Defining the Incremental Utility of Prostate Multiparametric Magnetic Resonance Imaging at Standard and Specialized Read in Predicting Extracapsular Extension of Prostate Cancer

Multiparametric magnetic resonance imaging (mpMRI) is increasingly used in staging early prostate cancer (PCa) but remains heavily reader-dependent. We aim to define the incremental utility of mpMRI over clinical parameters in determining the pathologic extracapsular extension (pECE) of PCa interpreted in a standard radiologic setting and when further over-read by a specialized reader. We retrospectively reviewed 120 men with clinically localized PCa undergoing mpMRI and radical prostatectomy. We obtained radiologic prediction of pECE from standard radiologic reports (standard read) and by a specialized reader blinded to clinical and pathologic findings (specialized read). We determined the incremental benefit of standard read and specialized read by sequential addition to a baseline clinical parameters-only logistic regression model predicting pECE. The sensitivity and specificity of standard read were 77% and 44%, respectively, whereas those of specialized read were 86% and 81%. The positive likelihood ratio was 1.7 at baseline, 1.7 adding standard read, and 6.5 adding specialized read. The negative likelihood ratio was 0.6 at baseline, 0.5 adding standard read, and 0.1 adding specialized read. Standard read modestly improved prediction of pECE, whereas specialized read improved it moderately. Patient summary: The incremental benefit of mpMRI over clinical information is small but increases to moderate with a specialized second opinion. This second opinion may be useful when considering active surveillance, nerve-sparing surgery, or focal therapy. (C) 2015 European Association of Urology. Published by Elsevier B.V. All rights reserved.