The thyroid C-cell signal: what the rodent finding means and what the label says

Every GLP-1 medicine carries a boxed warning about thyroid C-cell tumours. The warning comes from a real experimental finding, in rodents, and the question of what it means for people has a more specific answer than either "it's nothing" or "it causes cancer".

What happened in the rodents

C cells are the thyroid's calcitonin-producing cells, distinct from the follicular cells that make thyroid hormone. In 2010 Bjerre Knudsen and colleagues published the toxicology behind liraglutide: in rats and mice, two years of treatment produced calcitonin release, C-cell hyperplasia and eventually C-cell adenomas and carcinomas, in a dose-dependent way and at exposures overlapping the clinical range (PubMed 20203154). Rodent C cells carry dense GLP-1 receptors; activating them raised calcitonin within hours and, sustained for months, drove proliferation.

The same paper reported what happened in cynomolgus monkeys given liraglutide for 20 months at exposures more than 60 times the clinical dose: no calcitonin rise, no C-cell hyperplasia. Human and monkey C cells express far less GLP-1 receptor than rodent C cells, and human C cells did not release calcitonin in response to GLP-1 receptor activation in the assays used. The species difference is the crux.

What the human data show

Calcitonin. Hegedüs and colleagues pooled serial calcitonin measurements from more than 5,000 people in the liraglutide development programme over two years. There was no treatment-related increase, and the small number of people with raised calcitonin did not cluster on drug (PubMed 21209033). If human C cells responded like rodent C cells, calcitonin would have moved within hours; it did not move in two years.

Cancer registries. The data are not unanimous. Bezin and colleagues, using French national claims data, reported an increased risk of thyroid cancer, including medullary thyroid cancer, with one to three years of GLP-1 receptor agonist use (PubMed 36356111). The finding is debated because people on a drug with a thyroid warning get more neck examinations and imaging, which finds more small cancers regardless of cause. Pasternak and colleagues then followed about 145,000 GLP-1 receptor agonist users and 291,000 DPP-4 inhibitor users in Denmark, Norway and Sweden for a mean 3.9 years and found a hazard ratio of 0.93 (95% CI 0.66 to 1.31) for thyroid cancer, which rules out a large increase over that horizon but not a small one, and says nothing about decades (PubMed 38626916).

What the label says, exactly

The Wegovy boxed warning, mirrored across the class, says that semaglutide causes thyroid C-cell tumours in rodents; that it is unknown whether it causes them, including medullary thyroid carcinoma, in humans, because the human relevance of the rodent finding has not been determined; and that it is contraindicated in people with a personal or family history of medullary thyroid carcinoma or with Multiple Endocrine Neoplasia syndrome type 2. It asks prescribers to counsel patients on the symptoms of thyroid tumours and states that routine calcitonin monitoring or thyroid ultrasound is of uncertain value.

Compounded semaglutide and tirzepatide are not FDA approved and carry no FDA-reviewed labelling; the contraindication is a property of the molecule and applies regardless of source.

What is debated

Whether the rodent-human difference is complete (human C cells simply cannot respond) or quantitative (they respond weakly, and decades might matter) cannot be resolved with current data. This site grades the rodent finding C and the human reassurance B; the honest summary is that there is no human signal after roughly a decade of wide use, and that the contraindicated groups should stay contraindicated.

Where to go next

• How this site assigns grades: how we grade mechanism evidence.
• Each agent's approvals and warnings: GLP-1s Explained.
• The lab side, including when a clinician might check thyroid function: GLP-1 Lab Guide.