Integrative kinetics and machine learning modeling for prediction of outcome following immunotherapy in lung cancer - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Integrative kinetics and machine learning modeling for prediction of outcome following immunotherapy in lung cancer

Résumé

I will present recent results from COMPO (COMPutational pharmacology and clinical Oncology) ai- ming at combining mechanistic modeling and machine learning ("mechanistic learning") to integrate longitudinal, multi-modal and high-dimensional data into predictive models of outcome following im- munotherapy in non-small cell lung cancer (NSCLC). This will be based on two studies. The first leverages clinical trial data to help in drug development by predicting outcome of late-phase trials (e.g., phase 3) from early data (e.g., phase 2). The second is an integrative analysis of multi-modal deep-level biomarkers (multiplex immunohistochemistry, immune-monitoring, vasculo-monitoring, he- matology and biochemistry) collected during the RHU PIONeeR. The results show substantial im- provement of the predictive performances of classical markers (PDL1 expression, AUC = 0.64, tumor mutational burden, AUC = 0.65) using a novel kinetics-machine learning (kML) model (AUC = 0.86, c-index = 0.79, test set). The kML model was also able to predict the positive outcome of the phase 3 of the OAK trial (atezolizumab versus docetaxel) using 30 weeks on-study data (model HR = 0.802 (95% CI : 0.655 - 0.907)) while the observed data at this landmark time point was not conclusive (data HR = 1.04 (95% CI : 0.386 - 2.79)). References : [1] Benzekry et al., Supporting decision making and early prediction of survival for oncology drug development using a pharmacometrics-machine learning based model, PAGE 30 (2022) Abstr 10276 [2] Greillier, L. et al. Comprehensive biomarkers analysis to explain resistances to PD1-L1 ICIs : The precision immuno-oncology for advanced non-small cell lung cancer (PIONeeR) trial. Cancer Res 82, LB120LB120 (2022). [3] Barlesi, F. et al., Benzekry, S. Comprehensive biomarkers (BMs) analysis to predict efficacy of PD1/L1 immune checkpoint inhibitors (ICIs) in combination with chemotherapy : a subgroup analysis of the Precision Immuno-Oncology for advanced Non-Small CEll Lung CancER (PIONeeR) trial. Annals of Oncology, 16 (suppl_1), 2022
Fichier non déposé

Dates et versions

hal-04388478 , version 1 (11-01-2024)

Licence

Paternité

Identifiants

  • HAL Id : hal-04388478 , version 1

Citer

Sébastien Benzekry. Integrative kinetics and machine learning modeling for prediction of outcome following immunotherapy in lung cancer. SMAI 2023 congress - 11th French Biennial of Applied and Industrial Mathematics, May 2023, Le Gosier (Guadeloupe), France. ⟨hal-04388478⟩
28 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More