Characterization of Multi-Omics Landscapes and AI Pathological Prediction Model for Long-Term Survival in NSCLC Immunotherapy

Characterization of Multi-Omics Landscapes in Long-Term Survival Following Immunotherapy and Development of an AI Pathological Prediction Model for Long-Term Survival Based on H&E-Stained Images in Advanced Non-Small Cell Lung Cancer

This study is a retrospective, multicenter, observational cohort study in patients with advanced or locally advanced non-small cell lung cancer (NSCLC). The aim of this study was to establish a long-term survival (LTS) versus short-term survival (STS) real-world cohort, to systematically characterize the multi-omics landscapes, and to develop and validate an artificial intelligence (AI) pathological prediction model based on routine H&E-stained images for predicting immune microenvironment features and long-term survival outcomes following immunotherapy.

Inclusion Criteria:

• Patients with pathologically confirmed advanced or locally advanced non-small cell lung cancer (NSCLC).
• Patients derived from real-world data of multiple centers (including Cancer Hospital, Chinese Academy of Medical Sciences; Cancer Hospital of Shanxi, Chinese Academy of Medical Sciences [Shanxi Cancer Hospital]; and other participating centers) or from completed phase III clinical trials (e.g., Choice-01, Rationale-307, Rationale-304).
• Patients who received first-line or later-line immune checkpoint inhibitor (ICI) monotherapy or ICI-based combination therapy.
• Patients with complete clinical information and available follow-up data.

Exclusion Criteria:

• Patients whose systemic therapy did not include an immunotherapy regimen.
• Patients lost to follow-up.