Articol Exemplu

new england
The
journal of medicine
established in 1812 December 22, 2022 vol. 387 no. 25
Surgery or Endovascular Therapy for Chronic Limb-

Threatening Ischemia
A. Farber, M.T. Menard, M.S. Conte, J.A. Kaufman, R.J. Powell, N.K. Choudhry, T.H. Hamza, S.F. Assmann,*
M.A. Creager, M.J. Cziraky, M.D. Dake, M.R. Jaff, D. Reid, F.S. Siami, G. Sopko, C.J. White, M. van Over,
M.B. Strong, M.F. Villarreal, M. McKean, E. Azene, A. Azarbal, A. Barleben, D.K. Chew, L.C. Clavijo, Y. Douville,
L. Findeiss, N. Garg, W. Gasper, K.A. Giles, P.P. Goodney, B.M. Hawkins, C.R. Herman, J.A. Kalish,
M.C. Koopmann, I.A. Laskowski, C. Mena‑Hurtado, R. Motaganahalli, V.L. Rowe, A. Schanzer, P.A. Schneider,
J.J. Siracuse, M. Venermo, and K. Rosenfield, for the BEST-CLI Investigators†
a bs t r ac t
BACKGROUND
Patients with chronic limb-threatening ischemia (CLTI) require revascularization The authors’ full names, academic de-
to improve limb perfusion and thereby limit the risk of amputation. It is uncertain grees, and affiliations are listed in the
Appendix. Dr. Farber can be contacted at
whether an initial strategy of endovascular therapy or surgical revascularization alik.farber@bmc.org or at the Division of
for CLTI is superior for improving limb outcomes. Vascular and Endovascular Surgery, Bos-
ton Medical Center, Boston University
METHODS School of Medicine, Department of Sur-
gery, 85 E. Concord St., 3rd Fl., Rm. 3000,
In this international, randomized trial, we enrolled 1830 patients with CLTI and Boston, MA 02118.
infrainguinal peripheral artery disease in two parallel-cohort trials. Patients who
*Deceased.
had a single segment of great saphenous vein that could be used for surgery were
assigned to cohort 1. Patients who needed an alternative bypass conduit were as- †The BEST-CLI Investigators are listed in
the Supplementary Appendix, available
signed to cohort 2. The primary outcome was a composite of a major adverse limb at NEJM.org.
event — which was defined as amputation above the ankle or a major limb reinter-
Drs. Farber and Menard contributed
vention (a new bypass graft or graft revision, thrombectomy, or thrombolysis) — equally to this article.
or death from any cause.
This article was published on November
RESULTS 7, 2022, at NEJM.org.
In cohort 1, after a median follow-up of 2.7 years, a primary-outcome event oc- N Engl J Med 2022;387:2305-16.
curred in 302 of 709 patients (42.6%) in the surgical group and in 408 of 711 DOI: 10.1056/NEJMoa2207899
Copyright © 2022 Massachusetts Medical Society.
patients (57.4%) in the endovascular group (hazard ratio, 0.68; 95% confidence
interval [CI], 0.59 to 0.79; P<0.001). In cohort 2, a primary-outcome event occurred CME
in 83 of 194 patients (42.8%) in the surgical group and in 95 of 199 patients at NEJM.org
(47.7%) in the endovascular group (hazard ratio, 0.79; 95% CI, 0.58 to 1.06;
P = 0.12) after a median follow-up of 1.6 years. The incidence of adverse events was
similar in the two groups in the two cohorts.
CONCLUSIONS
Among patients with CLTI who had an adequate great saphenous vein for surgical
revascularization (cohort 1), the incidence of a major adverse limb event or death
was significantly lower in the surgical group than in the endovascular group.
Among the patients who lacked an adequate saphenous vein conduit (cohort 2), the
outcomes in the two groups were similar. (Funded by the National Heart, Lung,
and Blood Institute; BEST-CLI ClinicalTrials.gov number, NCT02060630.)
n engl j med 387;25 nejm.org December 22, 2022 2305

The New England Journal of Medicine
Downloaded from nejm.org on November 12, 2023. For personal use only. No other uses without permission.
Copyright © 2022 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e
C
hronic limb-threatening ische or the need for an alternative bypass conduit
mia (CLTI), the most severe manifesta- (cohort 2). The trial protocol was approved by
tion of peripheral artery disease, is de- the ethics committee or the national equivalent
fined by ischemic foot pain at rest, ischemic at each participating site. All the patients pro-
ulcerations, or gangrene.1 More than 200 million vided written informed consent.
A Quick Take
is available at people have peripheral artery disease worldwide; Enrollment began in August 2014 and contin-
NEJM.org CLTI affects up to 11% of this population.1,2 ued through October 2019. The patients in co-
Aside from the severe health outcomes associ- hort 1 were followed through October 2021, and
ated with CLTI, the economic effect of the condi- those in cohort 2 were followed through Decem-
tion is substantial, with an estimated annual ber 2019.
cost of approximately $12 billion in the United
States alone.3 Patient Population
Treatment for CLTI includes guideline-directed Eligible patients were at least 18 years old and
medical therapy to reduce cardiovascular risk, had received a diagnosis of CLTI, which was
revascularization to improve limb perfusion, and defined as arterial insufficiency of the lower
local care to control infection and improve wound limb with ischemic foot pain at rest, a nonheal-
healing.4 Without timely revascularization, the ing ischemic ulcer, or gangrene, as corroborated
incidence of limb amputation is approximately by hemodynamic criteria. Patients were excluded
25% at 1 year after diagnosis.5,6 Surgical bypass from the trial if they had excessive risk associ-
and endovascular therapy are the principal re- ated with open vascular surgery according to the
vascularization strategies used to treat CLTI.4 criteria of the American Heart Association and
The choice of surgery or endovascular therapy as the American College of Cardiology or according
the initial treatment varies greatly among pro- to the medical judgment of the investigator. De-
viders and is based on the patient’s arterial dis- tails regarding the representativeness of the pa-
ease pattern, surgical risk, availability of an autog- tient sample are provided in Table S1 in the
enous conduit for vein bypass, and patient Supplementary Appendix.12
preference, along with such physician factors as
training, skill set, and treatment bias.7-9 The ex- Randomization
tent to which this variability affects clinical Patients were enrolled into one of two parallel-
outcomes in patients with CLTI is unknown.7,9,10 trial cohorts according to prerandomization du-
We performed the Best Endovascular versus Best plex ultrasonography of the right and left great
Surgical Therapy in Patients with CLTI (BEST-CLI) saphenous veins. Within each cohort, eligible
trial to determine whether endovascular revas- patients were stratified according to clinical
cularization was superior to surgical revascular- criteria (ischemic rest pain or tissue loss) and
ization in patients with CLTI caused by infrain- anatomical criteria (presence or absence of con-
guinal peripheral artery disease who were judged siderable infrapopliteal arterial occlusive disease)
to be suitable candidates for both approaches. with the use of permuted randomized blocks.
The patients were then randomly assigned in a
1:1 ratio to receive surgical or endovascular treat-
Me thods
ment. All the patients were expected to receive
Trial Design their assigned treatment within 30 days after
BEST-CLI was an international, prospective, ran- randomization. An investigator with expertise in
domized, open-label, multicenter, superiority surgical bypass procedures had to agree with
trial, as described previously11 and in the trial another investigator with expertise in endovas-
protocol, available with the full text of this article cular revascularization procedures that clinical
at NEJM.org. Patients were enrolled at 150 sites equipoise existed in the randomization of each
in the United States, Canada, Finland, Italy, and patient.11,13
New Zealand (as detailed in the Supplementary In the surgical group, surgeons were allowed
Appendix, also available at NEJM.org). The trial to choose any bypass technique that was cur-
consisted of two parallel studies that were based rently being used in clinical practice. In the en-
on a preprocedural assessment of the availability dovascular group, interventionalists were allowed
of autogenous conduit for vein bypass: either a to choose any available endovascular technique.
single segment of great saphenous vein (cohort 1) Follow-up data were collected at 30 days after
2306 n engl j med 387;25 nejm.org December 22, 2022

