Oncological results of radical prostatectomy depending on the data of multiparametric magnetic resonance imaging and patient selection for nerve-sparing technique

Abstract

Background. Selecting patients for radical prostatectomy (RP) using nerve-sparing technique (NST) is crucial to optimize oncological and functional results. Correlation of multiparametric magnetic resonance imaging (mpMRI) data with RP results remains insufficiently studied, at the same time existing prognostic tools and nomograms show moderate effectiveness during third-party validation and have some drawbacks. Objective: to study pathomorphological results and evaluate the recurrence-free survival of patients after RP, depending on mpMRI data; develop a patient selection algorithm for NST. Materials and methods. The study included 95 patients with clinically localized prostate cancer (PCa), who underwent RP within 2012—2017. All mpMRI series were retrospectively reviewed and evaluated using the Prostate Imaging Reporting and Data System (PI-RADS) version 2 (v2) by one radiologist diagnostician who neither had access to database and clinical information about patients nor participated in data collection and statistical analysis. Patients were divided into 2 groups: low probability of PCa and suspected PCa (PI-RADS 2—3; n = 43); high and very high probability of PCa (PI-RADS 4—5; n = 52). We assessed the presence of positive surgical margins, as well as extracapsular extension and relapse-free survival. We also developed an algorithm to select patients for nerve-sparing PCa based on the PI-RADS category according to mpMRI data. Results. PI-RADS 2—3 group showed less positive surgical margins as compared to PI-RADS 4—5group (2.3 % versus 21.2 %; p = 0.025), as well as no cases of tumor extracapsular extension versus 36.5 % in PI-RADS 4—5group (p <0.001). Patients with Gleason score 6 after demonstrated the same trends: extracapsular extension on the tumor side was observed in 0 and 33.3 % of cases (p <0.001), positive surgical margins — in 2.4 and 15.2 % of cases (p = 0.046), respectively. An increase of the Gleason score after RP was observed in 12.2 % of patients of PI-RADS 2—3 group and in 30.3 % of PI-RADS 4—5 group (p = 0.04). Recurrence-free survival after 60 months was 93.0 and 71.1 %, respectively (p = 0.015). Conclusion. Risk categories for PCa according to mpMRI data are associated with pathomorphological results and recurrence-free survival after RP. Using PI-RADS v2 categories in the patient selection algorithm for NST optimizes the assessment of oncological safety and allows selecting a group of patients for a thorough individual analysis of the benefit/risk profile.