The Hashimoto’s Breakthrough: Somatic Mutations and the New Model of Autoimmune Disease

A landmark study from the Wellcome Sanger Institute has found that Hashimoto’s thyroiditis, the most common autoimmune disease in the world, is driven in part by acquired DNA mutations in immune cells that switch off the body’s natural restraint. The implications reach far beyond the thyroid.

Presented By Our Partners

For nearly a hundred years, autoimmune disease has been one of medicine’s most stubborn puzzles. Why do roughly five to eight percent of people, mostly women, eventually develop a condition in which the immune system turns on its own tissues? Why does the genetic risk so often fail to translate into actual disease? Why does it appear in one identical twin and not the other? The standard answer, that autoimmunity reflects an unlucky collision of inherited susceptibility and environmental trigger, has been useful but unsatisfying. It explains the shape of the problem without quite explaining the mechanism.

On April 14, 2026, a team of researchers at the Wellcome Sanger Institute, the University of Cambridge, and Cambridge University Hospitals published a study in Nature that may finally begin to close the gap. They examined the immune cells inside the inflamed thyroid glands of patients with Hashimoto’s thyroiditis and Graves’ disease, the two most common autoimmune disorders of the thyroid, and looked for something almost no one had seriously hunted for before. They looked for somatic mutations.

What they found is the kind of result that quietly redraws a map.

The mutations they were not supposed to find

Somatic mutations are DNA changes that occur during a person’s life rather than being inherited at conception. We covered the broader landscape in our pillar piece on the Mosaic Body; in short, every adult is a patchwork of cells with subtly different genomes, and the science of finding and counting these mutations has advanced dramatically in the last few years.

The most useful new tool is an ultra-accurate sequencing method called NanoSeq, developed at the Sanger Institute. Standard DNA sequencing has an error rate too high to spot a mutation present in only one cell in a thousand; NanoSeq can detect mutations at frequencies orders of magnitude lower, with fewer than five errors per billion bases. It is, in other words, the right instrument for finding rare genetic changes hidden inside otherwise normal-looking tissue.

The Cambridge and Sanger team applied NanoSeq, plus single-cell sequencing and spatial DNA analysis, to thyroid biopsies from consenting patients with Hashimoto’s and Graves’ disease. The results were striking. Inside the inflamed thyroid tissue, large populations of B cells, the immune cells that produce antibodies, were carrying acquired mutations that had not been inherited from any parent and would not be passed to any child. The mutations had emerged inside individual B cells during the patient’s life and then spread, as the descendants of those cells multiplied, into clones large enough to detect.

And the mutations were not random. They were concentrated in a small set of genes that, in healthy people, do something remarkable. They act as the brakes on the immune system.

The brakes that failed

The immune system runs hot by design. Its job is to detect and destroy threats, and the cellular machinery that lets a B cell or T cell recognize and attack a foreign invader is built to be aggressive. Without something to slow it down, that aggression would turn on the body itself within hours. So evolution layered on a set of brakes, molecules that sit on the surface of immune cells and signal “stand down” when activated.

The most famous of these brakes are called immune checkpoints. Two of them, PD-L1 (the protein encoded by the gene CD274) and HVEM (encoded by TNFRSF14), have been subjects of intense scientific attention for two decades, because cancer cells have a habit of exploiting them. Many tumors learn to display unusually high levels of PD-L1 on their surface, which docks with PD-1 on incoming T cells and tells them, in effect, “leave me alone.” The discovery that you could break this lie with a drug, by blocking the checkpoint and unleashing the immune system on the tumor, won the 2018 Nobel Prize in Physiology or Medicine and now powers a generation of cancer immunotherapies.

The Sanger team’s finding is the same trick, in reverse, inside an autoimmune disease.

In the inflamed thyroid tissue of Hashimoto’s patients, the researchers found B cells whose CD274 and TNFRSF14 genes had been knocked out by acquired mutations. The brakes were broken. And not in one or two cells, but in tens to hundreds of independent clones inside a single biopsy, each clone carrying its own distinct mutation. This is what biologists call convergent evolution: many separate cells, each finding its own way to disable the same restraint. Convergence on this scale is a signature of selection. The mutated cells were not random survivors. They were being actively favored by the local environment of the inflamed thyroid, outcompeting their unmutated neighbors, expanding their populations, and presumably driving the disease forward.

In some patients, individual B cell clones had accumulated as many as six successive mutations over years before the patient ever experienced symptoms. The pattern looks unnervingly familiar to oncologists. It is the slow, multistep accumulation of driver mutations that we usually associate with the gradual emergence of cancer.

