A Prostate Imaging-Reporting and Data System version 2.1-based predictive model for clinically significant prostate...

Area of Science:

• Radiology
• Oncology
• Urology

Background:

• Prostate cancer diagnosis relies on accurate risk stratification.
• Multiparametric magnetic resonance imaging (mpMRI) with Prostate Imaging-Reporting and Data System (PI-RADS) scoring improves detection.
• Validated predictive models are needed to integrate clinical and imaging data for optimal decision-making.

Purpose of the Study:

• To develop and validate a PI-RADS v2.1-based predictive model for clinically significant prostate cancer (csPCa).
• To integrate clinical and mpMRI data for enhanced diagnostic accuracy.
• To compare the novel model's performance against existing clinical and MRI-based algorithms.

Main Methods:

• Retrospective analysis of 1319 patients undergoing mpMRI and biopsy (April 2019-December 2023).
• Development of a 'Clinical Baseline' model (demographics, labs) and an 'MRI Added' model (including PI-RADS v2.1 scores, prostate volumes).
• Internal and external validation of models; comparison with prior algorithms using AUC and decision curve analysis.

Main Results:

• The 'MRI Added' model showed significantly improved discriminative ability (AUC_internal 0.88, AUC_external 0.79) versus the 'Clinical Baseline' model (AUC_internal 0.75, AUC_external 0.68).
• The 'MRI Added' model reduced unnecessary biopsies by 27% (internal) and 10% (external) at a 25% risk threshold.
• No significant difference was found between the new model and comparative PI-RADS v2/v1 models in predictive ability or biopsy reduction.

Conclusions:

• The PI-RADS v2.1-based mpMRI model significantly enhances csPCa prediction accuracy.
• This integrated approach effectively reduces unnecessary biopsies, improving patient management.
• The model shows promise for diverse populations, offering an updated strategy for prostate cancer detection.