Surgery or Endovascular Ther apy for Limb Ischemia
the procedure or 30 days after randomization if and a type I error rate of 0.05; the sample sizes
the index procedure had not been performed; were determined to allow for crossover between
follow-up was performed at 3 months, 6 months, groups, loss to follow-up, and the performance
and every 6 months thereafter up to 84 months of two interim analyses at prespecified intervals.
after randomization. Telephone visits in lieu of Trial enrollment was stopped early after 1830
clinic visits were planned at 30 months and patients had been enrolled owing to a lack of
every 12 months thereafter and at the end of the continued funding. Supplemental funding was
trial. raised to allow for 24 months of follow-up for all
the patients in cohort 1. Details regarding revi-
Outcomes sions to the statistical analysis plan are provided
The primary outcome was a composite of major in the trial protocol.
adverse limb events or death from any cause. A The primary and secondary outcome analyses
major adverse limb event was defined as above- were performed according to the intention-to-
ankle amputation of the index limb or a major treat principle. Analyses were carried out sepa-
index-limb reintervention (new bypass, interposi- rately in each cohort. Time-to-event outcomes
tion graft revision, thrombectomy, or thromboly- were described with the use of Kaplan–Meier
sis).14 The need for and timing of the reinterven- plots, and the two treatment groups were com-
tion was determined by the trial site investigator pared with the use of log-rank test statistics. We
on the basis of clinical assessment. All first used a prespecified covariate-adjusted Cox model
major reinterventions were adjudicated by an that was stratified according to randomization
independent, multidisciplinary clinical-events categories to calculate hazard ratios and their
committee. A modification of the criteria of the 95% confidence intervals. Missing baseline co-
Peripheral Academic Research Consortium was variates were imputed with the use of multiple
used to define technical success (see the Meth- imputation. We used Cox models that had been
ods section in the Supplementary Appendix).13 adjusted for the imputed covariates to calculate
Key secondary efficacy and safety outcomes were the results of secondary efficacy and safety analy-
the occurrence of a major adverse limb event at ses. In the primary analysis, a P value of less
any time or postoperative death within 30 days; than 0.045 after correction for two interim analy-
minor reinterventions; a major adverse cardio- ses was considered to indicate statistical signifi-
vascular event, which was defined as a compos- cance. The widths of the confidence intervals
ite of myocardial infarction, stroke, or death have not been adjusted for multiplicity, so confi-
from any cause; and serious adverse events (Ta- dence intervals should not be used for hypothe-
ble S2). Stroke and myocardial infarction were sis testing. All the analyses were performed with
adjudicated by the clinical-events committee. the use of SAS Enterprise Guide software, version
8.3 (SAS Institute), and R software, version 4.02.
Statistical Analysis Additional details regarding the statistical analy-
We originally determined that the enrollment of sis are provided in the Supplementary Appendix.
2100 patients (1620 in cohort 1 and 480 in co-
hort 2) would provide 85% power to detect a R e sult s
relative difference of 25% in the primary out-
come favoring the surgical group (i.e., an event Cohort 1
rate of 53.0% in the surgical group and 61.1% in Patients
the endovascular group) in cohort 1 and 80% From August 2014 through October 2019, a total
power to detect a relative difference of 30% in of 1434 patients with a single segment of great
the primary outcome favoring the endovascular saphenous vein underwent randomization in
group (i.e., an event rate of 53.0% in the surgical cohort 1 (718 to receive surgical treatment and
group and 45.3% in the endovascular group) in 716 to receive endovascular therapy) and were
cohort 2. In the two cohorts, the null hypothesis followed for up to 7 years, with a median follow-
was that there would be no difference in the up of 2.7 years (interquartile range, 1.6 to 4.0) in
time from randomization to a primary-outcome the surgical group and 2.7 years (interquartile
event between the surgical group and the endo- range, 1.6 to 4.1) in the endovascular group
vascular group. In the two cohorts, the calcula- (Fig. 1). Excluded from the primary analysis were
tions were to be based on 2.95 years of follow-up 14 patients (1.0%) — 9 in the surgical group and

2525 Patients were assessed for eligibility
501 Were ineligible

17 Were scheduled to undergo angiography
before consent date
465 Were scheduled to undergo angiography
on or after consent date
19 Had unknown angiogram status
166 Had unknown eligibility
11 Declined to participate or were found not to
be a candidate for both surgery and endo-
vascular therapy
1847 Underwent randomization
17 Were excluded after randomization

