Author + information
- Received September 5, 2014
- Accepted September 29, 2014
- Published online January 1, 2015.
- Eric J. Dippel, MD∗∗ (, )
- Prakash Makam, MD†,
- Richard Kovach, MD‡,
- Jon C. George, MD‡,
- Raghotham Patlola, MD§,
- D. Christopher Metzger, MD‖,
- Carlos Mena-Hurtado, MD¶,
- Robert Beasley, MD#,
- Peter Soukas, MD∗∗,
- Pedro J. Colon-Hernandez, MD††,
- Matthew A. Stark, PhD‡‡,
- Craig Walker, MD§§,
- EXCITE ISR Investigators
- ∗Genesis Medical Center, Davenport, Iowa
- †Cardiology Associates of Northwest Indiana, Munster, Indiana
- ‡Deborah Heart and Lung Center, Browns Mills, New Jersey
- §Cardiovascular Clinic of Hattiesburg, Hattiesburg, Mississippi
- ‖Wellmont Holston Valley, Kingsport, Tennessee
- ¶Section of Cardiovascular Medicine, Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut
- #Division of Radiology, Mount Sinai Heart Institute, Miami Beach, Florida
- ∗∗The Miriam Hospital, Providence, Rhode Island
- ††Centro Cardiovascular de Puerto Rico y el Caribe, San Juan, Puerto Rico
- ‡‡The Spectranetics Corporation, Colorado Springs, Colorado
- §§Cardiovascular Institute of the South, Houma, Louisiana
- ↵∗Reprint requests and correspondence:
Dr. Eric J. Dippel, Cardiovascular Medicine, PC, Genesis Medical Center, 1236 East Rusholme Street, Davenport, Iowa 52803.
Objectives The purpose of this study was to evaluate the safety and efficacy of excimer laser atherectomy (ELA) with adjunctive percutaneous transluminal angioplasty (PTA) versus PTA alone for treating patients with chronic peripheral artery disease with femoropopliteal bare nitinol in-stent restenosis (ISR).
Background Femoropopliteal stenting has shown superiority to PTA for lifestyle-limiting claudication and critical limb ischemia, although treating post-stenting artery reobstruction, or ISR, remains challenging.
Methods The multicenter, prospective, randomized, controlled EXCITE ISR (EXCImer Laser Randomized Controlled Study for Treatment of FemoropopliTEal In-Stent Restenosis) trial was conducted across 40 U.S. centers. Patients with Rutherford Class 1 to 4 and lesions of target lesion length ≥4 cm, vessel diameter 5 to 7 mm were enrolled and randomly divided into ELA + PTA and PTA groups by a 2:1 ratio. The primary efficacy endpoint was target lesion revascularization (TLR) at 6-month follow up. The primary safety endpoint was major adverse event (death, amputation, or TLR) at 30 days post-procedure.
Results Study enrollment was stopped at 250 patients due to early efficacy demonstrated at a prospectively-specified interim analysis. A total of 169 ELA + PTA subjects (62.7% male; mean age 68.5 ± 9.8 years) and 81 PTA patients (61.7% male; mean age 67.8 ± 10.3 years) were enrolled. Mean lesion length was 19.6 ± 12.0 cm versus 19.3 ± 11.9 cm, and 30.5% versus 36.8% of patients exhibited total occlusion. ELA + PTA subjects demonstrated superior procedural success (93.5% vs. 82.7%; p = 0.01) with significantly fewer procedural complications. ELA + PTA and PTA subject 6-month freedom from TLR was 73.5% versus 51.8% (p < 0.005), and 30-day major adverse event rates were 5.8% versus 20.5% (p < 0.001), respectively. ELA + PTA was associated with a 52% reduction in TLR (hazard ratio: 0.48; 95% confidence interval: 0.31 to 0.74).
