Drug-eluting stent thrombosis in routine clinical practice: two-year outcomes and predictors from the TAXUS ARRIVE registries.

BACKGROUND

Stent thrombosis (ST) is an uncommon but serious complication of drug-eluting and bare metal stents. To assess drug-eluting stent ST in contemporary practice, we analyzed 2-year data from the 7492-patient ARRIVE registry.

METHODS AND RESULTS

Patients were enrolled at the initiation of percutaneous coronary intervention with no inclusion/exclusion criteria beyond use of the paclitaxel-eluting TAXUS stent. Two-year follow-up was 94% with independent adjudication of major cardiac events. A second, autonomous committee adjudicated Academic Research Consortium (ARC) definite/probable ST. Cumulative 2-year ARC-defined ST was 2.6% (1.0% early ST [<30 days], 0.7% late ST [31 to 365 days], and 0.8% very late ST [>1 year]). Simple-use (single-vessel and single-stent) cases had lower rates than expanded use (broader patient/lesion characteristics, 2-year cumulative: 1.4% versus 3.3%, P<0.001; early ST: 0.4% versus 1.4%, P<0.001; late ST: 0.5% versus 0.8%, P=0.14; very late ST: 0.4% versus 1.0%, P=0.008). Within 7 days of ST, 23% of patients died; 28% suffered Q-wave myocardial infarction. Mortality was higher with early ST (39%) than late ST (12%, P<0.001) or very late ST (13%, P<0.001). Multivariate analysis showed anatomic factors increased early ST (lesion >28 mm, lesion calcification) and late ST (vessel <3.0 mm); biological factors increased very late ST (renal disease, prior brachytherapy). Although early ST (71.4%) and very late ST (23.1%) patients had dual antiplatelet therapy at the time of ST, premature thienopyridine discontinuation was a strong independent predictor of both.

CONCLUSIONS

The relative risks of early and late ST differ. Knowledge of ST risk for specific subgroups may guide revascularization options until the completion of randomized trials in these broad populations.