1

Introduction

Peripheral arterial disease (PAD) is estimated to be present in approximately 8.5 million Americans and is associated with high morbidity and both cardiovascular and all-cause mortality rates . The prevalence of PAD is approximately 12% in the adult population, with men being affected slightly more than women. In symptomatic patients, PAD has a significant impact on the quality of life due to claudication and atypical leg pain .

The most common artery diseased in the peripheral circulation is the femoropopliteal artery . Percutaneous transluminal balloon angioplasty (PTA) is the most common revascularization technique for SFA Disease . However, PTA of the SFA was reported to have high re-narrowing rate . Recently, drug-eluting stents (Zilver PTX) have been shown to be effective and were approved for the treatment of femoropopliteal artery disease . Nevertheless, the use of stents in the femoropopliteal region carries a risk of stent fracture, vessel injury, in-stent restenosis and in the case of drug eluting stent required prolonged antiplatelet therapy. Therefore, significant clinical need remains for a device that will result with acute and long-term patency. Drug coated balloon (DCB) technology was introduced to improve the outcome of PTA for SFA and results were promising when compared to PTA alone . DCB granted CE mark for the treatment of SFA in Europe since 2011.

The Lutonix drug coated balloon (DCB) (Lutonix, Inc., a subsidiary of C.R. Bard, New Hope, Minnesota) incorporates a standard PTA balloon with paclitaxel coating as the active pharmaceutical ingredient. Depending on the balloon size, a dose density of 2 μg/mm ² yields a variable total dose raging from 1.5 mg to 3.8 mg. Paclitaxel is carried by a non-polymer formulation consisting of polysorbate and sorbitol. This device is compatible with a 0.035” guidewire and is available in 75 cm, 100 cm and 130 cm catheter lengths. Balloon sizes range from 4.0 mm to 6.0 mm in diameter and from 40 mm to 100 mm in length. Devices are compatible with 5Fr (for the 4–5 mm balloon) and 6Fr (for the 6 mm balloon).

On June 12, 2014, the Food and Drug Administration’s (FDA) Circulatory System Devices Panel reviewed the Lutonix DCB premarket approval application (PMA) . This application was primarily based on data from the pivotal randomized, controlled and multicenter clinical trial (RCT), which compared the Lutonix DCB with PTA. The following is a summary of the discussions and recommendations made during the Circulatory System Devices Advisory Panel meeting.

2

Methodological trial design

The clinical testing of the Lutonix DCB was performed utilizing the Lutonix paclitaxel-coated balloon for the prevention of femoropopliteal restenosis (LEVANT) clinical program. This program includes the European randomized trial LEVANT 1 and the multicenter pivotal randomized trial LEVANT 2, which randomized patients inside and outside of the United States (US) (42 centers in the US and 12 in Europe). In addition, the program included two registries; a continued access cohort in sides participating in LEVANT 2, and a second cohort conducted at sites that did not participate in the pivotal trial.

Briefly, the LEVANT 1 trial randomized 101 European patients to receive the Lutonix DCB versus PTA with and without intended stenting. LEVANT 1 trial demonstrated a significantly lower late lumen loss at 6 months with the Lutonix DCB compared to standard PTA (p = 0.016) with comparable safety.

The LEVANT 2 study design was presented by Dr. Kenneth Rosenfield on behalf of the sponsor, and by Dr. Donna Buckley on behalf of the FDA. The LEVANT 2 pivotal trial was a global, prospective, multicenter, single blind, randomized, and controlled clinical trial comparing the Lutonix DCB with standard PTA for the treatment of femoropopliteal arteries. Selected inclusion and exclusion criteria are listed in Table 1 .

All patients enrolled underwent an angiogram followed by a mandatory pre-dilatation with a standard balloon 1 mm smaller than the reference vessel diameter. After initial lesion preparation, the investigator determined residual stenosis and vessel dissection. Patients with a residual restenosis < 70% and no flow limiting dissection or those with lesions not appropriate for stenting due to knee joint proximity were randomized 2:1 to receive either dilatation with the Lutonix DCB or standard PTA ( Fig. 1 ). Following angioplasty, bailout stenting was performed if necessary. Criteria for bailout stenting were well defined and rigorous (> 50% stenosis or flow limiting dissection and gradient refractory to prolonged dilatation). It should be noted that in this trial the investigators attempted to minimize stenting as much as possible as stenting may have been a confounder when analyzing the efficacy of the DCB versus standard PTA balloon. Bailout stenting was not counted as a failure in LEVANT 2 trial.

LEVANT 2 trial design.

LEVANT 2 enrolled a total of 543 patients (56 roll-in, 316 randomized to Lutonix DCB and 160 to PTA). The study was blinded to the patient and the investigator conducting the follow-up to prevent introduction of bias. Blinding of the treating physician was not possible due to significant differences in the appearance of the Lutonix DCB and a standard PTA balloon.

