Purpose: To examine the predictive value of tumor- and treatment-specific prognostic indicators of relapse-free survival (RFS) and overall survival (OS) in patients with high-risk breast cancer (HRBC) treated with high-dose chemotherapy (HDCT) and stem-cell rescue.
Patients and methods: Between June 1989 and September 1994, 114 patients with HRBC (stage II with > or = 10 axillary lymph nodes involved, stage IIIA, and stage IIIB inflammatory carcinoma) received adjuvant chemotherapy followed by HDCT with etoposide, cyclophosphamide, and either doxorubicin (CAVP) or cisplatin (CCVP). Variables analyzed included stage, tumor size, number of axillary nodes involved, grade and receptor status, and types of adjuvant chemotherapy and radiation therapy and HDCT.
Results: With a median follow-up time of 46 months (range, 23 to 93), Kaplan-Meier estimates of 3.5-year OS for stage II, IIIA, and IIIB HRBC are 82% (95% confidence interval [CI], 67% to 97%), 79% (95% CI, 67% to 91%), and 72% (95% CI, 53% to 91%); RFS estimates are 71% (95% CI, 56% to 85%), 57% (95% CI, 43% to 72%), and 50% (95% CI, 29% to 71%) irrespective of the HDCT regimen. In univariate analysis, the risk of relapse was lower for patients with progesterone receptor (PR)-positive tumors (risk ratio [RR], 0.43; 95% CI, 0.22 to 0.81; P = .01) and higher for patients with inflammatory carcinoma (RR, 2.20; 95% CI, 1.02 to 4.76; P = .05). OS was better for patients with PR (RR, 0.16; 95% CI, 0.05 to 0.55; P = .003) and estrogen receptor (ER)-positive tumors (RR, 0.42; 95% CI, 0.17 to 1.02; P = .05); OS was worse for patients with high-grade primary tumors (RR, 4.08; 95% CI, 1.21-13.7; P = .02). In multivariate analysis, PR positivity was associated with improved RFS (P = .01) and OS (P = .001).
Conclusion: HDCT in selected patients with HRBC is safe and warrants further evaluation. Patients with receptor-negative, high-grade, or inflammatory tumors require improvement in their therapeutic options. Better assessment of the role of HDCT awaits completion of ongoing randomized trials.