Locoregional first recurrence after mastectomy: prospective cohort studies with and without immediate chemotherapy.

Abstract

To evaluate prospectively the impact of combination chemotherapy in the combined modality treatment of isolated first locoregional recurrence (LRR) following mastectomy for breast cancer. Between 1979 and 1989, 120 chemotherapy-naive women with isolated LRR as first failure after mastectomy were prospectively identified, uniformly staged, and systematically followed. Treatment consisted of excision if feasible, radical locoregional radiotherapy, and a hormonal maneuver (unless estrogen receptor negative). The initial chemotherapy cohort also received 8 cycles of doxorubicin and cyclophosphamide. This was compared to a subsequent control cohort. For all patients, the 10-year actuarial relapse-free survival +/- 95% confidence interval was 42.1+/-9.2%, and overall survival was 56.8+/-9.1%. No difference was seen in locoregional control between cohorts. At 5 years, distant recurrence-free survival for chemotherapy and control cohort respectively was 75.4+/-10.8% and 60.7+/-12.5% (p = 0.33) and overall survival was 81.9%+/-9.6 and 74.3%+/-11.2 (p = 0.24). Univariate analysis showed no prognostic importance for any imbalance between cohorts. Cox modeling confirmed that complete resection was strongly associated with fewer LRR (hazard ratio [HR] 0.32, p = 0.001) and also with better overall survival (HR 1.82, p = 0.019). Chemotherapy produced a substantial reduction in risk of death (HR 0.72 CI 0.421-1.235, p = 0.23). In this prospective but nonrandomized study of treatment for first LRR, the risk of death in the later control cohort was 1.39 times the risk in the chemotherapy cohort but failed to reach statistical significance. The results justify further study.