Accuracy of Multiparametric Magnetic Resonance Imaging for Extracapsular Extension of Prostate Cancer in Community Practice

Introduction: The presence of extracapsular extension (ECE) in prostate cancer (PCa) can influence a surgeon's decision to perform a nerve-sparing approach during radical prostatectomy (RP). Preoperatively, multiparametric MRI (mp-MRI) is often used to stage PCa. More recently, the use of mp-MRI has gained wide acceptance in fusion biopsy of the prostate. In this framework, the reported accuracy of mp-MRI has been highly variable, with data often originating from large referral centers with experienced radiologists. We sought to determine the sensitivity and specificity of mp-MRI for detecting ECE in the community.

Materials and methods: We reviewed a prospectively maintained database of men with PCa who had undergone RP. We recorded the prevalence of ECE at RP and determined the sensitivity, specificity, positive predictive value, and negative predictive value of MRI for detecting ECE. We assessed these values according to the D'Amico risk groups and compared the predictive value of MRI to that of the Partin tables.

Results: The prevalence of ECE was 11.5%, 28.1%, and 47.1% in the low-, intermediate, and high-risk groups, respectively, with an overall prevalence of 24.1%. The overall sensitivity, specificity, positive predictive value, and negative predictive value of MRI was 12.5%, 93.1%, 36.4%, and 77.0%, respectively.

Conclusion: The reduction in the sensitivity of preoperative mp-MRI to determine ECE in the community setting is significant. Even with stratification using the D'Amico criteria and Partin tables, the performance of mp-MRI was not significantly improved. Because most cases of PCa are diagnosed and treated in the community, it is questionable whether mp-MRI is a suitable staging modality in the community.