Journal of the American Heart Association

ORIGINAL RESEARCH

Comparative Outcomes of Interventions for

Femoropopliteal Chronic Total Occlusion
Versus Non–­Chronic Total Occlusion
Lesions From the Multicenter XLPAD
Registry
Shirling Tsai , MD; Yulun Liu , PhD; Lawrence Hoang, MD; Michael Vu, DO; Hua Lu , MD;
Bala Ramanan , MBBS; David Fernandez Vazquez , MD; Zachary Rosol, MD; Sameh Sayfo, MD;
Mohamad Amer Alaiti, MD; Panagiotis Koutakis , PhD; Emmanouil S. Brilakis , MD PhD;
Mehdi H. Shishehbor , DO MPH; Subhash Banerjee , MD

BACKGROUND: Endovascular intervention of femoropopliteal chronic total occlusions (CTOs) is technically more complex.
However, there is lack of comparative analysis between CTO and non-­CTO femoropopliteal interventions.

METHODS AND RESULTS: We report procedural details and outcomes of patients treated for femoropopliteal CTO and non-­CTO
lesions in the XLPAD (Excellence in Peripheral Artery Disease) registry (NCT01904851) between 2006 and 2019. Primary out-
comes were procedural success and 1-­year major adverse limb events, a composite of all-­cause death, target limb revascu-
larization, or major amputation. Analysis included 2895 patients (CTO: n=1516 patients; non-­CTO: n=1379 patients) with 3658
lesions (CTO: n=1998 lesions; non-­CTO: n=1660 lesions). Conventional balloon angioplasty (20.86% versus 33.48%, P<0.001)
or drug-­coated balloon angioplasty (1.26% versus 2.93%, P<0.001) were more frequent in the non-­CTO group, whereas bare-­
metal stents (28.09% versus 20.22%, P<0.001) or covered stents (4.08% versus 1.83%, P<0.001) were more frequent in the
CTO group. Debulking procedures were more commonly performed in the non-­CTO group (41.44% versus 53.13%, P<0.001),
despite a similar degree of calcification between the 2 groups. Procedural success was higher in the non-­CTO group (90.12%
versus 96.79%, P<0.001). Procedural complications were higher in the CTO group (7.21% versus 4.66%, P=0.002), mainly
due to excess distal embolization (1.5% versus 0.6%, P=0.015). Adjusted 1-­year major adverse limb events were higher in the
CTO group (22.47% versus 18.77%, P=0.019), driven mainly by target limb revascularization (19.00% versus 15.34%, P=0.013).

CONCLUSIONS: Procedural success is lower for endovascular treatment of femoropopliteal CTO compared with non-­CTO le-
sions. CTO lesions are associated with higher rates of periprocedural complications and reinterventions after 1 year.

Key Words: chronic total occlusion ■ femoropopliteal disease ■ outcomes analysis ■ peripheral artery disease ■ prospective registry

T
he femoropopliteal artery constitutes the most fre- femoropopliteal lesions involve a chronic total occlusion
quent location for lower extremity peripheral artery (CTO).1 Endovascular treatment strategies for the femo-
interventions in patients with symptomatic pe- ropopliteal segment vary widely, particularly between pa-
ripheral artery disease (PAD). Moreover, nearly 40% of tients with femoropopliteal CTO and those without CTO

Correspondence to: Shirling Tsai, MD, University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, MC 9157, Dallas, TX 75390-­9157.
Email: shirling.tsai@utsouthwestern.edu
This article was sent to Tazeen H. Jafar, MD, MPH, Associate Editor, for review by expert referees, editorial decision, and final disposition.
For Sources of Funding and Disclosures, see page 9.
© 2023 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access article under the terms of the Creative
Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use
is non-commercial and no modifications or adaptations are made.
JAHA is available at: www.ahajournals.org/journal/jaha

