Purpose: To compare rates of locoregional recurrence (LRR) after mastectomy, doxorubicin-based chemotherapy, and radiation with those of patients receiving mastectomy and doxorubicin-based chemotherapy without radiation and to determine predictors of LRR after postmastectomy radiation.
Methods: Kaplan-Meier freedom-from-LRR rates were calculated for 470 patients treated with mastectomy, doxorubicin-based chemotherapy, and postmastectomy radiation in five single-institution clinical trials. The LRR rates in these patients were compared to previously reported rates in 1031 patients treated without radiation in the same trials.
Results: Median follow-up was 14 years. Irradiated patients had significantly less favorable prognostic factors for LRR than did unirradiated patients. Despite this, in all subsets of node-positive patients, postmastectomy radiation led to lower rates of LRR. This included patients with T1 or T2 tumors and one to three positive nodes (10-year LRR rates of 3% vs. 13%, p = 0.003). Multivariate analysis of LRR for patients with this stage of disease revealed that no radiation, close/positive margins, gross extracapsular extension, and dissection of <10 nodes predicted for increased LRR (hazard ratios 6.25, 4.61, 3.27, and 2.66, respectively). Significant predictors of LRR for patients treated with postmastectomy radiation were higher number and >or=20% positive nodes, larger tumor size, lymphovascular space invasion, and estrogen receptor (ER)-negative disease. Recursive partitioning analysis revealed ER-negative status to be the most powerful discriminator of LRR in irradiated patients.
Conclusions: Postmastectomy radiation decreases LRR for patients with breast cancer, including those with Stage II breast cancer and one to three positive lymph nodes.