RESUMO
BACKGROUND: In unselected patients, the incidence of restenosis is lower after placement of drug-eluting stents (DES) than bare-metal stents (BMS) without difference in safety at a time horizon of 4 years. However, DES appears less effective in "off label" patients. OBJECTIVES: The aim of the study was to assess available evidence of DES efficacy and safety by patient category to establish when DES placement may be recommended for reimbursement by the French national health insurance. METHODS: Based on a systematic review by patient category (January 2002 to August 2009), two health technology assessment (HTA) reports and thirty-eight clinical studies not covered by the HTA reports (eleven meta-analysis including ours, eleven randomized trials and sixteen cohort studies) were selected. After assessment of the methodological quality, the studies mostly comparing DES with BMS were reviewed by a panel of health professionals who defined a priori the most relevant end points of safety and efficacy. RESULTS: Seven to fourteen patients treated with DES were needed to avoid one target lesion revascularization (TLR) in patients with lesions >15 mm long, vessel diameter <3 mm, or diabetes, and with some complex lesions (total coronary occlusion, BMS in-stent restenosis multivessel disease, unprotected left main stenosis). DES appeared as safe as other alternatives over a follow-up of up to 4 years when dual antiplatelet therapy was continued for at least 1 year, but statistical power remains limited to conclude for some clinical features. CONCLUSIONS: For reimbursement, DES use should be limited to certain categories of patients. Treatment of particular cases requires a multidisciplinary approach.