10 Had data integrity issues at 1 site
5 Were assigned to surgery
5 Were assigned to endovascular therapy
5 Had consent issues
3 Were assigned to surgery
2 Were assigned to endovascular therapy
1 Was assigned to surgery but had duplicate
randomization
1 Was erroneously assigned to endovascular
therapy after death had occurred
1434 Had single segment of great saphenous 396 Needed alternative conduit and
vein and were included in cohort 1 were included in cohort 2
718 Were assigned to undergo 716 Were assigned to undergo 197 Were assigned to undergo 199 Were assigned to undergo
surgery endovascular therapy surgery endovascular therapy
662 Underwent surgery first 705 Underwent endovascular 188 Underwent surgery first 191 Underwent endovascular
25 Underwent endovascular therapy first 2 Underwent endovascular therapy first
therapy first 3 Underwent surgery first therapy first 4 Underwent surgery first
31 Did not undergo any 8 Did not undergo any 7 Did not undergo any 4 Did not undergo any
procedure procedure procedure procedure
718 Were included in the 716 Were included in the 197 Were included in the 199 Were included in the
intention-to-treat analysis intention-to-treat analysis intention-to-treat analysis intention-to-treat analysis
662 Were included in the 705 Were included in the 188 Were included in the 191 Were included in the
per-protocol analysis per-protocol analysis per-protocol analysis per-protocol analysis
Disposition at end of the trial: Disposition at end of the trial: Disposition at end of the trial: Disposition at end of the trial:
209 Died 248 Died 49 Died 47 Died
94 Withdrew 60 Withdrew 24 Withdrew 10 Withdrew
68 Were lost to follow-up 64 Were lost to follow-up 12 Were lost to follow-up 14 Were lost to follow-up
37 Did not consent to follow- 39 Did not consent to follow- 2 Did not consent to follow- 3 Did not consent to follow-
up after 48 mo up after 48 mo up after 48 mo up after 48 mo
27 Were followed until early 28 Were followed until early 5 Were followed until early 4 Were followed until early
site closure site closure site closure site closure
283 Completed the trial 277 Completed the trial 105 Completed the trial 121 Completed the trial

Figure 1 (facing page). Randomization and Outcomes.

Outcomes
The primary outcome of major adverse limb
Shown are data for patients with a single segment of events or death from any cause occurred in 302
great saphenous vein (who were included in cohort 1) of 709 patients (42.6%) in the surgical group and
and those who needed an alternative bypass conduit in 408 of 711 patients (57.4%) in the endovascu-
(who were included in cohort 2). The patients in each
lar group (hazard ratio, 0.68; 95% confidence
cohort subsequently underwent separate randomiza-
tion to undergo either surgery or endovascular therapy. interval [CI], 0.59 to 0.79; P<0.001) (Table 2 and
In the description of the patients’ disposition at the end Table S7). This result was similar in the per-
of the trial, the patients who provided limited consent protocol and the as-treated analyses (Table S8).
until follow-up at 48 months completed the 48-month The time until a primary-outcome event is
visit but did not consent to receive additional follow-up,
shown in Figure 2A. Major reinterventions oc-
as outlined in version 5.0 of the protocol.
curred in 65 of 709 patients (9.2%) in the surgi-
cal group and in 167 of 711 patients (23.5%) in
5 in the endovascular group — because of miss- the endovascular group (hazard ratio, 0.35; 95%
ing baseline data regarding diabetes, smoking CI, 0.27 to 0.47) (Fig. 2B). Above-ankle amputa-
status, end-stage renal disease, or previous in- tion of the index limb occurred in 74 of 709
frainguinal revascularization; the secondary effi- patients (10.4%) in the surgical group and in 106
cacy and safety analyses were adjusted for im- of 711 patients (14.9%) in the endovascular
puted covariates and did not exclude patients group (hazard ratio, 0.73; 95% CI, 0.54 to 0.98)
(Table S3). The characteristics of the patients (Fig. 2C). The incidences of death from any
were well balanced between the groups, with the cause and perioperative death were similar in
exception of more Black patients in the surgical the two groups (Fig. 2D and Fig. S1A).
group than in the endovascular group (Table 1 Subgroup analysis of the primary outcome
and Tables S4 and S5). suggested a treatment effect across most pre-
specified groups that appeared to favor the sur-
Index Procedure gical group as compared with the endovascular
The median time until the index procedure was group, with the exceptions of patients who were
4 days (interquartile range, 1 to 11) in the surgi- older than 80 years of age, Black patients, and
cal group and 1 day (interquartile range, 0 to 7) those with previous limb revascularization on
in the endovascular group. Procedures that were the same side, grade 3 wounds, or renal dysfunc-
performed in the surgical group included 307 tion (Fig. S2). Patients in the surgical group had
femoral–popliteal, 276 femoral–tibial or pedal, a lower incidence rate of new or recurrent CLTI
and 115 popliteal–tibial or pedal bypass opera- events than those in the endovascular group
tions; 85% of the procedures were performed (incidence rate ratio, 0.82; 95% CI, 0.70 to 0.95).
with a single segment of great saphenous vein
(Table S6). Procedures in the endovascular group Adverse Events
included 487 that were performed on the super- Major adverse cardiovascular events occurred in
ficial femoral artery, 382 on the popliteal artery, 56 of 1434 patients (3.9%) from randomization
and 381 on the tibial or pedal arteries. The type through 30 days after the index procedure and
of endovascular procedure varied depending on in 578 of 1434 patients (40.3%) through the end
the arterial segment that was treated (Table S6). of follow-up. There were no material between-
Endovascular interventions were performed by group differences in the incidence of major ad-
vascular surgeons in 73% of cases, by interven- verse cardiovascular events overall or at 30 days
tional cardiologists in 15% of cases, and by in- or of myocardial infarction or stroke (Table 2,
terventional radiologists in 13% of cases. The Table S7, and Fig. S1). From randomization
technical success of the index procedure was through 30 days after the procedure, 427 serious
98% in the surgical group and 85% in the endo- adverse events occurred in the surgical group
vascular group. Of the 108 cases of early techni- and 379 in the endovascular group, including in
cal failure in the endovascular group, 66 were 12 of 687 patients (1.7%) with perioperative
treated with a bypass operation within 30 days. death in the surgical group and in 9 of 708 pa-

2310
Table 1. Characteristics of the Patients at Baseline.*
Characteristic Cohort 1 Cohort 2
Overall Surgery Endovascular Therapy Overall Surgery Endovascular Therapy

