Introducing a prognostic score for pretherapeutic assessment of seminal vesicle invasion in patients with clinically localized prostate cancer.

PURPOSE: To identify prostate cancer patients who will have the most likely benefit from sparing the seminal vesicles during 3D conformal radiation therapy. METHODS AND MATERIALS: From 1988 to 2001, 532 patients underwent radical prostatectomy for clinically localized prostate cancer. Primary endpoint was the pathological evidence of seminal vesicle invasion. Variables for univariate and multivariate analyses were age, prostate weight, clinical stage, PSA level, Gleason score, number and site of positive prostate sextant biopsies. Multivariate logistic regression with backward stepwise variable selection was used to identify a set of independent predictors of seminal vesicle invasion, and the variable selection procedure was validated by non-parametric bootstrap. RESULTS: Seminal vesicle invasion was reported in 14% of the cases. In univariate analysis, all variables except age and prostate weight were predictors of seminal vesicle invasion. In multivariate analysis, only the number of positive biopsies (P<0.0001), Gleason score (P<0.007) and PSA (P<0.0001) were predictors for seminal vesicles invasion. Based on the multivariate model, we were able to develop a prognostic score for seminal vesicle invasion, which allowed us to discriminate two patient groups: A group with low risk of seminal vesicles invasion (5.7%), and the second with a higher risk of seminal vesicles invasion (32.7%). CONCLUSIONS: Using the number of positive biopsies, Gleason score and PSA, it is possible to identify patients with low risk of seminal vesicles invasion. In this population, seminal vesicles might be excluded as a target volume in radiation therapy of prostate cancer.