Negative multiparametric magnetic resonance imaging for prostate cancer: what's next?

Multiparametric magnetic resonance imaging (mpMRI) of the prostate has excellent sensitivity in detecting clinically significant prostate cancer (csPCa). Nevertheless, the clinical utility of negative mpMRI (nMRI) is less clearMultiparametric magnetic resonance imaging (mpMRI) of the prostate has excellent sensitivity in detecting clinically significant prostate cancer (csPCa). Nevertheless, the clinical utility of negative mpMRI (nMRI) is less clear. OBJECTIVE: To assess outcomes of men with nMRI and clinical follow-up after 7 yr of activity at a reference center. DESIGN, SETTING, AND PARTICIPANTS: All mpMRI performed from January 2010 to May 2015 were reviewed. We selected all patients with nMRI and divided them in group A (naïve patients) and group B (previous negative biopsy). All patients without a diagnosis of PCa had a minimum follow-up of 2 yr and at least two consecutive nMRI. Patients with positive mpMRI were also identified to assess their biopsy outcomes. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: A Kaplan-Meier analysis was performed to assess both any-grade PCa and csPCa diagnosis-free survival probabilities. Univariable and multivariable Cox regression models were fitted to identify predictors of csPCa diagnosis. RESULTS AND LIMITATIONS: We identified 1545 men with nMRI, and 1255 of them satisfied the inclusion criteria; 659 belonged to group A and 596 to group B. Any-grade PCa and csPCa diagnosis-free survival probabilities after 2 yr of follow-up were 94% and 95%, respectively, in group A; in group B, they were 96%. After 48 mo of follow-up, any-grade PCa diagnosis-free survival probability was 84% in group A and 96% in group B (log rank p<0.001). Diagnosis-free survival probability for csPCa was unchanged after 48 mo of follow-up. On multivariable Cox regression analysis, increasing age (p=0.005) was an independent predictor of lower csPCa diagnosis probability, while increasing prostate-specific antigen (PSA) and PSA density (<0.001) independently predicted higher csPCa diagnosis probability. The prevalence of and positive predictive value for csPCa were 31.6% and 45.5%, respectively. Limitations include limited follow-up and the inability to calculate true csPCa prevalence in the study population. CONCLUSIONS: mpMRI is highly reliable to exclude csPCa. Nevertheless, systematic biopsy should be recommended even after nMRI, especially in younger patients with high or raising PSA levels.