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284251

Tsai et al Treatment of CTO vs Non-­CTO Lesions

in Peripheral Artery Disease) registry to describe thera-

CLINICAL PERSPECTIVE peutic approaches to femoropopliteal CTO and non-­CTO
lesions and analyzed immediate and 1-­year outcomes in
What Is New? patients undergoing clinically indicated endovascular in-
• Analysis of patient, lesion, and endovascular terventions of these lesions.
procedural features of femoropopliteal occlu-
sive disease comprising chronic total occlu-
sion (CTO) and non-­CTO lesions demonstrate METHODS
that treatment of femoropopliteal CTO has been
acceptable; however, lower rates of procedural Study Design and Population
success compared with non-­CTO lesions have The XLPAD registry (NCT 01904851) is an ongoing,
been reported. multicenter registry of patients undergoing endovascu-
• Patients with femoropopliteal CTO are more lar revascularization for infrainguinal PAD.7 Data from
likely to suffer from periprocedural distal embo- 23 participating sites are entered in the REDCap web-­
lization and have higher rates of major adverse based data acquisition system, with a core laboratory–­
limb events and target lesion revascularization adjudicated angiographic review of index procedures.
at 1 year. This study is based on a cohort of 2895 patients un-
dergoing endovascular treatment for symptomatic
What Are the Clinical Implications? PAD with and without femoropopliteal CTO between
• These findings may be considered in clinical
decision making for cases involving femoro- 2006 and 2019. Periprocedural and 12-­ month out-
popliteal CTO. comes after the index procedure are collected during
follow-­up visits. Management and clinical decisions are
at the discretion of the care team.
The registry is conducted in accordance with
the Declaration of Helsinki and approved by the in-
Nonstandard Abbreviations and Acronyms stitutional review boards of participating hospitals.
Study participants were consented according to the
CTO chronic total occlusion protocols for each of the local site institutional re-
DCB drug-­coated balloon view boards. Deidentified imaging studies are inde-
ISR in-­stent restenosis pendently reviewed by the Veterans Affairs North Texas
MACE major adverse cardiac event Angiography and Ultrasound core laboratory. The
MALE major adverse limb event clinical coordinating center and data servers are lo-
XLPAD Excellence in Peripheral Artery Disease cated at the University of Texas Southwestern Medical
Center, Dallas, Texas. Periodic data audits as well as
remote and on-­site monitoring are performed and re-
ported to an independent Data Safety and Monitoring
Committee. Data that support the findings of this study
(non-­CTO).2 Although long-­ segment femoropopliteal
are available from the corresponding author upon rea-
CTO (Trans Atlantic Society Classification II type C and
sonable request.
D lesions) were previously managed primarily by surgical
bypass.3 In current practice, both endovascular revascu-
larization and surgical bypass are commonly performed Definitions and Outcome Measures
as an initial treatment strategy.4,5 Therefore, understand- The study population was divided into 2 groups based
ing procedural steps and treatment outcomes after in- on the presence of femoropopliteal CTO at the tar-
terventions for CTO compared with non-­CTO lesions is get lesion site. Baseline, target lesion, and proce-
clinically relevant. In the coronary literature, interventions dural characteristics were described for each group.
involving CTO are associated with lower rates of proce- Baseline characteristics included demographics, co-
dural success and higher rates of major adverse cardiac morbidities, and Rutherford classification, which was
events (MACE) when compared with interventions on dichotomized into intermittent claudication (Rutherford
non-­CTO lesions.6 However, no direct comparison of classification 1–­3) versus chronic limb-­threatening is-
femoropopliteal CTO versus non-­CTO interventions has chemia (Rutherford classification 4–­6). Target lesion
been reported, because such analyses are predicated characteristics included lesion length and presence
on inclusion of lesion and procedural details that are of severe calcification (at least 5 mm of calcification
often not available through administrative and third-­party on both sides of the vessel), in-­stent restenosis (ISR),
payer databases. diffuse disease (angiographic disease >30% diameter
Herein, we leveraged the unique features of the mul- stenosis for at least 20 mm in length), and number of
ticenter, core laboratory–­adjudicated XLPAD (Excellence run-­off below-­the-­knee vessels. CTO morphology was

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284252

Tsai et al Treatment of CTO vs Non-­CTO Lesions

Table 1. Baseline Patient Characteristics

Femoropopliteal non-CTO Femoropopliteal CTO

Variable (N=1379) Non-CTO, % (N=1516) CTO, % P value

Age, y
Mean±SD 67.42±9.83 65.85±9.84 Wilcoxon rank sum
test, <0.001*
Median (IQR) 67 (61–­74) 65 (59–­72)
Sex
Men 961 69.79% 1171 77.34% <0.001*
Women 416 30.21% 343 22.66%
Race and ethnicity
White (not 968 71.86% 1093 74.35% 0.294
Hispanic)
Black 239 17.74% 253 17.21%
Hispanic 121 8.98% 109 7.41%
Other (Asian, 19 1.41% 15 1.02%
Native American,
Other)
Smoking (any tobacco use)
Within the past 576 42.29% 825 55.07% <0.001*
year
>1 y ago 553 40.60% 489 32.64%
Never 233 17.11% 184 12.28%
Hypertension 1252 91.39% 1364 90.45% 0.420
Diabetes 759 55.97% 786 52.47% 0.066
Hyperlipidemia 1173 86.70% 1277 85.08% 0.236
(dyslipidemia)
Chronic kidney 212 15.37% 189 12.47% 0.027*
disease
Coronary artery 800 58.01% 863 56.93% 0.580
disease
Heart failure 159 11.53% 220 14.51% 0.020*
Prior nonfatal MI 264 19.14% 358 23.61% 0.004*
Prior stroke 96 6.96% 139 9.17% 0.035*
ABI
Mean±SD 0.71±0.24 0.61±0.23 Wilcoxon rank sum
test <0.001*
Median (IQR) 0.70 (0.58–­0.85) 0.60 (0.48–­0.72)
ABI (category)
ABI<0.9 750 82.42% 949 92.68% <0.001*
0.9≤ABI<1.4 148 16.26% 67 6.54%
1.4≤ABI 12 1.32% 8 0.78%
Ischemia level
CLTI 459 33.28% 601 39.64% <0.001*
Claudication 920 66.72% 915 60.36%

