Clinical Significance of Multiparametric Magnetic Resonance Imaging as a Preoperative Predictor of Oncologic Outcome in Very Low-Risk Prostate Cancer

Abstract

Currently, multiparametric magnetic resonance imaging (mpMRI) is not an indication for patients with very low-risk prostate cancer. In this study, we aimed to evaluate the usefulness of mpMRI as a diagnostic tool in these patients. We retrospectively analyzed the clinical and pathological data of individuals with very low-risk prostate cancer, according to the NCCN guidelines, who underwent mpMRI before radical prostatectomy at our institution between 2010 and 2016. Patients who did not undergo pre-evaluation with mpMRI were excluded. We analyzed the factors associated with biochemical recurrence (BCR) using Cox regression model, logistic regression analysis, and Kaplan⁻Meier curve. Of 253 very low-risk prostate cancer patients, we observed 26 (10.3%) with BCR during the follow-up period in this study. The median follow-up from radical prostatectomy was 53 months (IQR 33⁻74). The multivariate Cox regression analyses demonstrated that the only factor associated with BCR in very low-risk patients was increase in the pathologic Gleason score (GS) (HR: 2.185, p-value 0.048). In addition, multivariate logistic analyses identified prostate specific antigen (PSA) (OR: 1.353, p-value 0.010), PSA density (OR: 1.160, p-value 0.013), and suspicious lesion on mpMRI (OR: 1.995, p-value 0.019) as the independent preoperative predictors associated with the pathologic GS upgrade. In our study, the pathologic GS upgrade after radical prostatectomy in very low-risk prostate cancer patients demonstrated a negative impact on BCR and mpMRI is a good prognostic tool to predict the pathologic GS upgrade. We believe that the implementation of mpMRI would be beneficial to determine the treatment strategy for these patients.