GLP-1 Medications and Thyroid Cancer: What the Evidence Actually Shows

Introduction: The Warning That Worries Millions

Open any GLP-1 medication guide — whether for Ozempic, Wegovy, Mounjaro, or Zepbound — and you will encounter the same ominous language: a black box warning about thyroid C-cell tumors. It is the FDA’s strongest safety warning, reserved for drugs with serious or life-threatening risks.

For the millions of people now taking these medications for diabetes and weight loss, this warning raises an unsettling question: Am I trading obesity for cancer?

The short answer — backed by the most current evidence as of mid-2026 — is that the risk appears far smaller than the warning label might suggest, but it is not zero, and it is not the same for everyone. This article unpacks what the data actually shows, where the warning came from, and who genuinely needs to be concerned.

Part 1: Where the Black Box Warning Comes From (Hint: It’s About Rats)

The FDA’s black box warning for GLP-1 receptor agonists did not originate from a cluster of human cancer cases. It came from rodent toxicology studies conducted before these drugs ever reached pharmacy shelves.

The Key Discovery

In 2010, researchers at Novo Nordisk published a seminal paper demonstrating that GLP-1 receptor agonists directly activate thyroid C-cells in rodents PMID: 20203154. When liraglutide was administered to rats and mice, the researchers observed:

• GLP-1 receptors densely expressed on rodent thyroid C-cells
• Stimulation of calcitonin release
• Upregulation of calcitonin gene expression
• Development of C-cell hyperplasia (abnormal cell growth)
• Progression to C-cell adenomas and, in some cases, carcinomas

Follow-up carcinogenicity studies confirmed the pattern. In a 93-week rat study of dulaglutide (Trulicity), researchers found statistically significant increases in diffuse C-cell hyperplasia and adenomas at moderate doses, with numerical increases in C-cell carcinomas at the highest dose PMID: 25860029.

A Crucial Species Difference

Here is the critical detail that the black box warning cannot easily convey: rodents and humans are fundamentally different when it comes to GLP-1 receptors on thyroid C-cells.

A comprehensive review by Rosol (2013) established that GLP-1 receptor agonists produce C-cell hyperplasia and neoplasia in rats and mice through an on-target, species-specific mechanism PMID: 23471186. The study concluded:

“GLP-1 receptors have been demonstrated on normal rodent C-cells, but are either not present or occur in low numbers on C-cells of nonhuman primates and humans. Hyperplasia and neoplasia of C-cells in rodents treated with GLP-1 agonists represent a unique example of an on-target species-specific effect that may not have relevance to humans.”

In plain English: rats have lots of GLP-1 receptors on their thyroid C-cells. Humans have very few, if any. The rodent response is essentially a predictable pharmacological effect that does not translate to human biology.

This distinction was reinforced in 2025 by Hollingshead and Radi, who reviewed decades of pharmaceutical drug-induced thyroid tumor data in rats and found that GLP-1 agonist-associated C-cell tumors are the most notable exception — precisely because their human relevance remains uncertain and “a topic of ongoing clinical safety surveillance analysis” PMID: 40083171.

Part 2: The Human Data — What Do Large Studies Actually Show?

If the rodent data is not directly translatable, what do human studies tell us? The evidence falls into three categories: randomized controlled trials (RCTs), population-based observational studies, and pharmacovigilance databases.

Meta-Analyses of Randomized Trials

The most rigorous human evidence comes from meta-analyses pooling data across dozens of randomized controlled trials.

Silverii et al. (2025) analyzed 50 randomized controlled trials lasting at least 52 weeks, comparing GLP-1 receptor agonists to any comparator. The meta-analysis found that GLP-1RA treatment was not associated with a significant increase in overall cancer risk (MH-OR 1.05, 95% CI 0.98–1.13). However, there was one notable signal: a statistically significant increase in thyroid cancer specifically (MH-OR 1.55, 95% CI 1.05–2.27), which was more evident in longer-term trials PMID: 40437949.

This finding demands careful interpretation. The absolute number of thyroid cancer cases in these trials was small. With a baseline thyroid cancer incidence of approximately 13.5 per 100,000 person-years in the general population, an odds ratio of 1.55 translates to a very modest absolute risk increase — roughly 7–8 additional cases per 100,000 person-years. Furthermore, the meta-analysis could not distinguish between medullary thyroid cancer (the type seen in rodents) and the far more common papillary thyroid cancer.

