Explainable machine learning to predict immunotherapy outcomes in metastatic renal cell carcinoma - Meet-URO 15-AI study
15 affiliations
- 1Medical Oncology Unit, Ospedale San Paolo
- 2Department of Electronics, Information and Bioengineering, Polytechnic University of Milan
- 3Medical Oncology Unit, IRCCS Azienda Ospedaliera Metropolitana (IRCCS AOM), IRCCS Ospedale Policlinico San Martino
- 4Medical Oncology Unit, University Hospital of Parma
- 5Department of Health Sciences (DISSAL), Section of Biostatistics, University of Genova
- 6Department of Medical Oncology, Fondazione IRCCS Istituto Nazionale dei Tumori (INT)
- 7Department of Oncology, Portsmouth Hospitals University NHS Trust
- 8Translational and Clinical Research Institute, Centre for Cancer, Newcastle University
- 9Oncology Unit 3, Istituto Oncologico Veneto-IOV-IRCCS
- 10Oncology Unit 1, Istituto Oncologico Veneto-IOV-IRCCS
- 11Department of Medical Oncology, IRCCS Istituto Romagnolo per lo Studio dei Tumori (IRST) “Dino Amadori”
- 12Medical Oncology Unit, Fondazione IRCCS Policlinico San Matteo
- 13Medical Oncology Unit 2, Azienda Ospedaliera Universitaria Pisana
- 14Department of Biomedical Sciences, Humanitas University
- 15Barts Cancer Centre, St Bartholomew’s Hospital, Barts Health NHS Trust
Abstract
Metastatic renal cell carcinoma remains clinically challenging because of heterogeneous outcomes and limited predictive biomarkers for immunotherapy. We performed an explainable machine learning analysis using data from the multicenter retrospective Meet-URO 15 study, including 571 patients with metastatic renal cell carcinoma treated with second-line or later nivolumab. Clinical and inflammatory variables were used to develop classification models for disease control rate, progression-free survival at 3 and 9 months, and overall survival at 6, 18 and 24 months, as well as survival models for continuous progression-free and overall survival. Model performance was assessed using weighted F1-score for classification and concordance index for survival analysis, with interpretability provided through Shapley additive explanations. The best classification performance was observed for 6-month overall survival using a support vector machine model combined with minimum redundancy maximum relevance feature selection, achieving an F1-score of 0.81 on the test set and 0.77 in external validation. In survival analysis, random survival forest achieved a test-set concordance index of 0.68 for overall survival. Inflammatory indices, IMDC score, hemoglobin, lymphocytes and platelets consistently emerged as relevant prognostic features. These findings support explainable machine learning as a transparent approach to refine outcome prediction in immunotherapy-treated metastatic renal cell carcinoma.
Related
- Prognostic impact of the Gustave Roussy Immune score in metastatic renal cell carcinoma treated with first-line immunotherapy-based combinations: results from the Turkish Oncology Group Kidney Cancer Consortium (TKCC)
- Practice patterns and outcomes of local and systemic therapy for single-site metastatic non-small cell lung cancer in the era of immunotherapy.
- Multifactorial nutritional and metabolic risk profiles for behavioral outcomes in Chinese children: a cross-sectional study at a Health Management Center using explainable machine learning and principal component analysis
- Predictors of guideline-concordant treatment intensification in metastatic castration-sensitive prostate cancer: an explainable machine learning study
- Association between absolute lymphocyte count and outcomes of radiotherapy combined with immunotherapy in elderly patients with inoperable esophageal squamous cell carcinoma
BibTeX
@article{rebuzzi2026explainable,
title = {{Explainable machine learning to predict immunotherapy outcomes in metastatic renal cell carcinoma - Meet-URO 15-AI study}},
author = {Sara Elena Rebuzzi and Vanja Miskovic and Giuseppe Fornarini and Sara Ferri and Sebastiano Buti and Matteo Piceni and Alessio Signori and Leonardo Provenzano and Giuseppe Luigi Banna and Pasquale Rescigno and Marco Maruzzo and Davide Bimbatti and Beatrice Ramella Pollone and Umberto Basso and Ugo De Giorgi and Paolo Pedrazzoli and Luca Galli and Paolo Andrea Zucali and Fabrizio Di Costanzo and Aruni Ghose and Alessandra Laura Giulia Pedrocchi and Arsela Prelaj},
journal = {npj Precision Oncology},
year = {2026},
month = jul,
doi = {10.1038/s41698-026-01576-x},
url = {https://doi.org/10.1038/s41698-026-01576-x},
note = {CC BY},
}
RIS
TY - JOUR
TI - Explainable machine learning to predict immunotherapy outcomes in metastatic renal cell carcinoma - Meet-URO 15-AI study
AU - Sara Elena Rebuzzi
AU - Vanja Miskovic
AU - Giuseppe Fornarini
AU - Sara Ferri
AU - Sebastiano Buti
AU - Matteo Piceni
AU - Alessio Signori
AU - Leonardo Provenzano
AU - Giuseppe Luigi Banna
AU - Pasquale Rescigno
AU - Marco Maruzzo
AU - Davide Bimbatti
AU - Beatrice Ramella Pollone
AU - Umberto Basso
AU - Ugo De Giorgi
AU - Paolo Pedrazzoli
AU - Luca Galli
AU - Paolo Andrea Zucali
AU - Fabrizio Di Costanzo
AU - Aruni Ghose
AU - Alessandra Laura Giulia Pedrocchi
AU - Arsela Prelaj
PY - 2026
DA - 2026/07/01
JO - npj Precision Oncology
DO - 10.1038/s41698-026-01576-x
UR - https://doi.org/10.1038/s41698-026-01576-x
AB - Metastatic renal cell carcinoma remains clinically challenging because of heterogeneous outcomes and limited predictive biomarkers for immunotherapy. We performed an explainable machine learning analysis using data from the multicenter retrospective Meet-URO 15 study, including 571 patients with metastatic renal cell carcinoma treated with second-line or later nivolumab. Clinical and inflammatory variables were used to develop classification models for disease control rate, progression-free survival at 3 and 9 months, and overall survival at 6, 18 and 24 months, as well as survival models for continuous progression-free and overall survival. Model performance was assessed using weighted F1-score for classification and concordance index for survival analysis, with interpretability provided through Shapley additive explanations. The best classification performance was observed for 6-month overall survival using a support vector machine model combined with minimum redundancy maximum relevance feature selection, achieving an F1-score of 0.81 on the test set and 0.77 in external validation. In survival analysis, random survival forest achieved a test-set concordance index of 0.68 for overall survival. Inflammatory indices, IMDC score, hemoglobin, lymphocytes and platelets consistently emerged as relevant prognostic features. These findings support explainable machine learning as a transparent approach to refine outcome prediction in immunotherapy-treated metastatic renal cell carcinoma.
SN - 2397-768X
ER -