Low accuracy of routine ultrasound-guided systematic 12-core biopsies in prostate tumor mapping

Abstract

Introduction: To determine the accuracy of a 12-core biopsy protocol in assessing the location of prostate tumors within radical prostatectomy (RP) specimens.

Materials and methods: A consecutive series of patients with T1c stage prostate cancer who had undergone 12 ultrasound-guided prostate biopsies prior to RP was considered. The locations of the biopsies from prostate gland mapping were compared with the locations of tumor tissues obtained after analysis of the prostate specimens.

Results: Overall, 78 patients (27.4%) were included. The median PSA level was 6 ng/mL. The median prostate weight was 45 g (range 22 to 102). Overall, 936 biopsies were performed in the 78 men, of which 254 biopsies were positive. The mean number of positive biopsies per patient was 3.7 (range 1 to 12). Pathologic examination of the surgical specimens revealed that 58 (74.4%) patients had pT2 disease and 20 patients (25.6%) had locally advanced disease (pT3). The biopsy protocol's sensitivity, specificity and positive predictive value for tumor location were 0.34, 0.83 and 0.84. The performance of the protocol was modest in assessing the exact tumor location (area under curve (AUC) 0.581, 95% confidence interval (CI) 0.489-0.719).

Conclusions: Routine, ultrasound-guided, systematic 12-core biopsies lack precision in prostate tumor mapping.