According to one of the fundamental principles of statistics, correlation does not imply causation; that is, an association between two variables does not automatically mean that one directly causes the other. Even though all cancer cells contain genetic mutations, this circumstance does not indicate that every one of them is necessarily involved in the pathogenesis of the tumor.
A team of researchers at Harvard Medical School, led by Camilla Nakserova and David Chick, has developed a statistical methodology that distinguishes tumor-driving somatic mutations from genetic changes that arise gradually during the process of physiological aging in the human body. The paper, published in Nature Genetics in May 2026, establishes a new approach for studying the molecular mechanisms of carcinogenesis, where the patient’s age is considered a critical biological indicator for assessing the carcinogenic potential of a mutation.
Over the last decade, the exponential development of oncological genomics has enabled the analysis of genetic profiles from hundreds of thousands of tumor samples, which has laid the foundation for forming an extensive registry of potentially oncogenic mutations. However, the introduction of methodologies for sequencing normal tissues has revealed that a significant portion of these genetic changes is also recorded in tissues where no atypical processes are observed, and their frequency increases with advancing age. This circumstance has given rise to a significant epistemological dilemma: in the presence of identical genetic changes, by what criteria should causal mutations be differentiated from random correlational associations?
To answer this question, the Harvard University researchers developed a new statistical model that enables the identification of cancer-causing mutations with higher precision. The methodology is based on a comparative analysis of genetic data from tumor and normal tissues. Within the scope of the study, mutations in esophageal squamous cell carcinoma, acute myeloid leukemia, and colorectal cancer were analyzed. The obtained results showed that mutations with high carcinogenic potential occur more frequently in younger patients, whereas many mutations detected in elderly patients may merely represent a biological reflection of chronological aging.
Additionally, the analysis results revealed that only a small subgroup of previously classified carcinogenic somatic mutations participates in the pathogenesis of the tumor, confirming the growing opinion in the scientific community that current classifications may include false positives.
Of particular importance is the second part of the study, where the authors identified a statistically significant relationship between the patient’s age at diagnosis and the carcinogenic power of the mutation. Specifically, mutations found with high frequency in the tumors of younger patients were distinguished by a strong carcinogenic effect, while excessive mutations in elderly patients were characterized by a relatively weak causal influence. This connection is biologically logical, as a mutation with high causal potential determines the development of cancer at a relatively early age, whereas mutations with a weak carcinogenic effect, in combination with other co-factors, provoke a later manifestation of the disease. This conclusion is of particular clinical importance because age, as an analytical variable, is practically not integrated into existing models for the classification of cancer-associated mutations.
Scientists believe that the obtained results will have significant practical value for both the early diagnosis of cancer and the development of personalized therapy. More precise classification of mutations will allow clinicians to better determine individual patient risks and select more effective targeted therapies. Furthermore, this approach may assist the pharmaceutical industry in identifying genetic targets whose blockade is critically important for weakening tumor viability and, ultimately, for its elimination. Additionally, the study emphasizes that integrating the age factor into cancer genetics may become an essential prerequisite for the development of more precise and effective oncological medicine in the future.
Source: hms.harvard.edu
