The labels for Ozempic and Wegovy, both brands of semaglutide, open with the same black box about thyroid tumors. It is the most prominent warning on the label, and it rests on animal data. How the drug works is covered in the semaglutide guide.
This page sets out what the box says, what the animal studies found, who the label rules out, and what the human studies have measured since.
What the boxed warning says
The Ozempic label states that “in rodents, semaglutide causes dose-dependent and treatment-duration-dependent thyroid C-cell tumors at clinically relevant exposures” [1].
It then states the limit of that finding. “It is unknown whether OZEMPIC causes thyroid C-cell tumors, including medullary thyroid carcinoma (MTC), in humans,” because the human relevance “has not been determined.”
The Wegovy label carries the same boxed warning [2]. The wider list of side effects is in semaglutide side effects, and the same warning appears on the labels covered in tirzepatide side effects and liraglutide side effects.
What the rodent studies found
Section 13.1 of the label describes two lifetime studies [1]. Each ran for two years, with semaglutide injected under the skin every day.
In mice, the label reports “a statistically significant increase in thyroid C-cell adenomas and a numerical increase in C-cell carcinomas” in both sexes at clinically relevant exposures.
In rats, adenomas increased in both sexes “at all dose levels.” Carcinomas increased significantly in males at 0.01 mg/kg/day and above, doses the label places at clinically relevant exposures.
Adenomas are benign growths and carcinomas are cancers. Both arise from C-cells, the thyroid cells that make the hormone calcitonin. Medullary thyroid carcinoma is the human cancer of those same cells.
| Species | Length | Exposure vs. the 2 mg weekly human dose | C-cell finding |
|---|---|---|---|
| CD-1 mice | 2 years | Males 2 to 30 times; females 1 to 7 times | Adenomas up significantly; carcinomas up numerically |
| Sprague Dawley rats | 2 years | Below quantification to 3 times | Adenomas up at all doses; carcinomas up in males at 0.01 mg/kg/day and above |
Why the rodent result may not carry over
A 2010 study examined why these drugs act on rodent C-cells [3]. It tested liraglutide, an earlier drug in the class, and most of its authors worked for the drug’s developer.
It found the GLP-1 receptor on rat and mouse C-cells, where the drugs raised calcitonin and caused C-cell growth. In humans and cynomolgus monkeys, C-cells had “low GLP-1 receptor expression.”
Monkeys given liraglutide for 20 months, at more than 60 times human exposure, did not develop C-cell hyperplasia. The authors still wrote that the long-term effects in the human thyroid “remain unknown.”
That is why the label calls the risk unknown rather than absent. The label’s own nonclinical section says the human relevance “could not be determined by clinical studies or nonclinical studies” [1].
Who the label rules out
Ozempic is contraindicated in people with “a personal or family history of MTC” and in people with Multiple Endocrine Neoplasia syndrome type 2, or MEN 2 [1].
The label also tells prescribers to counsel patients about symptoms of thyroid tumors. It lists “a mass in the neck, dysphagia, dyspnea, persistent hoarseness,” meaning a lump, trouble swallowing, shortness of breath or a voice change that does not go away.
What the human studies show
No trial can answer this question alone, because medullary thyroid cancer is rare and trials run for a few years. The evidence comes from large health databases, and it points in different directions.
The label notes that MTC cases were reported after liraglutide reached the market. It says those reports are “insufficient to establish or exclude a causal relationship” [1].
A French study used the national health insurance database, matching 2,562 people with thyroid cancer to 45,184 controls [4]. All had type 2 diabetes and were on second-line diabetes drugs.
One to three years of GLP-1 drug use was associated with more thyroid cancer overall (adjusted hazard ratio 1.58, 95% CI 1.27 to 1.95). It was also associated with more medullary thyroid cancer (1.78, 95% CI 1.04 to 3.05).
A cohort from Denmark, Norway and Sweden compared 145,410 people starting a GLP-1 drug with 291,667 starting a DPP-4 inhibitor, another diabetes pill [5]. Mean follow-up was 3.9 years on the GLP-1 drugs.
Thyroid cancer occurred at 1.33 per 10,000 person-years on GLP-1 drugs and 1.46 on DPP-4 inhibitors. The hazard ratio was 0.93 (95% CI 0.66 to 1.31), which the authors read as no substantially increased risk.
For medullary cancer alone, that cohort’s hazard ratio was 1.19, with a confidence interval from 0.37 to 3.86. That interval is too wide to rule a risk in or out.
A 2025 multinational study followed 460,032 GLP-1 drug users with type 2 diabetes against users of three other drug classes [6]. It found no increased risk of thyroid tumors against any of them.
| Study | Design | Who | Finding |
|---|---|---|---|
| Bezin 2023 | Nested case-control, French national database | 2,562 cases, 45,184 controls, type 2 diabetes | 1 to 3 years of use: all thyroid cancer HR 1.58; medullary HR 1.78 |
| Pasternak 2024 | Cohort, Denmark, Norway and Sweden | 145,410 GLP-1 vs 291,667 DPP-4 starters | HR 0.93 (0.66 to 1.31); medullary HR 1.19 (0.37 to 3.86) |
| Morales 2025 | Cohort, international databases | 460,032 GLP-1 users with type 2 diabetes | No increased risk vs three other drug classes |
Reading the evidence together
The studies measured different things in different countries, and none was a randomized trial. None of them studied people taking semaglutide for weight loss without diabetes.
The Scandinavian authors state what their result can exclude. In the main comparison, the upper limit was consistent with “no more than a 31% increase in relative risk” [5].
The label’s position has not changed with these studies. It still says the risk in humans is unknown, keeps the contraindication for MTC and MEN 2, and tells prescribers to discuss neck symptoms.