Abstract
Telomerase activity plays an essential role in tumor growth and varies across cancers, typically classified as low or high based on its expression level. This variation is pertinent to cancer-related mechanisms and hallmarks of cellular aging. However, the relationship between distinct telomerase activity groups (low or high) and specific molecular programs across tumor types remains poorly defined, largely due to the absence of a robust classification framework. Here, we applied EXTEND, our previously validated computational model for quantifying telomerase activity, to stratify tumors into low and high telomerase activity groups across diverse cancer types using an unsupervised, data-driven approach. We analyzed over 10,000 tumor samples from bulk RNA sequencing data in The Cancer Genome Atlas (TCGA) and the Cancer Cell Line Encyclopedia (CCLE), as well as more than 10,000 single cells from single-cell and spatial transcriptomic datasets. Our analyses revealed that high telomerase activity group was strongly associated with genomic instability across majority of cancers, whereas low telomerase activity group was enriched for cellular senescence, inflammation, reactive oxygen species (ROS), and MAPK signaling pathways. Notably, cellular senescence, a hallmark of aging, was predominant in older individuals across cancers, normal tissues, and developmental stages. Together, our findings establish a comprehensive framework linking telomerase activity groups to distinct molecular and cellular phenotypes across human cancers and reveal that low telomerase activity corresponds to a senescence-like transcriptional program that is generally associated with favorable survival outcomes. Conclusively, our work provides a unifying framework for understanding telomerase-associated heterogeneity across a broad compendium of tumors.