(N = 1434) (N = 718) (N = 716) (N = 396) (N = 197) (N = 199)
Demographic
Age — yr 66.9±9.9 66.9±9.8 67.0±10.0 68.6±9.2 68.4±8.8 68.8±9.6
Female sex — no./total no. (%) 408/1434 (28.5) 201/718 (28.0) 207/716 (28.9) 111/396 (28.0) 56/197 (28.4) 55/199 (27.6)
The
Race or ethnic group — no./

total no. (%)†
White 1028/1423 (72.2) 500/711 (70.3) 528/712 (74.2) 275/390 (70.5) 143/194 (73.7) 132/196 (67.3)
Black 275/1423 (19.3) 156/711 (21.9) 119/712 (16.7) 96/390 (24.6) 40/194 (20.6) 56/196 (28.6)
Asian 20/1423 (1.4) 13/711 (1.8) 7/712 (1.0) 2/390 (0.5) 2/194 (1.0) 0/196
Other 100/1423 (7.0) 42/711 (5.9) 58/712 (8.1) 17/390 (4.4) 9/194 (4.6) 8/196 (4.1)
n engl j med 387;25

Hispanic 187/1433 (13.0) 82/717 (11.4) 105/716 (14.7) 53/396 (13.4) 28/197 (14.2) 25/199 (12.6)
Medical history
nejm.org
Body-mass index‡ 28.2±6.0 28.2±6.3 28.3±5.8 26.9±5.7 26.8±5.1 27.0±6.2
Coexisting condition — no./
n e w e ng l a n d j o u r na l
total no. (%)

of
Hypertension 1238/1424 (86.9) 620/712 (87.1) 618/712 (86.8) 350/395 (88.6) 171/196 (87.2) 179/199 (89.9)
Hyperlipidemia 1041/1423 (73.2) 521/712 (73.2) 520/711 (73.1) 299/395 (75.7) 147/196 (75.0) 152/199 (76.4)
Diabetes 1023/1424 (71.8) 513/712 (72.1) 510/712 (71.6) 238/395 (60.3) 122/196 (62.2) 116/199 (58.3)
December 22, 2022

Current smoking 509/1424 (35.7) 264/712 (37.1) 245/712 (34.4) 140/395 (35.4) 69/196 (35.2) 71/199 (35.7)
m e dic i n e
Coronary artery disease 617/1424 (43.3) 301/712 (42.3) 316/712 (44.4) 204/395 (51.6) 97/196 (49.5) 107/199 (53.8)
Congestive heart failure 79/1422 (5.6) 38/711 (5.3) 41/711 (5.8) 27/395 (6.8) 12/196 (6.1) 15/199 (7.5)
Stroke 190/1424 (13.3) 91/712 (12.8) 99/712 (13.9) 62/395 (15.7) 38/196 (19.4) 24/199 (12.1)
Chronic obstructive pulmonary 208/1424 (14.6) 100/712 (14.0) 108/712 (15.2) 69/395 (17.5) 34/196 (17.3) 35/199 (17.6)
disease
End-stage kidney disease 151/1423 (10.6) 67/712 (9.4) 84/711 (11.8) 45/395 (11.4) 25/196 (12.8) 20/199 (10.1)
Characteristic Cohort 1 Cohort 2
Overall Surgery Endovascular Therapy Overall Surgery Endovascular Therapy

(N = 1434) (N = 718) (N = 716) (N = 396) (N = 197) (N = 199)
Medication
Statin — no./total no. (%) 1001/1424 (70.3) 503/713 (70.5) 498/711 (70.0) 307/394 (77.9) 153/195 (78.5) 154/199 (77.4)
Aspirin — no./total no. (%) 953/1424 (66.9) 476/713 (66.8) 477/711 (67.1) 280/394 (71.1) 139/195 (71.3) 141/199 (70.9)
Clopidogrel — no./total no. (%) 312/1424 (21.9) 137/713 (19.2) 175/711 (24.6) 97/394 (24.6) 55/195 (28.2) 42/199 (21.1)
Prasugrel — no./total no. (%) 5/1424 (0.4) 2/713 (0.3) 3/711 (0.4) 1/394 (0.3) 0/195 1/199 (0.5)
Ticagrelor — no./total no. (%) 10/1424 (0.7) 4/713 (0.6) 6/711 (0.8) 4/394 (1.0) 0/195 4/199 (2.0)
Direct-acting oral anticoagulant 55/1424 (3.9) 27/713 (3.8) 28/711 (3.9) 22/394 (5.6) 9/195 (4.6) 13/199 (6.5)
— no./total no. (%)
Warfarin — no./total no. (%) 93/1424 (6.5) 46/713 (6.5) 47/711 (6.6) 31/394 (7.9) 12/195 (6.2) 19/199 (9.5)
Previous intervention
Tobacco cessation — no./ 97/1424 (6.8) 49/712 (6.9) 48/712 (6.7) 26/395 (6.6) 11/196 (5.6) 15/199 (7.5)
total no. (%)
Infrainguinal revascularization of in- 77/1423 (5.4) 40/711 (5.6) 37/712 (5.2) 40/393 (10.2) 20/194 (10.3) 20/199 (10.1)
dex limb — no./total no. (%)
n engl j med 387;25

Limb status
Ankle–brachial index in index limb§ 0.58±0.32 0.58±0.31 0.59±0.34 0.54±0.30 0.53±0.27 0.54±0.32
Ankle pressure — mm Hg¶ 84.9±47.7 85.2±46.2 84.5±49.2 81.3±49.6 80.4±47.3 82.2±51.8
nejm.org
Toe pressure — mm Hg‖ 36.3±25.7 36.5±27.7 36.1±23.5 31.0±21.7 37.0±23.5 25.5±18.4
* Plus–minus values are means ±SD. Cohort 1 included patients who had a single segment of great saphenous vein, and cohort 2 included patients who needed an alternative bypass
conduit.
† Race and ethnic group were reported by the patients.
‡ The body-mass index is the weight in kilograms divided by the square of the height in meters. Data regarding body-mass index were missing for 67 patients (36 in the surgical group

December 22, 2022
and 31 in the endovascular group) in cohort 1 and for 19 patients (9 in the surgical group and 10 in the endovascular group) in cohort 2.
§ The ankle–brachial index is the systolic blood pressure in the ankle divided by the systolic blood pressure in the arm; an index between 0.9 and 1.2 is considered to be normal. The index
value was missing for 463 patients (224 in the surgical group and 239 in the endovascular group) in cohort 1 and for 116 patients (62 in the surgical group and 54 in the endovascular
group) in cohort 2.
¶ Data regarding ankle pressure were missing for 426 patients (205 in the surgical group and 221 in the endovascular group) in cohort 1 and for 108 patients (57 in the surgical group
and 51 in the endovascular group) in cohort 2.

