significant difference in the
primary study outcome between the intensive and standard blood pressure treatment groups. The primary
outcome was the time to first occurrence after randomization of a heart attack, a stroke, or a cardiovascular
death. Thus, the primary hypothesis of the ACCORD BP trial was not supported. There was, however, a
significant reduction in the rate of strokes, although the numbers were relatively small. This reduction in
stroke was consistent with previous blood pressure lowering trials. Overall, however, the findings from the
ACCORD blood pressure trial suggest that, on average, the standard treatment for lowering blood pressure
was just as good as the intensive lowering treatment for cardiovascular outcomes.

The results of the lipid297 and the blood pressure298 trials were published in 2010. Overall, the fibrate and the
placebo groups did not differ in the rates of the combined outcome of heart attacks, strokes, or cardiovascular
death. The results, however, suggest that men may benefit from this treatment, but there was a trend toward
more cardiovascular problems in women receiving the combination therapy compared with those who
received statins only. Also, the group of patients who at the start of the trial had the lowest level of high-
density lipoprotein (HDL) cholesterol combined with the highest level of triglycerides (which represented
only 17% of the ACCORD participants) may have benefitted from this combined drug treatment.

More recently, the American College of Physicians published their glycemic control guidance statement to
guide clinicians in selecting targets for pharmacologic treatment of type 2 diabetes based on the AGREE II
(Appraisal of Guidelines for Research and Evaluation II) instrument, which was used to evaluate the
guidelines.299 The National Guideline Clearinghouse and the Guidelines International Network library were
searched (May 2017) for national guidelines published in English that addressed HbA1c targets for treating
type 2 diabetes in nonpregnant outpatient adults. The investigators also identified guidelines from the
National Institute for Health and Care Excellence and the Institute for Clinical Systems Improvement. In
addition, four commonly used guidelines were reviewed from the American Association of Clinical
Endocrinologists and the American College of Endocrinology, the American Diabetes Association, the

P121
51
 Diabetic Retinopathy PPP

Scottish Intercollegiate Guidelines Network, and the US Department of Veterans Affairs and Department of
Defense. They found that the ideal target that optimally balances benefits and harms remains uncertain. Their
four guidance statements emphasize the importance of personalizing the glycemic goals in patients with type
2 diabetes on the basis of the benefits/harms balance of pharmacotherapy, patient preference, and life
expectancy. They suggest an HbA1c goal range of 7% to 8% for most patients. These authors also recognized
the studies that showed that more intensive glycemic control likely requires a long time to manifest. Thus,
more stringent targets may be appropriate for patients who have a long life expectancy (>15 years). Further,
most of the guidelines noted that a target in the lower end of the range (7%) applied best to patients with
newly diagnosed diabetes and those without substantial diabetes-related complications.

P122
52
 Diabetic Retinopathy PPP

The Early Treatment of Diabetic Retinopathy Study (ETDRS) classification of diabetic retinopathy and
definitions of macular edema are in Tables A6-1.

P123
53
 Diabetic Retinopathy PPP

Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial: A large multicenter clinical trial that
evaluated intensive control of blood sugar, intensive control of blood pressure, and statin therapy (with or
without fibrate treatment) for the prevention of cardiovascular disease events among high-risk patients with
type 2 diabetes.

ACCORD: See Action to Control Cardiovascular Risk in Diabetes trial.

Anti-VEGF: See Anti-vascular endothelial growth factor.

Anti-vascular endothelial growth factor (anti-VEGF): Substances that inhibit the action of vascular
endothelial growth factor protein.

Bevacizumab or Laser Treatment (BOLT) study: A randomized trial that evaluated intravitreal bevacizumab
or conventional laser treatment for center-involved DME

BOLT: See Bevacizumab or Laser Treatment study.

Clinically significant macular edema (CSME): Retinal thickening at or within 500 µm of the center of the
macula; and/or hard exudates at or within 500 µm of the center of the macula, if associated with thickening
of the adjacent retina; and/or a zone or zones of retinal thickening 1 disc area in size, any part of which is
within 1 disc diameter of the center of the macula.

CSME: See Clinically significant macular edema.

ci-CSME: Center-involved CSME.

DA VINCI: See DME and VEGF Trap-Eye: Investigation of Clinical Impact study.

DCCT: See Diabetes Control and Complications Trial.

Diabetes Control and Complications Trial (DCCT): A multicenter, randomized, controlled trial designed to
study the connection between glycemic control and retinal, renal, and neurologic complications of type 1
diabetes mellitus. (See Appendix 5.)

Diabetes mellitus: According to the American Diabetes Association Expert Committee on the Diagnosis and
Classification of Diabetes Mellitus, the criteria for the diagnosis of diabetes mellitus are as follows.

◆ Fasting plasma glucose equal to or exceeding 126 mg/dL (7.0 mmol/L). Fasting is defined as no
caloric intake for at least 8 hours.
or
◆ Symptoms of hyperglycemia and a casual plasma glucose concentration equal to or exceeding 200
mg/dL (11.1 mmol/L). “Casual” is defined as any time of day without regard to time since last meal.
The classic symptoms of hyperglycemia include polyuria, polydipsia, and unexplained weight loss.
or
◆ A plasma glucose measurement at 2 hours postload equal to or exceeding 200 mg/dL (11.1 mmol/L)
during an oral glucose tolerance test. The test should be performed as described by the World Health
Organization, using a glucose load containing the equivalent of 75 g anhydrous glucose dissolved in
water. However, the expert committee has recommended against oral glucose tolerance testing for
routine clinical use. (Source: Report of the Expert Committee on the Diagnosis and Classification of
Diabetes Mellitus. Diabetes Care 2008;31 (suppl):55-60.)

Diabetic macular edema: The accumulation of fluid in the macula due to leaky blood vessels.

Diabetic Retinopathy Clinical Research Network (DRCR.net): A multicenter trial that is evaluating
different treatment modalities for diabetic retinopathy.

P124
 Diabetic Retinopathy PPP

Diabetic Retinopathy Study (DRS): A study designed to investigate the value of xenon arc and argon
photocoagulation surgery for patients with severe NPDR and PDR. (See Appendix 4.)

Diabetic Retinopathy Vitrectomy Study (DRVS): A study that investigated the role of vitrectomy in
managing eyes with very severe PDR. (See Appendix 4.)

DME: See Diabetic macular edema.

DME and VEGF Trap-Eye: Investigation of Clinical Impact (DA VINCI) study: A randomized trial of the
use of aflibercept for DME.

DRCR.net: See Diabetic Retinopathy Clinical Research Network.

DRS: See Diabetic Retinopathy Study.

DRVS: See Diabetic Retinopathy Vitrectomy Study.

Early Treatment Diabetic Retinopathy Study (ETDRS): A study that investigated the value of
photocoagulation surgery for patients with NPDR or PDR who did not have high-risk characteristics. (See
Appendix 4.)

Early proliferative diabetic retinopathy (i.e., proliferative retinopathy without DRS high-risk
characteristics): New vessels that do not meet the criteria of high-risk proliferative retinopathy.

EDIC: See Epidemiology of Diabetes Interventions and Complications study.

Epidemiology of Diabetes Interventions and Complications (EDIC) study: An observational study

following 95% of the DCCT subjects. (See Appendix 5.)

ETDRS: See Early Treatment Diabetic Retinopathy Study.

Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study: A large randomized controlled
type 2 diabetes mellitus.

FIELD study: See Fenofibrate Intervention and Event Lowering in Diabetes study.

Focal photocoagulation: A laser technique directed to abnormal blood vessels with specific areas of focal
leakage (i.e., microaneurysms) to reduce chronic fluid leakage in patients with macular edema.

Grid photocoagulation: A laser technique in which a grid pattern of scatter burns is applied in areas of
diffuse macular edema and nonperfusion. Typically, fluorescein angiograms of these areas show a diffuse
pattern rather than focal leakage.

High-risk proliferative diabetic retinopathy (PDR): New vessels on or within 1 disc diameter of the optic
disc equaling or exceeding standard photograph 10A (about one-quarter to one-third disc area), with or
without vitreous or preretinal hemorrhage; or vitreous and/or preretinal hemorrhage accompanied by new
vessels either on the optic disc less than standard photograph 10A or new vessels elsewhere equaling or
exceeding one-quarter disc area.

P125
55
 Diabetic Retinopathy PPP

Standard photograph 10A defines the lower border of moderate NVD. NVD
covers approximately one-third the area of the standard disc. This extent of
NVD alone would constitute PDR with high-risk characteristics.

Reprinted with permission from the Early Treatment Diabetic Retinopathy Study Research Group. Grading diabetic retinopathy from
stereoscopic color fundus photographs--an extension of the modified Airlie House classification: ETDRS report number 10.
Ophthalmology 1991;98:786-806.
ICD-9: International Statistical Classification of Diseases and Related Health Problems, Ninth Edition.

ICD-10: International Statistical Classification of Diseases and Related Health Problems, Tenth Edition.

Intraretinal microvascular abnormalities (IRMA): Tortuous intraretinal vascular segments, varying in

caliber from barely visible to 31 µm in diameter (one-quarter the width of a major vein at the disc margin);
they occasionally can be larger. Intraretinal microvascular abnormalities may be difficult to distinguish from
neovascularization.

IRMA: See Intraretinal microvascular abnormalities.

Macular edema: Thickening of the retina within 1 or 2 disc diameters of the center of the macula. (See
Clinically significant macular edema.) Any other thickening of the macula not within this area is non-CSME.

Mild nonproliferative diabetic retinopathy (NPDR): At least 1 microaneurysm and less than moderate
nonproliferative diabetic retinopathy.

Moderate nonproliferative diabetic retinopathy (NPDR): Hemorrhages and/or microaneurysms greater than
standard photograph 2A, and/or soft exudates, venous beading, or IRMA present but less than severe
nonproliferative retinopathy.

Moderate visual loss: The loss of 15 or more letters on the ETDRS visual acuity chart, or doubling of the
visual angle (e.g., 20/20 to 20/40, or 20/50 to 20/100).

nci-CSME: Non-center-involved CSME.

New vessels at the optic disc (NVD): New vessels at the optic disc; neovascularization on or within 1 disc
diameter of the optic disc.

New vessels elsewhere in the retina: New vessels elsewhere in the retina; neovascularization elsewhere in
the retina and greater than 1 disc diameter from the optic disc margin.

New vessels on the iris: New vessels on the iris; neovascularization of the iris.

Nonproliferative diabetic retinopathy (NPDR): The phases of diabetic retinopathy with no evidence of
retinal neovascularization.

NPDR: See Nonproliferative diabetic retinopathy.

NVD: See New vessels at the optic disc.

OCT: See Optical coherence tomography.

56
P126
 Diabetic Retinopathy PPP

Optical coherence tomography (OCT): A diagnostic test using low energy lasers that takes a cross-section
image of the retina, Used mostly to determine if there are membranes on the surface of the macula or fluid
within or beneath it.

Panretinal photocoagulation: A type of laser surgery used for patients with PDR. The surgery is delivered in
a scatter pattern throughout the peripheral fundus and is intended to lead to a regression of
neovascularization.

PDR: See Proliferative diabetic retinopathy.

Proliferative diabetic retinopathy (PDR): Advanced disease characterized by NVD and/or new vessels
elsewhere in the retina.

Quality-adjusted life year (QALY): A measure of health outcome that assigns to each year of a patient’s life
a weight (ranging from 0 to 1) corresponding to the health-related quality of life during that year, such that a
value of 1 indicates a year of optimal health and a value of 0 indicates a year in a health state judged
equivalent to death.

QALY: See Quality adjusted life year.

Ranibizumab for Edema of the mAcula in Diabetes (READ-2) study: A prospective multicenter randomized
controlled trial that compared 0.5 mg ranibizumab and laser photocoagulation surgery for the treatment of
DME.

READ-2: See Ranibizumab for Edema of the mAcula in Diabetes study.

Retinal hard exudate: Protein and lipid accumulation within the retina.

RIDE: A study of ranibizumab injection in subjects with CSME with center-involvement secondary to
diabetes mellitus.

RISE: A study of ranibizumab injection in subjects with clinically significant macular edema with center-
involvement secondary to diabetes mellitus.

Scatter photocoagulation: See Panretinal photocoagulation.

Severe nonproliferative diabetic retinopathy (NPDR): Using the 4-2-1 rule, the presence of at least one of
the following features: (1) severe intraretinal hemorrhages and microaneurysms, equaling or exceeding
standard photograph 2A, present in 4 quadrants; (2) venous beading in 2 or more quadrants (standard
photograph 6A); or (3) moderate IRMA equaling or exceeding standard photograph 8A in 1 or more
quadrants.

Standard photograph 2A, the standard for hemorrhages/microaneurysms.

Eyes with severe NPDR have this degree of severity of hemorrhages and
microaneurysms in all 4 midperipheral quadrants.

57
P127
 Diabetic Retinopathy PPP

Standard photograph 6A, less severe of two standards for venous beading.
Two main branches of the superior temporal vein show beading that is definite
but not severe.

Standard photograph 8A, the standard for moderate IRMA. Patients with
severe NPDR have moderate IRMA of at least this severity in at least 1
quadrant.

Reprinted with permission from the Early Treatment Diabetic Retinopathy Study Research Group. Grading diabetic retinopathy from
stereoscopic color fundus photographs--an extension of the modified Airlie House classification: ETDRS report number 10.
Ophthalmology 1991;98:786-806.

Severe visual loss: Occurrence of visual acuity worse than 5/200 at any two consecutive visits scheduled at
4-month intervals.

UKPDS: See United Kingdom Prospective Diabetes Study.

United Kingdom Prospective Diabetes Study (UKPDS): A randomized controlled clinical trial of blood
glucose control in patients with newly diagnosed type 2 diabetes. (See Appendix 5.)

VIVID: A randomized, double masked, active controlled, Phase III study of the efficacy and safety of
repeated doses of intravitreal VEGF Trap-Eye in subjects with DME.

VISTA: A randomized, double masked, active controlled, Phase III study of the efficacy and safety of
intravitreal administration of VEGF Trap-Eye in patients with DME.

WESDR: See Wisconsin Epidemiologic Study of Diabetic Retinopathy

Wisconsin Epidemiologic Study of Diabetic Retinopathy: A large epidemiologic study of complications

associated with diabetes and of risk factors associated with those complications

58
P128
 Diabetic Retinopathy PPP

Literature searches of the PubMed and Cochrane databases were conducted in April 2018; the search
strategies are provided at www.aao.org/ppp. Specific limited update searches were conducted after June 2019.
("Diabetic Retinopathy/epidemiology"[Mesh] OR "Diabetic
Retinopathy/ethnology"[Mesh])

("Diabetic Retinopathy"[Mesh]) AND ("Risk Factors"[Mesh])

"Diabetic Retinopathy"[Mesh] AND "natural history"[tiab]

"Diabetic Retinopathy/diagnosis"[Mesh]

"Diabetic Retinopathy/therapy"[Mesh]

"Diabetic Retinopathy"[Mesh] AND ((("Drug Therapy, Combination"[Mesh] OR "Drug

Combinations"[Mesh]) OR "Combined Modality Therapy"[Mesh]) OR (combination[tiab]
OR combined[tiab]))

"Diabetic Retinopathy"[Mesh] AND "Cost of Illness"[Mesh]

(("Diabetic Retinopathy"[Mesh] OR ("diabetic"[All Fields] AND "retinopathy") OR

"diabetic retinopathy") AND "Cost-Benefit Analysis"[Mesh])) OR ("Diabetic
Retinopathy/economics"[Mesh]

("Diabetic Retinopathy/therapy"[Mesh] AND ("Quality of Life"[Mesh]

"Diabetic Retinopathy"[Mesh] AND (("Quality of Life"[Mesh] NOT

("therapy"[Subheading] OR "therapy"[All Fields] OR "treatment" OR
"therapeutics"[MeSH Terms] OR "therapeutics"))

"Diabetic Retinopathy/genetics"[Mesh]

"Diabetic Retinopathy"[Mesh] AND (Guideline[ptyp]

P129
 Diabetic Retinopathy PPP

Basic and Clinical Science Course

Retina and Vitreous (Section 12, 2019–2020)

Clinical Statements –
Free download available at http://one.aao.org/guidelines-browse?filter=clinicalstatement.
Frequency of Ocular Examinations (2015)
International Clinical Classification System for Diabetic Retinopathy and Diabetic Macular Edema (2012)
Screening for Diabetic Retinopathy (2014)
Telemedicine for Ophthalmology Information Statement (2018)
Verifying the Source of Compounded Bevacizumab for Intravitreal Injections (2012)

Focal Points
Retinal Optical Coherence Tomography (2014)
Update on the Management of Diabetic Retinopathy (2011)

Ophthalmic Technology Assessment –

Published in Ophthalmology, which is distributed free to Academy members; links to full text available
at www.aao.org/ota.
Anti-VEGF Pharmacotherapy for Diabetic Macular Edema (2012)
Clinical Models and Algorithms for the Prediction of Retinopathy of Prematurity (2016)
Current Role of Cryotherapy in Retinopathy of Prematurity (2012)
Laser Scanning Imaging for Macular Disease (2007; reviewed for currency 2012)
Single Field Fundus Photography for Diabetic Retinopathy Screening (2004; reviewed for currency 2010)

Patient Education
Diabetic Retinopathy Brochure (2014)
Diabetic Retinopathy Brochure (Spanish: Retinopatía Diabetíca) (2014)
EyeSmart® What is Diabetic Retinopathy? Available at:
www.geteyesmart.org/eyesmart/diseases/diabetic-retinopathy/index.cfm

Preferred Practice Pattern® Guidelines – Free download available at www.aao.org/ppp.

Comprehensive Adult Medical Evaluation (2015)

To order any of these products, except for the free materials, please contact the Academy’s Customer Service
at 866.561.8558 (U.S. only) or 415.561.8540 or www.aao.org/store

P130
 Diabetic Retinopathy PPP

1. Ferris F. Early photocoagulation in patients with either type I or type II diabetes. Trans Am Ophthalmol
Soc. 1996;94:505-537.
2. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence
and strength of recommendations. BMJ. 2008;336(7650):924-926.
3. GRADE Working Group. Organizations that have endorsed or that are using GRADE.
http://www.gradeworkinggroup.org/. Accessed September 2019.
4. Shah AR, Gardner TW. Diabetic retinopathy: research to clinical practice. Clin Diabetes Endocrinol.
2017;3:9.
5. Abcouwer SF, Gardner TW. Diabetic retinopathy: loss of neuroretinal adaptation to the diabetic
metabolic environment. Ann N Y Acad Sci. 2014;1311:174-190.
6. Kawasaki R, Tanaka S, Abe S, et al. Japan Diabetes Complications Study Group. Risk of
cardiovascular diseases is increased even with mild diabetic retinopathy: the Japan Diabetes
Complications Study. Ophthalmology. 2013;120(3):574-582.
7. American Association of Clinical Endocrinologists. State of diabetes complications in America: a
comprehensive report issued by the American Association of Clinical Endocrinologists. Available at:
http://multivu.prnewswire.com/mnr/AACE/2007/docs/Diabetes_Complications_Report.pdf. Accessed
September 2019.
8. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care.
2010;33 Suppl 1:S62-69.
9. Klein R, Klein BE, Moss SE. Visual impairment in diabetes. Ophthalmology. 1984;91(1):1-9.
10. Eppens MC, Craig ME, Cusumano J, et al. Prevalence of diabetes complications in adolescents with
type 2 compared with type 1 diabetes. Diabetes Care. 2006;29(6):1300-1306.
11. Centers for Disease Control and Prevention. National diabetes statistics report, 2017. Available at:
https://www.cdc.gov/diabetes/pdfs/data/statistics/national-diabetes-statistics-report.pdf. Accessed
September 2019.
12. Cowie CC, Rust KF, Byrd-Holt DD, et al. Prevalence of diabetes and impaired fasting glucose in adults
in the U.S. population: National Health And Nutrition Examination Survey 1999-2002. Diabetes Care.
2006;29(6):1263-1268.
13. Diagnosing Diabetes and Learning about Prediabetes. 2014; http://www.diabetes.org/diabetes-
basics/diagnosis/. Accessed September 2019.
14. Acton KJ, Burrows NR, Moore K, Querec L, Geiss LS, Engelgau MM. Trends in diabetes prevalence
among American Indian and Alaska native children, adolescents, and young adults. Am J Public
Health. 2002;92(9):1485-1490.
15. Centers for Disease Control and Prevention. Prevalence of diagnosed diabetes among American
Indians/Alaskan Natives--United States, 1996. MMWR Morb Mortal Wkly Rep. 1998;47(42):901-904.
16. Liu L, Wu X, Geng J, Yuan Z, Shan Z, Chen L. Prevalence of diabetic retinopathy in mainland China:
a meta-analysis. PLoS One. 2012;7(9):e45264.
17. Namperumalsamy P, Kim R, Vignesh TP, et al. Prevalence and risk factors for diabetic retinopathy: a
population-based assessment from Theni District, south India. Br J Ophthalmol. 2009;93(4):429-434.
18. Narayan KM, Boyle JP, Thompson TJ, Sorensen SW, Williamson DF. Lifetime risk for diabetes
mellitus in the United States. JAMA. 2003;290(14):1884-1890.
19. Danaei G, Finucane MM, Lu Y, et al. National, regional, and global trends in fasting plasma glucose
and diabetes prevalence since 1980: systematic analysis of health examination surveys and
epidemiological studies with 370 country-years and 2.7 million participants. Lancet.
2011;378(9785):31-40.
20. Pinhas-Hamiel O, Zeitler P. The global spread of type 2 diabetes mellitus in children and adolescents. J
Pediatr. 2005;146(5):693-700.
21. Urakami T, Kubota S, Nitadori Y, Harada K, Owada M, Kitagawa T. Annual incidence and clinical
characteristics of type 2 diabetes in children as detected by urine glucose screening in the Tokyo
metropolitan area. Diabetes Care. 2005;28(8):1876-1881.

P131
 Diabetic Retinopathy PPP

22. Wei JN, Sung FC, Lin CC, Lin RS, Chiang CC, Chuang LM. National surveillance for type 2 diabetes
mellitus in Taiwanese children. JAMA. 2003;290(10):1345-1350.
23. Fagot-Campagna A, Pettitt DJ, Engelgau MM, et al. Type 2 diabetes among North American children
and adolescents: an epidemiologic review and a public health perspective. J Pediatr. 2000;136(5):664-
672.
24. McMahon SK, Haynes A, Ratnam N, et al. Increase in type 2 diabetes in children and adolescents in
Western Australia. Med J Aust. 2004;180(9):459-461.
25. Kaufman FR. Type 2 diabetes mellitus in children and youth: a new epidemic. J Pediatr Endocrinol
Metab. 2002;15 Suppl 2:737-744.
26. Harris MI, Flegal KM, Cowie CC, et al. Prevalence of diabetes, impaired fasting glucose, and impaired
glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988-
1994. Diabetes Care. 1998;21(4):518-524.
27. Harris MI, Klein R, Cowie CC, Rowland M, Byrd-Holt DD. Is the risk of diabetic retinopathy greater
in non-Hispanic blacks and Mexican Americans than in non-Hispanic whites with type 2 diabetes? A
U.S. population study. Diabetes Care. 1998;21(8):1230-1235.
28. Geiss LS, Cowie CC. Type 2 diabetes and persons at high risk of diabetes. In: Narayan KM, Williams
D, Gregg EW, Cowie CC, eds. Diabetes Public Health: From Data to Policy. New York: Oxford
University Press, Inc.; 2011:15-32.
29. Klein BE. Overview of epidemiologic studies of diabetic retinopathy. Ophthalmic Epidemiol.
2007;14(4):179-183.
30. Kempen JH, O'Colmain BJ, Leske MC, et al. The prevalence of diabetic retinopathy among adults in
the United States. Arch Ophthalmol. 2004;122(4):552-563.
31. Zhang X, Saaddine JB, Chou CF, et al. Prevalence of diabetic retinopathy in the United States, 2005-
2008. JAMA. 2010;304(6):649-656.
32. Chua J, Lim CXY, Wong TY, Sabanayagam C. Diabetic retinopathy in the Asia-Pacific. Asia Pac J
Ophthalmol (Phila). 2018;7(1):3-16.
33. Stram DA, Jiang X, Varma R, et al. Factors associated with prevalent diabetic retinopathy in Chinese
Americans: the Chinese American Eye Study. Ophthalmol Retina. 2018;2(2):96-105.
34. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The Wisconsin Epidemiologic Study of Diabetic
Retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years.
Arch Ophthalmol. 1984;102(4):520-526.
35. Varma R, Torres M, Pena F, Klein R, Azen SP. Prevalence of diabetic retinopathy in adult Latinos: the
Los Angeles Latino eye study. Ophthalmology. 2004;111(7):1298-1306.
36. Hirai FE, Knudtson MD, Klein BE, Klein R. Clinically significant macular edema and survival in type
1 and type 2 diabetes. Am J Ophthalmol. 2008;145(4):700-706.
37. West SK, Klein R, Rodriguez J, et al. Diabetes and diabetic retinopathy in a Mexican-American
population: Proyecto VER. Diabetes Care. 2001;24(7):1204-1209.
38. Tan GS, Gan A, Sabanayagam C, et al. Ethnic differences in the prevalence and risk factors of diabetic
retinopathy: the Singapore Epidemiology of Eye Diseases study. Ophthalmology. 2018;125(4):529-536.
39. Rudnisky CJ, Wong BK, Virani H, Tennant MTS. Risk factors for progression of diabetic retinopathy
in Alberta First Nations communities. Can J Ophthalmol. 2017;52 Suppl 1:S19-S29.
40. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The Wisconsin Epidemiologic Study of Diabetic
Retinopathy. III. Prevalence and risk of diabetic retinopathy when age at diagnosis is 30 or more years.
Arch Ophthalmol. 1984;102(4):527-532.
41. Xie X, Atkins E, Lv J, et al. Effects of intensive blood pressure lowering on cardiovascular and renal
outcomes: updated systematic review and meta-analysis. Lancet. 2016;387(10017):435-443.
42. Diabetes Control and Complications Trial Research Group. Progression of retinopathy with intensive
versus conventional treatment in the Diabetes Control and Complications Trial. Ophthalmology.
1995;102(4):647-661.
43. Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications
Research Group. Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of
intensive therapy. N Engl J Med. 2000;342(6):381-389.
44. Diabetes Control and Complications Trial Research Group. The relationship of glycemic exposure
(HbA1c) to the risk of development and progression of retinopathy in the Diabetes Control and
Complications Trial. Diabetes. 1995;44(8):968-983.

P132
62
 Diabetic Retinopathy PPP

45. Writing Team for the Diabetes Control and Complications Trial/Epidemiology of Diabetes
Interventions and Complications Research Group. Effect of intensive therapy on the microvascular
complications of type 1 diabetes mellitus. JAMA. 2002;287(19):2563-2569.
46. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas
or insulin compared with conventional treatment and risk of complications in patients with type 2
diabetes (UKPDS 33). Lancet. 1998;352(9131):837-853.
47. Kohner EM, Stratton IM, Aldington SJ, Holman RR, Matthews DR. Relationship between the severity
of retinopathy and progression to photocoagulation in patients with Type 2 diabetes mellitus in the
UKPDS (UKPDS 52). Diabet Med. 2001;18(3):178-184.
48. Wong TY, Liew G, Tapp RJ, et al. Relation between fasting glucose and retinopathy for diagnosis of
diabetes: three population-based cross-sectional studies. Lancet. 2008;371(9614):736-743.
49. White NH, Sun W, Cleary PA, et al. Prolonged effect of intensive therapy on the risk of retinopathy
complications in patients with type 1 diabetes mellitus: 10 years after the Diabetes Control and
Complications Trial. Arch Ophthalmol. 2008;126(12):1707-1715.
50. Buehler AM, Cavalcanti AB, Berwanger O, et al. Effect of tight blood glucose control versus
conventional control in patients with type 2 diabetes mellitus: a systematic review with meta-analysis
of randomized controlled trials. Cardiovasc Ther. 2013;31(3):147-160.
51. Do DV, Wang X, Vedula SS, et al. Blood pressure control for diabetic retinopathy. Cochrane Database
Syst Rev. 2015;1:CD006127.
52. Fullerton B, Jeitler K, Seitz M, Horvath K, Berghold A, Siebenhofer A. Intensive glucose control
versus conventional glucose control for type 1 diabetes mellitus. Cochrane Database Syst Rev.
2014(2):CD009122.
53. Virk SA, Donaghue KC, Wong TY, Craig ME. Interventions for Diabetic Retinopathy in Type 1
Diabetes: Systematic Review and Meta-Analysis. Am J Ophthalmol. 2015;160(5):1055-1064 e1054.
54. Davis MD, Fisher MR, Gangnon RE, et al. Risk factors for high-risk proliferative diabetic retinopathy
and severe visual loss: Early Treatment Diabetic Retinopathy Study report number 18. Invest
Ophthalmol Vis Sci. 1998;39(2):233-252.
55. Kilpatrick ES, Rigby AS, Atkin SL, Frier BM. Does severe hypoglycaemia influence microvascular
complications in Type 1 diabetes? An analysis of the Diabetes Control and Complications Trial
database. Diabet Med. 2012;29(9):1195-1198.
56. American Diabetes Association. Standards of medical care in diabetes--2013. Diabetes Care. 2013;36
Suppl 1:S11-66.
57. Varma R, Bressler NM, Doan QV, et al. Prevalence of and risk factors for diabetic macular edema in
the United States. JAMA Ophthalmol. 2014;132(11):1334-1340.
58. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and
microvascular complications in type 2 diabetes: UKPDS 38. BMJ. 1998;317(7160):703-713.
59. Snow V, Weiss KB, Mottur-Pilson C. The evidence base for tight blood pressure control in the
management of type 2 diabetes mellitus. Ann Intern Med. 2003;138(7):587-592.
60. van Leiden HA, Dekker JM, Moll AC, et al. Blood pressure, lipids, and obesity are associated with
retinopathy: The Hoorn Study. Diabetes Care. 2002;25(8):1320-1325.
61. Klein R, Sharrett AR, Klein BE, et al. The association of atherosclerosis, vascular risk factors, and
retinopathy in adults with diabetes: the atherosclerosis risk in communities study. Ophthalmology.
2002;109(7):1225-1234.
62. Lyons TJ, Jenkins AJ, Zheng D, et al. Diabetic retinopathy and serum lipoprotein subclasses in the
DCCT/EDIC cohort. Invest Ophthalmol Vis Sci. 2004;45(3):910-918.
63. Lopes-Virella MF, Baker NL, Hunt KJ, Lyons TJ, Jenkins AJ, Virella G. High concentrations of AGE-
LDL and oxidized LDL in circulating immune complexes are associated with progression of
retinopathy in type 1 diabetes. Diabetes Care. 2012;35(6):1333-1340.
64. Kang EY, Chen TH, Garg SJ, et al. Association of Statin Therapy With Prevention of Vision-
Threatening Diabetic Retinopathy. JAMA Ophthalmol. 2019.
65. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The Wisconsin Epidemiologic Study of Diabetic
Retinopathy. IX. Four-year incidence and progression of diabetic retinopathy when age at diagnosis is
less than 30 years. Arch Ophthalmol. 1989;107(2):237-243.
66. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. The Wisconsin Epidemiologic Study of Diabetic
Retinopathy. X. Four-year incidence and progression of diabetic retinopathy when age at diagnosis is
30 years or more. Arch Ophthalmol. 1989;107(2):244-249.

P133
63
 Diabetic Retinopathy PPP

67. Kriska AM, LaPorte RE, Patrick SL, Kuller LH, Orchard TJ. The association of physical activity and
diabetic complications in individuals with insulin-dependent diabetes mellitus: the Epidemiology of
Diabetes Complications Study--VII. J Clin Epidemiol. 1991;44(11):1207-1214.
68. Muni RH, Kohly RP, Lee EQ, Manson JE, Semba RD, Schaumberg DA. Prospective study of
inflammatory biomarkers and risk of diabetic retinopathy in the diabetes control and complications
trial. JAMA Ophthalmol. 2013;131(4):514-521.
69. Sumamo E, Ha C, Korownyk C, Vandermeer B, Dryden DM. In: Lifestyle Interventions for Four
Conditions: Type 2 Diabetes, Metabolic Syndrome, Breast Cancer, and Prostate Cancer. Rockville
(MD)2011.
70. American Diabetes Association. Standards of medical care in diabetes--2008. Diabetes Care. 2008;31
Suppl 1:S12-54.
71. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy: a systematic review. JAMA.
2007;298(8):902-916.
72. Shi R, Zhao L, Wang F, et al. Effects of lipid-lowering agents on diabetic retinopathy: a Meta-analysis
and systematic review. Int J Ophthalmol. 2018;11(2):287-295.
73. Mitchell SL, Neininger AC, Bruce CN, et al. Mitochondrial haplogroups modify the effect of diabetes
duration and HbA1c on proliferative diabetic retinopathy risk in patients with type 2 diabetes. Invest
Ophthalmol Vis Sci. 2017;58(14):6481-6488.
74. Liu E, Kaidonis G, Gillies MC, et al. Mitochondrial haplogroups are not associated with diabetic
retinopathy in a large Australian and British Caucasian sample. Nature Research Scientific Reports.
2019;9(612).
75. Chiefari E, Ventura V, Capula C, et al. A polymorphism of HMGA1 protects against proliferative
diabetic retinopathy by impairing HMGA1-induced VEGFA expression. Sci Rep. 2016;6:39429.
76. Grundy SM, Brewer HB, Jr., Cleeman JI, et al. Definition of metabolic syndrome: Report of the
National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues
related to definition. Circulation. 2004;109(3):433-438.
77. Lee MY, Hsiao PJ, Huang JC, Hsu WH, Chen SC, Shin SJ. Association between metabolic syndrome
and microvascular and macrovascular disease in type 2 diabetic mellitus. Am J Med Sci.
2018;355(4):342-349.
78. Lee CS, Lee AY, Baughman D, et al. The United Kingdom Diabetic Retinopathy Electronic Medical
Record Users Group: Report 3: Baseline retinopathy and clinical features predict progression of
diabetic retinopathy. Am J Ophthalmol. 2017;180:64-71.
79. Bressler SB, Beaulieu WT, Glassman AR, et al. Factors associated with worsening proliferative
diabetic retinopathy in eyes treated with panretinal photocoagulation or ranibizumab. Ophthalmology.
2017;124(4):431-439.
80. Ferris FL, III. How effective are treatments for diabetic retinopathy? JAMA. 1993;269(10):1290-1291.
81. Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of
diabetes on the development and progression of long-term complications in insulin-dependent diabetes
mellitus. N Engl J Med. 1993;329(14):977-986.
82. Diabetes Control and Complications Trial Research Group. The effect of intensive diabetes treatment
on the progression of diabetic retinopathy in insulin-dependent diabetes mellitus: the Diabetes Control
and Complications Trial. Arch Ophthalmol. 1995;113(1):36-51.
83. Epidemiology of Diabetes Interventions and Complications (EDIC) Research Group. Epidemiology of
Diabetes Interventions and Complications (EDIC): design, implementation, and preliminary results of a
long-term follow-up of the Diabetes Control and Complications Trial cohort. Diabetes Care.
1999;22(1):99-111.
84. Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications
(DCCT/EDIC) Research Group. Modern-day clinical course of type 1 diabetes mellitus after 30 years'
duration: the diabetes control and complications trial/epidemiology of diabetes interventions and
complications and Pittsburgh Epidemiology of Diabetes complications experience (1983-2005). Arch
Intern Med. 2009;169(14):1307-1316.
85. Diabetic Retinopathy Study Research Group. Indications for photocoagulation treatment of diabetic
retinopathy: Diabetic Retinopathy Study report number 14. Int Ophthalmol Clin. 1987;27(4):239-253.
86. Diabetic Retinopathy Study Research Group. Photocoagulation treatment of proliferative diabetic
retinopathy: the second report of Diabetic Retinopathy Study findings. Ophthalmology. 1978;85(1):82-
106.

P134
64
 Diabetic Retinopathy PPP

87. Early Treatment Diabetic Retinopathy Study Research Group. Grading diabetic retinopathy from
stereoscopic color fundus photographs--an extension of the modified Airlie House classification:
ETDRS report number 10. Ophthalmology. 1991;98(5 Suppl):786-806.
88. Early Treatment Diabetic Retinopathy Study Research Group. Fundus photographic risk factors for
progression of diabetic retinopathy: ETDRS report number 12. Ophthalmology. 1991;98(5 Suppl):823-
833.
89. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular
edema: Early Treatment Diabetic Retinopathy Study report number 1. Arch Ophthalmol.
1985;103(12):1796-1806.
90. Diabetic Retinopathy Vitrectomy Study Research Group. Early vitrectomy for severe vitreous
hemorrhage in diabetic retinopathy: four-year results of a randomized trial--Diabetic Retinopathy
Vitrectomy Study report 5. Arch Ophthalmol. 1990;108(7):958-964.
91. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic Study of diabetic
retinopathy. XIV. Ten-year incidence and progression of diabetic retinopathy. Arch Ophthalmol.
1994;112(9):1217-1228.
92. Scott R, Best J, Forder P, et al. FIELD Study Investigators. Fenofibrate Intervention and Event
Lowering in Diabetes (FIELD) study: baseline characteristics and short-term effects of fenofibrate
[ISRCTN64783481]. Cardiovasc Diabetol. 2005;4:13.
93. Goff DC, Jr., Gerstein HC, Ginsberg HN, et al. ACCORD Study Group. Prevention of cardiovascular
disease in persons with type 2 diabetes mellitus: current knowledge and rationale for the Action to
Control Cardiovascular Risk in Diabetes (ACCORD) trial. Am J Cardiol. 2007;99(12A):4i-20i.
94. Diabetic Retinopathy Clinical Research Network, Elman MJ, Qin H, et al. Intravitreal ranibizumab for
diabetic macular edema with prompt versus deferred laser treatment: three-year randomized trial
results. Ophthalmology. 2012;119(11):2312-2318.
95. Writing Committee for the Diabetic Retinopathy Clinical Research Network. Panretinal
photocoagulation vs intravitreous ranibizumab for proliferative diabetic retinopathy: a randomized
clinical trial. JAMA. 2015;314(20):2137-2146.
96. Diabetic Retinopathy Clinical Research Network. Aflibercept, bevacizumab, or ranibizumab for
diabetic macular edema. N Engl J Med. 2015;372(13):1193-1203.
97. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with
metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet.
1998;352(9131):854-865.
98. Early Treatment Diabetic Retinopathy Study Research Group. Fluorescein angiographic risk factors for
progression of diabetic retinopathy: ETDRS report number 13. Ophthalmology. 1991;98(5 Suppl):834-
840.
99. Early Treatment Diabetic Retinopathy Study Research Group. Classification of diabetic retinopathy
from fluorescein angiograms: ETDRS report number 11. Ophthalmology. 1991;98(5 Suppl):807-822.
100. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic Study of Diabetic
Retinopathy: XVII. The 14-year incidence and progression of diabetic retinopathy and associated risk
factors in type 1 diabetes. Ophthalmology. 1998;105(10):1801-1815.
101. Klein R, Moss SE, Klein BE, Davis MD, DeMets DL. The Wisconsin epidemiologic study of diabetic
retinopathy. XI. The incidence of macular edema. Ophthalmology. 1989;96(10):1501-1510.
102. Wang SY, Andrews CA, Herman WH, Gardner TW, Stein JD. Incidence and risk factors for
developing diabetic retinopathy among youths with type 1 or type 2 diabetes throughout the United
States. Ophthalmology. 2017;124(4):424-430.
103. Williams GA, Scott IU, Haller JA, Maguire AM, Marcus D, McDonald HR. Single-field fundus
photography for diabetic retinopathy screening: a report by the American Academy of Ophthalmology.
Ophthalmology. 2004;111(5):1055-1062.
104. Lin DY, Blumenkranz MS, Brothers RJ, Grosvenor DM. The sensitivity and specificity of single-field
nonmydriatic monochromatic digital fundus photography with remote image interpretation for diabetic
retinopathy screening: a comparison with ophthalmoscopy and standardized mydriatic color
photography. Am J Ophthalmol. 2002;134(2):204-213.
105. Larsen N, Godt J, Grunkin M, Lund-Andersen H, Larsen M. Automated detection of diabetic
retinopathy in a fundus photographic screening population. Invest Ophthalmol Vis Sci. 2003;44(2):767-
771.
106. Leese GP, Ellis JD, Morris AD, Ellingford A. Does direct ophthalmoscopy improve retinal screening
for diabetic eye disease by retinal photography? Diabet Med. 2002;19(10):867-869.

P135
65
 Diabetic Retinopathy PPP

107. Ahmed J, Ward TP, Bursell SE, Aiello LM, Cavallerano JD, Vigersky RA. The sensitivity and
specificity of nonmydriatic digital stereoscopic retinal imaging in detecting diabetic retinopathy.
Diabetes Care. 2006;29(10):2205-2209.
108. Velez R, Haffner S, Stern MP, Vanheuven WAJ. Ophthalmologist vs retinal photographs in screening
for diabetic retinopathy. Clinical Research. 1987;35(3):A363.
109. Pugh JA, Jacobson JM, Van Heuven WA, et al. Screening for diabetic retinopathy. The wide-angle
retinal camera. Diabetes Care. 1993;16(6):889-895.
110. Lawrence MG. The accuracy of digital-video retinal imaging to screen for diabetic retinopathy: an
analysis of two digital-video retinal imaging systems using standard stereoscopic seven-field
photography and dilated clinical examination as reference standards. Trans Am Ophthalmol Soc.
2004;102:321-340.
111. Abramoff MD, Folk JC, Han DP, et al. Automated analysis of retinal images for detection of referable
diabetic retinopathy. JAMA Ophthalmol. 2013;131(3):351-357.
112. Rudnisky CJ, Hinz BJ, Tennant MT, de Leon AR, Greve MD. High-resolution stereoscopic digital
fundus photography versus contact lens biomicroscopy for the detection of clinically significant
macular edema. Ophthalmology. 2002;109(2):267-274.
113. Cavallerano JD, Aiello LP, Cavallerano AA, et al. Nonmydriatic digital imaging alternative for annual
retinal examination in persons with previously documented no or mild diabetic retinopathy. Am J
Ophthalmol. 2005;140(4):667-673.
114. Fonda SJ, Bursell SE, Lewis DG, Garren J, Hock K, Cavallerano J. The relationship of a diabetes
telehealth eye care program to standard eye care and change in diabetes health outcomes. Telemed J E
Health. 2007;13(6):635-644.
115. Conlin PR, Fisch BM, Cavallerano AA, Cavallerano JD, Bursell SE, Aiello LM. Nonmydriatic
teleretinal imaging improves adherence to annual eye examinations in patients with diabetes. J Rehabil
Res Dev. 2006;43(6):733-740.
116. Diamond JP, McKinnon M, Barry C, et al. Non-mydriatic fundus photography: a viable alternative to
fundoscopy for identification of diabetic retinopathy in an Aboriginal population in rural Western
Australia? Aust N Z J Ophthalmol. 1998;26(2):109-115.
117. Klein R, Klein BE. Screening for diabetic retinopathy, revisited. Am J Ophthalmol. 2002;134(2):261-
263.
118. Maberley D, Walker H, Koushik A, Cruess A. Screening for diabetic retinopathy in James Bay,
Ontario: a cost-effectiveness analysis. CMAJ. 2003;168(2):160-164.
119. Farley TF, Mandava N, Prall FR, Carsky C. Accuracy of primary care clinicians in screening for
diabetic retinopathy using single-image retinal photography. Ann Fam Med. 2008;6(5):428-434.
120. Li HK, Horton M, Bursell SE, et al. Telehealth practice recommendations for diabetic retinopathy,
second edition. Telemed J E Health. 2011;17(10):814-837.
121. Lueder GT, Silverstein J. American Academy of Pediatrics Section on Ophthalmology and Section on
Endocrinology. Screening for retinopathy in the pediatric patient with type 1 diabetes mellitus.
Pediatrics. 2005;116(1):270-273. Reaffirmed 2014.
122. Diabetes Prevention Program Research Group. The prevalence of retinopathy in impaired glucose
tolerance and recent-onset diabetes in the Diabetes Prevention Program. Diabet Med. 2007;24(2):137-
144.
123. Klein BE, Moss SE, Klein R. Effect of pregnancy on progression of diabetic retinopathy. Diabetes
Care. 1990;13(1):34-40.
124. Chew EY, Mills JL, Metzger BE, et al. Metabolic control and progression of retinopathy. The Diabetes
in Early Pregnancy Study. National Institute of Child Health and Human Development Diabetes in
Early Pregnancy Study. Diabetes Care. 1995;18(5):631-637.
125. Diabetes Control and Complications Trial Research Group. Effect of pregnancy on microvascular
complications in the diabetes control and complications trial. Diabetes Care. 2000;23(8):1084-1091.
126. UK Prospective Diabetes Study (UKPDS). VIII. Study design, progress and performance.
Diabetologia. 1991;34(12):877-890.
127. Early Treatment Diabetic Retinopathy Study Research Group. Effects of aspirin treatment on diabetic
retinopathy: ETDRS report number 8. Ophthalmology. 1991;98(5 Suppl):757-765.
128. Chew EY, Klein ML, Murphy RP, Remaley NA, Ferris FL, III. Effects of aspirin on vitreous/preretinal
hemorrhage in patients with diabetes mellitus: Early Treatment Diabetic Retinopathy Study report
number 20. Arch Ophthalmol. 1995;113(1):52-55.

P136
66
 Diabetic Retinopathy PPP

129. Eye Disease Case-Control Study Group. Risk factors for idiopathic rhegmatogenous retinal
detachment. Am J Epidemiol. 1993;137(7):749-757.
130. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. Glycosylated hemoglobin predicts the incidence
and progression of diabetic retinopathy. JAMA. 1988;260(19):2864-2871.
131. Chew EY, Klein ML, Ferris FL, III, et al. Association of elevated serum lipid levels with retinal hard
exudate in diabetic retinopathy: Early Treatment Diabetic Retinopathy Study (ETDRS) report 22. Arch
Ophthalmol. 1996;114(9):1079-1084.
132. Early Treatment Diabetic Retinopathy Study Research Group. Early photocoagulation for diabetic
retinopathy: ETDRS report number 9. Ophthalmology. 1991;98(5 Suppl):766-785.
133. Klein R, Klein BE, Neider MW, Hubbard LD, Meuer SM, Brothers RJ. Diabetic retinopathy as
detected using ophthalmoscopy, a nonmydriatic camera and a standard fundus camera. Ophthalmology.
1985;92(4):485-491.
134. Klein R, Klein BE, Moss SE, Davis MD, DeMets DL. Retinopathy in young-onset diabetic patients.
Diabetes Care. 1985;8(4):311-315.
135. Frank RN, Hoffman WH, Podgor MJ, et al. Retinopathy in juvenile-onset diabetes of short duration.
Ophthalmology. 1980;87(1):1-9.
136. Krolewski AS, Warram JH, Rand LI, Christlieb AR, Busick EJ, Kahn CR. Risk of proliferative diabetic
retinopathy in juvenile-onset type I diabetes: a 40-yr follow-up study. Diabetes Care. 1986;9(5):443-
452.
137. Klein R, Klein BE, Moss SE. Epidemiology of proliferative diabetic retinopathy. Diabetes Care.
1992;15(12):1875-1891.
138. Gunderson EP, Lewis CE, Tsai AL, et al. A 20-year prospective study of childbearing and incidence of
diabetes in young women, controlling for glycemia before conception: the Coronary Artery Risk
Development in Young Adults (CARDIA) Study. Diabetes. 2007;56(12):2990-2996.
139. Hirano T, Kitahara J, Toriyama Y, Kasamatsu H, Murata T, Sadda S. Quantifying vascular density and
morphology using different swept-source optical coherence tomography angiographic scan patterns in
diabetic retinopathy. Br J Ophthalmol. 2018.
140. Nesper PL, Roberts PK, Onishi AC, et al. Quantifying microvascular abnormalities with increasing
severity of diabetic retinopathy using optical coherence tomography angiography. Invest Ophthalmol
Vis Sci. 2017;58(6):BIO307-BIO315.
141. Onishi AC, Nesper PL, Roberts PK, et al. Importance of considering the middle capillary plexus on
OCT angiography in diabetic retinopathy. Invest Ophthalmol Vis Sci. 2018;59(5):2167-2176.
142. Samara WA, Shahlaee A, Adam MK, et al. Quantification of diabetic macular ischemia using optical
coherence tomography angiography and its relationship with visual acuity. Ophthalmology.
2017;124(2):235-244.
143. Kaiser PK, Riemann CD, Sears JE, Lewis H. Macular traction detachment and diabetic macular edema
associated with posterior hyaloidal traction. Am J Ophthalmol. 2001;131(1):44-49.
144. Martidis A, Duker JS, Greenberg PB, et al. Intravitreal triamcinolone for refractory diabetic macular
edema. Ophthalmology. 2002;109(5):920-927.
145. Strom C, Sander B, Larsen N, Larsen M, Lund-Andersen H. Diabetic macular edema assessed with
optical coherence tomography and stereo fundus photography. Invest Ophthalmol Vis Sci.
2002;43(1):241-245.
146. McDonald HR, Williams GA, Scott IU, Haller JA, Maguire AM, Marcus DM. Laser scanning imaging
for macular disease: a report by the American Academy of Ophthalmology. Ophthalmology.
2007;114(6):1221-1228.
147. Virgili G, Menchini F, Dimastrogiovanni AF, et al. Optical coherence tomography versus stereoscopic
fundus photography or biomicroscopy for diagnosing diabetic macular edema: a systematic review.
Invest Ophthalmol Vis Sci. 2007;48(11):4963-4973.
148. Bressler NM, Edwards AR, Antoszyk AN, et al. Retinal thickness on Stratus optical coherence
tomography in people with diabetes and minimal or no diabetic retinopathy. Am J Ophthalmol.
2008;145(5):894-901.
149. Davis MD, Bressler SB, Aiello LP, et al. Comparison of time-domain OCT and fundus photographic
assessments of retinal thickening in eyes with diabetic macular edema. Invest Ophthalmol Vis Sci.
2008;49(5):1745-1752.
150. Do DV, Nguyen QD, Khwaja AA, et al. READ-2 Study Group. Ranibizumab for edema of the macula
in diabetes study: 3-year outcomes and the need for prolonged frequent treatment. JAMA Ophthalmol.
2013;131(2):139-145.

