MP83-12 SALVAGE ROBOT-ASSISTED RADICAL PROSTATECTOMY FOR RECURRENT PROSTATE CANCER.

Abstract

INTRODUCTION AND OBJECTIVES: To report short-term cancer control, functional and perioperative outcomes in a multi-institutional cohort of salvage robot-assisted radical prostatectomy (sRARP). METHODS: We retrospectively reviewed the records of 88 consecutivemenwho underwent sRARP, from July 2007 to June 2014, at 2 high-volume reference centers in the US. Failed primary therapy was: external beam radiotherapy (49%); brachytherapy (35%); HIFU (5%); cryoablation (9%) and others (2%). The endpoints were biochemical failure (BF), positive surgical margins (PSM), 30-day post-operative complications (POC), urinary continence (UC) and erectile function (EF). BF was defined as a PSA>0.2 ng/ml. UC and EF were defined as the use of 0padandSHIMscore>21, respectively. POCwere classifiedandgraded using the Clavien system. Kaplan-Meier method was conducted to estimate survivals for BF and exploratory univariate analysis and multivariable Cox regression for predictors of BF. Statistical significant if p<0.05. RESULTS: The mean (SD) patients’ age and PSA were 66 (7.5) years and 6.7 (7.8) ng/ml. Biopsy Gleason score was 6 (25%), 7 (39%), and 8 (36%). Clinical stage was cT1 (63%), cT2 (34%) and cT3 (4%). Mean (SD) estimated blood loss was 118 (61) ml and operative time was 166 (58) min. Mean (SD) length of hospital stay was 2.3 (3.7) days. There was no conversion to open approach. There was one (1%) intra-operative complication (rectal injury). Low grade (I-II) POC occurred in 7 (8%), high grade (III-IV)POC in4 (4%)and leakoncystogram in8 (9%)patients. PSM occurred in 17 (20%) patients. Pathology showed Gleason 8 in 52%, stage pT3 in 51%andpN1 in14%of the patients. In amedian (range) of 13 months (1-60 mon), BF occurred in 25 (28%) patients with a BF free-survival of 71%, 63% and 59% in 1, 2 and 3 years, respectively. No patient died. On univariate analysis pre-operative PSA, biopsy and pathology Gleason, cT and pT stage, and pN1 were predictors for BF. On multivariable analysis pre-operative PSA and pathology Gleason were independent predictors for BF. UC and EF were achieved in 48% and 11% within 1-year followup, respectively.Themain limitation is retrospectiveanalysis. CONCLUSIONS: To our knowledge, this is the largest cohort of sRARP demonstrating safety and feasibility. Operative time, perioperative morbidity, short-term cancer control and functional outcomes are acceptable and appear improved relative to historical data. This population remainsclinically challengingwithmajorityhaving local advanceddiseaseon final pathology. Gleason and PSA are independent predictors for BF.