Germline microRNA-based signatures predict toxicity and response to anti-CTLA-4 therapy

Area of Science:

• Oncology
• Immunotherapy
• Genetics

Background:

• Germline microRNA variants (mirSNPs) are known biomarkers for anti-PD1/PDL1 therapy response and toxicity.
• CTLA-4 inhibitors cause significant immune-related adverse events (irAEs), but predictive biomarkers are needed.
• The predictive potential of mirSNPs for anti-CTLA-4 therapy has not been previously investigated.

Purpose of the Study:

• To investigate the utility of mirSNPs in predicting irAEs and/or response to anti-CTLA-4 therapy alone.
• To identify genetic signatures for toxicity and tumor response in melanoma patients treated with anti-CTLA-4 inhibitors.

Main Methods:

• Evaluated genetic signatures in three melanoma patient cohorts (n=77) treated with anti-CTLA-4 therapy.
• Utilized DNA from blood samples analyzed with a custom mirSNP panel.
• Employed machine learning models (Elastic Net, Random Forest, Boosted Tree) with mirSNPs, demographics, and treatment variables.
• Conducted gene ontology (GO) pathway analysis to understand biological underpinnings.

Main Results:

• Developed two distinct mirSNP signatures with high predictive accuracy for toxicity (AUC=0.793) and response (AUC=0.842).
• Signatures were unique and did not overlap with each other or with previously reported anti-PD1/L1 toxicity signatures.
• GO analysis revealed shared pathways in pri-miRNA transcriptional regulation and unique differentiating pathways for each signature.

Conclusions:

• Germline mirSNPs are valuable predictive biomarkers for both toxicity and response in immune checkpoint therapy.
• Findings support the utility of mirSNPs for predicting outcomes in anti-CTLA-4 therapy.
• Future research will focus on larger cohorts and dual checkpoint inhibitor treatments.