Endovascular Treatment and Outcomes for Femoropopliteal In-Stent Restenosis: Insights from the XLPAD Registry

Background There is limited “real-world” evidence examining treatment modalities and outcomes in patients with symptomatic peripheral arterial disease undergoing endovascular treatment of femoropopliteal (FP) in-stent restenosis (ISR). Materials and Methods We compared outcomes in 2,895 patients from the XLPAD registry (NCT01904851) between 2006 and 2019 treated for FP ISR (n = 347) and non-ISR (n = 2,548) lesions. Primary endpoint included major adverse limb events (MALE) at 1 year, a composite of all-cause death, target limb repeat revascularization, or major amputation. Results ISR patients were more frequently on antiplatelet (94.5% vs 89.4%, p=0.007) and statin (68.9% vs 60.3%, p=0.003) therapies. Lesion length was similar (ISR: 145 ± 99 mm vs. non-ISR: 142 ± 99 mm, p=0.55). Fewer treated ISR lesions were chronic total occlusions (47.3% vs. 53.7%, p=0.02) and severely calcified (22.4% vs. 44.7%, p < 0.001). Atherectomy (63.5% vs. 45.0%, p < 0.001) and drug-coated balloons (DCB; 4.7% vs. 1.7%, p < 0.001) were more frequently used in ISR lesions. The distal embolization rate was higher in ISR lesions (2.4% vs. 0.9%, p=0.02). Repeat revascularization (21.5% vs. 16.7%, p=0.04; Figure) was higher and freedom from MALE at 1 year was significantly lower (87% vs. 92.5%, p < 0.001) in the ISR group. Conclusion Atherectomy and DCB are more frequently used to treat FP ISR lesions. Patients with FP ISR have more intraprocedural distal embolization, higher repeat revascularization procedures, and lower freedom from MALE at 1 year.


Introduction
Endovascular treatment of symptomatic PAD has emerged as the preferred revascularization modality with the advent of newer technology and combination therapies [1]. Treatment of femoropopliteal (FP) disease is a unique challenge as anatomical and mechanical forces in this vascular bed promote the development of atherosclerosis and in-stent restenosis (ISR; [2]). FP ISR is a pervasive complication with an incidence reaching 40% after 1 year of intervention and 60% at 3-years [3,4]. Stenting and balloon angioplasty are the current mainstays of treatment, but long-term durability is poor. Treatment failure remains elevated with stent fracture and recurrent non-drug-coated stent restenosis limiting long-term patency [3]. erefore, examining "realworld" data on the treatment and outcomes of FP ISR is highly pertinent as the optimal approach to ISR lesions remains elusive.
is study utilizes patients enrolled in the Excellence in Peripheral Artery Disease (XLPAD) registry to compare clinical outcomes of patients with symptomatic PAD undergoing endovascular treatment for FP ISR and non-ISR lesions. e primary endpoint included major adverse limb events (MALE) at 1 year, a composite of all-cause death, target limb repeat revascularization, or major amputation.

Study Design and Population.
e XLPAD registry is an ongoing, multicenter registry of patients undergoing endovascular revascularization for infrainguinal PAD. It utilizes the REDCap web-based data acquisition system to gather data across 23 hospital sites in the United States with a core-lab adjudicated angiographic review of enrolled patients. We analyzed data of 2,895 patients undergoing endovascular treatment for symptomatic PAD with and without FP ISR between 2006 and 2019. e ISR group included first time occurrence of ISR and both drug-eluting stent (DES) and bare metal stent (BMS) ISR. Eligibility criteria for this study were adult patients undergoing percutaneous intervention for FP PAD. Patient demographics, clinical characteristics, and medications were obtained from the review of the electronic medical record. Percutaneous intervention performed was at the operator's discretion. Lesion characteristics, procedural details, and device therapies used were obtained from procedure notes. Six and 12month outcomes after the index procedure were collected during follow-up visits. Management and clinical decisions were at the discretion of the care team.
e study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of participating hospitals. Patients from the prospective cohort were enrolled and consented prior to the index procedure.
e Institutional Review Boards of the participating hospitals waived the requirement for informed consent for the retrospective cohort. Deidentified procedural and angiographic details were independently verified by the Veterans Affairs North Texas Angiography and Ultrasound core laboratory. e clinical coordinating center and data servers are located at the University of Texas Southwestern Medical Center, Dallas, TX. Periodic data audits as well as remote and on-site monitoring were performed.

