Endovascular treatment of symptomatic atherosclerotic peripheral artery disease (PAD) has gained widespread acceptance and is now recommended as the primary revascularization strategy in many clinical and anatomic scenarios. 1–3 Percutaneous transluminal angioplasty (PTA) of the superficial femoral and popliteal artery has a high initial success rate, but restenosis occurs in up to 60% of cases. 4 Although randomized trials have demonstrated patency rates with bare metal stents and drug-eluting stents superior to those observed with PTA, 5–8 the optimal treatment for superficial femoral and popliteal artery disease remains controversial. Some practice guidelines advise against primary stenting in patients with intermittent claudication, 9 whereas others recommend primary stenting in short- or intermediate-length lesions 3 or in the event of acute PTA failure. 1–2 Despite the improved outcomes reported in some trials with stenting, the dynamic stresses applied by the superficial femoral and popliteal artery may result in stent fracture 10–11 or in-stent restenosis. 12 Given the limitations of stenting, there has been considerable interest in identifying the approaches that could improve patency without the need for a permanent metallic implant. Clinical Perspective on p 502 One approach to this challenge has been the development of the drug-coated balloon (DCB), which combines balloon dilatation with local delivery of an antiproliferative drug. Proof-of-concept evidence has demonstrated the utility of different DCB technologies in reducing both restenosis and the need for reintervention in comparison with PTA. 13–17 Promising primary patency and target lesion revascularization rates up to 2 years postimplantation have been reported. 18 However, robust evidence from large randomized, controlled trials is lacking. The IN.PACT SFA Trial was designed to test the safety and efficacy of the IN.PACT Admiral DCB for the treatment of patients with symptomatic PAD in the superficial femoral and proximal popliteal artery. Methods Study Design The IN.PACT SFA Trial is a multicenter, international, single-blinded, randomized, controlled trial to assess the safety and efficacy of the IN.PACT Admiral DCB (Medtronic Inc, Santa Rosa, CA) versus standard PTA balloons in patients with symptomatic superficial femoral and proximal popliteal artery disease. The trial was prospectively designed to be conducted in 2 phases: IN.PACT SFA I (conducted in Europe) and IN.PACT SFA II (conducted in the United States), which are jointly referred to as IN.PACT SFA. The IN.PACT SFA Trial was prospectively analyzed according to a single statistical analysis plan. The 2 phases occurred sequentially in time with enrollment completed in the IN.PACT SFA I phase before the initiation of the IN.PACT SFA II phase. Minor differences between the IN.PACT SFA I phase and the IN.PACT SFA II phase eligibility criteria exist and include subtle variations in concomitant inflow and contralateral limb treatment, along with differences in predilatation requirements. A prespecified poolability test for treatment-by-trial phase interaction was established, with planned data pooling across the 2 phases in the event that there was no significant treatment-by-trial interaction. Both protocols were approved by the institutional review boards or ethics committees at each trial site. All patients provided written informed consent before enrollment. Both trial phases were conducted in accordance with the Declaration of Helsinki, good clinical practice guidelines, and applicable laws as specified by all relevant governmental bodies. An independent clinical events committee adjudicated all major adverse events. Independent core laboratories analyzed all images, including duplex ultrasonography (VasCore, Massachusetts General Hospital, Boston, MA) and angiography (SynvaCor, Springfield, IL). Patient Population Patients were eligible for enrollment if they had moderate to severe intermittent claudication or ischemic rest pain (Rutherford 2–4) and stenosis of 70% to 99% with lesion lengths between 4 and 18 cm or occlusion with lengths of ≤10 cm involving the superficial femoral and proximal popliteal arteries, and met all other eligibility criteria. Randomization and Blinding Randomization occurred after successful crossing of the lesion in the IN.PACT SFA I phase and after successful crossing and predilatation with a standard PTA balloon 1 mm smaller than the reference vessel diameter in the IN.PACT SFA II phase. A patient was considered enrolled at the time of randomization. Subjects were randomly assigned by an Interactive Voice Response System with the use of a method of permuted blocks to ensure that a 2:1 ratio was maintained across sites (Figure 1). Figure 1. Trial flow diagram. The IN.PACT SFA Trial used a 2:1 randomized, control design, and intent-to-treat (ITT) analysis was conducted at 12 months. Three hundred thirty-one (331) patients with de novo or nonstented restenotic lesions in the superficial femoral and proximal popliteal artery were randomly assigned either to the IN.PACT Admiral drug-coated balloon or standard PTA treatment group. All subjects enrolled in the IN.PACT SFA Trial (n=331) will be followed for up to 5 years. Analysis at 1 year included subjects that provided end point data at the time of data snapshot. A subject was excluded under the following circumstances: (1) consent was withdrawn before the 1-year visit and no event had occurred before withdrawal or (2) there was no contact with the subject permitting a 1-year evaluation and no events had occurred before the 1-year evaluation. DCB indicates drug-coated balloon; PTA, percutaneous transluminal angioplasty; and SFA, superficial femoral artery. The patients and the trial sponsor were blinded to the treatment assignments through the completion of all 12-month follow-up evaluations. The