Can Retatrutide Cause Thyroid Cancer? What the Evidence Actually Shows

Introduction

Every GLP-1 receptor agonist approved to date — semaglutide (Ozempic, Wegovy), tirzepatide (Mounjaro, Zepbound), liraglutide (Saxenda) — carries the same black box warning about thyroid C-cell tumors. Retatrutide, when approved, will carry the same warning. The language is alarming: "medullary thyroid carcinoma," "C-cell tumors," "contraindicated."

For patients and their physicians, this warning raises an immediate and legitimate question: Is this a real risk, or a regulatory precaution based on animal data that may not apply to humans?

The answer is nuanced but knowable. The warning is real and appropriate — two specific patient populations should genuinely not use any GLP-1 medication including retatrutide. For the vast majority of patients without these specific risk factors, the picture is more reassuring than the warning language suggests.

This article explains where the warning came from, what human data shows, what remains uncertain, and who should take the warning most seriously.

What the Black Box Warning Actually Says

The FDA prescribing information for semaglutide states precisely:

"Semaglutide caused dose-dependent and treatment-duration-dependent thyroid C-cell tumors at clinically relevant exposures in both genders of rats and mice. It is unknown whether semaglutide causes thyroid C-cell tumors, including medullary thyroid carcinoma (MTC), in humans."

Three things to notice in that language:

"Rats and mice." The warning derives from animal studies, not human clinical trials.

"It is unknown whether." The FDA is not stating that these medications cause MTC in humans. It is stating that human causality has not been established.

"Clinically relevant exposures." The doses that caused tumors in rodents were at or near the range used in humans — not extreme overdoses, which would make the finding easier to dismiss.

The contraindication is specific: GLP-1 medications should not be used in patients with a personal or family history of medullary thyroid carcinoma (MTC), or in patients with Multiple Endocrine Neoplasia syndrome type 2 (MEN2).

Important context: Medullary thyroid carcinoma represents only 3–4% of all thyroid cancers. The warning does not apply to papillary thyroid carcinoma, follicular thyroid carcinoma, or other subtypes — which together constitute ~96–97% of thyroid cancers.

Why the Warning Exists: The Rodent Data

When the first GLP-1 receptor agonist (exenatide) entered development, regulatory agencies required standard two-year carcinogenicity studies in rodents. Rats and mice given long-acting GLP-1 agonists — liraglutide, semaglutide, dulaglutide, exenatide — developed dose- and duration-dependent C-cell hyperplasia, C-cell adenomas, and in some cases medullary thyroid carcinoma.

These findings were consistent enough across drugs in the class that the FDA mandated a black box warning for all GLP-1 receptor agonists. The warning is precautionary, based on the precautionary principle appropriately applied: when an animal finding is consistent, plausibly mechanistic, and involves a serious tumor type, regulatory agencies act conservatively.

The mechanism in rodents is understood: GLP-1 receptors on rodent thyroid C-cells are activated by these drugs, leading to calcitonin release and, over time, C-cell proliferation. The concern is that sustained GLP-1 receptor agonism in C-cells drives abnormal cell growth.

Why Rodents ≠ Humans: The Biological Difference

The warning exists because of rodent data. The primary reason many experts believe it doesn't translate directly to humans is a fundamental biological difference between species.

‍
As the Novo Nordisk FDA trial protocol directly states: "The C-cell changes in rodents are mediated by the GLP-1 receptor, which is not expressed in the normal human thyroid. Accordingly, the risk of GLP-1 receptor-mediated C-cell changes in humans is considered to be low."

This is not an excuse invented by a drug company. It reflects the position of the American Thyroid Association and independent endocrinologists: the specific pathway by which GLP-1 agonism drives C-cell tumors in rodents — receptor-mediated C-cell activation — operates on physiology that differs fundamentally between species.

What Human Data Shows: Calcitonin

Calcitonin is the biomarker for C-cell activity. If GLP-1 medications were stimulating human thyroid C-cells in the way they do in rodents, calcitonin levels should rise. Across multiple large clinical programs, they haven't.

A review of 93 clinical trials of liraglutide and semaglutide across 101,732 participants and 206,950 patient-years of exposure found no consistent clinically meaningful calcitonin elevation. The LEADER trial (liraglutide, 3.5–5 years of follow-up) specifically confirmed no effect on calcitonin levels and no increase in new MTC diagnoses. Multiple cardiovascular outcome trials covering thousands of participants for 3+ years showed only a few isolated thyroid malignancy events, at rates consistent with the general population's background incidence.

The Clayman Thyroid Center — one of the world's highest-volume thyroid cancer surgical programs, treating more than 2,000 thyroid cancer patients per year — published a clinical white paper in 2026 stating that they have not observed an MTC pattern attributable to GLP-1 use in their patient population.

What Human Data Shows: Population Studies

Where the Evidence Is Less Reassuring

Honesty requires acknowledging a conflicting signal.

A retrospective French cohort study (Bezin et al.) found a 58% increased risk of thyroid cancer with 1–3 years of GLP-1 exposure. A separate Diabetes Care analysis (July 2025) also found a signal for thyroid tumors in retrospective GLP-1 users.

These findings are real and cannot be entirely dismissed. However, the research community has identified a significant confounder: detection bias. Patients on GLP-1 medications — increasingly monitored by endocrinologists — receive more thyroid imaging and blood tests. More monitoring finds more pre-existing nodules and incidental cancers that were there before treatment began. This systematically inflates the apparent incidence in treated patients compared to controls who receive standard care monitoring.

