Percentage of positive biopsies predicts lymph node involvement in men with low-risk prostate cancer undergoing radical prostatectomy and extended pelvic lymphadenectomy.

Abstract

To evaluate preoperative predictive risk factors associated with lymph node metastases (LNM) in a cohort of low-risk prostate cancer (PCA) patients. The charts of 499 patients were retrospectively reviewed to identify prognostic risk factors for the presence of LNM. Pathohistological data and Gleason score of the radical prostatectomy (RP) specimen, number of removed nodes, number of positive lymph nodes, and anatomical distribution of LNM were tabulated and evaluated. A correlation between clinical stage, preoperative serum prostate-specific antigen (PSA), biopsy Gleason score, number of biopsies taken, percentage of positive biopsies and the presence of LNM were calculated. All 499 men underwent retropubic RP and extended pelvic lymphadenectomy (EPLND). LNM were identified in 29 (5.8%) patients. A prediction model based on clinical stage, PSA, and biopsy Gleason score had a predictive accuracy of 79.2%. The addition of number of positive biopsies and % positive cores improved its predictive accuracy to 81.5% and 87.8%, respectively. The predicted frequency of LNM by the original nomogram was 7.4% and differed by less than 3% with the actual observation of LNM. The predictive accuracy of the nomogram was 81.5% as compared with 87.8% of the prediction model of this study. The percentage of positive cores involved with PCA is the most reliable predictor of LNM and indicates the need for EPLND. The Briganti nomogram has been validated and a general applicability for predicting the presence of LNM was proven.