‖ Data regarding toe pressure were missing for 826 patients (418 in the surgical group and 408 in the endovascular group) in cohort 1 and for 227 patients (117 in the surgical group and
110 in the endovascular group) in cohort 2.
2311
Table 2. Efficacy and Safety Outcomes in Cohort 1.*
Endovascular Hazard Ratio

Outcome Surgery Therapy (95% CI)† P Value
Efficacy
Primary outcome: major adverse limb event or death from 302/709 (42.6) 408/711 (57.4) 0.68 (0.59–0.79) <0.001
any cause — no./total no. (%)‡
Secondary outcomes — no./total no. (%)
Death from any cause 234/709 (33.0) 267/711 (37.6) 0.98 (0.82–1.17)
Above-ankle amputation of the index limb 74/709 (10.4) 106/711 (14.9) 0.73 (0.54–0.98)
Intervention in index limb
Major 65/709 (9.2) 167/711 (23.5) 0.35 (0.27–0.47)
Minor 205/718 (28.6) 237/716 (33.1) 0.85 (0.70–1.02)
Perioperative death§ 12/687 (1.7) 9/708 (1.3) 1.54 (0.64–3.68)
Major adverse limb event or perioperative death 139/687 (20.2) 246/708 (34.7) 0.53 (0.43–0.65)
Myocardial infarction 75/718 (10.4) 85/716 (11.9) 0.97 (0.71–1.33)
Stroke 39/718 (5.4) 44/716 (6.1) 0.93 (0.60–1.43)
Safety
Major adverse cardiovascular event — no. of patients with
≥1 event/total no. of patients (%)
Event ≤30 days after procedure¶ 33/718 (4.6) 23/716 (3.2) 1.46 (0.86–2.50) 0.16
Event during follow-up 269/718 (37.5) 309/716 (43.2) 0.94 (0.80–1.11) 0.48
Serious adverse event
Event occurred ≤30 days after index procedure — no. of 244/718 (34.0) 226/716 (31.6) 0.34
patients with ≥1 event/total no. of patients (%)‖
No. of events ≤30 days after index procedure 427 379 0.10
No. of patients with ≥1 event/total no. of patients (%) 590/718 (82.2) 614/716 (85.8) 0.07
No. of events during follow-up 3141 3468 <0.001
Technical success of index procedure — no./total no. (%)** 651/662 (98.3) 596/704 (84.7)
Length of hospital stay after index procedure††
No. of days 7.5±6.2 5.9±7.3
Median no. of days (IQR) 6 (4–9) 3 (1–8)
* Plus–minus values are means ±SD. In various categories, denominators differ because of missing baseline covariates in the regression
model or the restriction of the analysis to patients who underwent the assigned index procedure. IQR denotes interquartile range.
† The widths of the confidence intervals have not been adjusted for multiplicity, so the confidence intervals should not be used for hypoth-
esis testing.
‡ Data for the outcomes of death from any cause and above-ankle amputation of the index limb were collected until the end of the trial. Data
for a major or minor reintervention in the index limb and major adverse cardiovascular events were collected until the end of the follow-up
period.
§ Perioperative death was defined as death from any cause within 30 days after the index procedure.
¶ Included in this category were major adverse cardiovascular events that occurred after randomization through 30 days after the index pro-
cedure or within 30 days after randomization if the index procedure was not performed.
‖ Serious adverse events were evaluated from the date of randomization through 30 days after the index procedure or within 30 days after
randomization if the index procedure was not performed.
** Technical success of the index procedure was defined according to prespecified criteria.
†† The length of the hospital stay was the number of days from the date of the index procedure through discharge or 30 days after the pro-
cedure, whichever came first. Data regarding the length of hospital stay were missing for 33 patients in the surgical group and for 125
patients in the endovascular group.

A Major Adverse Limb Events or Death B Major Reintervention

1.0 1.0
P<0.001 by log-rank test
0.8 0.8
Endovascular
therapy
Probability
Probability
0.6 0.6
0.4 0.4 Endovascular therapy

Surgery
0.2 0.2
Surgery
0.0 0.0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Years since Randomization Years since Randomization
No. at Risk No. at Risk
Endovascular 716 404 304 175 102 46 14 0 Endovascular 716 444 331 192 111 48 14 0
therapy therapy
Surgery 718 463 349 204 117 52 12 0 Surgery 718 500 385 227 128 58 13 0
C Above-Ankle Amputation D Death

1.0 1.0
0.8 0.8
Endovascular
Probability
Probability
0.6 0.6 therapy
0.4 0.4
Endovascular therapy Surgery
0.2 0.2
Surgery
0.0 0.0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Years since Randomization Years since Randomization
No. at Risk No. at Risk
Endovascular 716 501 387 239 142 64 17 1 Endovascular 716 586 462 298 182 85 23 1
therapy therapy
Surgery 718 502 387 229 131 58 15 0 Surgery 718 577 457 282 168 80 20 0
Figure 2. Kaplan–Meier Curves of the Primary Outcome and Its Components in Cohort 1.
Shown is the primary outcome — a composite of major adverse limb events or death from any cause — among patients in the surgical
group and the endovascular group in cohort 1 (which included patients who had a single segment of great saphenous vein) (Panel A).
The components of the primary outcome were a major index-limb reintervention, including a new bypass graft or graft revision, throm-
bectomy, or thrombolysis (Panel B); above-ankle amputation of the index limb (Panel C); and death from any cause (Panel D). Shading
indicates the 95% confidence interval.
tients (1.3%) in the endovascular group. The group and 1.6 years (interquartile range, 0.7 to
median length of hospital stay was longer in the 3.1) in the endovascular group. Excluded from
surgical group than in the endovascular group the primary analysis were 3 patients (0.8%) —
(6 days vs. 3 days). all in the surgical group — because of missing
baseline data regarding diabetes, smoking status,
Cohort 2 end-stage renal disease, or previous infrainguinal
Patients revascularization. The secondary efficacy and
In cohort 2, a total of 396 patients without a safety outcome analyses were adjusted for im-
single segment of great saphenous vein under- puted covariates and did not exclude patients.
went randomization (197 to receive surgical The characteristics of the patients were well bal-
treatment and 199 to receive endovascular ther- anced between the two groups except that the
apy) and were followed for a median of 1.6 years baseline toe pressures were higher in the surgi-
(interquartile range, 0.7 to 2.8) in the surgical cal group (Table 1).