Conclusions The EXCITE ISR trial is the first large, prospective, randomized study to demonstrate superiority of ELA + PTA versus PTA alone for treating femoropopliteal ISR. (Randomized Study of Laser and Balloon Angioplasty Versus Balloon Angioplasty to Treat Peripheral In-stent Restenosis [EXCITE ISR]; NCT01330628)
- femoropopliteal artery
- excimer laser atherectomy
- in-stent restenosis
- percutaneous transluminal angioplasty
- peripheral artery disease
- randomized controlled trial
Advancements in nitinol stents have revolutionized femoropopliteal treatment in patients with chronic peripheral arterial disease (PAD). The effectiveness of these methods is, in part, due to the ability of nitinol stents to effectively address historically problematic issues with conventional percutaneous transluminal angioplasty (PTA), including well-documented limitations such as elastic recoil, residual stenosis, and flow-limiting dissection (1–7). Despite notable improvements in stenting devices and techniques in the past decade, up to one-half of all patients that receive femoropopliteal stents will require secondary interventions (>115,000 procedures annually) due to in-stent restenosis (ISR), a form of arterial recurrent obstruction particularly common in long and complex lesions (8,9). As a result, there remains an unmet need for more effective ways to treat ISR.
Conventional treatment strategies for ISR include PTA with or without repeat stenting (bare-metal stents, stent grafts, or drug-eluting stents), drug-eluting or cutting balloon angioplasty, cryoplasty, and directional or laser atherectomy (10–18). The recent proliferation of treatment options has, however, complicated clinical treatment of ISR, resulting in a variety of reports with highly variable results that may be related to disparities in lesion crossing and repeat restenosis risk. Effectiveness of ISR treatments can be gauged by examining procedural success, complications, and post-treatment stenosis as well as the occurrence of major adverse events (MAEs) and target lesion revascularization (TLR) in the months following surgery.
Excimer laser atherectomy (ELA) allows for photoablation of residual hyperplasic tissues. A prospective, multicenter study of 90 patients with ISR of nitinol stents implanted in the femoropopliteal arteries conducted by Schmidt et al. (16) in 2014 demonstrated that ELA combined with adjunctive PTA resulted in 97% procedural success, 7% residual stenosis, 2.2% 30-day major complication rate, and 64% 1-year freedom from TLR. These findings suggest that ELA with adjunctive PTA may be more efficacious than PTA alone for treatment of femoropopliteal ISR, although no prior large, randomized studies have been completed. The current prospective, randomized, controlled trial was conducted for U.S. Food and Drug Administration submission to evaluate the safety and efficacy of ELA with adjunctive PTA versus PTA alone for the treatment of femoropopliteal ISR.
The EXCITE ISR (EXCImer Laser Randomized Controlled Study for Treatment of FemoropopliTEal In-Stent Restenosis) trial is a multicenter, prospective, randomized, controlled trial conducted across 40 centers in the United States with enrollment from June 2011 to February 2014. The expected enrollment was up to 318 chronic PAD patients with successful intraluminal guidewire crossing and 35 chronic PAD patients with unsuccessful intraluminal guidewire crossing, reported as a separate total occlusion registry component (not included in this report). The trial was conducted under an Investigational Device Exemption approved by the U.S. Food and Drug Administration. The study was approved by the institutional review board of each institution, and written informed consent was obtained from each patient prior to enrollment. This clinical trial was registered at www.clinicaltrials.gov (NCT01330628). A full list of the EXCITE ISR investigators can be found in the Online Appendix.
Patient selection and randomization
The study included patients age 18 to 85 years, with Rutherford Class 1 to 4 PAD, and with resting or abnormal exercise ankle-brachial index <0.9 in the target limb. Angiographic inclusion criteria included significant ISR ≥50% diameter stenosis (DS) by visual assessment within previously deployed femoropopliteal stent(s), target lesion length ≥4 cm, and reference vessel diameter between 5 and 7 mm. A minimum of 1 patent tibial vessel or peroneal artery to the foot containing <50% DS by visual estimate that did not require intervention prior to the 1-month follow-up was required. Additional anatomic inclusion criteria included: patent inflow vessel prior to study enrollment with or without treatment and peristent lesion not to exceed 3 cm. Exclusion criteria included pregnancy, acute limb ischemia, life expectancy <12 months, cerebrovascular accidents or myocardial infarction within 60 days prior to procedure, contraindications or allergies that could affect the procedure, uncontrolled hypercoagulability, systemic infection in target limb, previous treatment to the target vessel within 3 months prior to study procedure, serum creatinine ≥2.5 mg/dl unless dialysis-dependent, aneurysm within target lesion, drug-eluting stents or covered stents in the target lesion, planned or predicted cardiac surgery or interventions prior to completion of 30-day follow-up, or grade 4/5 stent fracture affecting target stent or proximal to the target stent.