The primary efficacy endpoint was primary patency of the target lesion at 1 year (primary patency = absence of target lesion binary restenosis by duplex ultrasound (DUS) adjudicated by blinded Doppler core-lab and freedom from target lesion revascularization (TLR) adjudicated by a clinical events committee). The primary safety endpoint was a composite of freedom from all-cause perioperative death (< 30 days), freedom from index limb amputation, index limb re-intervention and index limb-related death adjudicated by a clinical events committee. Key secondary endpoints included TLR, target vessel revascularization, Rutherford class, response to walking impairment questionnaire, and the SF-36 quality of life. All of these were included for hierarchical testing. The study was powered for both primary efficacy and safety endpoints (> 90% power). Statistical design was for superiority, with two sided, alpha = 0.05 for the primary efficacy endpoint (binary restenosis adjudicated by DUS and TLR) and non-inferiority, with 5% margin for the primary safety endpoint.

LEVANT 2 completed enrollment on July 2012 and the final 12-month follow-up occurred on July 2013. In addition to the LEVANT 2 randomized trial, the LEVANT 2 safety registry completed enrollment in July 2013. This safety registry enrolled 657 individuals. Currently a Global SFA registry is enrolling patients outside the US with a planned enrollment of up to 1000 patients.

Overall, at the time of the FDA panel meeting there were 286 patients with evaluable follow up data for safety and 264 for effectiveness at 12 months from the LEVANT 2 RCT. From the LEVANT safety registry, there were 228 patients evaluable for safety and 193 for effectiveness at 12 month follow up.

2

Methodological trial design

The clinical testing of the Lutonix DCB was performed utilizing the Lutonix paclitaxel-coated balloon for the prevention of femoropopliteal restenosis (LEVANT) clinical program. This program includes the European randomized trial LEVANT 1 and the multicenter pivotal randomized trial LEVANT 2, which randomized patients inside and outside of the United States (US) (42 centers in the US and 12 in Europe). In addition, the program included two registries; a continued access cohort in sides participating in LEVANT 2, and a second cohort conducted at sites that did not participate in the pivotal trial.

Briefly, the LEVANT 1 trial randomized 101 European patients to receive the Lutonix DCB versus PTA with and without intended stenting. LEVANT 1 trial demonstrated a significantly lower late lumen loss at 6 months with the Lutonix DCB compared to standard PTA (p = 0.016) with comparable safety.

The LEVANT 2 study design was presented by Dr. Kenneth Rosenfield on behalf of the sponsor, and by Dr. Donna Buckley on behalf of the FDA. The LEVANT 2 pivotal trial was a global, prospective, multicenter, single blind, randomized, and controlled clinical trial comparing the Lutonix DCB with standard PTA for the treatment of femoropopliteal arteries. Selected inclusion and exclusion criteria are listed in Table 1 .

All patients enrolled underwent an angiogram followed by a mandatory pre-dilatation with a standard balloon 1 mm smaller than the reference vessel diameter. After initial lesion preparation, the investigator determined residual stenosis and vessel dissection. Patients with a residual restenosis < 70% and no flow limiting dissection or those with lesions not appropriate for stenting due to knee joint proximity were randomized 2:1 to receive either dilatation with the Lutonix DCB or standard PTA ( Fig. 1 ). Following angioplasty, bailout stenting was performed if necessary. Criteria for bailout stenting were well defined and rigorous (> 50% stenosis or flow limiting dissection and gradient refractory to prolonged dilatation). It should be noted that in this trial the investigators attempted to minimize stenting as much as possible as stenting may have been a confounder when analyzing the efficacy of the DCB versus standard PTA balloon. Bailout stenting was not counted as a failure in LEVANT 2 trial.

LEVANT 2 trial design.

LEVANT 2 enrolled a total of 543 patients (56 roll-in, 316 randomized to Lutonix DCB and 160 to PTA). The study was blinded to the patient and the investigator conducting the follow-up to prevent introduction of bias. Blinding of the treating physician was not possible due to significant differences in the appearance of the Lutonix DCB and a standard PTA balloon.

The primary efficacy endpoint was primary patency of the target lesion at 1 year (primary patency = absence of target lesion binary restenosis by duplex ultrasound (DUS) adjudicated by blinded Doppler core-lab and freedom from target lesion revascularization (TLR) adjudicated by a clinical events committee). The primary safety endpoint was a composite of freedom from all-cause perioperative death (< 30 days), freedom from index limb amputation, index limb re-intervention and index limb-related death adjudicated by a clinical events committee. Key secondary endpoints included TLR, target vessel revascularization, Rutherford class, response to walking impairment questionnaire, and the SF-36 quality of life. All of these were included for hierarchical testing. The study was powered for both primary efficacy and safety endpoints (> 90% power). Statistical design was for superiority, with two sided, alpha = 0.05 for the primary efficacy endpoint (binary restenosis adjudicated by DUS and TLR) and non-inferiority, with 5% margin for the primary safety endpoint.

LEVANT 2 completed enrollment on July 2012 and the final 12-month follow-up occurred on July 2013. In addition to the LEVANT 2 randomized trial, the LEVANT 2 safety registry completed enrollment in July 2013. This safety registry enrolled 657 individuals. Currently a Global SFA registry is enrolling patients outside the US with a planned enrollment of up to 1000 patients.

Overall, at the time of the FDA panel meeting there were 286 patients with evaluable follow up data for safety and 264 for effectiveness at 12 months from the LEVANT 2 RCT. From the LEVANT safety registry, there were 228 patients evaluable for safety and 193 for effectiveness at 12 month follow up.