P137
67
 Diabetic Retinopathy PPP

151. Brown DM, Nguyen QD, Marcus DM, et al. RIDE and RISE Research Group. Long-term outcomes of
ranibizumab therapy for diabetic macular edema: the 36-month results from two phase III trials: RISE
and RIDE. Ophthalmology. 2013;120(10):2013-2022.
152. Wells JA, Glassman AR, Ayala AR, et al. Aflibercept, bevacizumab, or ranibizumab for diabetic
macular edema: two-year results from a comparative effectiveness randomized clinical trial.
Ophthalmology. 2016;123(6):1351-1359.
153. Dhoot DS, Baker K, Saroj N, et al. Baseline factors affecting changes in diabetic retinopathy severity
scale score after intravitreal aflibercept or laser for diabetic macular edema: post hoc analyses from
VISTA and VIVID. Ophthalmology. 2018;125(1):51-56.
154. Browning DJ, Glassman AR, Aiello LP, et al. Relationship between optical coherence tomography-
measured central retinal thickness and visual acuity in diabetic macular edema. Ophthalmology.
2007;114(3):525-536.
155. Browning DJ, Apte RS, Bressler SB, et al. Association of the extent of diabetic macular edema as
assessed by optical coherence tomography with visual acuity and retinal outcome variables. Retina.
2009;29(3):300-305.
156. Early Treatment Diabetic Retinopathy Study Research Group. Focal photocoagulation treatment of
diabetic macular edema. Relationship of treatment effect to fluorescein angiographic and other retinal
characteristics at baseline: ETDRS report number 19. Arch Ophthalmol. 1995;113(9):1144-1155.
157. Silva PS, Dela Cruz AJ, Ledesma MG, et al. Diabetic Retinopathy Severity and Peripheral Lesions Are
Associated with Nonperfusion on Ultrawide Field Angiography. Ophthalmology. 2015;122(12):2465-
2472.
158. Yannuzzi LA, Rohrer KT, Tindel LJ, et al. Fluorescein angiography complication survey.
Ophthalmology. 1986;93(5):611-617.
159. Sunness JS. The pregnant woman's eye. Surv Ophthalmol. 1988;32(4):219-238.
160. Kim AY, Chu Z, Shahidzadeh A, Wang RK, Puliafito CA, Kashani AH. Quantifying microvascular
density and morphology in diabetic retinopathy using spectral-domain optical coherence tomography
angiography. Invest Ophthalmol Vis Sci. 2016;57(9):OCT362-370.
161. Lu Y, Simonett JM, Wang J, et al. Evaluation of automatically quantified foveal avascular zone metrics
for diagnosis of diabetic retinopathy using optical coherence tomography angiography. Invest
Ophthalmol Vis Sci. 2018;59(6):2212-2221.
162. Hwang TS, Zhang M, Bhavsar K, et al. Visualization of 3 distinct retinal plexuses by projection-
resolved optical coherence tomography angiography in diabetic retinopathy. JAMA Ophthalmol.
2016;134(12):1411-1419.
163. Ashraf M, Nesper PL, Jampo L, Yu F, Fawzi AA. Statistical model of optical coherence tomography
angiography parameters that correlate with severity of diabetic retinopathy. Invest Ophthalmol Vis Sci.
2018: In press.
164. Hwang TS, Jia Y, Gao SS, et al. Optical coherence tomography angiography features of diabetic
retinopathy. Retina. 2015;35(11):2371-2376.
165. Couturier A, Mane V, Bonnin S, et al. Capillary plexus anomalies in diabetic retinopathy on optical
coherence tomography angiography. Retina. 2015;35(11):2384-2391.
166. Vujosevic S, Muraca A, Alkabes M, et al. Early microvascular and neural changes in patients with type
1 and type 2 diabetes mellitus without clinical signs of diabetic retinopathy. Retina. 2017:[Epub ahead
of print].
167. Russell JF, Shi Y, Hinkle JW, et al. Longitudinal Wide Field Swept Source OCT Angiography of
Neovascularization in Proliferative Diabetic Retinopathy After Panretinal Photocoagulation.
Ophthalmology. 2018.
168. Ishibazawa A, Nagaoka T, Yokota H, et al. Characteristics of Retinal Neovascularization in
Proliferative Diabetic Retinopathy Imaged by Optical Coherence Tomography Angiography. Invest
Ophthalmol Vis Sci. 2016;57(14):6247-6255.
169. Spaide RF, Fujimoto JG, Waheed NK. Image artifacts in optical coherence tomography angiography.
Retina. 2015;35(11):2163-2180.
170. Fawzi AA. Consensus on OCT angiography nomenclature: Do we need to develop and learn a new
language? JAMA Ophthalmol. 2018: In press.
171. Hirano T, Kakihara S, Toriyama Y, Nittala MG, Murata T, Sadda S. Wide-field en face swept-source
optical coherence tomography angiography using extended field imaging in diabetic retinopathy. Br J
Ophthalmol. 2017:[Epub ahead of print].

P138
68
 Diabetic Retinopathy PPP

172. Sawada O, Ichiyama Y, Obata S, et al. Comparison between wide-angle OCT angiography and ultra-
wide field fluorescein angiography for detecting non-perfusion areas and retinal neovascularization in
eyes with diabetic retinopathy. Graefes Arch Clin Exp Ophthalmol. 2018;256(7):1275-1280.
173. Schaal KB, Munk MR, Wyssmueller I, Berger LE, Zinkernagel MS, Wolf S. Vascular Abnormalities in
diabetic retinopathy assessed with swept-source optical coherence tomography angiography widefield
imaging. Retina. 2017:[Epub ahead of print].
174. Hutton DW, Stein JD, Bressler NM, et al. Cost-effectiveness of Intravitreous Ranibizumab Compared
With Panretinal Photocoagulation for Proliferative Diabetic Retinopathy: Secondary Analysis From a
Diabetic Retinopathy Clinical Research Network Randomized Clinical Trial. JAMA Ophthalmol.
2017;135(6):576-584.
175. Ross EL, Hutton DW, Stein JD, et al. Cost-effectiveness of Aflibercept, Bevacizumab, and
Ranibizumab for Diabetic Macular Edema Treatment: Analysis From the Diabetic Retinopathy Clinical
Research Network Comparative Effectiveness Trial. JAMA Ophthalmol. 2016;134(8):888-896.
176. National Diabetes Education Program. Redesigning the Health Care Team: Diabetes Prevention and
Lifelong Management. Bethesda, MD: CreateSpace Publishing; 2014.
177. Tuomilehto J, Lindstrom J, Eriksson JG, et al. Prevention of type 2 diabetes mellitus by changes in
lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344(18):1343-1350.
178. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with
lifestyle intervention or metformin. N Engl J Med. 2002;346(6):393-403.
179. Assurance NCfQ. State of Health Care Quality Report. 2017. https://www.ncqa.org/report-cards/health-
plans/state-of-health-care-quality-report/. Accessed September 2019.
180. Kraft SK, Marrero DG, Lazaridis EN, Fineberg N, Qiu C, Clark CM, Jr. Primary care physicians'
practice patterns and diabetic retinopathy: current levels of care. Arch Fam Med. 1997;6(1):29-37.
181. Paz SH, Varma R, Klein R, Wu J, Azen SP. Los Angeles Latino Eye Study Group. Noncompliance
with vision care guidelines in Latinos with type 2 diabetes mellitus: the Los Angeles Latino Eye Study.
Ophthalmology. 2006;113(8):1372-1377.
182. National Committee for Quality Assurance. Improving quality and patient experience: the state of
health care quality 2013. 2013:53. Available at:
www.ncqa.org/Portals/0/Newsroom/SOHC/2013/SOHC-web_version_report.pdf. Accessed September
2019.
183. Glassman AR, Beck RW, Browning DJ, Danis RP, Kollman C. Comparison of optical coherence
tomography in diabetic macular edema, with and without reading center manual grading from a clinical
trials perspective. Invest Ophthalmol Vis Sci. 2009;50(2):560-566.
184. Diabetic Retinopathy Clinical Research N, Brucker AJ, Qin H, et al. Observational study of the
development of diabetic macular edema following panretinal (scatter) photocoagulation given in 1 or 4
sittings. Arch Ophthalmol. 2009;127(2):132-140.
185. Colucciello M. Vision loss due to macular edema induced by rosiglitazone treatment of diabetes
mellitus. Arch Ophthalmol. 2005;123(9):1273-1275.
186. Ryan EH, Jr., Han DP, Ramsay RC, et al. Diabetic macular edema associated with glitazone use.
Retina. 2006;26(5):562-570.
187. Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab for diabetic macular edema: results from 2
phase III randomized trials: RISE and RIDE. Ophthalmology. 2012;119(4):789-801.
188. Olsen TW. Anti-VEGF pharmacotherapy as an alternative to panretinal laser photocoagulation for
proliferative diabetic retinopathy. JAMA. 2015;314(20):2135-2136.
189. Elman MJ, Bressler NM, Qin H, et al. Diabetic Retinopathy Clinical Research Network. Expanded 2-
year follow-up of ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for
diabetic macular edema. Ophthalmology. 2011;118(4):609-614.
190. Baker CW, Glassman AR, Beaulieu WT, et al. Effect of Initial Management With Aflibercept vs Laser
Photocoagulation vs Observation on Vision Loss Among Patients With Diabetic Macular Edema
Involving the Center of the Macula and Good Visual Acuity: A Randomized Clinical Trial. JAMA.
2019.
191. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular
edema: Early Treatment Diabetic Retinopathy Study report number 4. Int Ophthalmol Clin.
1987;27(4):265-272.
192. Chalam KV, Bressler SB, Edwards AR, et al. Retinal thickness in people with diabetes and minimal or
no diabetic retinopathy: Heidelberg Spectralis optical coherence tomography. Invest Ophthalmol Vis
Sci. 2012;53(13):8154-8161.

P139
69
 Diabetic Retinopathy PPP

193. Early Treatment Diabetic Retinopathy Study Research Group. Early Treatment Diabetic Retinopathy
Study design and baseline patient characteristics: ETDRS report number 7. Ophthalmology. 1991;98(5
Suppl):741-756.
194. Diabetic Retinopathy Clinical Research N, Elman MJ, Aiello LP, et al. Randomized trial evaluating
ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular
edema. Ophthalmology. 2010;117(6):1064-1077 e1035.
195. Do DV, Nguyen QD, Boyer D, et al. DA VINCI Study Group. One-year outcomes of the DA VINCI
Study of VEGF Trap-Eye in eyes with diabetic macular edema. Ophthalmology. 2012;119(8):1658-
1665.
196. Diabetic Retinopathy Clinical Research Network, Googe J, Brucker AJ, Bressler NM, et al.
Randomized trial evaluating short-term effects of intravitreal ranibizumab or triamcinolone acetonide
on macular edema after focal/grid laser for diabetic macular edema in eyes also receiving panretinal
photocoagulation. Retina. 2011;31(6):1009-1027.
197. Diabetic Retinopathy Clinical Research Network, Elman MJ, Aiello LP, Beck RW, et al. Randomized
trial evaluating ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for
diabetic macular edema. Ophthalmology. 2010;117(6):1064-1077.
198. Ho AC, Scott IU, Kim SJ, et al. Anti-vascular endothelial growth factor pharmacotherapy for diabetic
macular edema: a report by the American Academy of Ophthalmology. Ophthalmology.
2012;119(10):2179-2188.
199. Mitchell P, Bandello F, Schmidt-Erfurth U, et al. RESTORE Study Group. The RESTORE study:
ranibizumab monotherapy or combined with laser versus laser monotherapy for diabetic macular
edema. Ophthalmology. 2011;118(4):615-625.
200. Thomas BJ, Shienbaum G, Boyer DS, Flynn HW, Jr. Evolving strategies in the management of diabetic
macular edema: clinical trials and current management. Can J Ophthalmol. 2013;48(1):22-30.
201. US Food and Drug Administration, Center for Drug Evaluation and Research. Lucentis (ranibizumab
injection). BLA 25156. Available at:
www.accessdata.fda.gov/drugsatfda_docs/label/2012/125156s0069s0076lbl.pdf. Accessed September
2019.
202. Payne JF, Wykoff CC, Clark WL, et al. Randomized Trial of Treat and Extend Ranibizumab With and
Without Navigated Laser Versus Monthly Dosing for Diabetic Macular Edema: TREX-DME 2-Year
Outcomes. Am J Ophthalmol. 2019;202:91-99.
203. Nguyen QD, Shah SM, Khwaja AA, et al. READ-2 Study Group. Two-year outcomes of the
ranibizumab for edema of the mAcula in diabetes (READ-2) study. Ophthalmology.
2010;117(11):2146-2151.
204. Rajendram R, Fraser-Bell S, Kaines A, et al. A 2-year prospective randomized controlled trial of
intravitreal bevacizumab or laser therapy (BOLT) in the management of diabetic macular edema: 24-
month data: report 3. Arch Ophthalmol. 2012;130(8):972-979.
205. Regnier S, Malcolm W, Allen F, Wright J, Bezlyak V. Efficacy of anti-VEGF and laser
photocoagulation in the treatment of visual impairment due to diabetic macular edema: a systematic
review and network meta-analysis. PLoS One. 2014;9(7):e102309.
206. Lau PE, Jenkins KS, Layton CJ. Current Evidence for the Prevention of Endophthalmitis in Anti-VEGF
Intravitreal Injections. J Ophthalmol. 2018;2018:8567912.
207. Parke DW, II, Coleman AL, Rich WL, III, Lum F. Choosing Wisely: five ideas that physicians and
patients can discuss. Ophthalmology. 2013;120(3):443-444.
208. Eadie BD, Etminan M, Carleton BC, Maberley DA, Mikelberg FS. Association of Repeated
Intravitreous Bevacizumab Injections With Risk for Glaucoma Surgery. JAMA Ophthalmol.
2017;135(4):363-368.
209. Kahook MY, Ammar DA. In vitro effects of antivascular endothelial growth factors on cultured human
trabecular meshwork cells. J Glaucoma. 2010;19(7):437-441.
210. Yannuzzi NA, Patel SN, Bhavsar KV, Sugiguchi F, Freund KB. Predictors of sustained intraocular
pressure elevation in eyes receiving intravitreal anti-vascular endothelial growth factor therapy. Am J
Ophthalmol. 2014;158(2):319-327 e312.
211. Avery RL, Gordon GM. Systemic Safety of Prolonged Monthly Anti-Vascular Endothelial Growth
Factor Therapy for Diabetic Macular Edema: A Systematic Review and Meta-analysis. JAMA
Ophthalmol. 2016;134(1):21-29.

P140
70
 Diabetic Retinopathy PPP

212. Virgili G, Parravano M, Evans JR, Gordon I, Lucenteforte E. Anti-vascular endothelial growth factor
for diabetic macular oedema: a network meta-analysis. Cochrane Database Syst Rev.
2018;10:CD007419.
213. Early Treatment Diabetic Retinopathy Study Research Group. Techniques for scatter and local
photocoagulation treatment of diabetic retinopathy: Early Treatment Diabetic Retinopathy Study report
number 3. Int Ophthalmol Clin. 1987;27(4):254-264.
214. Early Treatment Diabetic Retinopathy Study Research Group. Treatment techniques and clinical
guidelines for photocoagulation of diabetic macular edema. Early Treatment Diabetic Retinopathy
Study report number 2. Ophthalmology. 1987;94(7):761-774.
215. Fong DS, Strauber SF, Aiello LP, et al. Comparison of the modified Early Treatment Diabetic
Retinopathy Study and mild macular grid laser photocoagulation strategies for diabetic macular edema.
Arch Ophthalmol. 2007;125(4):469-480.
216. Jorge EC, Jorge EN, Botelho M, Farat JG, Virgili G, El Dib R. Monotherapy laser photocoagulation for
diabetic macular oedema. Cochrane Database Syst Rev. 2018;10:CD010859.
217. Guyer DR, D'Amico DJ, Smith CW. Subretinal fibrosis after laser photocoagulation for diabetic
macular edema. Am J Ophthalmol. 1992;113(6):652-656.
218. Han DP, Mieler WF, Burton TC. Submacular fibrosis after photocoagulation for diabetic macular
edema. Am J Ophthalmol. 1992;113(5):513-521.
219. Fong DS, Segal PP, Myers F, Ferris FL, Hubbard LD, Davis MD. Early Treatment Diabetic
Retinopathy Study Research Group. Subretinal fibrosis in diabetic macular edema: ETDRS report no.
23. Arch Ophthalmol. 1997;115(7):873-877.
220. Lewis H, Schachat AP, Haimann MH, et al. Choroidal neovascularization after laser photocoagulation
for diabetic macular edema. Ophthalmology. 1990;97(4):503-510; discussion 510-501.
221. Diabetic Retinopathy Clinical Research Network, Chew E, Strauber S, et al. Randomized trial of
peribulbar triamcinolone acetonide with and without focal photocoagulation for mild diabetic macular
edema: a pilot study. Ophthalmology. 2007;114(6):1190-1196.
222. Diabetic Retinopathy Clinical Research Network. A randomized trial comparing intravitreal
triamcinolone acetonide and focal/grid photocoagulation for diabetic macular edema. Ophthalmology.
2008;115(9):1447-1459.
223. Diabetic Retinopathy Clinical Research Network. Three-year follow-up of a randomized trial
comparing focal/grid photocoagulation and intravitreal triamcinolone for diabetic macular edema. Arch
Ophthalmol. 2009;127(3):245-251.
224. Diabetic Retinopathy Clinical Research N, Elman MJ, Qin H, et al. Intravitreal ranibizumab for
diabetic macular edema with prompt versus deferred laser treatment: three-year randomized trial
results. Ophthalmology. 2012;119(11):2312-2318.
225. Boyer DS, Yoon YH, Belfort R, Jr., et al. Three-year, randomized, sham-controlled trial of
dexamethasone intravitreal implant in patients with diabetic macular edema. Ophthalmology.
2014;121(10):1904-1914.
226. Campochiaro PA, Brown DM, Pearson A, et al. Long-term benefit of sustained-delivery fluocinolone
acetonide vitreous inserts for diabetic macular edema. Ophthalmology. 2011;118(4):626-635 e622.
227. Maturi RK, Glassman AR, Liu D, et al. Effect of adding dexamethasone to continued ranibizumab
treatment in patients with persistent diabetic macular edema: a DRCR Network phase 2 randomized
clinical trial. JAMA Ophthalmol. 2018;136(1):29-38.
228. Mehta H, Hennings C, Gillies MC, Nguyen V, Campain A, Fraser-Bell S. Anti-vascular endothelial
growth factor combined with intravitreal steroids for diabetic macular oedema. Cochrane Database
Syst Rev. 2018;4:CD011599.
229. Massin P, Duguid G, Erginay A, Haouchine B, Gaudric A. Optical coherence tomography for
evaluating diabetic macular edema before and after vitrectomy. Am J Ophthalmol. 2003;135(2):169-
177.
230. Otani T, Kishi S. A controlled study of vitrectomy for diabetic macular edema. Am J Ophthalmol.
2002;134(2):214-219.
231. Yamamoto T, Hitani K, Tsukahara I, et al. Early postoperative retinal thickness changes and
complications after vitrectomy for diabetic macular edema. Am J Ophthalmol. 2003;135(1):14-19.
232. Diabetic Retinopathy Clinical Research Network Writing Committee, Haller JA, Qin H, Apte RS, et al.
Vitrectomy outcomes in eyes with diabetic macular edema and vitreomacular traction. Ophthalmology.
2010;117(6):1087-1093.

P141
71
 Diabetic Retinopathy PPP

233. Luttrull JK, Dorin G. Subthreshold diode micropulse laser photocoagulation (SDM) as invisible retinal
phototherapy for diabetic macular edema: a review. Curr Diabetes Rev. 2012;8(4):274-284.
234. Wu Y, Ai P, Ai Z, Xu G. Subthreshold diode micropulse laser versus conventional laser
photocoagulation monotherapy or combined with anti-VEGF therapy for diabetic macular edema: A
Bayesian network meta-analysis. Biomed Pharmacother. 2018;97:293-299.
235. Sahoo S, Barua A, Myint KT, Haq A, Abas AB, Nair NS. Topical non-steroidal anti-inflammatory
agents for diabetic cystoid macular oedema. Cochrane Database Syst Rev. 2015(2):CD010009.
236. Schatz H, Madeira D, McDonald HR, Johnson RN. Progressive enlargement of laser scars following
grid laser photocoagulation for diffuse diabetic macular edema. Arch Ophthalmol. 1991;109(11):1549-
1551.
237. Braun CI, Benson WE, Remaley NA, Chew EY, Ferris FL, III. Accommodative amplitudes in the Early
Treatment Diabetic Retinopathy Study. Retina. 1995;15(4):275-281.
238. Novak MA, Rice TA, Michels RG, Auer C. Vitreous hemorrhage after vitrectomy for diabetic
retinopathy. Ophthalmology. 1984;91(12):1485-1489.
239. Gupta B, Sivaprasad S, Wong R, et al. Visual and anatomical outcomes following vitrectomy for
complications of diabetic retinopathy: the DRIVE UK study. Eye (Lond). 2012;26(4):510-516.
240. Schachat AP, Oyakawa RT, Michels RG, Rice TA. Complications of vitreous surgery for diabetic
retinopathy. II. Postoperative complications. Ophthalmology. 1983;90(5):522-530.
241. Aaberg TM, Van Horn DL. Late complications of pars plana vitreous surgery. Ophthalmology.
1978;85(2):126-140.
242. Chu KM, Chen TT, Lee PY. Clinical results of pars plana vitrectomy in posterior-segment disorders.
Ann Ophthalmol. 1985;17(11):686-693.
243. Chew EY, Benson WE, Remaley NA, et al. Results after lens extraction in patients with diabetic
retinopathy: Early Treatment Diabetic Retinopathy Study report number 25. Arch Ophthalmol.
1999;117(12):1600-1606.
244. Gillies MC, Sutter FK, Simpson JM, Larsson J, Ali H, Zhu M. Intravitreal triamcinolone for refractory
diabetic macular edema: two-year results of a double-masked, placebo-controlled, randomized clinical
trial. Ophthalmology. 2006;113(9):1533-1538.
245. Chieh JJ, Roth DB, Liu M, et al. Intravitreal triamcinolone acetonide for diabetic macular edema.
Retina. 2005;25(7):828-834.
246. Diabetic Retinopathy Study Research Group. Photocoagulation treatment of proliferative diabetic
retinopathy: clinical application of Diabetic Retinopathy Study (DRS) findings, DRS report number 8.
Ophthalmology. 1981;88(7):583-600.
247. Diabetic Retinopathy Study Research Group. Four risk factors for severe visual loss in diabetic
retinopathy: the third report from the Diabetic Retinopathy Study. Arch Ophthalmol. 1979;97(4):654-
655.
248. Obeid A, Gao X, Ali FS, et al. Loss to follow-up in patients with proliferative diabetic retinopathy after
panretinal photocoagulation or intravitreal anti-VEGF injections. Ophthalmology. 2018:[Epub ahead of
print].
249. Diabetic Retinopathy Vitrectomy Study Research Group. Early vitrectomy for severe vitreous
hemorrhage in diabetic retinopathy: two-year results of a randomized trial--Diabetic Retinopathy
Vitrectomy Study report 2. Arch Ophthalmol. 1985;103(11):1644-1652.
250. Diabetic Retinopathy Vitrectomy Study Research Group. Early vitrectomy for severe proliferative
diabetic retinopathy in eyes with useful vision: clinical application of results of a randomized trial--
Diabetic Retinopathy Vitrectomy Study report 4. Ophthalmology. 1988;95(10):1321-1334.
251. Diabetic Retinopathy Vitrectomy Study Research Group. Early vitrectomy for severe proliferative
diabetic retinopathy in eyes with useful vision: results of a randomized trial--Diabetic Retinopathy
Vitrectomy Study report 3. Ophthalmology. 1988;95(10):1307-1320.
252. Writing Committee for the Diabetic Retinopathy Clinical Research N, Gross JG, Glassman AR, et al.
Panretinal Photocoagulation vs Intravitreous Ranibizumab for Proliferative Diabetic Retinopathy: A
Randomized Clinical Trial. JAMA. 2015;314(20):2137-2146.
253. Sun JK, Glassman AR, Beaulieu WT, et al. Rationale and Application of the Protocol S Anti-Vascular
Endothelial Growth Factor Algorithm for Proliferative Diabetic Retinopathy. Ophthalmology. 2018.
254. Gross JG, Glassman AR, Liu D, et al. Five-Year Outcomes of Panretinal Photocoagulation vs
Intravitreous Ranibizumab for Proliferative Diabetic Retinopathy: A Randomized Clinical Trial. JAMA
Ophthalmol. 2018;136(10):1138-1148.

P142
72
 Diabetic Retinopathy PPP

255. Sivaprasad S, Prevost AT, Vasconcelos JC, et al. Clinical efficacy of intravitreal aflibercept versus
panretinal photocoagulation for best corrected visual acuity in patients with proliferative diabetic
retinopathy at 52 weeks (CLARITY): a multicentre, single-blinded, randomised, controlled, phase 2b,
non-inferiority trial. Lancet. 2017;389(10085):2193-2203.
256. Ip MS, Domalpally A, Sun JK, Ehrlich JS. Long-term effects of therapy with ranibizumab on diabetic
retinopathy severity and baseline risk factors for worsening retinopathy. Ophthalmology.
2015;122(2):367-374.
257. Obeid A, Su D, Patel SN, et al. Outcomes of Eyes Lost to Follow-up with Proliferative Diabetic
Retinopathy That Received Panretinal Photocoagulation versus Intravitreal Anti-Vascular Endothelial
Growth Factor. Ophthalmology. 2019;126(3):407-413.
258. Parikh RN, Traband A, Kolomeyer AM, et al. Intravitreal Bevacizumab for the Treatment of Vitreous
Hemorrhage Due to Proliferative Diabetic Retinopathy. Am J Ophthalmol. 2017;176:194-202.
259. Diabetic Retinopathy Clinical Research N. Randomized clinical trial evaluating intravitreal
ranibizumab or saline for vitreous hemorrhage from proliferative diabetic retinopathy. JAMA
Ophthalmol. 2013;131(3):283-293.
260. Arevalo JF, Maia M, Flynn HW, Jr., et al. Tractional retinal detachment following intravitreal
bevacizumab (Avastin) in patients with severe proliferative diabetic retinopathy. Br J Ophthalmol.
2008;92(2):213-216.
261. Bressler SB, Qin H, Melia M, et al. Exploratory analysis of the effect of intravitreal ranibizumab or
triamcinolone on worsening of diabetic retinopathy in a randomized clinical trial. JAMA Ophthalmol.
2013;131(8):1033-1040.
262. Bressler SB, Liu D, Glassman AR, et al. Change in diabetic retinopathy through 2 years: secondary
analysis of a randomized clinical trial comparing aflibercept, bevacizumab, and ranibizumab. JAMA
Ophthalmol. 2017;135(6):558-568.
263. Ip MS, Domalpally A, Hopkins JJ, Wong P, Ehrlich JS. Long-term effects of ranibizumab on diabetic
retinopathy severity and progression. Arch Ophthalmol. 2012;130(9):1145-1152.
264. Wykoff CC, Eichenbaum DA, Roth DB, Hill L, Fung AE, Haskova Z. Ranibizumab Induces
Regression of Diabetic Retinopathy in Most Patients at High Risk of Progression to Proliferative
Diabetic Retinopathy. Ophthalmology Retina. 2018.
265. Brown DM, Schmidt-Erfurth U, Do DV, et al. Intravitreal aflibercept for diabetic macular edema: 100-
week results from the VISTA and VIVID studies. Ophthalmology. 2015;122(10):2044-2052.
266. Early vitrectomy for severe vitreous hemorrhage in diabetic retinopathy. Two-year results of a
randomized trial. Diabetic Retinopathy Vitrectomy Study report 2. The Diabetic Retinopathy
Vitrectomy Study Research Group. Arch Ophthalmol. 1985;103(11):1644-1652.
267. Early vitrectomy for severe proliferative diabetic retinopathy in eyes with useful vision. Results of a
randomized trial--Diabetic Retinopathy Vitrectomy Study Report 3. The Diabetic Retinopathy
Vitrectomy Study Research Group. Ophthalmology. 1988;95(10):1307-1320.
268. Recchia FM, Scott IU, Brown GC, Brown MM, Ho AC, Ip MS. Small-gauge pars plana vitrectomy: a
report by the American Academy of Ophthalmology. Ophthalmology. 2010;117(9):1851-1857.
269. Simunovic MP, Maberley DA. ANTI-VASCULAR ENDOTHELIAL GROWTH FACTOR THERAPY
FOR PROLIFERATIVE DIABETIC RETINOPATHY: A Systematic Review and Meta-Analysis.
Retina. 2015;35(10):1931-1942.
270. Zhao XY, Xia S, Chen YX. Antivascular endothelial growth factor agents pretreatment before
vitrectomy for complicated proliferative diabetic retinopathy: a meta-analysis of randomised controlled
trials. Br J Ophthalmol. 2018;102(8):1077-1085.
271. Smith JM, Steel DH. Anti-vascular endothelial growth factor for prevention of postoperative vitreous
cavity haemorrhage after vitrectomy for proliferative diabetic retinopathy. Cochrane Database Syst
Rev. 2015(8):CD008214.
272. Jacobson DR, Murphy RP, Rosenthal AR. The treatment of angle neovascularization with panretinal
photocoagulation. Ophthalmology. 1979;86(7):1270-1277.
273. Fong DS, Ferris FL, III, Davis MD, Chew EY. Early Treatment Diabetic Retinopathy Study Research
Group. Causes of severe visual loss in the early treatment diabetic retinopathy study: ETDRS report no.
24. Am J Ophthalmol. 1999;127(2):137-141.
274. Sivaprasad S, Crosby-Nwaobi R, Heng LZ, Peto T, Michaelides M, Hykin P. Injection frequency and
response to bevacizumab monotherapy for diabetic macular oedema (BOLT report 5). Br J
Ophthalmol. 2013;97(9):1177-1180.
275. Turner RC. The U.K. Prospective Diabetes Study. A review. Diabetes Care. 1998;21 Suppl 3:C35-38.

73
P143
 Diabetic Retinopathy PPP

276. Nathan DM, Bayless M, Cleary P, et al. DCCT/EDIC Research Group. Diabetes control and
complications trial/epidemiology of diabetes interventions and complications study at 30 years:
advances and contributions. Diabetes. 2013;62(12):3976-3986.
277. Ismail-Beigi F, Craven T, Banerji MA, et al. ACCORD Trial Group. Effect of intensive treatment of
hyperglycaemia on microvascular outcomes in type 2 diabetes: an analysis of the ACCORD
randomised trial. Lancet. 2010;376(9739):419-430. Erratum in: Lancet 2010;2376:1466.
278. Bressler SB, Qin H, Melia M, et al. Diabetic Retinopathy Clinical Research Network. Exploratory
analysis of the effect of intravitreal ranibizumab or triamcinolone on worsening of diabetic retinopathy
in a randomized clinical trial. JAMA Ophthalmol. 2013;131(8):1033-1040.
279. Diabetic Retinopathy Clinical Research Network Authors/Writing Committee. Macular edema after
cataract surgery in eyes without preoperative central-involved diabetic macular edema. JAMA
Ophthalmol. 2013;131(7):870-879.
280. Benson WE, Morse PH, Nantawan P. Late complications following cryotherapy of lattice degeneration.
Am J Ophthalmol. 1977;84(4):514-516.
281. Fontenot JL, Bona MD, Kaleem MA, et al. Vision Rehabilitation Preferred Practice Pattern®.
Ophthalmology. 2018;125(1):P228-P278.
282. Stelmack JA, Tang XC, Reda DJ, Rinne S, Mancil RM, Massof RW. LOVIT Study Group. Outcomes
of the Veterans Affairs Low Vision Intervention Trial (LOVIT). Arch Ophthalmol. 2008;126(5):608-
617.
283. Javitt JC, Canner JK, Sommer A. Cost effectiveness of current approaches to the control of retinopathy
in type I diabetics. Ophthalmology. 1989;96(2):255-264.
284. Javitt JC, Aiello LP, Bassi LJ, Chiang YP, Canner JK. Detecting and treating retinopathy in patients
with type I diabetes mellitus. Savings associated with improved implementation of current guidelines.
American Academy of Ophthalmology. Ophthalmology. 1991;98(10):1565-1573; discussion 1574.
285. Javitt JC, Aiello LP. Cost-effectiveness of detecting and treating diabetic retinopathy. Ann Intern Med.
1996;124(1 Pt 2):164-169.
286. Stein JD, Newman-Casey PA, Kim DD, Nwanyanwu KH, Johnson MW, Hutton DW. Cost-
effectiveness of various interventions for newly diagnosed diabetic macular edema. Ophthalmology.
2013;120(9):1835-1842.
287. Sharma S, Brown GC, Brown MM, Hollands H, Shah GK. The cost-effectiveness of grid laser
photocoagulation for the treatment of diabetic macular edema: results of a patient-based cost-utility
analysis. Curr Opin Ophthalmol. 2000;11(3):175-179.
288. Busbee BG, Brown MM, Brown GC, Sharma S. CME review: a cost-utility analysis of laser
photocoagulation for extrafoveal choroidal neovascularization. Retina. 2003;23(3):279-287.
289. Crijns H, Casparie AF, Hendrikse F. Continuous computer simulation analysis of the cost-effectiveness
of screening and treating diabetic retinopathy. Int J Technol Assess Health Care. 1999;15(1):198-206.
290. Ho T, Smiddy WE, Flynn HW, Jr. Vitrectomy in the management of diabetic eye disease. Surv
Ophthalmol. 1992;37(3):190-202.
291. Writing Committee for the Diabetic Retinopathy Clinical Research Network, Gross JG, Glassman AR,
et al. Panretinal photocoagulation vs intravitreous ranibizumab for proliferative diabetic retinopathy: a
randomized clinical trial. JAMA. 2015;314(20):2137-2146.
292. Heier JS, Korobelnik JF, Brown DM, et al. Intravitreal Aflibercept for Diabetic Macular Edema: 148-
Week Results from the VISTA and VIVID Studies. Ophthalmology. 2016;123(11):2376-2385.
293. Klein R. Hyperglycemia and microvascular and macrovascular disease in diabetes. Diabetes Care.
1995;18(2):258-268.
294. Diabetes Control and Complications Trial Research Group. Early worsening of diabetic retinopathy in
the Diabetes Control and Complications Trial. Arch Ophthalmol. 1998;116(7):874-886.
295. ACCORD Study Group. Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial: design
and methods. Am J Cardiol. 2007;99(12A):21i-33i.
296. Action to Control Cardiovascular Risk in Diabetes Study Group. Effects of intensive glucose lowering
in type 2 diabetes. N Engl J Med. 2008;358(24):2545-2559.
297. ACCORD Study Group. Effects of combination lipid therapy in type 2 diabetes mellitus. N Engl J Med.
2010;362(17):1563-1574. Erratum in: N Engl J Med 2010;1362:1748.
298. ACCORD Study Group. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl
J Med. 2010;362(17):1575-1585.

74 P144
 Diabetic Retinopathy PPP

299. Qaseem A, Wilt TJ, Kansagara D, et al. Hemoglobin A1c targets for glycemic control with
pharmacologic therapy for nonpregnant adults with type 2 diabetes mellitus: a guidance statement
update from the american college of physicians. Ann Intern Med. 2018;168(8):569-576.

P145
75
over

Posterior Vitreous
Detachment, Retinal
Breaks, and Lattice
Degeneration Preferred
Practice Pattern®

© 2019 by the American Academy of Ophthalmology http://dx.doi.org/10.1016/j.ophtha.2019.09.027

Published by Elsevier Inc. ISSN 0161-6420/19
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Secretary for Quality of Care

Timothy W. Olsen, MD

Academy Staff
Ali Al-Rajhi, PhD, MPH
Andre Ambrus, MLIS
Meghan Daly
Flora C. Lum, MD

Medical Editor: Susan Garratt

Approved by: Board of Trustees

September 7, 2019

© 2019 American Academy of Ophthalmology®

All rights reserved

AMERICAN ACADEMY OF OPHTHALMOLOGY and PREFERRED PRACTICE PATTERN are

registered trademarks of the American Academy of Ophthalmology. All other trademarks are the property of
their respective owners.

Preferred Practice Pattern® guidelines are developed by the Academy’s H. Dunbar Hoskins Jr., MD Center
for Quality Eye Care without any external financial support. Authors and reviewers of the guidelines are
volunteers and do not receive any financial compensation for their contributions to the documents. The
guidelines are externally reviewed by experts and stakeholders before publication.

Correspondence:
Ali A. Al-Rajhi, PhD, MPH American Academy of Ophthalmology, P. O. Box 7424, San Francisco, CA
94120-7424. E-mail: [email protected].

2
P224
 PVD, Retinal Breaks, and Lattice Degeneration PPP

The Retina/Vitreous Preferred Practice Pattern® Panel members wrote the Posterior Vitreous Detachment,
Retinal Breaks, and Lattice Degeneration Preferred Practice Pattern® (PPP) guidelines. The PPP Panel
members discussed and reviewed successive drafts of the document, meeting in person twice and conducting
other review by e-mail discussion, to develop a consensus over the final version of the document.

Retina/Vitreous Preferred Practice Pattern Panel 2018–2019

Ron A. Adelman, MD, MPH, MBA, FACS
Jennifer I. Lim, MD
Steven T. Bailey, MD, Retina Society Representative
Amani Fawzi, MD, Macula Society Representative
Gurunadh A. Vemulakonda, MD, American Society of Retina Specialists Representative
Gui-shang Ying, MD, PhD, Methodologist
Christina J. Flaxel, MD, Chair

We thank our partners, the Cochrane Eyes and Vision US Satellite (CEV@US), for identifying reliable systematic
reviews that we cite and discuss in support of the PPP recommendations.

The Preferred Practice Patterns Committee members reviewed and discussed the document during a meeting
in June 2019. The document was edited in response to the discussion and comments.

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair
Roy S. Chuck, MD, PhD
Steven P. Dunn, MD
Christina J. Flaxel, MD
Steven J. Gedde, MD
Francis S. Mah, MD
Randall J. Olson, MD
David K. Wallace, MD, MPH
David C. Musch, PhD, MPH, Methodologist

The Posterior Vitreous Detachment, Retinal Breaks, and Lattice Degeneration PPP was then sent for review to
additional internal and external groups and individuals in July 2019. All those returning comments were required to
provide disclosure of relevant relationships with industry to have their comments considered (indicated with an
asterisk below). Members of the Retina/Vitreous Preferred Practice Pattern Panel reviewed and discussed these
comments and determined revisions to the document.

P2253
 PVD, Retinal Breaks, and Lattice Degeneration PPP

In compliance with the Council of Medical Specialty Societies’ Code for Interactions with Companies
(available at www.cmss.org/codeforinteractions.aspx), relevant relationships with industry are listed. The
Academy has Relationship with Industry Procedures to comply with the Code (available at
http://one.aao.org/CE/PracticeGuidelines/PPP.aspx). A majority (88%) of the members of the Retina/Vitreous
Preferred Practice Pattern Panel 2018–2019 had no financial relationship to disclose.

Retina/Vitreous Preferred Practice Pattern Panel 2018–2019

Christina J. Flaxel, MD: No financial relationships to disclose
Ron A. Adelman, MD, MPH, MBA, FACS: No financial relationships to disclose
Steven T. Bailey, MD: No financial relationships to disclose
Amani Fawzi, MD: No financial relationships to disclose
Jennifer I. Lim, MD: Alcon Laboratories—Consultant/Advisor
Gurunadh A. Vemulakonda, MD: No financial relationships to disclose
Gui-shang Ying, MD, PhD: No financial relationships to disclose

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair: No financial relationships to disclose
Roy S. Chuck, MD, PhD: No financial relationships to disclose
Steven P. Dunn, MD: No financial relationships to disclose
Christina J. Flaxel, MD: No financial relationships to disclose
Steven J. Gedde, MD: No financial relationships to disclose
Francis S. Mah, MD: Alcon Laboratories—Consultant/Advisor
Randall J. Olson, MD: No financial relationships to disclose
David K. Wallace, MD, MPH: No financial relationships to disclose
David C. Musch, PhD, MPH, Methodologist: IRIDEX, Notal Vision—Consultant/Advisor

Secretary for Quality of Care

Timothy W. Olsen, MD: No financial relationships to disclose

Academy Staff
Ali Al-Rajhi, PhD, MPH: No financial relationships to disclose
Andre Ambrus, MLIS: No financial relationships to disclose
Meghan Daly: No financial relationships to disclose
Flora C. Lum, MD: No financial relationships to disclose

The disclosures of relevant relationships to industry of other reviewers of the document from January to
October 2019 are available online at www.aao.org/ppp.

4P226
 PVD, Retinal Breaks, and Lattice Degeneration PPP

OBJECTIVES OF PREFERRED PRACTICE PATTERN GUIDELINES ...........................................P229

METHODS AND KEY TO RATINGS ......................................................................................................P230
HIGHLIGHTED FINDINGS AND RECOMMENDATIONS FOR CARE ...........................................P231
INTRODUCTION .......................................................................................................................................P232
Disease Definition..........................................................................................................................................P232
Patient Population ..........................................................................................................................................P232
Clinical Objectives .........................................................................................................................................P232
BACKGROUND ..........................................................................................................................................P233
Posterior Vitreous Detachment ......................................................................................................................P233
Evolution of Retinal Breaks and Lattice Degeneration ..................................................................................P234
Asymptomatic Retinal Breaks...............................................................................................................P234
Symptomatic Retinal Breaks .................................................................................................................P234
Lattice Degeneration .............................................................................................................................P235
Incidence of Rhegmatogenous Retinal Detachment ......................................................................................P235
Risk Factors for Rhegmatogenous Retinal Detachment ................................................................................P235
Myopia ..................................................................................................................................................P236
Lattice Degeneration .............................................................................................................................P236
Cataract Surgery....................................................................................................................................P236
Trauma ..................................................................................................................................................P236
Rhegmatogenous Retinal Detachment in the Fellow Eye .....................................................................P236
Other Risk Factors ................................................................................................................................P237
CARE PROCESS .........................................................................................................................................P237
Patient Outcome Criteria................................................................................................................................P237
Diagnosis .......................................................................................................................................................P237
History ..................................................................................................................................................P237
Ophthalmic Examination ......................................................................................................................P238
Diagnostic Tests ....................................................................................................................................P238
Management...................................................................................................................................................P239
Prevention .............................................................................................................................................P239
Surgical Management ...........................................................................................................................P239
Complications of Treatment..................................................................................................................P242
Follow-up Evaluation ............................................................................................................................P242
Counseling and Referral ................................................................................................................................P244
Socioeconomic Considerations ......................................................................................................................P245
APPENDIX 1. QUALITY OF OPHTHALMIC CARE CORE CRITERIA ..........................................P246
APPENDIX 2. INTERNATIONAL STATISTICAL CLASSIFICATION OF DISEASES AND
RELATED HEALTH PROBLEMS (ICD) CODES ........................................................................P248
GLOSSARY .................................................................................................................................................P249
LITERATURE SEARCHES FOR THIS PPP...........................................................................................P251
RELATED ACADEMY MATERIALS .....................................................................................................P253
REFERENCES.............................................................................................................................................P254

P2275
 PVD, Retinal Breaks, and Lattice Degeneration PPP

PVD, Retinal Breaks, and Lattice Degeneration PPP

Background:
Posterior vitreous detachment (PVD) is a separation of the posterior vitreous cortex from the
internal limiting membrane of the retina. Retinal breaks are defined as full-thickness defects in
the retina. Lattice degeneration is a peripheral vitreoretinal condition characterized by retinal
thinning, overlying vitreous liquefaction, and firm vitreoretinal adhesions at the margins of
thinning.
A PVD typically occurs between the ages of 45 and 65 in the general population with earlier
onset in men; however, the posterior vitreous may detach earlier in trauma and myopia. PVD
may lead to vitreomacular traction (VMT), potentially leading to mechanical distortion of the
macula or to retinal tears and potentially retinal detachment. The recommendations of this
Preferred Practice Pattern (PPP) are based on Cochrane-identified reliable systematic reviews.
Rationale for treatment:
The patient outcome criteria and rational for treatment for PVD and RRD include the prevention
of visual loss and functional impairment and maintenance of quality of life. If patients are
familiar with the symptoms of retinal tears or detachment, they are recommended to report
promptly, thus improving the opportunity for successful treatment.

Care Process:
The care process for a patient starts with all the elements of a comprehensive adult eye
evaluation with attention to risk factors relevant for a PVD, retinal breaks, and lattice
degeneration. Additionally, a PVD should be evaluated for vitreous pigmented cells and includes
a thorough peripheral examination looking for retinal tears or holes.

Management recommendations are detailed in this PPP, which includes surgical management,
complications of treatment and follow-up evaluation. At the time of this published PPP, there are
no effective methods of preventing the factors that lead to PVD. All patients at increased risk of
retinal detachment should be instructed to notify their ophthalmologist as soon as possible if they
have a substantial change in symptoms (e.g., increased floaters, loss of visual field, or decrease
in visual acuity). Patients who undergo refractive surgery to reduce myopia should be informed
that they remain at risk of RRD despite reduction of their refractive error.

P228

P228
 PVD, Retinal Breaks, and Lattice Degeneration PPP

As a service to its members and the public, the American Academy of Ophthalmology has developed a series
of Preferred Practice Pattern® guidelines that identify characteristics and components of quality eye care.
Appendix 1 describes the core criteria of quality eye care.

The Preferred Practice Pattern® guidelines are based on the best available scientific data as interpreted by
panels of knowledgeable health professionals. In some instances, such as when results of carefully conducted
clinical trials are available, the data are particularly persuasive and provide clear guidance. In other instances,
the panels have to rely on their collective judgment and evaluation of available evidence.

These documents provide guidance for the pattern of practice, not for the care of a particular
individual. While they should generally meet the needs of most patients, they cannot possibly best meet the
needs of all patients. Adherence to these PPPs will not ensure a successful outcome in every situation. These
practice patterns should not be deemed inclusive of all proper methods of care or exclusive of other methods
of care reasonably directed at obtaining the best results. It may be necessary to approach different patients’
needs in different ways. The physician must make the ultimate judgment about the propriety of the care of a
particular patient in light of all of the circumstances presented by that patient. The American Academy of
Ophthalmology is available to assist members in resolving ethical dilemmas that arise in the course of
ophthalmic practice.

Preferred Practice Pattern® guidelines are not medical standards to be adhered to in all individual
situations. The Academy specifically disclaims any and all liability for injury or other damages of any kind,
from negligence or otherwise, for any and all claims that may arise out of the use of any recommendations or
other information contained herein.

References to certain drugs, instruments, and other products are made for illustrative purposes only and are
not intended to constitute an endorsement of such. Such material may include information on applications
that are not considered community standard, that reflect indications not included in approved U.S. Food and
Drug Administration (FDA) labeling, or that are approved for use only in restricted research settings. The
FDA has stated that it is the responsibility of the physician to determine the FDA status of each drug or
device he or she wishes to use, and to use them with appropriate patient consent in compliance with
applicable law.

Innovation in medicine is essential to ensure the future health of the American public, and the Academy
encourages the development of new diagnostic and therapeutic methods that will improve eye care. It is
essential to recognize that true medical excellence is achieved only when the patients’ needs are the foremost
consideration.

All Preferred Practice Pattern® guidelines are reviewed by their parent panel annually or earlier if
developments warrant and updated accordingly. To ensure that all PPPs are current, each is valid for 5 years
from the approved by date unless superseded by a revision. Preferred Practice Pattern guidelines are funded
by the Academy without commercial support. Authors and reviewers of PPPs are volunteers and do not
receive any financial compensation for their contributions to the documents. The PPPs are externally
reviewed by experts and stakeholders, including consumer representatives, before publication. The PPPs are
developed in compliance with the Council of Medical Specialty Societies’ Code for Interactions with
Companies. The Academy has Relationship with Industry Procedures (available at www.aao.org/about-
preferred-practice-patterns) to comply with the Code.

Appendix 2 contains the International Statistical Classification of Diseases and Related Health Problems
(ICD) codes for the disease entities that this PPP covers. The intended users of the Posterior Vitreous
Detachment, Retinal Breaks, and Lattice Degeneration PPP are ophthalmologists.

P2296
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Preferred Practice Pattern® guidelines should be clinically relevant and specific enough to provide
useful information to practitioners. Where evidence exists to support a recommendation for care, the
recommendation should be given an explicit rating that shows the strength of evidence. To accomplish
these aims, methods from the Scottish Intercollegiate Guideline Network1 (SIGN) and the Grading of
Recommendations Assessment, Development and Evaluation2 (GRADE) group are used. GRADE is a
systematic approach to grading the strength of the total body of evidence that is available to support
recommendations on a specific clinical management issue. Organizations that have adopted GRADE
include SIGN, the World Health Organization, the Agency for Healthcare Research and Policy, and the
American College of Physicians.3
◆ All studies used to form a recommendation for care are graded for strength of evidence individually, and
that grade is listed with the study citation.
◆ To rate individual studies, a scale based on SIGN1 is used. The definitions and levels of evidence to rate
individual studies are as follows:
I++ High-quality meta-analyses, systematic reviews of randomized controlled trials (RCTs), or
RCTs with a very low risk of bias
I+ Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias
I- Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias
II++ High-quality systematic reviews of case-control or cohort studies
High-quality case-control or cohort studies with a very low risk of confounding or bias and a
high probability that the relationship is causal
II+ Well-conducted case-control or cohort studies with a low risk of confounding or bias and a
moderate probability that the relationship is causal
II- Case-control or cohort studies with a high risk of confounding or bias and a significant risk that
the relationship is not causal
III Nonanalytic studies (e.g., case reports, case series)

◆ Recommendations for care are formed based on the body of the evidence. The body of evidence quality
ratings are defined by GRADE2 as follows:
Good quality Further research is very unlikely to change our confidence in the estimate of
effect
Moderate quality Further research is likely to have an important impact on our confidence in the
estimate of effect and may change the estimate
Insufficient quality Further research is very likely to have an important impact on our confidence in
the estimate of effect and is likely to change the estimate
Any estimate of effect is very uncertain

◆ Key recommendations for care are defined by GRADE2 as follows:

Strong Used when the desirable effects of an intervention clearly outweigh the
recommendation undesirable effects or clearly do not
Discretionary Used when the trade-offs are less certain—either because of low-quality evidence
recommendation or because evidence suggests that desirable and undesirable effects are closely
balanced

◆ The Highlighted Findings and Recommendations for Care section lists points determined by the PPP
Panel to be of particular importance to vision and quality of life outcomes.
◆ All recommendations for care in this PPP were rated using the system described above. Ratings are embedded
throughout the PPP main text in italics.
◆ Literature searches to update the PPP were undertaken in April 2018 and June 2019 in PubMed and the
Cochrane Library. Complete details of the literature searches are available online at www.aao.org/ppp.

7P230
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Acute horseshoe retinal tears and traumatic breaks usually require treatment.

Asymptomatic atrophic or operculated retinal breaks rarely need treatment. More generally, an eye that has
atrophic round holes within lattice lesions, has minimal subretinal fluid without progression, or lacks
evidence of posterior vitreous detachment (PVD) does not require treatment.

An early diagnosis of a retinal detachment is important because the rate of successful retinal reattachment is
higher and the visual results are better when repaired early, especially before the rhegmatogenous retinal
detachment (RRD) involves the macula.

Lattice degeneration is present in 6% to 8% of the population and increases the risk of retinal detachment.

Patients presenting with an acute PVD and no retinal breaks have a small chance (~2%) of developing retinal
breaks in the weeks that follow. Selected patients, particularly those with any degree of vitreous pigment,
vitreous or retinal hemorrhage, or visible vitreoretinal traction, should be asked to return for a second
examination promptly if they have new symptoms or within 6 weeks following the onset of PVD symptoms.

Between 5% and 14% of patients found to have an initial retinal break will develop additional breaks during
long-term follow-up. Cataract surgery is a risk factor for new retinal breaks.

Treatment of peripheral horseshoe tears should extend to the ora serrata if the tear cannot be surrounded
using laser or cryotherapy. The most common cause of failure is inadequate treatment, particularly along the
anterior border (where visualization is more difficult).

8 P231
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Posterior vitreous detachment (PVD) is a separation of the posterior vitreous cortex from the internal
limiting membrane of the retina.4 (See Glossary.) This separation may be complete or partial. Vitreous
traction at sites of significant vitreoretinal adhesion is responsible for most retinal breaks that lead to
retinal detachment. Retinal breaks are defined as full-thickness defects in the retina. Lattice
degeneration is a vitreoretinal degenerative process that predisposes to retinal tears and detachment. It
is a peripheral vitreoretinal condition characterized by retinal thinning, overlying vitreous
liquefaction, and firm vitreoretinal adhesions at the margins of thinning. Most lattice degenerations
are ovoid, with the long axes of lattice running parallel to the ora serrata. Perivascular lattice occurs
radially and is typically found adjacent to the retinal vessels. Radial lattice is associated with a much
higher risk of retinal detachment than circumferential lattice. Round holes occur frequently within
areas of lattice degeneration. Vitreomacular traction (VMT) may develop when the vitreous partially
separates from the macula, potentially leading to mechanical distortion of the macula that may
correspond to visual symptoms.4 (See Glossary.)

