Predictors of locoregional recurrence in patients with locally advanced breast cancer treated with neoadjuvant chemotherapy, mastectomy, and radiotherapy.

PURPOSE: To identify the clinical and pathologic factors predictive of locoregional recurrence (LRR) after neoadjuvant chemotherapy, mastectomy, and radiotherapy. METHODS AND MATERIALS: We retrospectively reviewed the hospital records of 542 patients treated on six consecutive institutional prospective trials using neoadjuvant chemotherapy and postmastectomy radiotherapy. The clinical stage (American Joint Committee on Cancer, 1988) was Stage II in 17%, Stage IIIA in 30%, Stage IIIB in 43%, and Stage IV (ipsilateral supraclavicular disease) in 10%. All LRRs were considered events, irrespective of the timing to distant metastases.
RESULTS: The median follow-up was 70 months. The 5-year and 10-year actuarial LRR rate was 9% and 11%, respectively. The clinical factors associated with LRR included combined clinical stage, clinical T stage, ipsilateral supraclavicular nodal disease, chemotherapy response, physical examination size after chemotherapy, and no tamoxifen use (p < or = 0.04 for all factors). The pathologic predictors of LRR included the number of positive nodes, dissection of <10 nodes, multifocal/multicentric disease, lymphovascular space invasion, extracapsular extension, skin/nipple involvement, and estrogen receptor-negative disease (p <or = 0.05 for all factors). Multivariate Cox regression analysis revealed that five factors independently predicted for LRR: skin/nipple involvement, supraclavicular nodal disease, no tamoxifen use, extracapsular extension, and estrogen receptor-negative disease (hazard ratio, 2.1-2.8; p < or = 0.02 for all factors). The 10-year LRR rate was only 4% for patients with one or none of these five independent factors, 8% for those with two factors, and 28% for those with three or more factors (p < 0.0001).
CONCLUSION: Although the long-term rate of LRR after neoadjuvant chemotherapy, mastectomy, and radiotherapy is low, we identified a number of factors that correlated independently with greater rates of LRR. Patients with three or more of these factors may benefit from research protocols investigating alternative treatment strategies.