Patients meeting the criteria were consecutively enrolled and randomly allocated into 2 groups, ELA with adjunctive PTA (ELA + PTA group) or PTA alone (PTA group) using a 2:1 allocation blocked by site, using random block sizes of 3 and 6.
Subjects were admitted and prepared according to the standard clinical practice at each participating institution, including necessary administrations of oral antiplatelet (antithrombotic) therapy. Pre-procedural and periprocedural diagnostic and guidance imaging was performed to determine eligibility. Vessel calcium was graded angiographically using the following criteria: none/mild = no radio-opacities; moderate = radio-opacities appear unilaterally on arterial wall or <1 cm (length) prior to contrast injection or digital subtraction angiography; and severe = radio-opacities appear bilaterally on arterial wall and extend >1 cm (length) prior to contrast injection or digital subtraction angiography. Stent fracture was assessed via straight-leg x-ray or fluoroscopy in 2 orthogonal views prior to intervention. Stent fractures were graded by severity score according to published criteria (19).
Intraluminal crossing with a guidewire was performed from a contralateral crossover approach or via antegrade femoral access utilizing standard endovascular techniques. Distal embolic protection was employed according to operator discretion. For ELA + PTA subjects, a pilot channel was created (if necessary) using a Turbo Elite laser catheter (Spectranetics Corp., Colorado Springs, Colorado). A Turbo Tandem laser catheter (Spectranetics Corp.) was then used to achieve maximum debulking, generally applied using 4 quadrant passes and starting parameters: Turbo Elite 45 Hz repetition rate and 25 mJ/mm2 fluence and Turbo Tandem 60 Hz repetition rate and 40 mJ/mm2 fluence. PTA was performed between nominal and rated burst pressures over the course of 2 to 5 min. The use of specialty balloons or other atherectomy devices was not permitted.
Patients underwent clinical follow-up at discharge and clinical follow-up with duplex ultrasound (DUS) at 30 days, 6 months, and 12 months. Stent integrity was investigated with straight-leg flat plate x-rays or fluoroscopy at 6 and 12 months.
The primary efficacy endpoint was freedom from TLR through 6-month follow-up. Secondary endpoints included acute procedural success, TLR through 12-month follow-up, DUS-assessed patency, ankle-brachial index, change in Rutherford score, and stent integrity.
Procedural success was defined as ≤30% residual diameter stenosis following assigned treatment without provisional or bailout procedures. Patency was defined as <50% DS as assessed by DUS in the absence of TLR, amputation, and/or surgical bypass. Peak systolic velocity ratio ≤2.5 was used as the patency threshold. Clinically-driven TLR was defined as a reintervention performed for >50% diameter stenosis or in the target lesion after documentation of recurrent clinical symptoms following the index procedure or bailout stenting during the index procedure.
The primary safety endpoint was MAE occurrence through 30-day follow-up (37 days post-procedure) defined as all-cause death, unplanned major amputation, or TLR. All adverse events (AEs), regardless of severity or relationship to the procedure, were recorded. AEs were classified by seriousness, severity, and relationship to procedure.
Data were recorded on case report forms using an electronic data capture. Trial oversight and adverse event adjudication was provided by a Data Safety Monitoring Board. Independent core laboratories analyzed angiographic (Synvacore, Springfield, Illinois) and DUS studies (VasCore, Boston, Massachusetts). Patients and readers at the DUS and angiographic imaging core laboratories were blinded to treatment assignment. Data management and monitoring was conducted by an independent contract research organization (Medpace, Cincinnati, Ohio).
A maximum sample size of 318 patients was determined to be required for this study, assuming 1-sided α = 0.025, β = 80%, assumed 70% and 53% freedom from TLR at 6 months in the ELA + PTA and PTA groups, respectively (16,20), and 10% attrition. A Bayesian hierarchical model was employed to prospectively assess results at 200, 250, and 300 patients, with potential early termination prior to enrolling the maximum subject population if the primary endpoint was achieved with sufficient statistical power (prospective stopping rules).
