The US labels for semaglutide, tirzepatide, liraglutide, dulaglutide and orforglipron all open with a boxed warning about thyroid C-cell tumors. The warning comes from studies in rats and mice. The labels say it is not known whether these drugs cause the same tumors in people. Human studies so far point in different directions: a French study found a higher risk, while a larger Scandinavian study found no clear increase. This article lays out what the labels say, what the human data show, and what neither can rule out.
The warning sits in a box at the very top of each label. On semaglutide (Ozempic, Wegovy, Rybelsus), tirzepatide (Mounjaro, Zepbound), liraglutide (Victoza, Saxenda) and dulaglutide (Trulicity), it makes the same three points:
The box ends with a line about testing, covered below (Ozempic label; Zepbound label; Boxed Warning and 5.1).
The wording differs a little from label to label, depending on which animals reacted. All of these come from section 13.1 of each label.
| Drug | Rats | Mice |
|---|---|---|
| Semaglutide | C-cell tumors | C-cell tumors (2-year study) |
| Tirzepatide | C-cell tumors | No tumors (6-month rasH2 study) |
| Liraglutide | C-cell tumors | C-cell tumors (104-week study) |
| Dulaglutide | C-cell tumors | No tumors (6-month rasH2 study) |
| Orforglipron | No tumors (drug not active in rats) | No tumors (drug not active in mice) |
The liraglutide label gives the clearest numbers. In a 104-week rat study, C-cell carcinomas, the malignant kind, were found in 2% of control males and in 8%, 6% and 14% of males on rising doses. The label notes these tumors are rare in standard rat cancer testing. It also says mouse studies showed the C-cell growth depended on the GLP-1 receptor (Victoza label, 13.1).
The newest drug is the odd one out. Orforglipron (Foundayo) does not act on the rat or mouse GLP-1 receptor, and it did not cause tumors in rodents. Its label still carries the boxed warning. The label notes that orforglipron does act on the human receptor, and that the rodent tumors seen with the other drugs are considered an on-target effect of the class (Foundayo label, Boxed Warning and 13.1). More on that drug in our orforglipron trial review.
Symptoms to report. The labels list the symptoms of thyroid tumors as a mass in the neck, dysphagia (trouble swallowing), dyspnea (shortness of breath), and persistent hoarseness. If you notice any of these, tell your prescriber.
Calcitonin and ultrasound. The labels say routine calcitonin blood tests or thyroid ultrasound are "of uncertain value" for catching MTC early in people on these drugs. Section 5.1 explains why: such testing may lead to unneeded procedures, because the calcitonin test often flags people who do not have cancer, and thyroid problems are common in general. It then adds three points (Zepbound label, 5.1):
So the label does not ask for routine screening, but it does not dismiss an abnormal result either. Whether testing makes sense for you is a decision for you and your prescriber.
Inside the liraglutide trials. The Victoza label reports 7 cases of papillary thyroid cancer on Victoza and 1 on comparator drugs in its adult diabetes trials (1.5 vs 0.5 per 1,000 patient-years). Most were under 1 cm and were found in tissue removed after surgery that was prompted by screening built into the trial. Average calcitonin was a little higher than on placebo, but the gap was about 0.1 ng/L or less. The label calls the clinical meaning of these findings unknown. It also says MTC cases have been reported after Victoza went on sale, and that these reports cannot establish or exclude a causal link (Victoza label, 5.1 and 6.1).
Pooled trials. Two meta-analyses, which combine many randomized trials, give a similar estimate but different conclusions:
France (Bezin 2023, Diabetes Care). This study used the French national health insurance database. It took 2,562 people with type 2 diabetes who developed thyroid cancer and compared their past drug use with 45,184 matched people who did not. Using a GLP-1 drug for 1 to 3 years was linked to a higher risk of all thyroid cancer (hazard ratio 1.58, 95% CI 1.27 to 1.95) and of MTC (hazard ratio 1.78, 95% CI 1.04 to 3.05) (Bezin 2023).
Denmark, Norway and Sweden (Pasternak 2024, BMJ). This study followed people who started a GLP-1 drug and compared them with people who started a DPP-4 inhibitor, another diabetes drug class. Comparing two groups who both started a new diabetes drug helps make the groups alike (Pasternak 2024).
| Pasternak 2024 | GLP-1 drugs | DPP-4 inhibitors |
|---|---|---|
| People | 145,410 | 291,667 |
| Thyroid cancers | 76 | 184 |
| Per 10,000 person-years | 1.33 | 1.46 |
| MTC cases | 4 | 11 |
The hazard ratio for any thyroid cancer was 0.93 (95% CI 0.66 to 1.31). The authors say the upper end of that range is consistent with no more than a 31% increase in relative risk. For MTC alone it was 1.19 (95% CI 0.37 to 3.86), a wide range because there were so few cases.
Two details matter when reading this study. Most users took liraglutide (57.3%) or semaglutide (32.9%), and the semaglutide users were followed for only 1.1 years on average. Tirzepatide was not included. The authors also found a non-significant rise in risk in the first year only (hazard ratio 1.47, 95% CI 0.74 to 2.93). They argue a cancer effect that early is unlikely to be real, and that it may reflect more thyroid cancers being found because people on these drugs get checked more. They suggest detection could also explain the French result. They add that a European Medicines Agency review concluded the evidence does not support a causal link.
The honest summary is that the human data neither confirm nor rule out a small increase in thyroid cancer. In the Scandinavian cohort, thyroid cancer was rare in both groups, about 1.3 to 1.5 cases per 10,000 person-years. The pooled trial estimates lean toward a higher risk but are imprecise. Data for tirzepatide and for long-term use are thin.
What the labels are clear on:
Dose Track can help you keep a clear record: every dose and dose change across your compounds, plus side effects logged on the day you notice them. That makes it easier to give your prescriber an accurate history. For the common side effects, see GLP-1 side effects week by week.
This article is general information, not medical advice. Talk to your prescriber before you start, change or stop any medicine.