Individuals may present with symptoms or signs suggestive of PVD, retinal breaks, vitreous
hemorrhage, retinal detachment, or VMT. A PVD typically occurs between the ages of 45 and 65 in
the general population; however, the posterior vitreous may detach earlier in trauma and myopia.5
Other individuals may not be symptomatic and, based on clinical examination findings, may have an
increased risk of retinal detachment as the vitreous separates.

◆ Identify patients at risk of developing a rhegmatogenous retinal detachment (RRD)

◆ Examine symptomatic patients who have an acute PVD to detect and treat associated retinal breaks or
tears
◆ Recognize the evolution of retinal breaks and lattice degeneration
◆ Manage patients at high risk of developing retinal detachment
◆ Educate high-risk patients about symptoms of PVD, retinal breaks, and retinal detachments as well as
the need for periodic follow-up
◆ Discuss treatment options available for VMT (See Idiopathic Epiretinal Membrane and Vitreomacular
Traction PPP)6
◆ Recognize the potential side effects of treatment of VMT (See Idiopathic Epiretinal Membrane and
Vitreomacular Traction PPP)6

9 P232
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Population-based studies that evaluate incidence and prevalence of PVD are difficult to conduct
owing to the lack of definite clinical signs and unreliable clinical tests. A PVD typically occurs
between the ages of 45 and 65 in the general population with earlier onset in men than women;
however, the posterior vitreous may detach earlier in trauma and myopia, or be precipitated by
ophthalmic surgical procedures.5,7-9 Posterior vitreous detachment leads to vitreous traction at the
vitreous base and in areas of lattice degeneration, and thereby, secondarily, is thought to cause most
symptomatic retinal breaks that may lead to an RRD. The symptoms of a PVD include light flashes
(photopsias) and floaters (myodesopias), and patients with such symptoms are at a higher risk for
retinal detachment.10-14 The stages of a PVD are described in Table 1.4 Patients typically report the
light flashes characteristic of a PVD as being most noticeable in the dark. Such photopsias are likely
the result of vitreous traction on the retina as the vitreous separates from the posterior retina toward
the vitreous base. The floaters may be due to blood from a torn or avulsed retinal vessel,
condensations of vitreous collagen, or the epipapillary glial tissue (Weiss ring) that is torn from the
optic nerve head and area adjacent to the optic nerve head. Between 8% and 22% of patients with
acute PVD symptoms have a retinal tear at the time of the initial examination.15-18 There is a direct
correlation between the amount of vitreous hemorrhage and the likelihood of a retinal tear.19 Patients
with an acute PVD who have no reported retinal breaks on presentation have a 2% to 5% chance of
experiencing a detected (missed or new) break in the weeks that follow.13,16,20

Approximately 80% of patients who presented without detected breaks, and then had breaks occur
subsequently, had either pigmented cells or hemorrhage in the vitreous or retina at the initial
evaluation, or new symptoms that prompted a return visit to the ophthalmologist.16

A spontaneous vitreous hemorrhage can be the presenting sign of PVD or may occur during the
evolution of the PVD. Two-thirds of patients who present with associated vitreous hemorrhage were
found to have at least one break. In this subgroup, one-third had more than one break and
approximately 88% of the breaks occurred in the superior quadrants.22

10 P233
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Precursors to RRDs are PVD, asymptomatic retinal breaks, symptomatic retinal breaks, lattice
degeneration, and cystic and zonular traction retinal tufts. (See Glossary.) Because spontaneous retinal
reattachment is rare, nearly all patients with a symptomatic clinical RRD will progressively lose
vision unless the detachment is repaired. Currently, more than 95% of uncomplicated RRDs can be
successfully repaired, although more than one procedure may be required.23 The prophylactic
treatment of high-risk breaks usually prevents RRD. An early diagnosis of an RRD is also important
because the rate of successful reattachment is higher and the visual results are better when repaired
early and especially before the RRD involves the macula.15,17 The goal of RRD treatment is to allow
patients to maintain their abilities to read, work, drive, care for themselves, and maintain their quality
of life.18

Asymptomatic operculated holes and atrophic round holes rarely lead to retinal detachment.
Byer followed 46 asymptomatic eyes with operculated retinal breaks over an average of 11
years.24 Davis followed 28 eyes for up to 5 years in subjects where 80% of the fellow eyes had
a retinal detachment.25,26 All combined, none of the 74 eyes from these studies progressed to
retinal detachment during the follow-up period.

Eyes with signs and symptoms of acute PVD may have atrophic retinal breaks with clinical
features, suggesting that they are unrelated to the acute vitreoretinal traction from the PVD.
Such breaks are considered to be pre-existing rather than symptomatic. Treatment may be
considered for these breaks in certain situations, although the literature provides little
guidance.26 Randomized clinical trials are not available; therefore, there is limited evidence to
support prophylactic therapy.26

Approximately 5% of eyes with asymptomatic horseshoe tears progress to retinal

detachment.24,27,28 Horseshoe tears discovered in asymptomatic fellow eyes are less likely than
symptomatic horseshoe tears to lead to clinical retinal detachment (See Glossary). Patients
should be encouraged to have follow-up. (See Follow-up Evaluation under Surgical
Management.)

A symptomatic retinal break is defined as a break caused by vitreoretinal traction in a patient

with a new PVD associated with new-onset flashes and/or floaters. At least half of untreated
symptomatic retinal breaks with persistent vitreoretinal traction (horseshoe or flap tears) will
lead to a clinical retinal detachment unless treatment is applied.25,29,30 (See Glossary.) Treatment
by prompt creation of a chorioretinal adhesion around these symptomatic tears reduces the risk
of retinal detachment to less than 5%.29-34 Traumatic dialyses and tears along the vitreous base
are managed similarly to symptomatic tears. Symptomatic operculated breaks usually do not

11 P234
 PVD, Retinal Breaks, and Lattice Degeneration PPP

progress to a clinical retinal detachment unless the vitreous remains adherent to the retina
surrounding the break.25,30

Generally, atrophic round holes within lattice lesions that are accompanied by minimal
subretinal fluid and no PVD do not require treatment. However, lattice degeneration is a risk
factor for developing an RRD either from round holes without PVD or tractional-related holes
associated with PVD. Small asymptomatic peripheral retinal detachments occurring secondary
to retinal holes in areas of lattice degeneration are termed subclinical detachments (see
Glossary).35,36 Although these can enlarge and progress to clinical retinal detachments, they
have been shown by observation alone to have a low likelihood of progression in most
patients.35 Prophylactic or interventional treatment should be considered when the detachments
are documented to become symptomatic, increase in size, or show other signs of
progression.27,37

One analysis studied 423 eyes with lattice degeneration in 276 patients over a period averaging
nearly 11 years.37 Of these, 150 eyes (35%) had atrophic holes in lattice, and 10 of these 150
eyes had subretinal fluid extending more than 1 disc diameter from the break (subclinical retinal
detachment). Six other eyes developed new subclinical retinal detachments during follow-up.
Clinical retinal detachments developed in 3 of the 423 eyes. Two were due to round retinal
holes in lattice lesions of patients in their mid-20s and one was due to a symptomatic tractional
tear. These data indicate that patients with lattice degeneration with or without round holes are
at a very low risk for progression to clinical retinal detachment without a previous RRD in the
fellow eye.

More commonly, RRD occurs in eyes with lattice degeneration when a PVD induces a
horseshoe tear. Such tears should be treated using either laser demarcation or cryotherapy.27,37

The annual incidence of RRD is approximately 10 to 18 per 100,000 persons.38-40 Of these, 20% to
40% have had cataract surgery and 10% have had ocular trauma.23,41,42 In a recent study from the
Netherlands, the annual RRD incidence was 18 per 100,000 people (95% CI, 11–19), with a peak
incidence of 53 per 100,000 people (95% CI, 29–57) between 55 and 59 years of age. The rate of
bilateral RRD was 1.7%. Prior cataract surgery was reported in 34% of RRD eyes.40

Aside from retinal breaks, risk factors for RRD include myopia, lattice degeneration, cataract or other
intraocular surgery, neodymium yttrium-aluminum-garnet (Nd:YAG) laser surgery, trauma, a history
of RRD in the other eye, certain genetic disorders such as Stickler syndrome, or family history of
retinal detachments in a first-degree relative. Combinations of these factors may increase the risk.

12 P235
 PVD, Retinal Breaks, and Lattice Degeneration PPP

More than half of nontraumatic RRD occurs in myopic eyes.43 As axial length increases, the
risk of RRD increases proportionately. One study found that individuals with low myopia (1–3
diopters) have a fourfold risk of RRD,43 and higher levels of myopia have higher risks
compared with nonmyopic individuals.43,44

Lattice degeneration is present in 6% to 8% of the population and increases the risk of retinal
detachment.37,45 Approximately 20% to 30% of patients with RRD have lattice degeneration.37
Perivascular or radial lattice is associated with a higher risk of retinal tear or detachment
formation.37 Perivascular lattice it is also frequently seen in Stickler syndrome.46

The overall risk of RRD after cataract surgery is approximately 1%.47-50 The following
conditions have been reported to increase the risk of RRD after cataract surgery: axial myopia,
pre-existing vitreoretinal disease, male gender, younger age, vitreous prolapse into the anterior
chamber, vitreous loss (ruptured posterior capsule/zonules), and spontaneous extension of the
capsulotomy at the time of surgery.51,52 One study suggests that in the absence of a posterior
capsular tear at the time of cataract surgery, subsequent Nd:YAG laser capsulotomy may not
increase the risk of retinal detachment.53 Other studies suggest that Nd:YAG laser capsulotomy
is associated with a fourfold increase in the risk of RRD, especially in myopic patients.41,42,54-61

Outside of complications at the time of surgery, risk of RRD after cataract surgery usually
occurs 1 to 2 years later. A 5-year study using B-scan ultrasonography reported that it was the
postoperative onset of a PVD that was the major risk factor for RRD (not the presence or
absence of lattice) after cataract surgery and that the majority of eyes after cataract surgery that
did not have a pre-existing PVD developed one at a median of 7 months after surgery.
Consequently, one can extrapolate that it is the absence of a PVD (in higher risk eyes such as
those with myopia and lattice) at the time of cataract surgery that is the major risk of RRD
later.60

Patients with blunt or penetrating ocular injuries that have altered the structure of the vitreous or
retina are at increased risk of RRD.62 Vitreoretinal interface changes caused by trauma may be
detected at the time of injury or even many years later.

Patients with a history of nontraumatic detachment in one eye have a 10% increased risk of
developing RRD in the fellow eye, since pathologic vitreoretinal changes are frequently
bilateral.27,39,63-65 The fellow eye in a patient with pseudophakic retinal detachment is also at
higher risk of developing a retinal detachment, whether the fellow eye is phakic or

13 P236
 PVD, Retinal Breaks, and Lattice Degeneration PPP

pseudophakic. Phakic fellow eyes in patients with pseudophakic retinal detachment have a 7%
risk of RRD, suggesting that the risk of developing RRD should not be attributed to cataract
surgery alone.66

Other risk factors that have been reported include prior retinopathy of prematurity67 and Stickler
syndrome.68,69

There are case reports of retinal detachment in patients who have had keratorefractive surgery;
however, large studies have not shown an increased risk in patients when compared with eyes
of a similar refractive error.70,71 It remains possible that the risk of vitreoretinal pathology is
different among particular keratorefractive techniques.9 Retinal detachment following refractive
lens exchange in patients with high myopia showed a cumulative increase from 2% to 8% over
a 7-year incidence.72 Phakic intraocular lenses have not been associated with increased risk of
retinal detachment compared with other intraocular interventions in highly myopic
patients.71,73,74

For management and treatment for PVD and RRD, the following outcomes are important:

◆ Prevention of visual loss and functional impairment

◆ Maintenance of quality of life

The initial evaluation of a patient with risk factors for retinal detachment or symptoms of a PVD
involves detection of vitreous pigment cells or debris and includes a thorough peripheral examination
looking for retinal tears or holes. It also includes all aspects of the comprehensive adult medical eye
evaluation,75 with particular attention to those aspects relevant to PVD, retinal breaks, and lattice
degeneration. The ophthalmologist should also consider other causes of vitreous cells or debris (e.g.,
uveitis, infection, inflammation, neoplasia).

A patient history should include the following elements:

◆ Symptoms of PVD10-14
◆ Family history of retinal detachment, genetic disorders (e.g., Stickler syndrome)68,69
◆ Prior eye trauma62
◆ Myopia43,76
◆ History of ocular surgery, including refractive lens exchange and cataract surgery41,42,64,77-79

14 P237
 PVD, Retinal Breaks, and Lattice Degeneration PPP

◆ History of YAG laser capsulotomy

◆ History of an intravitreal injection80

The eye examination should include the following elements:

◆ Confrontation visual field examination

◆ Visual acuity testing
◆ Pupillary assessment for the presence of a relative afferent pupillary defect
◆ Examination of the vitreous for hemorrhage, detachment, and pigmented cells10-14,16,81
◆ Careful examination of the peripheral fundus using scleral depression82 (see Table 3)

There are no symptoms that can reliably distinguish between a PVD with or without an
associated retinal break; therefore, a peripheral retinal examination is required.82 The preferred
method of evaluating patients for peripheral vitreoretinal pathology is to use an indirect
ophthalmoscope combined with scleral depression.83 Many patients with retinal tears have
blood and pigmented cells in the anterior vitreous. In fully dilated eyes, slit-lamp
biomicroscopy with a mirrored contact lens or a condensing lens is an alternative method in
fully dilated eyes instead of a scleral depressed indirect examination of the peripheral retina.

Optical coherence tomography may be helpful to evaluate and stage the PVD.4,21,84 If media
opacity or patient cooperation precludes an adequate examination of the peripheral retina, B-
scan ultrasonography should be performed to search for retinal tears, RRD, mass lesions, or
other causes of vitreous hemorrhage.85 Bilateral patching and/or elevation of the head while
sleeping may be used when attempting to clear the vitreous hemorrhage.86 If no abnormalities
are found, frequent follow-up examinations are recommended (i.e., every 1–2 weeks initially).
Wide-field color photography can detect some peripheral retinal breaks but does not replace
careful ophthalmoscopy and may be useful in patients not able to tolerate the exam.

Even if the vitreous hemorrhage is sufficiently dense to obscure the posterior pole, the
peripheral retina frequently can be examined using indirect ophthalmoscopy and scleral
depression. Patients who present with vitreous hemorrhage sufficient to obscure all retinal
details and have a negative B-scan ultrasonographic evaluation should be followed closely.
Often, patients are seen weekly until the vitreous hemorrhage resolves or until a thorough
indirect ophthalmoscopic depressed peripheral exam can be done to rule out an underlying
retinal tear. When a retinal tear is suspected, repeat ultrasonographic examination should be
performed within 1 to 2 weeks of the initial evaluation. There is considerable variation in the
reported sensitivity (44%–100%) of B-scan ultrasonography for detecting retinal tears in cases
of PVD-associated fundus-obscuring vitreous hemorrhage.85,87-89 Early vitrectomy (usually
defined as within 7 days of presentation) for dense PVD-associated vitreous hemorrhage has
been reported to have a low rate of complications and may be considered to reduce the risk of

15 P238
 PVD, Retinal Breaks, and Lattice Degeneration PPP

vision loss occurring secondary to macula-involving retinal detachment.89-91 Prompt

intervention is indicated if there is a tear seen on ultrasonography and the vitreous cavity
precludes a view.

There are no effective methods of preventing the vitreous syneresis and liquefaction that lead
to a PVD and possibly an RRD. If factors associated with an increased risk of retinal
detachment are discovered during a routine eye examination in an asymptomatic patient, a
careful peripheral fundus examination is recommended. Patients at high risk should also be
educated about the symptoms of PVD and retinal detachment as well as about the value of
periodic follow-up examinations.14 Patients with retinal or vitreous hemorrhage have an
increased risk of multiple retinal tears.92 Moreover, a systematic review performed in 2012
found that there is also no strong evidence in the literature to support or refute the use of 360-
degree laser intervention in the fellow eyes of patients with a unilateral giant retinal tear.93

Pharmacotherapy for the management of VMT has been developed. In a placebo-controlled trial
of microplasmin (a precursor of ocriplasmin) to induce a PVD, intravitreal injection of 125
micrograms of microplasmin led to a moderate increase in the likelihood of induction and
progression of PVD (10% vs. 31%).84 Please refer to the Idiopathic ERM and VMT PPP for a
detailed discussion.6 The analysis showed that ocriplasmin was better than sham or placebo for
inducing PVD, although adverse events were more common in the treated group. In addition,
20% still needed pars plana vitrectomy within 6 months.94

Pneumatic vitreolysis is also used to induce a PVD. In a meta-analysis, pneumatic vitreolysis

approached similar release rates of pars plana vitrectomy and was more effective than
ocriplasmin by day 28.95 Complications of this procedure are typically related to PVD
formation and include retinal tear, retinal detachment, epiretinal membrane, and lamellar or
macular hole formation.94,95

It is essential that clinical personnel be familiar with the symptoms of PVD and retinal
detachment and that they recognize the need for urgent ophthalmologic evaluation of
symptomatic patients.14 Patients with symptoms of possible or suspected PVD or retinal
detachment and related disorders should be examined as soon as is feasible by an
ophthalmologist skilled in binocular indirect ophthalmoscopy and supplementary techniques.
Patients with retinal breaks or detachments should be treated by an ophthalmologist with
experience in the management of these conditions. A Cochrane systematic review found low to
very low certainty evidence indicating little or no difference between pars plana vitrectomy and
scleral buckling in anatomical and visual acuity outcomes.96

16 P239
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Posterior vitreous detachment symptoms (e.g., flashes and floaters) usually diminish over time,
sometimes requiring several months. Appropriate reassurance and precautions regarding the
symptoms of retinal detachment should be given. However, some patients may be debilitated in
the absence of tears or detachments in the retina. The impact of floaters or floater-related visual
symptoms may have an adverse effect on a person’s vision-related quality of life. Pars plana
vitrectomy is an option if symptomatic floaters are still bothersome after several months. In
fact, it has been documented using contrast sensitivity function, that patients with a PVD have a
significant reduction in contrast sensitivity function.97 Laser treatments and pharmacotherapies
have been proposed to decrease these symptoms; however, such therapies currently lack
sufficient evidence to support their use.98 Pars plana vitrectomy has been used for removal of
floaters, and improvement in contrast sensitivity function has been documented.99 In a recent
review of series that compared pars plana vitrectomy and Nd:YAG laser for floaters,100 pars
plana vitrectomy showed evidence of greater patient satisfaction compared with only moderate
resolution of symptoms following the Nd:YAG laser procedure. Another study found YAG
vitreolysis to yield greater improvement in symptoms than sham laser.101

The goal of treatment for retinal breaks is to create a firm chorioretinal adhesion in the attached
retina immediately adjacent to and surrounding the retinal tear using cryotherapy or laser
photocoagulation surgery to halt the progression of subretinal fluid from detaching the
neurosensory retina.

Treatment of peripheral horseshoe tears should be extended to the ora serrata if the tear cannot
be surrounded using laser or cryotherapy.31,102,103 The most common cause of failure in treating
horseshoe tears is failure to adequately treat the tear, particularly at the anterior border.
Continued vitreous traction may extend the tear beyond the treated area and allow fluid to
dissect through the subretinal space to cause a clinical retinal detachment.31,102,103 Treatment of
dialyses must extend over the entire length of the dialysis, reaching the ora serrata beyond each
horn or end of the dialysis.

Sufficient evidence exists to warrant treating acute, symptomatic horseshoe tears.25 There is
insufficient evidence for management of other vitreoretinal abnormalities. A Cochrane
systematic review found that in making the decision to treat other vitreoretinal abnormalities,
including lattice degeneration and asymptomatic retinal breaks, that the risks that treatment
would be unnecessary, ineffective, or harmful must be weighed against the possible benefit of
reducing the rate of subsequent retinal detachment.26 (I+, Good quality, Strong
recommendation) Table 2 summarizes recommendations for management. A Cochrane
systematic review in 2014 shows that no randomized controlled clinical trials have been
performed to support treatment of asymptomatic retinal breaks of lattice degeneration.26 There
is no level 1 evidence to support the use of prophylactic laser to areas of lattice degeneration
prior to anterior segment surgery.63,65 A primary limitation of prophylactic therapy is that

17 P240
 PVD, Retinal Breaks, and Lattice Degeneration PPP

causative breaks leading to retinal detachment often occur during a PVD in areas that appear
normal prior to the PVD.65

18 P241
 PVD, Retinal Breaks, and Lattice Degeneration PPP

The surgeon should inform the patient of the risks, benefits, and alternatives to surgery.104,105
The treating surgeon is responsible for formulating a postoperative care plan and should inform
the patient of these arrangements.104,105

Retinal detachments may occur in spite of appropriate therapy. Traction is an important

component and may pull the tear from the treated area, especially when there are larger breaks
or bridging retinal blood vessels. The laser- or cryotherapy-induced treatment adhesion
(chorioretinal scar) may not be firm or complete for up to 1 month following treatment.31,33,102
Furthermore, 10% to 16% of patients will develop additional breaks during long-term follow-
up.33,106,107 Pseudophakic patients are more likely to require retreatment or to develop new
breaks.33

Proliferation of the epiretinal membrane (ERM), or macular pucker, has been occasionally
observed following treatment for a retinal break; however, a direct cause and effect relationship
of treatment of a retinal break to ERM remains unclear, since an ERM may also occur
spontaneously following the PVD. (See Glossary.) In one long-term follow-up study, the
percentage of eyes that developed macular pucker after treatment of retinal breaks was no
greater than the percentage of eyes observed to have macular pucker before treatment.31
Therefore, the method of creating a chorioretinal adhesion may be unrelated to the incidence of
postoperative macular pucker.108

The guidelines in Table 3 are recommendations for the timing of re-evaluation in the absence of
additional symptoms. Patients with new symptoms or a change in symptoms may require more
frequent evaluation. Patients with no positive findings at the initial examination should be seen
at the intervals recommended in the Comprehensive Adult Medical Eye Evaluation PPP.75 All
patients with risk factors should be advised to contact their ophthalmologist promptly if new
symptoms such as flashes, floaters, peripheral visual field loss, or decreased visual acuity
develop.41,42,77,109

19 P242
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Younger myopic patients who have lattice degeneration with holes need regular follow-up visits
to monitor for subclinical retinal detachments that may slowly enlarge to become clinical retinal
detachments. Treatment should be considered if the detachments progress in size.27,37

Patients presenting with an acute PVD and no retinal breaks have a small chance
(approximately 2%) of developing retinal breaks in the weeks that follow.13 Thus, selected
patients, particularly those with any degree of vitreous pigment, vitreous or retinal hemorrhage,
or visible vitreoretinal traction, should be asked to return for a second examination within 6
weeks following the onset of symptoms.13,107

20 P243
 PVD, Retinal Breaks, and Lattice Degeneration PPP

A patient history should identify changes in the following:

◆ Visual symptoms10-14,81
◆ Interval history of eye trauma, intraocular injection, or intraocular surgery42,62

The eye examination should emphasize the following elements:

◆ Measurement of visual acuity

◆ Evaluation of the vitreous status, with attention to the presence of pigment, hemorrhage, or
syneresis10-14,16,81
◆ Examination of the peripheral fundus using scleral depression82 or a fundus contact or non–
contact lens using the slit-lamp biomicroscope
◆ Wide-field photography may be helpful but does not replace careful ophthalmoscopy
◆ Optical coherence tomography if VMT is present4,21,84
◆ B-scan ultrasonography when the media is opaque85

If the treatment for treated patients appears satisfactory at the first follow-up visit at 1 to 2
weeks, indirect ophthalmoscopy and scleral depression at 2 to 6 weeks will determine the
adequacy of the chorioretinal scar, especially around the anterior boundary of the tear. If
the tear and the accompanying subretinal fluid are not completely surrounded by the
chorioretinal scar, additional treatment should be administered. At any postoperative visit,
additional treatment should be considered if subretinal fluid has accumulated beyond the
edge of treatment.31,33

Even when a patient has had adequate treatment, additional examinations are important.
Between 5% and 14% of patients found to have an initial retinal break will develop
additional breaks during long-term follow-up. These statistics appear to be similar
regardless of how the initial breaks were treated.33,106 New breaks may be particularly
likely in eyes that have had cataract surgery.33

All patients at increased risk of retinal detachment should be instructed to notify their ophthalmologist
as soon as possible if they have a substantial change in symptoms, such as an increase in floaters, loss
of visual field, or decrease in visual acuity.41,42,77,109 If patients are familiar with the symptoms of
retinal tears or detachment, they may be more likely to report promptly, thus improving the
opportunity for successful treatment and subsequent visual results.17 Patients who undergo refractive
surgery to reduce myopia should be informed that they remain at risk of RRD despite reduction of
their refractive error.

21 P244
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Limited data exist on the socioeconomic impacts of PVD, retinal breaks, or lattice degeneration.
However, research on the impact of the symptoms of these conditions (e.g., vitreous floaters) has
suggested that vitreous symptoms may have an unfavorable effect on a patient’s vision-related quality
of life.110,111 The modeled cost of evaluating a patient with PVD and treating associated pathology in
the facility/hospital (nonfacility/Ambulatory Surgery Centers)-based setting was $65 to $190 ($25–
$71) depending on whether a single or two-examination protocol was used. The cost per quality-
adjusted life year (QALY) saved was $255 to $638/QALY ($100–$293/QALY). Treatment of a
symptomatic horseshoe tear resulted in a net cost savings of $1,749 ($1,314) and improved utility,
whereas treatment of an asymptomatic horseshoe tear resulted in $2,981/QALY ($1,436/QALY).
Treatment of asymptomatic lattice degeneration in an eye in which the fellow eye had a history of RD
resulted in $4,414/QALY ($2,187/QALY).112

22 P245
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Providing quality care

is the physician's foremost ethical obligation, and is
the basis of public trust in physicians.
AMA Board of Trustees, 1986

Quality ophthalmic care is provided in a manner and with the skill that is consistent with the best interests of
the patient. The discussion that follows characterizes the core elements of such care.
The ophthalmologist is first and foremost a physician. As such, the ophthalmologist demonstrates
compassion and concern for the individual, and utilizes the science and art of medicine to help alleviate
patient fear and suffering. The ophthalmologist strives to develop and maintain clinical skills at the highest
feasible level, consistent with the needs of patients, through training and continuing education. The
ophthalmologist evaluates those skills and medical knowledge in relation to the needs of the patient and
responds accordingly. The ophthalmologist also ensures that needy patients receive necessary care directly or
through referral to appropriate persons and facilities that will provide such care, and he or she supports
activities that promote health and prevent disease and disability.
The ophthalmologist recognizes that disease places patients in a disadvantaged, dependent state. The
ophthalmologist respects the dignity and integrity of his or her patients and does not exploit their
vulnerability.
Quality ophthalmic care has the following optimal attributes, among others.
◆ The essence of quality care is a meaningful partnership relationship between patient and physician. The
ophthalmologist strives to communicate effectively with his or her patients, listening carefully to their
needs and concerns. In turn, the ophthalmologist educates his or her patients about the nature and
prognosis of their condition and about proper and appropriate therapeutic modalities. This is to ensure
their meaningful participation (appropriate to their unique physical, intellectual, and emotional state) in
decisions affecting their management and care, to improve their motivation and compliance with the
agreed plan of treatment, and to help alleviate their fears and concerns.
◆ The ophthalmologist uses his or her best judgment in choosing and timing appropriate diagnostic and
therapeutic modalities as well as the frequency of evaluation and follow-up, with due regard to the
urgency and nature of the patient's condition and unique needs and desires.
◆ The ophthalmologist carries out only those procedures for which he or she is adequately trained,
experienced, and competent, or, when necessary, is assisted by someone who is, depending on the
urgency of the problem and availability and accessibility of alternative providers.
◆ Patients are assured access to, and continuity of, needed and appropriate ophthalmic care, which can be
described as follows.
 The ophthalmologist treats patients with due regard to timeliness, appropriateness, and his or her own
ability to provide such care.
 The operating ophthalmologist makes adequate provision for appropriate pre- and postoperative
patient care.
 When the ophthalmologist is unavailable for his or her patient, he or she provides appropriate alternate
ophthalmic care, with adequate mechanisms for informing patients of the existence of such care and
procedures for obtaining it.
 The ophthalmologist refers patients to other ophthalmologists and eye care providers based on the
timeliness and appropriateness of such referral, the patient's needs, the competence and qualifications
of the person to whom the referral is made, and access and availability.

P246
23
 PVD, Retinal Breaks, and Lattice Degeneration PPP

 The ophthalmologist seeks appropriate consultation with due regard to the nature of the ocular or other
medical or surgical problem. Consultants are suggested for their skill, competence, and accessibility.
They receive as complete and accurate an accounting of the problem as necessary to provide efficient
and effective advice or intervention, and in turn they respond in an adequate and timely manner. The
ophthalmologist maintains complete and accurate medical records.
 On appropriate request, the ophthalmologist provides a full and accurate rendering of the patient's
records in his or her possession.
 The ophthalmologist reviews the results of consultations and laboratory tests in a timely and effective
manner and takes appropriate actions.
 The ophthalmologist and those who assist in providing care identify themselves and their profession.
 For patients whose conditions fail to respond to treatment and for whom further treatment is
unavailable, the ophthalmologist provides proper professional support, counseling, rehabilitative and
social services, and referral as appropriate and accessible.
◆ Prior to therapeutic or invasive diagnostic procedures, the ophthalmologist becomes appropriately
conversant with the patient's condition by collecting pertinent historical information and performing
relevant preoperative examinations. Additionally, he or she enables the patient to reach a fully informed
decision by providing an accurate and truthful explanation of the diagnosis; the nature, purpose, risks,
benefits, and probability of success of the proposed treatment and of alternative treatment; and the risks
and benefits of no treatment.
◆ The ophthalmologist adopts new technology (e.g., drugs, devices, surgical techniques) in judicious
fashion, appropriate to the cost and potential benefit relative to existing alternatives and to its
demonstrated safety and efficacy.
◆ The ophthalmologist enhances the quality of care he or she provides by periodically reviewing and
assessing his or her personal performance in relation to established standards, and by revising or altering
his or her practices and techniques appropriately.
◆ The ophthalmologist improves ophthalmic care by communicating to colleagues, through appropriate
professional channels, knowledge gained through clinical research and practice. This includes alerting
colleagues of instances of unusual or unexpected rates of complications and problems related to new
drugs, devices, or procedures.
◆ The ophthalmologist provides care in suitably staffed and equipped facilities adequate to deal with
potential ocular and systemic complications requiring immediate attention.
◆ The ophthalmologist also provides ophthalmic care in a manner that is cost effective without
unacceptably compromising accepted standards of quality.

Reviewed by: Council

Approved by: Board of Trustees
October 12, 1988
2nd Printing: January 1991
3rd Printing: August 2001
4th Printing: July 2005

P247
24
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Precursors to rhegmatogenous retinal detachment and related entities with the following ICD-9 and ICD-10
classifications (see Glossary):

•
•
•

P248
25
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Atrophic retinal breaks or holes: Full-thickness retinal defects, unrelated to vitreoretinal traction. These can
occur within lattice lesions or in areas of the retina that appear otherwise normal.
Clinical retinal detachment: A retinal detachment that either impairs a portion of the visual field or extends
more than 2 disc diameters posterior to the equator.
Cystic retinal tufts: Small congenital lesions of the peripheral retina. They are slightly elevated and usually
whitish in color with variable surrounding pigmentation. They are firmly attached to the overlying vitreous
cortex and are sometimes a cause of retinal tears following PVD.
Epiretinal membrane (ERM): See Macular pucker.
ERM: See Macular pucker.
Flap tear: A horseshoe tear.
Horseshoe tear: A retinal tear caused by vitreoretinal traction on the retina. The tear is horseshoe shaped
owing to a flap of torn tissue that remains attached to the detached vitreous gel.
ICD-9: International Statistical Classification of Diseases and Related Health Problems, Ninth Edition.
ICD-10: International Statistical Classification of Diseases and Related Health Problems, Tenth Edition.
Lattice degeneration: A peripheral vitreoretinal lesion characterized by retinal thinning, overlying vitreous
liquefaction, and firm vitreoretinal adhesions at its margins. Most lesions are ovoid with long axes parallel to
the ora serrata. Round holes frequently occur within the lattice lesion unassociated with PVD. If horseshoe
tears are present, they are seen at the development of PVD and usually are observed at the margins of lattice
lesions.
Macular pucker: Distortion of the retina in the macular region due to proliferation and contraction of a
fibrocellular membrane on the inner surface of the retina.
Operculated retinal tear or break: A defect in the retina caused by vitreoretinal traction at the site of the
lesion. The traction pulls a circular or oval piece of retinal tissue (the operculum) free from the retinal
surface. If this occurs during PVD, all traction in the vicinity of the retinal break is usually eliminated.
Posterior vitreous detachment (PVD): A separation of the posterior vitreous cortex from the internal surface
of the retina. This usually occurs as an acute event after substantial age-related liquefaction in the vitreous
gel; the separation usually extends rapidly to the posterior margin of the vitreous base in all quadrants.
Adhesions between the vitreous cortex and retina or retinal blood vessels may cause retinal breaks and/or
vessel rupture. Vitreous hemorrhage and/or localized intraretinal hemorrhage may accompany this event.
Posterior vitreous detachment is diagnosed by slit-lamp biomicroscopy, which will usually show a prominent
plane defining the posterior vitreous face. The presence of a glial annulus in the vitreous cavity (Weiss ring)
is strong evidence of PVD.
PVD: See Posterior vitreous detachment.
Retinal breaks: Full-thickness defects in the retina. Those caused by vitreoretinal traction are usually called
tears. Those that are round and unassociated with vitreoretinal traction are usually called holes.
Retinal dialysis: A specific type of crescentic peripheral retinal break at the ora serrata, usually associated
with trauma.
Rhegmatogenous retinal detachment (RRD): A separation of the retina from the retinal pigment epithelium
caused by fluid passing from the vitreous cavity into the subretinal space through a break in the retina (from
Greek rhegma, “rent”).
Round retinal hole: A round, full-thickness defect or break in the retina, unassociated with vitreoretinal
traction.
RRD: See Rhegmatogenous retinal detachment.

P249
26
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Stickler syndrome: The most common inherited vitreoretinal and systemic disorder associated with RRD.
Ocular features include (1) high myopia; (2) retrolental, transvitreal, and epiretinal membranes and strands;
(3) chorioretinal pigment alterations; (4) lattice degeneration, often with a perivascular component that
extends posteriorly; and (5) various other abnormalities including glaucoma and cataract. Systemic features
include a generalized skeletal dysplasia, often with a marfanoid habitus, flattened facies, high arched or cleft
palate, hearing loss, and other extracranial skeletal anomalies, many of which can be very subtle. The
inheritance pattern is autosomal dominant, and a gene defect has been related to COL2A1.
Subclinical retinal detachment: A retinal detachment that extends more than 1 disc diameter from the
posterior edge of the retinal break, less than 2 disc diameters from the equator, and does not impair the field
of vision.
Vitreoretinal adhesion (VMA): A firm attachment between the cortical vitreous and the inner surface of the
retina. Condensed vitreous strands adhering to the retina may sometimes be visualized using biomicroscopy
or indirect ophthalmoscopy and scleral depression. Traction of the vitreous on the retina during PVD may
cause retinal breaks to occur at these sites.
Vitreomacular traction (VMT): Partial vitreous separation from the retina resulting in mechanical distortion
of the macula.
VMA: See Vitreomacular adhesion
VMT: See Vitreomacular traction
Zonular traction retinal tufts: Small congenital lesions of the peripheral retina caused by thickened zonules
that have been displaced posteriorly to the anterior retina.

P250
27
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Literature searches of the PubMed and Cochrane databases were conducted in April 2018; the search
strategies are provided at www.aao.org/ppp. Specific limited update searches were conducted after June 2019.

(Retinal Detachment/epidemiology[mh]) AND (rhegmatogenous retinal detachment[tiab])

(Retinal Detachment/etiology[MAJR:noexp] OR Retinal

Perforations/etiology[MAJR:noexp] OR Vitreous Detachment/etiology[MAJR:noexp] OR
Retinal Degeneration/etiology[MAJR:noexp]) AND (rhegmatogenous retinal
detachment[tiab] OR posterior vitreous detachment[tiab] OR retinal break*[tiab] OR lattice
degeneration[tiab])

(Retinal Detachment/diagnosis[MAJR:noexp] OR Retinal

Perforations/diagnosis[MAJR:noexp] OR Vitreous Detachment/diagnosis[MAJR:noexp]
OR Retinal Degeneration/diagnosis[MAJR:noexp]) AND (posterior vitreous
detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

(Retinal Detachment[MAJR:noexp] OR Retinal Perforations[MAJR:noexp] OR Vitreous

Detachment[MAJR:noexp] OR Retinal Degeneration[MAJR:noexp]) AND (Risk
Factors[mh]) AND (rhegmatogenous retinal detachment[tiab] OR posterior vitreous
detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab] OR cataract*[tiab]
OR trauma*[tiab] OR injur*[tiab] OR fellow[tiab] OR retinopathy of prematurity[tiab] OR
Stickler[tiab] OR keratorefractive[tiab] OR refractive lens exchange*[tiab] OR phakic
intraocular lens*[tiab])

(Retinal Detachment/surgery[MAJR:noexp] OR Retinal Detachment/therapy[MAJR:noexp]

OR Retinal Detachment/drug therapy[MAJR:noexp] OR Retinal
Perforations/surgery[MAJR:noexp] OR Retinal Perforations/therapy[MAJR:noexp] OR
Retinal Perforations/drug therapy[MAJR:noexp] OR Vitreous
Detachment/surgery[MAJR:noexp] OR Vitreous Detachment/therapy [MAJR:noexp] OR
Vitreous Detachment/drug therapy[MAJR:noexp] OR Retinal
Degeneration/surgery[MAJR:noexp] OR Retinal Degeneration/therapy[MAJR:noexp] OR
Retinal Degeneration/drug therapy[MAJR:noexp]) AND (posterior vitreous
detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

(Retinal Detachment/surgery[MAJR:noexp] OR Retinal Detachment/therapy[MAJR:noexp]

OR Retinal Detachment/drug therapy[MAJR:noexp] OR Retinal
Perforations/surgery[MAJR:noexp] OR Retinal Perforations/therapy[MAJR:noexp] OR
Retinal Perforations/drug therapy[MAJR:noexp] OR Vitreous
Detachment/surgery[MAJR:noexp] OR Vitreous Detachment/therapy [MAJR:noexp] OR
Vitreous Detachment/drug therapy[MAJR:noexp] OR Retinal
Degeneration/surgery[MAJR:noexp] OR Retinal Degeneration/therapy[MAJR:noexp] OR
Retinal Degeneration/drug therapy[MAJR:noexp]) AND (posterior vitreous
detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

(Retinal Detachment [MAJR:noexp] OR Retinal Perforations [MAJR:noexp] OR Vitreous

Detachment [MAJR:noexp] OR Retinal Degeneration[MAJR:noexp]) AND (posterior
vitreous detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

(posterior vitreous detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

AND ((review*[tiab] AND (literature[tiab] OR systematic[tiab] OR search*[tiab])) OR
meta-analysis[tiab])

P251
 PVD, Retinal Breaks, and Lattice Degeneration PPP

(Retinal Detachment [mh] OR Retinal Perforations [mh] OR Vitreous Detachment [mh] OR

Retinal Degeneration[mh]) AND (Quality of Life[mh]) AND (rhegmatogenous retinal
detachment[tiab] OR posterior vitreous detachment[tiab] OR retinal break*[tiab] OR lattice
degeneration[tiab])

(Retinal Detachment [mh] OR Retinal Perforations [mh] OR Vitreous Detachment [mh] OR

Retinal Degeneration[mh]) AND (Cost-Benefit Analysis[mh] OR Cost of Illness[mh])
AND (posterior vitreous detachment[tiab] OR retinal break*[tiab] OR lattice
degeneration[tiab])

(Retinal Detachment/economics [mh] OR Retinal Perforations/economics [mh] OR

Vitreous Detachment/economics [mh] OR Retinal Degeneration/economics[mh]) AND
(posterior vitreous detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

(Retinal Detachment[mh:noexp] OR Retinal Perforations[mh:noexp] OR Vitreous Detachment[mh:noexp]

OR Retinal Degeneration[mh:noexp]) AND (Postoperative Complications[mh]) AND (posterior vitreous
detachment[tiab] OR retinal break*[tiab] OR lattice degeneration[tiab])

P252
29
 PVD, Retinal Breaks, and Lattice Degeneration PPP

Basic and Clinical Science Course

Retina and Vitreous (Section 12, 2019–2020)

Focal Points
Floaters and Flashes (2016)

Ophthalmic Technology Assessment

The Repair of Rhegmatogenous Retinal Detachments (1996; reviewed for currency 2006)

Patient Education Brochure

Detached and Torn Retina (2005)

Preferred Practice Pattern® Guidelines – Free download available at www.aao.org/ppp.

Comprehensive Adult Medical Eye Evaluation (2015)

To order any of these products, except for the free materials, please contact the Academy’s Customer Service
at 866.561.8558 (U.S. only) or 415.561.8540 or www.aao.org/store.

P253
 PVD, Retinal Breaks, and Lattice Degeneration PPP

1. Scottish Intercollegiate Guidelines Network (SIGN). SIGN 50: a guideline developer’s handbook.
Edinburgh: SIGN; 2015. (SIGN publication no. 50). [November 2015]. Available from URL:
http://www.sign.ac.uk.
2. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of
evidence and strength of recommendations. BMJ. 2008;336(7650):924-926.
3. GRADE Working Group. Organizations that have endorsed or that are using GRADE.
http://www.gradeworkinggroup.org/. Accessed September 2019.
4. Johnson MW. Posterior vitreous detachment: evolution and role in macular disease. Retina. 2012;32
Suppl 2:S174-178.
5. Snead MP, Snead DR, James S, Richards AJ. Clinicopathological changes at the vitreoretinal
junction: posterior vitreous detachment. Eye (Lond). 2008;22(10):1257-1262.
6. American Academy of Ophthalmology Retinal/Vitreous Panel. Preferred Practice Pattern
Guidelines. Idiopathic Epiretinal Membrane and Vitreomacular Traction. San Francisco, CA.
American Academy of Ophthalmology; 2019. www.aao.org/PPP.
7. Coppe AM, Lapucci G. Posterior vitreous detachment and retinal detachment following cataract
extraction. Curr Opin Ophthalmol. 2008;19(3):239-242.
8. Gavrilov JC, Gaujoux T, Sellam M, Laroche L, Borderie V. Occurrence of posterior vitreous
detachment after femtosecond laser in situ keratomileusis: Ultrasound evaluation. J Cataract
Refract Surg. 2011;37(7):1300-1304.
9. Osman MH, Khalil NM, El-Agha MS. Incidence of Posterior Vitreous Detachment After
Femtosecond LASIK Compared With Microkeratome LASIK. Cornea. 2017;36(9):1036-1039.
10. Boldrey EE. Risk of retinal tears in patients with vitreous floaters. Am J Ophthalmol.
1983;96(6):783-787.
11. Brod RD, Lightman DA, Packer AJ, Saras HP. Correlation between vitreous pigment granules and
retinal breaks in eyes with acute posterior vitreous detachment. Ophthalmology. 1991;98(9):1366-
1369.
12. Tasman WS. Posterior vitreous detachment and peripheral retinal breaks. Trans Am Acad
Ophthalmol Otolaryngol. 1968;72(2):217-224.
13. Dayan MR, Jayamanne DG, Andrews RM, Griffiths PG. Flashes and floaters as predictors of
vitreoretinal pathology: is follow-up necessary for posterior vitreous detachment? Eye. 1996;10:456-
458.
14. Byer NE. Natural history of posterior vitreous detachment with early management as the premier
line of defense against retinal detachment. Ophthalmology. 1994;101(9):1503-1514.
15. Tani P, Robertson DM, Langworthy A. Rhegmatogenous retinal detachment without macular
involvement treated with scleral buckling. Am J Ophthalmol. 1980;90(4):503-508.
16. Coffee RE, Westfall AC, Davis GH, Mieler WF, Holz ER. Symptomatic posterior vitreous
detachment and the incidence of delayed retinal breaks: case series and meta-analysis. Am J
Ophthalmol. 2007;144(3):409-413.
17. Benson WE, Grand MG, Okun E. Aphakic retinal detachment. Management of the fellow eye. Arch
Ophthalmol. 1975;93(4):245-249.
18. Scott IU, Smiddy WE, Merikansky A, Feuer W. Vitreoretinal surgery outcomes. Impact on bilateral
visual function. Ophthalmology. 1997;104(6):1041-1048.
19. Lincoff H, Stopa M, Kreissig I. Ambulatory binocular occlusion. Retina. 2004;24(2):246-253.
20. van Overdam KA, Bettink-Remeijer MW, Mulder PG, van Meurs JC. Symptoms predictive for the
later development of retinal breaks. Arch Ophthalmol. 2001;119(10):1483-1486.
21. Uchino E, Uemura A, Ohba N. Initial stages of posterior vitreous detachment in healthy eyes of
older persons evaluated by optical coherence tomography. Arch Ophthalmol. 2001;119(10):1475-
1479.
22. Sarrafizadeh R, Hassan TS, Ruby AJ, et al. Incidence of retinal detachment and visual outcome in
eyes presenting with posterior vitreous separation and dense fundus-obscuring vitreous hemorrhage.
Ophthalmology. 2001;108(12):2273-2278.
23. Adelman RA, Parnes AJ, Ducournau D. European Vitreo-Retinal Society (EVRS) Retinal
Detachment Study Group. Strategy for the management of uncomplicated retinal detachments: the

P254
 PVD, Retinal Breaks, and Lattice Degeneration PPP

European Vitreo-Retinal Society Retinal Detachment Study report 1. Ophthalmology.

2013;120(9):1804-1808.
24. Byer NE. What happens to untreated asymptomatic retinal breaks, and are they affected by posterior
vitreous detachment? Ophthalmology. 1998;105(6):1045-1050.
25. Davis MD. Natural history of retinal breaks without detachment. Arch Ophthalmol. 1974;92(3):183-
194.
26. Wilkinson CP. Interventions for asymptomatic retinal breaks and lattice degeneration for preventing
retinal detachment. Cochrane Database Syst Rev. 2014(9):CD003170.
27. Byer NE. Rethinking prophylactic therapy of retinal detachment. In: Stirpe M, ed. Advances in
Vitreoretinal Surgery. New York, NY: Ophthalmic Communications Society; 1992:399-411.
28. Neumann E, Hyams S. Conservative management of retinal breaks. A follow-up study of
subsequent retinal detachment. Br J Ophthalmol. 1972;56(6):482-486.
29. Shea M, Davis MD, Kamel I. Retinal breaks without detachment, treated and untreated. Mod Probl
Ophthalmol. 1974;12(0):97-102.
30. Colyear BH, Jr, Pischel D. Preventive treatment of retinal detachment by means of light
coagulation. Trans Pac Coast Oto-Ophthalmol Soc. 1960;41:193-217.
31. Robertson DM, Norton EW. Long-term follow-up of treated retinal breaks. Am J Ophthalmol.
1973;75(3):395-404.
32. Pollack A, Oliver M. Argon laser photocoagulation of symptomatic flap tears and retinal breaks of
fellow eyes. Br J Ophthalmol. 1981;65(7):469-472.
33. Smiddy WE, Flynn HW, Jr., Nicholson DH, et al. Results and complications in treated retinal
breaks. Am J Ophthalmol. 1991;112(6):623-631.
34. Verdaguer J, Vaisman M. Treatment of symptomatic retinal breaks. Am J Ophthalmol.
1979;87(6):783-788.
35. Byer NE. Subclinical retinal detachment resulting from asymptomatic retinal breaks: prognosis for
progression and regression. Ophthalmology. 2001;108(8):1499-1503; discussion 1503-1494.
36. Davis MD. The natural history of retinal breaks without detachment. Trans Am Ophthalmol Soc.
1973;71:343-372.
37. Byer NE. Long-term natural history of lattice degeneration of the retina. Ophthalmology.
1989;96(9):1396-1401; discussion 1401-1392.
38. Haimann MH, Burton TC, Brown CK. Epidemiology of retinal detachment. Arch Ophthalmol.
1982;100(2):289-292.
39. Wilkes SR, Beard CM, Kurland LT, Robertson DM, O'Fallon WM. The incidence of retinal
detachment in Rochester, Minnesota, 1970-1978. Am J Ophthalmol. 1982;94(5):670-673.
40. Van de Put MA, Hooymans JM, Los LI. The incidence of rhegmatogenous retinal detachment in
The Netherlands. Ophthalmology. 2013;120(3):616-622.
41. Javitt JC, Tielsch JM, Canner JK, Kolb MM, Sommer A, Steinberg EP. Cataract Patient Outcomes
Research Team. National outcomes of cataract extraction: increased risk of retinal complications
associated with Nd:YAG laser capsulotomy. Ophthalmology. 1992;99(10):1487-1498.
42. Tielsch JM, Legro MW, Cassard SD, et al. Risk factors for retinal detachment after cataract surgery:
a population-based case-control study. Ophthalmology. 1996;103(10):1537-1545.
43. Eye Disease Case-Control Study Group. Risk factors for idiopathic rhegmatogenous retinal
detachment. Am J Epidemiol. 1993;137(7):749-757.
44. Bernheim D, Rouberol F, Palombi K, Albrieux M, Romanet JP, Chiquet C. Comparative
prospective study of rhegmatogenous retinal detachments in phakic or pseudophakic patients with
high myopia. Retina. 2013;33(10):2039-2048.
45. Benson WE, Morse PH. The prognosis of retinal detachment due to lattice degeneration. Ann
Ophthalmol. 1978;10(9):1197-1200.
46. Parma ES, Korkko J, Hagler WS, Ala-Kokko L. Radial perivascular retinal degeneration: a key to
the clinical diagnosis of an ocular variant of Stickler syndrome with minimal or no systemic
manifestations. Am J Ophthalmol. 2002;134(5):728-734.
47. Erie JC, Raecker MA, Baratz KH, Schleck CD, Burke JP, Robertson DM. Risk of retinal
detachment after cataract extraction, 1980-2004: a population-based study. Ophthalmology.
2006;113(11):2026-2032.
48. Russell M, Gaskin B, Russell D, Polkinghorne PJ. Pseudophakic retinal detachment after
phacoemulsification cataract surgery: Ten-year retrospective review. J Cataract Refract Surg.
2006;32(3):442-445.

