See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/256333562

Proximal Humerus and Humeral Shaft Nonunions

Article  in  The Journal of the American Academy of Orthopaedic Surgeons · September 2013

DOI: 10.5435/JAAOS-21-09-538 · Source: PubMed

CITATIONS READS
44 1,674

5 authors, including:

Edwin R Cadet Brian Schulz

Raleigh Orthopaedic Clinic Kerlan-Jobe Orthopaedic Clinic
46 PUBLICATIONS   891 CITATIONS    24 PUBLICATIONS   237 CITATIONS   

SEE PROFILE SEE PROFILE

Christopher S Ahmad Melvin P Rosenwasser

Columbia University Columbia University
306 PUBLICATIONS   7,358 CITATIONS    189 PUBLICATIONS   4,623 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Partial Patellar Tendon Tears View project

Prevalence of vitamin d insufficiency in professional hockey players View project

All content following this page was uploaded by Edwin R Cadet on 31 December 2014.

The user has requested enhancement of the downloaded file.

 Review Article

Proximal Humerus and Humeral

Shaft Nonunions

Abstract
Edwin R. Cadet, MD The rate of nonunion is estimated to be 1.1% to 10% following
Bob Yin, MD closed treatment of proximal humerus fracture and 5.5% following
closed treatment of humeral shaft fracture. Surgical management
Brian Schulz, MD
should be considered for fractures that demonstrate no evidence of
Christopher S. Ahmad, MD progressive healing on consecutive radiographs taken at least 6 to
Melvin P. Rosenwasser, MD 8 weeks apart during the course of closed treatment. In the case of
proximal humerus nonunion, recent series have demonstrated
union in >90% of patients treated with reconstruction using locking
plates and autogenous bone graft. Shoulder arthroplasty is
reserved as a salvage option in cases in which the humeral head
is not viable or the proximal fragment will not support
osteosynthesis. For humeral shaft nonunions, open reduction and
internal fixation with compression plating and bone graft remains
From Raleigh Orthopaedic Clinic, the standard of care, with a >90% rate of union and good
Raleigh, NC (Dr. Cadet) and the
functional outcomes. Recent studies support the use of locked
Department of Orthopaedic Surgery,
Columbia University Medical Center, compression plates, dual plating, and cortical allograft struts in
New York, NY (Dr. Yin, Dr. Schulz, patients with osteopenic bone.
Dr. Ahmad, and Dr. Rosenwasser).

Dr. Ahmad or an immediate family

member serves as a paid consultant
to Acumed and Arthrex and has
received research or institutional
support from Arthrex, Major League
H umerus fractures comprise ap-
proximately 5% to 8% of all
fractures.1-3 Most humeral shaft frac-
ical comorbidities) and mechanical fac-
tors (eg, inadequate immobilization,
fracture pattern and displacement) may
Baseball, and Stryker.
tures and proximal humerus fractures cause nonunion. These factors must be
Dr. Rosenwasser or an immediate
family member has received initially are managed nonsurgically identified and addressed to maximize
royalties from Biomet, serves as a (>95% and 89.1%, respectively),4 with fracture healing.
paid consultant to Stryker, and some progressing to delayed union or In patients with humerus fracture
serves as a board member, owner,
officer, or committee member of the
nonunion. Humeral fractures that fail nonunions, the surgical goals are to
Osteosynthesis & Trauma Care nonsurgical management are challeng- provide a stable mechanical con-
Foundation. None of the following ing to treat surgically because the same struct that allows for early motion
authors or any immediate family patient-related risk factors that predis- and to create a biologic environment
member has received anything of
value from or has stock or stock posed the fractures to nonunion with favorable to fracture healing. Vari-
options held in a commercial closed treatment impede fracture heal- ous surgical strategies have been pro-
company or institution related ing postoperatively, as well. The hu- posed, but the standard of care is
directly or indirectly to the subject of
merus is among the most common lo- open reduction and internal fixation
this article: Dr. Cadet, Dr. Yin, and
Dr. Schulz. cations for fragility fractures, and (ORIF) with rigid compression plat-
surgeons must be prepared to manage ing and autogenous bone grafting.5
J Am Acad Orthop Surg 2013;21:
538-547 increasing numbers of humeral non- Other techniques that have been
unions. recently described include intramed-
http://dx.doi.org/10.5435/
JAAOS-21-09-538
The risk factors for developing non- ullary fibular strut allograft for atro-
union vary by fracture location. In gen- phic proximal nonunions and dual-
Copyright 2013 by the American
Academy of Orthopaedic Surgeons.
eral, however, both biologic factors (eg, compression plating for osteopenic
disrupted blood supply, smoking, med- humeral shaft nonunions.6-8

