LAUSR

PurposeTo clarify clinicopathological features of false-negative clinically significant prostate cancer (csPC) at multiparametric prostate MRI (mpMRI).Methods95 patients with 139 csPC undergoing 3T mpMRI before radical prostatectomy were included. Two radiologists were independently evaluated mpMR images using PI-RADS v2. Clinicopathological findings were compared between (a) detectable and undetectable lesions using overall mpMRI criteria (o-mpMRI criteria) and (b) lesions with early enhancement effect (EEE) and lesions without EEE at DCE-MRI.ResultsThe detection rate of csPS using cutoff value of category 3 or more in PI-RADS v2 for positive lesion was 72.1% (98/136 lesions). In 38 false-negative lesions with less than PI-RADS v2 category 3, the DCE-MRI detected 14 lesions. 17 undetectable lesions on o-mpMR criteria had lower PSA and D’amico risk classification, and higher tumor apparent diffusion coefficient (ADC) than those of 118 detectable lesions (p ≤ 0.048). 89 lesions with EEE showed higher PSA, tumor size, prostatectomy GS grade, frequency of lesions with GS ≥ 4 + 3 and lower tumor ADC than those in 38 lesions without EEE (p ≤ 0.046).ConclusionTumor detectability of csPC with PI-RADS v2 was limited compared with o-mpMRI criteria in Japanese men. Moreover, false-negative lesions on o-mpMRI criteria were characterized as small in size, low risk and low cellularity.