We compared the prognostic capabilities and clinical utility of using a cell cycle progression (CCP) gene expression classifier test, multiparametric MRI with PI-RADS scoring, and clinicopathologic features in a group of men newly diagnosed with prostate cancer (Cohort 1) and a group of men on active surveillance (Cohort 2).
Small, but statistically significant, correlations were found between PI-RADS and CCP, PI-RADS and CAPRA (UCSF Cancer of the Prostate Risk Assessment) score, as well as PI-RADS and CCR (clinical cell-cycle risk) score. CCR is a combination of CCP and CAPRA scores. These small correlations suggest that the prognostic information captured by these variables is at least somewhat independent.
Univariate and multivariable analyses evaluated the ability of CCP, CAPRA, and PI-RADS to predict post-radical prostatectomy Gleason scores in men undergoing definitive treatment. CCP and CAPRA were significant predictors of higher-grade tumors, but PI-RADS added no significant additional information.
When evaluating CCP, PI-RADS, and prostate-specific antigen (PSA) with tumor grade in diagnostic biopsies, only CCP and PI-RADS correlated significantly with tumor grade across all patients. Only CCP and CCR scores, as well as CAPRA, showed a significant impact on the decision for active surveillance versus definitive treatment.
Although it seems clear that mpMRI and PI-RADS scoring is useful in prostate cancer diagnosis, these results question its utility as a prognostic indicator.
Written by: David S Morris, J Scott Woods, Byard Edwards, Lauren Lenz, Jennifer Logan, Darl D Flake, Brent Mabey, Jay T Bishoff, Todd Cohen, Steven Stone
Urology Associates, PC, Nashville, TN. Electronic address: ., Urology Associates, PC, Nashville, TN., Premier Radiology, St. Thomas Health, Nashville, TN., Myriad Genetics, Inc., Salt Lake City, UT., Intermountain Healthcare, Salt Lake City, UT.
Read the Abstract