Definitions and Outcome Measures.
e study population was divided into two groups based on the presence of restenosis in the target stented lesion. ISR was defined as >50% diameter stenosis by contrast angiography or, if available, by >100% increase in the peak systolic velocity over the proximal segment or >400 cm per second via duplex ultrasound. Baseline, target lesion, and procedural characteristics were described for each group. Baseline characteristics included demographics, comorbidities, and Rutherford classification (RC): intermittent claudication (IC; RC 1-3) and critical limb ischemia (CLI; RC 4-6).
Target lesion characteristics included lesion length and presence of severe calcification (at least 5 mm of calcification on both sides of the vessel), chronic total occlusion (CTO), diffuse disease (angiographic disease >30% diameter stenosis for at least 20 mm in length), and number of runoff vessels. Procedural characteristics included type and number of balloons or stents and use of debulking or CTO crossing devices (CCD). Bail out stenting was defined as operator guided stenting of the target lesion when an initial nonstent-based strategy was unsuccessful.
We examined the following outcomes: procedural success, periprocedural complications, and 1-year major adverse limb events (MALE), and major adverse cardiovascular events (MACE). Procedural success was defined as ≤ 30% residual stenosis without complication after index intervention. Periprocedural complications (immediately following the index procedure to 30 days post-procedure) included residual dissection (flow or nonflow limiting), access site hematoma (>5 cm or <5 cm), retroperitoneal hematoma, distal embolization, bleeding diathesis, acute renal failure, perforation, emergency surgery, or their composite. MALE was defined as all-cause death, repeat revascularization (endovascular or surgical), or major amputation. MACE consisted of death, myocardial infarction (MI), or stroke. e individual components of MACE and MALE were also studied.

Statistical Analysis.
To compare baseline and lesion characteristics, procedure details, and procedure outcomes, univariate analysis was used. Continuous variables were described using mean ± standard deviation. Discrete and categorical variables were presented as numbers and percentages. Differences between the FP ISR and FP non-ISR groups were compared using either the t-test or the Wilcoxon rank sum test. Survival curves on the two study groups, with or without FP ISR, were generated by the Kaplan-Meier method and compared using the log-rank test. A two-tailed p value <0.05 was considered statistically significant. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc.) and R version 3.6.1 (the R Foundation for Statistical Computing).
Procedural outcomes are displayed in Table 4. Comparison of 1-year outcomes is shown in Table 5 and Figure 1. Procedural success for FP ISR and FP non-ISR lesions was similar (94.3% vs. 93.0%, p � 0.41). Even though the incidence of composite procedural complications was similar (6.78% vs. 5.97%, p � 0.62), the rate of distal embolization (2.43% vs. 0.94%, p � 0.02) was significantly higher in the ISR group.

