Background: In randomized clinical trials, the risk of thrombotic events with the absorb bioresorbable vascular scaffold (BVS) was significantly higher than with metallic drug-eluting stents. We evaluated predictors of scaffold thrombosis in the large-scale, multicenter German-Austrian ABSORB RegIstRy.
Methods and results: 3178 patients with treatment of 4252 lesions using 5020 scaffolds were included. Follow-up rate at 6 months was 97.4%. Forty-five (1.42%) patients experienced definite/probable scaffold thrombosis during follow-up. Multiple regression analysis showed implantation of absorb BVS in bifurcation lesions [odds ratio (OR): 4.43; 95% confidence interval (CI): 1.69-11.59; P=0.0024] or treatment in the years 2013/2014 (OR: 1.88; 95% CI: 1.02-3.47; P=0.04) to be significant predictors of scaffold thrombosis. Excluding bifurcation lesions, the incidence of definite/probable scaffold thrombosis decreased from 1.8% (95% CI: 1.17-2.64%) in 2013/2014 to 0.89% (95% CI: 0.5-1.46%) in 2015/2016. In the latter period, absorb BVS were implanted more often in younger patients with less complex de novo lesions, and debulking devices and postdilatation were used more frequently. Between the two treatment periods, there was a significant reduction in myocardial infarction (2.73-1.24%, P<0.01; OR: 0.45; 95% CI: 0.26-0.77), definite/probable scaffold thrombosis (1.79-0.88%, P<0.05; OR: 0.49; 95% CI: 0.26-0.93), and target lesion failure and revascularization during follow-up.
Conclusion: Improved procedural technique and more strict patient selection may explain a significant decrease in the absorb BVS thrombosis rates during the recruitment period of the large-scale German-Austrian ABSORB RegIstRy. In addition, treatment of bifurcation lesions was identified as an independent predictor of definite/probable scaffold thrombosis.