Index Procedure S10 and Fig. S5A through S5G). The median
The median time until the index procedure was length of hospital stay was longer in the surgical
4 days (interquartile range, 1 to 13) in the surgi- group.
cal group and 1 day (interquartile range, 0 to 7)
in the endovascular group. The technical success Discussion
was 100% in the surgical group and 80.6% in
the endovascular group. Of the 37 early cases of In recent years, the frequency of endovascular
technical failure in the endovascular group, 26 therapy as the initial revascularization strategy
patients underwent surgical bypass within 30 days. has increased. This trend notwithstanding, in
In the surgical group, 105 femoral–popliteal, our trial, we saw a compelling primary role for
86 femoral–tibial or pedal, and 18 popliteal– initial surgical revascularization in the treatment
tibial or pedal bypasses were performed. There of CLTI. In patients with a good-quality great
were 48 bypasses involving alternative autoge- saphenous vein for conduit (cohort 1), a surgery-
nous veins and 119 bypasses involving a pros- first strategy was associated with a 32% lower
thetic conduit. In 19% of cases, the surgeon un- risk of a composite of major adverse limb events
expectedly identified a single segment of great or death than was the endovascular strategy, a
saphenous vein that was suitable for bypass result that appeared to be driven by fewer major
surgery. Among the endovascular interventions, reinterventions and above-ankle amputations in
133 were performed on the superficial femoral the surgical group. However, in patients without
artery, 114 on the popliteal artery, and 86 on the a great saphenous vein for conduit (cohort 2),
tibial or pedal arteries (Table S6). overall efficacy and safety outcomes appeared to
be similar in the two treatment groups, findings
Outcomes that emphasize the importance of individualized
The primary outcome of major adverse limb patient-level decision making in patients without
events or death from any cause occurred in 83 an appropriate bypass conduit.
of 194 patients (42.8%) in the surgical group Patients in the two groups had similar inci-
and in 95 of 199 patients (47.7%) in the endo- dences of adverse cardiovascular events and
vascular group (hazard ratio, 0.79; 95% CI, death. In cohort 1, the 30-day mortality, inci-
0.58 to 1.06; P = 0.12) (Fig. S3A and Tables S9 dence of major adverse cardiovascular events,
and S10). The time until a major reintervention and long-term survival over a median 2.7 years
favored the surgical group (hazard ratio, 0.47; of follow-up were in line with the data in two
95% CI, 0.29 to 0.76) (Fig. S3B). There were no large registry studies.15,16 These findings indicate
material between-group differences in the time that in contemporary practice, revascularization
until above-ankle amputation or death from in appropriately selected patients with CLTI can
any cause (Fig. S3C and S3D and Tables S9 and be performed with low morbidity and high limb
S10). Similar results were obtained across sub- salvage. In agreement with the findings from
groups (Fig. S4) and in the per-protocol and one of these registry studies,15 our trial showed
as-treated analyses of the primary outcome that the majority of patients who were predicted
(Table S11). There was no material difference to have preferential benefit from surgical revas-
between the surgical group and the endovascu- cularization safely underwent surgery.
lar group regarding the incidence of new or In cohort 1, the between-group difference in
recurrent CLTI events (incidence rate ratio, reintervention was most pronounced during the
0.87; 95% CI, 0.64 to 1.17). first 6 months, and 99 of 233 first reintervention
events (42.5%) occurred within 30 days. This
Adverse Events early increase in major reintervention may have
Major adverse limb events occurred in 13 of 396 been related to a higher incidence of initial tech-
patients (3.3%) from the date of randomization nical failure in the endovascular group (15%)
through 30 days after the index procedure and than in the surgical group (<2%). Although it is
in 124 of 396 patients (31.3%) through the end likely that the majority of these major reinter-
of the trial. There were no material between- ventions were clinically driven, the choice and
group differences in the time until a major ad- timing of reintervention were based on the judg-
verse limb event overall or at 30 days or until ment of the treating physician. Causes of the
myocardial infarction or stroke (Tables S9 and technical failures, which may have influenced

the higher rate of major reinterventions in the underwent randomization were those in whom
endovascular group, are not known. the enrolling team believed there was equipoise
Overall, the findings from this large, interna- between endovascular intervention and bypass
tional trial suggest that preprocedural planning surgery. Although the majority of patients (66%)
of treatment in patients with CLTI should in- had substantial infrapopliteal-artery involvement,
clude a surgical risk assessment and a determi- an anticipated future review of angiographic
nation of saphenous-vein availability. Our find- data will elucidate the degree of anatomical
ings suggest that among the trial patients with complexity among these patients. Because inves-
an adequate saphenous vein who were suitable tigators used their preferred techniques, there
candidates for both surgical and endovascular was procedural heterogeneity within each trial
revascularization, bypass with a vein was a supe- group. The reliance on the judgment of individ-
rior initial strategy. However, many patients with ual operators in defining successful revascular-
CLTI who are appropriate candidates for limb- ization also could have influenced treatment
preserving interventions do not have adequate outcomes. The percentage of women in the trial
conduit, and others may still prefer an endovas- (28%) was lower than the targeted number. Be-
cular approach after fully informed, shared deci- cause of difficulties with enrollment, the planned
sion making. Additional analyses regarding ana- number of patients who were enrolled in the
tomical patterns of vascular disease, predictors trial was not met. Additional funds that were
of technical failure, effect on quality of life, cost, raised enabled the planned minimum of 24
and role of patient preference will further eluci- months of follow-up in cohort 1 but not in co-
date subgroups of patients who are most likely hort 2. Finally, a meta-analysis19 that had been
to benefit from each approach. published toward the end of the trial enrollment
In the randomized, controlled Bypass versus period aroused concern regarding a risk of death
Angioplasty in Severe Ischemia of the Leg (BASIL) associated with the use of paclitaxel-coated bal-
trial,17 which compared surgery with endovascu- loons and stents. These devices have been shown
lar therapy in 452 patients, investigators found no to reduce the need for reintervention in the super-
material between-group difference in the primary ficial femoral and proximal popliteal arteries.20
outcome of amputation-free survival. Analyses of This concern may have reduced the use of pacli-
secondary outcomes in our trial results add evi- taxel-coated balloons and stents in the trial.
dence to support complementary roles for surgi- In patients with CLTI who had an adequate
cal and endovascular procedures, as endorsed in single segment of great saphenous vein for con-
clinical practice guidelines.1,18,19 Factors such as duit and were considered to be suitable candidates
conduit availability for bypass, advanced age, and for both endovascular intervention and surgical
renal failure are key considerations in planning bypass, initial bypass surgery was associated with
revascularization procedures. Our data also high- a lower incidence of major adverse limb events or
light the importance of a team approach that le- death than initial endovascular intervention. In
verages experience with both strategies to most patients without a suitable great saphenous vein,
effectively treat patients with CLTI. results associated with initial endovascular inter-
Our study has several limitations. Trial results vention were not significantly different from those
may have been influenced by selection and op- associated with initial bypass surgery.
erator bias as a consequence of its pragmatic Supported by the National Heart, Lung, and Blood Institute.
design and implementation. Eligibility was de- Disclosure forms provided by the authors are available with
the full text of this article at NEJM.org.
termined locally and varied according to the site A data sharing statement provided by the authors is available
and the individual investigator; patients who with the full text of this article at NEJM.org.
Appendix
The authors’ full names and academic degrees are as follows: Alik Farber, M.D., Matthew T. Menard, M.D., Michael S. Conte, M.D.,
John A. Kaufman, M.D., Richard J. Powell, M.D., Niteesh K. Choudhry, M.D., Ph.D., Taye H. Hamza, Ph.D., Susan F. Assmann, Ph.D.,
Mark A. Creager, M.D., Mark J. Cziraky, Pharm.D., Michael D. Dake, M.D., Michael R. Jaff, D.O., Diane Reid, M.D., Flora S. Siami,
M.P.H., George Sopko, M.D., Christopher J. White, M.D., Max van Over, M.P.H., Michael B. Strong, M.A., Maria F. Villarreal, M.D.,
Michelle McKean, M.P.H., Ezana Azene, M.D., Ph.D., Amir Azarbal, M.D., Andrew Barleben, M.D., David K. Chew, M.D., Leonardo C.
Clavijo, M.D., Ph.D., Yvan Douville, M.D., Laura Findeiss, M.D., Nitin Garg, M.B., B.S., M.P.H., Warren Gasper, M.D., Kristina A. Giles,
M.D., Philip P. Goodney, M.D., Beau M. Hawkins, M.D., Christine R. Herman, M.D., Jeffrey A. Kalish, M.D., Matthew C. Koopmann,
M.D., Igor A. Laskowski, M.D., Ph.D., Carlos Mena‑Hurtado, M.D., Raghu Motaganahalli, M.D., Vincent L. Rowe, M.D., Andres Schan-
zer, M.D., Peter A. Schneider, M.D., Jeffrey J. Siracuse, M.D., Maarit Venermo, M.D., Ph.D., and Kenneth Rosenfield, M.D.