ABI indicates ankle–­brachial index; CLTI, chronic limb-­threatening ischemia; CTO, chronic total occlusion; IQR, interquartile range; and MI, myocardial
infarction.
*P<0.05.

assessed by examination of the proximal stump (blunt, as operator-­guided stenting of the target lesion when
concave versus tapered, convex), distal stump (blunt an initial non–­stent-­based strategy was unsuccessful.
versus concave), presence of side branches proximally We examined the following outcomes: proce-
and distally, and presence of collateral filling of the dis- dural success, periprocedural complications, and 1-­
tal vessels.8 Procedural characteristics included type year major adverse limb events (MALE) and MACE.
and number of balloons or stents and use of debulking Procedural success is defined as ≤30% target lesion
or CTO crossing devices. Bail-­out stenting was defined residual stenosis without complication after index

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284253

Tsai et al Treatment of CTO vs Non-­CTO Lesions

intervention. Periprocedural complications (immedi- variables was used. To display outcomes of MALE
ately following index procedure to 30 days after the and target limb revascularization between patient with
procedure) included residual dissection (flow-­or non-­ femoropopliteal CTO and femoropopliteal non-­CTO on
flow-­limiting), access site hematoma (>5 cm or <5 cm), survivals, adjusted Kaplan-­Meier survival curves were
retroperitoneal hematoma, distal embolization, bleed- generated for each outcome, controlling for potential
ing diathesis, acute renal failure, perforation, emer- confounders such as age, sex, smoking status, chronic
gency surgery, or their composite. MALE are defined renal insufficiency, heart failure, prior nonfatal MI, prior
as all-­cause death, repeat revascularization (endovas- stroke, and Rutherford class. To determine the associ-
cular or surgical), or major amputation. MACE consist ations between patients with femoropopliteal CTO and
of all-­cause death, nonfatal myocardial infarction (MI), femoropopliteal non-­ CTO and outcomes of interest
or stroke. The individual components of MACE and including MALE, MACE, repeat revascularization, and
MALE were also studied. repeat endovascular intervention, univariate and multi-
variable logistic regression models were performed to
Statistical Analysis compute unadjusted and adjusted odds ratios (ORs),
To compare baseline and lesion characteristics, pro- in which the multivariable logistic regression models
cedure details, and procedure outcomes, univariate were adjusted for age, sex, smoking status, chronic
analysis was used. Continuous variables were de- renal insufficiency, heart failure, prior nonfatal MI, prior
scribed using mean±SD or median with interquartile stroke, and Rutherford class (chronic limb-­threatening
range (IQR). Discrete and categorical variables were ischemia versus claudication). Regression results were
presented as numbers and percentages. Differences presented as ORs with 95% CIs. All statistical tests
between the femoropopliteal CTO and femoropopliteal were 2-­sided, with P<0.05 considered to be statisti-
non-­CTO groups were compared using a t test or cally significant. All statistical analyses were performed
Wilcoxon rank sum test for continuous variables, using R version 3.6.1 (The R Foundation for Statistical
whereas a χ2 test or Fisher exact test for categorical Computing).

Table 2. Lesion Characteristics

Femoropopliteal Femoropopliteal
Variable non-­CTO (N=1660) Non-­CTO, % CTO (N=1998) CTO, % P value

Lesions treated (continuous), per procedure

Mean±SD 1.38±0.70 1.47±0.80 Wilcoxon rank sum
test=0.003*
Vessel diameter, mm (continuous), per procedure
Mean±SD 5.17±1.09 5.21±5.34 Wilcoxon rank sum
test=0.020*
Median (IQR) 5.20 (4.91–­6.00) 5.14 (4.60–­5.83)
Lesion length, mm (categorical)
Lesion <100 mm 630 53.07% 467 30.52% <0.001*
100 mm ≤ lesion 384 32.35% 519 33.92%
<200 mm
Lesion ≥200 mm 173 14.57% 544 35.56%
Lesion length, mm (continuous), per procedure
Mean±SD 110.68±84.68 167.21±103.01 Wilcoxon rank sum
test <0.001*
Median (IQR) 90 (48–­150) 145 (86.13–­239.75)
Lesion characteristics
Severe calcification 691 41.63% 863 43.19% 0.357
(heavily calcified)
Diffuse disease 960 57.83% 1449 72.52% <0.001*
In-­stent restenosis 244 14.70% 352 17.62% 0.020*
Run-­off vessels (range 0–­3)
0/1 319 25.71% 351 27.02% 0.139
2 364 29.33% 414 31.87%
3 558 44.96% 534 41.11%

CTO indicates chronic total occlusion; and IQR, interquartile range.

