Estimation of DAPT Study Treatment Effects in Contemporary Clinical Practice: Findings from the EXTEND-DAPT Study.

Background: Differences in patient characteristics, changes in treatment algorithms, and advances in medical technology could each influence the applicability of older randomized trial results to contemporary clinical practice. The Dual Antiplatelet Therapy (DAPT) Study found that longer duration DAPT decreased ischemic events at the expense of greater bleeding, but subsequent evolution in stent technology and clinical practice may attenuate the benefit of prolonged DAPT in a contemporary population. We evaluated whether the DAPT Study population is different from a contemporary population of US patients receiving percutaneous coronary intervention (PCI), and estimated the treatment effect of extended duration antiplatelet therapy after PCI in this more contemporary cohort. Methods: We compared characteristics of drug-eluting stent (DES)-treated patients randomized in the DAPT Study to a sample of more contemporary DES-treated patients in the NCDR CathPCI Registry from July 2016-June 2017. After linking trial and registry data, we employed inverse-odds of trial participation weighting to account for patient and procedural characteristics and estimated a contemporary "real-world" treatment effect of 30 vs. 12 months of DAPT after coronary stent procedures. Results: The US DES-treated trial cohort included 8864 DAPT Study patients and the registry cohort included 568,540 patients. Compared to the trial population, registry patients had more comorbidities and were more likely to present with myocardial infarction and receive 2nd generation DES. After reweighting trial results to represent the registry population, there was no longer a significant effect of prolonged DAPT on reducing stent thrombosis (reweighted treatment effect: -0.40, 95% CI: -0.99%, 0.15%), major adverse cardiac and cerebrovascular events (reweighted treatment effect: -0.52, 95% CI: -2.62%, 1.03%), or myocardial infarction (reweighted treatment effect: -0.97%, 95% CI: -2.75%, 0.18%), but the increase in bleeding with prolonged DAPT persisted (reweighted treatment effect: 2.42%, 95% CI: 0.79%, 3.91%). Conclusions: Differences between patients and devices used in contemporary clinical practice compared with the DAPT Study were associated with attenuation of benefits and greater harms attributable to prolonged DAPT duration. These findings limit applicability of average treatment effects from the DAPT Study in modern clinical practice.