e12030 Background: The role of post-mastectomy radiation therapy (PMRT) after neo-adjuvant chemotherapy (NCT) in those with T3N0 disease, younger age, and different tumor subtypes is unclear. Methods: We conducted a single institution, retrospective analysis of patients treated with NCT and mastectomy from 1985-2010. The primary objectives were to (1) assess the association between PMRT and long-term outcomes, and (2) describe outcomes and practice patterns based on tumor subtype, pathologic response, and age at diagnosis. Secondary objectives were to analyze the benefit of PMRT in those with favorable tumor subtypes, cT3N0 disease, and patients younger than 40. Log-rank test and Cox regression was used to test the association of predictors with outcomes. IRB approval was obtained. Results: Of 131 evaluable patients, 115 (88%) received PMRT. Mean±SD age was 48±10 years (range 29-79). Of patients <40 years of age (n=33) 88% received PMRT. At diagnosis, 69.5% were cT3; 53.5% were cN1; 12.3% were cT3N0. Seventy-eight percent of patients received anthracycline-based NCT and 52% received both anthracycline and taxane NCT. Twenty-eight percent of all patients achieved pCR. With a median f/u of 39 months (range 8-177), 30.5% of patients had a recurrence event (6.9% local, 5.3% nodal, 16.8% distant). Ten year OS and DFS for all patients were 56.7% (95% CI, 43.5-73.8) and 66.1% (95% CI, 52.3-83.5), respectively. Those with pCR had a non-significant trend toward better outcome, with 10 year OS=68.0% vs. 54.0% (p=0.7); DFS=70.2% vs. 65.2% (p=1.0). In the Cox regression analysis, pathologic T-stage, triple negative disease, and tumor grade were significantly associated with OS; triple negative disease was also predictive of DFS. Benefit of PMRT was not analyzable in cT3N0 due to small numbers. Conclusions: Most patients had PMRT, with outcomes on par with historical data. Most patients <40 years received PMRT, with no events for the non-PMRT cohort most likely due to selection bias. Unfavorable subgroups were not more likely to receive PMRT, but of those who did, only triple negative was predictive of outcome. A future non-PMRT matched cohort study is necessary to further elucidate the role of PMRT.