IMPACT OF SELECTION PROCESS ON RESPONSE RATE AND LONG-TERM SURVIVAL OF POTENTIAL HIGH-DOSE CHEMOTHERAPY CANDIDATES TREATED WITH STANDARD-DOSE DOXORUBICIN-CONTAINING CHEMOTHERAPY IN PATIENTS WITH METASTATIC BREAST-CANCER

Purpose: Most of the data about high-dose chemotherapy (HDCT) for meta static breast cancer are derived from phase II studies. The interpreta tion of these data depends on comparisons with data from properly sele cted historical control patients treated with standard therapy under s imilar circumstances. We report the longterm results of patients with metastatic breast cancer who were eligible for HDCT but were treated w ith doxorubicin-containing standard-dose chemotherapy. Patients and Me thods: Prospectively collected data from 18 successive doxorubicin-con taining protocols for the treatment of metastatic breast cancer were e valuated. Using common eligibility criteria for HDCT, we identified pa tients who would have been candidates for HDCT. We analyzed response r ates, progression-free survival (PFS), and overall survival (OS) for a ll patients, potential HDCT candidates, and noncandidates. Results: A total of 1,581 patients was enrolled onto the 18 studies. Six hundred forty-five were HDCT candidates, and 936 were noncandidates. The compl ete response rate was 27% for HDCT candidates and 7% for noncandidates ; median PFS was 16 and 8 months and median OS was 30 and 17 months, r espectively. Survival rates for HDCT candidates and noncandidates, res pectively, were 21% and 6% at 5 years and 7% and 2% at 10 years. Concl usion: This study suggests that encouraging results of single-arm tria ls of HDCT could partially be due to selection of patients with better prognoses and further stresses the importance of completing ongoing r andomized trials of HDCT to assess the relative efficacy of HDCT in pa tients with metastatic breast cancer. (C) 1997 by American Society of Clinical Oncology.