Continuous variables were summarized as mean ± SD, and categorical variables were reported as counts and percentages. Between-group comparisons were performed using the 2-sample t test for continuous variables and Fisher exact test for categorical variables. Primary efficacy and safety endpoints were evaluated using the chi-square test. Freedom from primary patency, TLR, survival, and amputation were analyzed using the product-limit Kaplan-Meier and the Cox proportional hazards regression model. A stepwise backward elimination technique was used to determine independent predictors of target lesion revascularization. Variables were entered into the model at p values <0.10, and stepwise backward elimination multivariate analysis was conducted. A p value <0.05 was considered statistically significant. Cox proportional hazards models were used to estimate the hazard ratio for the randomized comparison in subgroups for dichotomous baseline covariates or at a specified level for continuous baseline covariates. The Cox proportional hazards models included all patients with independent variables at baseline covariate, the randomized group, and the interaction term between baseline covariate and randomized group. Statistical analysis was completed using SAS version 9.4 (SAS Institute, Inc., Cary, North Carolina).
A total of 250 subjects were randomized to ELA + PTA (n = 169) or PTA (n = 81). Seven patients with total occlusions uncrossable by guidewire were enrolled in a separate registry. Per the prospective Bayesian hierarchical model, assessments were performed when enrollment reached 200 and 250 patients. As a result, the study stopping criteria were met when the analysis was complete for the initial 250 subjects demonstrating early success for the primary safety and efficacy endpoints. Patient follow-up status at the interim analysis is shown in Figure 1. At time of data lock, less than one-third of available patients completed the 12-month follow-up; therefore, long-term results will be reported separately.
Baseline patient characteristics
Baseline patient characteristics were not significantly different between the ELA + PTA and PTA groups. Patients were predominantly elderly men with claudication and presented with hypertension, hyperlipidemia, and smoking history as the most common risk factors. The percentage of patients previously treated for ISR was not different between groups (32.7% vs. 30.0%) (Table 1). Mean lesion length was 19.6 ± 12 cm and 19.3 ± 12 cm in the ELA + PTA vs. PTA alone groups, respectively. The majority of lesions were classified as TransAtlantic Inter-Society Consensus C/D, and more than 70% of lesions were >10 cm in length. The ELA + PTA group presented with significantly higher frequency of minor stent fractures and with more extensive calcification at baseline (Table 2).
Procedural success was 93.5% in the ELA + PTA group and 82.7% in the PTA group (p = 0.01). Dissection (7.7% vs. 17.2%; p = 0.03), bailout stenting (4.1% vs. 11.1%; p = 0.05) and procedural TLR (5.3% vs. 16.0%; p = 0.008) occurred significantly less frequently in the ELA + PTA group compared with the PTA alone group. Distal protection use and occurrence of embolism were nonsignificantly higher in the ELA + PTA group (Table 3).
At 6-month follow-up, 117 of ELA + PTA and 56 of PTA patients were available for use in the primary efficacy endpoint analysis population. Freedom from TLR was 73.5% versus 51.8%, demonstrating that ELA + PTA was superior to PTA alone (p < 0.005) (Figure 2). ELA + PTA was also superior to PTA alone if bailout stenting was not considered a TLR for the primary efficacy endpoint analysis, 78.1% versus 61.7% (p < 0.05) (Figure 3). ELA + PTA maintained superiority in freedom from TLR (p < 0.003) and patency (p < 0.005) throughout the available follow-up period (Figures 3 and 4⇓). ELA + PTA treatment was also superior to PTA alone for several patient and lesion groups analyzed by Cox proportional hazards (Figure 5). Clinical status at 6 months was similar among both groups; however, patients in the ELA + PTA group demonstrated significantly better improvement or maintenance of Rutherford class (p = 0.008) (Table 4).
At 30-day follow up, 155 of ELA + PTA and 73 of PTA patients were available for use in the safety analysis. Freedom from any MAE was 94.2% in the ELA + PTA group and 79.5% in the PTA group (p < 0.001), demonstrating that ELA + PTA was superior to PTA alone for the primary safety endpoint. The primary safety endpoint was primarily driven by TLR; however, 1 PTA patient died within 30 days due to hypotensive shock and respiratory failure.
