Association of PI-RADS Score with Clinically Significant Prostate Cancer and Histopathological Upgrading in Patients Undergoing Radical Prostatectomy: A Retrospective Single-Centre Study

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Background and Objectives: Multiparametric magnetic resonance imaging (mpMRI) is widely used in prostate cancer diagnosis, but biopsy findings may still underestimate tumor aggressiveness. This study evaluated the association of Prostate Imaging Reporting and Data System (PI-RADS) assessment with clinically significant prostate cancer (csPCa) and histopathological upgrading after radical prostatectomy. Methods: The retrospective analysis comprised 564 patients who underwent mpMRI–transrectal ultrasound fusion-guided biopsy followed by radical prostatectomy between 2021 and 2025. All included patients subsequently underwent radical prostatectomy, so the cohort represents a surgically selected population. Clinical variables, PI-RADS score, biopsy ISUP Grade Group, and final prostatectomy pathology were assessed. Clinically significant disease was defined as ISUP Grade Group ≥ 2. Histopathological upgrading was defined as an increase of at least one ISUP Grade Group between biopsy and final prostatectomy pathology. Results: csPCa was found in 385 patients (68.3%). Patients with csPCa had higher prostate-specific antigen density (PSAD) values than those without csPCa. PI-RADS score was significantly associated with csPCa and remained independently associated with csPCa in multivariable analysis. Histopathological upgrading occurred in 164 patients (29.1%), and PI-RADS score was also independently associated with upgrading. Model discrimination was limited. Conclusions: Preoperative mpMRI with PI-RADS assessment was associated with both csPCa and biopsy underestimation. Although PI-RADS was independently associated with both outcomes, its modest discriminatory performance indicates that it should be interpreted together with other clinical and pathological parameters rather than as a standalone predictive tool.