538 Journal of the American Academy of Orthopaedic Surgeons

 Edwin R. Cadet, MD, et al

an internist or endocrinologist for pre- ing injury.11 The median time to union
Proximal Humerus operative optimization before any sur- or bridging callus of nonsurgically
Nonunion gical procedure.11,13 managed proximal humerus fractures
is 13 weeks, and an appropriate
Proximal humerus fractures are
among the most common orthopae-
Patient Evaluation workup should be performed at that
Patients with proximal humerus time in the absence of healing.4,10 The
dic injuries worldwide,9,10 with the
nonunion typically report pain, stiff- diagnosis of nonunion in the proxi-
vast majority being low-energy, non-
ness, and disability associated with mal humerus can also be made when
displaced or minimally displaced
shoulder dysfunction. Physical exam- there is no evidence of interval heal-
fractures. Such fractures heal un-
ination usually reveals diminished ing on two consecutive radiographs
eventfully without surgical interven-
taken 6 to 8 weeks apart during the
tion, but those that progress to non- forward elevation, with or without
period of closed treatment. Surgical
union have a negative effect on disuse atrophy of the deltoid and
management is recommended at ap-
overall glenohumeral function and periscapular musculature. Axillary
proximately 3 to 6 months following
the ability to perform activities of nerve function must be assessed, and
injury if an impending nonunion
daily living as early as 6 months fol- electromyography is warranted if
is suspected, given patient- and
lowing injury.4,11 Hanson et al10 pro- neurologic injury is suspected. True
fracture-related risk factors (eg, pre-
spectively followed 124 patients with AP radiographs are taken in the
existing osteopenia or significant
proximal humerus fractures that scapular plane with the shoulder in
fracture displacement with disrup-
were managed nonsurgically. At neutral, internal rotation, and exter-
tion of the soft-tissue envelope). By
1-year follow-up, only 3% required nal rotation. Outlet and axillary ra-
intervening at this time point, such
surgery for fracture nonunion. Simi- diographs should also be included in
action may help to prevent disabling
larly, Court-Brown and McQueen4 the radiographic series.
glenohumeral dysfunction that is as-
reported a nonunion rate of 1.1% in The type of nonunion (eg, hyper-
sociated with chronic proximal hu-
their prospective study of patients trophic versus atrophic) should be
merus nonunions.
treated nonsurgically for proximal defined. Radiographically, hypertro-
humerus fracture. phic nonunions are characterized by
hypertrophic and sclerotic bone ends
Nonsurgical Management
Several risk factors have been impli-
with fracture callus, whereas atro- Nonsurgical management for symp-
cated in the development of nonunion
phic nonunions appear osteopenic tomatic proximal humerus non-
following proximal humerus fracture.
with the absence of callus. In gen- unions is typically reserved for pa-
Court-Brown and McQueen4 found an
eral, hypertrophic nonunions de- tients with medical comorbidities
8% rate of nonunion in patients with
velop when insufficient mechanical that place them at an unacceptable
metaphyseal comminution and a 10%
stability and/or axial alignment ex- risk for surgical management and for
rate in patients with surgical neck
ists and the vascularity and biologic patients who may be at risk for non-
translation between 33% and 100%.
environment for fracture healing is compliance with postoperative reha-
Fracture pattern may contribute to the
preserved. With atrophic nonunion, bilitation and precautions. Patients
risk of nonunion. Two-part surgical
vascularity and the biologic environ- with minimal pain and mild func-
neck fracture is the most common frac-
ment are often compromised, which tional losses may be appropriate can-
ture pattern associated with fracture
causes an inadequate fracture heal- didates for nonsurgical manage-
nonunion.11-14 It is possible that these
ing response. Radiographs also ment.15
seemingly disparate risk factors lead to
nonunion via the same general mech- should be evaluated for evidence of
anism—all are signs of increased dis- osteonecrosis of the humeral head, Surgical Management
ruption to the medial soft tissues and pathologic fracture, and extent of Osteosynthesis
blood supply that are important for bone loss. Comparison views of the Osteosynthesis using locking plate
fracture healing. Persons who smoke contralateral shoulder may be help- fixation techniques is preferred in the
are at 5.5 times higher risk than non- ful. CT is a useful modality if the di- presence of good bone quality and a
smokers for developing nonunion.10 agnosis of nonunion is unclear. viable humeral head in the absence
Significant medical comorbidities (eg, of significant medial calcar commi-
diabetes, osteopenia, obesity) may con- Timing of Surgery nution or osteopenia that may com-
tribute to nonunion, and patients with Nonunion in long bone fractures is typ- promise adequate fixation. Clinical
such conditions should be referred to ically diagnosed 6 to 9 months follow- and radiographic assessment of tu-