*P<0.05.

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284254

Tsai et al Treatment of CTO vs Non-­CTO Lesions

167.21±103.01 mm versus 110.68±84.68 mm; P<0.001;

RESULTS median [IQR], 145 mm [86.13–­239.75] versus 90 mm [48–­
A total of 2895 patients with femoropopliteal PAD 150]), and greater proportion of patients with diffuse dis-
were included: 1516 in the CTO and 1379 in the non-­ ease and ISR. However, number of below-­the-­knee run-­off
CTO groups, respectively, with 3658 lesions (femo- vessels were similar between the 2 groups. Analysis of
ropopliteal CTO: n=1998 lesions and femoropopliteal characteristics of the most proximal and most distal CTO
non-­ CTO: n=1660 lesions). Baseline demographics lesions (1562 lesions total) was conducted. The proximal
and comorbidities are described in Table 1. Current stump was blunt in 666 lesions (42.63%) and tapered in
smoking (within the past year) was more prevalent 387 lesions (24.77%). A side branch was present at the
in the CTO group, as was heart failure, prior MI, and proximal stump in 561 lesions (35.91%) and absent in 634
prior stroke. The CTO group had a lower mean ankle–­ lesions (40.58%). The distal stump was blunt in 703 lesions
brachial index (ABI; 0.61±0.23 versus 0.71±0.24, (45.01%) and tapered in 311 lesions (19.91%). A distal side
P<0.001) and a higher percentage of patients with branch was present in 388 lesions (24.84%) and absent in
chronic limb-­ threatening ischemia (39.64% versus 796 lesions (50.96%). Collateral filling of the distal vessel
33.28%, P<0.001). Antiplatelet and statin medication was observed in 1155 lesions (73.94%). A prior attempt to
use was similar between the 2 groups. cross the CTO was performed in 37 cases (2.37%).
In terms of lesion characteristics (Table 2), patients The most common crossing strategy for CTO le-
in the CTO group had longer lesion lengths (mean±SD, sions was a wire-­catheter combination (68%). Of the

Table 3. Procedural Details

Femoropopliteal Femoropopliteal
Variable non-­CTO (N=1660) Non-CTO, % CTO (N=1998) CTO, % P value

Conventional balloon 542 33.48% 412 20.86% <0.001*

Drug-­coated balloon 48 2.93% 25 1.26% <0.001*
Balloon length, mm
Mean±SD 85.10±53.28 102.39±49.55 Wilcoxon rank sum test <0.001*
Median (IQR) 80 (40–­110) 100 (66.67–­126.67)
No. of balloons
0 116 7.09% 152 7.67% <0.001*
1 680 41.56% 642 32.39%
≥2 840 51.34% 1188 59.94%
Bare-­mental stent 331 20.22% 557 28.09% <0.001*
Drug-­eluting stent 93 5.68% 146 7.36% 0.05
Covered stent 30 1.83% 81 4.08% <0.001*
Vascular mimetic stent 122 7.45% 161 8.12% 0.496
Stent length, mm
Mean±SD 85.52±41.31 99.87±37.51 Wilcoxon rank sum test <0.001*
Median (IQR) 80 (50–­120) 100 (80–­120)
No. of stents
Mean±SD 0.67±0.98 1.34±1.40 Wilcoxon rank sum test <0.001*
Median (IQR) 0 (0–­1) 1 (0–­2)
Bail-­out stenting 92 5.54% 146 7.31% 0.037*
Treatment (unit: procedure-­level)
Debulking
All types 882 53.13% 828 41.44% <0.001*
CTO crossing devices
All types 103 6.20% 1670 83.58% <0.001*
Fluoroscopy time
Mean±SD 24.82±15.39 38.75±20.90 Wilcoxon rank sum test <0.001*
Median (IQR) 21.30 (14.40– ­30.35) 34.54
(24.10–­48.50)

CTO indicates chronic total occlusion; and IQR, indicates interquartile range.
*P<0.05.

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284255

Tsai et al Treatment of CTO vs Non-­CTO Lesions

Table 4. Periprocedural Outcomes

Femoropopliteal non-­ Femoropopliteal CTO

Variable CTO (N=1660) Non-CTO % (N=1998) CTO, % P value

Procedural success 1600 96.79% 1797 90.12% <0.001*

Procedural 77 4.66% 144 7.21% 0.002*
complications
Complications
Residual dissection (flow-­limiting)
Yes 18 1.09% 29 1.45% 0.407
Residual dissection (non-­flow-limiting)
Yes 23 1.39% 37 1.85% 0.332
Access site hematoma <5 cm
Yes 3 0.18% 5 0.25% 0.736
Access site hematoma >5 cm
Yes 6 0.36% 6 0.30% 0.973
Retroperitoneal hematoma
Yes 5 0.30% 6 0.30% 1.000
Distal embolization
Yes 10 0.60% 30 1.50% 0.015*
Bleeding
Yes 2 0.12% 3 0.15% 1.000
Acute renal failure
Yes 4 0.24% 2 0.10% 0.420
Perforation
Yes 5 0.30% 14 0.70% 0.109
Emergency surgery
Yes 1 0.06% 4 0.20% 0.385
Composite hematoma (access site <5 cm, >5 cm, retroperitoneal)
Yes 14 0.84% 17 0.85% 1.000

CTO indicates chronic total occlusion.