independent core laboratories and clinical events committee will remain blinded to the treatment assignments throughout the 60-month follow-up duration. Because of the visual difference between the IN.PACT DCB and standard PTA balloon, treating physicians, research coordinators, and catheterization laboratory staff were not blinded to the treatment assignment. Treating physicians, research coordinators, and catheterization laboratory staff received detailed and specific instructions and training on how to preserve the patients’ blinded status. Treatment and Medical Therapy Patients randomly assigned to the experimental arm were treated with the IN.PACT Admiral DCB. The IN.PACT DCB has a dual mode of action, comprising mechanical dilatation by the angioplasty balloon plus local drug delivery to the arterial wall intended to inhibit restenosis. The IN.PACT DCB coating includes paclitaxel as the antiproliferative agent at a dose of 3.5 μg/mm2, with urea as the excipient. Available IN.PACT Admiral DCB sizes included 4-, 5-, 6-, and 7-mm diameters and 20-, 40-, 60-, 80 and 120-mm lengths (the 7-mm diameter device was not available in the 120-mm length). A minimum balloon inflation time of 180 seconds was required for both treatment groups. To avoid geographic miss, DCB length was chosen to exceed the target lesion length by 10 mm at the proximal and distal edges. The IN.PACT DCB is a single-inflation device, and, when treatment required multiple balloons, an overlap of 10 mm was applied for contiguous balloon inflations. Premedication included a loading dose of aspirin 300 to 325 mg and clopidogrel 300 mg within 24 hours of the index procedure or 2 hours postprocedure. Heparin was administered at the time of the procedure to maintain an activated clotting time ≥250 seconds. Postdilatation with a standard PTA balloon was allowed at the discretion of the operator. In both treatment groups, provisional stenting was allowed only in the case of PTA failure after repeated and prolonged PTA inflations. PTA failure was defined as a residual stenosis ≥50% or major (≥grade D) flow-limiting dissection confirmed by a peak translesional systolic pressure gradient of >10 mm Hg. In both arms, postprocedure medical therapy included aspirin 81 to 325 mg daily (for a minimum of 6 months) and clopidogrel 75 mg daily for a minimum duration of 1 month for nonstented patients and 3 months for patients who received stents. Usage of aspirin and antiplatelet drugs did not differ between treatment arms at discharge (97.6%), 30 days (87.6%), or 12 months (51.5%). Follow-Up For the primary end point analysis, patients were followed by the treating physician at 30 days, 6 months, and 12 months, including office visits with duplex ultrasonography functional testing and adverse event assessment. Reinterventions, if required within 12 months of the procedure, were performed according to standard practice by using PTA balloons and provisional stenting. Study End Points The primary efficacy end point was primary patency at 12 months following the index procedure, defined as freedom from clinically driven target lesion revascularization and restenosis as determined by a duplex ultrasonography–derived peak systolic velocity ratio of ≤2.4. 19 Each component of the primary efficacy end point was independently adjudicated by the blinded Clinical Events Committee (for clinically driven target lesion revascularization) or by the core laboratories (for restenosis). Specifically, the independent Clinical Events Committee determined whether reinterventions at the target lesion were clinically driven on the basis of objective testing (ankle-brachial index decrease ≥20% or >0.15 in comparison with postprocedure ankle-brachial index) or symptoms of exertional limb discomfort. Safety end points included 30-day device- and procedure-related death, all-cause death, major target limb amputation, and target vessel thrombosis. These events were site reported and Clinical Events Committee adjudicated. Additional efficacy end points included acute procedural success, target vessel revascularization at 12 months, and primary sustained clinical improvement (defined as freedom from target limb amputation, target vessel revascularization, and increase in Rutherford class at 12 months). Functional assessments included general appraisal through administration of a 5-dimension (EQ-5D) health-related quality-of-life questionnaire and specific evaluation of walking capacity by using a Walking Impairment Questionnaire. A Six-Minute Walk Test was additionally conducted in the IN.PACT SFA II phase only. Statistical Analysis The planned enrollment of 330 subjects provided a power of 80% to detect a 50% improvement in the primary end point at 12 months (from 40% 4 in the PTA group to 60% in the DCB group) with a 1-sided type I error of 2.5%. From its inception, the trial was intended to have 2 phases under a single statistical analysis plan. Poolability of subjects across trial phases for the primary end point analysis was tested by using Cox proportional hazards regression. For this poolability analysis, model covariates included treatment group, phase, and a treatment-by-phase interaction effect. Because the treatment-by-trial phase interaction value for the primary end point was nonsignificant (P=0.428), the 2 trial phases were pooled for all analyses. All analyses were based on the intention-to-treat principle. Continuous variables are described as mean±standard deviation and were compared by t tests. Categorical variables are described as proportions and were compared by the Z test owing to the 1-sided testing. The Z test was used to test the hypothesis of equality of proportions in achieving the primary end point. Multiple imputation was performed by using the logistic regression approach for patients with missing primary end point data (29 DCB, 7 PTA). The following