The role of clinicopathologic and molecular prognostic factors in the post-mastectomy radiotherapy (PMRT): a retrospective analysis of 912 patients.

OBJECTIVE To assess the association of clinicopathologic and molecular features with loco-regional recurrence (LRR) in post-mastectomy breast cancer patients with or without adjuvant radiotherapy (PMRT). PATIENTS AND METHODS We retrospectively reviewed data of patients undergone to mastectomy followed or not by PMRT between January 2004 and June 2013. The patients were divided according to clinicopathologic and molecular sub-classification features. LRR and Cancer Specific Survival (CSS) were calculated using the Kaplan-Meier method; the prognostic factors were compared using long-rank tests and Cox regression model. RESULTS A total of 912 patients underwent to mastectomy of whom 269 (29.5%) followed by PMRT and 643 (70.5%) not; among the PMRT group, 77 underwent to the chest wall (CW) and 202 to the chest wall and lymphatic drainage (CWLD) irradiation. The median follow-up was 54 months (range, 3-118). No significant difference in terms of LRR and CSS was found between non-PMRT and PMRT group (p=0.175; and p=0.628). The multivariate analysis of LRR for patients who did not undergo PMRT showed a significant correlation with the presence of extracapsular extension (ECE) (p=0.049), Ki-67>30% (p=0.048) and triple negative status (p=0.001). In the PMRT group, triple negative status resulted as the only variable significantly correlated to LRR (p=0.006) at the multivariate analysis and T-stage also showed a trend to significance (p=0.073). Finally, no difference in LRR control was shown between CW and CWLD-PMRT (p=0.078). CONCLUSIONS After mastectomy ECE, a cut off of Ki-67>30% and triple negative status were strictly correlated with LRR regardless of clinicopathologic stage. PMRT has a positive impact in decreasing LRR in patients with this molecular profile. Besides, CW might represent a valid option for patients with one to three positive nodes.