PREDICTION OF EXTRAPROSTATIC EXTENSION IN THE NEUROVASCULAR BUNDLE BASED ON PROSTATE NEEDLE BIOPSY PATHOLOGY, SERUM PROSTATE SPECIFIC ANTIGEN AND DIGITAL RECTAL EXAMINATION

Publisher Website
Google Scholar
Abstract
Purpose: There are few studies on predictors of extraprostatic extension (EPE) in the region of the neurovascular bundle (NVB). We investigated whether clinical information and prostate biopsy data could predict EPE of clinical localized prostate cancer. Materials and Methods: Through a retrospective analysis of the pathology database we identified 2,660 cases of clinically localized prostate cancer treated with radical retropubic prostatectomy without preoperative adjuvant therapy at The Johns Hopkins Hospital. The study sample involved a total of 3,006 lobes with prostate cancer including 2,070 with organ confined disease, 620 with EPE in the NVB at the posterolateral edge of the prostate and 316 with EPE in a region other than the NVB (EPE elsewhere). Through univariate and multivariate logistic regression analysis we determined whether patient age, year of surgery, serum prostate specific antigen, digital rectal examination, biopsy highest Gleason score, perineural invasion, percent of side specific biopsy cores with cancer, percent of each core involved with cancer and the maximum percent of a core involved with cancer was predictive of EPE in the NVB. Results: Prostate specific antigen (10 or greater vs less than 10), biopsy Gleason score (7 or greater vs 6 or less), digital rectal examination (abnormal vs normal), percent of side specific cores with tumor (greater than 33.3% vs 33.3% or less) and average percent involvement of each positive core (greater than 20% vs 20% or less) were all found to be statistically significant independent predictors of NVB penetration in multivariate analysis. The generated model stratifies each of these variables into high and low risk. The probability of EPE in the NVB was less than 10% in cases with 1 or fewer of the higher risk variables and was 10% or greater in cases with more than 1 of the higher risk variables. Conclusions: The model generated in this study allows for the preoperative identification of patients with 10% or greater probability of EPE in the NVB. Our algorithm will help provide objective parameters that aid in the decision to spare the NVB safely.