September 2013, Vol 21, No 9 539

Proximal Humerus and Humeral Shaft Nonunions

berosity function and its integrity is recently, large volumes of autoge- tients treated with ORIF using a
critical in deciding whether osteosyn- nous bone graft have been harvested blade plate or Humerusblock device
thesis is the appropriate treatment using the Reamer-Irrigator-Aspirator (Synthes) without bone grafting to
choice for proximal humerus non- (RIA) system (Synthes) to fill large manage humeral surgical neck frac-
unions. With surgical neck non- cortical defects, facilitate osseous fu- tures (P < 0.01). Radiographic evi-
union, rigid fixation can be achieved sions, and in the treatment of dence of fracture healing was seen in
with a variety of plates, including nonunions.17-19 The RIA is an in- 51 patients (93%).
3.5- and 4.5-mm plates made for the tramedullary canal reaming system
proximal humerus, blade plates, and that collects large amounts of autog- Fixed-angle Locked Plating With
4.5-mm T plates. Fixed-angle lock- enous corticocancellous bone from Fibular Strut Allograft
ing or blade plates provide a biome- the intramedullary canal in a rela- Use of an intercalary strut allograft
chanically stable construct in the set- tively minimally invasive manner. with fixed-angle locked plating to
ting of osteoporotic bone. This technique may lessen the donor manage proximal humerus fracture
Isolated greater and lesser tuberos- site morbidity that is sometimes seen nonunion was first described by Bad-
ity nonunions are less common than with iliac crest autograft harvest. man et al15 in 2006. Fibular strut al-
surgical neck nonunions. The bone Compared with iliac crest bone graft, lograft has several advantages. It
quality of the tuberosity fragment RIA may also generate a larger vol- provides additional biologic and
and rotator cuff function are critical ume of autogenous bone graft mate- structural support to the often poor-
components in determining the most rial and pluripotential mesenchymal quality bone found at the proximal
appropriate surgical option. In pa- stem cells, especially in elderly pa- humerus, it is mechanically stronger
tients with large tuberosity fragments tients.20,21 Free vascularized fibular than cancellous bone autograft or al-
and a viable rotator cuff, osteosyn- autograft may be considered for pa- lograft, and it avoids the donor site
thesis may be achieved with lag tients who need significant biologic morbidity associated with autograft
screw compression and/or buttress augmentation along with mechanical harvesting. This technique is useful
plating with autogenous bone graft. support. in both the acute proximal humerus
Tension band techniques, transosse- ORIF with osteosynthesis and fracture setting and chronic non-
ous suture fixation, or current suture bone graft has yielded good results. union scenario when medial calcar
anchor configurations used in mod- Healy et al13 reported union in 12 of support is compromised secondary
ern rotator cuff repair techniques 13 patients following ORIF with to significant medial calcar commi-
that provide compression across the bone graft in a retrospective review nution or osteopenia (Figure 1).
fracture site with autogenous bone of 25 patients with proximal hu-
grafting augmentation can be used merus nonunions. Ring et al22 used Intramedullary Nailing
for comminuted tuberosity frag- blade plating and autogenous iliac Historically, results were disappoint-
ments, only if rotator cuff function is crest cancellous bone graft (ICBG) to ing following intramedullary nailing
determined to be intact clinically. A treat 25 patients with proximal hu- to manage proximal humerus non-
deltoid-splitting or deltopectoral ap- merus fracture nonunion. Fracture unions. Early intramedullary devices
proach can be used for greater tuber- union was achieved in 23 patients were prone to postoperative sub-
osity osteosynthesis. A deltopectoral (92%), and functional results were acromial impingement, necessitating
approach is suggested for lesser tu- classified as good to excellent in 20 a second surgery following union for
berosity nonunions. Arthroscopic (80%). Two patients had complica- nail removal. Most patients, how-
techniques have also been described tions due to iliac crest harvest. Al- ever, progressed to union and re-
for managing greater tuberosity non- lende and Allende23 reported union gained good shoulder function.11
unions.16 in all seven patients treated with a Recently, Yamane et al14 published
Autogenous or allograft bone aug- locking 90° blade plate (average encouraging results with the use of
mentation is recommended to facili- follow-up, 22 months). The average interlocking intramedullary nails to
tate osteosynthesis. Large amounts time to union was 5.9 months. Aver- manage proximal humerus fracture
of cancellous bone autograft can be age Disabilities of the Arm, Shoulder, nonunion in 13 patients. The average
obtained from the iliac crest, but the and Hand and Constant scores at follow-up was 37.8 months. All pa-
patient must be advised of the possi- latest follow-up were 25 and 72.7 tients achieved union. All patients
bility of donor site pain. Allograft points, respectively. More recently, were satisfied with the results and
may be used instead if donor site Tauber et al24 reported significantly had improved shoulder range of mo-
morbidity is unacceptable.11 More improved Constant scores in 55 pa- tion postoperatively. Two patients

540 Journal of the American Academy of Orthopaedic Surgeons

 Edwin R. Cadet, MD, et al

subsequently underwent removal of Figure 1

proximal interlocking screws be-
cause of screw back-out.

Unconstrained and Reverse

Total Shoulder Arthroplasty
The decision to perform uncon-
strained arthroplasty (ie, hemiarthro-
plasty, total shoulder arthroplasty) to
manage proximal humerus fractures
depends in part on the degree of os-
teopenia present, the viability of the
humeral head and, most important,
tuberosity integrity and position as
well as rotator cuff functional status.
Total shoulder replacement is consid-
ered in the setting of concomitant
glenohumeral osteoarthrosis with a
functional rotator cuff.
Boileau et al25 investigated factors
important to successful patient selec-
tion for unconstrained arthroplasty
(ie, hemiarthroplasty, total shoulder
arthroplasty) in the setting of proxi-
mal humerus malunion or nonunion.
They retrospectively reviewed 203
consecutive patients with sequelae of
proximal humerus fractures that had
been managed with unconstrained
glenohumeral arthroplasty. Of the
unconstrained arthroplasties per-
formed, 59% were hemiarthroplasty.
Total shoulder arthroplasty was indi-
cated for patients with preexisting
pain secondary to glenohumeral os-
teoarthrosis or glenoid erosions
Intraoperative AP clinical photograph (A) and corresponding fluoroscopic
noted at the time of surgery. The au- view (B) of an acute surgical neck fracture in an elderly patient with severe
thors suggested that tuberosity integ- osteoporosis who was treated with proximal humeral locking plate fixation
rity and anatomic position is critical and cortical intramedullary fibular strut allograft. B, A locking screw was
inserted through the plate (dashed arrow) to medialize the intramedullary
for a good functional outcome fol- allograft (solid arrow). Other locking screws were inserted through the
lowing unconstrained arthroplasty. allograft, and this push screw was exchanged for a longer locking screw that
Furthermore, they recommended re- also crossed the graft. Intraoperative fluoroscopic AP (C) and axillary (D)
verse total shoulder arthroplasty in views demonstrating multiple locking screws placed through the allograft to
enhance construct rigidity. The medialized position of the allograft provided
cases in which tuberosity osteotomy additional medial calcar support. (Copyright Center for Shoulder, Elbow and
is unavoidable (eg, type 4 fractures). Sports Medicine at Columbia University, New York, NY.)
Although arthroplasty has been
shown to reliably relieve pain in pa- proximal humerus fracture non- preinjury level of activity. Antuña
tients with proximal humerus non- union. All eventually were able to et al12 published the results of 25
union, return to preinjury function is perform activities of daily living and shoulders managed with uncon-
less predictable.11,13 Nayak et al26 ret- had less pain as well as increased strained arthroplasty (mean follow-
rospectively reviewed seven patients function and range of motion. How- up, 6 years). Twenty-one patients un-
who underwent hemiarthroplasty for ever, no patients returned to their derwent hemiarthroplasty, and 4