P255
32
 PVD, Retinal Breaks, and Lattice Degeneration PPP

49. Jahn CE, Richter J, Jahn AH, Kremer G, Kron M. Pseudophakic retinal detachment after uneventful
phacoemulsification and subsequent neodymium: YAG capsulotomy for capsule opacification. J
Cataract Refract Surg. 2003;29(5):925-929.
50. Kassem R, Greenwald Y, Achiron A, et al. Peak Occurrence of Retinal Detachment following
Cataract Surgery: A Systematic Review and Pooled Analysis with Internal Validation. J
Ophthalmol. 2018;2018:9206418.
51. Ranta P, Tommila P, Kivela T. Retinal breaks and detachment after neodymium: YAG laser
posterior capsulotomy: five-year incidence in a prospective cohort. J Cataract Refract Surg.
2004;30(1):58-66.
52. Koch DD, Liu JF, Gill EP, Parke DW, II. Axial myopia increases the risk of retinal complications
after neodymium-YAG laser posterior capsulotomy. Arch Ophthalmol. 1989;107(7):986-990.
53. Tuft SJ, Minassian D, Sullivan P. Risk factors for retinal detachment after cataract surgery: a case-
control study. Ophthalmology. 2006;113(4):650-656.
54. Ramos M, Kruger EF, Lashkari K. Biostatistical analysis of pseudophakic and aphakic retinal
detachments. Semin Ophthalmol. 2002;17(3-4):206-213.
55. Ranta P, Kivela T. Retinal detachment in pseudophakic eyes with and without Nd:YAG laser
posterior capsulotomy. Ophthalmology. 1998;105(11):2127-2133.
56. Glacet-Bernard A, Brahim R, Mokhtari O, Quentel G, Coscas G. Retinal detachment following
posterior capsulotomy using Nd:YAG laser. Retrospective study of 144 capsulotomies [in French]. J
Fr Ophtalmol. 1993;16(2):87-94.
57. Ficker LA, Vickers S, Capon MR, Mellerio J, Cooling RJ. Retinal detachment following Nd:YAG
posterior capsulotomy. Eye. 1987;1 (Pt 1):86-89.
58. Arya AV, Emerson JW, Engelbert M, Hagedorn CL, Adelman RA. Surgical management of
pseudophakic retinal detachments: a meta-analysis. Ophthalmology. 2006;113(10):1724-1733.
59. Lois N, Wong D. Pseudophakic retinal detachment. Surv Ophthalmol. 2003;48(5):467-487.
60. Ripandelli G, Coppe AM, Parisi V, et al. Posterior vitreous detachment and retinal detachment after
cataract surgery. Ophthalmology. 2007;114(4):692-697.
61. Mirshahi A, Hoehn F, Lorenz K, Hattenbach LO. Incidence of posterior vitreous detachment after
cataract surgery. J Cataract Refract Surg. 2009;35(6):987-991.
62. Cooling RJ. Traumatic retinal detachment--mechanisms and management. Trans Ophthalmol Soc
UK. 1986;105:575-579.
63. Folk JC, Arrindell EL, Klugman MR. The fellow eye of patients with phakic lattice retinal
detachment. Ophthalmology. 1989;96(1):72-79.
64. Rowe JA, Erie JC, Baratz KH, et al. Retinal detachment in Olmsted County, Minnesota, 1976
through 1995. Ophthalmology. 1999;106(1):154-159.
65. Mastropasqua L, Carpineto P, Ciancaglini M, Falconio G, Gallenga PE. Treatment of retinal tears
and lattice degenerations in fellow eyes in high risk patients suffering retinal detachment: a
prospective study. Br J Ophthalmol. 1999;83(9):1046-1049.
66. Sharma MC, Chan P, Kim RU, Benson WE. Rhegmatogenous retinal detachment in the fellow
phakic eyes of patients with pseudophakic rhegmatogenous retinal detachment. Retina.
2003;23(1):37-40.
67. Kaiser RS, Trese MT, Williams GA, Cox MS, Jr. Adult retinopathy of prematurity: outcomes of
rhegmatogenous retinal detachments and retinal tears. Ophthalmology. 2001;108(9):1647-1653.
68. Snead MP, Payne SJ, Barton DE, et al. Stickler syndrome: correlation between vitreoretinal
phenotypes and linkage to COL 2A1. Eye. 1994;8 (Pt 6):609-614.
69. Brown DM, Graemiger RA, Hergersberg M, et al. Genetic linkage of Wagner disease and erosive
vitreoretinopathy to chromosome 5q13-14. Arch Ophthalmol. 1995;113(5):671-675.
70. Loewenstein A, Goldstein M, Lazar M. Retinal pathology occurring after excimer laser surgery or
phakic intraocular lens implantation: evaluation of possible relationship. Surv Ophthalmol.
2002;47(2):125-135.
71. Ruiz-Moreno JM, Alio JL. Incidence of retinal disease following refractive surgery in 9,239 eyes. J
Refract Surg. 2003;19(5):534-547.
72. Colin J, Robinet A, Cochener B. Retinal detachment after clear lens extraction for high myopia:
seven-year follow-up. Ophthalmology. 1999;106(12):2281-2284; discussion 2285.
73. Chang JS, Meau AY. Visian Collamer phakic intraocular lens in high myopic Asian eyes. J Refract
Surg. 2007;23(1):17-25.
74. Ruiz-Moreno JM, Montero JA, de la Vega C, Alio JL, Zapater P. Retinal detachment in myopic
eyes after phakic intraocular lens implantation. J Refract Surg. 2006;22(3):247-252.

P256
33
 PVD, Retinal Breaks, and Lattice Degeneration PPP

75. Feder RS, Olsen TW, Prum BE, Jr., et al. Comprehensive adult medical eye evaluation Preferred
Practice Pattern® Guidelines. Ophthalmology. 2016;123(1):P209-236.
76. Austin KL, Palmer JR, Seddon JM, et al. Case-control study of idiopathic retinal detachment. Int J
Epidemiol. 1990;19(4):1045-1050.
77. Norregaard JC, Thoning H, Andersen TF, Bernth-Petersen P, Javitt JC, Anderson GF. Risk of
retinal detachment following cataract extraction: results from the International Cataract Surgery
Outcomes Study. Br J Ophthalmol. 1996;80(8):689-693.
78. Javitt JC, Vitale S, Canner JK, Krakauer H, McBean AM, Sommer A. National outcomes of cataract
extraction. I. Retinal detachment after inpatient surgery. Ophthalmology. 1991;98(6):895-902.
79. Kraff MC, Sanders DR. Incidence of retinal detachment following posterior chamber intraocular
lens surgery. J Cataract Refract Surg. 1990;16(4):477-480.
80. Geck U, Pustolla N, Baraki H, Atili A, Feltgen N, Hoerauf H. Posterior vitreous detachment
following intravitreal drug injection. Graefes Arch Clin Exp Ophthalmol. 2013;251(7):1691-1695.
81. Boldrey EE. Vitreous cells as an indicator of retinal tears in asymptomatic or not recently
symptomatic eyes. Am J Ophthalmol. 1997;123(2):263-264.
82. Brockhurst RJ. Modern indirect ophthalmoscopy. Am J Ophthalmol. 1956;41:265-272.
83. Natkunarajah M, Goldsmith C, Goble R. Diagnostic effectiveness of noncontact slitlamp
examination in the identification of retinal tears. Eye. 2003;17(5):607-609.
84. Benz MS, Packo KH, Gonzalez V, et al. A placebo-controlled trial of microplasmin intravitreous
injection to facilitate posterior vitreous detachment before vitrectomy. Ophthalmology.
2010;117(4):791-797.
85. DiBernardo C, Blodi B, Byrne SF. Echographic evaluation of retinal tears in patients with
spontaneous vitreous hemorrhage. Arch Ophthalmol. 1992;110(4):511-514.
86. Wilkinson CP. What ever happened to bilateral patching? Retina. 2005;25(4):393-394.
87. Rabinowitz R, Yagev R, Shoham A, Lifshitz T. Comparison between clinical and ultrasound
findings in patients with vitreous hemorrhage. Eye (Lond). 2004;18(3):253-256.
88. Sandinha MT, Kotagiri AK, Owen RI, Geenen C, Steel DH. Accuracy of B-scan ultrasonography in
acute fundus obscuring vitreous hemorrhage using a standardized scanning protocol and a dedicated
ophthalmic ultrasonographer. Clin Ophthalmol. 2017;11:1365-1370.
89. Tan HS, Mura M, Bijl HM. Early vitrectomy for vitreous hemorrhage associated with retinal tears.
Am J Ophthalmol. 2010;150(4):529-533.
90. Dhingra N, Pearce I, Wong D. Early vitrectomy for fundus-obscuring dense vitreous haemorrhage
from presumptive retinal tears. Graefes Arch Clin Exp Ophthalmol. 2007;245(2):301-304.
91. Melamud A, Pham H, Stoumbos Z. Early Vitrectomy for Spontaneous, Fundus-Obscuring Vitreous
Hemorrhage. Am J Ophthalmol. 2015;160(5):1073-1077 e1071.
92. Karahan E, Karti O, Er D, et al. Risk factors for multiple retinal tears in patients with acute posterior
vitreous detachment. Int Ophthalmol. 2018;38(1):257-263.
93. Ang GS, Townend J, Lois N. Interventions for prevention of giant retinal tear in the fellow eye.
Cochrane Database Syst Rev. 2012(2):CD006909.
94. Chan CK, Mein CE, Crosson JN. Pneumatic vitreolysis for management of symptomatic focal
vitreomacular traction. J Ophthalmic Vis Res. 2017;12(4):419-423.
95. Yu G, Duguay J, Marra KV, et al. Efficacy and safety of treatment options for vitreomacular
traction: a case series and meta-analysis. Retina. 2016;36(7):1260-1270.
96. Znaor L, Medic A, Binder S, Vucinovic A, Marin Lovric J, Puljak L. Pars plana vitrectomy versus
scleral buckling for repairing simple rhegmatogenous retinal detachments. Cochrane Database Syst
Rev. 2019;3:CD009562.
97. Garcia GA, Khoshnevis M, Yee KMP, Nguyen-Cuu J, Nguyen JH, Sebag J. Degradation of contrast
sensitivity function following posterior vitreous detachment. Am J Ophthalmol. 2016;172:7-12.
98. Kokavec J, Wu Z, Sherwin JC, Ang AJ, Ang GS. Nd:YAG laser vitreolysis versus pars plana
vitrectomy for vitreous floaters. Cochrane Database Syst Rev. 2017;6:CD011676.
99. Sebag J, Yee KM, Wa CA, Huang LC, Sadun AA. Vitrectomy for floaters: prospective efficacy
analyses and retrospective safety profile. Retina. 2014;34(6):1062-1068.
100. Ivanova T, Jalil A, Antoniou Y, Bishop PN, Vallejo-Garcia JL, Patton N. Vitrectomy for primary
symptomatic vitreous opacities: an evidence-based review. Eye (Lond). 2016;30(5):645-655.
101. Shah CP, Heier JS. YAG Laser Vitreolysis vs Sham YAG Vitreolysis for Symptomatic Vitreous
Floaters: A Randomized Clinical Trial. JAMA Ophthalmol. 2017;135(9):918-923.
102. Benson WE, Morse PH, Nantawan P. Late complications following cryotherapy of lattice
degeneration. Am J Ophthalmol. 1977;84(4):514-516.

P257
34
 PVD, Retinal Breaks, and Lattice Degeneration PPP

103. Delaney WV, Jr. Retinal tear extension through the cryosurgical scar. Br J Ophthalmol.
1971;55(3):205-209.
104. American Academy of Ophthalmology. Preoperative assessment: responsibilities of the
ophthalmologist. 2012; https://www.aao.org/about/policies. Accessed September 2019.
105. American Academy of Ophthalmology. An ophthalmologist's duties concerning postoperative care -
2012. 2012; https://www.aao.org/clinical-statement/ophthalmologists-duties-concerning-
postoperative-c. Accessed September 2019.
106. Goldberg RE, Boyer DS. Sequential retinal breaks following a spontaneous initial retinal break.
Ophthalmology. 1981;88(1):10-12.
107. Sharma MC, Regillo CD, Shuler MF, Borrillo JL, Benson WE. Determination of the incidence and
clinical characteristics of subsequent retinal tears following treatment of the acute posterior vitreous
detachment-related initial retinal tears. Am J Ophthalmol. 2004;138(2):280-284.
108. Saran BR, Brucker AJ. Macular epiretinal membrane formation and treated retinal breaks. Am J
Ophthalmol. 1995;120(4):480-485.
109. Singh AJ, Seemongal-Dass RR. The influence of counselling on patient return following
uncomplicated posterior vitreous detachment. Eye. 2001;15(Pt 2):152-154.
110. Wagle AM, Lim WY, Yap TP, Neelam K, Au Eong KG. Utility values associated with vitreous
floaters. Am J Ophthalmol. 2011;152(1):60-65.
111. Sebag J. Floaters and the quality of life. Am J Ophthalmol. 2011;152(1):3-4.
112. Yannuzzi NA, Chang JS, Brown GC, Smiddy WE. Cost-Utility of Evaluation for Posterior Vitreous
Detachment and Prophylaxis of Retinal Detachment. Ophthalmology. 2018;125(1):43-50.

P258
35
over

Adult Strabismus
Preferred Practice
Pattern®

© 2019 by the American Academy of Ophthalmology http://dx.doi.org/10.1016/j.ophtha.2019.09.023

Published by Elsevier Inc. ISSN 0161-6420/19
 Adult Strabismus PPP

Secretary for Quality of Care

Timothy W. Olsen, MD

Academy Staff
Ali Al-Rajhi, PhD, MPH
Andre Ambrus, MLIS
Meghan Daly
Flora C. Lum, MD
Doris Mizuiri

Medical Editor: Susan Garratt

Approved by: Board of Trustees

September 7, 2019

© 2019 American Academy of Ophthalmology®

All rights reserved

AMERICAN ACADEMY OF OPHTHALMOLOGY and PREFERRED PRACTICE PATTERN are

registered trademarks of the American Academy of Ophthalmology. All other trademarks are the property of
their respective owners.

Preferred Practice Pattern® guidelines are developed by the Academy’s H. Dunbar Hoskins Jr., M.D. Center
for Quality Eye Care without any external financial support. Authors and reviewers of the guidelines are
volunteers and do not receive any financial compensation for their contributions to the documents. The
guidelines are externally reviewed by experts and stakeholders before publication.

American Academy of Ophthalmology®

P. O. Box 7424
San Francisco, CA 94120-7424
415.561.8500

P322
 Adult Strabismus PPP

The Pediatric Ophthalmology/Adult Strabismus Preferred Practice Pattern® Panel of the American
Association for Pediatric Ophthalmology and Strabismus Adult Strabismus Task Force members wrote the
Adult Strabismus Preferred Practice Pattern® guidelines (“PPP”). The PPP Panel members discussed and
reviewed successive drafts of the document, meeting in person and conducting other review by e-mail
discussion, to develop a consensus over the final version of the document.

Pediatric Ophthalmology/Adult Strabismus Preferred Practice Pattern Panel of the American

Association of Pediatric Ophthalmology and Strabismus 2017–2019

Chair Linda R. Dagi, MD

Vice Chair: Federico G. Velez, MD

Jonathan M. Holmes, MD
Steven M. Archer, MD
Stacy L. Pineles, MD
Mitchell B. Strominger, MD
Matthew Simon Pihlblad, MD
Evelyn A. Paysse, MD
David Stager Jr, MD
David R. Stager Sr, MD
Hatice Tuba Atalay, MD
Hilda Capo, MD
Natalie C. Kerr, MD
Burton J. Kushner, MD
Sarah E. MacKinnon, CO, COMT
Brian N. Campolattaro, MD
Jason H. Peragallo, MD
Reecha S. Bahl, MD

The Preferred Practice Patterns Committee members reviewed and discussed the document during a series
of meetings spanning 2017-2019. The document was edited in response to the discussion and comments.

Preferred Practice Patterns Committee 2019

Robert S. Feder, MD, Chair
Roy S. Chuck, MD, PhD
Steven P. Dunn, MD
Christina J. Flaxel, MD
Steven J. Gedde, MD
Francis S. Mah, MD
Randall J. Olson, MD
David K. Wallace, MD, MPH
David C. Musch, PhD, MPH, Methodologist

The Adult Strabismus PPP was then sent for review to additional internal and external groups and individuals
in August 2019. All those returning comments were required to provide disclosure of relevant relationships
with industry to have their comments considered. Members of the Pediatric Ophthalmology/Strabismus
Preferred Practice Pattern Panel reviewed and discussed these comments and determined revisions to the
document. The following organizations and individuals returned comments.

P323
 Adult Strabismus PPP

In compliance with the Council of Medical Specialty Societies’ Code for Interactions with Companies
(available at www.cmss.org/codeforinteractions.aspx), relevant relationships with industry are listed. The
Academy has Relationship with Industry Procedures to comply with the Code (available at www.aao.org/about-
preferred-practice-patterns). A majority (75%) of the members of the Adult Strabismus Preferred Practice
Pattern Panel 2017–2019 had no financial relationship to disclose.

Adult Strabismus Preferred Practice Pattern Panel 2017–2019

Steven M. Archer, MD: No financial relationships to disclose
Hatice Tuba Atalay, MD: No financial relationships to disclose
Reecha S. Bahl, MD:
Brian N. Campolattaro, MD: No financial relationships to disclose
Hilda Capo, MD: No financial relationships to disclose
Linda R. Dagi, MD: Boston Neurosciences
Jonathan M. Holmes, MD: Grant support from National Institutes of Health and Research to Prevent
Blindness
Natalie C. Kerr, MD: No financial relationships to disclose
Burton J. Kushner, MD: NovaSight – Scientific Advisory Board member
Sarah E. MacKinnon, CO, COMT: No financial relationships to disclose
Evelyn A. Paysse, MD: No financial relationships to disclose
Jason H. Pergallo, MD: No financial relationships to disclose
Matthew Simon Pihlblad, MD: No financial relationships to disclose
Stacy L. Pineles, MD: No financial relationships to disclose
Mitchell B. Strominger, MD: No financial relationships to disclose
David R. Stager Sr, MD: No financial relationships to disclose
David Stager Jr, MD: No financial relationships to disclose
Federico G. Velez, MD: Grant support from Omeros, Bausch+Lomb, Retrophin, Research to Prevent
Blindness

Pediatric Ophthalmology/Strabismus Preferred Practice Patterns Committee 2019

David K. Wallace, MD, PhD: No financial relationships to disclose
Stephen P. Christiansen, MD: No financial relationships to disclose
Katherine A. Lee, MD, PhD: No financial relationships to disclose
Christie L. Morse, MD: Grant Support from Luminopia
Michael X. Repka, MD, MBA: No financial relationships to disclose
Derek T. Sprunger, MD: No financial relationships to disclose
Michele Melia, ScM: No financial relationships to disclose

Preferred Practice Patterns Committee 2019

Roy S. Chuck, MD, PhD: No financial relationships to disclose
Steven P. Dunn, MD: No financial relationships to disclose
Robert S. Feder, MD: No financial relationships to disclose
Christina J. Flaxel, MD: No financial relationships to disclose
Steven J. Gedde, MD: No financial relationships to disclose
Francis S. Mah, MD: No financial relationships to disclose
Randall J. Olson, MD: No financial relationships to disclose
David K. Wallace, MD, MPH: No financial relationships to disclose
David C. Musch, PhD, MPH: No financial relationships to disclose

Secretary for Quality of Care

Timothy W. Olsen, MD: No financial relationships to disclose

Academy Staff
Ali Al-Rajhi, PhD, MPH: No financial relationships to disclose
Andre Ambrus, MLIS: No financial relationships to disclose
Susan Garratt, Medical Editor: No financial relationships to disclose
Meghan Daly: No financial relationships to disclose
Flora C. Lum, MD: No financial relationships to disclose
Doris Mizuiri: No financial relationships to disclose

P324
 Adult Strabismus PPP
The disclosures of relevant relationships to industry of other reviewers of the document from January to
October 2019 are available online at www.aao.org/ppp

OBJECTIVES OF PREFERRED PRACTICE PATTERN GUIDELINES ..........................................................P332

METHODS AND KEY TO RATINGS .....................................................................................................................P333
HIGHLIGHTED FINDINGS & RECOMMENDATIONS FOR CARE ...............................................................P334
SECTION I. ADULT STRABISMUS OVERVIEW
INTRODUCTION .......................................................................................................................................................P335
Disease Definition .........................................................................................................................................................P336
Patient Population..........................................................................................................................................................P336
Clinical Objectives ........................................................................................................................................................P336
BACKGROUND ..........................................................................................................................................................P336
Prevalence .....................................................................................................................................................................P336
Rationale for Treatment ………………………………………………………………………………………………..P336

SECTION II. COMMON AND CLINICALLY IMPORTANT MANIFESTATIONS OF ADULT STRABISMUS

SECTION IIa. PERSISTENT OR RECURRENT CHILDHOOD STRABISMUS
INTRODUCTION .......................................................................................................................................................P339
Disease Definition .........................................................................................................................................................P339
Patient Population..........................................................................................................................................................P339
Clinical Objectives ........................................................................................................................................................P340
BACKGROUND ..........................................................................................................................................................P340
Prevalence .....................................................................................................................................................................P340
Natural History ..............................................................................................................................................................P340
Rationale for Treatment .................................................................................................................................................P340
CARE PROCESS ........................................................................................................................................................P341
Patient Outcome Criteria ...............................................................................................................................................P341
Diagnosis .......................................................................................................................................................................P341
History ...................................................................................................................................................................P341
Examination...........................................................................................................................................................P341
Management ..................................................................................................................................................................P343
Monitor/Observe....................................................................................................................................................P343
Nonsurgical ...........................................................................................................................................................P343
Surgical..................................................................................................................................................................P343
Provider and Setting ......................................................................................................................................................P343
Counseling and Referral ................................................................................................................................................P343
SECTION IIb. SENSORY STRABISMUS
INTRODUCTION .......................................................................................................................................................P344
Disease Definition .........................................................................................................................................................P344
Patient Population..........................................................................................................................................................P344
Clinical Objectives ........................................................................................................................................................P344
BACKGROUND ..........................................................................................................................................................P344
Prevalence .....................................................................................................................................................................P344
Natural History ..............................................................................................................................................................P344
Rationale for Treatment .................................................................................................................................................P345
CARE PROCESS ........................................................................................................................................................P345
Patient Outcome Criteria ...............................................................................................................................................P345
Diagnosis .......................................................................................................................................................................P345
History ...................................................................................................................................................................P345
Examination...........................................................................................................................................................P345
Management ..................................................................................................................................................................P346
Monitor/Observe....................................................................................................................................................P346
Nonsurgical ...........................................................................................................................................................P346
Surgical..................................................................................................................................................................P346
Provider and Setting ......................................................................................................................................................P346
Counseling and Referral ................................................................................................................................................P346
SECTION IIc. CONVERGENCE INSUFFICIENCY
INTRODUCTION .......................................................................................................................................................P347
Disease Definition .........................................................................................................................................................P347
Patient Population..........................................................................................................................................................P347

P325
 Adult Strabismus PPP
Clinical Objectives ........................................................................................................................................................P347
BACKGROUND ..........................................................................................................................................................P347
Incidence .......................................................................................................................................................................P347

Risk Factors ...................................................................................................................................................................P347

Natural History ..............................................................................................................................................................P347
Rationale for Treatment .................................................................................................................................................P348
CARE PROCESS ........................................................................................................................................................P348
Patient Outcome Criteria ...............................................................................................................................................P348
Diagnosis .......................................................................................................................................................................P348
History ...................................................................................................................................................................P348
Examination...........................................................................................................................................................P348
Management ................................................................................................................................................................. P349
Monitor/Observe...................................................................................................................................................P349
Nonsurgical ...........................................................................................................................................................P349
Surgical..................................................................................................................................................................P349
Provider and Setting ......................................................................................................................................................P349
Counseling and Referral ................................................................................................................................................P349

SECTION IId ACQUIRED STRABISMUS RELATED TO AGING AND MYOPIA

INTRODUCTION .......................................................................................................................................................P350
Disease Definition .........................................................................................................................................................P350
Divergence Insufficiency .......................................................................................................................................P350
Sagging Eye Syndrome .........................................................................................................................................P350
Strabismus fixus (“Heavy Eye Syndrome”)...........................................................................................................P351
Patient Population..........................................................................................................................................................P351
Clinical Objectives ........................................................................................................................................................P351
BACKGROUND ..........................................................................................................................................................P351
Prevalence and Risk Factors ..........................................................................................................................................P351
Divergence Insufficiency .......................................................................................................................................P351
Sagging Eye Syndrome .........................................................................................................................................P351
Strabismus fixus or “Heavy Eye” ..........................................................................................................................P352
Natural History ..............................................................................................................................................................P352
Rationale for Treatment .................................................................................................................................................P352
CARE PROCESS ........................................................................................................................................................P352
Patient Outcome Criteria ...............................................................................................................................................P352
Diagnosis .......................................................................................................................................................................P352
History ...................................................................................................................................................................P352
Examination...........................................................................................................................................................P353
Management ..................................................................................................................................................................P354
Divergence Insufficiency .......................................................................................................................................P354
Sagging Eye Syndrome .........................................................................................................................................P355
Strabismus fixus ....................................................................................................................................................P355
Provider and Setting ......................................................................................................................................................P356
Counseling and Referral ................................................................................................................................................P356
Divergence Insufficiency/Sagging Eye Syndrome ................................................................................................P356
Strabismus fixus ....................................................................................................................................................P356

SECTION IIe. THYROID EYE DISEASE

INTRODUCTION .......................................................................................................................................................P358
Disease Definition .........................................................................................................................................................P358
Patient Population..........................................................................................................................................................P358
Clinical Objectives ........................................................................................................................................................P358
BACKGROUND ..........................................................................................................................................................P358
Incidence ......................................................................................................................................................................P358
Risk Factors ...................................................................................................................................................................P358
Natural History ..............................................................................................................................................................P359
Rationale for Treatment .................................................................................................................................................P359
CARE PROCESS ........................................................................................................................................................P359
Patient Outcome Criteria ...............................................................................................................................................P359
Diagnosis .......................................................................................................................................................................P359
History ...................................................................................................................................................................P359
Examination...........................................................................................................................................................P359
Management ..................................................................................................................................................................P360

P326
 Adult Strabismus PPP

Observation ...........................................................................................................................................................P361
Nonsurgical ...........................................................................................................................................................P361
Surgical..................................................................................................................................................................P361
Provider and Setting ......................................................................................................................................................P363
Counseling and Referral ................................................................................................................................................P363

SECTION IIf. STRABISMUS AFTER ORBITAL TRAUMA

INTRODUCTION .......................................................................................................................................................P364
Disease Definition .........................................................................................................................................................P364
Patient Population..........................................................................................................................................................P364
Clinical Objectives ........................................................................................................................................................P364
BACKGROUND ..........................................................................................................................................................P365
Prevalence and Risk Factors ..........................................................................................................................................P365
Natural History ..............................................................................................................................................................P365
Rationale for Treatment .................................................................................................................................................P365
CARE PROCESS ........................................................................................................................................................P365
Diagnosis .......................................................................................................................................................................P365
History ...................................................................................................................................................................P365
Examination...........................................................................................................................................................P365
Management ..................................................................................................................................................................P366
General Guidelines ................................................................................................................................................P366
Guidelines for Treating Strabismus .......................................................................................................................P367
Provider and Setting ......................................................................................................................................................P368
Counseling and Referral ................................................................................................................................................P368

SECTION IIg. STRABISMUS ASSOCIATED WITH OTHER OPHTHALMIC SURGERY

INTRODUCTION .......................................................................................................................................................P369
Disease Definition
Cataract Extraction or Keratoplasty .......................................................................................................................P369
Glaucoma Filtering Procedures (including trabeculectomy and glaucoma plate reservoir surgery) ......................P370
Scleral Buckling Procedures ..................................................................................................................................P370
Pterygium Surgery .................................................................................................................................................P370
Blepharoplasty and Eyelid Procedures ..................................................................................................................P370
Patient Population..........................................................................................................................................................P371
Clinical Objectives ........................................................................................................................................................P371
BACKGROUND ..........................................................................................................................................................P371
Cataract Extraction ................................................................................................................................................P371
Glaucoma Filtering Procedures .............................................................................................................................P371
Scleral Buckling Procedures ..................................................................................................................................P372
Blepharoplasty and Eyelid Procedures ..................................................................................................................P372
Natural History ..............................................................................................................................................................P372
Rationale for Treatment .................................................................................................................................................P372
CARE PROCESS ........................................................................................................................................................P373
Patient Outcome Criteria ...............................................................................................................................................P373
Diagnosis .......................................................................................................................................................................P373
History ...................................................................................................................................................................P373
Examination...........................................................................................................................................................P373
Management ..................................................................................................................................................................P374
Monitor/Observe....................................................................................................................................................P374
Nonsurgical ...........................................................................................................................................................P374
Surgical..................................................................................................................................................................P374
Provider and Setting ......................................................................................................................................................P375
Counseling and Referral ................................................................................................................................................P375

SECTION IIh. SUPERIOR OBLIQUE PALSY

INTRODUCTION ........................................................................................................................................................P376
Disease Definition ..........................................................................................................................................................P376
Patient Population...........................................................................................................................................................P376
Clinical Objectives .........................................................................................................................................................P376
BACKGROUND ...........................................................................................................................................................P377
Incidence ........................................................................................................................................................................P377
Risk Factors ....................................................................................................................................................................P377
Natural History ...............................................................................................................................................................P377
Rationale for Treatment ..................................................................................................................................................P377
CARE PROCESS .........................................................................................................................................................P377
P327
 Adult Strabismus PPP
Patient Outcome Criteria ...............................................................................................................................................P377
Diagnosis .......................................................................................................................................................................P377
History ...................................................................................................................................................................P377
Patient Outcome Criteria ...............................................................................................................................................P377
Examination...........................................................................................................................................................P378
Diagnosis .......................................................................................................................................................................P377
Management ..................................................................................................................................................................P378
History ...................................................................................................................................................................P377
Monitor/Observe....................................................................................................................................................P378
Examination...........................................................................................................................................................P378
Nonsurgical..................................................................................................................................................................P378
Management ...........................................................................................................................................................P378
Surgical ..................................................................................................................................................................P378
Monitor/Observe....................................................................................................................................................P378
Provider and Setting ......................................................................................................................................................P379
Nonsurgical ...........................................................................................................................................................P378
CounselingSurgical ..................................................................................................................................................................P378
and Referral ................................................................................................................................................P379
Provider and Setting ......................................................................................................................................................P379
SECTION
Counseling IIi.andSKEW ReferralDEVIATION
................................................................................................................................................P379
INTRODUCTION .......................................................................................................................................................P380
SECTION IIi. SKEW DEVIATION
Disease Definition .........................................................................................................................................................P380
INTRODUCTION .......................................................................................................................................................P380
Patient Population..........................................................................................................................................................P381
Disease Definition .........................................................................................................................................................P380
Clinical Objectives ........................................................................................................................................................P381
Patient Population..........................................................................................................................................................P381
BACKGROUND
Clinical Objectives ..........................................................................................................................................................P381
........................................................................................................................................................P381
Prevalence .....................................................................................................................................................................P381
BACKGROUND ..........................................................................................................................................................P381
Risk Prevalence
Factors ...................................................................................................................................................................P381
.....................................................................................................................................................................P381
Natural
RiskHistory
Factors..............................................................................................................................................................P381
...................................................................................................................................................................P381
Rationale forHistory
Natural Treatment .................................................................................................................................................P381
..............................................................................................................................................................P381
CARE Rationale
PROCESS for Treatment .................................................................................................................................................P381
........................................................................................................................................................P381
CARE
Patient OutcomePROCESS Criteria ........................................................................................................................................................P381
...............................................................................................................................................P381
Patient.......................................................................................................................................................................P382
Diagnosis Outcome Criteria ...............................................................................................................................................P381
Diagnosis .......................................................................................................................................................................P382
History ...................................................................................................................................................................P382
History ...................................................................................................................................................................P382
Examination...........................................................................................................................................................P382
Examination...........................................................................................................................................................P382
Management ..................................................................................................................................................................P383
Management ..................................................................................................................................................................P383
Provider and and
Provider Setting.........................................................................................................................................................P383
Setting.........................................................................................................................................................P383
Counseling and
Counseling and Referral
Referral................................................................................................................................................P383
................................................................................................................................................P383

SECTION
SECTION IIj. IIj.
ABDUCENS
ABDUCENS PALSY
PALSY
INTRODUCTION
INTRODUCTION .......................................................................................................................................................P384
.......................................................................................................................................................P384
Disease Definition
Disease Definition.........................................................................................................................................................P384
.........................................................................................................................................................P384
Patient
Patient Population
Population ..........................................................................................................................................................P385
..........................................................................................................................................................P385
Clinical
Clinical Objectives
Objectives ........................................................................................................................................................P385
........................................................................................................................................................P385
BACKGROUND ..........................................................................................................................................................P385
BACKGROUND ..........................................................................................................................................................P385
Incidence .......................................................................................................................................................................P385
Incidence .......................................................................................................................................................................P385
Risk Factors ...................................................................................................................................................................P385
Risk Factors ...................................................................................................................................................................P385
Rationale for Treatment .................................................................................................................................................P385
Rationale for Treatment .................................................................................................................................................P385
CARE PROCESS ........................................................................................................................................................P385
CARE PROCESS
Patient Outcome........................................................................................................................................................P385
Criteria ...............................................................................................................................................P385
Patient Outcome
Diagnosis Criteria ...............................................................................................................................................P385
.......................................................................................................................................................................P385
DiagnosisHistory
.......................................................................................................................................................................P385
...................................................................................................................................................................P386
History ...................................................................................................................................................................P386
Examination ...........................................................................................................................................................P386
Examination
Ancillary...........................................................................................................................................................P386
Testing ...................................................................................................................................................P386
Management
Ancillary Testing..................................................................................................................................................................P387
...................................................................................................................................................P386
ManagementMonitor/Observe....................................................................................................................................................P387
..................................................................................................................................................................P387
Nonsurgical ...........................................................................................................................................................P387
Monitor/Observe....................................................................................................................................................P387
Surgical..................................................................................................................................................................P387
Nonsurgical ...........................................................................................................................................................P387
Provider and Setting ......................................................................................................................................................P388
Surgical..................................................................................................................................................................P387
Counseling and Referral ................................................................................................................................................P388
Provider and Setting ......................................................................................................................................................P388
Counseling
SECTION and Referral ................................................................................................................................................P388
IIk. OCULOMOTOR PALSY
INTRODUCTION .......................................................................................................................................................P389
SECTION IIk. OCULOMOTOR
Disease Definition PALSY
.........................................................................................................................................................P389
INTRODUCTION
Patient Population .......................................................................................................................................................P389
..........................................................................................................................................................P389
Disease Definition
Clinical Objectives.........................................................................................................................................................P389
........................................................................................................................................................P389
Patient Population..........................................................................................................................................................P389
BACKGROUND ..........................................................................................................................................................P389
Incidence
Clinical .......................................................................................................................................................................P389
Objectives ........................................................................................................................................................P389
Risk Factors ...................................................................................................................................................................P389
BACKGROUND ..........................................................................................................................................................P389
Rationale
Incidence for Treatment .................................................................................................................................................P390
.......................................................................................................................................................................P389
CARE PROCESS ........................................................................................................................................................P390
Risk Factors ...................................................................................................................................................................P389
Patient Outcome Criteria ...............................................................................................................................................P390
Rationale for Treatment .................................................................................................................................................P390
Diagnosis .......................................................................................................................................................................P390
CARE PROCESS ........................................................................................................................................................P390
History ...................................................................................................................................................................P391
Patient Outcome Criteria ...............................................................................................................................................P390
Diagnosis .......................................................................................................................................................................P390
P328
History ...................................................................................................................................................................P391
 Adult Strabismus PPP
Examination...........................................................................................................................................................P391
Management ..................................................................................................................................................................P392
Monitor/Observe....................................................................................................................................................P392
Nonsurgical ...........................................................................................................................................................P392
Surgical..................................................................................................................................................................P392
Provider and Setting ......................................................................................................................................................P393
Counseling and Referral................................................................................................................................................P393
________________________________________________________________________________________________

SECTION IIl. MYASTHENIA GRAVIS

INTRODUCTION .......................................................................................................................................................P394
Disease Definition .........................................................................................................................................................P394
Patient Population..........................................................................................................................................................P394
Clinical Objectives ........................................................................................................................................................P394
BACKGROUND ..........................................................................................................................................................P395
Prevalence .....................................................................................................................................................................P395
Risk Factors ...................................................................................................................................................................P395
Natural History ..............................................................................................................................................................P395
Rationale for Treatment .................................................................................................................................................P395
CARE PROCESS ........................................................................................................................................................P395
Patient Outcome Criteria ...............................................................................................................................................P395
Diagnosis .......................................................................................................................................................................P395
History ...................................................................................................................................................................P396
Examination...........................................................................................................................................................P396
Management ..................................................................................................................................................................P397
Provider and Setting ......................................................................................................................................................P397
Counseling and Referral ................................................................................................................................................P397

SECTION IIm. FIXATION SWITCH DIPLOPIA

INTRODUCTION ........................................................................................................................................................P398
Disease Definition ..........................................................................................................................................................P398
Patient Population...........................................................................................................................................................P398
Clinical Objectives .........................................................................................................................................................P398
BACKGROUND ...........................................................................................................................................................P398
Prevalence and Risk Factors ...........................................................................................................................................P398
Natural History ...............................................................................................................................................................P399
Rationale for Treatment ..................................................................................................................................................P399
CARE PROCESS .........................................................................................................................................................P399
Patient Outcome Criteria ................................................................................................................................................P399
Diagnosis ........................................................................................................................................................................P399
History ....................................................................................................................................................................P399
Examination............................................................................................................................................................P400
Management ...................................................................................................................................................................P400
Monitor/Observe.....................................................................................................................................................P400
Nonsurgical ............................................................................................................................................................P400
Surgical...................................................................................................................................................................P400
Provider and Setting .......................................................................................................................................................P400
Counseling and Referral .................................................................................................................................................P401

SECTION IIn. RETINAL MISREGISTRATION (BINOCULAR RETINAL DIPLOPIA)

INTRODUCTION ........................................................................................................................................................P402
Disease Definition ..........................................................................................................................................................P402
Patient Population...........................................................................................................................................................P402
Clinical Objectives .........................................................................................................................................................P402
BACKGROUND ...........................................................................................................................................................P402
Prevalence and Risk Factors ...........................................................................................................................................P402
Natural History ...............................................................................................................................................................P402
Rationale for Treatment ..................................................................................................................................................P403
CARE PROCESS .........................................................................................................................................................P403
Patient Outcome Criteria ................................................................................................................................................P403
Diagnosis ........................................................................................................................................................................P403
History ....................................................................................................................................................................P403
Examination............................................................................................................................................................P403
P329
 Adult Strabismus PPP

Management ...................................................................................................................................................................P404
Monitor/Observe.....................................................................................................................................................P404
Nonsurgical ............................................................................................................................................................P404
Surgical...................................................................................................................................................................P404
Provider and Setting .......................................................................................................................................................P405
Counseling and Referral .................................................................................................................................................P405

SECTION III. COMPLICATIONS OF PERFORMING ADULT STRABISMUS SURGERY

INTRODUCTION .......................................................................................................................................................P406
BACKGROUND ..........................................................................................................................................................P406
Prevalence and Risk Factors ..........................................................................................................................................P406
CARE PROCESS ........................................................................................................................................................P406
Postoperative Concerns .................................................................................................................................................P406
Mild Concerns .......................................................................................................................................................P406
Moderate Concerns ................................................................................................................................................P407
Major Concerns .....................................................................................................................................................P407
Provider and Setting ......................................................................................................................................................P408
Counseling and Referral ................................................................................................................................................P408
SECTION IV. TECHNICAL CONSIDERATIONS WHEN PERFORMING ADULT STRABISMUS SURGERY
INTRODUCTION .......................................................................................................................................................P409
CARE PROCESS ........................................................................................................................................................P409
Surgical Planning and Management ..............................................................................................................................P409
Anticoagulants .......................................................................................................................................................P409
Adjustable Sutures .................................................................................................................................................P409
Microtropias ..........................................................................................................................................................P410
Chemodenervation .................................................................................................................................................P410
Anesthesia .............................................................................................................................................................P410
Complex Strabismus ..............................................................................................................................................P410
Intraoperative Issues ..............................................................................................................................................P411
Provider and Setting ......................................................................................................................................................P411
Counseling and Referral ................................................................................................................................................P411
APPENDIX 1. GLOSSARY OR TERMS ..................................................................................................................P412
APPENDIX 2. ALGORITHM FOR APPROACHING ADULT STRABISMUS BASED ON PRESENTATION
OF DEVIATION .............................................................................................................................................................P413
APPENDIX 3. LITERATURE SEARCHES FOR THIS PPP .................................................................................P414
APPENDIX 4. RELATED ACADEMY MATERIALS ............................................................................................P415
REFERENCES ............................................................................................................................................................P416

P330
 Adult Strabismus PPP
Adult Strabismus PPP

Background:

Strabismus has an estimated incidence of 4% in the adult population. Commonest causes include unresolved or
recurrent childhood strabismus, sensory strabismus, divergence insufficiency or sagging eye syndrome, Graves
orbitopathy, 4th and 6th cranial nerve palsies and convergence insufficiency. Other notable causes include orbital
trauma, third nerve palsy, myasthenia gravis, myopic strabismus fixus, skew deviation, retinal misregistration
(“dragged-fovea diplopia”) and strabismus occurring after other ophthalmic procedures.

This Preferred Practice Pattern (PPP) details evaluation and management of all these disorders, surgical
considerations unique to adults with strabismus, and improvements in quality of life often experienced by those
successfully treated. The PPP recommendations are based on Cochrane-identified reliable systematic reviews.

Rationale for treatment:

Treating adult strabismus can improve binocular function including stereoacuity and binocular fields and can reduce
binocular inhibition, diplopia and confusion, associated compensatory head posture, and asthenopia. Successful
treatment yields psychosocial benefits due to restoration of normal eye alignment and eye contact.

Care Process:

Pre-existing suppression increases risk for the development of fixation switch diplopia, which can occur when the
previously nondominant eye becomes the dominant eye as a result of monovision or optical changes associated with
cataract surgery. Fixation switch diplopia may also occur if vision loss from diseases such as macular degeneration
or diabetic retinopathy leaves the previously nondominant eye with better acuity.

The risk of new-onset diplopia after strabismus surgery in adults undergoing treatment for unresolved childhood
strabismus without previous diplopia is under 1%, and the risk of serious sight-threatening complications is
exceedingly rare. Marked improvement in health-related quality of life after successful strabismus surgery is
reported in both diplopic and nondiplopic patients, including improvements in general function, reading, self-
perception and social interactions.

P331

P331
 Adult Strabismus PPP

As a service to its members and the public, the American Academy of Ophthalmology has developed a series
of Preferred Practice Pattern® guidelines that identify characteristics and components of quality eye care.
Appendix 1 describes the core criteria of quality eye care.

The Preferred Practice Pattern® guidelines are based on the best available scientific data as interpreted by
panels of knowledgeable health professionals. In some instances, such as when results of carefully conducted
clinical trials are available, the data are particularly persuasive and provide clear guidance. In other instances,
the panels have to rely on their collective judgment and evaluation of available evidence.

These documents provide guidance for the pattern of practice, not for the care of a particular
individual. While they should generally meet the needs of most patients, they cannot possibly best meet the
needs of all patients. Adherence to these PPPs will not ensure a successful outcome in every situation. These
practice patterns should not be deemed inclusive of all proper methods of care or exclusive of other methods
of care reasonably directed at obtaining the best results. It may be necessary to approach different patients’
needs in different ways. The physician must make the ultimate judgment about the propriety of the care of a
particular patient in light of all of the circumstances presented by that patient. The American Academy of
Ophthalmology is available to assist members in resolving ethical dilemmas that arise in the course of
ophthalmic practice.

Preferred Practice Pattern® guidelines are not medical standards to be adhered to in all individual
situations. The Academy specifically disclaims any and all liability for injury or other damages of any kind,
from negligence or otherwise, for any and all claims that may arise out of the use of any recommendations or
other information contained herein.

References to certain drugs, instruments, and other products are made for illustrative purposes only and are
not intended to constitute an endorsement of such. Such material may include information on applications
that are not considered community standard, that reflect indications not included in approved U.S. Food and
Drug Administration (FDA) labeling, or that are approved for use only in restricted research settings. The
FDA has stated that it is the responsibility of the physician to determine the FDA status of each drug or
device he or she wishes to use, and to use them with appropriate patient consent in compliance with
applicable law.

Innovation in medicine is essential to ensure the future health of the American public, and the Academy
encourages the development of new diagnostic and therapeutic methods that will improve eye care. It is
essential to recognize that true medical excellence is achieved only when the patients’ needs are the foremost
consideration.

All Preferred Practice Pattern® guidelines are reviewed by their parent panel annually or earlier if
developments warrant and updated accordingly. To ensure that all PPPs are current, each is valid for 5 years
from the “approved by” date unless superseded by a revision. Preferred Practice Pattern guidelines are
funded by the Academy without commercial support. Authors and reviewers of PPPs are volunteers and do
not receive any financial compensation for their contributions to the documents. The PPPs are externally
reviewed by experts and stakeholders, including consumer representatives, before publication. The PPPs are
developed in compliance with the Council of Medical Specialty Societies’ Code for Interactions with
Companies. The Academy has Relationship with Industry Procedures (available at www.aao.org/about-
preferred-practice-patterns) to comply with the Code.

Appendix 2 contains the International Statistical Classification of Diseases and Related Health Problems
(ICD) codes for the disease entities that this PPP covers. The intended users of the Esotropia and Exotropia
PPP are ophthalmologists.

P332
 Adult Strabismus PPP

Preferred Practice Pattern® guidelines should be clinically relevant and specific enough to provide useful
information to practitioners. Where evidence exists to support a recommendation for care, the
recommendation should be given an explicit rating that shows the strength of evidence. To accomplish these
aims, methods from the Scottish Intercollegiate Guideline Network1 (SIGN) and the Grading of
Recommendations Assessment, Development and Evaluation2 (GRADE) group are used. GRADE is a
systematic approach to grading the strength of the total body of evidence that is available to support
recommendations on a specific clinical management issue. Organizations that have adopted GRADE include
SIGN, the World Health Organization, the Agency for Healthcare Research and Policy, and the American
College of Physicians.3
◆ All studies used to form a recommendation for care are graded for strength of evidence individually, and
that grade is listed with the study citation.
◆ To rate individual studies, a scale based on SIGN1 is used. The definitions and levels of evidence to rate
individual studies are as follows:
I++ High-quality meta-analyses, systematic reviews of randomized controlled trials (RCTs), or
RCTs with a very low risk of bias
I+ Well-conducted meta-analyses, systematic reviews of RCTs, or RCTs with a low risk of bias
I- Meta-analyses, systematic reviews of RCTs, or RCTs with a high risk of bias
II++ High-quality systematic reviews of case-control or cohort studies
High-quality case-control or cohort studies with a very low risk of confounding or bias and a
high probability that the relationship is causal
II+ Well-conducted case-control or cohort studies with a low risk of confounding or bias and a
moderate probability that the relationship is causal
II- Case-control or cohort studies with a high risk of confounding or bias and a significant risk that
the relationship is not causal
III Nonanalytic studies (e.g., case reports, case series)

◆ Recommendations for care are formed based on the body of the evidence. The body of evidence quality
ratings are defined by GRADE2 as follows:
Good quality Further research is very unlikely to change our confidence in the estimate of
effect
Moderate quality Further research is likely to have an important impact on our confidence in the
estimate of effect and may change the estimate
Insufficient quality Further research is very likely to have an important impact on our confidence in
the estimate of effect and is likely to change the estimate
Any estimate of effect is very uncertain

◆ Key recommendations for care are defined by GRADE 2 as follows:

Strong Used when the desirable effects of an intervention clearly outweigh the
recommendation undesirable effects or clearly do not
Discretionary Used when the trade-offs are less certain—either because of low-quality
recommendation evidence or because evidence suggests that desirable and undesirable effects are
closely balanced

◆ The Highlighted Findings and Recommendations for Care section lists points determined by the PPP
Panel to be of particular importance to vision and quality of life outcomes.
◆ All recommendations for care in this PPP were rated using the system described above. Ratings are embedded
throughout the PPP main text in italics.
◆ Literature searches to update the PPP were undertaken in March 2016, February 2017, and June 2019 in
the PubMed and Cochrane databases. Complete details of the literature searches are available in
Appendix 3.
P333
 Adult Strabismus PPP

Strabismus in adults has profound negative effects on quality of life and many aspects of day-to-day
function. Strabismus surgery very often improves quality of life and function, and there are instruments to
assess these aspects of evaluation and treatment. Patients with diplopia tend to have greatest improvements
in functional domains, and nondiplopic patients tend to have greatest improvements in psychosocial
domains.

Recessions of the restricted muscles are the mainstay of surgical correction in thyroid eye disease.
Resection is generally avoided in restrictive disease out of concern for further reducing ductions and
operating on a rectus muscle that would best be spared to provided ciliary artery supply to the anterior
segment. However, it can be a useful adjunct in select cases, particularly when extremely large recessions
have not fully corrected the alignment.

Screening for a history of childhood amblyopia or strabismus, checking spectacles for prism, and
performing a cycloplegic refraction and cover testing are recommended for all patients undergoing
refractive or cataract surgery, especially for those patients for whom monovision is planned. Patients with a
history of childhood strabismus and suppression are particularly at risk for developing fixation switch
diplopia. This form of diplopia results when the previously nondominant eye becomes the dominant eye as
a result of intended or unintended monovision by refractive manipulation or cataract surgery. A trial of
monovision with contact lenses is prudent prior to corneal or lenticular refractive surgery to determine
whether surgically induced monovision will result in new-onset diplopia. The development of asymmetric
vision loss from other common diseases such as macular degeneration, myopia with axial elongation in the
previously dominant eye, or diabetic retinopathy when the nondominant eye is left with better acuity may
also result in fixation switch diplopia.

P334
 Adult Strabismus PPP

Strabismus is misalignment of the eyes and may be congenital or acquired. Although more typically associated
with the pediatric population, strabismus is quite common among adults, with an estimated incidence of 4% in
this population.4-6 The causes of strabismus in the adult population are numerous, in part because the challenges
to ocular alignment common in the pediatric population persist, and because new disorders destabilizing
alignment occur secondary to aging, vision loss, myopia, endocrine and neurologic disorders, and iatrogenic and
non-iatrogenic trauma to the globe or orbit. Notable causes of strabismus in the adult population include:
• Recurrent or unresolved childhood strabismus
• Sensory strabismus
• Convergence insufficiency
• Divergence insufficiency
• Sagging eye syndrome
• Strabismus associated with high axial myopia
• Strabismus fixus
• Graves’ disease
• Orbital trauma
• Strabismus associated with other ophthalmic surgery
• Fourth nerve palsy
• Skew deviation
• Sixth nerve palsy
• Third nerve palsy
• Myasthenia gravis
• Fixation switch diplopia
• Retinal Misregistration or Binocular Retinal Diplopia

Accordingly, accurate diagnosis of the etiology of strabismus in the adult population requires the expertise to
recognize associated signs and symptoms and familiarity with studies that may confirm the diagnosis. In
addition, the goal(s) of treatment may impact the plan for surgical or nonsurgical intervention. Adult patients
often have unique concerns associated with functional vision as well as psychosocial concerns that affect
quality of life. In the past, most emphasis was on improving motor alignment, but it is now understood that the
goals of treatment should be much broader and include sensory recovery when possible as well as gains in
psychosocial and functional domains of vision-related quality of life.

P335
 Adult Strabismus PPP

Success rates depend on the subpopulation studied and the goals of surgery, but overall they are approximately
80% after one procedure5,7,8 and may exceed 95% if a second procedure is performed on eyes failing the first
intervention.4 Intractable postoperative diplopia in primary gaze for adult patients without diplopia prior to
intervention occurs in 1% or fewer cases.4,7,9,10 Satisfactory resolution of diplopia in primary position for those
who present for treatment of diplopia is possible more than 95% of the time, although diplopia in eccentric gaze
may persist.5

The patient population is adults with strabismus.

