Integrated Germline and Somatic Features Reveal Divergent Immune Pathways Driving Response to Immune Checkpoint...

Area of Science:

• Oncology
• Immunology
• Computational Biology

Background:

• Immune checkpoint blockade (ICB) has transformed cancer therapy.
• Understanding the mechanisms behind patient response to ICB is crucial.
• Biomarkers for predicting ICB efficacy are actively sought.

Purpose of the Study:

• To predict ICB response using machine learning on germline and somatic biomarkers.
• To interpret machine learning models for uncovering mechanisms of superior ICB outcomes.
• To investigate differential ICB responses based on MHC class-I (MHC-I) and MHC class-II neoantigen reliance.

Main Methods:

• Applied machine learning models to predict ICB response.
• Analyzed germline and somatic biomarkers.
• Investigated tumor immune microenvironment characteristics.
• Examined immune checkpoint expression, including LAG3.

Main Results:

• Higher T-follicular helper cell infiltration correlated with ICB response, even with MHC-I defects.
• MHC II-reliant responses showed significantly longer durable clinical benefits compared to MHC I-reliant responses.
• Tumor immune microenvironment features differed based on MHC neoantigen reliance.
• LAG3 was identified as a potential therapeutic target in MHC II-reliant responses.

Conclusions:

• Interpretable machine learning can elucidate biological mechanisms driving ICB response.
• MHC neoantigen reliance is a key factor in ICB efficacy and durability.
• Targeting LAG3 may benefit patients with MHC II-reliant tumors.