September 2013, Vol 21, No 9 541

Proximal Humerus and Humeral Shaft Nonunions

Figure 2 union was noted in only 35 shoul-

ders (52%). Anatomic or near ana-
tomic union of the tuberosity was a
significant factor in achieving greater
active forward elevation (P = 0.02).
Reverse total shoulder arthroplasty
is a viable option in the setting of
proximal humerus nonunion or
malunion with humeral head col-
lapse and/or a clinically dysfunc-
tional rotator cuff, radiologic rotator
cuff atrophy (Goutallier stage 2 or
greater), or radiographic evidence of
severe tuberosity malunion or re-
sorption28 (Figure 2). In a study of
18 patients treated with reverse total
shoulder arthroplasty for proximal
humerus nonunion, Martinez et al29
reported significant improvements in
average active forward elevation
(35° to 90°; P < 0.0001), external ro-
tation (15° to 30°; P < 0.0001), and
internal rotation (25° to 55°; P <
0.0001) at an average follow-up of
28 months. Fourteen patients were
either satisfied or very satisfied with
the result of the operation.

Humeral Shaft Nonunion

The vast majority of humeral shaft

fractures heal uneventfully with
functional bracing. Residual angula-
tion of up to 20° in the AP and lat-
eral planes is well tolerated by most
patients and typically does not lead
AP (A) and axillary (B) radiographs demonstrating longstanding right to functional deficits.5,30 However, in
proximal humeral nonunion with humeral head osteonecrosis and resorption
of the tuberosities in an 81-year-old right hand–dominant woman. a recent systematic review of diaphy-
Preoperative range of motion was limited to 40° of forward elevation. AP (C) seal fractures managed nonsurgically,
and axillary (D) radiographs obtained following reverse total shoulder Papasoulis et al31 showed an overall
arthroplasty. Forward elevation had improved to 160° by 4 months
nonunion rate of 5.5%. Studies pub-
postoperatively. (Copyright Center for Shoulder, Elbow and Sports Medicine
at Columbia University, New York, NY.) lished in the past decade that in-
cluded more than 50 patients dem-
underwent total shoulder arthro- merus nonunion. Fifty-four patients onstrate a 10% to 23% rate of
plasty. Although significant pain re- underwent hemiarthroplasty, and the humeral shaft nonunion following
lief and subjective patient satisfac- remaining 13 underwent total shoul- functional bracing.32-35 This is signifi-
tion were achieved, function was not der arthroplasty. Significant im- cantly greater than the rate of zero to
completely restored. Duquin et al27 provement was noted in active for- 2% reported in earlier studies by
reported the results of 67 patients ward elevation (46° to 104°) and Sarmiento and colleagues.30,36
treated with unconstrained shoulder external rotation (26° to 50°) (P < Many attempts have been made to
arthroplasty to manage proximal hu- 0.001). Radiographic tuberosity identify which fracture patterns pre-