• Perform a clinical examination and ancillary testing as indicated to diagnose the cause of strabismus
• Counsel the patient on the diagnosis and treatment options
• Consult other medical providers if the diagnosis indicates the need for multidisciplinary management
• Establish priorities in the goal-directed management of strabismus (improved eye contact and appearance
of alignment (reconstructive), enhancing binocular potential, or reduction or resolution of diplopia and/or
compensatory head posture11,12)
• Inform the patient’s primary care and other health care providers of the diagnosis and treatment plan

Individual practice estimates vary based on setting and provider. Intelligent Research in Sight (IRIS) registry
data document a prevalence of 2.7% in ophthalmology practices.4-6,13

In general, the goal of adult strabismus surgery is to optimize the functional visual status of patients while also
addressing psychosocial concerns. The potential benefits of strabismus surgery in adults include:
• Improvement in binocularity (ranging from simultaneous perception to stereopsis)1,4,7,14-17
• Improvement in diplopia or compensatory head position4,5,7
• Normalization of binocular visual field18
• Improvement in binocular summation or resolution of binocular inhibition 19
• Reduction of asthenopia
• Psychosocial benefits related to the restoration, rehabilitation, and improvement in an abnormal facial
appearance secondary to ocular misalignment4,6,7,20-26

P336
 Adult Strabismus PPP

The overall rate of success for achieving satisfactory ocular alignment cannot easily be summarized because of
the vast heterogeneity within this group and the lack of consensus on what constitutes “success.” However,
analysis of several studies comprising multiple strabismus subtypes suggests that the overall success rate
approaches 80%,4,7,27 with resolution of diplopia in a similar proportion.4,7,27 Functional improvement in
binocularity can also be achieved in many patients. Although the highest level of binocularity (high-grade
stereopsis) requires good vision in each eye, orthotropia and baseline binocular potential, lower levels of
binocularity (such as simultaneous perception or peripheral fusion) can be achieved in patients with poorer
vision or longstanding strabismus. Additionally, patients with poor vision in one eye or even infantile-onset
strabismus can achieve improvement in binocular summation after surgery.19 Successful strabismus surgery
may also increase the binocular visual field in patients with esotropia.18

For individual patients, a realistic functional goal for surgical success is dependent on their strabismus subtype
as well as the duration of strabismus. For example, in patients with strabismus acquired during adulthood, the
functional goal of surgery is likely to be resolution of or improvement in diplopia or compensatory head
position and high-grade stereopsis. However, in adults with a longstanding history of infantile-onset strabismus,
achieving bifoveal fixation is not a realistic goal, and functional goals would therefore include normalization of
visual fields, improvement in binocular summation, or improvement of diplopia or compensatory head position,
if present. It is important to measure and document diplopia in addition to ocular alignment,28 and there are now
methods to quantify diplopia. The field of binocular single vision can be plotted on a Goldmann perimeter,29 a
cervical range-of-motion device can be used to record diplopia in specific positions of gaze,30,31 or the Diplopia
Questionnaire can be used to assess the frequency of diplopia in specific positions of gaze.32

Aside from the functional goals of surgery described above, psychosocial concerns are also important. Patients
with strabismus endure both psychosocial and economic hardships, such as difficulty obtaining employment,33
receiving promotions,34 and overcoming negative social bias.25 Assessment of psychosocial aspects of the
impact of strabismus is best achieved using patient-reported outcomes measures (PROMs). These PROMs are
typically questionnaires completed by the patient and then evaluated and scored. The Adult Strabismus-2035
questionnaire was developed using a rigorous method of conducting individual patient interviews to identify
concerns and to generate potential questions. A master questionnaire was then refined by eliminating questions
that were not broadly applicable, that had marked ceiling or floor effects, or that were limited by socioeconomic
status or location.35 Subsequent studies revealed that the questions should be further subdivided into two
psychosocial domains (self-perception and interactions) and two function domains (reading function and
general function).36

Using PROM questionnaires, several studies have reported marked improvement in health-related quality of life
after strabismus surgery in both diplopic and nondiplopic patients.22,37-39 In some patients, this improvement has

P337
 Adult Strabismus PPP

been found to correlate with improvements in binocular function.40 In general, patients with diplopia tend to
have greatest improvements in function domains, and nondiplopic patients tend to have greatest improvements
in psychosocial domains.38 Nevertheless, even adults with limited binocular potential have been found to have
function-related benefits from strabismus surgery.41 The use of PROM questionnaires adds to previous methods
of assessing residual motor angle of deviation and the presence or absence of diplopia.28,39 Additionally, even
some patients who might have been classified as surgical failures, based on residual angle and diplopia, report
dramatic improvements in health-related quality of life, often attributable to improvements in psychosocial
issues.42 In addition to improvements in health-related quality of life, other facets of mental health such as
mood, depression, anxiety, social avoidance, and self-esteem have been found to improve after strabismus
surgery in adults.22,43-45 Strabismus surgery may also have a positive impact of other aspects of overall health,
such as reducing falls in the elderly.46

Nevertheless, despite improvements in binocular function and psychosocial issues for most adults undergoing
strabismus surgery, there are some patients who do not perceive an improvement in quality of life. Such lack of
improvement is likely multifactorial, but it may be due to depressive symptoms that are not at the level of
clinical depression,47 other mental illnesses,48 or unrealistic expectations.49 Therefore, physicians must counsel
patients preoperatively to ensure that there is mutual understanding about realistic goals of surgery and—if
there are mental health issues—that these are identified and appropriate referrals made.

Although the primary goal of strabismus surgery is typically ocular realignment,27 patient-specific surgical
success metrics should be based on achieving a goal mutually set by the physician and the patient. This goal
should incorporate the patient’s primary concerns as well as the physician’s determination of which goals are
achievable considering the underlying etiology, previous treatment history, and disease duration. Specific, goal-
determined metrics for surgical success have been suggested for patients with esotropia and exotropia.11,12

These suggested metrics11,12 combine ratings of the following four goals:

• Binocular potential
• Reconstruction of ocular alignment
• Resolution or reduction of diplopia
• Improvement of compensatory head position, or ocular torticollis

In summary, strabismus in adults has profound negative effects on quality of life and many aspects of day-to-
day function. There is excellent evidence that strabismus surgery very often improves quality of life and
function, and there are now questionnaires to assess these aspects of evaluation and treatment.

P338
 Adult Strabismus PPP

Persistent or recurrent childhood strabismus in adults is strabismus that originated in early childhood before
visual maturity (most often defined as before 9 years of age). 5,7,8 This type of adult strabismus arises in several
different ways:
• Childhood strabismus that was untreated
• Decompensation of previously asymptomatic strabismus
• Recurrence of a previously treated strabismus
• Evolution of consecutive strabismus after childhood treatment

The patient population is adults who had onset of strabismus in early childhood.

• Counsel the patient on the diagnosis and treatment options

• Provide goal-directed management of strabismus (to restore eye alignment for improved self-image, restore
binocular vision, if possible, and reduce diplopia or compensatory head posture, if present)
• Inform the patient’s other health care providers about the diagnosis and treatment plan

P339
 Adult Strabismus PPP

Prevalence varies according to setting. The authors of one report found that 60% of cases in their adult
strabismus clinic had recurrent or previously untreated childhood strabismus.5,7

The ocular misalignment in adults with persistent or recurrent childhood strabismus will not resolve
spontaneously. However, signs and symptoms may range in severity from mild (that is psychosocially
acceptable) to large-angle misalignment of the eyes, with or without diplopia.

Psychosocial concerns are among the many reasons for considering surgical intervention in cases of childhood
strabismus presenting or re-presenting in the adult years.4,5,7,11,12,16-19 These include poor self-perception,
impaired social interactions, and poor eye contact. The psychosocial benefits of such intervention have been
extensively reviewed.6,8,20-22,24-26,50 In addition, increased difficulty in maintaining binocular fusion or the
development of increasing compensatory head positioning to maintain binocular fusion, and episodic diplopia
are reasons to consider intervention as well.

Although there is a vast range of presenting patterns, consecutive exotropia after prior surgery for esotropia is a
particularly common form of this type of strabismus. Whereas anisometropia, amblyopia, hypermetropia, and
dissociated vertical deviation (DVD) are risk factors for consecutive exotropia in childhood,51,52 adduction
deficits are more typically associated with consecutive exotropia in the adult population and require a specific
surgical approach.53,54

Adults with a history of childhood strabismus often have a successful outcome from surgery later in life. Over
90% of adults treated with strabismus surgery demonstrate improvement in alignment or in binocular function,
and as many as 25% achieve both after surgery in their adult years. 5 In a study of 359 adults with longstanding
strabismus, an absence of diplopia prior to strabismus surgery, and attainment of postoperative motor alignment
within 8 prism diopters of orthophoria, 86% demonstrated postoperative binocularity based on testing with
Bagolini lenses (striated lenses used to test for suppression and for normal or abnormal retinal
correspondence).16,55 In another study of adults with a history of infantile esotropia and failure to obtain motor
alignment by 2 years of age but who achieved motor alignment within 8 prism diopters of orthophoria as adults,
all achieved peripheral fusion (Worth 4-Dot at near) and nearly half achieved 200 or more arc seconds of
stereopsis.17 It is not always possible to predict which patients will acquire sensory fusion. However, patients
with a history of congenital esotropia and a positive response to prism adaptation often demonstrate
postoperative fusion along with good motor alignment. 56 In a review of outcomes of strabismus surgery,

P340
 Adult Strabismus PPP

postoperative diplopia is a concern but, although transient diplopia after surgery is common,9 patients can be
reassured that persistent diplopia after successful surgery occurs in 1% or fewer patients, 4,7,10 even in those
whose preoperative prism testing suggests that it may occur.

Treatment in adult strabismus is goal directed, and targeted outcomes may include:
• Restoration of normal ocular alignment and appearance, and improved self-image
• Restoration of binocular function
• Reduction of diplopia (if present)
• Reduction of compensatory head position (if present)

• Ocular misalignment in an adult with a history of persistent or recurrent childhood strabismus.

• Strabismus angle and direction may be the same as original presentation in childhood or different as
a result of aging of the extraocular muscles or the impact of prior strabismus surgery
• History of chronicity is important. An acute change or newly symptomatic diplopia may warrant
evaluation for other causes such as cranial nerve palsy, thyroid eye disease (TED), or fixation switch
diplopia. (See appropriate subsections of this PPP.) Formal neuro-ophthalmic evaluation may be
indicated in some cases where presentation suggests other neurologic or systemic disease. An attempt
should be made to acquire and review past clinical and surgical records as well as any prior imaging
studies.

The examination should include the following elements:

• A careful review of the current optical correction (glasses) and the presence of any ground-in or
overlay prism, and the impact that the current correction might have on alignment

• Manifest refraction to identify refractive barriers to binocular alighnment or fusion

• Assessment of alignment by light reflex testing (e.g., Krimsky) to compare with cover test
measurements and identification of abnormal angle kappa

P341
 Adult Strabismus PPP

• Dry manifest and cycloplegic refraction, which may reveal anisometropia or high hyperopia, providing
clues to original ocular motor disturbance. Consideration should be given to changing the current
correction, which might resolve strabismus

• Complete motility examination, including cover-uncover and alternate-cover testing as well as testing
for binocular fusion and stereopsis.57 Strabismus patterns suggestive of cranial nerve palsy, skew
deviation, or the presence of new-onset nystagmus, proptosis or inflammatory changes may indicate
central nervous system or endocrine pathology (see sections on cranial nerve palsies, skew deviation
and thyroid eye disease). These disorders can, of course, impact the population of patients with
unrelated childhood strabismus and warrant further evaluation and potentially different treatment.

• Inspection of the ocular surface for conjunctival scars (prior incision sites) and exposure of the thinned
sclera behind anatomical insertions (evidence of likely muscle recession), because inspection can
provide clues indicating details of prior extraocular muscle surgery

• Inspection of the interpalpebral fissures for evidence of possible prior vertical or horizontal rectus
muscle resection (smaller interpalpebral fissure) or recession (larger interpalpebral fissure)

• Prism testing to simulate the desired postsurgical alignment and the range of overcorrection and
undercorrection comfortably tolerated and unlikely to result in diplopia. With adult strabismus
originating in early childhood, however, the response to prism can be misleading owing to frequent
coexistent anomolous retinal correspondence, which often changes postoperatively. Often a patient
will have diplopia when the misalignment is corrected with prism but will rapidly readapt to the same
correction when induced by surgery, resulting in a very low incidence of new longstanding
postoperative diplopia.10
• Assessment for ocular torsion by sensory testing or by anatomic evidence of torsion noted during
indirect opthalmoscopy, particularly in patients with vertical strabismus. Fortunately, it is very rare for
torsional diplopia to be problematic in adult patients with early childhood onset strabismus due to
either suppression or torsional anomalous retinal correspondence, which nearly always readapts
postoperatively.
• Imaging, such as computed tomography (CT), magnetic resonance imaging (MRI), or orbital
ultrasound may aid in localization of previously operated extraocular muscles, although nearly all these
cases can be managed without imaging using careful preoperative and intraoperative assessment.

P342
 Adult Strabismus PPP

The patient should be monitored/observed if symptoms are mild, occasional, and well tolerated or if the
patient is opposed to treatment.

For refractive error, consider whether alignment might be improved using all the information gained from
dry and cycloplegic refraction, for example, by correction of hyperopia and appropriate bifocal or
progressive lenses for adults approaching presbyopia. Reversal of monovision may be necessary and may
resolve symptoms for some patients (see Section IIm. Fixation Switch Diplopia). The impact of this
intervention can often be appraised in-office.4 Prisms to address some forms of diplopia and orthoptic
exercises to address some forms of diplopia and asthenopia can also be considered in some cases.

Correction of childhood strabismus in adults is generally surgical but, because a broad range of conditions
may be responsible, the specifics of the surgical treatment will vary. However, there are often sequelae of
previous surgery (such as underaction of a recessed muscle, restriction of a scarred muscle, or unsightly
conjunctival scarring) that should be addressed to optimize the postoperative alignment. Many patients,
particularly those with a history of infantile esotropia, have poor fusion along with accompanying sensory
adaptations such as suppression and anomalous retinal correspondence. As adults, the majority readily
adapt to the new ocular alignment resulting from strabismus surgery, although some may require more time
to adapt to this change.

Diagnosis and management of persistent or recurrent childhood strabismus require the training and clinical
judgment of an experienced ophthalmologist. Working under the supervision of an ophthalmologist, orthoptists
can be an asset in the examination, diagnosis, and nonsurgical management of adult strabismus. Surgical
treatment of childhood-onset strabismus in adults is often challenging because of pre-existing surgical scarring,
uncertainty about extraocular muscle attributes and location, possible limited fusional skills, and (in rare cases)
impaired ability to adapt to new alignment.

The ophthalmologist should discuss the findings, explain the disorder, provide a diagnosis, and discuss
management options with the patient and any caregivers.

P343
 Adult Strabismus PPP

Sensory strabismus denotes an ocular misalignment in the setting of vision loss in one or both eyes. There are
myriad causes of sensory strabismus, which include the following:
• Congenital structural ocular defects (i.e., optic nerve hypoplasia)
• Amblyopia from anisometropia or deprivation (i.e., pediatric cataract)
• Acquired vision loss from ophthalmic disease or trauma (e.g., glaucoma, retinal detachment, or optic
atrophy)

The patient population is adults with sensory strabismus.

• Counsel the patient on the diagnosis and treatment options

• Provide goal-directed management of strabismus (reconstruction of ocular alignment and, in some cases,
restoration of binocular function and reduction of diplopia)
• Inform the patient’s other health care providers about the diagnosis and treatment plan

About 7% of adults over the age of 60 who have been treated with strabismus surgery have sensory strabismus.7

Loss of visual acuity in one or both eyes and loss of binocular function often results in compromised alignment.
Whereas esotropia is more common in early childhood, later onset vision loss is more typically associated with
exotropia. Sixty-nine percent of this population present with exotropia versus 31% who present with esotropia.58
Coexisting vertical misalignment is not uncommon.59,60

P344
 Adult Strabismus PPP

Although improvements in binocular function are uncommon after the surgical treatment of sensory strabismus,
there are significant benefits to patients following ocular realignment. In esotropic patients, ocular realignment
can yield improvement in peripheral vision and visual field expansion.8 Also, enhanced emotional health, self-
esteem, employment opportunities, and social interactions have been well documented after the correction of
sensory strabismus.5,6,20-22,24-26 Long-lasting improvement in sensory strabismus is common despite the lack of
binocular potential in most patients.23

• Restoration of normal ocular alignment and appearance, with improved self-image

• Potential for improvement in other psychosocial domains

A detailed medical and ocular history, including specific questions about patient symptoms and perception
of ocular misalignment, may be helpful.

Complete sensorimotor examination should be performed, and misalignment should be measured by prism
and alternate cover test, if possible. However, in the setting of significant vision loss, an estimation of
binocular alignment is often best accomplished using the Krimsky, modified Krimsky, or Hirschberg tests
at distance and near.57 Misalignment is appraised when the better sighted eye is fixated on an
accommodative target, because that is often the angle that warrants repair with surgical intervention. If
there is eccentric fixation with the poorer seeing eye, the prism and alternate cover test may be misleading.
A modified Krimsky or Hirschberg test, using a muscle light at 1/3 meter, may better guide surgical
management.

The presence of treatable causes of vision loss should be determined prior to strabismus surgery and treated
as possible or appropriate. Sometimes this will improve the ability to maintain normal alignment as may
occur after cataract extraction in a patient with a history of prior excellent binocular fusion. At other times,
treatment may only increase symptoms of diplopia and patients may defer or delay this intervention in
order to minimize diplopia awareness.

P345
 Adult Strabismus PPP

The patient should be monitored/observed if symptoms are mild, occasional, and well tolerated or if the
patient is opposed to treatment.

Chemodenervation (botulinum toxin) can yield long-term improvement in ocular alignment in some
cases.61

Surgical treatment for sensory strabismus includes the complete realm of current strabismus surgical
techniques, with some surgeons favoring the use of adjustable sutures.7 Patients with sensory strabismus
may also suffer from asymmetric eyelid or globe position (i.e., ptosis, proptosis, enophthalmos, or
exophthalmos). In some cases, these eyelid abnormalities may be iatrogenic and associated with prior
orbital or retinal surgery or trauma. In such cases, use of large extraocular muscle recessions can also
improve concomitant ptosis or enophthalmos via its effect on globe position within the orbit. Similarly,
large extraocular resections can reduce proptosis and exophthalmos by repositioning the globe within the
orbit. Such dual benefits can be particularly important to patients who may not have the means to afford
elective oculoplastic surgery.62

Diagnosis and management of sensory strabismus requires the training and clinical judgment of an experienced
ophthalmologist. Working under the supervision of an ophthalmologist, orthoptists can be an asset with the
examination and diagnosis of sensory adult strabismus.

The ophthalmologist should discuss the findings, explain the disorder, provide diagnosis, and discuss
management options with the patient and any caregivers. The risks of strabismus surgery in patients with
sensory strabismus are like those of any patient undergoing strabismus surgery and include the remote risk of
scleral perforation, new onset diplopia, and (in select cases) disruption of extraocular muscles or dislodging of
orbital implants. (See Section IIg. Iatrogenic Strabismus.)

P346
 Adult Strabismus PPP

Convergence insufficiency is a binocular disorder associated with eyestrain, headaches, blurred vision,
horizontal diplopia, or a need to close one eye to avoid diplopia while reading or engaged in near activities. 50
Difficulty concentrating, movement of print, and loss of comprehension after short periods of reading are
described as well.50 It is also associated with exophoria or exotropia at near.63

The patient population is adults with convergence insufficiency.

• Counsel the patient on the diagnosis and treatment options

• Provide goal-directed management of symptoms (reduction of asthenopia, improved reading, binocular
vision, and/or reduction of diplopia)
• Communicate with the patient’s other health care providers on the diagnosis and treatment plan

Convergence insufficiency has an annual incidence of 8.4 per 100,000 people and represents 15.7% of new-
onset adult strabismus cases. Though the median age of new-onset adult cases is 69 years old, some adults
present after decompensation of convergence insufficiency that began earlier in life. 64,63

A history of concussion,65 recurrent sub-concussive events,66 and certain central nervous system disorders such
as Parkinson’s disease67 are associated with convergence insufficiency.

In general, convergence insufficiency does not improve, but symptoms may vary as demands for near work
change over time. Cases of convergence insufficiency secondary to concussion may improve over time.65,68

P347
 Adult Strabismus PPP

Treatment for convergence insufficiency is offered to ameliorate asthenopia and intermittent diplopia at near
and to enhance ability to perform activities requiring binocularity at near.

Treatment of convergence insufficiency in adult strabismus is symptom directed, and targeted outcomes may
include the following:
• Resolution of asthenopia or increasing comfort with reading and near work
• Control of diplopia at near
• Improvement in binocular function at near

The diagnosis of a patient with convergence insufficiency includes a comprehensive examination of adult
strabismus patient and the following:
• Detailed sensorimotor evaluation
• Assessment of refractive status
• Dilated fundus exam

A detailed medical history should include information/medical records on prior ocular surgery and diseases
or trauma to the central nervous system. Assessment and quantification of patient symptoms can be achieved
using the Convergence Insufficiency Symptom Survey and/or the Diplopia Questionnaire. 69

A sensorimotor exam in the presence of convergence insufficiency demonstrates the following: 64,70,71

• Exophoria greater at near than distance

• Insufficient positive fusional vergence at near (<20 PD mean positive fusional vergence blur or failing
Sheards’ criterion; mean positive fusional vergence measured less than twice the near phoria
magnitude)
• Distant near point convergence (> 6 cm)

It should always be verified that the optical correction is appropriate for near activities.

P348
 Adult Strabismus PPP

The patient should be monitored/observed if symptoms are mild, occasional, and not bothering the patient
or if the patient is opposed to treatment.

Convergence orthoptic exercises are often helpful. Office-based training appears to be more effective than
home-based exercises in children. However, in young adults (19–30 years old) office-based training was
found to be more effective than home-based training at improving positive fusional vergence but not near
point of convergence or patient symptoms.50,72 In general, results of convergence training are less consistent
in the adult population than in children.72 Nonsurgical treatment can also include prism reading glasses.73

Strabismus surgery may be helpful in some cases, particularly when a manifest distance deviation is
present, symptoms are consistent, and prism reading glasses are not satisfactory. Lateral rectus recession
(based on distance deviation) and/or medial rectus resection (based on near deviation) have been reported
to be successful for surgical management,74,75 although other options (including chemodenervation,76
slanting insertions and augmentation of lateral rectus recessions) have been studied and have possible
advantages in certain populations.77

Diagnosis and management of convergence insufficiency requires the training, clinical judgment, and
experience of an ophthalmologist familiar with this diagnosis, and treatment may benefit from the assistance of
an orthoptist or an optometrist.

The ophthalmologist should discuss the findings, explain the disorder, provide a diagnosis, and discuss
management options with the patient and any caregivers.

P349
 Adult Strabismus PPP

Acquired strabismus related to aging and myopia focuses on forms of strabismus typically associated with aging
of the extraocular muscle and pulley complex and with exacerbation of this in the myopic eye—divergence
insufficiency, sagging eye syndrome, and strabismus fixus (“heavy eye”).

Divergence insufficiency is characterized by an esotropia that increases with distance fixation and may also
be known as age-related distance esotropia. Patients present with an acquired comitant esotropia that is at
least 10 prism diopters larger at distance than at near78 and complain about progressive or subacute-onset
double vision at distance (e.g., while driving or watching television). They rarely have diplopia at near.
Primary divergence insufficiency may be associated with decompensation of a long-standing esodeviation,
loss of fusion, or mechanical changes to the extraocular muscles. 78 Secondary divergence insufficiency is
associated with neurological disease and typically caused by a lesion to a supranuclear divergence center,
subtle sixth nerve palsy, neurologic abnormality associated with pontine tumors, or head trauma.79,80 (See
Section IIi. Abducens Palsy.)

Some cases of divergence insufficiency may be seen in association with an age-related degeneration of the
lateral rectus-superior rectus (SR-LR) connective tissue band known as sagging eye syndrome.81 This band
typically maintains a fixed distance between the lateral rectus and the superior rectus as they course
posteriorly, but age or other degeneration can result in a medial shift of the superior rectus and an inferior
shift of the lateral rectus due to failure of this connective tissue band. As the lateral rectus sags downward,
abduction becomes impaired and esotropia results. The esotropia is often greater at distance than at near.
This downward shift of the lateral rectus muscle may be responsible for an often-associated small-angle
hyopotropia, although other extraocular muscle and pulley abnormalities have been reported in addition to
the inferior sagging of the lateral rectus and the disruption of the SR-LR band.81,82

Sagging eye syndrome also shares some imaging characteristics with myopic strabismus fixus; however,
myopic strabismus fixus results in more extreme displacement of the lateral rectus, nasal displacement of
the superior rectus muscle, and superotemporal prolapse of the globe.

P350
 Adult Strabismus PPP

Strabismus fixus is progressive large-angle esotropia and hypotropia with a limitation in ocular rotation in
patients with longstanding pathological high myopia. As a result of the increased axial length, there is
development of a staphyloma or a staphylomatous-like change in globe dimensions that results in rupture of
the SR-LR band. An inferior shift of the lateral rectus muscle83 and a nasal shift of the superior rectus
muscles ensue.84,85 86,87 Globe dislocation (prolapse) into the superotemporal orbit and outside the confines
defined by the lateral rectus and superior rectus occurs. Ocular rotations in myopic strabismus fixus are
much more severely altered than in sagging eye syndrome; there is a component of mechanical restriction
not seen in sagging eye syndrome, likely secondary to secondary medial rectus contracture as well as globe
prolapse into the superotemporal orbit. Not at all cases of axial high axial myopic result in strabismus fixus.
In some cases the globe elongation is axial only, superotemporal globe dislocation does not occur, and the
patient may be treated as if he or she has classic divergence insufficiency.86

The patient population is adults with divergence insufficiency, sagging eye syndrome, or strabismus fixus.

• Counsel the patient on the diagnosis and treatment options

• Provide goal-directed management of strabismus (restoration of binocularity and control of diplopia and,
when severe, reconstruction of normal appearing alignment, improved eye contact, and self-image)
• Communicate with the patient’s other health care providers about the diagnosis and treatment plan

The annual incidence of 6.0 per 100,000 people represents 10.6% of new-onset adult strabismus cases.63
The median age of presentation is 74 years.63 Divergence insufficiency typically presents after 50 years of
age, and it frequency increases with age.63

The prevalence of sagging eye syndrome as a distinct variant of divergence insufficiency is unknown
because the clinical description is relatively new, and not all practitioners distinguish this entity from the

P351
 Adult Strabismus PPP

more general descriptor, divergence insufficiency. Sagging eye occurs more frequently in myopic patients,
and the most common age of participants in research studies is 66 to 69.81,88,89

The prevalence globally is unknown but has been estimated to be 2.65% among high myopes in Japan.90
Progressive strabismus fixus is most often associated with middle or older age, with less severe divergence
insufficiency predating this progression.91,92 Severe myopia with an axial length of more than 27mm is
characteristic.83,84,87

The strabismus and diplopia caused by divergence insufficiency, sagging eye syndrome, and strabismus fixus
will not resolve spontaneously. Strabismus fixus is often the most progressive in severity.86,92

Diplopia, the inability to make eye contact, and severe ocular misalignment preventing proper eye examinations
or other needed ophthalmic surgery (in the case of strabismus fixus) warrant treatment.

Treatment of adult strabismus is symptom directed, and targeted outcomes may include:
• Reduction of diplopia
• Improving/returning binocular vision
• Reconstruction of ocular alignment
• Reduction of compensatory head position (if present)

Patients present with a history of subacute onset of binocular diplopia from esotropia, worse at distance,
and sometimes associated with modest vertical misalignment. This group of disorders is progressive, and
may begin to impact alignment at near. Detailed medical history should include the use of prism correction
in glasses.93 Symptoms first become apparent with distant fixation (e.g., while driving) and most commonly
after age 50 years for classic divergence insufficiency 63 . Symptoms are similar for sagging eye syndrome,
with associated small-angle vertical and torsional diplopia at times; presentation with sagging eye
syndrome is more common in the seventh decade of life or older. Associated changes of blepharoptosis

P352
 Adult Strabismus PPP

(29%), deep superior lid sulcus defect (64%) and a history of blepharoplasty, brow lift, or facelift surgery
has been noted in some cases.81 Heavy eye or myopic strabismus fixus is associated with very high axial
myopia of more than 27mm. Onset may begin in the earlier decades of adult life with chronic progression
of a more extreme esotropia and possible hypotropia often associated with mechanical restriction.83,84,87

• More esotropia at distance than at near (usually 10 prism diopters or more)

• Similar esotropia in right and left gaze in the absence of pontine or other neurological pathology
or significant abduction nystagmus
• No limitation on forced duction under anesthesia

• Esotropia, often greater at distance, and sometimes accompanied by a small vertical deviation and
cyclotropia81,82
• Deviations are nearly comitant, except for possible distance-near disparity
• Mild to no limitation of ocular rotation
• Associated changes of blepharoptosis (29%), deep superior lid sulcus defect (64%) with a
previous history of blepharoplasty, brow lift, or facelift surgery in some cases81

• Severe myopia, typically -8.00 diopters or more but often much greater
• Axial length of more than 27 mm
• Large-angle ocular deviation of esotropia with or without hypotropia
• Limitation of ocular rotations such as abduction or elevation
• An SR-LR dislocation angle of 121° ± 7° (measuring the angle between the centroids of the
superior rectus muscle and lateral rectus muscle on quasicoronal imaging)86

A complete ocular motility examination for these syndromes should include cover testing at near and
distance in primary and secondary gaze positions and evaluation of torsion for patients with vertical
misalignment. In patients with myopic strabismus fixus there may be limited rotations and
measurements can be obtained only with the modified prisms light reflex (Krimsky) test.
High-resolution orbital imaging is useful for evaluation of patients with likely sagging eye or myopic
strabismus fixus to look for evidence of disruption of the LR-SR band. This characteristic is best seen
on coronal T1-weighted images obtained without fat suppression. The lateral rectus sags downward,

P353
 Adult Strabismus PPP

creating a large angle between the lateral rectus and superior rectus (typically 104° ± 11) for patients
with sagging eye syndrome. 86
There is fat prolapse within this potential space.

In contrast, for patients with myopic strabismus fixus (“heavy eye syndrome”) the angle between the
lateral rectus and the superior rectus is much larger (described as 121° ± 7°) and there is frank
prolapse of the globe within this space. The globe itself demonstrates axial elongation and often
posterior enlargement.81-84 ,86

Brain imaging and further neurologic workup is indicated if the patient presents with significant
abduction nystagmus and lateral incomitance suggesting the alternative diagnosis of sixth nerve palsy,
or with signs or symptoms of elevated intracranial pressure such as headache and papilledema or other
neurologic decline which may result in 6th nerve palsy. (See Chapter IIj)

Conditions resembling myopic strabismus fixus and presenting with restrictive esotropia with or
without hypotropia include orbital trauma, chronic complete 6th nerve palsy, thyroid eye disease and
acquired fibrosis of the extraocular muscles but are readily distinguished based on history and
characteristic orbital imaging.

The patient should be monitored/observed if symptoms are mild and occasional or if the patient is
opposed to treatment.

Fresnel or ground-in prisms can be temporary (prior to surgical correction) or long-term treatment
options.93 Patients should be counseled that the magnitude of prism correction needed to control
diplopia may increase over time and that the ability to control the diplopia when glasses are not used
will likely decline. There is little evidence to support the use of fusional exercise for the treatment of
divergence insufficiency. In fact, previous studies have reported no benefit. 94

Lateral rectus resection95 as well as medial rectus recession95,96 have proven successful at resolving
divergence insufficiency esotropia. Surgery can be performed bilaterally or unilaterally.78,97

P354
 Adult Strabismus PPP

Typically, a greater surgical dose (than indicated in the standard tables) is needed for this type of
esotropia. Therefore, some surgeons find adjustable sutures very useful with a target angle of a 2 to 4
prism diopters of exophoria in the distance.

It is useful to check convergence amplitudes at near prior to surgery. The presence of robust
convergence amplitudes at near generally protects the patient from developing postoperative diplopia
with convergence insufficiency at near.96

The patient should be monitored/observed if symptoms are mild and occasional or if the patient is
opposed to treatment.

Fresnel or ground-in prisms are helpful for patients with new-onset deviation, intermittent or constant
and small in angle.

Surgical prognosis for this group of patients is quite favorable and can be addressed using a variety of
techniques, including marginal insertional tenotomies,89 recessions, resections, and plications.98
Surgical resolution or reduction of diplopia has been particularly beneficial for patients who don’t
otherwise need glasses for distance vision.88 As described for patients with divergence insufficiency
unrelated to sagging eye syndrome, more than typical recession of the medial rectus muscles is usually
needed. 96

The patient should be monitored/observed if he or she is opposed to treatment.

Prisms are not typically an option to correct the baseline strabismus but may be used postoperatively in
patients with residual deviation

In general, surgery is the only alternative to correct the large-angle esotropia and frequent hypotropia
and to normalize ocular rotations. Referral to a retina specialist for preoperative evaluation should be
considered to ensure that progressive myopic degeneration is not associated with active retinal

P355
 Adult Strabismus PPP

pathology requiring treatment prior to strabismus surgery. Measuring the angle between the SR
muscle the LR muscle (dislocation angle) is helpful in determining the severity of the disease and for
planning the surgery.83-85 A forced duction test to determine if limitation to ocular rotation is the
result of a medial rectus muscle restriction or, more rarely, an orbital mechanical restriction resulting
from the increased axial length and adjacency of the globe to the lateral orbital wall should be
performed at the onset of surgery. If the forced duction test confirms restriction to abducting rotation,
the medial rectus muscle should be recessed. However, some orbital mechanical restriction may be
present even after medial rectus recession. In such cases, restriction to abduction may persist, limiting
the outcome of the surgery.87,92,99,100 Medial rectus recession is typically accompanied by one of
several muscle union procedures that join the superior rectus to the lateral rectus to reposition the
muscle paths. The classical approach is the Yokoyama procedure, a vessel-sparing suture union of the
superior rectus and lateral rectus 15 mm posterior to their insertions, typically accompanied by a
medial rectus recession.87 Many variants have been described, including union by a silicone sleeve101
and the use of a three-suture lateral rectus-superior rectus union to successfully reduce more extreme
degrees of esotropia.102

In some cases of very high axial myopia and esotropia, prolapse of the globe beyond the confines of
the superior and lateral rectus may not be present. Such cases can be surgically managed like sagging
syndrome or basic divergence insufficiency and do not typically require loop myopexy to resolve
misalignment and diplopia.86

Diagnosis and management of divergence insufficiency and associated sagging eye syndrome and myopic
strabismus fixus benefit from the training and clinical judgment of an experienced ophthalmologist. Working
under the supervision of an ophthalmologist, orthoptists can be an asset in the examination, diagnosis, and
nonsurgical management when appropriate.

Magnetic resonance imaging and referral for central nervous system evaluation should be considered if
other signs, such as abduction nystagmus, papilledema, or neurologic decline, are present.

Magnetic resonance imaging is often helpful in confirming the diagnosis of strabismus fixus. Referral to a
retina specialist for preoperative evaluation should be considered to ensure that progressive myopic
degeneration is not associated with active retinal pathology that, untreated, might increase risk of post-

P356
 Adult Strabismus PPP

operative retinal tear or detachment. Exposure keratopathy should be monitored and treated in cases of
associated exophthalmos sometimes seen with high axial myopia or referred to an oculoplastic specialistfor
more definitive care. The ophthalmologist should discuss the findings, explain the disorder, provide a
diagnosis, and discuss management options with the patient and any caregivers.

P357
 Adult Strabismus PPP

Thyroid eye disease is an autoimmune disorder characterized by congestion and inflammation of the orbit and
surrounding tissues.103 Typical ocular findings include soft tissue congestion with enlargement of the preseptal
fat pads, eyelid retraction, exophthalmos, restrictive extraocular myopathy, and optic neuropathy. 104 Vision can
be compromised from corneal exposure or optic neuropathy or as a result of development of diplopia secondary
to congestive and restrictive extraocular muscle myopathy.

The patient population is adults with strabismus and thyroid eye disease.

• Counsel the patient on the diagnosis and treatment options

• Carefully monitor patients at risk for vision loss secondary to compressive optic neuropathy due to
crowding at the orbital apex or severe exposure keratopathy from eyelid retraction with or without
exophthalmos
• Manage binocular diplopia, compensatory head posture, and sometimes associated elevation in IOP
• Refer to an orbital specialist if the patient experiences vision-threatening disease, moderate-to-severe
exophthalmos and/or significant orbital pain related to thyroid eye disease
• Inform the patient’s other health care providers about the diagnosis and work collaboratively on the
treatment plan

The incidence of TED in an adult Caucasian population is 16 per 100,000 per year in women and 2.9 per
100,000 per year in men.105

Thyroid eye disease can occur at any age, but onset is most often in the fourth to fifth decade of life.106 There is
a predilection for TED among women (8:1). 107 Risk of TED causing restrictive strabismus increases with

P358
 Adult Strabismus PPP

age.108,109 Smoking, diseases resulting in reduced oxygen saturation (such as emphysema), and exposure to
ionizing radiation increase the risk for and severity of orbitopathy.110,111 A history of prior orbital
decompression is associated with an increased risk of strabismus,112 an association that worsens with age.108
Radioactive iodine treatment, high anti-thyroid antibody titer and serum vitamin D deficiency are independent
risk factors for developing TED.113,114

Between 30% and 50% of patients with TED develop restrictive myopathy. Ocular motility is restricted initially
by inflammatory edema and later by fibrosis. The most frequently affected muscle in TED is the inferior rectus,
followed by the medial rectus,115 although more global extraocular muscle involvement is common.

Motility impairment causing diplopia and compensatory head posture is a frequent manifestation of TED. This
impacts visual function, affects quality of life, and can have profound socioeconomic consequences.116

Treatment of thyroid eye disease is symptom directed, and targeted outcomes may include the following:
• Recognition and prevention of primary vision loss due to optic neuropathy, exposure keratopathy, and
elevation in intraocular pressure (IOP)
• Restoration of ocular alignment
• Reduction of diplopia
• Improvement of compensatory head position

The goal of the strabismus surgeon is to re-establish single binocular vision in primary gaze and reading
position and a substantial usable field of single binocular vision.

A detailed medical and ocular history should include specific questions about weight gain or loss,
tachycardia, and unexplained fatigue potentially secondary to hyperthyroidism or hypothyroidism.
Strabismus is likely to be associated with diplopia and, in some cases, a compensatory head posture (i.e.,

P359
 Adult Strabismus PPP

ocular torticollis, often a chin-up position in this population). Because the incidence of myasthenia is
increased in patients with TED, assessment for the coexistence of myasthenia is suggested if indicated by
findings on clinical examination.

The examination should include a full sensorimotor examination, noting evidence of mechanical restriction.
Forced ductions can confirm restriction but classic features, such as fixation duress (brow elevation and
augmentation of eyelid retraction on attempted ocular elevation), may confirm the restrictive nature of the
associated strabismus. Hypotropia of the affected eye(s) with esotropia is the most frequent deviation
owing to involvement of the inferior and medial rectus muscles, although other rectus muscles are
commonly involved, resulting in other strabismus patterns. More global extraocular muscle involvement
may be apparent on orbital imaging or with forced duction testing.117,118

The examination should also include the following elements:

• Binocular field testing to map the region of single binocular vision. Duction deficits can be measured
and monitored using uniocular fields of fixation.
• Measurement of the exophthalmos by exophthalmometer, which is important diagnostically and to
follow clinical course.
• Surveillance for optic neuropathy, paying close attention to acuity, color vision, visual fields, pupillary
exam, and fundus exam (for disc edema or optic atrophy).
• Ocular coherence tomography and Humphrey visual fields, which provide ancillary information and
are used to screen for or monitor possible associated compressive optic neuropathy.

Screening and management of possible associated optic neuropathy may fall within the expertise of the co-
managing orbital specialist, e.g. neuro-ophthalmologist or oculoplastic surgeon.

Orbital CT or MRI can be performed to confirm the diagnosis and evaluate the size of the extraocular
muscles and the volume of orbital fat. Imaging features include tendon-sparing enlargement of one or more
of the extraocular muscles and proptosis. Muscle involvement and proptosis are often bilateral but
frequently asymmetric.118 Imaging may also help determine whether the superior oblique is involved and
demonstrate the degree of orbital apex crowding. The size of the rectus muscles as measured by CT does
not however directly correlate with the degree of muscle dysfunction.119

Thyroid function and associated immune studies may be ordered, although this is typically handled by the
associated primary care or endocrinology specialist; patients may be hyperthyroid, hypothyroid, or
euthyroid at the time of presentation with TED.120

P360
 Adult Strabismus PPP

Ocular lubricants are almost always needed to combat exposure related to eyelid retraction and proptosis.
Punctal plugs can also be useful. Patients should be counseled to cease smoking.107

Selenium supplementation in patients with selenium deficiency has been shown to reduce some inflammatory
symptoms in patients with milder TED,121 perhaps in part through an antioxidant effect.122 It does not impact
control of hyperthyroidism in populations, at least in populations that are not selenium deficient.123 Treatment
with teprotumamab, a human monoclonal antibody inhibitor of IGF-IR demonstrated to attenuate the actions of
IGF-I, thyrotropin, thyroid-stimulating immunoglobulins, and immunoglobulins in patients with Graves’
disease reduces proptosis and the clinical activity score of disease in many with active disease.124

Orbital decompression, high-dose steroids, or radiation treatment should be considered, as indicated, for severe
proptosis or optic neuropathy.125 In patients with concurrent proptosis and significant strabismus, the predictable
worsening of proptosis following recession of fibrotic muscles may warrant consultation with an orbital
specialist to consider orbital decompression surgery prior to strabismus surgery. If orbital decompression is
indicated, strabismus repair should be delayed until after the decompression because new onset or exacerbated
strabismus may result.126-128 Shift of the extraocular muscles and the globe into newly available space, or, in
some cases, disinsertion and reinsertion of the inferior oblique muscle to improve exposure, can significantly
impact alignment.129-131 There is some evidence that preserving the orbital strut may diminish the risk of new-
onset strabismus.126,132

133

Patients can be observed if there is no diplopia in primary gaze or reading position and should be observed
if the angle of strabismus is not yet stable. Observation is also necessary if the patient requires orbital
decompression surgery, as this procedure often changes ocular alignment.134,135

Fresnel or ground-in prism can provide temporary relief from diplopia while awaiting definitive treatment
or may be suitable for modest residual deviation after surgical correction.107 Chemodenervation during the
active phase of the disease is advocated by some for temporary relief of diplopia. Its use may favorably
reduce the misalignment left after the initial inflammatory stage of the disease.136-139

Surgery is required in most patients with persistent diplopia in primary or reading positions of gaze. This is
usually undertaken when the inflammatory stage has subsided and the angle of deviation has been stable for
at least 6 months,118,140 but earlier surgery in select patients can also have satisfactory outcomes and shorten
the period of disability.141 Quiescence is usually determined on clinical grounds.142 Less frequently, MRI

P361
 Adult Strabismus PPP

sequences T2, FSE or fast spin echo, postcontrast T1 (T1Gad) signal intensity ratios and normalized-
apparent diffusion coefficient (n-ADC)143 and water content on STIR (Short TI Inversion Recovery)
sequences may be used to determine if the disease has become quiescent.144

Recessions of the restricted muscles are the mainstay of surgical correction. Resection is generally avoided
in restrictive disease out of concern for further reducing ductions; 118, but can be a useful adjunct in select
cases,145 particularly when extremely large recessions have not fully corrected the alignment.

The rectus muscles of the thyroid patient can be severely tight. Thyroid eye disease is the most commonly
identified risk factor for development of “pulled in two syndrome,”146 an unplanned horizontal transection
of a rectus muscle that seems to spontaneously occur while it is hooked during surgery. Because this
transection takes place typically 4 to 9 mm or further from the anatomical insertion, recovery of the already
taught distal end can prove quite challenging.146 Care should be taken during surgery to avoid this
complication.

Surgery for horizontal deviations is more successful than surgery for vertical strabismus (84% vs. 66%
success after the primary surgery).140 Because the orbitopathy often affects the inferior rectus muscle,
vertical deviations are more common.147 Recession of the inferior rectus muscle is frequently complicated
by consecutive hypertropia, in part, because of suboptimal contact of the recessed muscle with the globe,148
ipsilateral superior rectus involvement,149-151 or imbalanced contralateral disease.152,153 Use of the Kushner
semi-adjustable suture technique may help minimize unanticipated muscle drift,150 as may intentional
undercorrection of the inferior rectus recession 117 with the use of adjustable sutures,154 wider spreading of
the muscle to prevent sagging of the center of the insertion over time,151 and the use of permanent polyester
sutures. Large bilateral inferior rectus recession may result in an A-pattern with unanticipated exotropia in
downgaze as well as intorsion. This occurs because the superior oblique becomes the dominant
infraductor;107,112 and may be exacerbated by primary involvement of the superior oblique muscle in the
disease process.155 Because managing vertical deviations in TED is complex, a step-by-step approach
appraising preoperative and intraoperative forced ductions of all extraocular muscles and attention to
preoperative and intraoperative ocular torsion may significantly minimize the risk of an undesired
postoperative result.151,155 Recession of the inferior rectus muscle, even with careful dissection from the
lower lid retractors, may result in lower lid retraction. Releasing restriction with recession of rectus muscles
may result in increased proptosis of the globe. 156 Preoperative counseling with the patient and
communication with the treating orbital surgeon are important in this regard. Recession of tenon’s capsule
from overlying conjunctiva may augment the effect of rectus muscle recession and in turn improve post-
operative ductions and the range of single binocular vision in select cases.157,158

P362
 Adult Strabismus PPP

Surgeons generally prefer an absorbable suture (6–0 polyglactin) for isolating the muscle and reattaching it
to the globe. Some surgeons favor non-absorbable sutures for large recessions of the inferior rectus muscle
because they may reduce the risk of postoperative overcorrection.120 The role of adjustable suture surgery
in TED remains controversial. Some surgeons claim more successful results,159 whereas others never use
the technique because of concern over possible overcorrection and/or muscle slippage.160 Some surgeons
advocate a “relaxed muscle positioning technique” where the muscles are recessed to a position where they
rest freely on the globe without tension.161 In patients with concurrent proptosis and significant strabismus,
the predictable worsening of proptosis following recession of fibrotic muscles may warrant consultation
with an orbital specialist for orbital decompression surgery prior to strabismus surgery. 156

Diagnosis and management of strabismus in a patient with TED requires the training and clinical judgment of
an experienced ophthalmologist comfortable with very complex strabismus and managing severe restrictive
disease. Working under the supervision of an ophthalmologist, orthoptists can be an asset with the examination,
diagnosis, and nonsurgical management of TED.

A multidisciplinary approach to TED treatment is recommended, combining the expertise of endocrinology,

oculoplastics, and neuro-ophthalmology as indicated. The ophthalmologist should discuss the findings, explain
the disorder, provide diagnosis, and discuss management options with the patient and any caregivers.

P363
 Adult Strabismus PPP

Strabismus after orbital trauma is common and may be present along with other ophthalmic, medical, and
surgical conditions. Although all life-threatening and vision-threatening conditions need to be addressed before
treating strabismus, the presence and pattern of ocular misalignment may indicate other orbital or central
nervous system trauma guiding primary triage, imaging, and treatment.

Determining the cause of diplopia after orbital trauma requires a detailed history, examination, and, often,
imaging. The etiology of the strabismus may be multifactorial, including direct damage to the extraocular
muscles and surrounding orbital structures or contemporaneous trauma to the central nervous system. Rectus or
oblique muscle avulsion, partial or complete “loss,” flap tear,162,163 hemorrhage, edema, or paresis may occur in
addition to orbital hemorrhage or edema, soft tissue swelling, and fracture of orbital bones, with and without
entrapment of muscle, fat, or muscle pulleys. Cranial neuropathies, 164 disorders of accommodation or
convergence,165 disruption of fusion, fusional amplitudes, saccades and smooth pursuit, decompensation of
previous heterophoria, and other supranuclear defects may result in diplopia.166

Diplopia is very common after orbital trauma and reported to occur in 58% to 68% of blowout fractures.167-169
Strabismus surgery was required in 7% to 24% of cases in two series of patients with orbital floor
fractures.170,171 Less commonly, strabismus can occur iatrogenically after sinus surgery 172,173 or other periocular
surgeries, including surgery to the eyelids (See Section IIg Strabismus Associated With Other Ophthalmic
Surgery).174

The patient population is adults with strabismus following orbital trauma.

• Counsel the patient on the diagnosis and treatment options

• Provide goal-directed management of strabismus (reconstructive restoration of binocular vision, or
treatment of diplopia or compensatory head posture)
• Inform the patient’s other health care providers of the diagnosis and treatment plan.

P364
 Adult Strabismus PPP

In self-reported population studies, the incidence of trauma in adults ranged from 1.7% to 19.8%.175-177 In all of
these studies, younger age and male sex were identified as risk factors. 175-177

Diplopia may be transient following ocular trauma, but if it has not resolved within 6 months it is unlikely that
it will resolve on its own.

Diplopia, loss of binocular vision, compensatory head position, and inability to make eye contact all warrant
treatment.

A detailed medical and ocular history, including specific questions about symptoms of diplopia, vision loss,
and details on the nature and timing of the trauma are helpful.

A thorough examination should include vision testing, refraction (if possible), pupillary examination, IOP
measurement, confrontational visual field testing, slit-lamp examination, a dilated (if safe) or undilated
fundus examination (with attention to any torsion), testing of facial sensation, and exophthalmometry. The
goal of the primary examination is to rule out any globe injury or sight-threatening injury. Patients with
orbital or cranial trauma may also have decreased vision (secondary to corneal scars, traumatic cataract,
optic neuropathy, or other damage to the retina or other ocular structures) or decreased visual fields. Care
should be taken to note these limitations, as it may help guide test selection during the examination.

A detailed sensorimotor exam should be performed, with attention to versions, ductions, saccades, pursuit,
vergence, and near reflex, along with alignment in multiple gaze positions with attention to primary and
secondary deviations.178 Forced duction and forced generation testing may help distinguish restriction from
paresis of the extraocular muscles.179 Double Maddox rod, Lancaster red-green, Hess screen, or
synoptophore (especially if there is concern for disrupted central fusion) may also be helpful.

P365
 Adult Strabismus PPP

Vital signs looking for any bradycardia or heart block, along with symptoms of dizziness, nausea, vomiting,
or loss of consciousness in a patient with orbital trauma may indicate an entrapped muscle causing the
oculocardiac reflex.180 This can be life threatening and usually requires medical and surgical treatment.

Imaging studies are frequently indicated. Computed tomography rather than MRI is required if there is any
concern about a possible ferrous-metallic foreign body and often provides sufficient information about the
presence of orbital fracture and entrapment. Magnetic resonance imaging provides more precise imaging of
the extraocular muscles and surrounding tissues, including the pulley system, and reduces radiation
exposure. It can sometimes be performed dynamically and provides additional useful information for
planning an intervention.181-184 Occasionally, patients with atypical strabismus demonstrate occult fractures
absent a history of trauma recalled by the patient.185

All life-threatening and vision-threatening conditions need to be treated before the strabismus. In one series
of 379 patients who underwent surgical repair of facial fractures, 5.5% had complete loss of vision in one
eye and 0.8% had complete loss of vision in both eyes.186 Another series of blowout fractures demonstrated
serious ocular injury in 24% of cases.187

Timing of surgical repair of orbital fractures has been debated.168,169 Current recommendations for timing of
repair of isolated orbital floor fractures are as follows:188
• Immediate repair:
o Immediate repair is indicated for patients with CT evidence of an entrapped muscle or
periorbital tissue associated with a nonresolving oculocardiac reflex
o “White-eyed blow-out fracture”189, a form of trap-door fracture with muscle entrapment and
oculocardiac reflex( seen in children) requires urgent repair as well.
o Globe subluxation into the maxillary sinus, a rare occurance, demands immediate surgical
repair.190,191
• Repair within 2 weeks:
o Symptomatic diplopia with positive forced ductions or entrapment on CT and minimal
improvement over time is best repaired within about 2 weeks. Significant fat or periorbital
tissue entrapment can also result in permanent strabismus in the absence of muscle
entrapment.192
o Large floor fractures, hypoglobus, and progressive infraorbital hypoesthesia are also best
addressed within about 2 weeks.
o Early enophthalmos or hypoglobus causing facial asymmetry will not resolve and are best
addressed within about 2 weeks

P366
 Adult Strabismus PPP

• Delayed Repair:
o Restrictive strabismus and unresolved enophthalmos may benefit from delayed repair
• Observation:
o Observation may be considered in cases of minimal diplopia (not in primary or downgaze),
and good ocular motility without significant enophthalmos or hypoglobus.

