Development and validation of a mutation-based model to predict immunotherapeutic efficacy in NSCLC

Area of Science:

• Oncology
• Immunology
• Genetics

Background:

• Immunotherapy is crucial in perioperative non-small cell lung cancer (NSCLC) treatment.
• Predicting response to immune checkpoint inhibitors (ICIs) remains a challenge.

Purpose of the Study:

• Develop a mutation-based model to predict ICI efficacy in NSCLC patients.
• Identify key mutated genes associated with immunotherapy outcomes.

Main Methods:

• Random Forest (RF) classifiers identified 18 key mutated genes.
• A predictive model classified patients into mutant (MT) and wild-type (WT) groups.
• Survival analyses were performed on training and validation cohorts.

Main Results:

• The MT group (mutant gene ≥ 2) showed significantly better overall survival (OS) than the WT group (median OS not reached vs. 9 months, P < 0.0001).
• The model demonstrated robust predictive performance across three validation cohorts (AUCs 0.70, 0.57, 0.64).
• The MT group exhibited higher tumor neoantigen burden, increased immune cell infiltration, and upregulated immune checkpoints.

Conclusions:

• The developed 18-gene model accurately predicts immunotherapy efficacy in NSCLC.
• Mutant gene profiles within the model correlate with improved treatment outcomes.
• Immune-related features and cell expression support the model's predictive capability.