Can the optimal type of stent be predicted based on clinical risk factors? A subgroup analysis of the randomized BASKET-PROVE trial

BACKGROUND:The randomized BASKET-PROVE study showed no significant differences between sirolimus-eluting stents (SES), everolimus-eluting stents (EES), and bare-metal stents (BMS) with respect to the primary end point, rates of death from cardiac causes, or myocardial infarction (MI) at 2 years of follow-up, in patients requiring stenting of a large coronary artery. Clinical risk factors may affect clinical outcomes after percutaneous coronary interventions. We present a retrospective analysis of the BASKET-PROVE data addressing the question as to whether the optimal type of stent can be predicted based on a cumulative clinical risk score. METHODS:A total of 2,314 patients (mean age 66 years) who underwent coronary angioplasty and implantation of ≥1 stents that were ≥3.0 mm in diameter were randomly assigned to receive SES, EES, or BMS. A cumulative clinical risk score was derived using a Cox model that included age, gender, cardiovascular risk factors (hypercholesterolemia, hypertension, family history of cardiovascular disease, diabetes, smoking), presence of ≥2 comorbidities (stroke, peripheral artery disease, chronic kidney disease, chronic rheumatic disease), a history of MI or coronary revascularization, and clinical presentation (stable angina, unstable angina, ST-segment elevation MI). RESULTS:An aggregate drug-eluting stent (DES) group (n = 1,549) comprising 775 patients receiving SES and 774 patients receiving EES was compared to 765 patients receiving BMS. Rates of death from cardiac causes or nonfatal MI at 2 years of follow-up were significantly increased in patients who were in the high tertile of risk stratification for the clinical risk score compared to those who were in the aggregate low-mid tertiles. In patients with a high clinical risk score, rates of death from cardiac causes or nonfatal MI were lower in patients receiving DES (2.4 per 100 person-years, 95% CI 1.6-3.6) compared with BMS (5.5 per 100 person-years, 95% CI 3.7-8.2, hazard ratio 0.45, 95% CI 0.26-0.80, P = .007). However, they were not significantly different between receivers of DES and BMS in patients in the low-mid risk tertiles. CONCLUSIONS:This exploratory analysis suggests that, in patients who require stenting of a large coronary artery, use of a clinical risk score may identify those patients for whom DES use may confer a clinical advantage over BMS, beyond lower restenosis rates.