INTRODUÇÃO: Reportamos a incidência e preditores de nova revascularização da lesão-alvo (RLA) numa coorte de pacientes do mundo real tratados com stents farmacológicos (SF) MÉTODO: Entre maio de 2002 e fevereiro de 2010, 3.320 pacientes de único centro foram submetidos a intervenção percutânea (ICP), eletiva ou de urgência, com implante de 5.320 SF (Cypher, 83%; Taxus, 9%; Xience V/Promus, 3%; Endeavor, 4%; BioMatrix, 1%). O seguimento clínico de até oito anos (mediana, 3,4 anos) foi completo em 98% do grupo e 129 pacientes (3,88%) foram submetidos a RLA RESULTADOS: Entre os pacientes com RLA, 33% tinham diabetes, 10% eram portadores de insuficiência renal e 61%, de doença em múltiplos vasos. A artéria descendente anterior foi o vaso mais frequentemente tratado (33%), e 72,3% das lesões eram de alta complexidade (tipo B2/C). Nesse grupo, foram implantados 214 SF em 191 lesões e 45% dos pacientes receberam múltiplos stents. Na análise univariada, os preditores de RLA foram: diabetes melito [razão de risco (RR) 1,88, intervalo de confiança (IC) de 95% 1,08-3,27; P = 0,025), uso do stent Taxus (RR 1,73, IC 95% 1,20-2,50; P = 0,003), lesão ostial (RR 1,82, IC 95% 1,1-2,98; P = 0,018), ponte de safena (RR 2,36; IC 95% 1,55-3,50; p = 0,0001), e lesões reestenóticas (RR 1,95, IC 95% 1,17-3,24; P = 0,009). Após análise multivariada, permaneceram como preditores independentes: diabetes melito, uso do stent Taxus e tratamento de lesões em pontes de safena (RR 1,64, IC 95% 1,14-2,35; P = 0,008) CONCLUSÃO: No registro DESIRE, a RLA em pacientes complexos tratados com SF foi baixa, e o diabetes melito, o tratamento de lesões em ponte de safena e o uso de SF Taxus foram os únicos preditores de RLA no acompanhamento clínico de até oito anos. BACKGROUND: We report the incidence and predictors of target lesion revascularization (TLR) in a cohort of realworld patients treated with drug eluting stents (DES) METHOD: From May 2002 to February 2010, 3,320 patients were treated with elective or emergency percutaneous coronary intervention (PCI) and 5,320 DES (Cypher, 83%; Taxus, 9%; Xience V/Promus, 3%; Endeavor, 4%; BioMatrix, 1%) were implanted in a single center. The clinical follow-up of up to eight years (median time 3.4 years) was concluded in 98% patients of the group and 129 patients (3.88%) were submitted to TLR RESULTS: Of the patients with TLR, 33% had diabetes, 10% had chronic kidney disease and 61% had multivessel disease. The left anterior descending artery was the most frequent target vessel (33%) and 72.3% of the lesions were complex lesions (type B2/C). In this group, 214 DES were implanted in 191 lesions and 45% of the patients received multiple stents. In the univariate analysis, TLR predictors were: diabetes mellitus [relative risk (RR) 1.88, 95% confidence interval (CI) 1.08-3.27; P = 0.025], use of Taxus stents (RR 1.73, 95% CI 1.20-2.50; P = 0.003), ostial lesion (RR 1.82, 95% CI 1.1-2.98; P = 0.018), saphenous bypass graft (RR 2.36, 95% CI 1.55-3.50; P = 0.0001), and restenotic lesions (RR 1.95, 95% CI 1.17-3.24; P = 0.009). After the multivariate analysis, the independent predictors were: diabetes mellitus, use of Taxus stents and treatment of lesions in saphenous bypass grafts (RR 1.64, 95% CI, 1.14-2.35; P = 0.008) CONCLUSION: In the DESIRE registry, TLR in complex patients treated with DES was low and diabetes mellitus, treatment of lesions in saphenous bypass grafts and the use of Taxus stents were the only TLR predictors in the clinical follow-up of up to 8 years.