Prebiopsy Biparametric MRI for Clinically Significant Prostate Cancer Detection With PI-RADS Version 2: A Multicenter Study.

Abstract

The purpose of this study was to evaluate the diagnostic performance of the Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) with respect to prebiopsy MRI with and without dynamic contrast enhancement in the detection of clinically significant cancer (CSC). A total of 113 patients with prostate cancer who underwent radical prostatectomy and prebiopsy multiparametric 3-T MRI (mpMRI) that included T2-weighted imaging, DWI, and dynamic contrast-enhanced MRI (DCE-MRI) were enrolled in a retrospective study conducted at two institutions. For detecting CSC at prebiopsy mpMRI with DCE-MRI and biparametric MRI (bpMRI) without DCE-MRI, two independent radiologists using PI-RADSv2 scored suspicious lesions in all patients. CSC was identified in 74.3% (84/113) of patients. For CSC detection rate, no statistical differences between bpMRI and mpMRI were found for any PI-RADS score (p > 0.05). For cancer in the peripheral zone, reader 1 upgraded 22 lesions and reader 2 upgraded 13 lesions from PI-RADS score 3 at bpMRI to PI-RADS 4 (3 + 1) at mpMRI. The CSC detection rate of PI-RADS 3 + 1 lesions at mpMRI (reader 1, 63.6%; reader 2, 69.2%) was slightly greater than that of PI-RADS 3 lesions at bpMRI (reader 1, 53.8%; reader 2, 60.0%), which was not statistically different (p > 0.05). Interreader agreement on PI-RADS scoring was moderate for both bpMRI (κ = 0.540) and mpMRI (κ = 0.478). For detecting CSC, the diagnostic performance of prebiopsy bpMRI without DCE-MRI is similar to that of mpMRI with DCE-MRI.