Multi‐institutional nomogram predicting benign prostate pathology on magnetic resonance/ultrasound fusion biopsy in men with a prior negative 12‐core systematic biopsy

• Published in: Cancer Vol. 124; no. 2; pp. 278 - 285
• Main Authors: Truong, Matthew, Wang, Bokai, Gordetsky, Jennifer B., Nix, Jeffrey W., Frye, Thomas P., Messing, Edward M., Thomas, John V., Feng, Changyong, Rais‐Bahrami, Soroush
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 15.01.2018
• Subjects: Benign; Biopsy; Cancer; Decision analysis; Decision theory; diagnostic imaging; early detection of cancer; Health care facilities; Magnetic resonance imaging; Multivariate analysis; NMR; Nomograms; Nuclear magnetic resonance; Oncology; Pathology; Patients; Prostate; Prostate cancer; Prostate-specific antigen; prostatic neoplasms; Resonance; Ultrasonic imaging; Ultrasound; magnetic resonance imaging; early detection of cancer; nomograms; diagnostic imaging; prostatic neoplasms
• ISSN: 0008-543X; 1097-0142; 1097-0142
• DOI: 10.1002/cncr.31051

BACKGROUND Prostate multiparametric magnetic resonance imaging (mpMRI) may be recommended for patients with a prior negative systematic biopsy (SB). However, a proportion of these patients will continue to have no prostate cancer (PCa) identified on magnetic resonance/ultrasound fusion biopsy (FB) despite abnormal mpMRI findings. METHODS In this multi‐institutional, retrospective study, clinical and mpMRI parameters were assessed for 285 consecutive patients with at least 1 prior negative biopsy who underwent FB for a Prostate Imaging Reporting and Data System (PI‐RADS) score of 3 to 5 at the University of Rochester Medical Center from December 2014 to December 2016, or at the University of Alabama at Birmingham from February 2014 to February 2017. Nomograms were generated for predicting benign prostate pathology on both the targeted biopsy and the concurrent SB. RESULTS Benign pathology was found in 132 of 285 patients (46.3%). In a multivariate analysis, the predictors of benign prostate pathology on FB were age, prostate‐specific antigen, prostate volume, and PI‐RADS score. The predicted probabilities were plotted on a receiver operating characteristic curve, and the area under the curve was 0.825. The nomogram demonstrated excellent calibration and a high net benefit in a decision curve analysis. With a theoretical cutoff probability of ≥0.7 used to recommend deferment of FB, 61 of 285 patients (21.4%) would have avoided an unnecessary biopsy, and only 4 of 285 patients (1.4%) with PCa with a Gleason score ≥ 3 + 4 would have been missed. CONCLUSIONS False‐positive mpMRI examinations may occur in up to 46.3% of patients with a prior negative biopsy. Thus, a multi‐institutional nomogram has been developed and validated for predicting benign pathology after FB in patients with a prior negative biopsy, and this may help to reduce the number of unnecessary biopsies in the setting of abnormal mpMRI findings. Cancer 2018;124:278‐85. © 2017 American Cancer Society. Prostate multiparametric magnetic resonance imaging is subject to false‐positives. A multi‐institutional nomogram for predicting benign pathology after magnetic resonance/ultrasound fusion–targeted biopsy can reduce the number of unnecessary biopsies.