External Validation of the Nomogram Predicting Locally Advanced Prostate Cancer in Patients Undergoing Robot-Assisted...

Area of Science:

• Urology
• Oncology
• Medical Informatics

Background:

• Locally advanced prostate cancer (PCa) requires accurate prediction for treatment planning.
• Existing nomograms may need external validation for specific surgical techniques like robot-assisted radical prostatectomy (RARP).

Purpose of the Study:

• To externally validate the Memorial Sloan Urology Group (MSUG) nomogram for predicting locally advanced PCa (pathological T stage ≥ 3) in patients with clinical T stage ≤ 2 undergoing RARP.
• To compare the predictive accuracy of the MSUG nomogram with the Memorial Sloan Kettering Cancer Center (MSKCC) nomogram in this cohort.

Main Methods:

• External validation using a cohort of 8194 patients who underwent RARP.
• Calculation of receiver operating characteristic (ROC) curves and area under the curve (AUC) to assess predictive accuracy.
• Calibration plots to evaluate nomogram over- and underestimation of risk.

Main Results:

• The MSUG nomogram achieved an AUC of 0.66, while the MSKCC nomogram achieved an AUC of 0.65 in the validation cohort.
• Both nomograms demonstrated overestimation of locally advanced PCa risk at certain probability thresholds: >70% for MSUG and >30% for MSKCC.
• Locally advanced PCa was present in 26.8% of the MSUG cohort and 25.3% of the validation cohort.

Conclusions:

• The MSUG nomogram is a clinically applicable tool for predicting locally advanced PCa in patients undergoing RARP.
• External validation confirms its utility, showing comparable accuracy to the MSKCC nomogram.
• Calibration requires attention, particularly at higher predicted risk probabilities.