Frontiers in Oncology (Jan 2023)
Real-world data to build explainable trustworthy artificial intelligence models for prediction of immunotherapy efficacy in NSCLC patients
- Arsela Prelaj,
- Arsela Prelaj,
- Edoardo Gregorio Galli,
- Edoardo Gregorio Galli,
- Edoardo Gregorio Galli,
- Vanja Miskovic,
- Mattia Pesenti,
- Giuseppe Viscardi,
- Giuseppe Viscardi,
- Benedetta Pedica,
- Laura Mazzeo,
- Laura Mazzeo,
- Laura Mazzeo,
- Achille Bottiglieri,
- Achille Bottiglieri,
- Leonardo Provenzano,
- Leonardo Provenzano,
- Andrea Spagnoletti,
- Andrea Spagnoletti,
- Roberto Marinacci,
- Alessandro De Toma,
- Claudia Proto,
- Roberto Ferrara,
- Marta Brambilla,
- Mario Occhipinti,
- Sara Manglaviti,
- Giulia Galli,
- Diego Signorelli,
- Diego Signorelli,
- Claudia Giani,
- Claudia Giani,
- Teresa Beninato,
- Teresa Beninato,
- Chiara Carlotta Pircher,
- Chiara Carlotta Pircher,
- Alessandro Rametta,
- Alessandro Rametta,
- Sokol Kosta,
- Michele Zanitti,
- Maria Rosa Di Mauro,
- Arturo Rinaldi,
- Settimio Di Gregorio,
- Martinetti Antonia,
- Marina Chiara Garassino,
- Marina Chiara Garassino,
- Filippo G. M. de Braud,
- Filippo G. M. de Braud,
- Marcello Restelli,
- Giuseppe Lo Russo,
- Monica Ganzinelli,
- Francesco Trovò,
- Alessandra Laura Giulia Pedrocchi
Affiliations
- Arsela Prelaj
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Arsela Prelaj
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Edoardo Gregorio Galli
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Edoardo Gregorio Galli
- Niguarda Cancer Center, Grande Ospedale Metropolitano Niguarda, Milan, Italy
- Edoardo Gregorio Galli
- Oncology Department, University of Milan, Milan, Italy
- Vanja Miskovic
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Mattia Pesenti
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Giuseppe Viscardi
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Giuseppe Viscardi
- Medical Oncology Unit, Department of Precision Medicine, University of Campania “Luigi Vanvitelli”, Naples, Italy
- Benedetta Pedica
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Laura Mazzeo
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Laura Mazzeo
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Laura Mazzeo
- Oncology Department, University of Milan, Milan, Italy
- Achille Bottiglieri
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Achille Bottiglieri
- Oncology Department, University of Milan, Milan, Italy
- Leonardo Provenzano
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Leonardo Provenzano
- Oncology Department, University of Milan, Milan, Italy
- Andrea Spagnoletti
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Andrea Spagnoletti
- Oncology Department, University of Milan, Milan, Italy
- Roberto Marinacci
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Alessandro De Toma
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Claudia Proto
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Roberto Ferrara
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Marta Brambilla
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Mario Occhipinti
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Sara Manglaviti
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Giulia Galli
- Medical Oncology Unit, Policlinico San Matteo Fondazione IRCCS, Pavia, Italy
- Diego Signorelli
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Diego Signorelli
- Niguarda Cancer Center, Grande Ospedale Metropolitano Niguarda, Milan, Italy
- Claudia Giani
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Claudia Giani
- Oncology Department, University of Milan, Milan, Italy
- Teresa Beninato
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Teresa Beninato
- Oncology Department, University of Milan, Milan, Italy
- Chiara Carlotta Pircher
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Chiara Carlotta Pircher
- Oncology Department, University of Milan, Milan, Italy
- Alessandro Rametta
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Alessandro Rametta
- Oncology Department, University of Milan, Milan, Italy
- Sokol Kosta
- Department of Electronic System, Aalborg University, Copenhagen, Aalborg, Denmark
- Michele Zanitti
- Department of Electronic System, Aalborg University, Copenhagen, Aalborg, Denmark
- Maria Rosa Di Mauro
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Arturo Rinaldi
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Settimio Di Gregorio
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Martinetti Antonia
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Marina Chiara Garassino
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Marina Chiara Garassino
- Thoracic Oncology Program, Section of Hematology/Oncology, University of Chicago, Chicago, IL, United States
- Filippo G. M. de Braud
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Filippo G. M. de Braud
- Oncology Department, University of Milan, Milan, Italy
- Marcello Restelli
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Giuseppe Lo Russo
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Monica Ganzinelli
- Medical Oncology Department, Fondazione IRCCS Istituto Nazionale Tumori, Milan, Italy
- Francesco Trovò
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- Alessandra Laura Giulia Pedrocchi
- Department of Electronics, Information and Bioengineering, Politecnico di Milano, Milan, Italy
- DOI
- https://doi.org/10.3389/fonc.2022.1078822
- Journal volume & issue
-
Vol. 12
Abstract
IntroductionArtificial Intelligence (AI) methods are being increasingly investigated as a means to generate predictive models applicable in the clinical practice. In this study, we developed a model to predict the efficacy of immunotherapy (IO) in patients with advanced non-small cell lung cancer (NSCLC) using eXplainable AI (XAI) Machine Learning (ML) methods.MethodsWe prospectively collected real-world data from patients with an advanced NSCLC condition receiving immune-checkpoint inhibitors (ICIs) either as a single agent or in combination with chemotherapy. With regards to six different outcomes - Disease Control Rate (DCR), Objective Response Rate (ORR), 6 and 24-month Overall Survival (OS6 and OS24), 3-months Progression-Free Survival (PFS3) and Time to Treatment Failure (TTF3) - we evaluated five different classification ML models: CatBoost (CB), Logistic Regression (LR), Neural Network (NN), Random Forest (RF) and Support Vector Machine (SVM). We used the Shapley Additive Explanation (SHAP) values to explain model predictions.ResultsOf 480 patients included in the study 407 received immunotherapy and 73 chemo- and immunotherapy. From all the ML models, CB performed the best for OS6 and TTF3, (accuracy 0.83 and 0.81, respectively). CB and LR reached accuracy of 0.75 and 0.73 for the outcome DCR. SHAP for CB demonstrated that the feature that strongly influences models’ prediction for all three outcomes was Neutrophil to Lymphocyte Ratio (NLR). Performance Status (ECOG-PS) was an important feature for the outcomes OS6 and TTF3, while PD-L1, Line of IO and chemo-immunotherapy appeared to be more important in predicting DCR.ConclusionsIn this study we developed a ML algorithm based on real-world data, explained by SHAP techniques, and able to accurately predict the efficacy of immunotherapy in sets of NSCLC patients.
Keywords
- non-small cell lung cancer
- immunotherapy
- machine learning
- explainable artificial intelligence
- treatment