Identifying cancer-driving mutations is complicated by the fact that the signals of positive selection that are used to detect driver mutations in cancer genomes also occur in healthy aging tissues. To address this, Cheek and colleagues developed a statistical framework that integrates cancer genomes, normal tissue mutation data, and patient age to distinguish carcinogenic mutations from those that simply expand with age through positive selection. The authors quantified each mutation's carcinogenic effect by estimating how strongly the mutation increased the probability that a normal cell would give rise to cancer, using differences in mutation frequency between tumors and normal tissues as a proxy. Applying this framework across multiple malignancies revealed striking variation in carcinogenic potency. Canonical drivers such as TP53 and KRAS showed strong cancer-promoting effects, whereas other previously classified driver genes, including NOTCH1, exhibited minimal or even inhibitory effects on carcinogenesis. The authors then asked whether patient age at diagnosis could provide comparable information about causation when normal tissue data are unavailable. Evolutionary modeling predicted that highly carcinogenic mutations would appear preferentially in younger patients by accelerating tumor onset, while mutations under positive selection that do not directly promote cancer would accumulate in older individuals. In acute myeloid leukemia (AML), this prediction held: Mutations arising in younger patients showed stronger carcinogenic effects, whereas mutations positively selected in normal blood appeared more frequently in older patients. Importantly, carcinogenic effect estimates remained relatively stable between pediatric and adult AML despite substantial differences in mutation frequencies, suggesting that age captures distinct biological information beyond mutation prevalence alone. Extending the framework across additional cancer types revealed consistent age-dependent patterns for both point mutations and chromosomal alterations. Together, these findings establish patient age as a meaningful biological signal that, combined with genomic data, may help separate cancer causation from positive selection, with implications for prevention, risk stratification, and therapeutic prioritization.Cheek D, Blohmer M, Nowak MA, Antal T, Naxerova K. Age distinguishes selection from causation in cancer genomes. Nat Genet 2026 May 5 Epub ahead of print.Note: Research Watch is written by Cancer Discovery editorial staff. Readers are encouraged to consult the original articles for full details. For more Research Watch, visit Cancer Discovery online at https://aacrjournals.org/cdnews.
A Fri, study studied this question.