Correlation of percent of core biopsies positive for prostate cancer and biochemical outcome following I-125 prostate brachytherapy or external beam radiation.

Abstract

40 Background: The percentage of positive cores (PPC) on biopsy for prostate cancer has been identified as a predictor of outcome following definitive local treatment. We aim to identify whether this observation holds true for a modern cohort of patients (pts) treated at Cleveland Clinic with permanent prostate brachytherapy (PB) or external beam radiation therapy (EBRT). Methods: We retrospectively reviewed pathology reports of pts treated with either PB or EBRT from our IRB-approved prospective prostate cancer registry. No pts underwent both PB and EBRT. The number of biopsy cores sampled, number of cores positive for prostate cancer, and maximum length of any core positive for prostate cancer were collected. Cox proportional hazards regression was used to analyze biochemical relapse free survival (bRFS) using the nadir + 2 ng/ml definition. Results: We identified 1253 PB and 879 EBRT pts with complete pathology and clinical information. Among PB pts, 46% were low risk, 40% intermediate risk, and 14% high risk, while 78% had <50% PPC, and 22% had >=50% PPC. The 5-year bRFS for PB was 92.0% for <50% PPC, vs. 83.1% for >=50% PPC (HR 2.1, p=0.0005). For PB pts, significant predictors of bRFS on univariate analysis included: PPC, clinical T stage, PSA, biopsy Gleason score, androgen deprivation, and frequency of PSA testing. On multivariate analysis, only PPC, biopsy Gleason score, and PSA frequency remained significant predictors following PB. Among EBRT pts, 11% were low risk, 36% intermediate risk, and 53% high risk, while 55% had <50% PPC, and 45% had >=50% PPC. The 5-year bRFS for EBRT was 85.6% for <50% PPC, vs. 77.1% for >=50% PPC (HR 1.8, p<0.0001). For EBRT pts, significant predictors of bRFS on univariate analysis included: PPC, clinical T stage, PSA, biopsy Gleason score, androgen deprivation, EBRT dose, and frequency of PSA testing. On multivariate analysis, only PPC, biopsy Gleason score, and PSA frequency remained significant predictors following EBRT. Conclusions: Following PB or EBRT, the percent of positive cores for prostate cancer was a significant predictor of bRFS on multivariate analysis, more so than conventional predictors such as T stage and PSA. No significant financial relationships to disclose.