*P<0.05.

CTO crossed, 71.25% were reported to be true lumen or drug-­coated balloon (DCB) angioplasty (1.26% ver-
crossing, whereas 19.78% were subintimal crossing. sus 2.93%, P<0.001) (Table 3). CTO lesions were more
Reentry from a subintimal dissection plane involved often treated with bare-­metal stents (28.09% versus
use of a reentry device in 52.31% of cases. In 69.59% 20.22%, P<0.001), but there was no difference in use
of cases, the CTO was crossed from an antegrade of drug-­eluting stents. Lesions with CTO were also
approach. Retrograde crossing was used in 18.95% more often treated with covered stents (4.08% versus
of cases, and a combined antegrade–­retrograde ap- 1.83%, P<0.001). The need for bail-­out stenting was
proach was used in 3.89% of cases. US Food and expectedly higher in the CTO group (7.31% versus
Drug Administration–­ approved crossing catheters 5.54%, P=0.037).
were more commonly used in CTO lesions as com- The CTO group had a lower procedural success
pared with non-­CTO lesions (83.58% versus 6.20%, (90.12% versus 96.79%, P<0.001) and higher over-
P<0.001). Examination of CTO crossing device use re- all rate of procedural complications (7.21% versus
vealed a broad array of device mechanisms (ultrasonic, 4.66%, P=0.002) compared with the non-­CTO group
microdissection, spinning distal tip). Atherectomy was (Table 4). However, mean ABI improved in both groups
less common in the CTO group as compared with the at 6 months (0.83±0.23 versus 0.87±0.29, P=0.089).
non-­CTO group (41.44% versus 53.13%, P<0.001). A At 12 months, mean ABI, although higher from preop-
variety of atherectomy devices was used, including or- erative value, was significantly lower in the CTO group
bital atherectomy (5.76%), laser atherectomy (7.62%), (0.81±0.24 versus 0.85±0.24, P=0.01). The higher com-
and directional atherectomy (11.27%). Distal embolic plication rate was driven mostly by nearly 3 times as
protection devices were used in 37.58% of CTO cases. many distal embolization events. One-­year follow-­up
CTO lesions were less often treated with either plain revealed higher composite MALE rate in the CTO group
balloon angioplasty (20.86% versus 33.48%, P<0.001) (22.47% versus 18.77%, P=0.019); however, there was

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284256

Tsai et al Treatment of CTO vs Non-­CTO Lesions

Table 5. Twelve-­Month Outcomes

Femoropopliteal non-­ Femoropopliteal CTO

Variable CTO (N=1379) Non-­CTO, % (N=1516) CTO, % P value

MALE
Death 30 2.34% 46 3.17% 0.230
Repeat endovascular 175 13.69% 233 16.10% 0.088
intervention
Surgical target limb 30 2.35% 52 3.59% 0.074
revascularization
Amputation in target limb 24 1.88% 35 2.42% 0.404
(major)
MI 17 1.33% 29 2.00% 0.231
Stroke 4 0.31% 8 0.55% 0.378
Composite MALE (death/repeat endovascular or sugical intervention/major amputation)
Yes 241 18.77% 327 22.47% 0.019*
Composite MACE (death/MI/stroke)
Yes 46 3.59% 75 5.16% 0.057
Composite revascularization (repeat/surgical)
Yes 196 15.34% 275 19.00% 0.013*

CTO indicates chronic total occlusion; MACE, major adverse cardiac events; MALE, major adverse limb events; and MI, myocardial infarction.
*P<0.05.

