Feed

IntroductionBiparametric magnetic resonance imaging (MRI) preserves enough information to enable the prediction of prostate cancer (PCa). This fast, cost-effective, and non-invasive modality includes acquisition of T2-weighted images, and accelerated diffusion-weighted imaging (DWI) sequences with corresponding apparent diffusion coefficient (ADC) maps. In this proof-of-concept study, we aimed to assess the prediction of PCa using a tumor location-(L) dependent risk score (LADCT₂) generated from an ADC and T2 images - based model.MethodsThe single-center institutional retrospective cohort study used 113 patients who underwent multiparametric MRI (mpMRI) for the diagnosis and management of PCa. A discovery cohort (n = 58) and an evaluation cohort (n = 55) were identified from a prospectively maintained institutional cancer registry. The discovery cohort consisted of patients who underwent MRI-guided TRUS biopsies, whereas the evaluation cohort consisted of patients who received only standard TRUS biopsy. Among the discovery cohort, we developed a predictive risk score (LADCT₂) using a multivariable logistic regression model that incorporated tumor location (L) with normalized mean signal differences of T2-and ADC- tumor region of interest. The primary outcome assessed the predictive accuracy of the LADCT₂ risk score in predicting PCa.ResultsOur results demonstrated that the LADCT₂ score exhibited excellent predictive accuracy for PCa among both the evaluation (AUC = 0.84, OR = 2.80 [95% CI, 1.04-7.52]; P = .04), and discovery (AUC = 0.77, OR = 2.71 [95% CI, 1.38-5.35]; P = .003) cohorts. Additionally, it also predicted for clinically significant PCa among both the discovery (AUC = 0.71, OR = 2.11 [95% CI, 1.16-3.84]; P = .01), and evaluation (AUC = 0.65, OR = 1.94 [95% CI, 1.02-3.69]; P = .04) cohorts.ConclusionThe novel LADCT₂ risk score may function as an effective risk stratification tool to support clinical decision-making in the management of PCa.