Comparative validation of nomograms predicting clinically insignificant prostate cancer

Area of Science:

• Urology
• Oncology
• Medical Informatics

Background:

• Accurate prediction of insignificant prostate cancer is crucial for treatment decisions.
• Existing nomograms require validation in contemporary patient cohorts and diverse tumor locations.

Purpose of the Study:

• To validate and compare the accuracy of prostate cancer nomograms.
• To analyze nomogram performance based on tumor location (anterior-apical vs. posterior-basal).

Main Methods:

• Retrospective analysis of 370 radical prostatectomy patients with Gleason ≤6 prostate cancer.
• Evaluation of nomogram performance using discrimination, calibration, predictive accuracy, and net benefit.
• Subgroup analysis based on tumor location: anterior-apical (AA) and posterior-basal (PB).

Main Results:

• 141 patients (38%) had insignificant prostate cancer.
• Kattan and Steyerberg nomograms showed superior performance (AUC 0.768-0.770) with good calibration and net benefit.
• Nomogram accuracy was higher for PB tumors than AA tumors, especially at higher predicted probabilities.

Conclusions:

• Kattan and Steyerberg nomograms are the best predictors of insignificant prostate cancer in this cohort.
• Nomogram accuracy is reduced for anterior-apical tumors and for predicting insignificant disease versus significant disease.