The use of high-dose cyclophosphamide, carmustine, and thiotepa plus autologous hematopoietic stem cell transplantation as consolidation therapy for high-risk primary breast cancer after primary surgery or neoadjuvant chemotherapy

Abstract

We assessed the 5-year results of a high-dose cyclophosphamide, carmustine, and thiotepa (CBT) regimen plus autologous hematopoietic stem cell transplantation (AHST) as an adjuvant consolidation therapy for high-risk primary breast cancer patients with > or =10 positive axillary lymph nodes after primary surgery or > or =4 positive axillary lymph nodes after neoadjuvant chemotherapy and surgery. The associations of various potential prognostic factors with the relapse-free survival (RFS) rate and overall survival (OS) rate were determined. Between October 1992 and March 2000, 177 eligible patients (median age, 46 years) were given high-dose CBT followed by AHST. At a median follow-up of 63 months, the acute treatment-related mortality was 4.5%. Estimated 5-year RFS and OS rates were 62% and 68%, respectively, for all patients. For patients with > or =10 positive axillary lymph nodes after primary surgery, the 5-year RFS and OS rates were 71% and 70%, respectively, and for patients with > or =4 positive axillary lymph nodes after neoadjuvant chemotherapy, the 5-year RFS and OS rates were 53% and 66%, respectively. In 2-sided log-rank tests, earlier disease stage, a lower lymph node ratio, and a lower tumor score were associated with a prolonged RFS and OS. In a multivariate proportional hazards model, disease stage and lymph node ratio remained significant. We concluded that high-dose CBT with AHST for high-risk primary breast cancer is feasible, with comparable efficacy to other phase II studies. More than a 50% estimated 5-year survival rate was seen in all high-risk primary breast cancer patients. In accordance with results from recent randomized studies, we need to continue high-dose chemotherapy with AHST for patients with high-risk primary breast cancer in the phase III randomized setting.