Discussion
is analysis from the XLPAD registry demonstrates that patients with FP ISR experience more distal embolization,     were independent predictors of distal embolization [6].
In our analysis, a debulking strategy was more frequently used by operators to treat ISR lesions. However, the use of atherectomy devices may increase the risk for distal embolization [7][8][9]. Mechanical disruption of ISR lesions can generate downstream debris and potential embolic material. However, multiple studies support the use of atherectomy devices for ISR treatment with favorable rates of procedural success and freedom from target lesion revascularization at 1   year [10][11][12][13]. Debulking is a widely accepted approach for treating ISR lesions as reducing smooth muscle proliferation theoretically avoids the need for repeat revascularization and restenting. is combined with distal embolization protection and antirestenotic measures has been described as a successful strategy to treat ISR [14]. Implementation of a debulking strategy is further supported by the larger proportion of 0-1 run-off vessels in the FP ISR group. With poor outflow, there is increased concern that stenting may lead to increased risk of occlusion and stent thrombosis [15].
Our study is unique as it utilizes a core-lab adjudicated angiographic review of registry patients. Fewer lesions treated for FP ISR were severely calcified, CTOs, or with diffuse disease, and there was a larger proportion with 0 to 1 runoff vessels. Known predictors of distal embolization are diffuse and/or long lesions, CTOs, and severe calcification. ese parameters were not observed within the ISR group; however, distal embolization remained significantly elevated. ISR segments may have the propensity for distal embolization due to poor outflow and high utilization of atherectomy devices by operators. We suggest that ISR can be considered as a novel risk factor for distal embolization.
Within our registry, the use of DCBs (4.63% vs. 1.72%, p < 0.001) and DES (10.9% vs. 6.12%, p < 0.001) was significantly greater in the FP ISR group. is is supported by the current literature as DCBs are the most extensively investigated treatment device for ISR. Multiple randomized controlled trials display superior patency rates and freedom from target lesion revascularization at 1 year compared to percutaneous transluminal angioplasty (PTA; [16][17][18][19]). According to the 2017 European Society of Cardiology Guidelines, there is a Class IIb recommendation with Level B evidence for the use of DCBs in the treatment of ISR [20].
DES provide direct delivery of pharmacologic agents to the site of vessel injury to prevent neointimal hyperplasia and smooth muscle cell proliferation reducing the risk of restenosis [21,22]. A limited number of prior studies have demonstrated the safety and effectiveness of DES in the treatment of FP ISR. Zeller et al. demonstrated that paclitaxeleluting stents achieved favorable rates of primary patency (95.7% at 6-months and 78.8% at 1 year) and freedom from target lesion revascularization (96.2% at 6-months, 81.0% at 1 year, and 60.8% at 2 years). Additionally, improvements in ABI and RC were observed at 2-year follow-up (mean ABI preprocedure: 0.60 ± 0.28 vs. mean ABI at 2-years: 0.84 ± 0.22; median RC preprocedure: 3 vs median RC at 2years: 1; [23]). A subgroup analysis comparing the use of paclitaxel-eluting stents to PTA to treat FP ISR exhibited lower rates of restenosis (44.1% vs. 90.3%, p < 0.001) and MALE (25.5% vs. 53.6%, p < 0.00). e use of DES was independently associated with a reduced risk of recurrent restenosis at 1 year (OR 0.2, 95% CI 0.1 to 0.6, p � 0.006; [24]). A postmarket surveillance study on the use of the Zilver PTX stent (Cook Medical, Bloomington, IN) in treating FP ISR suggested a 5-year freedom from TLR of 73.4% [25].
BMS were infrequently used (11.7% vs. 26.0%, p � 0.001) in the ISR group. Current evidence suggests that BMS do not provide superior patency rates at 1 year compared to PTA and carry an increased risk of restenosis [12,26,27]. It would be disadvantageous to use a BMS in a lesion that is already at-risk for further restenosis. Our study provides real-world evidence that operators preferentially use atherectomy and drug-coated devices to address FP ISR lesions.
is is supported by the limited, but existing literature on endovascular strategies for FP PAD and ISR.
In our cohort, patients with FP ISR underwent significantly higher rates of repeat revascularization (21.5% vs. 16.7%, p � 0.04; Figure 1). With Kaplan-Meier analysis, survival from MALE at 12 months did not reach statistical significance (p � 0.17); however, freedom from target limb revascularization did (p � 0.019; Figure 2). e ISR group initially had superior MALE and TLR-free survival, but at 8 months or 240 days, a crossover occurs with the non-ISR group achieving higher survival rates. DCBs and DES were preferentially used in the ISR group. is study suggests that these devices may carry short-term clinical benefits due to the drug-eluting/coated technology; however, it did not translate into long-term patency in our ISR population. is may explain the crossover in the Kaplan-Meier curves, with FP non-ISR lesions having a higher percentage of MALEfree survival and target limb revascularization-free survival beyond 8 months. As MALE was not statistically significant with both the survival model and chi-squared testing, propensity score matching was not performed.
Treatment for ISR remains an active field of investigation. Despite acute procedural success, prior studies demonstrate poor patency rates and high rates of target vessel and lesion revascularization [28,29]. Several factors contribute to the unsatisfactory long-term outcomes. ISR is predominately driven by smooth muscle cell proliferation [30]. Intervention at an existing site of vessel injury creates a local inflammatory response that promotes neointimal hyperplasia [31]. Additionally, FP stents are at further risk of restenosis given the mechanical stressors of elongation, torsion, flexion, and extension that are unique and inherent to this vascular bed [32]. A lesion predisposed to inflammation, injury, and smooth muscle cell hypertrophy is subject to a high likelihood of revascularization after index intervention. Our incidence of revascularization (21.5% for ISR lesions) is similar when compared to prior studies. Listro et al. had a target lesion revascularization rate of 13.6% at 1 year in the DCB group and 31% in the PTA group [33]. Dippel et al. demonstrated a similar incidence of 26.5% in the laser atherectomy with PTA at 6-month follow-up [12].
Despite being performed at experienced centers and the use of novel treatment strategies, revascularization rates remain elevated and the optimal treatment modality for FP ISR is yet to be determined. Promising results from Kokkinidis et al. demonstrated that combination therapy of laser atherectomy with DCB yielded superior rates of freedom from TLR at 1 year (72.5% vs. 50.5%, p � 0.043) and its use was associated with reduced risk of reocclusion (HR 0.08, 95% CI 0.17 to 0.38; p � 0.002) when compared to laser atherectomy with PTA in the treatment of complex FP ISR lesions [34]. Future studies comparing clinical outcomes of patients with FP ISR undergoing endovascular monotherapy and varying combination therapies (laser atherectomy with DES or DCB, excisional atherectomy with DES or DCB) would be of interest.
A higher proportion of FP ISR patients in our study were on GDMT, antiplatelet (94.5% vs. 89.4%, p � 0.007) and statin therapy (68.9% vs. 60.3%, p � 0.003), and female. However, a large proportion of the non-ISR group did not receive optimal medical therapy. It has been demonstrated that a disproportionate number of PAD patients are not prescribed GDMT. In the Reduction of Atherothrombosis for Continued Health registry, out of 8273 PAD patients, only 70% received lipid-lowering therapy and only 82% received antiplatelet therapy [35]. Factors that may contribute to under-prescription include African-American race, male gender, care by nonvascular specialists, and claudicants with normal ABI's [36]. Female gender has been identified as an independent risk factor for the development of ISR and is associated with a higher prevalence of CLI, poor stent patency, and frequent restenosis in TASC C/D lesions [37]. Further research and efforts are needed to bridge the disparity observed in PAD population and specifically, female patients with PAD and restenosis.
ere are important limitations to note. Our study is a retrospective observational study and investigators are responsible for enrolling patients undergoing endovascular treatment for symptomatic PAD. erefore, it is subject to recall and selection bias and cannot account for unmeasured confounders. Conclusions must be considered as hypothesis-generating and do not determine causation. Data on the use of distal embolic filters or rate of stent fractures are not recorded and are not reflected in our analysis of periprocedural and 1-year outcomes.

