MP14-07 PATTERNS AND PREDICTORS OF CLINICAL RECURRENCE FOLLOWING EARLY SALVAGE RADIATION THERAPY IN PATIENTS WITH PSA RISE AFTER RADICAL PROSTATECTOMY: A LONG TERM MULTI-INSTITUTIONAL ANALYSIS

INTRODUCTION AND OBJECTIVES: Prostate cancer (PCa) patients with seminal vesicle invasion (SVI) are at increased risk of recurrence and mortality after radical prostatectomy (RP). We aimed at assessing the impact of adjuvant radiotherapy (aRT) on cancer-specific mortality (CSM) in patients with SVI after adjusting for the risk of dying from other causes METHODS: A total of 1,218 patients with localized PCa undergoing RP at a single center between 1990 and 2015 were identified. All patients included in the study had SVI at final pathology. Cumulative incidence CSM rates were generated for the overall population and after stratifying patients according to aRT status and were compared using the Gray test. Competing-risks regression analyses assessed the impact of aRT on CSM after adjusting for the risk of dying from other causes. Covariates consisted of age, positive surgical margins, pathologic Gleason score, adjuvant hormonal therapies, and nodal status RESULTS: Median age was 66.4 years. Overall, 10.5, 42.3, and 47.2% of the patients included had pathologic Gleason score 6, 7, and 8, respectively. Overall, 539 (44.3%) patients had positive surgical margins. Overall, 641 (52.6%) patients had node positive disease. Median follow-up was 84 months. Overall, 222 patients experienced clinical recurrence (CR). The 15-year CR-free survival rate was 55.5%. Among patients who experienced CR, 19.7, 20.7, 8.0, and 51.6% had involvement of the prostatic fossa, pelvic lymph nodes, retroperitoneal nodes, and systemic metastases. Overall, 98 and 110 patients died from PCa and other causes, respectively. Overall, the 15-year CSM rate was 18.8%. When patients were stratified according to aRT status, the 15-year CSM rates were significantly higher among those who did not receive aRT as compared to those treated with aRT (22.9 vs. 13.7%, respectively; P1⁄40.02). In competing-risks regression analyses, aRT (Hazard ratio [HR]: 0.51; P1⁄40.01), pathologic Gleason score 8-10 (HR: 2.36; P1⁄40.01); and pN1 (HR: 1.79; P1⁄40.02) represented independent predictors of CSM CONCLUSIONS: A substantial proportion of patients with SVI is at risk of experiencing systemic recurrence after surgery. Adjuvant RT significantly reduced the risk of CSM after accounting for the risk of dying from other causes. These results suggest that maximising local control would improve oncologic outcomes in these patients. Conversely, men with node positive disease and higher Gleason score are at higher risk of CSM and might benefit more from systemic approaches