542 Journal of the American Academy of Orthopaedic Surgeons

 Edwin R. Cadet, MD, et al

dispose to delayed union or non- upper extremity for activities of daily Timing of Surgery
union so that the patient can be living and for work. Higher-energy Although most authors define de-
counseled regarding the chance of mechanisms of injury as well as pa- layed union as no radiographic signs
achieving bony union with nonsurgi- tient comorbidities—both medical of osseous union 4 months after in-
cal treatment. In the classic report on and psychosocial—increase the risk jury and nonunion 6 months after in-
diaphyseal nonunions by Healy of nonunion. On physical examina- jury, the exact length of time recom-
et al,5 transverse fractures were most tion, a thorough assessment of the mended before functional bracing is
susceptible to nonunion (11 of 26 functional brace for proper fit and discontinued in favor of surgery var-
[42%]), followed by short oblique associated skin irritation will im- ies widely in the literature.2 Toivanen
fractures (7 of 26 [27%]). In one prove patient compliance with use of et al35 recommended surgery if frac-
study, 9 of 67 fractures managed the orthosis. A careful neurovascular tures show no clinical or radio-
with functional bracing went on to examination is needed to rule out graphic signs of consolidation after
require surgery for nonunion; 6 of neurologic injury, particularly in the only 6 weeks of functional bracing.
those 9 fractures exhibited a trans- radial, median, and ulnar nerve dis- In a single-surgeon series by Rutgers
verse fracture pattern.33 In contrast, tributions. Gross motion at the frac- and Ring,34 of 52 diaphyseal frac-
Ring et al37 reported that, of the 32 ture site is an obvious sign of non- tures managed nonsurgically, 5 es-
consecutive humeral shaft fracture union; however, this finding may be tablished nonunions resulted; these
nonunions managed at their institu- difficult to elicit in the presence of fractures were managed surgically at
tion over a 10-year period that went significant arm edema, arm girth, an average of 28 weeks after injury
on to surgical management, most and pain. (range, 10 to 55 weeks). In a Swed-
had spiral or oblique fracture pat- On plain radiographs, nonunion ish series of 78 nonsurgically man-
terns (84.4%), with only 12.5% hav- may demonstrate lack of fracture aged diaphyseal fractures, 9 patients
ing transverse fracture patterns. In a consolidation, increased interfrag- failed to heal with bracing and un-
study of 78 fractures, Ekholm et al32 mentary diastasis or angulation, and derwent surgery at a mean of 8.7
observed a trend toward more fre- the presence of hypertrophic callus months (range, 1.8 to 15.5 months)
quent nonunions in Orthopaedic without clinical stability. When it is after injury.32 The systematic review
Trauma Association type A fractures difficult to evaluate fracture consoli- by Papasoulis et al31 of 15 English-
(simple, >90% cortical contact) com- dation on plain radiographs (eg, hy- language studies on outcomes of
pared with type B (wedge) and type pertrophic callus, comminuted angu- nonsurgically managed diaphyseal
C (complex) fractures (P = 0.08). lar deformity), CT may be used to fractures showed a mean time to
The location of the fracture may evaluate for bridging callus. Routine union of 10.7 weeks. These results
also play a role in the development laboratory studies should include a suggest that surgeons should broach
of nonunion. Several studies have complete blood count and a basic the topic of surgical treatment to pa-
noted that fractures in the proximal metabolic panel to evaluate for ac- tients with fractures that have not
diaphysis are at greater risk of non- tive medical conditions that may im- healed by 10 to 12 weeks after in-
union compared with middle- or pede fracture healing. In the setting jury. We recommend maintaining a
distal-third shaft fractures, presum- of previous open fracture or surgery, high index of suspicion for nonunion
ably due to greater deforming forces infection must always be excluded in humeral shaft fractures that have
generated by the deltoid and pectora- and appropriate laboratory workup demonstrated no interval healing on
lis insertions, greater risk of muscle (eg, complete blood count with dif- consecutive radiographs taken 6 to 8
and long of head of the biceps ten- ferential, erythrocyte sedimentation weeks apart.
don interposition within the fracture, rate, C-reactive protein level) ob-
and difficulty with immobilization of tained. For the patient whose nutri- Nonsurgical Management
the proximal humerus fracture frag- tional status is in question, total se- Nonsurgical management of estab-
ments with current functional brace rum protein and serum albumin tests lished diaphyseal nonunion is un-
designs.7,34,35,37 may be performed. More advanced likely to result in healing and should
markers of bone metabolism (eg, vi- be reserved for select patients with
Patient Evaluation tamin D level) and an endocrinology medical comorbidities that place
Patients with established nonunions consult may be needed to investigate them at high risk for surgical and
most often report an inability to en- biologic causes of poor osteosynthe- anesthesia-related complications and
gage in repetitive activities using the sis potential. for patients with asymptomatic non-

September 2013, Vol 21, No 9 543

Proximal Humerus and Humeral Shaft Nonunions

unions or pseudarthroses and low can be extended proximally and in- The use of bone graft from intramed-
functional demands. Bone stimula- corporated into the deltopectoral in- ullary reaming and/or autogenous
tion is another nonsurgical treatment terval. A posterior approach to the ICBG is strongly recommended to fa-
option. These devices transmit low- humerus can be used for fractures in- cilitate healing. In hypertrophic cases,
pulsed direct current, pulsed electro- volving the distal three fourths of the the existing callus can augment autog-
magnetic fields, capacitive couplings, humerus and in cases of suspected enous bone or RIA grafts.39 Success has
and/or ultrasound electromagnetic entrapment of the radial nerve on its been reported with demineralized bone
signals to augment fracture healing. course along the posterior humeral matrix (DBM) with or without bone
Bone stimulators are contraindicated shaft. morphogenetic protein (BMP).7,41-43
in the presence of a synovial pseudar- Fracture reduction must correct Hierholzer et al41 retrospectively com-
any angular deformity and achieve pared two consecutive cohorts of pa-
throsis, fracture gapping >5 mm, and
good joint alignment, with maximal tients with aseptic delayed union or
poor vascular supply to the fracture
cortical contact to enhance mechani- nonunion of the humeral shaft that was
site.38
cal stability and osteosynthesis. The managed with rigid plate fixation with
The role of bone stimulators in fa-
fracture ends should be débrided and either autogenous ICBG or DBM and
cilitating osseous union of humeral
lightly decorticated to stimulate bone found no differences in union rates or
shaft nonunion or fractures is not
healing. Depending on the type of functional outcome and found only a
well defined in the literature. Ultra- fracture exposure, rigid fixation is slight difference in union rate and no
sound devices are approved for use obtained with a laterally, anteriorly, difference in overall functional out-
in acute fractures, but all other bone or posteriorly placed 4.5-mm com- come between the two groups. How-
stimulators have been approved for pression plate with at least six bicor- ever, 44% of patients in the ICBG
use by the FDA only in established tical fixation points proximal and group had donor site morbidity; as a
nonunions, that is, fracture non- distal to the fracture site. Eight fixa- result, DBM was adopted as the stan-
union that persists ≥9 months fol- tion points are preferred. Special at- dard bone graft option for humeral
lowing injury, or a fracture that dem- tention should be paid to plate shaft nonunions at that institution.
onstrates no visibly progressive length. To ensure balanced fixation Marti et al39 observed 100% union
radiographic signs of healing ≥3 across the construct, we recommend after 1 year in 51 patients with dia-
months following injury.38 that the plate be long enough to span physeal nonunion who were treated
the nonunion site a distance of at with ORIF with compression plat-
Surgical Management least two to three times the cortical ing; ICBG was added in atrophic
ORIF With Compression Plating diameter, both proximal and distal to cases. Livani et al40 also reported
and Bone Graft the fracture site. If a 3.5-mm plate is 100% union using a similar treat-
ORIF using a broad, 4.5-mm com- used for a patient with a particularly ment algorithm. Thus, we recom-
pression plate along with autogenous narrow humeral diaphysis, a mini- mend the use of autogenous ICBG in
bone graft is the standard of care for mum of eight points of proximal and patients who can withstand the pos-
humeral diaphyseal nonunion in the distal cortical fixation and a long sibility of donor site morbidity at the
absence of prior surgery. An exten- plate should be used because rigid harvest site.
sile approach was advocated by fixation is critical for successful frac-
Healy et al5 and in more recent re- ture healing (Figure 3). Dual Plating
ports.39,40 With an anterolateral ap- Depending on the fracture configu- Two studies have supported the use
proach, the radial nerve should be ration, the plate or lag screw inser- of dual orthogonal plating, which in-
identified between the brachialis and tion can be prestressed to maximize volves the use of an additional plate
brachioradialis muscles and pro- interfragmentary compression. In a in the presence of micromotion at
tected along its course well beyond study by Ring et al,37 all 32 consecu- the fracture site following prelimi-
the zone of nonunion. Neurolysis tive diaphyseal nonunions in adults nary plate fixation.7,44 The additional
can be performed if the nerve is en- who were treated with functional plate may be particularly useful in
cased in scar tissue. Depending on bracing for ≥4 months achieved cases of proximal shaft nonunion in
the location of the fracture site, a di- union with ORIF. However, three which fixation is limited by the short
rect anterior approach is also a utili- older patients with osteoporosis re- proximal fragment, for distal shaft
tarian approach to the humerus. In quired a second procedure to address fractures that present with a short
cases of proximal extension of the either failed fixation or persistent distal segment, in the presence of
fracture line, the anterior approach nonunion. poor metaphyseal bone quality, and