Featured Partner

Invest in the Infrastructure Behind Modern Medicine

As healthcare expands beyond hospital walls, the buildings and campuses supporting that shift are generating compelling returns for investors who move early. The Healthcare Real Estate Fund offers qualified investors direct access to a curated portfolio of medical office, outpatient, and specialty care facilities.

Learn More →

Hashimoto’s thyroiditis, in this picture, is not so much a single event as a slow somatic evolution.

Why this changes the model

The dominant frame for autoimmune disease has long been a two-factor model: inherited predisposition, plus environmental trigger, equals immune system attacking the body. The model has carried real explanatory weight, but it has always struggled with the most basic clinical observations. Hashimoto’s affects somewhere between five and ten percent of women in iodine-sufficient countries, but the inherited risk variants associated with it are far more common than the disease itself. Identical twins, who share their full inherited genome, show only modest concordance for autoimmune thyroid disease. The two-factor model could not quite explain why one body crossed the line and another did not.

The somatic mutation model fills part of that gap. The unlucky cell, in this view, is not the patient as a whole. It is one of their B cells, or one of a handful of B cells, that happens to acquire a mutation in a checkpoint gene at the wrong moment. That cell has a fitness advantage in the local immune environment. Its descendants expand. Over years, perhaps decades, more such mutations accumulate, more clones expand, and at some point the inflammatory pressure on the thyroid crosses a clinical threshold. The patient gets a diagnosis.

This reframing matters in three ways.

First, it links autoimmunity to cancer at the level of cellular evolution. Both are stories about cells that acquire mutations giving them a fitness advantage and then expand. They differ in tissue and in clinical consequence, but they share a deep mechanism. The same family of techniques that has transformed cancer diagnosis and treatment in the last decade may now be poised to do something similar for autoimmunity.

Second, it changes what we might mean by “early diagnosis.” Today, Hashimoto’s is typically detected once thyroid antibodies appear in the blood, or once the gland has been damaged enough to push thyroid hormone levels out of range. The Sanger team’s data raise the possibility that, with the right ultra-sensitive sequencing, we might one day detect the slow accumulation of checkpoint mutations in immune cells well before symptoms appear, perhaps even decades before. What would be done with that information is a separate question, but the diagnostic horizon has just shifted.

Third, it points toward a different kind of treatment. The current standard of care for severe autoimmune disease is broad immunosuppression: drugs that turn down the entire immune system, often at the cost of leaving patients vulnerable to infection. If autoimmune disease is driven, at least in part, by specific clonal populations of mutated B cells, then treatments could in principle target those clones directly, much as modern cancer therapy targets the molecular drivers of a tumor. You would not need to silence the whole immune system. You would only need to remove the rogue clones.

This is still a way off. The Sanger paper itself is careful: the authors describe their work as the beginning of a new phase in understanding autoimmunity, not as a clinical roadmap. Replication in other autoimmune diseases, and other patient populations, will matter enormously. But the conceptual ground has shifted.

What it means if the finding generalizes

The thyroid is a useful starting point because it is small, accessible, and frequently biopsied. But there is no obvious reason the same pattern of clonal checkpoint mutations should be confined to thyroid disease. Many of the most common autoimmune disorders, including type 1 diabetes, lupus, multiple sclerosis, rheumatoid arthritis, and inflammatory bowel disease, share key features with Hashimoto’s: female predominance, gradual onset over years, only partial concordance in identical twins, partial response to broad immunosuppression. If the somatic mutation model holds for thyroid autoimmunity, it would be surprising if it did not hold, in some form, for at least several others.

The implications for women’s health are particularly significant. Autoimmune diseases affect women at roughly three to four times the rate of men, and Hashimoto’s specifically may affect as many as one in ten adult women in some populations. The mechanisms behind that disparity have remained incompletely understood. A model in which slow-accumulating somatic mutations interact with female-specific immune biology, including the sustained immunological flexibility associated with reproduction, may finally provide a coherent explanation.

The implications for cancer biology are also worth pausing on. Cancer immunotherapy works by blocking PD-1 and other checkpoints to free the immune system to attack tumors. About six to twenty percent of patients on these drugs develop thyroid dysfunction as a side effect, in essence, drug-induced autoimmune thyroiditis. The Sanger finding suggests the underlying mechanism is more than coincidence. The drugs are doing chemically what somatic mutation does genetically. They are taking the brakes off. The same biology, approached from opposite directions.

What this means for your health

If you have Hashimoto’s, or any other autoimmune condition, the most important thing this study offers is not yet a treatment but a different mental frame. Autoimmune disease is not a moral failure of the body, nor a punishment for some unidentified lifestyle choice. It is, on this evidence, a slow somatic evolution unfolding inside immune tissue, driven by mutations that no one chose and that accumulate over years. That reframing, on its own, may be useful.

