A negative multiparametric magnetic resonance imaging finding does not guarantee the absence of significant cancer among biopsy-proven prostate cancer patients: a real-life clinical experience

We analyzed the data of consecutive patients who had preoperative multiparametric magnetic resonance imaging (mpMRI) and underwent radical prostatectomy (RP) to evaluate the actual performance of mpMRI among biopsy-proven prostate cancer (PCa) patients in predicting favorable pathology in the real-life clinical setting. Among a total 730 biopsy-proven PCa patients underwent RP, the preoperative mpMRIs of 534 patients were positive, demonstrating one or more PI-RADs V2 grade ≥ 2 lesion(s). Other 196 mpMRIs were classified as negative, without any suspicious lesion. Pathology was classified to be unfavorable when showing Gleason score (GS) 4/5 or pT3/N1 features. Significant cancer was defined as non-organ-confined, GS 4/5, or cancer volume of ≥ 0.5 mL. Among a total 196 negative preoperative mpMRI patients, final RP pathology showed that 20 (10.2%) had pT3 disease and 2 (1.0%) had pN1 disease. Regarding the pathologic Gleason score, 117 (59.7%) had GS 3 + 4 and 44 (22.4%) had GS ≥ 4 + 3. The rate of a favorable PCa and an insignificant cancer was as low as 14.3% and 10.2%. Even among only the 101 D’Amico low-risk patients with negative MRI, the rates of a favorable pathology and an insignificant cancer were only 18.2% and 12.7%. The sensitivity, specificity, positive, and negative predictive value of mpMRI to predict a significant cancer were 74.3%, 45.5%, 95.5%, and 10.2%, respectively. In the real-life clinical setting, mpMRI demonstrated limited performance in the prediction of favorable and insignificant prostate cancer as a negative mpMRI could not guarantee the absence of unfavorable pathology among PCa patients.