The New Approaches to Coronary Intervention (NACI) registry was established to define the role of new coronary devices in overcoming the limitations of balloon angioplasty. The purpose of the present study was to evaluate the acute and long-term efficacy of the transluminal extraction catheter (TEC) device utilizing data from the NACI registry and identify clinical and anatomic patient subsets who may benefit from this device. From 1990-1994, >4,300 patients from 39 clinical sites enrolled consecutive patients treated with one of the 7 new devices to the NACI registry. The study population consists of 331 patients (385 lesions) treated with planned TEC as the sole new device. Of these patients, 243 (292 lesions) were treated for saphenous vein graft (SVG) disease and 88 (93 lesions) for native disease. Patients undergoing SVG treatment were older and more likely to be male. They had lower ventricular function, more unstable angina, and a higher incidence of congestive heart failure. Multivessel disease was more prevalent in the SVG cohort, as was evidence of thrombus before treatment. Although device success was achieved in 50% of SVG lesions and 41% of native lesions, lesion success was achieved in 90% and 78%, respectively, after adjunctive balloon angioplasty, and procedure success rates were 86% and 79%, respectively. The in-hospital major complication (death/Q-wave myocardial infarction/emergency coronary artery bypass graft [CABG] surgery) rate was higher in the SVG cohort (6.2% vs 2.3%), mainly due to higher mortality rate (5.3% vs 1.1%). Multivariate analysis showed that SVG was not an independent predictor for either an in-hospital major complication or clinical failure. The risk factors for major in-hospital complications were history of congestive heart failure (odds ratio = 3.17) and thrombus (odds ratio = 3.36). For clinical failure the risk factors were diabetes (odds ratio = 1.88), thrombus (odds ratio = 2.08), and calcium (odds ratio = 3.09). One-year rates of death, Q-wave myocardial infarction, or any repeat revascularization were 51% in the SVG cohort and 41% in the native cohort. Following adjustment, patients treated for SVG disease did not have a higher risk when compared with those treated for native disease. The factors significantly associated with this composite event at 1 year are male (relative risk = 1.41), patients with history of congestive heart failure (relative risk = 1.56), and total occlusions (relative risk = 1.52). This study shows that for both SVG and native cohorts, device success rates were low with TEC alone, but acceptable lesion success rates were achieved when adjunctive PTCA was used. In-hospital as well as 1-year major complications were higher in the SVG cohort. However, after adjusting for other risk factors, SVG attempt was not significantly associated with either in-hospital or 1-year events.