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

Abstract

OBJECTIVE\nThe 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).\n\n\nMATERIALS AND METHODS\nA 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.\n\n\nRESULTS\nCSC 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).\n\n\nCONCLUSION\nFor detecting CSC, the diagnostic performance of prebiopsy bpMRI without DCE-MRI is similar to that of mpMRI with DCE-MRI.