Even with repair or observation of orbital fractures, strabismus and diplopia can persist. In a series of 54
patients who underwent repair of orbital blowout fractures, 86% had had diplopia preoperatively and 37%
postoperatively.193

Some forms of strabismus after orbital trauma will improve with time, and watchful waiting is often a
reasonable approach. A short burst of oral steroids can hasten recovery and uncover strabismus that
will persist despite resolution of orbital edema/hematoma.

In the absence of muscle entrapment, waiting 4 to 6 months after orbital trauma is advised because
strabismus may resolve on its own unless substantial fat and orbital pulley entrapment suggests benefit
in earlier repair.194 Conservative treatment such as occlusion, filters, Fresnel prisms, and prism glasses
may provide temporary or permanent relief of diplopia.

For patients who do not meet the criteria for early surgical intervention, waiting 4 to 6 months after the
injury is advised to ensure stability of the misalignment prior to repair.171 The goals of surgical
intervention should be clearly discussed with the patient and are often to eliminate diplopia in the
primary position and downgaze and to enlarge the field of binocular single vision. More than one
operation and supplementation with prism glasses may be required. Every situation is unique, and
more than one method of repair is possible.195 Preoperative forced generation and preoperative and
intraoperative forced duction testing is very important.

Fat entrapment can prove nearly as challenging as extraocular muscle entrapment, resulting in fibrotic
and adhesion syndromes not readily relieved with dissection around the involved muscle. 192 Adhesions
and entrapment may extend well into the deeper orbit, out of reach of the strabismus surgeon.
Adhesion of extraocular muscles, particularly to porous implants, can be equally problematic at
times,196,197 as can iatrogenic disinsertion of the inferior oblique at its origin, which may be required to
place an implant along the nasal wall of the orbit.198

P367
 Adult Strabismus PPP

Some general principles include an attempt to recover “lost” muscles whenever possible. 199 If recovery
of a muscle proves impossible or if a muscle is determined to be severely paretic, a transposition
procedure may be indicated. Matching restriction with the use of posterior fixation sutures or Scott
procedures (recess/resect)200 on the unaffected eye are established methods of expanding the field of
single binocular vision. Adjustable sutures are often helpful in these cases.

Because the etiology of the strabismus may be multifactorial and the misalignment incomitant,
complete elimination of diplopia is often difficult, if not impossible.

A multidisciplinary approach may be required and it may be applied in different settings—an emergency
department or inpatient or outpatient facilities. Diagnosis and management of strabismus caused by ocular
trauma requires the training and clinical judgment of an experienced ophthalmologist. Working under the
supervision of an ophthalmologist, orthoptists can be an asset in the examination, diagnosis, and nonsurgical
management.

Consultation or comanagement with comprehensive ophthalmology, oculoplastics, neuro-ophthalmology as

well as plastics, otolaryngology, emergency department, and neurosurgery may prove necessary, depending on
the circumstances of the injury and the timing of the strabismus evaluation.

P368
 Adult Strabismus PPP

This section focuses on strabismus acquired after the following procedures:

• Cataract extraction or Keratoplasty
• Glaucoma filtering
• Scleral buckling
• Pterygium excision
• Blepharoplasty and Ptosis Repair

Strabismus sometimes resulting in diplopia has been reported after virtually any periocular procedure including
major orbital surgery as well as dacryocystorhinostomy. A detailed discussion on management of these causes of
strabismus in the adult population is outside the purview of this PPP. However, determining whether muscle or
nerve damage is responsible and whether paresis versus restriction is the major cause is a common theme that
informs the evaluation and treatment of all forms of strabismus associated with other ophthalmic surgery.201

Iatrogenic strabismus can occur following different procedures or treatments.

Anesthetic myotoxicity following retrobulbar or peribulbar anesthesia for any ocular procedure including
cataract extraction or keratoplasty may occur. Initial presentation is usually transient paresis followed by
deviation of the eye into the field of action of the involved extraocular muscle. The hypertropia or
hypotropia may result from restriction of the involved extraocular muscle when damage is extensive, or as
an overaction pattern due to focal contracture or muscle hypertrophy.202-206 Superior and inferior rectus
muscles are the most commonly injured by regional and local anesthesia, in retrobulbar or peribulbar
blocks,207 but any muscle including the superior and inferior oblique muscles, may be involved as well.208-
211

Surgical manipulation of orbital tissue from a bridle suture under a rectus muscle can also result in small-
angle strabismus due to nicking the rectus muscle, local bleeding, and scar formation. Likewise, temporary
disinsertion and subsequent reattachment of an extraocular muscle in the course of repairing a globe
laceration may result in strabismus. Causes of binocular diplopia after cataract surgery or keratoplasy in

P369
 Adult Strabismus PPP

addition to local anesthetic myotoxity and direct surgical manipulation of the extraocular muscles and
periorbital tissue include: previously undiagnosed TED;212 pre-existing/unmasked binocular disorders; and
fixation switch diplopia. These conditions are addressed in other sections of this PPP.

Motility disturbances occur from a variety of causes, including mass effect from the implant or associated
bleb, scarring, fat adhesion, or incorporation of adjacent extraocular muscles.213 These complications result
in a wide array of motility patterns, including Brown syndrome, partial disinsertion of the superior oblique
muscle, and restriction of gaze in the direction of the adjacent muscle.214-218

Following scleral buckling procedures, the root cause of the resulting strabismus may be multifactorial.
Patients may develop strabismus secondary to loss of fusion, poor vision, aniseikonia, direct trauma to the
muscles, myotoxicity from retrobulbar anesthesia, malposition of a detached muscle, scar tissue formation,
muscle slippage or disinsertion, entanglement of the superior oblique tendon, and mass effect.219-222
Fusional disturbance, aniseikonia, and torsional diplopia may sometimes accompany repair of retinal
detachment, even without scleral buckling. These threats to binocular fusion increase the complexity of
treating the mechanical aspect of strabismus induced by a buckling procedure. 223

New-onset diplopia can occur after pterygium surgery as a result of direct injury to the adjacent medial
rectus or florid scaring associated with recurrence.224-227 This can result in consecutive exotropia from
medial rectus damage or restrictive incomitant esotropia from scarring. 224-227

New onset diplopia can occur after upper lid blepharoplasty, sometimes associated with trochlear
damage,228,229 and after lower lid blepharoplasty, sometimes due to damage to the inferior rectus. 230
Scarring and hemorrhage in adjacent tissue or fat adherence231 has been implicated among other
mechanisms.

New-onset diplopia can occur after ptosis repair as a result of late recognition of previously existing
strabismus. The classic example would be double elevator palsy with a limitation of upgaze that is
recognized only after the associated ptosis is repaired.

P370
 Adult Strabismus PPP

The patient population is adults who have strabismus associated with other ophthalmic surgeries that may
disrupt normal extraocular muscle function or sensory fusion.

• Counsel the patient on the diagnosis and treatment options

• Provide goal-directed management of strabismus (to restore binocular vision, or minimize diplopia or
compensatory head posture, and in some cases for the reconstructive benefit to improve eye alignment, eye
contact, and self-image)
• Inform the patient’s other health care providers, and in particular their ophthalmic specialists, of the
diagnosis and treatment plan

There is a 0.18% incidence of secondary strabismus232 and a 0.23% incidence of secondary strabismus for
cases involving retrobulbar block anesthesia used most often for cataract surgery. Diplopia may be transient
or persist.232

Risk factors that increase the incidence of secondary strabismus after retrobulbar or peribulbar block are
injection by a nonophthalmologist, left eye injection, and the absence of hyaluronidase in the block.233,234
Topical anesthesia is associated with a lower risk of diplopia than regional anesthesia. It is more typically
associated with other factors that threaten binocular stability, such as the choice of monovision for cataract
surgery or fixation switch diplopia.

Fixation switch diplopia occurs if a previously suppressed eye becomes the dominant eye. (See Section Iik.
Oculumotor Palsy.)

Transient strabismus occurs in 1% of patients after trabeculectomy201,235 and in 3% of patients following

glaucoma plate reservoir surgery.201,235,236 Persistent motility disorders can range from 2% to 77%235
following glaucoma plate reservoir surgery depending, in part, on the type of implant used, the length of

P371
 Adult Strabismus PPP

follow-up, and whether the setting was one that afforded an adequate audit of postoperative alignment.
One center reported a prevalence of binocular diplopia of 21% in those treated with plate reservoir surgery
and 3% in those treated with trabeculectomy.213

Advanced age is a risk factor for the development of binocular diplopia.235 Superior placement of plate
reservoir versus inferior placement may be associated with less risk of diplopia in the reading
position.201,235

Ocular motility disturbance is reported in up to 60% of patients after retinal detachment repaired by scleral
buckle.237 The strabismus may be transient and resolve within 6 months. Persistent strabismus, however,
has been reported in up to 23% of patients treated with an encircling scleral buckle procedure, although not
all required strabismus surgery.238 In another single center study of 1030 patients treated with a scleral
buckling procedure, only 3.8% had persistent strabismus and were referred for intervention. 239

The risk of strabismus after placement of a radial scleral buckle is substantially less than after placement of
an encircling band.238

The incidence of strabismus is very low, estimated at under 3% (three cases in 920 procedures) and much
lower in association with other eyelid procedures.240

Excessive cautery, hemorrhage, and misdirected dissection likely causes strabismus, but there are no other
identified risk factors.

As noted above, diplopia may be transient following other ocular surgeries, but if it has not resolved or
demonstrated improvement within 6 months of the procedure, diplopia is less likely to resolve on its own.

P372
 Adult Strabismus PPP

Diplopia, loss of binocular vision, compensatory head position, and inability to make eye contact with poor eye
alignment all warrant treatment.

• Reduction of diplopia
• Improvement in or return of binocular vision
• Improvement of compensatory head position
• Restoration of ocular alignment, improved eye contact, and self-esteem.

A detailed medical and ocular history, including specific questions about particular surgical procedures,
anesthetic used, and onset of patient symptoms should be obtained.

A detailed sensorimotor evaluation, assessment of refractive status, and dilated fundus examination should
be performed to assess the oculomotor status, including possible ocular torsion impacting fusion. In-office
forced ductions, if tolerated, may help determine the significance of the restrictive component. If this is not
possible, the impact of mechanical restriction can be determined at the time of surgery.

Strabismus following retrobulbar or peribulbar block is characterized by an initial paresis with underaction
of the affected muscle or more extensive contracture resulting in limitation strabismus, followed by
segmental contracture of the affected muscle resulting in overaction of the muscle. The most commonly
affected muscle is the inferior rectus muscle (70%).202 Strabismus following glaucoma plate reservoir
surgery usually presents within 3 months of glaucoma implant placement, often within the first month.218,236
Superotemporal implants more frequently cause hypertropia and exotropia, and restriction or scarring may
create a “pseudoresection” effect on adjacent muscles.218,241 Superonasal implants more frequently cause
hypotropia.215,217,218,236 In patients who present with ocular deviation away from the implant, the mechanical
bulk of the implant (e.g., in the superior nasal quadrant) may be responsible.133,214,217,218

In patients who have had scleral buckling, strabismus is usually incomitant and restrictive with limitations
to ocular rotations. Approximately 50% of the patients with diplopia have a torsional component, usually
excyclotropia.219 Evaluation on the synoptophore is helpful, using fusible targets with square contours, to
determine how much torsional correction is needed to restore single vision.

P373
 Adult Strabismus PPP

In patients postblepharoplasty or other eyelid surgery, strabismus may be transient. Many cases improve
after 8 to 15 months if it is the result of local injury or hemorrhage without permanent loss of function or
significant scarring.240 As diplopia is often incomitant, prism correction is rarely satisfactory in the long run
and a combination of imaging, forced ductions, and surgical exploration may be required to maximize the
outcome.

The patient should be monitored/observed if symptoms are mild, occasional, and not bothering the patient
or if the patient is opposed to treatment. If the potential compromise to optic nerve function with removal
of the reservoir is too great for patients with a glaucoma plate reservoir, other options such as prism
correction, occlusion, and contralateral eye surgery may need to be considered. Close communication with
the physician treating the patient’s glaucoma often clarifies a course of action that facilitates surgical
treatment of the secondary strabismus, along with a method of controlling IOP.

Transient strabismus is common, so early treatment with prisms may be offered. Prisms may also be
helpful for residual diplopia following strabismus surgery.

For patients with persistent strabismus, prisms and surgery are recommended. Patients may require multiple
surgeries, including surgery in the non-affected eye. In general, surgery of the affected eye is indicated in
patients who have limited ocular rotations in the affected eye. Surgery on the non-affected eye may be
considered when, for example, surgery designed to match restrictions seems preferable, or when there is
minimal mechanical strabismus and surgery on the contralateral eye is preferable to reduce the risk of
destabilizing the achieved outcome for the involved eye.218

For patients with a glaucoma plate reservoir, if the potential compromise to optic nerve function with
removal of the reservoir is too great, other options such as prism correction or occlusion may need to be
considered. Close communication with the physician treating the patient’s glaucoma often clarifies a course
of action that facilitates surgical treatment of the secondary strabismus, along with a method of controlling
IOP. When operating in the field of a glaucoma implant, the surgeon should be prepared to manage
hypotony and the added risk of intraocular infection if an intraocular breach occurs.

P374
 Adult Strabismus PPP

Generally, careful evaluation with forced ductions at the time of surgery will clarify the mechanism of the
induced strabismus. High resolution MRI may provide additional insight into the multiple causes of
misalignment and thus inform management.242

In patients with a scleral buckle, factors associated with better surgical outcomes include small preoperative
deviation and minimally restricted ocular rotations. In patients with ocular torsion, exploration of the
superior oblique tendon and the inferior rectus muscles should be considered.219 The benefits of removing a
scleral buckle are controversial. Some surgeons believe it is almost never necessary to remove the buckle
and they will recess, resect, and transpose muscles and tendons over, under, and around the hardware,
whereas other surgeons prefer to remove the buckle at the time of strabismus surgery.243 244 The risk of
retinal re-detachment is about 8% after scleral buckle removal.245 Preoperative discussion with the
specialist who placed the periocular hardware, or another colleague in this subspecialty, should address the
risks associated with displacement of this hardware. Some procedures might benefit from having both
surgeons present and scrubbed for the strabismus repair.

Chemodenervation (botulinum toxin) has proven successful in some cases.246 Regardless of the etiology, a
careful sensorimotor evaluation with attention to torsional strabismus and incomitance, forced ductions,
and surgical exploration may be necessary to optimize outcomes because the details of repair will be
determined on a case-by-case basis,

Diagnosis and management of strabismus occurring after other ocular surgery requires the training and clinical
judgment of an experienced ophthalmologist. Communication, as possible, should be established with the
surgeon who is performing the original surgical procedure that is associated with the development or
exacerbation of strabismus. Working under the supervision of an ophthalmologist, orthoptists can be an asset in
the examination, diagnosis, and nonsurgical management. Optometrists may add additional support in the
management of aniseikonia, or contact lens fitting, as appropriate.

The ophthalmologist should discuss the findings, explain the disorder, provide diagnosis, and discuss
management options with the patient, the referring ophthalmic surgeon, and any caregivers. When surgical
treatment of the strabismus is considered, consultation with the specialist who performed the initial surgery, or
another ophthalmologist with expertise in the relevant subspecialty is recommended. This should include a
discussion of the potential impact of removing or disrupting implanted hardware. Review of the surgical record
from the index case can provide additional information on the approach to the surgical field

P375
 Adult Strabismus PPP

Superior oblique palsy (SOP) is due to partial or complete paralysis of the fourth cranial nerve (trochlear nerve),
resulting in weakness of the superior oblique muscle. The motility disturbance creates a combination of vertical,
torsional, and, to a lesser degree, horizontal incomitant strabismus pattern. Both congenital and acquired forms
exist, with the presumed congenital type often presenting for the initial evaluation in adulthood.247

Patients will show hypertropia of the involved eye, often accompanied by a compensatory head tilt away from
the affected eye. Complaints include diplopia, asthenopia, and/or anomalous head posture. Patients with
decompensated congenital SOP may have facial asymmetry, demonstrating a shorter maxilla on the opposite the
side of the SOP (and thus reduced distance between the corner of the mouth and lateral canthus). 248

Head trauma (including concussion) is one of the most common identifiable causes of both unilateral and
bilateral acquired superior oblique palsy, although in extremely rare cases isolated trochlear schwannomas and
giant cell arteritis can result in acquired SOP.249-251, 247

Skew deviation has a motility pattern that can resemble the misalignment seen with superior oblique palsy.
Distinguishing characteristics, particularly related to fundus torsion, symptoms of ocular tilt and, at times, the
impact of upright versus supine positioning of the patient are described at length in Section IIi. Skew
Deviation.252 Distinction is critical as skew deviation is more frequently associated with less benign etiology.

The patient population includes adults with strabismus associated with congenital or acquired superior oblique
palsy.

• Determine if SOP is congenital or acquired and unilateral or bilateral

• Counsel the patient on the diagnosis and the treatment options
• Provide goal-directed management of strabismus (reconstructive to enhance binocular vision, reduce
diplopia or a compensatory head posture)
• Inform the patient’s other health care providers of the diagnosis and treatment plan

P376
 Adult Strabismus PPP

Superior oblique palsy is one of the common types of vertical strabismus seen in adults. Annual incidence has
been reported to be 6.3 cases per 100,000 people, with a higher incidence among males than females.63

Risk factors include head trauma and age-related decompensation of congenital weakness.

Patients may experience no diplopia or only intermittent diplopia in cases of longstanding and presumed
congenital superior oblique palsy. A longstanding and effective compensatory head position minimizes diplopia
awareness by avoiding the field of action of the weak superior oblique muscle.

Diplopia, compensatory head position sometimes associated with neck pain, and asthenopia warrant treatment.

Treatment in adult strabismus is symptom directed, and targeted outcomes may include:
• Reconstruction of ocular alignment
• Improved binocular vision
• Improved control of diplopia, especially in the primary position and in side gaze
• Reduction in compensatory head position

A detailed medical and ocular history, including specific questions about prior head trauma, diplopia, and
compensatory head posture is important. Sometimes a review of prior photos demonstrates a pattern of
longstanding and potentially worsening compensatory head posture.

P377
 Adult Strabismus PPP

A sensorimotor evaluation with special attention is recommended to measure torsion and determine
position of gaze with the greatest deviation. The Parks Bielschowsky three-step test is often used to confirm
the diagnosis. Hypertropia is greatest in opposite lateral gaze and head tilt to the same side. Assessing
whether torsion is a barrier to fusion is an essential part of the evaluation. Quantifying the magnitude and
direction of torsional diplopia enhances surgical planning. As the superior oblique intorts, excyclotorsion is
commonly found and may be documented with Double Maddox Rod or Lancaster Red Green testing and
by noting fundus torsion.

Refractive status should be included in the examination. A neurological evaluation and neuroimaging are
controversial but are rarely indicated in cases of isolated unilateral SOP or bilateral cases in which trauma
is the known cause. Symptomatic diplopia with poor fusional amplitudes may be an indication for brain
and/or orbital MRI with contrast, because it may indicate an acquired SOP due to trochlear schwannoma.
Neuroimaging should always be performed if there are additional central nervous system signs or
symptoms.253 Occult thyroid eye disease can present as new-onset vertical diplopia with a positive three-
step test254 and giant cell arteritis remains within the differential diagnosis.

Exaggerated forced ductions often reveal a lax superior oblique tendon in many patients with presumed
congenital SOP.255 Forced ductions done in-office or at the time of surgery may reveal evidence of an
ipsilateral limitation of downgaze due to superior rectus contracture or contralateral restriction of elevation
due to inferior rectus contracture.

The patient should be monitored/observed if symptoms are mild, occasional, and not bothering the patient
or if the patient is opposed to treatment.

Small vertical deviations may be managed with prism glasses, although lateral incomitance, if present,
often precludes successful implementation.

Torsional surgery will be needed if torsion is a barrier to fusion (typically in bilateral cases or severe
unilateral cases due to trauma or cases associated with neurosurgery). Larger and incomitant deviations
most often require surgical intervention. Some patients with smaller deviations may prefer surgical repair
or prism correction. The goal of surgery is to improve the vertical and torsional deviation, improve the head
posture, and improve the area (field) of single binocular vision. A variety of surgical choices exist, and the

P378
 Adult Strabismus PPP

decision for which muscle(s) to operate on depends on the magnitude of deviation in primary gaze, the
degree of torsion, and the field of gaze with greatest deviation. The most common procedures used include
the following (in combination or alone):
• Inferior oblique weakening by recession or myectomy (ipsilateral)
• Inferior rectus weakening (contralateral) 256
• Superior oblique strengthening, for example, tuck (recommended for lax tendons in presumed
congenital SOP255), advancement for vertical and torsional deviations,257 or a version of the Harada-Ito
procedure, advancing only the anterior SO fibers that affect torsion (ipsilateral)258,259
• Superior rectus weakening (ipsilateral)260

Many of the above procedures may be performed using adjustable suture techniques; the target angle
should be a small undercorrection because overcorrection, particularly in downgaze, is extremely poorly
tolerated.

Management of SOP, specifically surgery, requires the training and clinical judgment of an experienced
ophthalmologist and in some cases a neuro-ophthalmologist.256 Working under the supervision of an
ophthalmologist, orthoptists can be an asset in the examination, diagnosis, and nonsurgical management of
SOP.

If the SOP is not isolated or is associated with other neurological signs and symptoms, neuroimaging and
referral to a neurologist or neuro-ophthalmologist should be considered. Small vertical fusional amplitudes in a
case of isolated SOP, absent a history of trauma, may indicate an acquired cause and should be evaluated with
neuroimaging. The ophthalmologist should discuss the findings, explain the disorder, provide diagnosis, and
discuss management options with the patient and, as appropriate, with a neuro-ophthalmologist, or other
specialist, as indicated.

P379
 Adult Strabismus PPP

Skew deviation is a vertical strabismus associated with disorders of the end-organ vestibular pathways within
the ear (such as acute vestibular neuronitis) or associated supranuclear utriculo-ocular pathways within the
posterior fossa including both the brainstem and cerebellum.261-265 The characteristic vertical strabismus may be
comitant or incomitant and results in vertical diplopia associated with ocular torsion, torticollis, and a tilt of the
vertical visual field referred to as the ocular tilt reaction. 252,266 The ocular tilt reaction produces a head tilt
towards the shoulder of the hypotropic eye and both eyes rotate in the direction of the head tilt. If the patient
experiences tilt of the vertical visual field, it will be in the same direction as the head tilt.267 With skew
deviation, the hypertropic eye will demonstrate fundus incyclotorsion and the hypotropic eye will demonstrate
fundus excyclotorsion.267 This characteristic helps to distinguish it from fourth nerve palsy, where the
hypertropic eye is most typically excyclotorted. (In the absence of skew deviation, head tilt results in a
compensatory fundus rotation opposite the direction of the head tilt.) Ocular torsion in skew may be conjugate
or dysconjugate.

It is critical to distinguish skew deviation from fourth nerve palsy because causes of skew deviation demand
immediate recognition and sometimes treatment (e.g., acute vestibular neuronitis, demyelination, or stroke) and
urgent imaging of the brain and brainstem (MRI with and without contrast seeking evidence of demyelination,
stroke, or mass lesion) may be needed. Most cases of isolated fourth nerve palsy (discussed in detail in Section
IIh. Superior Oblique Palsy) are more benign in etiology. Because vertical misalignment in superior oblique
nerve palsy may become comitant over time and skew may be comitant or incomitant, and because both
characteristically demonstrate head tilt away from the hypertropic side, the classic three-step test may not
distinguish the two. The upright-supine test (a fourth step for the three-step test) has been proposed after
demonstration that hypertropia reduces by 50% in the supine position for patients with skew deviation
(sensitivity 80%; specificity, 100%). Subsequent to introduction of the upright-supine test, other investigators
demonstrated that this distinctive decrease in hypertropia was not reliably found in patients with acute onset
skew deviation.268 These authors concluded that the upright-supine test was not a reliable method to distinguish
acute skew deviation (onset within 2 months) from SOP, although this test might prove useful in the setting of
more longstanding skew deviation.

P380
 Adult Strabismus PPP

The patient population consists of adults with acute peripheral vestibulopathy (vestibular neuronitis), or
demyelination, ischemia, or mass lesions affecting the vestibular supranuclear pathways within the region of the
brainstem and cerebellum

• Recognition of skew deviation and institution of appropriate ancillary testing and referral to colleagues
with expertise in otolaryngology or neurology, as indicated for further evaluation and management

The prevalence of skew deviation is unknown because it is associated with a variety of etiologies.

Because there are multiple disparate causes, risk factors vary for skew deviation. Vestibular neuronitis can
occur at any age. Demyelinating disease affects females more than males and both demyelination and trauma
are more common in younger adults. Cardiovascular risk factors associated with stroke, such as a history of
hypertension, diabetes, hyperlipidemia and smoking, have a greater impact in the aging population.

Vestibular neuronitis may be self-limiting and respond to medical intervention, and transient ischemia in the
posterior fossa may, likewise, cause a transient skew deviation that resolves over time.252 More profound
ischemic damage or mass lesions often result in long-lasting symptoms of skew deviation.

Control of disabling diplopia and perceived tilt of the subjective visual vertical warrant treatment.

Treatment in adult strabismus is symptom directed, and targeted outcomes may include:
• Improved binocular vision
• Improved control of diplopia

P381
 Adult Strabismus PPP

• Reduced tilt of the subjective visual vertical

Skew deviation is almost always acute to subacute in onset. However, if the etiology is demyelination or a
slow-growing tumor, the onset may be more insidious. Skew associated with acute vestibular neuronitis
will likely be associated with severe vertigo, dizziness, nausea, and vomiting, whereas skew in association
with brainstem or cerebellar demyelination, stroke, or a mass lesion, will demonstrate associated
neurologic features that may include nystagmus, INO, hemiparesis, sensory loss, ataxia, and Horner’s
syndrome, to name a few, depending on the locus of the pathology, and may be acute to subacute in onset.

A history of neurological symptoms (motor or sensory changes, ataxia, headache) or symptoms of vertigo
and nausea, in addition to the classic vertical diplopia and head tilt may help guide additional evaluation.
Vertical diplopia and commonly torticollis and subjective tilting of the perceived visual world are
characteristic.252,266

The examination should include the following elements:

• Complete ophthalmic examination with emphasis on the sensorimotor evaluation and completion of
the three-step test and consideration of the upright-supine test.
• Careful checking for other neuro-ophthalmic signs and symptoms looking for Horner’s syndrome,
cranial nerve palsy, internuclear ophthalmoplegia (INO), nystagmus, and hearing loss
• Fundus examination to check for papilledema or optic atrophy
• Visual field testing, which may provide additional information on the etiology
• Abnormalities in the rostral pons and midbrain (e.g., INO) will result in contralateral hypotropia and
head tilt, whereas abnormalities in the vestibular periphery, medulla, and more caudal pons will result
in ipsilateral hypotropia and head tilt.267

The primary diagnostician should refer the patient for evaluation by indicated neurology or otolaryngology
specialists. Initial treatment with prism may be helpful to manage diplopia while waiting for possible recovery,
and botox, prism, or strabismus surgery may be considered for cases that do not resolve. The goals of surgery
should be customized for the individual patient, who may be bothered to different degrees by diplopia, head tilt,
and perceived tilting of the visual vertical. These interventions may help reduce both vertical and torsional
diplopia, they may reduce the perceived tilting of the visual field, and in select cases they may prove successful
at resolving head tilt.269-271

P382
 Adult Strabismus PPP

Diagnosis and management of skew deviation requires the training and clinical judgment of an experienced
ophthalmologist and in some cases a neuro-ophthalmologist. Working under the supervision of an
ophthalmologist, orthoptists can be an asset in the examination, diagnosis, and nonsurgical management of
SOP.

Referral to specialists in neurology, otolaryngology, or neuro-otology may be indicated based on the likely
cause of the skew deviation.

P383
 Adult Strabismus PPP

Abducens palsy is also known as sixth nerve palsy. It typically presents with an acute onset of horizontal double
vision, worse at distance than at near and worse laterally toward the side of the affected nerve. Most often the
double vision is noticed by the patients in primary position, but in partial paralysis it may only be noted on
lateral gaze. Some patients present with a head turn to compensate for the diplopia caused by the paralysis. The
incomitant horizontal misalignment described can result in disabling and disturbing diplopia at onset.272
Symptoms may be less bothersome, or even subtle, if the palsy is partial, gradual, and chronic, permitting the
development of horizontal fusional amplitudes. The diplopia may become more apparent with exhaustion or
systemic illness, or in extreme horizontal gaze.

The majority of acute sixth nerve palsies in the adult population are vasculopathic, associated with risk factors
such as diabetes and hypertension.273 The onset of double vision is acute, may be accompanied by pain, and
does not have any associated neurologic or ocular findings. Most palsies resolve after 6 months, and about one-
third resolve within 8 weeks. If no recovery is apparent by 6 months, approximately 40% of patients
demonstrate a serious underlying pathology warranting further evaluation.274 The elderly who present with
symptoms of scalp or temporal region tenderness, or pain with chewing (jaw claudication) may have giant cell
arteritis, a more serious vasculopathic disorder that can result in permanent visual loss if not promptly
diagnosed and treated.

Other common causes of sixth nerve palsy are trauma and neoplasm. Traumatic etiology is usually self-evident
and may include a history of head injury, typically involving a basilar skull fracture, or an acute rise in
intracranial pressure from an intracranial bleed.275,276

A sixth nerve palsy caused by an intracranial neoplasm may be either insidious or acute. Neurologic changes
may include other motor deficits, depending on the topographic location detailed below. Evaluation for facial
and extremity motor weakness, third- and fourth-cranial nerve involvement, visual field defect, and central
acuity from optic nerve involvement, IOP, and proptosis all help in localization. Bilateral sixth-nerve
involvement can be seen in the clival chordoma as well as increased intracranial pressure or a meningeal
process.275,277 A sixth nerve palsy can be associated with demyelinating diseases such as multiple sclerosis.
Because lesions typically involve the pons, other neurologic findings are present, most notably a facial palsy,
because the seventh nerve curves over the sixth-nerve nucleus. However, isolated cranial nerve VI palsy is the
most common in adults. Postviral sixth nerve palsy can occur, but it is typically a diagnosis of exclusion.278

P384
 Adult Strabismus PPP

The patient population is adults with strabismus caused by sixth nerve palsy.

• Determine the etiology of the sixth nerve palsy

• Counsel the patient on the diagnosis and treatment options
• Provide goal-directed management of strabismus (typically to reduce symptoms of diplopia or secondary
compensatory head posture diplopia)
• Inform the patient’s other health care providers of the diagnosis and treatment plan

The annual incidence of new-onset sixth nerve palsy is approximately 11 per 100,000.273

Risk factors relate to etiology but, as vascular etiology is the most common, both hypertension and diabetes
increase the risk.

Diplopia, loss of binocular vision, compensatory head position, and inability to make eye contact all warrant
treatment.

Treatment in adult strabismus is symptom directed, and targeted outcomes may include:
• Reduction of diplopia
• Resolution of torticollis
• Reconstruction of ocular alignment
• Improvement of binocular vision

Understanding the path of the sixth nerve aids localization of the lesion as well as the cause. 279 The sixth nerve
originates in the abducens nucleus of the pons. Thus, lesions involving the sixth nerve nucleus often are
accompanied by an ipsilateral facial paralysis or a complete horizontal gaze deficit. Within the pons, it courses
through the corticospinal tracts, and thus, lesions involving the sixth in this region can lead to a contralateral

P385
 Adult Strabismus PPP

hemiparesis. The nerve then ascends in the subarachnoid space and passes underneath the petroclinoid ligament
to enter into the cavernous sinus. Tethering of the nerve along this pathway make it susceptible to stretching
from increased intracranial pressure, cerebrospinal fluid inflammation, infection, infiltration, or compression
from clival tumors.

In the cavernous sinus, the sixth nerve is in close proximity to the intracavernous carotid artery; ocular
sympathetics; and cranial nerves three, four, and the first division of the trigeminal nerve. Thus, carotid artery
aneurysms, inflammatory cavernous sinus processes, or other structural lesions may include a concomitant
Horner’s syndrome, third nerve palsies and fourth-nerve palsies, or facial pain.280 The sixth nerve then passes
through the superior orbital fissure into the orbit. 281 Orbital infections, mass lesions, or inflammation can affect
the sixth nerve as well as the lateral recuts muscle itself. Associated proptosis and optic neuropathy may be
present. Venous congestion from either an orbital process or back pressure from cavernous sinus lesions can
cause chemosis with increased IOP.

A detailed medical history should include information and medical records on prior ocular surgery, a
history of diabetes or hypertension, and complaints of specific additional symptoms, including facial
weakness, motor weakness, headache, fever, proptosis, vision loss, scalp pain or jaw claudication that
might reflect the anatomical locus and cause of sixth nerve pathology.

The examination should include the following elements:

• Complete ophthalmic evaluation with emphasis on best-corrected acuity, a check for afferent defect,
and color acuity to screen for orbital and cavernous sinus pathology
• Sensorimotor examination demonstrating incomitant esotropia, typically greater at distance, and
possibly associated with abduction nystagmus
• Fundus examination to look for papilledema or optic atrophy as indicators of elevated intracranial
pressure

Neuroimaging should be considered in all young patients or in any patient manifesting other cranial
neuropathies, other neurologic change or elevated IOP, or if there are no compelling vasculopathic risk
factors. Study should seek possible evidence of neoplasm, demyelination, stroke, vascular abnormality, or
signs of infectious or inflammatory etiology. Referral to a neuro-ophthalmologist or neurologist for further
evaluation may be indicated in some cases.

P386
 Adult Strabismus PPP

In an elderly patients with hypertension, hyperlipidemia or diabetes, and without temporal tenderness, jaw
claudication or scalp pain, evaluation may be limited to determining blood pressure, serum glucose level,
and hemoglobin A1c. Follow-up is necessary to determine if the palsy spontaneously resolves. In the
absence of resolution or improvement, MRI of the brain with and without contrast may be indicated. In the
elderly with a history of scalp tenderness, jaw claudication, or pain, the erythrocyte sedimentation rate and
C-reactive protein should be checked immediately and a temporal artery biopsy performed if the results
indicate possible giant cell arteritis. Consultation with a neuro-ophthalmologist may facilitate a biopsy and
initiation of treatment.

Evidence of increased intracranial pressure with papilledema, bilateral sixth nerve palsy, or meningeal
signs (stiff neck with headache) suggests a need for lumbar puncture following neuroimaging to measure
intracranial pressure and look for meningitis (infectious, inflammatory, or carcinomatous) or
demyelination. Systemic serology in this instance should include Lyme and syphilis testing. 282

A detailed treatment of the sixth nerve palsy is initially directed to the primary cause. Neuroimaging should be
considered in all young patients or in any patient manifesting other cranial neuropathies, other neurologic
change or elevated IOP, or signs of elevated intracranial pressure, or if there is no compelling vasculopathic risk
factor. In the aging population with vasculopathic risk factors, lack of resolution over time suggests a need for
neuroimaging. At onset, a history of scalp tenderness, jaw claudication, or pain should prompt evaluation for
possible giant cell arteritis. From the ocular motility standpoint, symptomatic intervention is to reduce diplopia
and torticollis and to restore binocular vision.

The patient should be monitored/observed if symptoms are mild or if the patient is opposed to treatment.

Nonsurgical treatment for abducens palsy can include occlusion (with a patch over one eye or with a
Bangerter filter or satin tape applied to a lens of the glasses) unless a comfortable region of single binocular
vision is achieved with minimal compensatory head posture. Prisms can also be used temporarily, and if the
deviation remains stable and fusion can be achieved, a ground-in base-out prism can be incorporated into
the lenses.283 Chemodenervation (botulinum toxin) of the medial rectus muscle may help reduce secondary
contracture and the severity of compensatory head position. 284,285

Strabismus surgical management is generally offered when the deviation persists after 6 months from onset.
A small residual deviation with some residual abducting force of the lateral rectus past the midline usually
responds well to a medial rectus recession and lateral rectus resection. Other options include a contralateral
medial rectus recession, with or without posterior fixation, with or without ipsilateral lateral rectus
resection, to balance ductions. Larger deviations with no abducting force of the lateral rectus (confirmed by

P387
 Adult Strabismus PPP

force generation testing) usually necessitate some form of transposition procedure of the vertical recti
laterally (often combined with medial rectus weakening by recession or injection of botulinum toxin) either
initially or subsequently.

Many techniques for vertical rectus transposition have been described, including full-tendon or partial-
tendon transfer, of one or both vertical rectus muscles, with or without an augmentation suture.278,286-289
Partial tendon transfer, or loop myopexy, may decrease the incidence of anterior segment ischemia, since
three extraocular muscles are not operated on at the same time. 289,290 Superior rectus muscle transposition
alone with simultaneous medial rectus recession has similar advantages. When performing transposition of
both vertical rectus muscles, staging the procedure by performing transpositions followed months later by
medial rectus recession also decreases the risk of anterior segment ischemia.291

Diagnosis and management of sixth nerve palsy requires the training and clinical judgment of an experienced
ophthalmologist. Working under the supervision of an ophthalmologist, orthoptists can be an asset in the
examination, diagnosis, and nonsurgical management of sixth nerve palsy.

Neuroimaging should be considered in all young patients or in any patient manifesting other cranial
neuropathies, other neurologic change or elevated IOP, or signs of elevated intracranial pressure, or if there are
no compelling vasculopathic risk factors. In the aging population with risk factors, lack of resolution over time
suggests a need for neuroimaging. Referral should also be considered to those providers who have expertise in
neuro-ophthalmology and adult strabismus, if the primary diagnosis is uncertain, and for directed treatment. The
ophthalmologist should discuss the findings, explain the disorder, provide diagnosis, and discuss management
options with the patient and, as appropriate, with a neuro-ophthalmologist, or other subspecialist, as indicated.
Despite the number of strabismus surgical options for repair, patients should be advised that the goal of
treatment is to eliminate diplopia in the primary position and to create a reasonable field of single binocular
vision. Patients will most likely continue to have diplopia in more extreme lateral gaze because of the difficulty
in balancing ductions given the underlying neurological deficit.

P388
 Adult Strabismus PPP

Oculomotor palsy is also known as a third nerve palsy and is the second most common cranial neuropathy.292
Patients usually present with diplopia secondary to misalignment that has both a horizontal and vertical
component, and sometimes with difficulty reading secondary to accommodative deficiency. Because the eyelid
can be partially or completely ptotic, the subjective complaint of diplopia may be lessened because visual axis
may be occluded.

The presentation is an incomitant deviation. In addition to ptosis from levator dysfunction, paresis of the
superior rectus, inferior oblique, medial rectus, and the inferior rectus muscles typically leaves the eye abducted
and infraducted as a result of preserved lateral rectus and superior oblique muscle function. Variants of
misalignment may be present, particularly if the location of the inciting pathology is intracavernous or
intraorbital, since the nerve has already divided into an upper and lower division. The pupil may or may not be
involved, and it is important to note this characteristic, because it helps direct the evaluation.

The patient population is adults with strabismus caused by oculomotor palsy.

• Determine the etiology of the third nerve palsy

• Counsel the patient on the diagnosis and treatment options
• Provide goal-directed management of strabismus (reconstruction, binocular vision, or reduction of diplopia)
• Inform the patient’s other health care providers of the diagnosis and treatment plan

The annual incidence of acquired third-nerve palsy is approximately 4 per 100,000.293

Patients with hyperlipidemia, hypertension, and diabetes are at greater risk of vasculopathic third nerve
palsy.293,294 Elderly with symptoms of scalp or temporal tenderness or jaw claudication are at risk for third nerve
palsy due to giant cell arteritis.

P389
 Adult Strabismus PPP

Diplopia, loss of binocular vision, compensatory head position, and reconstruction of alignment all warrant
treatment consideration.

Treatment in adult strabismus is symptom directed, and targeted outcomes may include:
• Reconstruction of ocular alignment with improved eye contact and self-image
• Improvement of binocular vision
• Reduction of diplopia
• Reduction of torticollis

The third nerve follows a course from its origination in the midbrain to the orbit. 295 The third-nerve nucleus lies
near the midline of the midbrain. It is composed of four paired subnuclei and one unpaired subnucleus. The
unpaired central caudal nucleus innervates both ipsilateral and contralateral elevator palpebral muscles. If
complete third nerve palsy is accompanied by contralateral superior rectus weakness, the lesion is nuclear.
Localization of the lesion by associated findings and symptoms is important in determining further workup,
establishing an etiology, and directing treatment. Lesions involving the superior cerebellar peduncle cause
ipsilateral cerebellar ataxia,296 lesions involving the red nucleus cause ipsilateral flapping hand tremor and
ataxia,297 and lesions involving the cerebral peduncle cause ipsilateral hemiplegia or hemiparesis.

Within the subarachnoid space, the third nerves pass close to the tentorial edge and lateral to the posterior
communicating artery. Aneurysms in this location cause a pupil-involving third nerve palsy, although the pupil
may appear normal at presentation.249,298 The nerve passes along the tentorial edge and adjacent to the most
medial aspect of the temporal lobe (the uncus). Mass lesions or intracranial bleeding can force the uncus
through the tentorial notch, causing compression. 299

The nerve then enters the cavernous sinus, and pathology there may be accompanied by sixth nerve and fourth
nerve paralysis.300 Within the cavernous, sinus the third nerve separates into superior and inferior divisions.
Both enter into the orbit through the superior orbital fissure. The superior division innervates the superior rectus
and the levator. The inferior division sends parasympathetic fibers into the ciliary ganglion, the pupillary
sphincter, the ciliary (accommodation), the inferior oblique, the medial rectus, and inferior rectus. Divisional
palsy typically localizes the lesion to the orbit. If the lesion involves the orbital apex or orbit, there may be
associated optic neuropathy and proptosis.

P390
 Adult Strabismus PPP

A detailed medical and ocular history should include specific questions about patient symptoms, speed of
onset of the strabismus, and possible associated unilateral or bilateral ptosis or other neurologic symptoms
(such as ataxia, tremor, hemiplegia, sixth nerve or fourth nerve palsy, and noted pupillary asymmetry, and
vision loss).

A comprehensive eye examination with particular attention to sensorimotor exam, evidence of ptosis and
anisocoria, and pupillary responses in bright and dim illumination. It should also include a fundus
examination to evaluate for the presence of papilledema or optic atrophy.

Evaluation of the third nerve palsy depends on the presumed location of the lesion and is based on other
accompanying neurologic findings. Isolated third nerve palsy, however, is what most clinicians encounter.
The important characteristics are whether the pupil is involved and the extent of the motility disorder and
ptosis. A classic pupil-sparing third nerve palsy has normal pupillary function, complete ptosis, and
complete (related) motility dysfunction. In this situation, the etiology is almost always secondary to
microvascular disease with associated diabetes, hypertension, or hyperlipidemia.293 However, even if the
pupil is unaffected, when there is partial extraocular muscle involvement or incomplete ptosis, one cannot
be certain of a microvascular etiology; a compressive lesion, for example, might present this way. In this
situation, it is recommended to proceed either with neuroimaging including an MRI with gadolinium and
magnetic resonance angiography (MRA) or computed tomography angiography (CTA). In some instances
of vasculopathic third nerve palsy, the pupil may be involved, although mildly.

Pupil-involving third nerve palsy is more concerning. A compressive lesion must be ruled out, urgently,
especially an aneurysm involving the posterior communicating artery. Active force generation testing in the
office may help identify muscles with residual function (which may be responsive to resection) vs.
completely paretic muscles (where resection will not be effective and the muscle should be left intact to
preserve ciliary circulation).301 Neuroimaging, including MRA or CTA is recommended.302 If there is a
high suspicion of aneurysm despite a normal MRA or CTA, then a catheter angiogram should be
considered after a brain MRI with and without contrast and specific attention to the third nerve is
performed. Tumors, including meningioma, schwannoma, and metastatic lesions, are in the differential.
Other causes include trauma, subarachnoid hemorrhage, viral illnesses, demyelinating disease, and
leptomeningeal disorders.293 If the neuroimaging is normal, the next step would be serologic testing for
infectious diseases (such as syphilis and Lyme), with consideration given to lumbar puncture that would
include glucose, protein, cell count, and cytology and culture.

P391
 Adult Strabismus PPP

Management of a third nerve palsy is directed towards the cause and then management of the diplopia.

With complete ptosis, many patients are not troubled by diplopia until their lid is elevated.

If the patient experiences diplopia, chemodenervation (botulinum toxin) of an opposing extraocular muscle
or the levator or use of occlusion can temporize while waiting for further recovery of function. 303 Prisms,
either press-on or ground-in, are often helpful, either while waiting for further recovery or treating residual
diplopia (following partial recovery or following surgical correction). Despite all efforts to alleviate
diplopia in the primary position, the incomitant nature of the deficit results in diplopia with minimal shift of
gaze from the primary position. Many patients wear an eye patch or occlusive contact lens or MIN lens
(Fresnel, Inc.) for times when the diplopia is most bothersome. A deficit of accommodation may cause
difficulty reading. Uniocular progressive lenses or bifocals in younger patients can be prescribed to aid
accommodation.

Surgical management is complicated, and success often depends on the amount of residual deficit.304 A
residual exotropia with ability to adduct past the midline typically responds well to recession of the lateral
rectus muscle combined with resection of the medial rectus muscle (with or without adjustment), with or
without vertical transposition.305,306 Likewise, in partial third nerve palsy, ipsilateral weakening of the
superior oblique muscle, or its anterior intorting fibers might reduce the hypotropia and intorsion as a result
of inferior oblique weakness. Recession with or without posterior fixation of muscles on the contralateral
eye can also be used to expand the field of binocular single vision.

For compete paralysis, many other techniques have been tried with variable success. 307 Some form of
weakening procedure of the lateral rectus muscle is typically required, whether supramaximal recession,
extirpation, or suturing to the periosteum of the lateral orbital rim. 308 Sometimes this is combined with
nasal transposition of the superior oblique muscle toward the medial rectus muscle insertion or maximal
medial rectus resection for its tethering effect.309,310 Recently, nasal transposition of the split lateral rectus
muscle to the medial rectus muscle insertion has been performed with some success.311

Post-operative prism correction may be required to allow for fusion in the primary or reading position.
Ptosis surgery should be tempered if the Bell’s response is extremely impaired, increasing the risk of
exposure keratopathy. Patients treated surgically often have significant diplopia awareness outside of a
region of binocular fusion and may benefit from part-time occlusion or a fogging contact lens for high-risk
activities like driving a motor vehicle. Patients with concomitant injury to other structures in the brain or

P392
 Adult Strabismus PPP

brainstem may have central fusion disruption and be unable to fuse despite otherwise satisfactory post-
operative alignment.

Diagnosis and management of third nerve palsy requires the training, clinical judgment, and the experience of
providers with expertise in neuro-ophthalmology and adult strabismus. Working under the supervision of an
ophthalmologist, orthoptists can be an asset in the examination, diagnosis, and nonsurgical management of
patients with third nerve palsy.

The ophthalmologist should discuss the findings, explain the disorder, provide diagnosis, and discuss
management options with the patient and any caregivers.

P393
 Adult Strabismus PPP

The term myasthenia gravis refers to a group of B-cell mediated autoimmune disorders that profoundly affect
activity at the neuromuscular junction causing variable weakness exacerbated by fatigue. 312 Although generalized
myasthenia gravis affects large motor groups and sometimes includes ocular features, ocular myasthenia gravis
affects only the levator, orbicularis oculi, and the extraocular muscles. It is known as the great masquerader because
presentation can mimic many types of incomitant strabismus with or without ptosis.

Acetylcholine is released at the neuromuscular junction at the onset of an action potential and migrates across
the synapse to reach the associated striated muscle. Acetylcholine receptor antibodies have been demonstrated
in nearly all patients with generalized myasthenia and in 40% to 77% of patients with ocular myasthenia.
Twitch fibers in extraocular muscles are thought to be particularly susceptible to fatigue, and this characteristic
as well as the fewer number of acetylcholine receptor antibodies in these muscles may be responsible for the
common ocular manifestations.313

The patient population includes adults with myasthenia gravis without geographic or racial predilection,
although this disorder occurs in children as well in the form of transient neonatal myasthenia (transplacental
transmission), congenital myasthenia (not immune mediated), and juvenile autoimmune myasthenia. Patients are
at risk for having or developing thymoma.

• Recognize the disorder to initiate appropriate workup and referral to other subspecialties, including
neurology
• Ensure proper medical and surgical treatment, and initiate short-term and subsequent long-term
management of associated ptosis and strabismus
Thymectomy is indicated, certainly in the presence of thymoma, but also for specific age-based and immune-
based characteristics for which thymectomy may substantially reduce clinical symptoms.312

P394
 Adult Strabismus PPP

The incidence ranges from 0.04 to 5/100,000 per year, and prevalence estimates are 0.5 to 12.5/100,000 per
year.314

There is increased risk for myasthenia gravis in the presence of autoimmune thyroid disease and thymoma.315

Onset in adults is usually in the third to fourth decade of life but sometimes much later in males. If the disorder is
generalized, it may involve bulbar, limb, and respiratory muscles, which can lead to life threatening respiratory
failure. Fifty percent of patients with myasthenia present with ocular symptoms only (typically ptosis and
variable strabismus and diplopia), and of these 50% to 80% develop generalized systemic myasthenia within a
few years.316,317

Minimizing extraocular motility impairment, diplopia, and compensatory head position that can affect visual
function, quality of life, and its socioeconomic consequences warrant treatment.

Treatment is goal directed, and targeted outcomes may include:

• Recognition of disease and prevention of morbidity and mortality from generalized myasthenia gravis
• Reduction of diplopia (if present)
• Restoration of normal ocular alignment, appearance, and improved self-image
• Restoration of binocular function
• Reduction of compensatory head position (if present)

Variable incomitant strabismus may be accompanied by variable ptosis, both of which worsen with fatigue.
Strabismus and associated diplopia may take on an entirely different pattern with fatigue or repetition of
examination. Additionally, a period of rest or ice pack test in the physician’s office may temporarily reduce or

P395
 Adult Strabismus PPP

eliminate both the extraocular motility disturbances and ptosis. Ptosis worsens with fatigue (prolonged upgaze)
and may have associated Cogan lid-twitch sign. Ptosis of the contralateral eye may worsen upon manual
elevation of the eye, with more severe ptosis due to Hering’s law. Ocular saccades are often slow. Patients may
also present with lid retraction as a result of associated TED.318 Pupils are typically not affected but may in rare
cases show impaired or slow responses.319,320

Patients may present with a history of acquired variable strabismus, diplopia and ptosis. Those with
antecedent generalized disease may have variable motor weakness, respiratory fatigue, difficulty with
chewing, swallowing, holding up the head and slurred speech.

Complete sensorimotor examination and external examination is critical, with attention to the presence of
strabismus, which changes over the course of prolonged examination, and variable ptosis with possible
Cogan lid-twitch sign, and slow saccades. The Ice Test-application of an ice pack over the closed eyes for 2
minutes in the case of ptosis and for 5 minutes in the case of strabismus may demonstrate a reduction of
ptosis of about 2 mm and a reduction of misalignment. This phenomenon, thought to relate to diminished
anticholinesterase activity, is highly specific to this disorder .321,322 A rest test without an ice pack can also
be used to aid in diagnosis.323

Tensilon (edrophonium) testing may be considered but is best performed by a practitioner experienced with
proper intravenous administration, because associated muscarinic activity (excess tearing, salivation,
sweating, abdominal cramping, bradycardia, bronchospasm hypotension, and syncope) can occur. The test
should be performed in a monitored setting and atropine available for potential administration. It is 95%
sensitive for generalized myasthenia and 86% sensitive in cases of ocular myasthenia.324

The presence of antiacetylcholine receptor antibody (AChR-Ab-binding, blocking, or modulating) can

confirm the diagnosis; however, about 20% of patients with generalized myasthenia and about half of those
with ocular myasthenia are seronegative. About one-third of these seronegative patients will be seropositive
for muscle-specific kinase (anti MuSKAb) ,and lipoprotein-related protein 4 (LRP4) has been associated
with generalized and ocular myasthenia gravis as well.325

Repetitive nerve stimulation testing (positive in only one-third of patients with ocular myasthenia) and the
far more sensitive single fiber electromyography (positive in over 90% of patients with ocular myasthenia)
may also assist in diagnosis.326 In many centers, single fiber EMG is considered the gold standard for
diagnosis.