no significant difference in all-­cause mortality or major reintervention (endovascular or surgical) (OR, 1.278
amputation rates between the CTO and non-­ CTO [95% CI, 1.039–­1.573]; P=0.020) at 1 year (Table 6).
groups (Table 5). There was a significantly higher rate Finally, adjusted Kaplan-­Meier analysis was performed
of repeat revascularization (including both endovascu- to demonstrate differences in MALE-­free and target
lar and surgical procedures) in the CTO group (19.00% limb revascularization–­free survival over 1 year of fol-
versus 15.34%, P=0.013). low-­up (Figure).
In univariate logistic regression models, com-
posite MALE, composite MACE, and any revascu-
larization procedure (endovascular or surgical) were
significantly more likely in patients with femoropopli- DISCUSSION
teal CTO (Table 6). On multivariate analysis, after ad- The current analysis provides an in-­depth insight into
justing for patient-­level risk factors (age, sex, smoking patient, lesion, and procedural features of femoro-
status, chronic renal insufficiency, heart failure, prior popliteal PAD comprising CTO and non-­CTO lesions
nonfatal MI, prior stroke, and Rutherford classifica- treated by endovascular intervention. These data con-
tion), patients with femoropopliteal CTO had a higher firm that treatment of femoropopliteal CTO has accept-
risk of composite MALE at 1 year (OR, 1.217 [95% CI, able, but lower, rates of procedural success compared
1.002–­1.477]; P=0.048) and higher likelihood for any with non-­ CTO lesions. Although mortality and limb

Table 6. Results of Univariate and Multivariable Logistic Regression Models for Comparing CTO/Femoropopliteal Versus
Non-­CTO/Femoropopliteal on Outcomes

Univariate logistic regression Multivariable logistic regression

Outcome Odds ratio (95% CI) P value Odds ratio (95% CI)† P value

Composite MALE (death/repeat endovascular or surgical intervention/ 1.255 (1.041–­1.512) 0.017* 1.217 (1.002–­1.477) 0.048*
major amputation)
Composite MACE (death/MI/stroke) 1.463 (1.005–­2.128) 0.047* 1.389 (0.938–­2.055) 0.101
Composite revascularization (repeat endovascular or surgical) 1.295 (1.059–­1.584) 0.012* 1.278 (1.039–­1.573) 0.020*
Repeat endovascular intervention 1.210 (0.978–­1.496) 0.079 1.220 (0.980–­1.518) 0.076

MACE indicates major adverse cardiac events; MALE, major adverse limb events; and MI, myocardial infarction.
*P<0.05.
†
Odds ratios were adjusted by the following confounders: age, sex, smoking, chronic kidney disease, heart failure, prior nonfatal MI, prior stroke, and
Rutherford class.

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284257

Tsai et al Treatment of CTO vs Non-­CTO Lesions

Figure. Adjusted Kaplan-­Meier curves for freedom from MALE (A) and freedom from target limb revascularization (B) in
patients with FP CTO and non-­CTO.
CTO indicates chronic total occlusion; FP, femoropopliteal; and MALE, major adverse limb events.

salvage at 1 year are not significantly different between stenting after suboptimal results from DCB. Cost and
the groups, patients with CTO are more likely to suffer operator preference likely contributed to the low use of
from periprocedural distal embolization and undergo drug-­eluting technology in this registry. Although the
more frequent target lesion revascularization at 1 year. DCBs were Food and Drug Administration approved
The unique features of the XLPAD registry data in late 2014, the studies demonstrating improved ef-
provide exceptional granularity into procedural details ficacy over conventional percutaneous balloon angio-
that are often absent in other registries. As such, the plasty with midterm follow-­up were published in mid to
findings that procedural success is lower and com- late 2015.10,11 This may account for some later use of
plication rate is higher in the CTO group is not sur- DCBs. The meta-­analysis12 by Katsanos et al may have
prising. However, the current analysis shows that this abruptly decreased use of paclitaxel-­eluting or coated
is largely through higher rates of distal embolization, devices in 2019, but this only accounted for a small
despite the use the distal embolic protection devices percentage of cases over the time course of the study,
in one-­third of patients with CTO. This is perhaps and was unlikely the sole cause of the low use of DCBs
due to release of downstream embolic debris during and drug-­eluting stents in this cohort.
crossing of CTO lesions that precedes deployment of Another notable finding was the relatively low fre-
any embolic protection device. The higher prevalence quency of stenting, even in the CTO group. An ear-
of patients with chronic limb-­ threatening ischemia lier analysis of cases in the XLPAD registry suggests
and ISR lesions in the CTO group could also account that ≈50% of femoropopliteal procedures involved
for this difference.9 any type of stent, including bare metal, covered, and
Another interesting finding is that atherectomy was drug-­eluting stents.7 Similarly, a comparison of stent-­
performed more often in the non-­CTO group. This may versus nonstent-­ based interventions in the femoro-
have been related to more frequent subintimal cross- popliteal segment from the XLPAD found that 46% of
ings in the CTO group, which would have deterred the femoropopliteal interventions were stent-­ based, and
use of atherectomy devices. All types of atherectomy that stents were associated with significantly higher
were used; however, the use of directional atherectomy procedural costs.13 Consistent with the prior reports,
was most frequent, because it may be favored in cases in the current report, a total of 47.65% of femoropop-
with eccentric calcified nonocclusive plaque. Notably, liteal CTO lesions were treated with stenting, plus an
the use of a DCB as the sole treatment modality was low additional 7.31% that were treated with bail-­out stent-
in both the CTO group and non-­CTO group. This does ing. The greater proportion of patients with ISR in
not, however, account for combination treatments, the CTO group may also contribute to the lower fre-
such as the use of atherectomy before DCB or bail-­out quency of stenting, because patients with ISR were