The authors’ affiliations are as follows: the Division of Vascular and Endovascular Surgery, Boston Medical Center, Boston Univer-
sity School of Medicine (A.F., M.F.V., J.A. Kalish, J.J.S.), the Division of Vascular and Endovascular Surgery (M.T.M., M.B.S.) and the
Department of Medicine (N.K.C.), Brigham and Women’s Hospital, Harvard Medical School (M.R.J.), and the Section of Vascular
Medicine and Intervention, Massachusetts General Hospital (K.R.), Boston, HealthCore, Watertown (T.H.H., S.F.A., M.J.C., M.O.,
M.M.), and the Division of Vascular Surgery, UMass Memorial Health, Worcester (A.S.) — all in Massachusetts; the Division of Vascu-
lar and Endovascular Surgery, University of California, San Francisco, San Francisco (M.S.C., W.G., P.A.S.), the Division of Vascular
Surgery, UC San Diego Health, La Jolla (A.B.), Coastal Cardiology, French Hospital Medical Center, San Luis Obispo (L.C.C.), and the
Division of Vascular Surgery and Endovascular Therapy, University of Southern California, Los Angeles (V.L.R.) — all in California; the
Department of Interventional Radiology (J.A. Kaufman) and the Division of Vascular Surgery, School of Medicine (A.A.), Oregon Health
and Science University, and the Operative Care Division, Portland VA Medical Center (M.C.K.) — all in Portland; the Heart and Vascular
Center, Dartmouth Hitchcock Medical Center, Geisel School of Medicine at Dartmouth, Lebanon, NH (R.J.P., M.A.C., P.P.G.); the De-
partment of Medical Imaging, University of Arizona Health Sciences, Tucson (M.D.D.), and the Division of Vascular Surgery, Peak Heart
and Vascular, Glendale (N.G.) — both in Arizona; the National Heart, Lung, and Blood Institute, Bethesda, MD (D.R., G.S.); the
Medical Device Innovation Consortium, National Evaluation System for Health Technology Coordinating Center, Arlington, VA (F.S.S.);
the Department of Cardiovascular Diseases, Ochsner Clinical School, University of Queensland, Brisbane, Australia (C.J.W.); the Depart-
ment of Interventional Radiology, Gundersen Health System, La Crosse, WI (E.A.); the Iowa Heart Center, Des Moines (D.K.C.); the
Division of Vascular Surgery, Centre Hospitalier Universitaire de Quebec–Université Laval, Quebec City (Y.D.), and Dalhousie Univer-
sity, Division of Cardiac and Vascular Surgery, Department of Surgery, Nova Scotia Health Association, Halifax (C.R.H.) — both in
Canada; the Department of Radiology and Imaging Sciences, Emory University School of Medicine, Atlanta (L.F.); the Division of Vas-
cular and Endovascular Surgery, Maine Medical Center, Portland (K.A.G.); the Cardiovascular Section, University of Oklahoma Health
Sciences Center, Oklahoma City (B.M.H.); Westchester Medical Center and the Department of Surgery, Section of Vascular Surgery, New
York Medical College — both in Valhalla, NY (I.A.L.); the Vascular Medicine Outcomes Program, Yale University, New Haven, CT (C.M.-
H.); the Division of Vascular Surgery, Indiana University School of Medicine, Indianapolis (R.M.); and the Department of Vascular
Surgery, University of Helsinki, and Helsinki University Hospital — both in Helsinki (M.V.).
References
1. Conte MS, Bradbury AW, Kolh P, et al. guinal chronic limb-threatening ischemia. chronic limb-threatening ischemia who
Global vascular guidelines on the manage- J Vasc Surg 2019;69:6S:126S-136S. undergo infrainguinal revascularization.
ment of chronic limb-threatening ische 8. Bradbury A, Wilmink T, Lee AJ, et al. J Vasc Surg 2019;69(6S):137S-151S.e3.
mia. J Vasc Surg 2019;69(6S):3S-125S.e40. Bypass versus angioplasty to treat severe 16. Iida O, Takahara M, Soga Y, Kodama
2. Fowkes FGR, Rudan D, Rudan I, et al. limb ischemia: factors that affect treat- A, Terashi H, Azuma N. Three-year out-
Comparison of global estimates of preva- ment preferences of UK surgeons and in- comes of surgical versus endovascular
lence and risk factors for peripheral artery terventional radiologists. J Vasc Surg 2004; revascularization for critical limb ische
disease in 2000 and 2010: a systematic 39:1026-32. mia: the SPINACH study (Surgical Recon-
review and analysis. Lancet 2013; 382: 9. Farber A, Eberhardt RT. The current struction Versus Peripheral Intervention
1329-40. state of critical limb ischemia: a system- in Patients With Critical Limb Ischemia).
3. Mustapha JA, Katzen BT, Neville RF, atic review. JAMA Surg 2016;151:1070-7. Circ Cardiovasc Interv 2017; 10(12):
et al. Determinants of long-term out- 10. Goodney PP, Travis LL, Nallamothu e005531.
comes and costs in the management of BK, et al. Variation in the use of lower 17. Adam DJ, Beard JD, Cleveland T, et al.
critical limb ischemia: a population-based extremity vascular procedures for critical Bypass versus angioplasty in severe ischae
cohort study. J Am Heart Assoc 2018; limb ischemia. Circ Cardiovasc Qual Out- mia of the leg (BASIL): multicentre, ran-
7(16):e009724. comes 2012;5:94-102. domised controlled trial. Lancet 2005;
4. Farber A. Chronic limb-threatening 11. Menard MT, Farber A, Assmann SF, 366:1925-34.
ischemia. N Engl J Med 2018;379:171-80. et al. Design and rationale of the Best En- 18. Gerhard-Herman MD, Gornik HL,
5. Hirsch AT, Haskal ZJ, Hertzer NR, et al. dovascular Versus Best Surgical Therapy Barrett C, et al. 2016 AHA/ACC guideline
ACC/AHA guidelines for the management for Patients With Critical Limb Ischemia on the management of patients with low-
of patients with peripheral arterial dis- (BEST-CLI) trial. J Am Heart Assoc 2016; er extremity peripheral artery disease:
ease (lower extremity, renal, mesenteric, 5(7):e003219. a report of the American College of Car-
and abdominal aortic): a collaborative re- 12. Fleisher LA, Fleischmann KE, Auer- diology/American Heart Association Task
port from the American Associations for bach AD, et al. 2014 ACC/AHA guideline Force on Clinical Practice Guidelines. Cir-
Vascular Surgery/Society for Vascular Sur- on perioperative cardiovascular evalua- culation 2017;135(12):e726-e779.
gery, Society for Cardiovascular Angiog- tion and management of patients under- 19. Katsanos K, Spiliopoulos S, Kitrou P,
raphy and Interventions, Society for Vas- going noncardiac surgery: executive sum- Krokidis M, Karnabatidis D. Risk of death
cular Medicine and Biology, Society of mary: a report of the American College of following application of paclitaxel-coated
Interventional Radiology, and the ACC/ Cardiology/American Heart Association balloons and stents in the femoropopli-
AHA Task Force on Practice Guidelines Task Force on Practice Guidelines. Circu- teal artery of the leg: a systematic review
(writing committee to develop guidelines lation 2014;130:2215-45. and meta-analysis of randomized con-
for the management of patients with pe- 13. Powell R, Menard M, Farber A, et al. trolled trials. J Am Heart Assoc 2018;
ripheral arterial disease) — summary of Comparison of specialties participating 7(24):e011245.
recommendations. J Vasc Interv Radiol in the BEST-CLI trial to specialists treat- 20. Aboyans V, Ricco J-B, Bartelink M-LEL,
2006;17:1383-97. ing peripheral arterial disease nationally. et al. 2017 ESC guidelines on the diagno-
6. Creager MA, Matsushita K, Arya S, et al. J Vasc Surg 2019;69:1505-9. sis and treatment of peripheral arterial
Reducing nontraumatic lower-extremity 14. Conte MS, Geraghty PJ, Bradbury AW, diseases, in collaboration with the Euro-
amputations by 20% by 2030: time to get et al. Suggested objective performance pean Society for Vascular Surgery (ESVS):
to our feet: a policy statement from the goals and clinical trial design for evaluat- document covering atherosclerotic dis-
American Heart Association. Circulation ing catheter-based treatment of critical ease of extracranial carotid and vertebral,
2021;143(17):e875-e891. limb ischemia. J Vasc Surg 2009;50(6): mesenteric, renal, upper and lower ex-
7. Almasri J, Adusumalli J, Asi N, et al. A 1462-73.e1-3. tremity arteries. Eur Heart J 2018;39:763-
systematic review and meta-analysis of 15. Simons JP, Schanzer A, Flahive JM, 816.
revascularization outcomes of infrain- et al. Survival prediction in patients with Copyright © 2022 Massachusetts Medical Society.