The Clayman Thyroid Center's white paper and the CancerNetwork review of the Novo Nordisk data both conclude that this detection bias is the most plausible explanation for the observational signal. Prospective trials with calcitonin monitoring — which would capture a real C-cell effect — consistently show no signal.

The honest summary: retrospective observational data shows a signal; prospective trials and calcitonin data do not. Detection bias is the strongest explanation. Human causality for MTC is not established, but cannot be called impossible.

Retatrutide Phase 2 Data

The Phase 2 retatrutide trial (Jastreboff et al., NEJM, 2023, N=338, 48 weeks) followed participants with calcitonin monitoring as a standard safety measure. No cases of medullary thyroid carcinoma were reported. No clinically meaningful calcitonin elevation was documented.

This is consistent with every other GLP-1 drug's Phase 2 and Phase 3 human data. It is also a short-term dataset (48 weeks, ~338 participants) — insufficient to definitively rule out a long-term, rare risk. The TRIUMPH Phase 3 trials include calcitonin monitoring as part of their exclusion and discontinuation criteria, which means safety signals would be captured across 2,000+ additional participants over longer durations.

Who Should Genuinely Avoid Retatrutide and All GLP-1 Medications

Two populations face a genuine, firm contraindication — not a theoretical concern:

Personal history of medullary thyroid carcinoma: If you have ever been diagnosed with MTC, no GLP-1 receptor agonist should be used. This is an absolute contraindication.

Family history of medullary thyroid carcinoma: First-degree relatives of MTC patients have significantly elevated genetic risk. GLP-1 agonists are contraindicated.

Multiple Endocrine Neoplasia syndrome type 2 (MEN2): MEN2 is a hereditary condition that includes MTC as a core component. GLP-1 agonists are contraindicated.

For these patients, the theoretical mechanism — even if not confirmed in humans — represents an unacceptable risk given the alternative options and the pre-existing vulnerability.

For all other patients, the contraindication does not apply. The Clayman Thyroid Center's position is representative of expert consensus: evidence-based counseling should address the narrow MTC/MEN2 risk specifically, without extending concern to all thyroid cancers, most of which have no established mechanistic relationship with GLP-1 receptor agonism.

See our eligibility guide for the full list of retatrutide contraindications.

What This Means for Patients Without MTC/MEN2 Risk

If you do not have a personal or family history of MTC, and do not have MEN2:

The black box warning reflects a real animal finding that led to appropriate regulatory caution. It does not reflect established human causality. Human calcitonin data, primate data, and multiple large population studies do not support a meaningful increase in MTC risk for this population.

The same is true for papillary and follicular thyroid cancer — the large majority of thyroid cancers — for which no mechanistic relationship to GLP-1 agonism has been established. The FDA warning is narrow and specific to MTC; broader extrapolation to all thyroid cancer types is not scientifically justified per current evidence.

For a patient facing significant obesity-related health risks — cardiovascular disease, type 2 diabetes, joint damage, sleep apnea — the risk-benefit calculation for retatrutide does not meaningfully change because of a black box warning that derives from rodent physiology that differs fundamentally from human thyroid biology.

Conclusion

The thyroid cancer warning on retatrutide and all GLP-1 medications is real, appropriate, and based on consistent animal data. Two patient groups — those with personal or family history of MTC, and MEN2 patients — should not use these medications.

For everyone else, the evidence is substantially more reassuring. Human C-cells express far fewer GLP-1 receptors than rodent C-cells. Calcitonin monitoring across tens of thousands of trial participants has not shown clinically meaningful elevation. Multiple prospective cardiovascular outcome trials covering 3+ years of exposure show no MTC signal. The Scandinavian population study of 100,000 GLP-1 users found no clinically meaningful thyroid cancer increase.

What remains honest to acknowledge: observational retrospective studies have shown an elevated signal for thyroid cancer, likely explained by detection bias but not definitively resolved. Long-term human data on retatrutide specifically will accumulate through the TRIUMPH trials and post-marketing surveillance after approval.

The bottom line for most patients: the black box language is alarming, but it reflects a precautionary response to rodent findings that have not yet manifested as a confirmed human risk in any prospective trial.

Sources

• Jastreboff AM, et al. "Triple-Hormone-Receptor Agonist Retatrutide for Obesity — A Phase 2 Trial." New England Journal of Medicine, 2023;389:514–526.
• Feier CVI, et al. "Assessment of thyroid carcinogenic risk and safety profile of GLP-1 receptor agonist semaglutide therapy for diabetes mellitus and obesity: a systematic literature review." International Journal of Molecular Sciences, 2024;25(8):4346.
• Novo Nordisk/Balachandra S, et al. "Assessment of thyroid cancer risk associated with glucagon-like peptide 1 receptor agonist use." PMC12803555.
• Clayman Thyroid Center White Paper. "GLP-1 Medications and Thyroid Cancer Risk: 2026 Evidence Review." February 2026. thyroidcancer.com.
• Scandinavian cohort study: GLP-1 receptor agonists and thyroid cancer. BMJ, 2025. (~100,000 GLP-1 users, Denmark/Norway/Sweden)
• Eli Lilly press release: TRIUMPH-4 Phase 3 topline results. December 11, 2025.
• Eli Lilly press release: TRIUMPH-1 Phase 3 topline results. May 21, 2026.
• FDA prescribing information: Wegovy (semaglutide) and Zepbound (tirzepatide). Boxed warning language.