Predictors of Drug-Coated Balloon Efficacy in Femoropopliteal Intervention: Insights From a Patient-Level Pooled Analysis

During a session on Monday, June 6, that focused on novel approaches and insights for treatment of complex femoropopliteal disease, Prakash Krishnan, MD, from Mount Sinai Health System in New York, shared the results of a study that evaluated the predictors of drug-coated balloon (DCB) outcomes in a large, rigorous dataset with the use of core labs and CEC adjudication. While a few factors have been identified that impact DCB efficacy for peripheral arterial disease, including Rutherford class, lesions over 150 mm in length, body mass index, TASC II class, coronary artery disease, and calcium scores, the multivariable outcomes across studies are inconsistent.

The study used the DCB arm of the IN.PACT SFA (EU and US) randomized control trial (RCT) (n = 220), the DCB arm of the IN.PACT Japan RCT (n = 68), the IN.PACT Global Long Lesion Imaging Cohort (n = 157), and the IN.PACT CTO Imaging Cohort (n = 126); total subjects, N = 557. Twenty-eight clinical factors, 15 anatomical factors, and 82 procedural factors were used, and the entry criteria was strict: P<.20 on univariate and variables with <20% missing data; the focus was on clinical and procedural relevance.

The safety and effectiveness outcomes showed a primary patency through 1 year rate of 82.5%, a freedom from clinically driven target lesion revascularization (CD-TLR) through 1 year rate of 94.7%, and a freedom from binary restenosis through 1 year rate of 84.4%. There were no device/procedure-related deaths after 30 days or major target limb amputations after 1 year. “The predictors that we found of CD-TLR were residual stenosis greater than 30% and Rutherford class greater than 3,” Dr. Krishnan stated. “As far as binary restenosis, it was very clear that residual stenosis was the clear predictor of binary restenosis. And when you look at primary patency failure, the clear predictor was residual stenosis of less than 30%.” Interestingly, the presence of both risk factors increased the risk of CD-TLR and binary restenosis.

In summary, using an expanded set of clinical, anatomical, and procedural risk factors, only residual stenosis over 30%, Rutherford category greater than 3, and smaller preprocedure reference vessel diameter were significant multivariable predictors of reduced DCB effectiveness. “My feeling is that vessel prep may be important—the jury’s still out—however postprocedure residual stenosis greater than 30% was most strongly associated with DCB efficacy,” Dr. Krishnan said. “It increased the risk of CD-TLR by 4.7-fold, binary restenosis by 1.7-fold, and loss of patency by 1.9-fold.” Longer lesion length, severe calcification, and total occlusion were not statistically significant predictors in the current multivariable analysis; however, longer lesions were more commonly treated with provisional stents, so conclusions cannot be fully made. With relatively few TLR and patency failure events, these results will need to be replicated in a larger dataset that is sufficiently powered to determine the algorithm of clinical, anatomical, and procedural risk factors most associated with DCB effectiveness.