The ELA + PTA group maintained superiority in freedom from an MAE throughout the study period (Figure 6). TLR was the most common MAE in both groups. The incidence of major amputation and mortality through 6 months was low in both groups (Table 4). There was no change in stent fracture grade in any patient during the study.
Analysis of TLR risk factors
Univariate and multivariate predictors of TLR are presented in Table 5. Previous treatment for ISR in the target limb (hazard ratio [HR]: 1.64, 95% confidence interval [CI]: 1.06 to 2.54, p = 0.03) and an increased lesion length (HR: 1.04, 95% CI: 1.02 to 1.05, p < 0.001) were found to increase the occurrence of TLR, whereas increased age (HR: 0.77, 95% CI: 0.61 to 0.96, p = 0.02), increased reference vessel diameter (HR: 0.71, 95% CI: 0.53 to 0.95, p = 0.02) and treatment with ELA + PTA (HR: 0.48, 95% CI: 0.31 to 0.74, p = 0.001) were found to decrease the occurrence of TLR. Lesion length was the only significant interaction term identified. ELA + PTA performed proportionately better compared with PTA alone in longer lesions (Figure 7).
As the first randomized, controlled atherectomy study of its kind, the EXCITE ISR trial showed that ELA + PTA reliably achieved superior safety and efficacy compared with PTA alone. Overall, this study demonstrated that the benefit-risk profile of ELA with adjunctive PTA is favorable in the treatment of femoropopliteal ISR, with statistically significant reductions in procedural complications, MAEs, and TLR compared with PTA alone. Freedom from TLR remained significantly better in the ELA + PTA group compared with PTA alone when bailout stenting was not considered a procedural TLR. Furthermore, there is no evidence of late convergence in either the MAE or TLR survival curves, suggesting that there is sustained benefit through 12 months.
Excimer laser debulking may be advantageous for ISR patients because of the ability of this technique to effectively remove hyperplastic tissue (21). The predicate device to the Turbo Tandem laser catheter showed significant plaque removal and increased vessel compliance in de novo lesions of the femoropopliteal artery (22). However, the use of self-expanding nitinol stents limits positive remodeling that would otherwise occur in an unconstrained vessel or in an underexpanded balloon-expandable stent used in the coronary arteries. Within the stent lumen, neointimal hyperplasia and extracellular matrix are primary contributors to ISR occurrence and severity (23). ELA treatment has previously demonstrated ablation efficacy reducing baseline ISR lesion stenosis by over 60% (16). The hypothesis that ELA ISR debulking with adjunctive PTA could reduce TLR was confirmed in this study.
A covariate analysis for common factors considered to have an effect on TLR, such as occlusion, diabetes, sex, and smoking, showed that lesion length, reference vessel diameter, age, and previous history of ISR were predictors of TLR. However, lesion length was found to be the only significant interaction term. Average lesion length in historical ISR reports (PTA only) ranges from 7 to 20 cm (10,18,24,25). Collectively, lesions in this series were complex, averaging 20 cm in length with over 20% of lesions >30 cm in length and more than 30% occlusion. A large retrospective study on ISR treatment with PTA reported freedom from TLR at 12 months to be 69% in stenoses (average length 9 cm) and 23% in long occlusions (25). In this study, the 6-month freedom from TLR was 53% in the PTA only group, similar to previous studies in long complex ISR lesions. Overall, ELA + PTA reduced the occurrence of TLR by over 50% compared with PTA alone.
Primary patency remained significantly better in the ELA + PTA group as compared with PTA alone throughout the study. Reintervention occurred less frequently than would be expected by 6-month DUS examination in both groups. Per protocol, the decision to undergo reintervention was driven by clinical symptoms in addition to DUS binary restenosis. Indeed, recent studies suggest that DUS alone may not accurately reflect ISR stenosis and that the common threshold for binary restenosis, peak systolic velocity ratio ≤2.5, may be inappropriate (26,27). Clinical surveillance of ISR patients is recommended to determine the need for reintervention.
Procedural complications were low in both groups. ELA + PTA treatment demonstrated significantly fewer procedural major dissections, residual stenosis >30%, or need for bailout stenting. There were no new or deteriorated stent fractures due to laser–stent interaction. Clinically-relevant embolism occurred at rates comparable to previous studies in complex lesions (16,28–30). All events were treated during the index procedure and resolved without complication.