544 Journal of the American Academy of Orthopaedic Surgeons

 Edwin R. Cadet, MD, et al

Figure 3 in cases at risk of glenohumeral joint

penetration. Rubel et al44 noted no
difference in healing rates and func-
tional outcome in their retrospective
comparison of two cohorts of pa-
tients with nonunion that was man-
aged with either single- or dual-
compression plating. The overall
healing rate was 92% at an average
of 4.8 months. Prasarn et al7 treated
19 elderly patients (mean age, 70
years) who presented with proximal
shaft nonunions. Dual plating was
used in 11 patients to achieve ade-
quate fixation in the small metaphys-
eal proximal fragment, and all
achieved union at an average of 15.2
weeks. No revision surgery was re-
quired.

Cortical Strut Allograft

and Autograft
Augmentation of compression plat-
ing and bone graft with a cortical
strut allograft can be considered in
patients with severe osteopenia due
to a combination of advanced age,
disuse of the limb, or previous surgi-
cal treatment.45,46 The cortical al-
lograft is placed in an intramedullary
location or against the medial hu-
meral cortex, opposite the lateral
compression plate, and the screws
are inserted from lateral to medial to
sandwich the host humeral bone
with the strut, thereby providing ad-
ditional cortical fixation. Two recent
series on atrophic humeral shaft non-
unions demonstrated high rates of
fracture union (95% to 100%) after
fixation with dynamic compression
plating augmented by intramedullary
A, AP radiograph demonstrating an acute comminuted midshaft left humerus
fracture in a 19-year-old right hand–dominant woman who was initially fibular struts.45,46
treated with functional bracing. B, AP radiograph obtained 12 weeks after
injury demonstrating no radiographic signs of fracture healing. The patient Biologic Augmentation
continued to have pain and gross motion at the fracture site, so the decision Biologic augmentation of humeral
was made to proceed with surgical treatment. C, AP radiograph obtained
following open reduction and internal fixation of the atrophic nonunion. A nonunions with BMP-2 and BMP-7
conventional 4.5-mm plate was used to bridge the fracture comminution, and has been performed. Although litera-
the fracture site was supplemented with local bone intramedullary autograft ture exists to support the use of
and demineralized bone matrix. D, AP radiograph demonstrating bony union BMPs in tibial fracture and non-
6 months postoperatively. (Courtesy of Melvin P. Rosenwasser, MD, Training
Trauma Center, Columbia University, New York, NY.) union, we know of no study that has
reported on the use of BMP in hu-

September 2013, Vol 21, No 9 545

Proximal Humerus and Humeral Shaft Nonunions

meral nonunion.47,48 A recent Co- 2. Schemitsch EH, Bhandar M, Talbot M: 15. Badman B, Mighell M, Drake G:
Fractures of the humeral shaft, in Proximal humeral nonunions: Surgical
chrane review concluded that there is Browner BD, Levine AM, Jupiter JB, technique with fibular strut allograft and
limited evidence to support the use Trafton PG, Krettek C, eds: Skeletal fixed-angle locked plating. Techniques in
of BMP for fracture healing in Trauma: Basic Science, Management, Shoulder and Elbow Surgery 2006;7(2):
and Reconstruction, ed 4. Philadelphia, 95-101.
adults; further, it indicated that all PA, Saunders Elsevier, 2009, pp 1593-
16. Gartsman GM, Taverna E: Arthroscopic
randomized controlled trials to date 1622. treatment of rotator cuff tear and greater
have a high risk of bias due to indus- 3. Green A, Norris TR: Proximal humeral tuberosity fracture nonunion.
try involvement.49 The role of BMPs fractures and glenohumeral dislocation, Arthroscopy 1996;12(2):242-244.

