Résumé : PURPOSE: The 4-year results of this trial demonstrated that a higher dose of epirubicin with cyclophosphamide (HEC) is superior to a lower dose of epirubicin, 60 mg/m(2) (EC), for event-free survival (EFS; 27% reduction), but is not superior to classical oral cyclophosphamide, methotrexate, and fluorouracil (CMF) in the adjuvant treatment of node-positive breast cancer. Herein we report the 15-year data on efficacy and long-term toxicity of this three-arm Belgian multicenter trial. PATIENTS AND METHODS: Between March 1988 and December 1996, 777 eligible patients were randomly assigned to six cycles of CMF, eight cycles of EC, or eight cycles HEC. RESULTS: The 15-year EFS was 45% for patients who received CMF, 39% for patients who received EC, and 50% for patients who received HEC. The hazard ratios (HR) were 0.77 for HEC versus EC (95% CI, 0.60 to 0.98; P = .03), 0.90 for HEC versus CMF (P = .39), and 0.86 for EC versus CMF (P = .21). No difference in overall survival (OS) was seen. Cardiac toxicity was more frequent with HEC than with CMF (11 patients v 1 patient; P = .006), but no more than with EC (P = .21). CONCLUSION: Treatment with HEC demonstrated superior EFS when compared with lower-dose epirubicin. However, we do not recommend the use of HEC regimen in daily clinical practice, mainly because of the higher risk of cardiotoxicity related to the cumulative doses of epirubicin and the lack of superiority of anthracyclines over CMF in our study.