Association of artificial intelligence-based immunoscore with the efficacy of chemoimmunotherapy in patients with advanced non-squamous non-small cell lung cancer: a multicentre retrospective study

Front Immunol. 2024 Nov 6:15:1485703. doi: 10.3389/fimmu.2024.1485703. eCollection 2024.

Abstract

Purpose: Currently, chemoimmunotherapy is effective only in a subset of patients with advanced non-squamous non-small cell lung cancer. Robust biomarkers for predicting the efficacy of chemoimmunotherapy would be useful to identify patients who would benefit from chemoimmunotherapy. The primary objective of our study was to develop an artificial intelligence-based immunoscore and to evaluate the value of patho-immunoscore in predicting clinical outcomes in patients with advanced non-squamous non-small cell lung cancer (NSCLC).

Methods: We have developed an artificial intelligence-powered immunoscore analyzer based on 1,333 whole-slide images from TCGA-LUAD. The predictive efficacy of the model was further validated in the CPTAC-LUAD cohort and the biomarker cohort of the ORIENT-11 study, a randomized, double-blind, phase 3 study. Finally, the clinical significance of the patho-immunoscore was evaluated using the ORIENT-11 study cohort.

Results: Our immunoscore analyzer achieved good accuracy in all the three cohort mentioned above (TCGA-LUAD, mean AUC: 0.783; ORIENT-11 cohort, AUC: 0.741; CPTAC-LUAD cohort, AUC: 0.769). In the 259 patients treated with chemoimmunotherapy, those with high patho-immunoscore (n = 146) showed significantly longer median progression-free survival than those with low patho-immunoscore (n = 113) (13.8 months vs 7.13 months, hazard ratio [HR]: 0.53, 95% confidence interval [CI]: 0.38 - 0.73; p < 0.001). In contrast, no significant difference was observed in patients who were treated with chemotherapy only (5.07 months vs 5.07 months, HR: 1.04, 95% CI: 0.71 - 1.54; p = 0.83). Similar trends were observed in overall survival.

Conclusion: Our study indicates that AI-powered immunoscore applied on LUAD digital slides can serve as a biomarker for survival outcomes in patients with advanced non-squamous NSCLC who received chemoimmunotherapy. This methodology could be applied to other cancers and facilitate cancer immunotherapy.

Keywords: NSCLC; artificial intelligence; immunoscore; immunotherapy; pathology.

Publication types

  • Multicenter Study

MeSH terms

  • Aged
  • Antineoplastic Combined Chemotherapy Protocols / therapeutic use
  • Artificial Intelligence*
  • Biomarkers, Tumor
  • Carcinoma, Non-Small-Cell Lung* / drug therapy
  • Carcinoma, Non-Small-Cell Lung* / immunology
  • Carcinoma, Non-Small-Cell Lung* / mortality
  • Carcinoma, Non-Small-Cell Lung* / therapy
  • Female
  • Humans
  • Immunotherapy / methods
  • Lung Neoplasms* / drug therapy
  • Lung Neoplasms* / immunology
  • Lung Neoplasms* / mortality
  • Lung Neoplasms* / therapy
  • Male
  • Middle Aged
  • Retrospective Studies
  • Treatment Outcome

Substances

  • Biomarkers, Tumor

Grants and funding

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. This work was supported by: Chinese National Natural Science Foundation project (Grant No. 82172713, 82373262, 82102864, 82241232, 82272789), Guangzhou Basic and Applied Basic Research Foundation (Grant No. 2023A04J2133, 2024A04J4082), the Natural Science Foundation of Guangdong Province (Grant No. 2022A1515012582), Young Talents Program of Sun Yat-sen University Cancer Center, the Postdoctoral Fellowship Program of CPSF (Grant No. GZB20240900).