in humeral nonunions remains unde- in Browner BD, Levine AM, Jupiter JB, 17. Newman JT, Stahel PF, Smith WR,
Trafton PG, Krettek C, eds: Skeletal Resende GV, Hak DJ, Morgan SJ: A new
fined. Thus, use of BMP should be Trauma: Basic Science, Management, minimally invasive technique for large
guarded given the issues of addi- and Reconstruction, ed 4. Philadelphia, volume bone graft harvest for treatment
tional cost without substantiated ef- PA, Saunders Elsevier, 2009, pp 1623- of fracture nonunions. Orthopedics
1754. 2008;31(3):257-261.
ficacy compared with ICBG or local
autograft. 4. Court-Brown CM, McQueen MM: 18. Kanakaris NK, Morell D, Gudipati S,
Nonunions of the proximal humerus: Britten S, Giannoudis PV: Reaming
Their prevalence and functional Irrigator Aspirator system: Early
outcome. J Trauma 2008;64(6):1517- experience of its multipurpose use.
Summary 1521. Injury 2011;42(suppl 4):S28-S34.

5. Healy WL, White GM, Mick CA, 19. Herscovici D Jr, Scaduto JM: Use of the
Historically, proximal humerus and Brooker AF Jr, Weiland AJ: Nonunion of reamer-irrigator-aspirator technique to
the humeral shaft. Clin Orthop Relat obtain autograft for ankle and hindfoot
humeral shaft nonunions have been arthrodesis. J Bone Joint Surg Br 2012;
Res 1987;(219):206-213.
regarded to be relatively rare inju- 94(1):75-79.
ries; however, more recent literature 6. Badman BL, Mighell M, Kalandiak SP,
Prasarn M: Proximal humeral nonunions 20. Cox G, Jones E, McGonagle D,
has suggested that such nonunions treated with fixed-angle locked plating Giannoudis PV: Reamer-irrigator-
and an intramedullary strut allograft. aspirator indications and clinical results:
may be more prevalent than previ- A systematic review. Int Orthop 2011;
J Orthop Trauma 2009;23(3):173-179.
ously expected. Primary, nonsurgical 35(7):951-956.
management for aligned fracture pat- 7. Prasarn ML, Achor T, Paul O, Lorich
DG, Helfet DL: Management of 21. Cox G, McGonagle D, Boxall SA,
terns with proper immobilization re- nonunions of the proximal humeral Buckley CT, Jones E, Giannoudis PV:
The use of the reamer-irrigator-aspirator
mains the first line of treatment for diaphysis. Injury 2010;41(12):1244-
to harvest mesenchymal stem cells.
1248.
fractures of the proximal humerus J Bone Joint Surg Br 2011;93(4):517-
and humeral shaft. The precise time 8. Cole PA: Endosteal allograft plating for 524.
the treatment of recalcitrant nonunions.
frame for diagnosing nonunion is Techniques in Orthopaedics 2004;18(4): 22. Ring D, McKee MD, Perey BH, Jupiter
JB: The use of a blade plate and
controversial. However, the standard 344-355.
autogenous cancellous bone graft in the
of care is achievement of a mechani- 9. Court-Brown CM, Caesar B: treatment of ununited fractures of the
cally stable and biologically friendly Epidemiology of adult fractures: A proximal humerus. J Shoulder Elbow
review. Injury 2006;37(8):691-697. Surg 2001;10(6):501-507.
construct with compression plating
10. Hanson B, Neidenbach P, de Boer P, 23. Allende C, Allende BT: The use of a new
with the use of regional or autoge- Stengel D: Functional outcomes after locking 90 degree blade plate in the
nous bone grafting. Tuberosity posi- nonoperative management of fractures of treatment of atrophic proximal humerus
tion and function are critical in the proximal humerus. J Shoulder Elbow nonunions. Int Orthop 2009;33(6):1649-
Surg 2009;18(4):612-621. 1654.
deciding between osteosynthesis, un-
11. Galatz LM, Iannotti JP: Management of 24. Tauber M, Brugger A, Povacz P, Resch
constrained arthroplasty, and reverse
surgical neck nonunions. Orthop Clin H: Reconstructive surgical treatment
total shoulder arthroplasty for the North Am 2000;31(1):51-61. without bone grafting in nonunions of
management of proximal humerus humeral surgical neck fractures.
12. Antuña SA, Sperling JW, Sánchez-Sotelo J Orthop Trauma 2011;25(7):392-398.
nonunions. J, Cofield RH: Shoulder arthroplasty for
proximal humeral nonunions. J Shoulder 25. Boileau P, Chuinard C, Le Huec JC,
Elbow Surg 2002;11(2):114-121. Walch G, Trojani C: Proximal humerus
fracture sequelae: Impact of a new
References 13. Healy WL, Jupiter JB, Kristiansen TK, radiographic classification on
White RR: Nonunion of the proximal arthroplasty. Clin Orthop Relat Res
humerus: A review of 25 cases. J Orthop 2006;442:121-130.
References printed in bold type are Trauma 1990;4(4):424-431.
those published within the past 5 26. Nayak NK, Schickendantz MS, Regan
14. Yamane S, Suenaga N, Oizumi N, WD, Hawkins RJ: Operative treatment
years. Minami A: Interlocking intramedullary of nonunion of surgical neck fractures of
nailing for nonunion of the proximal the humerus. Clin Orthop Relat Res
1. Volgas DA, Stannard JP, Alonso JE: humerus with the Straight Nail System. 1995;(313):200-205.
Nonunions of the humerus. Clin Orthop J Shoulder Elbow Surg 2008;17(5):755-
Relat Res 2004;(419):46-50. 759. 27. Duquin TR, Jacobson JA, Sanchez-

