Followup of Men with PI-RADS™ 4 or 5 Abnormality on Prostate Magnetic Resonance Imaging and Nonmalignant Pathological Findings on Initial Targeted Prostate Biopsy

Purpose:A benign magnetic resonance imaging targeted prostate biopsy in the setting of a PI-RADS™ 4/5 abnormality presents a clinical dilemma for future management. We evaluated benign histological features on magnetic resonance imaging targeted prostate biopsy to determine if they predict the likelihood of missed cancer on subsequent biopsy.Materials and Methods:Between June 2012 and September 2016, 1,595 men were enrolled in a prospective study of magnetic resonance imaging targeted and systematic biopsy outcomes. We re-reviewed pathology from benign biopsies of PI-RADS 4/5 abnormalities and divided them into 5 groups for comparison to outcomes of clinical followup: inflammation (38%), stroma/glandular hyperplasia (9%), normal prostate tissue (28%), atypical small acinar proliferation/high grade prostatic intraepithelial neoplasia (9%) and cancer in adjacent systematic cores (16%).Results:Of 497 men 88 (18%) with PI-RADS 4/5 abnormality prior to initial biopsy had no cancer on magnetic resonance imaging targeted prostate biopsy. On followup, 45 men underwent repeat magnetic resonance imaging: 12 (27%) had persistent PI-RADS 4/5 abnormalities, 17 (38%) had PI-RADS 2/3, 16 (35%) had PI-RADS 1. On repeat magnetic resonance imaging targeted prostate biopsy, cancer was found in 62.5% of men with PI-RADS 4/5 and 23% of men with PI-RADS 2/3. Histological groups on initial biopsy were not predictive of the likelihood of PI-RADS downgrade on repeat magnetic resonance imaging or cancer detection on repeat biopsy.Conclusions:Among men with no cancer on magnetic resonance imaging targeted prostate biopsy performed for PI-RADS abnormality, downgrade of PI-RADS score is noted in 73% on repeat magnetic resonance imaging. Persistence of PI-RADS 4/5 predicts a higher risk of missed cancer, warranting prompt re-biopsy. While histological findings such as inflammation may underlie some PI-RADS 4/5 abnormalities, initial histology is a poor predictor of cancer likelihood on repeat biopsy.