variables were included in the imputation model as covariates: age, sex, diabetes mellitus, lesion length, total occlusion, and Rutherford class at baseline. Five data sets were imputed from these covariates that mimic different realizations of the missing data. Within each imputed data set for the end point, the proportion experiencing the end point was statistically compared between treatment groups by using the 2-sample Z test. From these, an overall test statistic for the end point and its associated P value were calculated for the imputed data. The imputed difference (95% confidence interval) and P value are reported along with the as-observed numerator and denominator. A sensitivity analysis of the as-observed rates revealed a similar highly significant P value (P 70% at 1 to 2 years, although recent results have suggested improved outcomes with peripheral drug-eluting stents 28 and DCB. 29,30 Use of a DCB (and avoidance of stent implantation) does not limit future treatment options, an important consideration given the chronic and progressive nature of PAD. These findings compare favorably with other randomized clinical trials in this patient population. Despite the inclusion of longer lesion lengths that are at a higher risk of treatment failure, the 2.4% target lesion revascularization rate experienced in this trial is the lowest reported for an SFA device trial at 12 months. Clinically driven target lesion revascularization rates of 12.7% and 9.5% were reported in 2 recent randomized trials of bare metal and drug-eluting stents, despite their inclusion of shorter lesions (average lesion lengths of 7.0 and 5.4 cm, respectively). 7,8 IN.PACT DCB was associated with a low complication rate, including the absence of major amputations and a low rate of vessel thrombosis. Study Limitations The trial was deliberately and prospectively conducted in 2 sequential phases. The blinding of phase I was rigorously maintained until the completion of phase II. When the data were analyzed, there were no statistical differences between the 2 phases. Although improvements in the functional assessments of quality of life, walking impairment, and walking distance were observed in both treatment groups, the interpretation of these measures is complicated by the subjective nature of patient questionnaires and the influence of comorbidities, including progressive disease in nontreated vessels. The results of this trial cannot be generalized to patients not included in this trial. Future studies should encompass longer lesions and consider comparison with bare metal stents, drug-eluting stents, and bypass, and optimal medical therapy and exercise, as well. Longer-term follow-up is needed to confirm the durability of the benefit. Conclusions In conclusion, in this large, prospective, multicenter, international, randomized trial, DCB was superior to PTA and had a favorable safety profile for the treatment of patients with symptomatic superficial femoral and proximal popliteal artery PAD. The IN.PACT DCB demonstrated impressive patency rates with low repeat revascularization rates in comparison with other modern endovascular therapies. DCB stands to become an important treatment option for patients with superficial femoral and popliteal artery disease. Acknowledgments We thank Judith Greengard, Victoria Rendon, and Melissa Hasenbank for editing assistance. Contributions: Drs Tepe, Jaff, and Laird prepared the first draft of this manuscript, which was then reviewed and edited by the other coauthors. Drs Snead and Cohen undertook the statistical analysis. All the authors made substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data; and also contributed to drafting the article or revising it critically for important intellectual content. Source of Funding The study was funded by Medtronic, Inc. Disclosures Dr Tepe holds research grants from Bard Peripheral Vascular, B. Braun, Biotronic, Covidien, Medrad, and Medtronic. He is a compensated advisory board member for Medtronic, and receives speaking honoraria from Bard Peripheral Vascular, Biotronic, Covidien, Medrad, and Medtronic. Dr Laird holds research grants from W.L. Gore and Medtronic. He is a compensated advisory board member and consultant for Abbott Vascular, Bard Peripheral Vascular, Boston Scientific, Covidien, and Medtronic. Dr Micari holds research grants from Medtronic, and is a compensated consultant for Medtronic. Dr Metzger receives speaking honoraria from Bard Peripheral Vascular and Medtronic and is compensated for participation in training courses sponsored by Abbott Vascular and Medtronic. He is a compensated consultant for Abbott Vascular. Dr Scheinert holds research grants from Abbott Vascular, Angioslide, Atheromed, Biotronik, Boston Scientific, Cook Medical, Cordis, Covidien, CR Bard, Gardia Medical, Heoteq, Intact Vascular, Medtronic, TriReme Medical, and Upstream Peripheral Technologies. He is a compensated consultant/advisory board member for Abbott Vascular, Angioslide, Atheromed, Biotronik, Boston Scientific, Cook Medical, Cordis, Covidien, CR Bard, Gardia Medical, Heoteq, Intact Vascular, Medtronic, TriReme Medical, and Upstream Peripheral Technologies. Dr Zeller holds research grants from Bard-Lutonix, Biotronik, Cook Medical, and Medtronic. He receives speaking honoraria from Bard-Lutonix, Biotronik, Cook Medical, and Medtronic. Dr Cohen holds research grants from Abbott Vascular, Boston Scientific, Covidien, and Medtronic and is a compensated consultant for Abbott Vascular and Medtronic. Dr Snead, B. Alexander, and M. Landini are full-time employees of Medtronic. Dr Jaff is a noncompensated Advisor for Medtronic and is the Medical Director of VasCore, the Vascular Ultrasound Core Laboratory. He is a compensated member of VIVA Physicians, a not-for-profit 501c3 education/research organization. The other authors report no conflicts.