P396
 Adult Strabismus PPP

Pyridostigmine bromide administered orally two to four times a day is the first-line treatment for myasthenia
gravis, but about half of patients with strabismus-associated myasthenia show minimal response. In contrast,
about 66% to 85% of patients show a positive response to corticosteroids.327 For some patients, various forms of
immunosuppressive therapy with azathioprine, known to be effective, and other agents under current
investigation may be offered by treating neurologists. Thymectomy is indicated in some cases, always in the
presence of thymoma, and may substantially reduce symptoms for certain subpopulations with myasthenia
gravis.312

Diplopia and strabismus are highly variable and not readily remedied with prism. Remission or stabilization of
the disease is often possible after 2 to 3 years of treatment,328 and at that point surgical intervention for
strabismus may be considered if desired or if prism use is insufficient. Particular care is indicated in the use of
anesthetic agents given any evidence of associated weakness of the respiratory muscles. Surgical management,
with and without the use of adjustable sutures, has met with modest success in cases where there is a stabilized
primary deviation, sometimes exacerbated by fatigue. 329-334 More than one procedure may prove necessary.

Diagnosis and management of myasthenia gravis requires the training and clinical judgment of an experienced
ophthalmologist typically working in concert with a treating neurologist.

Counselling and referral to a neurologist or neuro-ophthalmologist, and sometimes a general surgeon, is often
indicated in the management of myasthenia. The ophthalmologist should discuss the findings, explain the
disorder, provide a diagnosis, and discuss management options with the patient and any caregivers and be aware
of any comorbidities such as respiratory distress that might present with generalization of the disease.

P397
 Adult Strabismus PPP

In patients with childhood strabismus who have a suppression scotoma in their nondominant eye, fixation
switch diplopia may occur because of a change in ocular fixation preference.335 Prior to their change in eye
fixation preference, such patients do not perceive diplopia as the result of a suppression scotoma that was
formed during childhood in the previously nondominant eye. However, when visual acuity in the previously
dominant eye declines, fixation preference may switch to the previously nondominant eye. The previously
dominant eye does not have a suppression scotoma, so for the first time, patients with “childhood” strabismus
may experience double vision.

The largest study of fixation switch diplopia included 16 patients with a history of childhood strabismus who
later presented with diplopia.336 The most common underlying etiologies were the use of monovision for the
treatment of presbyopia in 38%, the development of myopia in the previously preferred eye in 25%, and
incorrect refractive correction in 38% of patients. A trial of monovision with contact lenses is prudent prior to
corneal or lenticular refractive surgery to determine whether surgically induced monovision will result in new-
onset diplopia. Other published rare causes of fixation switch diplopia include a corneal ulcer and resultant scar
in the dominant eye of a patient with monofixation syndrome and amblyopia 337 and retinal detachment in the
dominant eye of patients with strabismus and amblyopia. 335

The patient population includes adults with fixation switch diplopia.

• Counsel the patient on the diagnosis and treatment options

• Manage diplopia
• Inform the patient’s other health care providers of the diagnosis and treatment plan

Prevalence and incidence are unknown. Risk factors include any ophthalmic disorder or intervention that
switches fixation to the previously nondominant eye in a patient with latent or manifest strabismus.

P398
 Adult Strabismus PPP

Fixation switch diplopia is a subacute complaint of new onset diplopia after any ophthalmic disorder or
intervention that switches fixation to the previously nondominant eye in a patient with latent or manifest
strabismus.

Diplopia warrants treatment.

• Reduction or elimination of symptomatic diplopia

Fixation switch diplopia is the development of new-onset diplopia when fixation preference has shifted to the
previously nondominant eye.

A detailed medical history includes information on past ocular conditions and ocular fixation preference.
Careful documentation of symptoms experienced when fixating with the previously preferred eye, and when
fixing with the newly preferred eye, can confirm the diagnosis. A positive family history of strabismus is
common.338 Kushner and Kowal339 recommend a minimal screening procedure consisting of obtaining a history
of childhood eye disease or treatments (such as patching or eye muscle surgery), checking spectacles for prism,
cycloplegic refraction, and cover testing for all patients undergoing corneal or lenticular refractive surgery,
especially for those planning to induce monovision.

Common scenarios in which fixation switch diplopia occurs include:

• Monovision has been induced by optical means or by refractive surgery
• Myopia with axial elongation has developed in a previously dominant eye
• Following refractive surgery, if the outcome favors fixation with the previously nondominant eye
• An unbalanced refraction has been used that encourages fixation with the nondominant eye
• Following cataract surgery in a previously nondominant eye when the surgery results in better visual acuity
of the nondominant eye compared with the dominant eye336

P399
 Adult Strabismus PPP

• Following asymmetric vision loss from other common diseases such as macular degeneration or diabetic
retinopathy when the nondominant eye is left with better acuity

Comprehensive eye examinations of adult strabismus patients need to include:

• Detailed sensorimotor evaluation
• Assessment of refractive status
• Dilated fundus examination
Enabling fixation with the previously dominant eye can be done to confirm a history of monofixation and
suppression before the change in ocular fixation preference occurred.

The patient should be monitored/observed if symptoms are mild or occasional or if the patient is opposed to
treatment.

An attempt can be made to “switch” fixation back to the dominant eye, by pursuing refractive correction.
Patients will gradually experience less diplopia if dominance is re-established in the previously dominant
eye. In cases where the fixation cannot be switched back due to permanent and untreatable vision loss,
vision in the newly dominant eye should be optimized with refractive correction and the use of prisms.
Rarely, fixation switch diplopia cannot be alleviated, and in those cases occlusion must be contemplated as
a last resort.

In cases where the fixation cannot be switched back because of permanent and untreatable vision loss,
strabismus surgery may be an option if there is a significant angle of misalignment and a demonstrated
ability to relieve diplopia with prism correction by establishing a different alignment at which the patient
can suppress.

Fixation switch diplopia should be managed by an ophthalmologist with expertise in the diagnosis of
longstanding childhood strabismus and abnormalities of binocular function. Working under the supervision of
an ophthalmologist, orthoptists can be an asset in the examination, diagnosis, and nonsurgical management of
fixation switch diplopia.

P400
 Adult Strabismus PPP

Patients should be advised that they have an excellent prognosis if fixation can be switched back to the
dominant eye. They should be counseled to avoid any procedures or refractive corrections that result in a switch
of fixation to their nondominant eye. The ophthalmologist should discuss the findings, explain the disorder,
provide diagnosis, and discuss management options with the patient and any caregivers.

P401
 Adult Strabismus PPP

Retinal disease that causes distortion or displacement of the fovea in one or both eyes—such as subretinal or
epiretinal membranes or after retinal detachment involving the macula—can cause binocular diplopia. In these
cases, binocular retinal misregistration (often manifesting as metamorphopsia, micropsia or macropsia, or
foveal ectopia) renders the foveal images in the two eyes too dissimilar to fuse.340,341. When binocular retinal
misregistration results in diplopia, central and peripheral fusion are in conflict, described as central-peripheral
rivalry.342,343 In this situation, most often the stronger peripheral contribution to motor fusion brings the
peripheral retinas into alignment and leaves the fovea of each eye misaligned.

The patient population includes adults with binocular retinal diplopia.

• Counsel the patient on the diagnosis and treatment options

• Manage diplopia
• Inform the patient’s other health care providers of the diagnosis and treatment plan

The prevalence of subretinal neovascular membranes and macular epiretinal membranes is 2% for individuals
under 60 years of age and increases to as much as 12% for those over 70.344,345 Between 16% and 37% of those
diagnosed with epiretinal membrane or other maculopathy have binocular retinal diplopia.343,346 With aging of
the population, it is likely that retinal disease will be an increasingly important cause of diplopia.

Within days to weeks of development of maculopathy or foveal abnormality, binocular retinal diplopia may
develop, and this only rarely improves over time.

P402
 Adult Strabismus PPP

Diplopia warrants treatment.

Treatment for binocular retinal diplopia includes diminishing diplopia.

Binocular diplopia develops within days to weeks of worsening maculopathy or foveal abnormality because the
foveal image in the two eyes becomes too dissimilar to fuse.

A detailed medical history should include information on ocular conditions and specifically retinal disease
involving the macula.

An examination should include any history of retinal disease affecting the macula of one or both eyes. A
full orthoptic examination is warranted because a proportion of patients who have maculopathy and
diplopia have other forms of treatable strabismus.347 Prism alternate cover testing demonstrates either no
strabismus or a small vertical deviation. Any diplopia relief from prism correction is most often transient,
although in rare patients it may be helpful. Metamorphopsia may be identified on the Amsler grid test and
quantified using M-Charts, and aniseikonia (unequal images) can be documented and quantified using the
Awaya test. (See Glossary)

For dragged-fovea diplopia syndrome, the lights on/off test is pathognomonic. In this test, a small white-
on-black test letter is shown and, when peripheral fusion cues are eliminated by darkening the room, central
fusion allows the test letter to be seen singly. When the room lights are turned on, peripheral fusion
reasserts itself, the images of the test letter separate and diplopia resumes. 343 This test is not effective unless
the entire room is darkened with no cues to peripheral fusion such as door frames or background lighting
on a computer monitor. An alternative to the lights-on lights-off test for those who are unable to eliminate
all peripheral cues in their examination rooms is the optotype-frame test.347(See Glossary) For the optotype-
frame test, the patient is requested to fixate an isolated (uncrowded) Snellen optotype on an illuminated

P403
 Adult Strabismus PPP

monitor and to describe whether the letter itself is single or double. Assuming it is single, the patient is
asked whether the frame of the monitor is single or double. If maintaining a single frame causes diplopia of
the letter, peripheral fusion has taken over and foveal diplopia syndrome is present. Stronger peripheral
fusion makes it impossible to maintain central single binocular vision when foveal image disparity has
become too great.

The patient should be monitored/observed if symptoms are mild or occasional or if the patient is opposed to
treatment.

Although cover testing in patients with binocular retinal diplopia often demonstrates small-angle
strabismus that is nearly always vertical, prismatic or surgical correction of this deviation is not curative
because it does not resolve the mismatch of distorted macular images or the conflict between foveal and
peripheral alignment. Even so, intervention can sometimes reduce symptoms. Reduced diplopia awareness
has been reported with the use of prism correction or fogging in some patients.348 Fogging the vision in one
eye, which eliminates the foveal conflict by producing a central scotoma, 349 has been the most successful.
Blenderm surgical tape applied to the spectacle lens has been used,3 but Scotch Satin tape343 or
Bangerter foils349 are generally better tolerated, or use of an occlusive contact lenses. A small amount of
prism in addition to a Bangerter foil can provide better relief from diplopia in some difficult cases.350

If superimposition of the foveal images by synoptophore or prism does not satisfactorily resolve or
diminish subjective diplopia, strabismus surgery is unlikely to be successful. It may be considered if
improved binocular alignment reduces symptoms incompletely addressed by fogging or optical correction
alone, or if superimposition of the foveal images by synoptophore diminishes subjective diplopia. It is
important to remember that some patients with retinal misregistration have treatable causes of diplopia as
well347 . There is emerging evidence that peeling of the epiretinal membrane may be effective in addressing
binocular retinal diplopia in a proportion of patients, 348 but retinal surgery for epiretinal membrane appears
to be a double-edged sword because some nondiplopic patients become diplopic following this retinal
procedure.351

P404
 Adult Strabismus PPP

A complaint of diplopia normally prompts referral to a pediatric ophthalmologist, orthoptist or neuro-

ophthalmologist, many of whom are familiar with the diagnosis and management of this perplexing problem.

Patients with binocular retinal diplopia are generally already under the care of a retina specialist, from whom
they are typically referred. While the underlying retinal disease often needs ongoing care, surgical treatment
such as membrane peeling improves diplopia only in a proportion of patients. The strabismus specialist may
need to discuss the potential benefits of membrane peeling with the retina specialists. The potential value of
prism improved refractive correction and strabismus surgery to reduce diplopia awareness in some cases should
not be overlooked. The patient should be counseled, however, that this condition usually does not improve
spontaneously, and fogging of one eye may be the most suitable long-term solution. The ophthalmologist
should discuss the findings, explain the diagnosis, and discuss management options with the patient and any
caregivers.

P405
 Adult Strabismus PPP

Although complications can occur as a result of any surgical procedure, the risk of sight-threatening complications is
particularly low with strabismus surgery.27,352,353 The more common complications are minor, often self-limited, or
treated with topical medications. More serious complications are fortunately rare.

Severe complications from strabismus surgery have been estimated to be 1 in 400 (globe perforation, severe
infection, slipped or lost muscle and scleritis), with 1 in 2,400 resulting in a poor or very poor outcome.353

The following postoperative concerns are common to adults as well as children after strabismus surgery.
Postoperative concerns with a higher incidence in the adult population are indicated.

These are best treated with observation or a short course of common supportive therapy and include: 354
• Foreign body sensation
• Corneal abrasion
• Allergic reaction
• Delle (2.2%–18.9%, higher risk in reoperations and transpositions) 355,356
• Subconjunctival hemorrhage
• Chemosis
• Mydriasis
• Reduced accommodation
• Conjunctival scar
• Visible muscle insertion
• Thin visible/dark sclera

P406
 Adult Strabismus PPP

• Persistent injection over the surgical site

• Postoperative nausea
• Pyogenic granuloma (2.1%)357
• Tenon’s prolapse
• Epithelial inclusion cyst (0.25%)358
• Advancing plica semilunaris
• Significant conjunctival scar

Persistent Tenon’s prolapse, pyogenic granuloma, epithelial inclusion cyst, advancing plica semilunaris,
and conjunctival scar are sometimes addressed with minor surgical intervention.

• Overcorrection or undercorrection (possibly treated with additional strabismus surgery)

• Limitation of eye ductions (possibly treated with additional strabismus surgery)
• Loss of binocular function (may benefit from additional strabismus surgery, orthoptic
intervention, or use of a prism)

• Globe perforation (0.08%–5.1%; although serious, most cases have no sequelae but place patient
at higher risk for retinal detachment, vitreous hemorrhage, and endophthalmitis).353,359-362 Some
cases are observed, others are treated with retinopexy.
• Altered eyelid position, common with surgery on the inferior or superior rectus muscles and more
notable in patients with thyroid ophthalmopathy (who already may have eyelid retraction). If
persistent or of concern, may be addressed with eyelid surgery.
• Oculocardiac reflex (67.9%) or asystole (0.11%) (both usually brief and of no consequence).363,364

Rare, but very serious complications include:

• Endophthalmitis (1 in 30,000–1 in 185,000)359,365
• Subconjunctival abscess, preseptal/orbital cellulitis (1 in 1,100–1 in 1,900)353,365,366
• Surgically induced necrotizing scleritis (1 in 4,000; more likely in adults)353
• Slipped muscle (1 in 1,500)353
• Lost muscle (1 in 4,500; more likely in adults)353
• Pulled-in-two syndrome (1 in 14,000; more likely in adults)352
• Retinal detachment (1 in 10,000–1 in 40,000)359,367

P407
 Adult Strabismus PPP

• Adhesive syndrome (fat adherence syndrome) 368

• Anterior segment ischemia (1 in 6,000; higher risk with age, vascular risk factors, and operating on
three or more muscles)357
• Diplopia, rarely intractable (0.8%; more likely in adults)10
• Postoperative orbital hemorrhage (incidence unknown, exceedingly rare, potentially greater risk for
patients on anticoagulation)369,370

• Operating on the wrong eye or muscle (1 in 2,506)371

Diagnosis and management of complications from adult strabismus surgery requires the training, clinical
judgment, and experience of a pediatric ophthalmologist or strabismologist.

The ophthalmologist should discuss and explain the findings and discuss management options with the patient.
In some cases, treatment may be best managed with the additional expertise of other ophthalmic colleagues with
advanced retina, oculoplastics, or immunology expertise, as indicated.

P408
 Adult Strabismus PPP

Strabismus surgery in adults is often more challenging than in children for a variety of reasons. A well thought
out surgical plan with preparation for the unexpected is important. Attention should be given to the following
special considerations.

Many strabismus surgeons do not routinely stop anticoagulants for strabismus surgery. Temporary
discontinuation of anticoagulation medication may reduce intraoperative bleeding, but consultation with the
physician prescribing anticoagulation is advised. Reasons for anticoagulation differ and, in some cases,
bridging therapy is indicated. Anticoagulation should be restarted immediately or shortly after the
procedure.372 It is possible to perform strabismus surgery on patients who are on anticoagulants. 373
Absorbable gelatin sponges with or without thrombin may also be useful in difficult cases but are rarely
required for more routine procedures.

Adjustable sutures are often used by strabismus surgeons, particulary for adult patients, although many
surgeons achieve excellent results without the use of adjustable sutures. Advocates for adjustable sutures
point to several advantages over fixed sutures, including a second chance at obtaining satisfactory
alignment, and the potential to minimize risk of post-operative diplopia. A variety of techniques (bowtie,
noose, semiadjustable) exist, and the timing of the adjustment varies with surgeon preference, ranging from
immediately in the operation room, several hours after the surgery, and to up to several days following the
procedure. This technique may be most helpful in reoperations and unpredictable cases, such as those with
restrictive or paralytic strabismus.150,154,200,257,311,374,375

P409
 Adult Strabismus PPP

Surgeons will frequently encounter microtropias in adults. Small deviations (<8 prism diopters horizontally
and <3 prism diopters vertically) may prove clinically significant causing diplopia or asthenopia. Small
horizontal or vertical deviations may be successfully treated using a tenotomy procedure (partial, central
snip) to obviate the need for prism correction. 376 377

At present, intramuscular injection of botulinum toxin into an extraocular muscle may be used alone or in
combination with strabismus surgery to treat new onset deviations, to enhance the impact of traditional
surgery on large deviations, to address a residual deviation after prior strabismus surgery, and to treat
small-angle deviations.61,76,139,303,378 Botulinum toxin can also be used to prevent contracture of an
ipsilateral antagonist muscle in cases of paralytic strabismus while waiting the requisite period of time for
spontaneous resolution before recommending strabismus surgery. A Cochrane Systematic Review found
that it was difficult to assess outcomes given the limited number of RCTs. 378

Retrobulbar or peribulbar anesthesia and monitored sedation or pure topical anesthesia are possibilities in
adults having unilateral surgery, though general anesthesia may be preferable in the case of reoperation,
and complex, bilateral, and longer procedures.

A higher percentage of adults have complex strabismus as a result of scarring from previous strabismus
surgery, restriction from orbital trauma, or restriction from other ocular surgeries including glaucoma seton
or scleral buckle surgery, the placement of orbital wall implants after trauma, blepharoplasty,
dacryocystorhinostomy, pterygium excision or sinus surgery as well as myotoxicity from local
anesthesia..201 (Many of these concerns are addressed in Section IIg. Strabismus Associated with Other
Ophthalmic Surgery.) Scleromalacia may increase the complexity of both surgery and postoperative
recovery.

A “stretched scar” (weak attachment between muscle and globe) is a common occurrence in adults who
have undergone horizontal rectus muscle surgery in childhood. Suggestive clinical findings include
increased deviation in the field of action of the previously operated muscle. It is important to distinguish
weak, non-muscular attachments from more robust muscle tissue intraoperatively, as placement of sutures
in this non-muscular tissue will lead to early recurrence of the strabismus and further loss of muscle force.
Slipped or tenuously attached muscle (which sometimes appears like a stretched scar), and lost muscles are
more common in adults and make surgical outcomes less predictable.379 They should be approached with

P410
 Adult Strabismus PPP

great care. A meticulous technique is often required to ensure a successful outcome. An experienced
assistant can be very helpful. Sometimes a slipped or lost muscle is readily retrieved by tracing its natural
path or by taking advantage of fine attachments, if not yet severed, to adjacent extraocular muscles. For
example, a slipped inferior rectus muscle may still retain its attachments to the adjacent inferior oblique
muscle. If the lost muscle cannot be found, waiting a month until likely adhesion to the globe may facilitate
recovery of the previously lost segment, as the muscle can now be hooked. High-resolution orbital imaging
may facilitate surgical planning. If this proves technically impossible, a transposition procedure may be
considered to replace missing function.352,353,380,381

The surgeon should also pay careful attention to specific intraoperative issues that are far more common in
adults. A second strabismus surgeon or skilled assistant may be needed and ample time allotted.
Conjunctiva and Tenon’s are frail in the older population. Thin and delicate conjunctiva may make wound
closure difficult. Amniotic membrane grafts may sometimes be used in extreme cases. Likewise, careful
technique is necessary to prevent pulled-in-two syndrome, in which the muscle belly is torn in two
fragments from too vigorous traction on weakened extraocular muscles. 146

Retrobulbar, intramuscular, and intraocular hemorrhage may occur in adult strabismus surgery as a result of
the surgery or periorbital anesthesia.369,370,382 Although rare, they are potentially vision-threatening
complications, and the surgeon should be knowledgeable in their management.

Whereas children will have formed vitreous, liquefied vitreous is common in adults. Consequently, adults
have higher risk for retinal detachment following inadvertent scleral perforation of the needle. Various
techniques such as short scleral passes may reduce this risk.

Diagnosis and management of adult strabismus requires the training, clinical judgment, and experience of a
pediatric ophthalmologist or strabismologist.

The ophthalmologist should discuss the findings, explain the disorder, provide diagnosis, and discuss
management options with the patient and any caregivers.

P411
 Adult Strabismus PPP

Awaya Test: A test of aniseikonia that has the patient compare adjacent calibrated halfmoons—one red and one
green—to determine the degree of difference in image size and appearance. The images can be presented in any
meridian enabling quantification of vertical, horizontal, and oblique aniseikonia.

Bagolini Lenses: Striated lenses used to test for suppression, and normal or abnormal retinal correspondence.

Kushner Semi-Adjustable Technique: A modification of the adjustable suture technique that limits undesired
excess recession over time by allowing only a modest degree of secondary adjustment to an otherwise fixed
recession. (See reference.150)

Loop Myopexy: A strabismus surgical procedure wherein suture (typically nonabsorbable) is used to create an
adhesion between two adjacent muscles without additional adhesion to sclera. Examples of common use of the loop
myopexy technique include loop myopexy between the superior and lateral recti to treat myopic strabismus fixus,
and between the superior or inferior rectus muscles and the lateral rectus muscle to improve abduction in the
treatment of 6th nerve palsy or Duane syndrome.

SR-LR (Superior Rectus-Lateral Rectus) Band: This band maintains a soft tissue connection between the superior
rectus and the lateral rectus muscles as they course posteriorly. Age or staphylomatous growth of the globe can
result in a medial shift of the superior rectus and an inferior shift of the lateral rectus as this band of connective
tissue fails.

Optotype Frame Test: This is a test of central versus peripheral fusion. The patient is requested to fixate an isolated
(uncrowded) Snellen optotype on an illuminated monitor and to describe whether the letter itself is single or double.
Assuming it is single, the patient is asked whether the frame of the monitor is single or double. If maintaining a
single frame causes diplopia of the letter, peripheral fusion has taken over and foveal diplopia syndrome is present.
This is a disorder wherein stronger peripheral fusion makes it impossible to maintain single binocular vision
centrally and occurs when foveal image disparity is sufficient to disrupt central fusion.

Sagging Eye Syndrome: Some cases of divergence insufficiency may be seen in association with an age-related
degeneration of the superior rectus–lateral rectus (SR-LR) connective tissue band. The strabismus is often
accompanied by mild ptosis (or a history of treated ptosis) and deep superior lid sulcus defect associated with aging.
Divergence insufficiency is common and sometimes there is associated, modest, vertical misalignment of the eyes.

Synoptophore An instrument that measures the manifest strabismus angle in all planes and allows stimuli to be
presented to both eyes at once. The misalignment can be “fixed” by the device and the ability of the patient to fuse
the superimposed images can be tested. This device can also be used for orthoptic training, exercising fusional
vergences and for predicting whether strabismus surgery may result in binocular fusion.

P412
 Adult Strabismus PPP

Algorithm for approaching adult strabismus based on presentation of deviation

Comitant

No
Diplopia
diplopia

Horizontal Vertical Sensory

ET XT Recurrent
childhood
strabismus
Decompensated E Decompensated X Decompensated H

Age-
CI Skew
related

Divergence insuff MG MG

TED TED TED

Sagging
eye

Fixation switch

Post retinal
surgery

P413
 Adult Strabismus PPP

Incomitant

Diplopia No diplopia

Paralytic Restrictive CCDDs

CN 3,4,6 Post trauma/eye Longstanding CN 4

palsy surgery palsy

MG TED DVD

Sagging Sensory
eye

Strabismus
Fixus

Skew

APPENDIX 3.

Literature searches of the PubMed and Cochrane databases were conducted in February 2017; the search strategies
were as follows. Specific limited update searches were conducted after June 2019.

"divergence insufficiency" OR "divergence insufficiency esotropia" OR "divergence insufficiency pattern" OR

"divergence insufficiency pattern esotropia" OR “divergence insufficience”
(convergence[tw] and insufficiency[tw]) OR (("convergence insufficiency" OR "convergence insufficiency and
excess" OR "convergence insufficiency and reading study cirs group" OR "convergence insufficiency symptom" OR
"convergence insufficiency symptom survey" OR "convergence insufficiency symptom survey ciss" OR
"convergence insufficiency symptom survey questionnaire" OR "convergence insufficiency symptom survey score"
OR "convergence insufficiency symptoms" OR "convergence insufficiency treatment" OR "convergence
insufficiency treatment trial" OR "convergence insufficiency treatment trial citt investigator group" OR
"convergence insufficiency treatment trial citt study group" OR "convergence insufficiency treatment trial executive
committee" OR "convergence insufficiency treatment trial group" OR "convergence insufficiency treatment trial
study group" OR "convergence insufficiency type" OR "convergence insufficiency type intermittent exotropia"))
(superior oblique palsy) OR (("fourth nerve"[tw] OR "trochlear"[tw] OR "fourth cranial nerve"[tw] OR "cranial
nerve iv"[tw])) AND palsy[tw]) OR ("superior oblique palsy"[tw] OR "superior oblique paralysis"[tw] OR "fourth
nerve palsy"[tw] OR "fourth nerve paralysis"[tw] OR "trochlear palsy"[tw] OR "trochlear paralysis"[tw] OR "cranial
nerve iv palsy"[tw]

P414
 Adult Strabismus PPP

APPENDIX 4

Basic and Clinical Science Course

External Disease and Cornea (Section 8, 2019–2020)

Focal Points 2018

Module: Adult Strabismus

Clinical Statements – Free download available at http://one.aao.org/guidelines-

browse?filter=clinicalstatement.
Adult Strabismus Surgery - 2017

Preferred Practice Pattern® Guidelines – Free download available at www.aao.org/ppp.

Amblyopia (2017)
Esotropia and Exotropia (2017)

To order any of these products, except for the free materials, please contact the Academy’s Customer Service at
866.561.8558 (U.S. only) or 415.561.8540 or www.aao.org/store.

Orge FH. Strabismus stimulator. 2015; https://www.aao.org/interactive-tool/strabismus-simulator. Accessed August

13, 2019.

Complex Strabismus Simulator

Oct 24, 2018 by Faruk H. Orge, MD; K. David Epley, MD
This simulator expands on the basic Strabismus Simulator, allowing the exploration of more complex eye deviations
such as the alphabet patterns, cranial nerve...
Interactive / Tool

P415
 Adult Strabismus PPP

1. Bagolini B. Sensorial anomalies in strabismus. (suppression, anomalous correspondence,

amblyopia). Documenta ophthalmologica Advances in ophthalmology. 1976;41(1):1-22.
2. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating
quality of evidence and strength of recommendations. Bmj. 2008;336(7650):924-926.
3. Buehl W, Sacu S, Schmidt-Erfurth U. Retinal vein occlusions. Dev Ophthalmol.
2010;46:54-72.
4. Kushner BJ. The benefits, risks, and efficacy of strabismus surgery in adults. Optometry
and vision science : official publication of the American Academy of Optometry.
2014;91(5):e102-109.
5. Hertle RW. Clinical characteristics of surgically treated adult strabismus. Journal of
pediatric ophthalmology and strabismus. 1998;35(3):138-145; quiz 167-138.
6. Nelson BA, Gunton KB, Lasker JN, Nelson LB, Drohan LA. The psychosocial aspects of
strabismus in teenagers and adults and the impact of surgical correction. Journal of
AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2008;12(1):72-76 e71.
7. Scott WE, Kutschke PJ, Lee WR. 20th annual Frank Costenbader Lecture--adult
strabismus. Journal of pediatric ophthalmology and strabismus. 1995;32(6):348-352.
8. Mills MD, Coats DK, Donahue SP, Wheeler DT. Strabismus surgery for adults: a report
by the American Academy of Ophthalmology. Ophthalmology. 2004;111(6):1255-1262.
9. Scott WE, Kutschke PJ, Lee WR. Diplopia in adult strabismus. Am Orthoptic J.
1994;44:66-69.
10. Kushner BJ. Intractable diplopia after strabismus surgery in adults. Arch Ophthalmol.
2002;120(11):1498-1504.
11. Chang YH, Melvin P, Dagi LR. Goal-determined metrics to assess outcomes of exotropia
surgery. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2015;19(4):304-310.
12. Ehrenberg M, Nihalani BR, Melvin P, Cain CE, Hunter DG, Dagi LR. Goal-determined
metrics to assess outcomes of esotropia surgery. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2014;18(3):211-216.
13. Repka MX, Lum F, Burugapalli B. Strabismus, Strabismus Surgery, and Reoperation
Rate in the United States: Analysis from the IRIS Registry. Ophthalmology.
2018;125(10):1646-1653.
14. Birch EE, Fawcett S, Stager DR. Why does early surgical alignment improve stereoacuity
outcomes in infantile esotropia? J AAPOS. 2000;4(1):10-14.
15. Fawcett S, Leffler J, Birch EE. Factors influencing stereoacuity in accommodative
esotropia. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2000;4(1):15-20.
16. Kushner BJ, Morton GV. Postoperative binocularity in adults with longstanding
strabismus. Ophthalmology. 1992;99(3):316-319.
17. Morris RJ, Scott WE, Dickey CF. Fusion after surgical alignment of longstanding
strabismus in adults. Ophthalmology. 1993;100(1):135-138.
18. Kushner BJ. Binocular field expansion in adults after surgery for esotropia. Arch
Ophthalmol. 1994;112(5):639-643.

P416
 Adult Strabismus PPP

19. Pineles SL, Demer JL, Isenberg SJ, Birch EE, Velez FG. Improvement in binocular
summation after strabismus surgery. JAMA ophthalmology. 2015;133(3):326-332.
20. Burke JP, Leach CM, Davis H. Psychosocial implications of strabismus surgery in adults.
Journal of pediatric ophthalmology and strabismus. 1997;34(3):159-164.
21. Hatt SR, Leske DA, Kirgis PA, Bradley EA, Holmes JM. The effects of strabismus on
quality of life in adults. American journal of ophthalmology. 2007;144(5):643-647.
22. Jackson S, Harrad RA, Morris M, Rumsey N. The psychosocial benefits of corrective
surgery for adults with strabismus. The British journal of ophthalmology.
2006;90(7):883-888.
23. Kraft SP. Outcome criteria in strabismus surgery. Canadian journal of ophthalmology
Journal canadien d'ophtalmologie. 1998;33(4):237-239.
24. Menon V, Saha J, Tandon R, Mehta M, Khokhar S. Study of the psychosocial aspects of
strabismus. Journal of pediatric ophthalmology and strabismus. 2002;39(4):203-208.
25. Olitsky SE, Sudesh S, Graziano A, Hamblen J, Brooks SE, Shaha SH. The negative
psychosocial impact of strabismus in adults. Journal of AAPOS : the official publication
of the American Association for Pediatric Ophthalmology and Strabismus.
1999;3(4):209-211.
26. Satterfield D, Keltner JL, Morrison TL. Psychosocial aspects of strabismus study. Arch
Ophthalmol. 1993;111(8):1100-1105.
27. Mills MD, Coats DK, Donahue SP, Wheeler DT, American Academy of O. Strabismus
surgery for adults: a report by the American Academy of Ophthalmology.
Ophthalmology. 2004;111(6):1255-1262.
28. Hatt SR, Leske DA, Liebermann L, Holmes JM. Comparing outcome criteria
performance in adult strabismus surgery. Ophthalmology 2012;119(9):1930-1936.
29. Sullivan TJ, Kraft SP, Burack C, O'Reilly C. A functional scoring method for the field of
binocular single vision. Ophthalmology. 1992;99(4):575-581.
30. Holmes JM, Leske DA, Kupersmith MJ. New methods for quantifying diplopia.
Ophthalmology. 2005;112(11):2035-2039.
31. Kushner BJ. The usefulness of the cervical range of motion device in the ocular motility
examination. Archives of Ophthalmology. 2000;118:946-950.
32. Holmes JM, Liebermann L, Hatt SR, Smith SJ, Leske DA. Quantifying diplopia with a
questionnaire. Ophthalmology. 2013;120(7):1492-1496.
33. Coats DK, Paysse EA, Towler AJ, Dipboye RL. Impact of large angle horizontal
strabismus on ability to obtain employment. Ophthalmology. 2000;107(2):402-405.
34. Goff MJ, Suhr AW, Ward JA, Croley JK, O'Hara MA. Effect of adult strabismus on
ratings of official U.S. Army photographs. Journal of AAPOS : the official publication of
the American Association for Pediatric Ophthalmology and Strabismus. 2006;10(5):400-
403.
35. Hatt SR, Leske DA, Bradley EA, Cole SR, Holmes JM. Development of a quality-of-life
questionnaire for adults with strabismus. Ophthalmology. 2009;116(1):139-144
36. Leske DA, Hatt SR, Liebermann L, Holmes JM. Evaluation of the Adult Strabismus-20
(AS-20) Questionnaire using Rasch analysis. Investigative Ophthalmology & Visual
Science. 2012;53(6):2630-2639.
37. Dickmann A, Aliberti S, Rebecchi MT, et al. Improved sensory status and quality-of-life
measures in adult patients after strabismus surgery. Journal of AAPOS : the official

P417
 Adult Strabismus PPP

publication of the American Association for Pediatric Ophthalmology and Strabismus.

2013;17(1):25-28.
38. Hatt SR, Leske DA, Holmes JM. Responsiveness of health-related quality of life
questionnaires in adults undergoing strabismus surgery. Ophthalmology.
2010;117(12):2322-2328.
39. Hatt SR, Leske DA, Liebermann L, Holmes JM. Changes in health-related quality of life
1 year following strabismus surgery. American Journal of Ophthalmology.
2012;153(4):614-619.
40. Tandon AK, Velez FG, Isenberg SJ, Demer JL, Pineles SL. Binocular inhibition in
strabismic patients is associated with diminished quality of life. Journal of AAPOS : the
official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2014;18(5):423-426.
41. Liebermann L, Hatt SR, Leske DA, Holmes JM. Improvement in specific function-
related quality-of-life concerns after strabismus surgery in nondiplopic adults. Journal of
AAPOS: American Association for Pediatric Ophthalmology & Strabismus.
2014;18(2):105-109.
42. Hatt SR, Leske DA, Liebermann L, Holmes JM. Incorporating health-related quality of
life into the assessment of outcome following strabismus surgery. American Journal of
Ophthalmology. 2016;164:1-5.
43. Xu M, Yu H, Chen Y, Xu J, Zheng J, Yu X. Long-Term Quality of Life in Adult Patients
with Strabismus after Corrective Surgery Compared to the General Population. PloS one.
2016;11(11):e0166418.
44. Durnian JM, Owen ME, Marsh IB. The psychosocial aspects of strabismus: correlation
between the AS-20 and DAS59 quality-of-life questionnaires. Journal of AAPOS : the
official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2009;13(5):477-480.
45. McBain HB, MacKenzie KA, Au C, et al. Factors associated with quality of life and
mood in adults with strabismus. The British journal of ophthalmology. 2014;98(4):550-
555.
46. Pineles SL, Repka MX, Yu F, Lum F, Coleman AL. Risk of musculoskeletal injuries,
fractures, and falls in medicare beneficiaries with disorders of binocular vision. JAMA
ophthalmology. 2015;133(1):60-65.
47. Hatt SR, Leske DA, Liebermann L, Philbrick KL, Holmes JM. Depressive symptoms
associated with poor health-related quality of life in adults with strabismus.
Ophthalmology. 2014;121(10):2070-2071.
48. Mohney BG, McKenzie JA, Capo JA, Nusz KJ, Mrazek D, Diehl NN. Mental illness in
young adults who had strabismus as children. Pediatrics. 2008;122(5):1033-1038.
49. McBain H, MacKenzie K, Hancox J, Ezra DG, Adams GG, Newman SP. What do
patients with strabismus expect post surgery? The development and validation of a
questionnaire. The British journal of ophthalmology. 2016;100(3):415-419.
50. Convergence Insufficiency Treatment Trial Study Group. Randomized clinical trial of
treatments for symptomatic convergence insufficiency in children. Arch Ophthalmol.
2008;126(10):1336-1349.
51. Yurdakul NS, Ugurlu S. Analysis of risk factors for consecutive exotropia and review of
the literature. Journal of pediatric ophthalmology and strabismus. 2013;50(5):268-273.

P418
 Adult Strabismus PPP

52. Han SY, Han J, Rhiu S, Lee JB, Han SH. Risk factors for consecutive exotropia after
esotropia surgery. Japanese journal of ophthalmology. 2016;60(4):333-340.
53. Ganesh A, Pirouznia S, Ganguly SS, Fagerholm P, Lithander J. Consecutive exotropia
after surgical treatment of childhood esotropia: a 40-year follow-up study. Acta
ophthalmologica. 2011;89(7):691-695.
54. Hatt SR, Leske DA, Jung JH, Holmes JM. Intraoperative Findings in Consecutive
Exotropia with and without Adduction Deficit. Ophthalmology. 2017;124(6):828-834.
55. Wikipedia. Bagolini Striated Glasses Test. 2019;
https://en.wikipedia.org/wiki/Bagolini_Striated_Glasses_Test. Accessed August 19,
2019.
56. Kutschke PJ, Scott WE. Prism adaptation in visually mature patients with esotropia of
childhood onset. Ophthalmology. 2004;111(1):177-179.
57. Orge FH. Strabismus stimulator. 2015; https://www.aao.org/interactive-tool/strabismus-
simulator. Accessed August 13, 2019.
58. Magramm I, Schlossman A. Strabismus in patients over the age of 60 years. Journal of
pediatric ophthalmology and strabismus. 1991;28(1):28-31.
59. Havertape SA, Cruz OA, Chu FC. Sensory strabismus--eso or exo? Journal of pediatric
ophthalmology and strabismus. 2001;38(6):327-330; quiz 354-325.
60. Von Noorden GK, Campos EC. Binocular vision and ocular motility : theory and
management of strabismus. In: 6th ed. St. Louis, Mo.: Mosby; 2002:345-346.
61. Flanders M, Tischler A, Wise J, Williams F, Beneish R, Auger N. Injection of type A
botulinum toxin into extraocular muscles for correction of strabismus. Canadian journal
of ophthalmology Journal canadien d'ophtalmologie. 1987;22(4):212-217.
62. Helveston EM. Complications of strabismus surgery. Transactions of the New Orleans
Academy of Ophthalmology. 1986;34:61-71.
63. Martinez-Thompson JM, Diehl NN, Holmes JM, Mohney BG. Incidence, types, and
lifetime risk of adult-onset strabismus. Ophthalmology. 2014;121(4):877-882.
64. Rouse MW, Borsting E, Hyman L, et al. Frequency of convergence insufficiency among
fifth and sixth graders. The Convergence Insufficiency and Reading Study (CIRS) group.
Optometry and vision science : official publication of the American Academy of
Optometry. 1999;76(9):643-649.
65. Storey EP, Master SR, Lockyer JE, Podolak OE, Grady MF, Master CL. Near Point of
Convergence after Concussion in Children. Optometry and vision science : official
publication of the American Academy of Optometry. 2017;94(1):96-100.
66. Kawata K, Rubin LH, Lee JH, et al. Association of Football Subconcussive Head Impacts
With Ocular Near Point of Convergence. JAMA ophthalmology. 2016;134(7):763-769.
67. Biousse V, Skibell BC, Watts RL, Loupe DN, Drews-Botsch C, Newman NJ.
Ophthalmologic features of Parkinson's disease. Neurology. 2004;62(2):177-180.
68. Gallaway M, Scheiman M, Mitchell GL. Vision Therapy for Post-Concussion Vision
Disorders. Optometry and vision science : official publication of the American Academy
of Optometry. 2017;94(1):68-73.
69. Holmes JM, Liebermann L, Hatt SR, Smith SJ, Leske DA. Quantifying diplopia with a
questionnaire. Ophthalmology. 2013;120(7):1492-1496.
70. Pediatric Eye Disease Investigator G. Home-Based Therapy for Symptomatic
Convergence Insufficiency in Children: A Randomized Clinical Trial. Optometry and

P419
 Adult Strabismus PPP

vision science : official publication of the American Academy of Optometry.

2016;93(12):1457-1465.
71. Scheiman M, Mitchell GL, Cotter S, et al. A randomized clinical trial of treatments for
convergence insufficiency in children. Arch Ophthalmol. 2005;123(1):14-24.
72. Scheiman M, Gwiazda J, Li T. Non-surgical interventions for convergence insufficiency.
The Cochrane database of systematic reviews. 2011(3):CD006768.
73. Teitelbaum B, Pang Y, Krall J. Effectiveness of base in prism for presbyopes with
convergence insufficiency. Optometry and vision science : official publication of the
American Academy of Optometry. 2009;86(2):153-156.
74. Yang HK, Hwang JM. Surgical outcomes in convergence insufficiency-type exotropia.
Ophthalmology. 2011;118(8):1512-1517.
75. Kraft SP, Levin AV, Enzenauer RW. Unilateral surgery for exotropia with convergence
weakness. Journal of pediatric ophthalmology and strabismus. 1995;32(3):183-187.
76. Saunte JP, Holmes JM. Sustained improvement of reading symptoms following
botulinum toxin A injection for convergence insufficiency. Strabismus. 2014;22(3):95-
99.
77. Farid MF, Abdelbaset EA. Surgical outcomes of three different surgical techniques for
treatment of convergence insufficiency intermittent exotropia. Eye (Lond). 2017.
78. Kirkeby L. Update on divergence insufficiency. International ophthalmology clinics.
2014;54(3):21-31.
79. Jacobson DM. Divergence insufficiency revisited: natural history of idiopathic cases and
neurologic associations. Arch Ophthalmol. 2000;118(9):1237-1241.
80. Reche Sainz JA, Espinet Badia R, Puig Ganau T. Divergence insufficiency and
demyelinating disorder. European journal of ophthalmology. 2002;12(3):238-240.
81. Chaudhuri Z, Demer JL. Sagging eye syndrome: connective tissue involution as a cause
of horizontal and vertical strabismus in older patients. JAMA ophthalmology.
2013;131(5):619-625.
82. Demer JL. The Apt Lecture. Connective tissues reflect different mechanisms of
strabismus over the life span. Journal of AAPOS : the official publication of the American
Association for Pediatric Ophthalmology and Strabismus. 2014;18(4):309-315.
83. Krzizok TH, Schroeder BU. Measurement of recti eye muscle paths by magnetic
resonance imaging in highly myopic and normal subjects. Investigative ophthalmology &
visual science. 1999;40(11):2554-2560.
84. Aoki Y, Nishida Y, Hayashi O, et al. Magnetic resonance imaging measurements of
extraocular muscle path shift and posterior eyeball prolapse from the muscle cone in
acquired esotropia with high myopia. American journal of ophthalmology.
2003;136(3):482-489.
85. Demer JL. Muscle paths matter in strabismus associated with axial high myopia.
American journal of ophthalmology. 2010;149(2):184-186 e181.
86. Tan RJ, Demer JL. Heavy eye syndrome versus sagging eye syndrome in high myopia.
Journal of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2015;19(6):500-506.
87. Yamaguchi M, Yokoyama T, Shiraki K. Surgical procedure for correcting globe
dislocation in highly myopic strabismus. American journal of ophthalmology.
2010;149(2):341-346 e342.

P420
 Adult Strabismus PPP

88. Chaudhuri Z, Demer JL. Long-term Surgical Outcomes in the Sagging Eye Syndrome.
Strabismus. 2018:1-5.
89. Chaudhuri Z, Demer JL. Graded vertical rectus tenotomy for small-angle cyclovertical
strabismus in sagging eye syndrome. The British journal of ophthalmology.
2016;100(5):648-651.
90. Nakao Y, Kimura T. Prevalence and anatomic mechanism of highly myopic strabismus
among Japanese with severe myopia. Japanese journal of ophthalmology.
2014;58(2):218-224.
91. Rowe FJ, Noonan CP. Surgical treatment for progressive esotropia in the setting of high-
axial myopia. Journal of AAPOS : the official publication of the American Association
for Pediatric Ophthalmology and Strabismus. 2006;10(6):596-597.
92. Hayashi T, Iwashige H, Maruo T. Clinical features and surgery for acquired progressive
esotropia associated with severe myopia. Acta ophthalmologica Scandinavica.
1999;77(1):66-71.
93. Haller T. Evaluation and Prism Management of Divergence Insufficiency Esotropia. The
American orthoptic journal. 2015;65:40-43.
94. Daum KM. The course and effect of visual training on the vergence system. American
journal of optometry and physiological optics. 1982;59(3):223-227.
95. Pineles SL. Divergence Insufficiency Esotropia: Surgical Treatment. The American
orthoptic journal. 2015;65:35-39.
96. Chaudhuri Z, Demer JL. Medial rectus recession is as effective as lateral rectus resection
in divergence paralysis esotropia. Arch Ophthalmol. 2012;130(10):1280-1284.
97. Bothun ED, Archer SM. Bilateral medial rectus muscle recession for divergence
insufficiency pattern esotropia. Journal of AAPOS : the official publication of the
American Association for Pediatric Ophthalmology and Strabismus. 2005;9(1):3-6.
98. Yadav S, Young J, Voas-Clarke C, Marsh IB, Durnian JM. Treatment of age-related
distance esotropia with unilateral lateral rectus resection. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2014;18(5):446-448.
99. Ranka MP, Steele MA. Esotropia associated with high myopia. Current opinion in
ophthalmology. 2015;26(5):362-365.
100. Bagheri A, Adhami F, Repka MX. Bilateral recession-resection surgery for convergent
strabismus fixus associated with high myopia. Strabismus. 2001;9(4):225-230.
101. Shenoy BH, Sachdeva V, Kekunnaya R. Silicone band loop myopexy in the treatment of
myopic strabismus fixus: surgical outcome of a novel modification. The British journal of
ophthalmology. 2015;99(1):36-40.
102. Farid MF, Elbarky AM, Saeed AM. Superior rectus and lateral rectus muscle union
surgery in the treatment of myopic strabismus fixus: three sutures versus a single suture.
Journal of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2016;20(2):100-105.
103. Burch HB, Wartofsky L. Graves' ophthalmopathy: current concepts regarding
pathogenesis and management. Endocrine reviews. 1993;14(6):747-793.
104. Bartley GB, Gorman CA. Diagnostic criteria for Graves' ophthalmopathy. American
journal of ophthalmology. 1995;119(6):792-795.

P421
 Adult Strabismus PPP

105. Bartley GB, Fatourechi V, Kadrmas EF, et al. The incidence of Graves' ophthalmopathy
in Olmsted County, Minnesota. American journal of ophthalmology. 1995;120(4):511-
517.
106. Wiersinga WM, Smit T, van der Gaag R, Mourits M, Koornneef L. Clinical presentation
of Graves' ophthalmopathy. Ophthalmic research. 1989;21(2):73-82.
107. Dagi LR, Elliott AT, Roper-Hall G, Cruz OA. Thyroid eye disease: honing your skills to
improve outcomes. Journal of AAPOS : the official publication of the American
Association for Pediatric Ophthalmology and Strabismus. 2010;14(5):425-431.
108. Wu CY, Kahana A. Geriatric patients are predisposed to strabismus following thyroid-
related orbital decompression surgery: A multivariate analysis. Orbit. 2017;36(2):95-101.
109. Perros P, Crombie AL, Matthews JN, Kendall-Taylor P. Age and gender influence the
severity of thyroid-associated ophthalmopathy: a study of 101 patients attending a
combined thyroid-eye clinic. Clin Endocrinol (Oxf). 1993;38(4):367-372.
110. Stan MN, Bahn RS. Risk factors for development or deterioration of Graves'
ophthalmopathy. Thyroid : official journal of the American Thyroid Association.
2010;20(7):777-783.
111. Bartalena L, Marcocci C, Tanda ML, et al. Cigarette smoking and treatment outcomes in
Graves ophthalmopathy. Annals of internal medicine. 1998;129(8):632-635.
112. Del Monte MA. 2001 an ocular odyssey: lessons learned from 25 years of surgical
treatment for graves eye disease. The American orthoptic journal. 2002;52:40-57.
113. Jang SY, Shin DY, Lee EJ, Lee SY, Yoon JS. Relevance of TSH-receptor antibody levels
in predicting disease course in Graves' orbitopathy: comparison of the third-generation
TBII assay and Mc4-TSI bioassay. Eye (Lond). 2013;27(8):964-971.
114. Takakura A, Kirkeby K, Earle K, Silkiss RZ. Predicting the Development of Orbitopathy
in Graves Thyroidopathy Patients: The Potential Role of TSI Testing. Ophthalmic Plast
Reconstr Surg. 2015;31(5):369-372.
115. Cruz OA. Restricted motility for vertical strabismus in Graves' ophthalmopathy.
Ophthalmology. 2001;108(2):242-243.
116. Farid M, Roch-Levecq AC, Levi L, Brody BL, Granet DB, Kikkawa DO. Psychological
disturbance in graves ophthalmopathy. Arch Ophthalmol. 2005;123(4):491-496.
117. Barker L, Mackenzie K, Adams GG, Hancox J. Long-term Surgical Outcomes for
Vertical Deviations in Thyroid Eye Disease. Strabismus. 2017;25(2):67-72.
118. Buckley EG. Restrictive Strabismus. In: Plager DA, ed. Strabismus surgery: basic and
advanced strategies. Oxford: Oxford University Press; 2004:117-154.
119. Dagi LR, Zoumalan CI, Konrad H, Trokel SL, Kazim M. Correlation between
extraocular muscle size and motility restriction in thyroid eye disease. Ophthalmic Plast
Reconstr Surg. 2011;27(2):102-110.
120. Kerr NC. The role of thyroid eye disease and other factors in the overcorrection of
hypotropia following unilateral adjustable suture recession of the inferior rectus (an
American Ophthalmological Society thesis). Transactions of the American
Ophthalmological Society. 2011;109:168-200.
121. Marcocci C, Kahaly GJ, Krassas GE, et al. Selenium and the course of mild Graves'
orbitopathy. The New England journal of medicine. 2011;364(20):1920-1931.
122. Rotondo Dottore G, Leo M, Casini G, et al. Antioxidant Actions of Selenium in Orbital
Fibroblasts: A Basis for the Effects of Selenium in Graves' Orbitopathy. Thyroid : official
journal of the American Thyroid Association. 2017;27(2):271-278.