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284258

Tsai et al Treatment of CTO vs Non-­CTO Lesions

more commonly managed with balloon angioplasty, as In conclusion, procedural success is lower for
noted in prior publications.13 endovascular treatment of femoropopliteal CTO
The adjusted analysis controlled for clinically rel- compared with non-­CTO lesions. Periprocedural com-
evant patient-­ level risk factors but not other lesion plications are higher for CTO lesions, driven mostly by
characteristics or procedural details. CTO versus distal embolization events. Although there are no dif-
non-­CTO is the major lesion-­level driver of the out- ferences in mortality or limb salvage at 1 year, patients
comes of interest (limb salvage and death); therefore, with femoropopliteal CTO were more likely to require
other lesion characteristics (eg, lesion length, severity reintervention.
of calcification, CTO morphology) and treatment de-
tails (eg, atherectomy, use of DCBs or covered stents)
were not included in the adjusted regression analysis. ARTICLE INFORMATION
Furthermore, many treatment variables may be depen- Received February 2, 2023; accepted May 2, 2023.

dent on lesion characteristics, and therefore would not Affiliations

be truly independent variables for analysis. North Texas Veterans Affairs Health Care Systems, Dallas, TX (S.T., L.H.,
On univariate analysis, it appeared that 1-­year out- M.V., H.L., B.R., D.F.V., S.B.); University of Texas Southwestern Medical
Center, Dallas, TX (S.T., Y.L., H.L., B.R.); Methodist Health System, Dallas,
comes, both MACE and MALE, were worse for patients TX (L.H., M.V.); Baylor Scott and White Hospital, Dallas, TX (Z.R., S.B., S.S.);
with CTO. However, after adjusting for patient-­level risk Medical City Plano, Plano, TX (M.A.A.); Baylor University, Waco, TX (P.K.);
factors, only MALE, driven by an increased rate of re- Minnesota Heart Institute, Minneapolis, MN (E.S.B.); and University Hospital
Harrington Heart and Vascular Institute, Cleveland, OH (M.H.S.).
intervention, were significantly different for the CTO
group. This suggests that the worse outcomes are at Acknowledgments
least partially attributable to higher atherosclerotic dis- The authors extend their gratitude to the members of the XLPAD registry
across all sites that make our work possible.
ease burden in the CTO group. Additionally, the higher
rate of reintervention suggests that the CTO group may Sources of Funding
require more careful surveillance and aggressive med- This work was supported by institutional research grant support from
AngioSafe Incorporated–­XLPAD Registry.
ical management.
Our data for the first time provide real-­world esti- Disclosures
mates of procedural success, periprocedural compli- Dr Banerjee reports the following consulting relationships: Medtronic, Cordis,
Livmor, Kaneka, and AstraZeneca. Dr Alaiti reports the following consulting
cations, and need for reintervention. In both the CTO
relationships: Abbott Vascular, Abiomed, and AngioSafe. Dr Brilakis reports
and non-­CTO groups, mean ABI increased after the the following consulting relationships: Medtronic, Abbott, Amgen, Asahi,
procedure, and the improvement was sustained for Siemens, GE, Teleflex, and InfraRedex. Dr Shishehbor reports the follow-
ing consulting relationships: Abbot Vascular, Boston Scientific, Medtronic,
1 year. However, although at 6 months there was no dif-
Philips, and Terumo-­Advisory Board. The remaining authors have no disclo-
ference in ABI between the CTO and non-­CTO groups, sures to report.
at 12 months the mean ABI for the CTO group was sig-
nificantly lower than the non-­CTO group, which also
suggests that interventions for CTO lesions are less REFERENCES
durable. These findings will inform both providers and 1. Banerjee S, Shishehbor MH, Mustapha JA, Armstrong EJ, Ansari
patients, because femoropopliteal CTOs are frequently M, Rundback JH, Fisher B, Pena CS, Brilakis ES, Lee AC, et al. A
encountered in patients with symptomatic PAD. The percutaneous crossing algorithm for femoropopliteal and tibial ar-
tery chronic total occlusions (PCTO algorithm). J Invasive Cardiol.
success rates of >90% in our registry perhaps speaks 2019;31:111–­119.
to the more generalized CTO skills of endovascular in- 2. Banerjee S, Jeon-­Slaughter H, Tsai S, Mohammad A, Foteh M, Abu-­
terventionalists, unlike coronary CTO operators who Fadel M, Gigliotti OS, Cawich I, Rodriguez G, Kumbhani D, et al.
Comparative assessment of procedure cost and outcomes between
are more specialized. However, these results need to guidewire and crossing device strategies to cross peripheral artery
be interpreted considering the limitations of this obser- chronic total occlusions. JACC Cardiovasc Interv. 2016;9:2243–­2252.
vational registry. doi: 10.1016/j.jcin.2016.08.010
3. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, Fowkes FG;
An important limitation of this study is selection TASC II Working Group. Inter-­society consensus for the management of
bias. The registry requirement for sites to report con- peripheral arterial disease (TASC II). J Vasc Surg. 2007;45:S5–­S67.
secutive patients limits this bias. The selection bias on 4. Abu Dabrh AM, Steffen MW, Asi N, Undavalli C, Wang Z, Elamin MB,
Conte MS, Murad MH. Bypass surgery versus endovascular interven-
the use of lesion-­crossing techniques and devices re- tions in severe or critical limb ischemia. J Vasc Surg. 2016;63:244–­253.
mains. Absence of longer follow-­up is another import- e11. doi: 10.1016/j.jvs.2015.07.068
ant limitation of the XLPAD registry. However, the vast 5. Conte MS, Bradbury AW, Kolh P, White JV, Dick F, Fitridge R, Mills JL,
Ricco JB, Suresh KR, Murad MH, et al. Global vascular guidelines on
amounts of in-­depth patient level and core laboratory–­ the management of chronic limb-­threatening ischemia. J Vasc Surg.
adjudicated procedural data in the XLPAD registry 2019;58:S1–­S109.e33.
highlight the diverse approaches to revascularization of 6. Brilakis ES, Banerjee S, Karmpaliotis D, Lombardi WL, Tsai TT,
Shunk KA, Kennedy KF, Spertus JA, Holmes DR Jr, Grantham JA.
femoropopliteal CTO and non-­CTO lesions in patients Procedural outcomes of chronic total occlusion percutaneous cor-
with symptomatic PAD. onary intervention: a report from the NCDR (National Cardiovascular

