The thymus - long dismissed after puberty - may predict who lives longest and who responds to cancer treatment

Adults with a healthy thymus - the small chest organ that trains immune T cells - had roughly half the risk of dying over the study period compared to those with poor thymic health. They had 63% lower cardiovascular mortality. They had 36% lower rates of lung cancer. And in a separate study of cancer patients, those with better thymic health responded markedly better to immunotherapy, with 37% lower risk of cancer progression and 44% lower risk of death.

These findings, from two papers published simultaneously in Nature by a Mass General Brigham research team led by Hugo Aerts, PhD, challenge a medical assumption that has persisted for decades: that the thymus becomes functionally irrelevant after puberty and can be safely ignored in adult medicine.

An organ hiding in plain sight on CT scans

The thymus sits in the upper chest, behind the breastbone. In children, it is a busy organ, training T cells to recognize foreign invaders while learning to tolerate the body's own tissues. But the thymus begins shrinking after puberty, gradually replaced by fat tissue in a process called involution. By adulthood, much of the organ appears to be fatty and inactive on imaging. For decades, this visible shrinkage led physicians to conclude that the adult thymus was essentially vestigial - a childhood organ with no continuing role in adult health.

The problem with that assumption, the new research suggests, is that nobody had actually looked at the data in large populations. Previous studies linking T cell diversity to aging and immune decline relied on small, blood-based analyses. The Mass General Brigham team took a fundamentally different approach: they used artificial intelligence to analyze routine CT scans that had already been performed for other purposes, extracting detailed measurements of thymic size, shape, and tissue composition that human radiologists would not typically assess.

25,000 adults and a thymic health score

The first study examined more than 25,000 adults from a national lung cancer screening trial and over 2,500 participants from the Framingham Heart Study - one of the longest-running population health studies in the world. The AI system analyzed each person's thymus on their CT scan and generated a composite "thymic health" score reflecting the organ's apparent structural integrity.

The associations with health outcomes were striking in their consistency and magnitude. People with high thymic health scores had approximately 50% lower risk of death from any cause compared to those with low scores. Cardiovascular death risk was 63% lower. Lung cancer incidence was 36% lower. These associations held after adjusting for age, sex, smoking status, and other established health factors - meaning the thymic health signal was not simply a proxy for being generally healthier or younger.

The researchers also examined what factors were associated with poorer thymic health. Chronic inflammation, smoking, and high body weight all tracked with lower thymic health scores. This raises the possibility that some of the immune decline attributed to normal aging may actually be driven by modifiable factors - lifestyle and systemic inflammation degrading an organ that, contrary to longstanding belief, continues to matter.

Thymic health and immunotherapy response

The second study applied the same AI-based thymic assessment to a different question: does the thymus influence how well cancer patients respond to immune checkpoint inhibitors - the class of immunotherapy drugs that work by unleashing the patient's own T cells against tumors?

The logic was straightforward. If the thymus continues producing or maintaining T cells in adulthood, and if T cell quality matters for immunotherapy effectiveness, then thymic health should predict treatment response. The researchers analyzed CT scans and outcomes from more than 1,200 patients treated with immunotherapy across multiple cancer types.

The results supported the hypothesis. Patients with stronger thymic health had 37% lower risk of cancer progression and 44% lower risk of death during follow-up, even after accounting for other patient characteristics, tumor features, and treatment variables. In a field where identifying reliable biomarkers of immunotherapy response remains a major challenge, the thymus emerged as a previously unrecognized factor with substantial predictive power.

Why the thymus was overlooked for so long

The thymus's decades-long neglect in adult medicine reflects a common pattern in biomedical thinking: visible changes in anatomy are assumed to mean loss of function. The thymus clearly shrinks with age. It clearly produces fewer new T cells in adults than in children. The inference - that it therefore does nothing important in adults - seemed logical but was never rigorously tested in large populations.

Several factors reinforced the assumption. The thymus is difficult to study in living humans without invasive procedures. Blood-based T cell analyses, while accessible, capture only circulating cells and may miss the organ-level picture. And CT scans, while routinely performed on millions of adults for other indications, were never systematically analyzed for thymic features because no one thought to look.

The AI-based approach changed that calculus. By training machine learning models to extract thymic measurements from existing scans, the researchers bypassed the need for dedicated imaging studies and gained access to datasets of a size that previous thymus research could not approach.

From correlation to mechanism: the gap that remains

The researchers are transparent about what their studies do and do not establish. The findings are observational. The associations between thymic health and longevity, cardiovascular risk, cancer incidence, and immunotherapy response are statistically robust and consistent across multiple cohorts. But they do not prove that the thymus directly causes these outcomes.

It remains possible that thymic health is a marker of overall biological aging rather than an independent contributor to health outcomes. People who age slowly in general may retain thymic tissue as part of a broader pattern of preserved organ function. In that case, the thymus would be a useful indicator but not a therapeutic target.

The imaging method itself is not yet validated for routine clinical use. The AI model was developed and tested within research settings, and translating it to clinical practice would require standardization, prospective validation, and integration into radiology workflows. The researchers caution that the tool is not ready for patient-level decision-making.

While lifestyle factors such as smoking, inflammation, and obesity were associated with poorer thymic health, the studies did not test whether modifying those factors actually improves thymic function. That interventional evidence - showing that doing something about thymic health changes outcomes - does not yet exist.

Unintended radiation and the thymus

The team is currently pursuing several follow-up investigations. One study examines whether unintended radiation exposure to the thymus during lung cancer treatment - a common occurrence given the organ's location in the radiation field - affects patient outcomes. If the thymus contributes to anti-tumor immunity, inadvertently damaging it during radiation therapy could undermine the very immune response that treatment depends on.

Aerts suggests that improving understanding and monitoring of thymic health could eventually help physicians better assess disease risk and guide treatment decisions - particularly for immunotherapy, where patient selection remains imprecise and many patients receive treatments that do not benefit them.

The broader implication is that the medical community may need to revisit its assumptions about which organs matter in adult health. The thymus may not be the only organ dismissed prematurely as vestigial or irrelevant. The tools to ask these questions at scale - AI, large imaging databases, population cohorts - now exist. The thymus, it turns out, was simply the first organ to benefit from someone thinking to look.