P422
 Adult Strabismus PPP

123. Leo M, Bartalena L, Rotondo Dottore G, et al. Effects of selenium on short-term control
of hyperthyroidism due to Graves' disease treated with methimazole: results of a
randomized clinical trial. Journal of endocrinological investigation. 2017;40(3):281-287.
124. Smith TJ, Kahaly GJ, Ezra DG, et al. Teprotumumab for Thyroid-Associated
Ophthalmopathy. The New England journal of medicine. 2017;376(18):1748-1761.
125. Bartalena L, Baldeschi L, Boboridis K, et al. The 2016 European Thyroid
Association/European Group on Graves' Orbitopathy Guidelines for the Management of
Graves' Orbitopathy. European thyroid journal. 2016;5(1):9-26.
126. Nunery WR, Nunery CW, Martin RT, Truong TV, Osborn DR. The risk of diplopia
following orbital floor and medial wall decompression in subtypes of ophthalmic Graves'
disease. Ophthalmic Plast Reconstr Surg. 1997;13(3):153-160.
127. Leong SC, White PS. Outcomes following surgical decompression for dysthyroid
orbitopathy (Graves' disease). Current opinion in otolaryngology & head and neck
surgery. 2010;18(1):37-43.
128. Paridaens D, Hans K, van Buitenen S, Mourits MP. The incidence of diplopia following
coronal and translid orbital decompression in Graves' orbitopathy. Eye (Lond). 1998;12 (
Pt 5):800-805.
129. Abramoff MD, Kalmann R, de Graaf ME, Stilma JS, Mourits MP. Rectus extraocular
muscle paths and decompression surgery for Graves orbitopathy: mechanism of motility
disturbances. Investigative ophthalmology & visual science. 2002;43(2):300-307.
130. Trokel SL, Cooper WC. Symposium: extraocular muscle problems associated with
graves' disease. Orbital decompression: effect on motility and globe position.
Ophthalmology. 1979;86(12):2064-2070.
131. Serafino M, Fogagnolo P, Trivedi RH, Saunders RA, Nucci P. Torsional diplopia after
orbital decompression and strabismus surgery. European journal of ophthalmology.
2010;20(2):437-441.
132. Finn AP, Bleier B, Cestari DM, et al. A Retrospective Review of Orbital Decompression
for Thyroid Orbitopathy with Endoscopic Preservation of the Inferomedial Orbital Bone
Strut. Ophthalmic Plast Reconstr Surg. 2017;33(5):334-339.
133. Schotthoefer EO, Wallace DK. Strabismus associated with thyroid eye disease. Current
opinion in ophthalmology. 2007;18(5):361-365.
134. Zloto O, Ben Simon G, Didi Fabian I, et al. Association of orbital decompression and the
characteristics of subsequent strabismus surgery in thyroid eye disease. Canadian journal
of ophthalmology Journal canadien d'ophtalmologie. 2017;52(3):264-268.
135. Mainville NP, Jordan DR. Effect of orbital decompression on diplopia in thyroid-related
orbitopathy. Ophthalmic plastic and reconstructive surgery. 2014;30(2):137-140.
136. Scott AB. Injection treatment of endocrine orbital myopathy. Documenta
ophthalmologica Advances in ophthalmology. 1984;58(1):141-145.
137. Dunn WJ, Arnold AC, O'Connor PS. Botulinum toxin for the treatment of dysthyroid
ocular myopathy. Ophthalmology. 1986;93(4):470-475.
138. Lyons CJ, Vickers SF, Lee JP. Botulinum toxin therapy in dysthyroid strabismus. Eye
(Lond). 1990;4 ( Pt 4):538-542.
139. Granet DB, Hodgson N, Godfrey KJ, et al. Chemodenervation of extraocular muscles
with botulinum toxin in thyroid eye disease. Graefe's archive for clinical and
experimental ophthalmology = Albrecht von Graefes Archiv fur klinische und
experimentelle Ophthalmologie. 2016;254(5):999-1003.

P423
 Adult Strabismus PPP

140. Al Qahtani ES, Rootman J, Kersey J, Godoy F, Lyons CJ. Clinical Pearls and
Management Recommendations for Strabismus due to Thyroid Orbitopathy. Middle East
African journal of ophthalmology. 2015;22(3):307-311.
141. Coats DK, Paysse EA, Plager DA, Wallace DK. Early strabismus surgery for thyroid
ophthalmopathy. Ophthalmology. 1999;106(2):324-329.
142. Mourits MP, Prummel MF, Wiersinga WM, Koornneef L. Clinical activity score as a
guide in the management of patients with Graves' ophthalmopathy. Clin Endocrinol
(Oxf). 1997;47(1):9-14.
143. Politi LS, Godi C, Cammarata G, et al. Magnetic resonance imaging with diffusion-
weighted imaging in the evaluation of thyroid-associated orbitopathy: getting below the
tip of the iceberg. European radiology. 2014;24(5):1118-1126.
144. Bailey CC, Kabala J, Laitt R, et al. Magnetic resonance imaging in thyroid eye disease.
Eye (Lond). 1996;10 ( Pt 5):617-619.
145. Yoo SH, Pineles SL, Goldberg RA, Velez FG. Rectus muscle resection in Graves'
ophthalmopathy. Journal of AAPOS : the official publication of the American Association
for Pediatric Ophthalmology and Strabismus. 2013;17(1):9-15.
146. Ellis EM, Kinori M, Robbins SL, Granet DB. Pulled-in-two syndrome: a multicenter
survey of risk factors, management and outcomes. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2016;20(5):387-391.
147. Nardi M. Squint surgery in TED -- hints and fints, or why Graves' patients are difficult
patients. Orbit. 2009;28(4):245-250.
148. Chatzistefanou KI, Kushner BJ, Gentry LR. Magnetic resonance imaging of the arc of
contact of extraocular muscles: implications regarding the incidence of slipped muscles.
Journal of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2000;4(2):84-93.
149. Hudson HL, Feldon SE. Late overcorrection of hypotropia in Graves ophthalmopathy.
Predictive factors. Ophthalmology. 1992;99(3):356-360.
150. Kushner BJ. An evaluation of the semiadjustable suture strabismus surgical procedure.
Journal of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2004;8(5):481-487.
151. Dagi LR. Understanding and managing vertal strabismus from thyroid eye disease.
Journal of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2018.
152. Flanders M, Hastings M. Diagnosis and surgical management of strabismus associated
with thyroid-related orbitopathy. Journal of pediatric ophthalmology and strabismus.
1997;34(6):333-340.
153. Wei Y, Kang XL, Del Monte MA. Enlargement of the superior rectus and superior
oblique muscles causes intorsion in Graves' eye disease. The British journal of
ophthalmology. 2016;100(9):1280-1284.
154. Volpe NJ, Mirza-George N, Binenbaum G. Surgical management of vertical ocular
misalignment in thyroid eye disease using an adjustable suture technique. Journal of
AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2012;16(6):518-522.
155. Holmes JM, Hatt SR, Bradley EA. Identifying masked superior oblique involvement in
thyroid eye disease to avoid postoperative A-pattern exotropia and intorsion. Journal of

P424
 Adult Strabismus PPP

AAPOS : the official publication of the American Association for Pediatric

Ophthalmology and Strabismus. 2012;16(3):280-285.
156. Gomi CF, Yang SW, Granet DB, et al. Change in proptosis following extraocular muscle
surgery: effects of muscle recession in thyroid-associated orbitopathy. Journal of AAPOS
: the official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2007;11(4):377-380.
157. Zoumalan CI, Lelli GJ, Jr., Kazim M. Tenon recession: a novel adjunct to improve
outcome in the treatment of large-angle strabismus in thyroid eye disease. Ophthalmic
Plast Reconstr Surg. 2011;27(4):287-292.
158. Scofield-Kaplan SM, Dunbar K, Stein G, Kazim M. Improvement in Both Primary and
Eccentric Ocular Alignment After Thyroid Eye Disease-Strabismus Surgery With
Tenon's Recession. Ophthalmic Plast Reconstr Surg. 2018;34(4S Suppl 1):S85-S89.
159. Kraus DJ, Bullock JD. Treatment of thyroid ocular myopathy with adjustable and
nonadjustable suture strabismus surgery. Transactions of the American Ophthalmological
Society. 1993;91:67-79; discussion 79-84.
160. Peragallo JH, Velez FG, Demer JL, Pineles SL. Postoperative drift in patients with
thyroid ophthalmopathy undergoing unilateral inferior rectus muscle recession.
Strabismus. 2013;21(1):23-28.
161. Nicholson BP, De Alba M, Perry JD, Traboulsi EI. Efficacy of the intraoperative relaxed
muscle positioning technique in thyroid eye disease and analysis of cases requiring
reoperation. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2011;15(4):321-325.
162. Ludwig IH, Brown MS. Flap tear of rectus muscles: an underlying cause of strabismus
after orbital trauma. Ophthalmic plastic and reconstructive surgery. 2002;18(6):443-449;
discussion 450.
163. Ludwig IH, Brown MS. Strabismus due to flap tear of a rectus muscle. Transactions of
the American Ophthalmological Society. 2001;99:53-62; discussion 62-53.
164. Wojno TH. The incidence of extraocular muscle and cranial nerve palsy in orbital floor
blow-out fractures. Ophthalmology. 1987;94(6):682-687.
165. al-Qurainy IA. Convergence insufficiency and failure of accommodation following
midfacial trauma. The British journal of oral & maxillofacial surgery. 1995;33(2):71-75.
166. Baker RS, Epstein AD. Ocular motor abnormalities from head trauma. Survey of
ophthalmology. 1991;35(4):245-267.
167. al-Qurainy IA, Stassen LF, Dutton GN, Moos KF, el-Attar A. Diplopia following
midfacial fractures. The British journal of oral & maxillofacial surgery. 1991;29(5):302-
307.
168. Putterman AM, Stevens T, Urist MJ. Nonsurgical management of blow-out fractures of
the orbital floor. American journal of ophthalmology. 1974;77(2):232-239.
169. Hawes MJ, Dortzbach RK. Surgery on orbital floor fractures. Influence of time of repair
and fracture size. Ophthalmology. 1983;90(9):1066-1070.
170. Cole HG, Smith B. Eye muscle imbalance complicating orbital floor fractures. American
journal of ophthalmology. 1963;55:930-935.
171. Van Eeckhoutte L, De Clippeleir L, Apers R, Van Lammeren M, Janssens H, Baekeland
L. A protocol for extraocular muscle surgery after orbital floor fracture ("blow-out").
Binocular vision & strabismus quarterly. 1998;13(1):29-36.

P425
 Adult Strabismus PPP

172. Thacker NM, Velez FG, Demer JL, Rosenbaum AL. Strabismic complications following
endoscopic sinus surgery: diagnosis and surgical management. Journal of AAPOS : the
official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2004;8(5):488-494.
173. Bhatti MT, Stankiewicz JA. Ophthalmic complications of endoscopic sinus surgery.
Survey of ophthalmology. 2003;48(4):389-402.
174. Galli M. Diplopia following cosmetic surgery. The American orthoptic journal.
2012;62:19-21.
175. Wong TY, Klein BE, Klein R. The prevalence and 5-year incidence of ocular trauma.
The Beaver Dam Eye Study. Ophthalmology. 2000;107(12):2196-2202.
176. Loon SC, Tay WT, Saw SM, Wang JJ, Wong TY. Prevalence and risk factors of ocular
trauma in an urban south-east Asian population: the Singapore Malay Eye Study. Clinical
& experimental ophthalmology. 2009;37(4):362-367.
177. Wang JD, Xu L, Wang YX, You QS, Zhang JS, Jonas JB. Prevalence and incidence of
ocular trauma in North China: the Beijing Eye Study. Acta ophthalmologica.
2012;90(1):e61-67.
178. Henson KJ. Diagnostic tools in the evaluation of strabismus secondary to trauma.
American Orthoptic Journal. 2004;54:2-6.
179. Mauriello JA, Jr., Antonacci R, Mostafavi R, et al. Combined paresis and restriction of
the extraocular muscles after orbital fracture: a study of 16 patients. Ophthalmic plastic
and reconstructive surgery. 1996;12(3):206-210.
180. Borumandi F, Rippel C, Gaggl A. Orbital trauma and its impact on the heart. BMJ case
reports. 2014;2014.
181. Shin GS, Demer JL, Rosenbaum AL. High resolution, dynamic, magnetic resonance
imaging in complicated strabismus. Journal of pediatric ophthalmology and strabismus.
1996;33(6):282-290.
182. Demer JL, Clark RA, Kono R, Wright W, Velez F, Rosenbaum AL. A 12-year,
prospective study of extraocular muscle imaging in complex strabismus. Journal of
AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2002;6(6):337-347.
183. Ela-Dalman N, Velez FG, Rosenbaum AL. Importance of sagittal orbital imaging in
evaluating extraocular muscle trauma following endoscopic sinus surgery. The British
journal of ophthalmology. 2006;90(6):682-685.
184. Ward TP, Thach AB, Madigan WP, Jr., Berland JE. Magnetic resonance imaging in
posttraumatic strabismus. Journal of pediatric ophthalmology and strabismus.
1997;34(2):131-134.
185. Ortube MC, Rosenbaum AL, Goldberg RA, Demer JL. Orbital imaging demonstrates
occult blow out fracture in complex strabismus. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2004;8(3):264-273.
186. Manfredi SJ, Raji MR, Sprinkle PM, Weinstein GW, Minardi LM, Swanson TJ.
Computerized tomographic scan findings in facial fractures associated with blindness.
Plastic and reconstructive surgery. 1981;68(4):479-490.
187. Emery JM, Noorden GK, Sclernitzauer DA. Orbital floor fractures: long-term follow-up
of cases with and without surgical repair. Transactions - American Academy of

P426
 Adult Strabismus PPP

Ophthalmology and Otolaryngology American Academy of Ophthalmology and

Otolaryngology. 1971;75(4):802-812.
188. Burnstine MA. Clinical recommendations for repair of isolated orbital floor fractures: an
evidence-based analysis. Ophthalmology. 2002;109(7):1207-1210; discussion 1210-1201;
quiz 1212-1203.
189. Jordan DR, Allen LH, White J, Harvey J, Pashby R, Esmaeli B. Intervention within days
for some orbital floor fractures: the white-eyed blowout. Ophthalmic plastic and
reconstructive surgery. 1998;14(6):379-390.
190. Abrishami M, Aletaha M, Bagheri A, Salour SH, Yazdani S. Traumatic subluxation of
the globe into the maxillary sinus. Ophthalmic Plast Reconstr Surg. 2007;23(2):156-158.
191. Pelton RW, Rainey AM, Lee AG. Traumatic subluxation of the globe into the maxillary
sinus. AJNR American journal of neuroradiology. 1998;19(8):1450-1451.
192. Silverman N, Spindle J, Tang SX, et al. Orbital floor fracture with entrapment: Imaging
and clinical correlations in 45 cases. Orbit. 2017;36(5):331-336.
193. Biesman BS, Hornblass A, Lisman R, Kazlas M. Diplopia after surgical repair of orbital
floor fractures. Ophthalmic plastic and reconstructive surgery. 1996;12(1):9-16;
discussion 17.
194. Clark RA. The Role of Extraocular Muscle Pulleys in Incomitant Non-Paralytic
Strabismus. Middle East African journal of ophthalmology. 2015;22(3):279-285.
195. Hall LS, McCann JD, Goldberg RA, Santiago AP, Rosenbaum AL. Strabismus after
orbital fractures and sinus surgery. In: Rosenbaum AL, Santiago AP, eds. Clinical
strabismus management Philadelphia: W.B. Saunders Company; 1999:309-322.
196. Shah HA, Shipchandler T, Vernon D, et al. Extra-ocular movement restriction and
diplopia following orbital fracture repair. American journal of otolaryngology.
2018;39(1):34-36.
197. Kersey TL, Ng SG, Rosser P, Sloan B, Hart R. Orbital adherence with titanium mesh
floor implants: a review of 10 cases. Orbit. 2013;32(1):8-11.
198. Tiedemann LM, Lefebvre DR, Wan MJ, Dagi LR. Iatrogenic inferior oblique palsy:
intentional disinsertion during transcaruncular approach to orbital fracture repair. Journal
of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2014;18(5):511-514.
199. Pineles SL, Laursen J, Goldberg RA, Demer JL, Velez FG. Function of transected or
avulsed rectus muscles following recovery using an anterior orbitotomy approach.
Journal of AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2012;16(4):336-341.
200. Thacker NM, Velez FG, Rosenbaum AL. Combined adjustable rectus muscle resection--
recession for incomitant strabismus. Journal of AAPOS : the official publication of the
American Association for Pediatric Ophthalmology and Strabismus. 2005;9(2):137-140.
201. Sobol EK, Rosenberg JB. Strabismus After Ocular Surgery. Journal of pediatric
ophthalmology and strabismus. 2017;54(5):272-281.
202. Capo H, Guyton DL. Ipsilateral hypertropia after cataract surgery. Ophthalmology.
1996;103(5):721-730.
203. Rainin EA, Carlson BM. Postoperative diplopia and ptosis. A clinical hypothesis based
on the myotoxicity of local anesthetics. Arch Ophthalmol. 1985;103(9):1337-1339.

P427
 Adult Strabismus PPP

204. Porter JD, Edney DP, McMahon EJ, Burns LA. Extraocular myotoxicity of the
retrobulbar anesthetic bupivacaine hydrochloride. Investigative ophthalmology & visual
science. 1988;29(2):163-174.
205. Scott AB, Alexander DE, Miller JM. Bupivacaine injection of eye muscles to treat
strabismus. The British journal of ophthalmology. 2007;91(2):146-148.
206. Debert I, Miller JM, Danh KK, Scott AB. Pharmacologic injection treatment of comitant
strabismus. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2016;20(2):106-111 e102.
207. Capo H, Roth E, Johnson T, Munoz M, Siatkowski RM. Vertical strabismus after cataract
surgery. Ophthalmology. 1996;103(6):918-921.
208. Hunter DG, Lam GC, Guyton DL. Inferior oblique muscle injury from local anesthesia
for cataract surgery. Ophthalmology. 1995;102(3):501-509.
209. Erie JC. Acquired Brown's syndrome after peribulbar anesthesia. American journal of
ophthalmology. 1990;109(3):349-350.
210. Spierer A, Schwalb E. Superior oblique muscle paresis after sub-Tenon's anesthesia for
cataract surgery. J Cataract Refract Surg. 1999;25(1):144-145.
211. Costa PG, Debert I, Passos LB, Polati M. Persistent diplopia and strabismus after cataract
surgery under local anesthesia. Binocular vision & strabismus quarterly. 2006;21(3):155-
158.
212. Hamed LM, Lingua RW. Thyroid eye disease presenting after cataract surgery. Journal
of pediatric ophthalmology and strabismus. 1990;27(1):10-15.
213. Sun PY, Leske DA, Holmes JM, Khanna CL. Diplopia in Medically and Surgically
Treated Patients with Glaucoma. Ophthalmology. 2017;124(2):257-262.
214. Dobler-Dixon AA, Cantor LB, Sondhi N, Ku WS, Hoop J. Prospective evaluation of
extraocular motility following double-plate molteno implantation. Arch Ophthalmol.
1999;117(9):1155-1160.
215. Frank JW, Perkins TW, Kushner BJ. Ocular motility defects in patients with the Krupin
valve implant. Ophthalmic surgery. 1995;26(3):228-232.
216. Coats DK, Paysse EA, Orenga-Nania S. Acquired Pseudo-Brown's syndrome
immediately following Ahmed valve glaucoma implant. Ophthalmic Surg Lasers.
1999;30(5):396-397.
217. Dobler AA, Sondhi N, Cantor LB, Ku S. Acquired Brown's syndrome after a double-
plate Molteno implant. American journal of ophthalmology. 1993;116(5):641-642.
218. Roizen A, Ela-Dalman N, Velez FG, Coleman AL, Rosenbaum AL. Surgical treatment of
strabismus secondary to glaucoma drainage device. Arch Ophthalmol. 2008;126(4):480-
486.
219. Munoz M, Rosenbaum AL. Long-term strabismus complications following retinal
detachment surgery. Journal of pediatric ophthalmology and strabismus. 1987;24(6):309-
314.
220. Hwang JM, Wright KW. Combined study on the causes of strabismus after the retinal
surgery. Korean journal of ophthalmology : KJO. 1994;8(2):83-91.
221. Macleod JD, Morris RJ. Detached superior rectus following scleral buckling: anatomy
and surgical management. Eye (Lond). 1997;11 ( Pt 1):30-32.
222. Chang MY, Yulek F, Pineles SL, Velez FG. Surgery for superior oblique tendon
anteriorization and entrapment following scleral buckle. Journal of AAPOS : the official

P428
 Adult Strabismus PPP

publication of the American Association for Pediatric Ophthalmology and Strabismus.

2016;20(2):165 e161-162.
223. Wright LA, Cleary M, Barrie T, Hammer HM. Motility and binocularity outcomes in
vitrectomy versus scleral buckling in retinal detachment surgery. Graefe's archive for
clinical and experimental ophthalmology = Albrecht von Graefes Archiv fur klinische
und experimentelle Ophthalmologie. 1999;237(12):1028-1032.
224. Raab EL, Metz HS, Ellis FD. Medial rectus injury after pterygium excision. Arch
Ophthalmol. 1989;107(10):1428.
225. Ela-Dalman N, Velez FG, Rosenbaum AL. Incomitant esotropia following pterygium
excision surgery. Arch Ophthalmol. 2007;125(3):369-373.
226. Laria C, Shokida F, Tatarchuck P, Pinero DP, Gonzalez X. New diplopic restrictive
strabismus as a sequela after conjunctival surgery for conjunctival lesions: a series of 3
cases, management and outcome. Binocular vision & strabology quarterly, Simms-
Romano's. 2012;27(2):113-121.
227. Ugrin MC, Molinari A. Disinsertion of the medial rectus following pterygium surgery:
signs and management. Strabismus. 1999;7(3):147-152.
228. Neely KA, Ernest JT, Mottier M. Combined superior oblique paresis and Brown's
syndrome after blepharoplasty. American journal of ophthalmology. 1990;109(3):347-
349.
229. Syniuta LA, Goldberg RA, Thacker NM, Rosenbaum AL. Acquired strabismus following
cosmetic blepharoplasty. Plastic and reconstructive surgery. 2003;111(6):2053-2059.
230. Pirouzian A, Goldberg RA, Demer JL. Inferior rectus pulley hindrance: a mechanism of
restrictive hypertropia following lower lid surgery. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2004;8(4):338-344.
231. Jameson NA, Good WV, Hoyt CS. Fat adherence simulating inferior oblique palsy
following blepharoplasty. Arch Ophthalmol. 1992;110(10):1369.
232. Johnson DA. Persistent vertical binocular diplopia after cataract surgery. American
journal of ophthalmology. 2001;132(6):831-835.
233. Brown SM, Brooks SE, Mazow ML, et al. Cluster of diplopia cases after periocular
anesthesia without hyaluronidase. J Cataract Refract Surg. 1999;25(9):1245-1249.
234. Hamada S, Devys JM, Xuan TH, et al. Role of hyaluronidase in diplopia after peribulbar
anesthesia for cataract surgery. Ophthalmology. 2005;112(5):879-882.
235. Rauscher FM, Gedde SJ, Schiffman JC, et al. Motility disturbances in the tube versus
trabeculectomy study during the first year of follow-up. American journal of
ophthalmology. 2009;147(3):458-466.
236. Abdelaziz A, Capo H, Banitt MR, et al. Diplopia after glaucoma drainage device
implantation. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2013;17(2):192-196.
237. Berk AT, Saatci AO, Kir E, Durak I, Kaynak S. Extraocular muscle imbalance after
scleral buckling. Strabismus. 1996;4(2):69-75.
238. Smiddy WE, Loupe D, Michels RG, Enger C, Glaser BM, deBustros S. Extraocular
muscle imbalance after scleral buckling surgery. Ophthalmology. 1989;96(10):1485-
1489; discussion 1489-1490.

P429
 Adult Strabismus PPP

239. Goezinne F, Berendschot TT, van Daal EW, et al. Diplopia was not predictable and not
associated with buckle position after scleral buckling surgery for retinal detachment.
Retina. 2012;32(8):1514-1524.
240. Hayworth RS, Lisman RD, Muchnick RS, Smith B. Diplopia following blepharoplasty.
Annals of ophthalmology. 1984;16(5):448-451.
241. Munoz M, Parrish RK, 2nd. Strabismus following implantation of Baerveldt drainage
devices. Arch Ophthalmol. 1993;111(8):1096-1099.
242. Wu TE, Rosenbaum AL, Demer JL. Severe strabismus after scleral buckling: multiple
mechanisms revealed by high-resolution magnetic resonance imaging. Ophthalmology.
2005;112(2):327-336.
243. Wong V, Kasbekar S, Young J, Stappler T, Marsh IB, Durnian JM. The effect of scleral
exoplant removal on strabismus following retinal detachment repair. Journal of AAPOS :
the official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2011;15(4):331-333.
244. Rabinowitz R, Velez FG, Pineles SL. Risk factors influencing the outcome of strabismus
surgery following retinal detachment surgery with scleral buckle. Journal of AAPOS : the
official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2013;17(6):594-597.
245. Chen CJ, Kosek K, Benvenutti E. Outcomes and complications of hydrogel scleral
explant removal. Ophthalmic surgery, lasers & imaging : the official journal of the
International Society for Imaging in the Eye. 2012;43(5):383-387.
246. Petitto VB, Buckley EG. Use of botulinum toxin in strabismus after retinal detachment
surgery. Ophthalmology. 1991;98(4):509-512; discussion 512-503.
247. Dosunmu EO, Hatt SR, Leske DA, Hodge DO, Holmes JM. Incidence and Etiology of
Presumed Fourth Cranial Nerve Palsy: A Population-based Study. American journal of
ophthalmology. 2018;185:110-114.
248. Akbari MR, Khorrami Nejad M, Askarizadeh F, Pour FF, Ranjbar Pazooki M,
Moeinitabar MR. Facial asymmetry in ocular torticollis. Journal of current
ophthalmology. 2015;27(1-2):4-11.
249. Rucker CW. Paralysis of the third, fourth and sixth cranial nerves. American journal of
ophthalmology. 1958;46(6):787-794.
250. Rucker CW. The causes of paralysis of the third, fourth and sixth cranial nerves.
American journal of ophthalmology. 1966;61(5 Pt 2):1293-1298.
251. Rush JA, Younge BR. Paralysis of cranial nerves III, IV, and VI. Cause and prognosis in
1,000 cases. Arch Ophthalmol. 1981;99(1):76-79.
252. Brodsky MC, Donahue SP, Vaphiades M, Brandt T. Skew deviation revisited. Survey of
ophthalmology. 2006;51(2):105-128.
253. Mollan SP, Edwards JH, Price A, Abbott J, Burdon MA. Aetiology and outcomes of
adult superior oblique palsies: a modern series. Eye (Lond). 2009;23(3):640-644.
254. Chen VM, Dagi LR. Ocular misalignment in Graves disease may mimic that of superior
oblique palsy. Journal of neuro-ophthalmology : the official journal of the North
American Neuro-Ophthalmology Society. 2008;28(4):302-304.
255. Plager DA. Tendon laxity in superior oblique palsy. Ophthalmology. 1992;99(7):1032-
1038.
256. Archer SM. Management of paretic vertical deviations. The American orthoptic journal.
2011;61:6-12.

P430
 Adult Strabismus PPP

257. Bata BM, Leske DA, Holmes JM. Adjustable Bilateral Superior Oblique Tendon
Advancement for Bilateral Fourth Nerve Palsy. American journal of ophthalmology.
2017;178:115-121.
258. Fells P. Management of paralytic strabismus. The British journal of ophthalmology.
1974;58(3):255-265.
259. Nejad M, Thacker N, Velez FG, Rosenbaum AL, Pineles SL. Surgical results of patients
with unilateral superior oblique palsy presenting with large hypertropias. Journal of
pediatric ophthalmology and strabismus. 2013;50(1):44-52.
260. Aseff AJ, Munoz M. Outcome of surgery for superior oblique palsy with contracture of
ipsilateral superior rectus treated by superior rectus recession. Binocular vision &
strabismus quarterly. 1998;13(3):177-180.
261. Dieterich M, Brandt T. Ocular torsion and tilt of subjective visual vertical are sensitive
brainstem signs. Annals of neurology. 1993;33(3):292-299.
262. Tilikete C, Vighetto A. Internuclear ophthalmoplegia with skew deviation. Two cases
with an isolated circumscribed lesion of the medial longitudinal fasciculus. European
neurology. 2000;44(4):258-259.
263. Zwergal A, Cnyrim C, Arbusow V, et al. Unilateral INO is associated with ocular tilt
reaction in pontomesencephalic lesions: INO plus. Neurology. 2008;71(8):590-593.
264. Brandt T, Dieterich M. Pathological eye-head coordination in roll: tonic ocular tilt
reaction in mesencephalic and medullary lesions. Brain : a journal of neurology.
1987;110 ( Pt 3):649-666.
265. Dieterich M, Brandt T. Thalamic infarctions: differential effects on vestibular function in
the roll plane (35 patients). Neurology. 1993;43(9):1732-1740.
266. Rabinovitch HE, Sharpe JA, Sylvester TO. The ocular tilt reaction. A paroxysmal
dyskinesia associated with elliptical nystagmus. Arch Ophthalmol. 1977;95(8):1395-
1398.
267. Brandt T, Dieterich M. Skew deviation with ocular torsion: a vestibular brainstem sign of
topographic diagnostic value. Annals of neurology. 1993;33(5):528-534.
268. Lemos J, Subei A, Sousa M, et al. Differentiating Acute and Subacute Vertical
Strabismus Using Different Head Positions During the Upright-Supine Test. JAMA
ophthalmology. 2018;136(4):322-328.
269. Buckley SA, Elston JS. Surgical treatment of supranuclear and internuclear ocular
motility disorders. Eye (Lond). 1997;11 ( Pt 3):377-380.
270. Lee J. Management of selected forms of neurogenic strabismus. In: Rosenbaum AL,
Santiago AP, eds. Clinical Strabismus Management. Principles and Surgical Techniques.
Philadelphia: W. B. Saunders Company; 1999:389-391.
271. Brodsky MC, Holmes JM. Torsional augmentation for the treatment of lateropulsion and
torticollis in partial ocular tilt reaction. Journal of AAPOS : the official publication of the
American Association for Pediatric Ophthalmology and Strabismus. 2012;16(2):141-144.
272. Zadro I, Barun B, Habek M, Brinar VV. Isolated cranial nerve palsies in multiple
sclerosis. Clinical neurology and neurosurgery. 2008;110(9):886-888.
273. Patel SV, Mutyala S, Leske DA, Hodge DO, Holmes JM. Incidence, associations, and
evaluation of sixth nerve palsy using a population-based method. Ophthalmology.
2004;111(2):369-375.
274. King AJ, Stacey E, Stephenson G, Trimble RB. Spontaneous recovery rates for unilateral
sixth nerve palsies. Eye (Lond). 1995;9 ( Pt 4):476-478.

P431
 Adult Strabismus PPP

275. Durkin SR, Tennekoon S, Kleinschmidt A, Casson RJ, Selva D, Crompton JL. Bilateral
sixth nerve palsy. Ophthalmology. 2006;113(11):2108-2109.
276. Holmes JM, Droste PJ, Beck RW. The natural history of acute traumatic sixth nerve
palsy or paresis. Journal of AAPOS : the official publication of the American Association
for Pediatric Ophthalmology and Strabismus. 1998;2(5):265-268.
277. Rassi MS, Hulou MM, Almefty K, et al. Pediatric Clival Chordoma: A Curable Disease
that Conforms to Collins' Law. Neurosurgery. 2017.
278. Santiago AP, Rosenbaum AL. Sixth cranial nerve palsy. In: Rosenbaum AL, Santiago
AP, eds. Clinical strabismus management : principles and surgical techniques.
Philadelphia: Saunders; 1999:259-271.
279. Li C, Tang Y, Ge H, et al. Sectional anatomy of the abducens nerve: according to 3D-
SPACE magnetic resonance sequences correlated with cryosectional specimens. Surgical
and radiologic anatomy : SRA. 2015;37(8):921-929.
280. Linskey ME, Sekhar LN, Hirsch W, Jr., Yonas H, Horton JA. Aneurysms of the
intracavernous carotid artery: clinical presentation, radiographic features, and
pathogenesis. Neurosurgery. 1990;26(1):71-79.
281. Morard M, Tcherekayev V, de Tribolet N. The superior orbital fissure: a microanatomical
study. Neurosurgery. 1994;35(6):1087-1093.
282. Elder C, Hainline C, Galetta SL, Balcer LJ, Rucker JC. Isolated Abducens Nerve Palsy:
Update on Evaluation and Diagnosis. Current neurology and neuroscience reports.
2016;16(8):69.
283. Holmes JM, Leske DA, Christiansen SP. Initial treatment outcomes in chronic sixth
nerve palsy. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2001;5(6):370-376.
284. Scott AB, Kraft SP. Botulinum toxin injection in the management of lateral rectus
paresis. Ophthalmology. 1985;92(5):676-683.
285. Biglan AW, Burnstine RA, Rogers GL, Saunders RA. Management of strabismus with
botulinum A toxin. Ophthalmology. 1989;96(7):935-943.
286. Mehendale RA, Dagi LR, Wu C, Ledoux D, Johnston S, Hunter DG. Superior rectus
transposition and medial rectus recession for Duane syndrome and sixth nerve palsy.
Arch Ophthalmol. 2012;130(2):195-201.
287. Rosenbaum AL. Costenbader Lecture. The efficacy of rectus muscle transposition
surgery in esotropic Duane syndrome and VI nerve palsy. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2004;8(5):409-419.
288. Foster RS. Vertical muscle transposition augmented with lateral fixation. Journal of
AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 1997;1(1):20-30.
289. Gunton KB. Vertical rectus transpositions in sixth nerve palsies. Current opinion in
ophthalmology. 2015;26(5):366-370.
290. Britt MT, Velez FG, Thacker N, Alcorn D, Foster RS, Rosenbaum AL. Partial rectus
muscle-augmented transpositions in abduction deficiency. Journal of AAPOS : the
official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2003;7(5):325-332.

P432
 Adult Strabismus PPP

291. Holmes JM, Leske DA. Long-term outcomes after surgical management of chronic sixth
nerve palsy. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2002;6(5):283-288.
292. Richards BW, Jones FR, Jr., Younge BR. Causes and prognosis in 4,278 cases of
paralysis of the oculomotor, trochlear, and abducens cranial nerves. American journal of
ophthalmology. 1992;113(5):489-496.
293. Fang C, Leavitt JA, Hodge DO, Holmes JM, Mohney BG, Chen JJ. Incidence and
Etiologies of Acquired Third Nerve Palsy Using a Population-Based Method. JAMA
ophthalmology. 2017;135(1):23-28.
294. Fujiwara S, Fujii K, Nishio S, Matsushima T, Fukui M. Oculomotor nerve palsy in
patients with cerebral aneurysms. Neurosurgical review. 1989;12(2):123-132.
295. Park HK, Rha HK, Lee KJ, Chough CK, Joo W. Microsurgical Anatomy of the
Oculomotor Nerve. Clinical anatomy. 2017;30(1):21-31.
296. Derakhshan I, Sabouri-Deylami M, Kaufman B. Bilateral Nothnagel syndrome. Clinical
and roentgenological observations. Stroke. 1980;11(2):177-179.
297. Duncan GW, Weindling SM. Posterior cerebral artery stenosis with midbrain infarction.
Stroke. 1995;26(5):900-902.
298. Elmalem VI, Hudgins PA, Bruce BB, Newman NJ, Biousse V. Underdiagnosis of
posterior communicating artery aneurysm in noninvasive brain vascular studies. Journal
of neuro-ophthalmology : the official journal of the North American Neuro-
Ophthalmology Society. 2011;31(2):103-109.
299. Chakeres DW, Kapila A. Radiology of the ambient cistern. Part I: Normal.
Neuroradiology. 1985;27(5):383-389.
300. Yasuda A, Campero A, Martins C, Rhoton AL, Jr., de Oliveira E, Ribas GC.
Microsurgical anatomy and approaches to the cavernous sinus. Neurosurgery. 2005;56(1
Suppl):4-27; discussion 24-27.
301. Metz HS. Forced duction, active force generation, and saccadic velocity tests.
International ophthalmology clinics. 1976;16(3):47-73.
302. Tantiwongkosi B, Hesselink JR. Imaging of Ocular Motor Pathway. Neuroimaging
clinics of North America. 2015;25(3):425-438.
303. Talebnejad MR, Sharifi M, Nowroozzadeh MH. The role of Botulinum toxin in
management of acute traumatic third-nerve palsy. Journal of AAPOS : the official
publication of the American Association for Pediatric Ophthalmology and Strabismus.
2008;12(5):510-513.
304. Cabrejas L, Hurtado-Cena FJ, Tejedor J. Predictive factors of surgical outcome in
oculomotor nerve palsy. Journal of AAPOS : the official publication of the American
Association for Pediatric Ophthalmology and Strabismus. 2009;13(5):481-484.
305. Mazow ML. Third cranial nerve palsy: Diagnosis and management strategies. In:
Rosenbaum AL, Santiago AP, eds. Clinical strabismus management : principles and
surgical techniques. Philadelphia: Saunders; 1999:251-258.
306. Yanovitch T, Buckley E. Diagnosis and management of third nerve palsy. Current
opinion in ophthalmology. 2007;18(5):373-378.
307. Deutsch JA, Greenwald MJ, Lingua RW, Nelson LB. Surgical Approaches to Strabismus
After Third Nerve Palsy. Journal of pediatric ophthalmology and strabismus.
2015;52(6):326-330.

P433
 Adult Strabismus PPP

308. Velez FG, Thacker N, Britt MT, Alcorn D, Foster RS, Rosenbaum AL. Rectus muscle
orbital wall fixation: a reversible profound weakening procedure. Journal of AAPOS : the
official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2004;8(5):473-480.
309. Singh A, Pandey PK, Mittal SK, Agrawal A, Bahuguna C, Kumar P. Impact of Superior
Oblique Transposition on Primary Position Deviation, a Pattern and Intorsion in Third
Nerve Palsy. Strabismus. 2016;24(4):173-177.
310. Eraslan M, Cerman E, Onal S, Ogut MS. Superior Oblique Anterior Transposition with
Horizontal Recti Recession-Resection for Total Third-Nerve Palsy. Journal of
ophthalmology. 2015;2015:780139.
311. Shah AS, Prabhu SP, Sadiq MA, Mantagos IS, Hunter DG, Dagi LR. Adjustable nasal
transposition of split lateral rectus muscle for third nerve palsy. JAMA ophthalmology.
2014;132(8):963-969.
312. Gilhus NE, Verschuuren JJ. Myasthenia gravis: subgroup classification and therapeutic
strategies. The Lancet Neurology. 2015;14(10):1023-1036.
313. Pal J, Rozsa C, Komoly S, Illes Z. Clinical and biological heterogeneity of autoimmune
myasthenia gravis. Journal of neuroimmunology. 2011;231(1-2):43-54.
314. O'Brien MD. The miracle at St Alfege's: seventy years on. Journal of the Royal Society of
Medicine. 2007;100(6):257.
315. Chen YL, Yeh JH, Chiu HC. Clinical features of myasthenia gravis patients with
autoimmune thyroid disease in Taiwan. Acta neurologica Scandinavica.
2013;127(3):170-174.
316. Antonio-Santos AA, Eggenberger ER. Medical treatment options for ocular myasthenia
gravis. Current opinion in ophthalmology. 2008;19(6):468-478.
317. Grob D, Arsura EL, Brunner NG, Namba T. The course of myasthenia gravis and
therapies affecting outcome. Annals of the New York Academy of Sciences.
1987;505:472-499.
318. Chen CS, Lee AW, Miller NR, Lee AG. Double vision in a patient with thyroid disease:
what's the big deal? Survey of ophthalmology. 2007;52(4):434-439.
319. Lepore FE, Sanborn GE, Slevin JT. Pupillary dysfunction in myasthenia gravis. Annals of
neurology. 1979;6(1):29-33.
320. Yamazaki A, Ishikawa S. Abnormal pupillary responses in myasthenia gravis. A
pupillographic study. The British journal of ophthalmology. 1976;60(8):575-580.
321. Kubis KC, Danesh-Meyer HV, Savino PJ, Sergott RC. The ice test versus the rest test in
myasthenia gravis. Ophthalmology. 2000;107(11):1995-1998.
322. Chatzistefanou KI, Kouris T, Iliakis E, et al. The ice pack test in the differential diagnosis
of myasthenic diplopia. Ophthalmology. 2009;116(11):2236-2243.
323. Odel JG, Winterkorn JM, Behrens MM. The sleep test for myasthenia gravis. A safe
alternative to Tensilon. Journal of clinical neuro-ophthalmology. 1991;11(4):288-292.
324. Phillips LH, 2nd, Melnick PA. Diagnosis of myasthenia gravis in the 1990s. Seminars in
neurology. 1990;10(1):62-69.
325. Barton JJ, Fouladvand M. Ocular aspects of myasthenia gravis. Seminars in neurology.
2000;20(1):7-20.
326. Padua L, Stalberg E, LoMonaco M, Evoli A, Batocchi A, Tonali P. SFEMG in ocular
myasthenia gravis diagnosis. Clinical neurophysiology : official journal of the
International Federation of Clinical Neurophysiology. 2000;111(7):1203-1207.

P434
 Adult Strabismus PPP

327. Kupersmith MJ, Moster M, Bhuiyan S, Warren F, Weinberg H. Beneficial effects of

corticosteroids on ocular myasthenia gravis. Archives of neurology. 1996;53(8):802-804.
328. Oosterhuis HJ. The natural course of myasthenia gravis: a long term follow up study.
Journal of neurology, neurosurgery, and psychiatry. 1989;52(10):1121-1127.
329. Acheson JF, Elston JS, Lee JP, Fells P. Extraocular muscle surgery in myasthenia gravis.
The British journal of ophthalmology. 1991;75(4):232-235.
330. Davidson JL, Rosenbaum AL, McCall LC. Strabismus surgery in patients with
myasthenia. Journal of pediatric ophthalmology and strabismus. 1993;30(5):292-295.
331. Bentley CR, Dawson E, Lee JP. Active management in patients with ocular
manifestations of myasthenia gravis. Eye (Lond). 2001;15(Pt 1):18-22.
332. Ohtsuki H, Hasebe S, Okano M, Furuse T. Strabismus surgery in ocular myasthenia
gravis. Ophthalmologica Journal international d'ophtalmologie International journal of
ophthalmology Zeitschrift fur Augenheilkunde. 1996;210(2):95-100.
333. Morris OC, O'Day J. Strabismus surgery in the management of diplopia caused by
myasthenia gravis. The British journal of ophthalmology. 2004;88(6):832.
334. Peragallo JH, Velez FG, Demer JL, Pineles SL. Long-term follow-up of strabismus
surgery for patients with ocular myasthenia gravis. Journal of neuro-ophthalmology : the
official journal of the North American Neuro-Ophthalmology Society. 2013;33(1):40-44.
335. Hamed LM. Strabismus after adult cataract surgery. In: Rosenbaum AL, Santiago, A.P.,
eds. Clinical Strabismus Management. Philadelphia: W.B. Saunders Company;
1999:371-379.
336. Kushner BJ. Fixation switch diplopia. Arch Ophthalmol. 1995;113(7):896-899.
337. Arnoldi K. Case corner: complicated fixation switch diplopia. The American orthoptic
journal. 2009;59:111-115.
338. Scott MH, Noble AG, Raymond WRt, Parks MM. Prevalence of primary monofixation
syndrome in parents of children with congenital esotropia. Journal of pediatric
ophthalmology and strabismus. 1994;31(5):298-301; discussion 302.
339. Kushner BJ, Kowal L. Diplopia after refractive surgery: occurrence and prevention. Arch
Ophthalmol. 2003;121(3):315-321.
340. Bixenman WW, Joffe L. Binocular diplopia associated with retinal wrinkling. Journal of
pediatric ophthalmology and strabismus. 1984;21(6):215-219.
341. Benegas NM, Egbert J, Engel WK, Kushner BJ. Diplopia secondary to aniseikonia
associated with macular disease. Arch Ophthalmol. 1999;117(7):896-899.
342. Burgess D, Roper-Hall G, Burde RM. Binocular diplopia associated with subretinal
neovascular membranes. Arch Ophthalmol. 1980;98(2):311-317.
343. De Pool ME, Campbell JP, Broome SO, Guyton DL. The dragged-fovea diplopia
syndrome: clinical characteristics, diagnosis, and treatment. Ophthalmology.
2005;112(8):1455-1462.
344. Ting FS, Kwok AK. Treatment of epiretinal membrane: an update. Hong Kong medical
journal = Xianggang yi xue za zhi. 2005;11(6):496-502.
345. Pournaras CJ, Donati G, Brazitikos PD, Kapetanios AD, Dereklis DL, Stangos NT.
Macular epiretinal membranes. Seminars in ophthalmology. 2000;15(2):100-107.
346. Veverka KK, Hatt SR, Leske DA, et al. Prevalence and Associations of Central-
Peripheral Rivalry-Type Diplopia in Patients With Epiretinal Membrane. JAMA
ophthalmology. 2017;135(12):1303-1309.

P435
 Adult Strabismus PPP

347. Veverka KK, Hatt SR, Leske DA, Brown WL, Iezzi R, Jr., Holmes JM. Causes of
Diplopia in Patients With Epiretinal Membranes. American journal of ophthalmology.
2017;179:39-45.
348. Hatt SR, Leske DA, Klaehn LD, Kramer AM, Iezzi R, Jr., Holmes JM. Treatment for
central-peripheral rivalry-type diplopia ("dragged-fovea diplopia syndrome"). American
journal of ophthalmology. 2019.
349. Silverberg M, Schuler E, Veronneau-Troutman S, Wald K, Schlossman A, Medow N.
Nonsurgical management of binocular diplopia induced by macular pathology. Arch
Ophthalmol. 1999;117(7):900-903.
350. Iacobucci IL, Furr BA, Archer SM. Management of binocular diplopia due to
maculopathy with combined bangerter filter and fresnel prism. The American orthoptic
journal. 2009;59:93-97.
351. De Pool ME, Iezzi R, Jr., Hatt SR, Leske DA, Holmes JM. The Double-edged Sword of
Epiretinal Membrane Surgery: Improving versus Inducing Diplopia. IOVS. 2018;59.
352. Wan MJ, Hunter DG. Complications of strabismus surgery: incidence and risk factors.
Seminars in ophthalmology. 2014;29(5-6):421-428.
353. Bradbury JA, Taylor RH. Severe complications of strabismus surgery. Journal of AAPOS
: the official publication of the American Association for Pediatric Ophthalmology and
Strabismus. 2013;17(1):59-63.
354. Escardo-Paton JA, Harrad RA. Duration of conjunctival redness following adult
strabismus surgery. Journal of AAPOS : the official publication of the American
Association for Pediatric Ophthalmology and Strabismus. 2009;13(6):583-586.
355. Tessler HH, Urist MJ. Corneal dellen in the limbal approach to rectus muscle surgery.
The British journal of ophthalmology. 1975;59(7):377-379.
356. Fresina M, Campos EC. Corneal 'dellen' as a complication of strabismus surgery. Eye
(Lond). 2009;23(1):161-163.
357. Espinoza GM, Lueder GT. Conjunctival pyogenic granulomas after strabismus surgery.
Ophthalmology. 2005;112(7):1283-1286.
358. Guadilla AM, de Liano PG, Merino P, Franco G. Conjunctival cysts as a complication
after strabismus surgery. Journal of pediatric ophthalmology and strabismus.
2011;48(5):298-300.
359. Simon JW, Lininger LL, Scheraga JL. Recognized scleral perforation during eye muscle
surgery: incidence and sequelae. J Pediatr Ophthalmol Strabismus. 1992;29(5):273-275.
360. Dang Y, Racu C, Isenberg SJ. Scleral penetrations and perforations in strabismus surgery
and associated risk factors. Journal of AAPOS : the official publication of the American
Association for Pediatric Ophthalmology and Strabismus. 2004;8(4):325-331.
361. Noel LP, Bloom JN, Clarke WN, Bawazeer A. Retinal perforation in strabismus surgery.
Journal of pediatric ophthalmology and strabismus. 1997;34(2):115-117.
362. Morris RJ, Rosen PH, Fells P. Incidence of inadvertent globe perforation during
strabismus surgery. The British journal of ophthalmology. 1990;74(8):490-493.
363. Apt L, Isenberg S, Gaffney WL. The oculocardiac reflex in strabismus surgery. American
journal of ophthalmology. 1973;76(4):533-536.
364. Min SW, Hwang JM. The incidence of asystole in patients undergoing strabismus
surgery. Eye (Lond). 2009;23(4):864-866.
365. Ing MR. Infection following strabismus surgery. Ophthalmic surgery. 1991;22(1):41-43.

P436
 Adult Strabismus PPP

366. Kivlin JD, Wilson ME, Jr. Periocular infection after strabismus surgery. The Periocular
Infection Study Group. Journal of pediatric ophthalmology and strabismus.
1995;32(1):42-49.
367. Mittelman D, Bakos IM. The role of retinal cryopexy in the management of experimental
perforation of the eye during strabismus surgery. Journal of pediatric ophthalmology and
strabismus. 1984;21(5):186-189.
368. Parks MM. The weakening surgical procedures for eliminating overaction of the inferior
oblique muscle. American journal of ophthalmology. 1972;73(1):107-122.
369. Todd B, Sullivan TJ, Gole GA. Delayed orbital hemorrhage after routine strabismus
surgery. American journal of ophthalmology. 2001;131(6):818-819.
370. Ares C, Superstein R. Retrobulbar hemorrhage following strabismus surgery. Journal of
AAPOS : the official publication of the American Association for Pediatric
Ophthalmology and Strabismus. 2006;10(6):594-595.
371. Shen E, Porco T, Rutar T. Errors in strabismus surgery. JAMA ophthalmology.
2013;131(1):75-79.
372. Eke T. Preoperative Preparation and Anesthesia for Trabeculectomy. Journal of current
glaucoma practice. 2016;10(1):21-35.
373. Kemp PS, Larson SA, Drack AV. Strabismus surgery in patients receiving warfarin
anticoagulation. Journal of AAPOS : the official publication of the American Association
for Pediatric Ophthalmology and Strabismus. 2014;18(1):84-86.
374. Nihalani BR, Hunter DG. Adjustable suture strabismus surgery. Eye (Lond).
2011;25(10):1262-1276.
375. Saxena R, Singh D, Chandra A, Sharma P. Adjustable anterior and nasal transposition of
inferior oblique muscle in case of torsional diplopia in superior oblique palsy. Indian
Journal of Clinical and Experimental Ophthalmology. 2015;1(2):104-106.
376. Yim HB, Biglan AW, Cronin TH. Graded partial tenotomy of vertical rectus muscles for
treatment of hypertropia. Transactions of the American Ophthalmological Society.
2004;102:169-175; discussion 175-166.
377. Singh J, Choi CS, Bahl R, Archer SM. Partial tendon recession for small-angle vertical
strabismus. Journal of AAPOS : the official publication of the American Association for
Pediatric Ophthalmology and Strabismus. 2016;20(5):392-395.
378. Rowe FJ, Noonan CP. Botulinum toxin for the treatment of strabismus. The Cochrane
database of systematic reviews. 2017;3:CD006499.
379. Simon JW. Complications of strabismus surgery. Current opinion in ophthalmology.
2010;21(5):361-366.
380. Al-Haddad C, Abdul Fattah M. Slipped extraocular muscles: characteristics and surgical
outcomes. Canadian journal of ophthalmology Journal canadien d'ophtalmologie.
2017;52(1):42-47.
381. Parks MM, Bloom JN. The "slipped" muscle. Ophthalmology. 1979;86(8):1389-1396.
382. Olitsky SE, Juneja RG. Orbital hemorrhage after the administration of sub-Tenon's
infusion anesthesia. Ophthalmic Surg Lasers. 1997;28(2):145-146.

P437