Articol Exemplu

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Articol Exemplu

Uploaded by

Copyright:

Available Formats

new england

Surgery or Endovascular Therapy for Chronic Limb-

n engl j med 387;25 nejm.org December 22, 2022 2305

2306 n engl j med 387;25 nejm.org December 22, 2022

The New England Journal of Medicine

n engl j med 387;25 nejm.org December 22, 2022 2307

2525 Patients were assessed for eligibility

501 Were ineligible

1847 Underwent randomization

17 Were excluded after randomization

2308 n engl j med 387;25 nejm.org December 22, 2022

The New England Journal of Medicine

Figure 1 (facing page). Randomization and Outcomes.

n engl j med 387;25 nejm.org December 22, 2022 2309

Characteristic Cohort 1 Cohort 2

Overall Surgery Endovascular Therapy Overall Surgery Endovascular Therapy

Race or ethnic group — no./

n engl j med 387;25

total no. (%)

The New England Journal of Medicine

December 22, 2022

Copyright © 2022 Massachusetts Medical Society. All rights reserved.

Overall Surgery Endovascular Therapy Overall Surgery Endovascular Therapy

n engl j med 387;25

The New England Journal of Medicine

Copyright © 2022 Massachusetts Medical Society. All rights reserved.

Table 2. Efficacy and Safety Outcomes in Cohort 1.*

Endovascular Hazard Ratio

2312 n engl j med 387;25 nejm.org December 22, 2022

The New England Journal of Medicine

A Major Adverse Limb Events or Death B Major Reintervention

0.4 0.4 Endovascular therapy

C Above-Ankle Amputation D Death

n engl j med 387;25 nejm.org December 22, 2022 2313

2314 n engl j med 387;25 nejm.org December 22, 2022

The New England Journal of Medicine

n engl j med 387;25 nejm.org December 22, 2022 2315

2316 n engl j med 387;25 nejm.org December 22, 2022

The New England Journal of Medicine

You might also like