Search

Immune checkpoint blockade (ICB) has revolutionized cancer treatment; however, the mechanisms determining patient response remain poorly understood. Here, we used machine learning to predict ICB response from germline and somatic biomarkers and interpreted the learned model to uncover putative mechanisms driving superior outcomes. Patients with higher infiltration of T-follicular helper cells had responses even in the presence of defects in the MHC class-I (MHC-I). Further investigation uncovered different ICB responses in tumors when responses were reliant on MHC-I versus MHC-II neoantigens. Despite similar response rates, MHC II-reliant responses were associated with significantly longer durable clinical benefits (discovery: median overall survival of 63.6 vs. 34.5 months; P = 0.0074; validation: median overall survival of 37.5 vs. 33.1 months; P = 0.040). Characteristics of the tumor immune microenvironment reflected MHC neoantigen reliance, and analysis of immune checkpoints revealed LAG3 as a potential target in MHC II-reliant but not MHC I-reliant responses. This study highlights the value of interpretable machine learning models in elucidating the biological basis of therapy responses., (©2024 The Authors; Published by the American Association for Cancer Research.)

Insights into oncogenesis derived from cancer susceptibility loci (SNP) hold the potential to facilitate better cancer management and treatment through precision oncology. However, therapeutic insights have thus far been limited by our current lack of understanding regarding both interactions of these loci with somatic cancer driver mutations and their influence on tumorigenesis. For example, although both germline and somatic genetic variation to the p53 tumor suppressor pathway are known to promote tumorigenesis, little is known about the extent to which such variants cooperate to alter pathway activity. Here we hypothesize that cancer risk-associated germline variants interact with somatic TP53 mutational status to modify cancer risk, progression, and response to therapy. Focusing on a cancer risk SNP (rs78378222) with a well-documented ability to directly influence p53 activity as well as integration of germline datasets relating to cancer susceptibility with tumor data capturing somatically-acquired genetic variation provided supportive evidence for this hypothesis. Integration of germline and somatic genetic data enabled identification of a novel entry point for therapeutic manipulation of p53 activities. A cluster of cancer risk SNPs resulted in increased expression of prosurvival p53 target gene KITLG and attenuation of p53-mediated responses to genotoxic therapies, which were reversed by pharmacologic inhibition of the prosurvival c-KIT signal. Together, our results offer evidence of how cancer susceptibility SNPs can interact with cancer driver genes to affect cancer progression and identify novel combinatorial therapies. SIGNIFICANCE: These results offer evidence of how cancer susceptibility SNPs can interact with cancer driver genes to affect cancer progression and present novel therapeutic targets., (©2021 American Association for Cancer Research.)