546 Journal of the American Academy of Orthopaedic Surgeons

 Edwin R. Cadet, MD, et al

Sotelo J, Sperling JW, Cofield RH: treatment of closed humeral shaft nonunions with allograft, demineralized
Unconstrained shoulder arthroplasty for fractures. Int Orthop 2005;29(1):10-13. bone matrix, and plate fixation.
treatment of proximal humeral Osteosynthesis and Trauma Care 2005;
nonunions. J Bone Joint Surg Am 2012; 36. Sarmiento A, Zagorski JB, Zych GA, 13(2):105-112.
94(17):1610-1617. Latta LL, Capps CA: Functional bracing
for the treatment of fractures of the 44. Rubel IF, Kloen P, Campbell D, et al:
28. Cadet ER, Ahmad CS: Hemiarthroplasty humeral diaphysis. J Bone Joint Surg Am Open reduction and internal fixation of
for three- and four-part proximal 2000;82(4):478-486. humeral nonunions: A biomechanical
humerus fractures. J Am Acad Orthop and clinical study. J Bone Joint Surg Am
Surg 2012;20(1):17-27. 37. Ring D, Chin K, Taghinia AH, Jupiter 2002;84(8):1315-1322.
JB: Nonunion after functional brace
29. Martinez AA, Bejarano C, Carbonel I,
treatment of diaphyseal humerus 45. Vidyadhara S, Vamsi K, Rao SK,
Iglesias D, Gil-Albarova J, Herrera A:
fractures. J Trauma 2007;62(5):1157- Gnanadoss JJ, Pandian S: Use of
The treatment of proximal humerus
nonunions in older patients with the 1158. intramedullary fibular strut graft: A
reverse shoulder arthroplasty. Injury novel adjunct to plating in the treatment
38. Nelson FR, Brighton CT, Ryaby J, et al: of osteoporotic humeral shaft nonunion.
2012;43 suppl 2:S3-S6. Use of physical forces in bone healing. Int Orthop 2009;33(4):1009-1014.
30. Sarmiento A, Kinman PB, Galvin EG, J Am Acad Orthop Surg 2003;11(5):
Schmitt RH, Phillips JG: Functional 344-354. 46. Willis MP, Brooks JP, Badman BL,
bracing of fractures of the shaft of the Gaines RJ, Mighell MA, Sanders RW:
39. Marti RK, Verheyen CC, Besselaar PP: Treatment of atrophic diaphyseal
humerus. J Bone Joint Surg Am 1977;
59(5):596-601. Humeral shaft nonunion: Evaluation of humeral nonunions with compressive
uniform surgical repair in fifty-one locked plating and augmented with an
31. Papasoulis E, Drosos GI, Ververidis AN, patients. J Orthop Trauma 2002;16(2):
intramedullary strut allograft. J Orthop
Verettas DA: Functional bracing of 108-115.
Trauma 2013;27(2):77-81.
humeral shaft fractures: A review of
clinical studies. Injury 2010;41(7):e21- 40. Livani B, Belangero W, Medina G,
47. Friedlaender GE, Perry CR, Cole JD,
e27. Pimenta C, Zogaib R, Mongon M:
Anterior plating as a surgical alternative et al: Osteogenic protein-1 (bone
32. Ekholm R, Tidermark J, Törnkvist H, in the treatment of humeral shaft non- morphogenetic protein-7) in the
Adami J, Ponzer S: Outcome after closed union. Int Orthop 2010;34(7):1025- treatment of tibial nonunions. J Bone
functional treatment of humeral shaft 1031. Joint Surg Am 2001;83(suppl 1 pt 2):
fractures. J Orthop Trauma 2006;20(9): S151-S158.
591-596. 41. Hierholzer C, Sama D, Toro JB, Peterson
M, Helfet DL: Plate fixation of ununited 48. Govender S, Csimma C, Genant HK,
33. Koch PP, Gross DF, Gerber C: The humeral shaft fractures: Effect of type of et al: Recombinant human bone
results of functional (Sarmiento) bracing bone graft on healing. J Bone Joint Surg morphogenetic protein-2 for treatment
of humeral shaft fractures. J Shoulder Am 2006;88(7):1442-1447. of open tibial fractures: A prospective,
Elbow Surg 2002;11(2):143-150. controlled, randomized study of four
42. Ring D, Kloen P, Kadzielski J, Helfet D, hundred and fifty patients. J Bone Joint
34. Rutgers M, Ring D: Treatment of Jupiter JB: Locking compression plates Surg Am 2002;84(12):2123-2134.
diaphyseal fractures of the humerus for osteoporotic nonunions of the
using a functional brace. J Orthop diaphyseal humerus. Clin Orthop Relat 49. Garrison KR, Shemilt I, Donell S, et al:
Trauma 2006;20(9):597-601. Res 2004;(425):50-54. Bone morphogenetic protein (BMP) for
fracture healing in adults. Cochrane
35. Toivanen JA, Nieminen J, Laine HJ, 43. Taylor NL, Crow SA, Heyworth BE, Database Syst Rev 2010;(6):CD006950.
Honkonen SE, Järvinen MJ: Functional Rosenwasser MP: Treatment of humeral

September 2013, Vol 21, No 9 547

View publication stats