Prediction of biochemical recurrence after radical prostatectomy with PI-RADS version 2 in prostate cancers: initial results.

Abstract

To determine whether the Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) helps predict biochemical recurrence (BCR) after radical prostatectomy for prostate cancer (PCa). We included 158 patients with PCa who underwent magnetic resonance imaging (MRI) and radical prostatectomy (RP). Clinical (prostate-specific antigen, greatest percentage of core, and percentage of positive core number), PI-RADSv2 score on MRI, and surgical parameters (Gleason score, extracapsular extension, seminal vesicle invasion, and tumour volume) were investigated. Univariate and multivariate analyses using Cox's proportional hazards model were performed to assess parameters predictive of BCR (two consecutive prostate specific antigens ≥0.2 ng/ml). Kaplan-Meier survival curves were analyzed. The rate of BCR was 13.3 % (21/158) after surgery (median follow-up, 25 months; range, 12-36). No subject with a PI-RADS score <4 had BCR. In univariate analysis, all parameters were significant for BCR (p < 0.05), except seminal vesicle invasion (p = 0.254). Meanwhile, PI-RADS score was the only independent parameter for BCR in multivariate analysis (p < 0.05). Two-year, BCR-free survival post-RP was significantly lower for PI-RADS ≥4 (84.7-85.5 %) than for PI-RADS <4 (100 %; p < 0.05). As a preoperative imaging tool, PI-RADSv2 may be useful to predict BCR after radical prostatectomy for PCa. • No subject with PI-RADS <4 had BCR after RP • PI-RADSv2 was the only predictor of BCR in multivariate analysis • Two-year, BCR-free survival following RP was lower for PI-RADS≥4 than for PI-RADS<4 • Inter-rater agreement was good for PI-RADS ≥4 or not.