Paclitaxel drug-coated balloons (DCBs) have shown superior prevention of restenosis in de novo and ISR lesions of the femoropopliteal artery compared with PTA alone (31–34). However, DCB application suffers from the same procedural complications as PTA, dissection and residual stenosis necessitating bailout stenting, occurring in up to 20% of procedures in de novo and 16% of ISR lesions (24,31,34). As reported in this study, ELA treatment provides a significant procedural advantage to PTA alone. Therefore, vessel preparation with atherectomy prior to balloon-dependent drug delivery, especially in the context of peripheral ISR, may provide interesting research avenues to reduce procedural complications and improve drug delivery for long-term results (35–37). With the eagerly-anticipated approval of DCB in the United States, larger randomized controlled trials are needed to test the potential clinical benefits of these novel combinatorial therapies.
The study originally planned to enroll a maximum of 318 randomized patients. Although the stopping criteria were reached at 250 patients, enabling statistically-powered assessments of both primary efficacy and safety endpoints, this study may be underpowered for examining other secondary variables. The estimated HRs for subgroups may be prone to inflated type I error as these analyses were not pre-specified. Conclusive statements would require validation in a prospectively-designed study to address the specific questions. Finally, there was limited follow-up at 12 months at the time of data analysis. Long-term data will be reported when available.
The EXCITE ISR trial is a first-of-its-kind large, prospective, randomized atherectomy study. ELA with adjunctive balloon angioplasty results in significantly better acute and midterm efficacy and safety outcomes for treatment of peripheral femoropopliteal ISR compared with conventional PTA alone in all lesion types examined in this research.
The authors thank Larry Miller, PhD, for providing critical review and editorial assistance; Teresa Nelson, MS, for statistical modeling assistance; and Christopher DeMorett and Hailey Austin, MS, for technical assistance.
The EXCITE ISR trial was supported by the Spectranetics Corporation. Dr. Dippel is a consultant for and shareholder in the Spectranetics Corporation. Drs. Beasley, George, Kovach and Walker are consultants for the Spectranetics Corporation. Dr. Metzger has received symposium honoraria from the Spectranetics Corporation, Bard, Abbott Vascular, and Boston Scientific; has received payment for teaching courses from Abbott Vascular; and was a proctor for courses with Medtronic. Dr. Stark is an employee of the Spectranetics Corporation. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- diameter stenosis
- duplex ultrasound
- in-stent restenosis
- major adverse event
- peripheral artery disease
- percutaneous transluminal angioplasty
- target lesion revascularization
- Received September 5, 2014.
- Accepted September 29, 2014.
- 2015 American College of Cardiology Foundation
- Krankenberg H.,
- Schlüter M.,
- Steinkamp H.J.,
- et al.
- Schillinger M.,
- Sabeti S.,
- Dick P.,
- et al.
- Rocha-Singh K.J.,
- Jaff M.R.,
- Crabtree T.R.,
- Bloch D.A.,
- Ansel G.,
- VIVA Physicians, Inc
- Armstrong E.J.,
- Saeed H.,
- Alvandi B.,
- et al.
- Laird J.R. Jr..,
- Yeo K.K.,
- Rocha-Singh K.,
- et al.
- Zeller T.,
- Dake M.D.,
- Tepe G.,
- et al.
- Virga V.,
- Stabile E.,
- Biamino G.,
- et al.
- Zeller T.,
- Rastan A.,
- Sixt S.,
- et al.
- Laird J.R.,
- Katzen B.T.,
- Scheinert D.,
- et al.
- Mehran R.,
- Mintz G.S.,
- Satler L.F.,
- et al.
- Liistro F.,
- Angioli P.,
- Porto I.,
- et al.
- Tosaka A.,
- Soga Y.,
- Iida O.,
- et al.
- Werk M.,
- Albrecht T.,
- Meyer D.-R.,
- et al.
- Scheinert D.,
- Duda S.,
- Zeller T.,
- et al.
- Stabile E.,
- Virga V.,
- Salemme L.,
- et al.
- Gandini R.,
- Del Giudice C.,
- Merolla S.,
- Morosetti D.,
- Pampana E.,
- Simonetti G.