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.0284259

Tsai et al Treatment of CTO vs Non-­CTO Lesions

Data Registry). JACC Cardiovasc Interv. 2015;8:245–­253. doi: 10.1016/j. effect using a drug-­ coated balloon for femoropopliteal lesions: 24-­
jcin.2014.08.014 month results of IN.PACT SFA. J Am Coll Cardiol. 2015;66:2329–­2338.
7. Banerjee S, Sarode K, Mohammad A, Gigliotti O, Baig MS, Tsai S, doi: 10.1016/j.jacc.2015.09.063
Shammas NW, Prasad A, Abu-­Fadel M, Klein A, et al. Femoropopliteal 11. Rosenfield K, Jaff MR, White CJ, Rocha-­Singh K, Mena-­Hurtado C,
artery stent thrombosis: report from the Excellence in Peripheral Artery Metzger DC, Brodmann M, Pilger E, Zeller T, Krishnan P, et al. Trial of
Disease registry. Circ Cardiovasc Interv. 2016;9:e002730. doi: 10.1161/ a paclitaxel-­coated balloon for femoropopliteal artery disease. N Engl J
CIRCI​NTE​RVE​NTI​ONS.115.002730 Med. 2015;373:145–­153. doi: 10.1056/NEJMoa1406235
8. Saab F, Jaff MR, Diaz-­Sandoval LJ, Engen GD, McGoff TN, Adams 12. Katsanos K, Spiliopoulos S, Kitrou P, Krokidis M, Karnabatidis D.
G, Al-­Dadah A, Goodney PP, Khawaja F, Mustapha JA. Chronic total Risk of death following application of paclitaxel-­coated balloons and
occlusion crossing approach based on plaque cap morphology: stents in the femoropopliteal artery of the leg: a systematic review
the CTOP classification. J Endovasc Ther. 2018;25:284–­291. doi: and meta-­analysis of randomized controlled trials. J Am Heart Assoc.
10.1177/1526602818759333 2018;7:e011245. doi: 10.1161/JAHA.118.011245
9. Shrikhande GV, Khan SZ, Hussain HG, Dayal R, McKinsey JF, Morrissey 13. Banerjee S, Jeon-­Slaughter H, Armstrong EJ, Bajzer C, Abu-­Fadel M,
N. Lesion types and device characteristics that predict distal emboliza- Khalili H, Prasad A, Bou Dargham B, Kamath P, Addo T, et al. Clinical
tion during percutaneous lower extremity interventions. J Vasc Surg. outcomes and cost comparisons of stent and non-­stent interventions
2011;53:347–­352. doi: 10.1016/j.jvs.2010.09.008 in Infrainguinal peripheral artery disease: insights from the Excellence
10. Laird JR, Schneider PA, Tepe G, Brodmann M, Zeller T, Metzger C, in Peripheral Artery Disease (XLPAD) registry. J Invasive Cardiol.
Krishnan P, Scheinert D, Micari A, Cohen DJ, et al. Durability of treatment 2019;31:1–­9.

J Am Heart Assoc. 2023;12:e028425. DOI: 10.1161/JAHA.122.02842510