Nomograms Combining PHI and PI-RADS in Detecting Prostate Cancer: A Multicenter Prospective Study

Area of Science:

• Urology
• Oncology
• Medical Informatics

Background:

• Prostate cancer (PCa) detection rates need improvement, especially in Asian populations.
• Current diagnostic methods may lead to unnecessary biopsies.
• Development of accurate predictive models is crucial.

Purpose of the Study:

• To construct and validate nomograms for detecting PCa and clinically significant PCa (CSPCa) in Asian men.
• To identify key predictive factors for PCa and CSPCa.
• To create web-based dynamic nomograms for clinical use.

Main Methods:

• A multicenter prospective study involving 293 patients.
• Univariable and multivariable logistic regression for risk factor identification and nomogram construction.
• Assessment of nomogram performance using discrimination, calibration, and clinical utility metrics.
• Development of web-based dynamic nomograms.

Main Results:

• Four independent predictors for PCa: age, Prostate Health Index (PHI), prostate volume (PV), and PI-RADS score.
• Four independent predictors for CSPCa: age, PHI, PI-RADS score, and Log PSA Density (PSAD).
• High Area Under the ROC Curve (AUC) values for PCa (0.902 training, 0.869 validation) and CSPCa (0.896 training, 0.890 validation).

Conclusions:

• Combining PHI and PI-RADS scores improves PCa and CSPCa detection rates and can reduce unnecessary biopsies.
• Nomograms incorporating age, PHI, PV, and PI-RADS enhance PCa detection.
• Nomograms incorporating age, PHI, PI-RADS, and Log PSAD improve CSPCa detection.