The familiar foundations still apply, and may matter more than we previously appreciated. The same factors that drive harmful somatic mutation across the body, including chronic inflammation, oxidative stress, disrupted sleep, and metabolic dysfunction, are exactly the conditions that any sensible plan for managing autoimmune disease already aims to reduce. Whole-food nutrition that stabilizes blood sugar and reduces inflammatory load. Sleep adequate to support DNA repair. Movement that improves immune surveillance. Stress regulation that lowers the cortisol and inflammatory tone in which mutated cells appear to thrive. These are not magical fixes for autoimmunity. They are, in light of the Mosaic Body view, plausible ways to slow the rate at which the underlying clonal evolution proceeds.

The clinical horizon will move, probably faster than most people expect. Within five to ten years, we are likely to see ultra-sensitive sequencing tests for checkpoint mutation burden in autoimmune disease. We are likely to see early-stage trials of clone-targeted therapies that aim to eliminate specific mutated B cell populations rather than suppressing the whole immune system. We are likely to see, in select centers, risk stratification of asymptomatic people based on the clonal mosaic of their immune cells.

The question Hashimoto’s patients have been asking for a century, why my body, why now, may finally have the beginnings of an answer. It was never your body in the abstract. It was a small, slow accumulation of mutations in a few of your immune cells, expanding quietly, until the inflammation they drove became impossible to ignore.

Frequently asked questions

What did the Nature study actually find?

Researchers at the Wellcome Sanger Institute and the University of Cambridge found that B cells inside the inflamed thyroid glands of patients with Hashimoto’s thyroiditis and Graves’ disease carry acquired somatic mutations that disable immune checkpoint genes, particularly CD274 (PD-L1) and TNFRSF14 (HVEM). These mutations were not inherited and were present in tens to hundreds of independent B cell clones within a single biopsy.

Are somatic mutations a cause of Hashimoto’s thyroiditis?

The study suggests they are a major contributing driver, not necessarily the sole cause. Hashimoto’s likely involves a combination of inherited risk, environmental factors, and slowly accumulating somatic mutations in immune cells that progressively disable the brakes on the immune response. The somatic mutation finding fills in a key piece of mechanism that earlier two-factor models did not adequately explain.

How is this different from inherited genetic risk for autoimmune disease?

Inherited genetic risk for autoimmune disease comes from variants in genes such as the HLA region that are present in every cell of the body and passed to offspring. The somatic mutations described in this study arise during a person’s lifetime in single immune cells, are present only in specific clonal populations, cannot be inherited, and appear to be selected for inside the inflamed tissue itself.

Could this lead to better treatments for Hashimoto’s?

Potentially yes, though not immediately. Current treatments either replace lost thyroid hormone (levothyroxine) or broadly suppress the immune system. If the clonal somatic mutation model holds up, future therapies could in principle target the specific mutated B cell clones driving the disease, similar to how modern cancer therapy targets molecular drivers of tumors. Clinical applications are likely a decade or more away.

Does this finding apply to other autoimmune diseases?

The study examined only autoimmune thyroid disease, but the mechanism it describes, somatic loss of immune checkpoint function in lymphocytes, is biologically plausible for many other autoimmune conditions including type 1 diabetes, rheumatoid arthritis, lupus, multiple sclerosis, and inflammatory bowel disease. Confirmation will require similar studies in those conditions.

Why are immune checkpoint genes important?

Immune checkpoint genes encode proteins that act as brakes on the immune response, preventing it from attacking the body’s own tissues. The two genes most affected in this study, CD274 (PD-L1) and TNFRSF14 (HVEM), are the same genes that cancer immunotherapy drugs target to release the immune system against tumors. When somatic mutations disable these brakes in healthy people, the result appears to be the opposite problem: the immune system attacking the body’s own tissues.

This piece is part of an ongoing series on the Mosaic Body, the emerging view of human biology in which acquired somatic mutations shape health, disease, and aging. See also our pillar essay on why your cells are quietly mutating, and why it’s rewriting medicine, and forthcoming pieces on clonal hematopoiesis and cardiovascular risk and mosaic aging.

Reference: Nicola, P. A., Lawson, A. R. J., et al. (2026). Polyclonal selection of immune checkpoint mutations in thyroid autoimmunity. Nature. DOI: 10.1038/s41586-026-10493-9

Free Daily Briefing

The Latest Longevity Science.
Delivered Every Morning.

Join researchers, physicians, and health professionals getting daily breakthroughs in AI-driven medicine, epigenetics, and longevity research.

Support the research that powers this editorial

Subscribe Free

No spam. Unsubscribe anytime. We respect your inbox.