External Validation of the Nomogram Predicting Locally Advanced Prostate Cancer in Patients Undergoing Robot-Assisted Radical Prostatectomy (the MSUG94 Group).

Abstract

A clinically applicable nomogram predicting locally advanced prostate cancer (PCa) (defined as pathological T stage ≥ 3) of patients with clinical T stage ≤ 2 was released (MSUG nomogram). We performed external validation, ensuring its applicability to patients undergoing robot-assisted radical prostatectomy (RARP). Therefore, we also compared the external validation for the Memorial Sloan Kettering Cancer Center (MSKCC) nomogram. We collected the data for 8194 patients who underwent RARP at Daimonji Clinical Application Database Group (Dai-CAD) as the validation cohort and performed the external validation using this cohort. The primary endpoint was the accuracy of the MSUG nomogram, and the secondary endpoint was comparison with the MSKCC nomogram. The receiver operating characteristic (ROC) curve and area under the curve (AUC) were calculated to quantify the accuracy of the nomogram at predicting pT ≥ 3. A calibration plot was performed to evaluate the extent of over- and underestimation. Locally advanced PCa was diagnosed in 677 of 2530 patients (26.8%) in the MSUG cohort and 1472 of 5799 patients (25.3%) in the validation cohort. The ROC curve for the validation cohorts fit to the MSUG nomogram and the MSKCC nomogram, with AUC of 0.66 and 0.65, respectively. For calibration plots, it overestimated the risk of locally advanced PCa when probability thresholds are over 70% in the MSUG nomogram, while it may overestimate when probability thresholds are over 30% in the MSKCC nomogram. We conducted external validation of a clinically applicable nomogram that predicts the probability of locally advanced PCa in patients undergoing RARP using available clinical parameters.