Conclusion
Our large, multicenter registry demonstrates that atherectomy and drug-coated devices were more frequently used to treat FP ISR lesions. Compared to patients without ISR, patients with FP ISR experience significantly higher revascularization procedures and lower freedom from MALE at 1 year. Further studies evaluating more refined strategies to address FP PAD and ISR are needed to improve long-term patency and freedom from revascularization.

Data Availability
Due to the nature of this research, participants of this study did not agree for their data to be shared publicly, so supporting data are not available. e data are stored within the XLPAD registry where IRB approval is required for researchers and clinicians to gain access to registry for data analysis and entry.

Conflicts of Interest
Mitul Patel, MD, has received consulting honoraria from Abbott Vascular, Medtronic, Cardiovascular Systems, Inc., and Chiesi. Research/Institutional grant support was obtained from Abiomed and Boston Scientific; Chris Metzger, MD, is nominal symposia/proctor at Abbott Vascular, Shockwave, Cardiovascular Systems, Inc., Penumbra; Mehdi H. Shishehbor, DO, MPH, PhD, has received Advisory Board/Consulting fees from Medtronic, Abbott Vascular, Phillips, Terumo, and Boston Scientific; Emmanouil S. Brilakis, MD, has received consulting/speaker honoraria from Abbott Vascular, American Heart Association (associate editor circulation), Amgen, Biotronik, Boston Scientific, Cardiovascular Innovations Foundation (Board of Directors), ControlRad, CSI, Ebix, Elsevier, GE Healthcare, InfraRedx, Medtronic, Siemens, and Teleflex; research support was obtained from Regeneron and Siemens. Owner, Hippocrates LLC, shareholder: MHI Ventures; Nicolas W. Shammas, MD, has received consulting fees from Intact Vascular, Bards, and Boston Scientific. Research and Institutional Grants were obtained from Intact Vascular, Bard, Boston Scientific, Phillips; Peter Monteleone, MD, has received consulting fees from Medtronic and Biotronik. Institutional grant support was obtained from Medtronic and Abbott; Subhash Banerjee is consultant for Medtronic, Cordis, Kaneka, and Institutional research grants were obtained from Boston Scientific Corporation and Chiesi. e rest of the authors declare that there are no conflicts of interest regarding publication of this paper.