Abstract 3881: Comprehensive analysis of somatic mutagenesis in normal tissues across 19,643 samples reveals insights into clonal evolution and tumorigenesis

The detection of somatic mutagenesis in normal tissues has transformed our understanding of clonal evolution, revealing a mosaic of somatic mutations, including those in cancer driver genes, that influence cellular behavior and tissue homeostasis. However, key questions remain about how these mutational patterns contribute to tumorigenesis and the role of cancer driver mutations in the transition from normal to malignant states. To address this, we performed a combined computational analysis of the largest normal-tissue DNA-sequencing and clinical dataset to date, spanning 21 published datasets, 37 tissue types, 521 patients, and 19,643 samples, and integrated these with precancer and cancer tissue data to identify tissue-specific mutational changes potentially critical for tumorigenesis. We identified tissue-specific differences in mutation burden, with epithelial tissues exhibiting significantly higher mutation rates than other tissue categories (p<2x10-16). These differences were further reflected in the mutational signatures: UV-light signatures dominated sun-exposed epithelial tissues, while gastrointestinal epithelial tissues showed mutation burden driven by intrinsic factors. Regression analysis revealed that, alongside aging, mutations in specific driver genes (e.g., ARID2, ATM, TP53), but not others, were selectively associated with higher mutation rates, highlighting nuanced contributions to tissue mutagenesis. Further MutSig2CV and dN/dS analyses identified positively selected driver genes with distinct patterns based on tumorigenic pathways: tumors that develop through normal-to-dysplastic sequences (e.g., colon cancer) shared driver genes across all progression stages, reflecting clonal evolution within the same tissue lineage. In contrast, tumors arising from metaplastic processes (e.g., esophageal adenocarcinoma) shared mutated driver genes between cancer and precancer metaplastic tissues, but not normal tissues, suggesting selective pressures emerge during metaplasia. Across the cohort, most driver genes were highly mutated in cancer and present at lower mutation frequency in normal tissues. In contrast, similar to NOTCH1 mutations, CHEK2 mutations were significantly enriched in the normal esophagus but absent in cancer, hinting at potential tumor-suppressive effects. Taken together, our large-scale analysis has allowed us to detect previously unrecognized phenomena through increased statistical power. The analysis of normal tissue mutagenesis underscores its potential for predicting tissue-specific cancer risks. In addition, our integration of normal tissue data into the cBioPortal platform provides an important resource to support and advance future cancer research. Citation Format: Leonie Stockschläder, Eldar Abdullaev, Siddharth Annaldasula, Ino de Bruijn, Baby Anusha Satravada, Ritika Kundra, Rima Madupuri, Manuela Benary, Dieter Beule, Tim H. Coorens, Regine Armann, Nikolaus Schultz, Chris Sander, Kirsten Kubler. Comprehensive analysis of somatic mutagenesis in normal tissues across 19,643 samples reveals insights into clonal evolution and tumorigenesis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3881.