A Single Center Evaluation of the Diagnostic Accuracy of Multiparametric Magnetic Resonance Imaging against Transperineal Prostate Mapping Biopsy: An Analysis of Men with Benign Histology and Insignificant Cancer following Transrectal Ultrasound Biopsy.

PURPOSE: In this study we evaluated the diagnostic performance of transrectal ultrasound guided biopsy and multiparametric magnetic resonance imaging to detect prostate cancer against transperineal prostate mapping biopsy as the reference test.

MATERIALS AND METHODS: Transrectal ultrasound guided biopsy, multiparametric magnetic resonance imaging and transperineal prostate mapping biopsy were performed in 426 patients between April 2012 and January 2016. Patients initially underwent systematic 12 core transrectal ultrasound guided biopsy followed 3 months later by 1.5 Tesla, high resolution T2, diffusion-weighted, dynamic contrast enhanced multiparametric magnetic resonance imaging. Two specialist uroradiologists blinded to the results of transperineal prostate mapping biopsy allocated a PI-RADS™ (Prostate Imaging-Reporting and Data System) score to each multiparametric magnetic resonance imaging study. Transperineal prostate mapping biopsy with 5 mm interval sampling, which was performed within 6 months of multiparametric magnetic resonance imaging, served as the reference test.

RESULTS: Transrectal ultrasound guided biopsy identified 247 of 426 patients with prostate cancer and 179 of 426 with benign histology. Transperineal prostate mapping biopsy detected prostate cancer in 321 of 426 patients. On transperineal prostate mapping biopsy 94 of 179 patients with benign transrectal ultrasound guided biopsy had prostate cancer and 95 of 247 with prostate cancer on transrectal ultrasound guided biopsy were identified with cancer of higher grade. Using a multiparametric magnetic resonance imaging PI-RADS score of 3 or greater to detect significant prostate cancer, defined as any core containing Gleason 4 + 3 or greater prostate cancer on transperineal prostate mapping biopsy, the ROC AUC was 0.754 (95% CI 0.677-0.819) with 87.0% sensitivity (95% CI 77.3-97.0), 55.3% specificity (95% CI 50.2-60.4) and 97.1% negative predictive value (95% CI 94.8-99.4).

CONCLUSIONS: Multiparametric magnetic resonance imaging is a more accurate diagnostic test than transrectal ultrasound guided biopsy. However, a significant proportion of ISUP (International Society of Urological Pathology) Grade Group 2 prostate cancer remained undetected following multiparametric magnetic resonance imaging. Although multiparametric magnetic resonance imaging could avoid unnecessary biopsy in many patients with ISUP Grade Group 3 or greater prostate cancer, at less stringent definitions of significant cancer a substantial proportion of prostate cancer would remain undetected after multiparametric magnetic resonance imaging.