PD10-07: Can We Rely on Available Models to Identify Candidates for Extended Pelvic Lymph Node Dissection (ePLND) in Men Staged With PSMA-PET? External Validation of the Briganti Nomograms and Development of a Novel Tool to Identify Optimal Candidates for ePLND
Assistant Professor Department of Urology, San Raffaele Hospital
Introduction: Nomograms to identify prostate cancer (PCa) patients candidate for extended pelvic lymph node dissection (ePLND) during radical prostatectomy (RP) do not account for PSMA-PET. We hypothesized that available tools may be suboptimal and be recalibrated in cN0 disease at PSMA-PET due to its high negative predictive value (NPV) for lymph node invasion (LNI). Methods: 662 PCa patients who received a PSMA-PET and subsequent RP + ePLND at 9 referral centres between 2016-2022 were identified. All patients received ePLND regardless of PSMA-PET. Information on preoperative MRI, biopsies and pathologic data were available for all patients. Calibration and discrimination of available nomograms predicting LNI (Briganti 2012, Briganti 2017 and Briganti 2019) was assessed at external validation. The same process was repeated in men with cN0M0 disease at PSMA-PET. The Briganti 2019 model was recalibrated using the multivariable logistic regression coefficients of the variables included in the nomogram in men with cN0M0 disease at PSMA-PET. Calibration plots, the ROC-derived AUC, and decision-curve analyses were used to determine calibration, discrimination, and net benefit associated with the recalibrated nomogram and compare it with available tools. Results: Overall, 115 (17.4%) patients had positive pelvic spots at PSMA PET/CT. LNI rate was 19.5%. The discrimination of the Briganti 2012, 2017, and 2019 nomograms was 75, 75, and 71%. In patients with negative PSMA-PET (n= 537, LNI rate: 10%), the discrimination of the Briganti 2012, 2017, and 2019 nomograms was even lower (72, 72, and 70%). In these patients, the maximum index lesion diameter, seminal vesicle invasion at MRI, and a target biopsy ISUP group 4-5 were associated with a risk of LNI (all p < 0.05). A nomogram based on the coefficients of this multivariable regression had a discrimination of 80%, superior calibration characteristics and higher net-benefit compared to other models. A 7% cut-off in the recalibrated nomogram in cN0M0 disease at PSMA-PET would have spared 55% ePLNDs (vs. 29% for the Briganti 2019 nomogram),missing only 3.4% (vs. 2.9%) of LNIs. Conclusions: Due to its elevated NPV, in men staged with PSMA-PET the Briganti nomogram results in a high number of unnecessary ePLNDs. A novel recalibrated version of the nomogram should be used to identify candidates for ePLND, reducing unnecessary procedures, without missing additional LNIs in men with cN0M0 PCa at PSMA-PET. SOURCE OF Funding: NA
