External Validation of Nomograms for the Identification of Pelvic Nodal Dissection Candidates Among Prostate Cancer Patients with Negative Preoperative Prostate-specific Membrane Antigen Positron Emission Tomography

Abstract

Background and objective

Extended pelvic lymph node dissection (ePLND) is recommended in selected radical prostatectomy (RP) prostate cancer (PCa) patients for staging purposes. We aim to externally validate available tools to predict lymph node invasion (LNI) in men with negative preoperative prostate-specific membrane antigen positron emission tomography (miN0).

Methods

Overall, 282 intermediate- to high-risk PCa patients with miN0 disease undergoing RP and ePLND at ten centers between 2016 and 2023 were identified. The Memorial Sloan Kettering Cancer Center (MSKCC); Amsterdam-Brisbane-Sydney; and Briganti 2017, 2019, and 2023 tools predicting LNI were validated externally using calibration plots, C-indexes, and decision-curve analyses to assess calibration, discrimination, and net benefit.

Key findings and limitations

Overall, 36 (13%) patients had LNI. The C-indexes of the MSKCC, Briganti 2017, Briganti 2019, Amsterdam-Brisbane-Sydney, and Briganti 2023 nomograms were 64%, 69%, 72%, 64%, and 77%, respectively. The Briganti 2023 nomogram exhibited higher net benefit than the other available nomograms, and the use of a 5% cutoff would have spared 47% ePLND procedures (vs 14% and 4.3% for the Briganti 2019 and Amsterdam-Brisbane-Sydney nomograms, respectively) at the cost of missing only five (3.8%) LNI cases. Heterogeneity in patient selection and imaging protocols represents the main limitations.

Conclusions and clinical implications

The Briganti 2023 nomogram outperformed other available tools in predicting LNI in men with miN0 PCa. The use of this tool resulted in a considerable number of unnecessary ePLND procedures spared and optimization of ePLND recommendations in a contemporary clinical setting.