Abstract A002: Epigenetic aging as a predictive biomarker of genetic and pharmacological vulnerabilities in cancer cells

Abstract Epigenetic clocks are mathematical models that estimate biological age based on DNA methylation patterns. These clocks have become powerful tools in the field of aging research, providing insights into biological processes related to aging that are not captured by chronological age alone. While current epigenetic clocks have been effective in predicting the age of healthy samples, they perform poorly on tumor data due to the altered epigenetic patterns present in cancer. The aim of this study is to develop a novel cancer-specific epigenetic clock which provides an accurate estimation of epigenetic aging states in various cancers then to take advantage of this model to predict the epigenetic age of cancer in vitro models, thus to identify and dissect age-related genetic vulnerabilities, explore therapeutic opportunities, and predict drug responses, by exploiting the wealth of data from the Cancer Dependency Map (DepMap) project repositories. We developed a novel epigenetic clock represented by an elastic net regression model trained on a pan-cancer cohort of DNA methylation primary tumor samples from The Cancer Genome Atlas (TCGA). This new model, trained exclusively on cancer data, can accurately predict age in both malignant and healthy samples, with Pearson correlations of 0.781 and 0.857, respectively. These results suggest that DNA methylation patterns may potentially affect common age-related biological pathways that are preserved even under altered conditions of cancer. The epigenetic signature provided by the model, consisting of 5378 CpGs, was used to investigate therapeutic potential and analyze genetic vulnerabilities in a large panel of multi-omics characterized cancer cell line models from the Cell Model Passport platform. Given the intrinsic limitations of using chronological age data in cell lines, where “age at sampling” information cannot reflect the true aging of the cells due to numerous confounding factors of in vitro models, the age predicted by our epigenetic clock serves as a new measure for assessing aging-related epigenetic effects in the quest for biomarkers of cancer drug response and genetic dependencies on a very large scale. By correlating the predicted age of cancer cell lines with drug response from the Genomics of Drug Sensitivity in Cancer (GDSC) database and gene essentiality data from large-scale CRISPR screenings in the DepMap, we uncovered potential age-related specific vulnerabilities and therapeutic responses in both pan-cancer and cancer-specific settings. This study highlights the potential of epigenetic aging as a molecular signature in the context of cancer, opening new directions for identifying age-related drug response patterns and genetic dependencies. Citation Format: Irene Fernández-Rebollz, Alessandro Digilio, Athanasios Oikonomou, Lucia Trastulla, Manel Esteller, Francesco Iorio. Epigenetic aging as a predictive biomarker of genetic and pharmacological vulnerabilities in cancer cells [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: DNA Methylation, Clonal Hematopoiesis, and Cancer; 2025 Feb 1-4; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2025;85(3 Suppl):Abstract nr A002.