Abstract A5: Genome-wide methylation pattern predicts clinical benefit of immune checkpoint blockade therapy in NSCLC patients

Abstract

Abstract Immune checkpoint blockade (ICB) therapy for non-small cell lung cancer (NSCLC) has been gaining wide attention as the next-generation cancer therapy that can bolster patients’ immune system to better fight cancer. However, only a small subset of patients reap clinical benefit. Thus, it is crucial to unravel molecular determinants of response to ICB therapy. Although several predictive biomarkers of clinical response to ICB therapy have been proposed, such as mutation burden and expression of key genes (e.g., PD-L1), these biomarkers are not sufficient to accurately predict response to ICB therapy. Here, we aim to facilitate prediction of therapeutic response by analyzing DNA methylation pattern difference between responders and nonresponders on a genome-wide scale. To this end, we performed DNA methylation array (Illumina 850K/EPIC platform), exome-, and RNA-seq experiments for 60 NSCLC patients who have received ICB therapy from Samsung Medical Center. We found 434 differentially methylated genes between nonresponders and responders where 73 and 361 were hyper- and hypo- methylated, respectively, in nonresponders. Notably, immune pathways enriched with hyper-methylated genes significantly overlapped with those enriched with underexpressed genes in nonresponders, suggesting that promoter methylation-mediated immune pathway silencing affects immune response and clinical benefit. Furthermore, we identified 474 protein interaction networks where two or three hypermethylated genes showed significantly mutually exclusive pattern across nonresponders. Importantly, 34 protein networks were associated with progression-free survival. Finally, based on the methylation data from our cohort, we have built LASSO-Cox regression model to predict prognosis of patients. The model, which comprises 8 genes, was validated in two independent cohorts (81 NSCLC patients receiving ICB therapy from Bellvitge Biomedical Research Institute and 479 NSCLC patients from TCGA), demonstrating the importance of the methylation pattern of 8 genes in determining clinical benefit. In conclusion, methylation pattern can provide insight into clinical benefit of ICB therapy. In this study, we have extensively characterized methylation difference between responders and nonresponders on a genome-wide scale. We also demonstrate that gene hypermethylation significantly correlates with immune-related pathways’ silencing. Furthermore, our machine learning model built upon methylation data provides predictive power of clinical benefit. Thus, in addition to known biomarkers (e.g., mutation burden), methylation pattern should be considered when predicting benefit of ICB therapy. Citation Format: Jeongyeon Kim, Hyeon Gu Kang, Jae Soon Park. Genome-wide methylation pattern predicts clinical benefit of immune checkpoint blockade therapy in NSCLC patients [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2019 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2020;8(3 Suppl):Abstract nr A5.