PURPOSE: To define clinical and pathologic predictors of local-regional recurrence (LRR) for patients treated with neoadjuvant chemotherapy and mastectomy without radiation. PATIENTS AND METHODS: We analyzed the outcome of the 150 breast cancer cases treated on prospective institutional trials with neoadjuvant chemotherapy and mastectomy without postmastectomy radiation. Clinical stage at diagnosis was I in 1%, II in 43%, IIIA in 23%, IIIB in 25%, and IV in 7%. No patient had inflammatory breast cancer. RESULTS: The median follow-up period of surviving patients was 4.1 years. The 5- and 10-year actuarial rates of LRR were both 27%. Pretreatment factors that positively correlated with LRR were increasing T stage (P < .0001) and increasing combined clinical stage (P < .0001). Pathologic and treatment factors that positively correlated with LRR were size of the residual primary tumor (P = .0048), increasing number of involved lymph nodes (P < .0001), and no use of tamoxifen (P = .0013). The LRR rate for the 18 patients with a pathologic complete response of both the primary tumor and lymph nodes (pCR) was 19% (95% confidence interval, 6% to 48%). In a forward stepwise Cox logistic regression analysis, clinical stage IIIB or greater (hazard ratio of 4.5, P < .001), pathologic involvement of four or more lymph nodes (hazard ratio of 2.7, P = .008), and no use of tamoxifen (hazard ratio of 3.9, P = .027) independently predicted for LRR. CONCLUSION: Advanced disease at presentation and positive lymph nodes after chemotherapy predict for clinically significant rates of LRR. Achievement of pCR does not preclude the need for postmastectomy radiation if warranted by the pretreatment stage of the disease.