Ko et al. (2026) conducted the most comprehensive systematic review and meta-analysis to date, published in the Annals of Internal Medicine. Their analysis encompassed both GLP-1 receptor agonists and dual GIP/GLP-1 agonists (such as tirzepatide), examining cancer risk across multiple organ systems. The study provided the most up-to-date risk stratification across individual cancer types PMID: 41359966.

The Novo Nordisk Comprehensive Assessment

In 2026, Vilsbøll et al. published a uniquely thorough risk assessment drawing from three distinct data sources:

1. Clinical trial data: 93 liraglutide and semaglutide phase 2 and 3 trials, including six cardiovascular outcome trials (CVOTs)
2. Post-marketing surveillance: The Novo Nordisk global safety database
3. Real-world data: The US Merative MarketScan Commercial Database

This analysis evaluated the association between GLP-1RA use and thyroid cancer risk using the most comprehensive dataset assembled for this question. Hazard ratios were estimated for overall thyroid cancer risk across treatment groups PMID: 41287564.

Population-Based Cohort Studies

Bea et al. (2024) conducted a population-based cohort study comparing thyroid cancer risk between GLP-1RA users and DPP-4 inhibitor users among patients with type 2 diabetes. Using a large national database, the study accounted for multiple confounders and provided real-world risk estimates outside the controlled environment of clinical trials PMID: 37735822.

Pharmacovigilance Data: A Note of Caution

The FDA Adverse Event Reporting System (FAERS) has consistently shown disproportionality signals for thyroid cancer with GLP-1 receptor agonists. These pharmacovigilance analyses report elevated reporting odds ratios — sometimes dramatically so — for semaglutide, liraglutide, dulaglutide, and tirzepatide.

However, FAERS data has well-known limitations: it is a voluntary reporting system subject to stimulated reporting bias (media attention drives reporting spikes), cannot establish causation, lacks denominator data (we do not know how many people are taking the drug), and does not adjust for confounders like obesity itself being a risk factor for several cancers.

These signals are worth monitoring, but they do not constitute evidence of causation.

Part 3: Medullary vs. Papillary Thyroid Cancer — Why the Distinction Matters

One of the most important concepts for patients and clinicians to understand is that not all thyroid cancers are the same.

Medullary Thyroid Cancer (MTC)

MTC arises from the parafollicular C-cells of the thyroid — the same cell type affected in rodent studies. It accounts for only 3–5% of all thyroid cancers. MTC can be sporadic or hereditary (associated with MEN2 syndromes).

• The rodent concern is specifically about MTC. The C-cell hyperplasia → adenoma → carcinoma pathway observed in rats mirrors the theoretical risk pathway for MTC in humans.
• The black box warning focuses on MTC risk, explicitly contraindicated in patients with personal or family history of MTC or MEN2.
• Human data on GLP-1 and MTC specifically is extremely limited because MTC is rare, making it difficult to detect even in large studies.

Papillary and Follicular Thyroid Cancer (Differentiated Thyroid Cancer, DTC)

DTC arises from follicular cells (not C-cells) and accounts for over 90% of thyroid cancers. There is no known mechanistic pathway by which GLP-1 receptor activation would cause DTC, as follicular cells do not express GLP-1 receptors in any meaningful quantity.

However, some observational data has suggested an association between GLP-1 use and DTC. A 2025 clinical review by Kelly and Sipos in the Journal of Clinical Endocrinology & Metabolism addressed this head-on PMID: 39400117:

“Emerging data also suggest an increased incidence of differentiated thyroid cancer (DTC) in patients treated with these agents. Other studies, though, have not confirmed an association between GLP-1RAs and DTC. With conflicting results concerning thyroid cancer risk, there is no clear consensus.”

A 2026 study from Memorial Sloan Kettering Cancer Center specifically examined whether GLP-1RA exposure affects structural progression or recurrence in patients who already have differentiated thyroid cancer PMID: 41693146.

The plausible explanation for any observed DTC association is detection bias, not causation: patients on GLP-1s see doctors more frequently, undergo more physical examinations, and may receive more thyroid imaging — all of which increases the likelihood of incidental discovery of small, indolent thyroid cancers that might otherwise never be detected.

Part 4: Who Should NOT Take GLP-1 Medications

Based on the current evidence, the contraindications are clear and specific:

Absolute Contraindications

• Personal history of medullary thyroid carcinoma (MTC)
• Family history of medullary thyroid carcinoma
• Multiple Endocrine Neoplasia syndrome type 2 (MEN2A or MEN2B)

These contraindications are based on the theoretical risk from rodent data and the precautionary principle. If you have any of these conditions, GLP-1 receptor agonists are not appropriate for you, and alternative treatments should be discussed with your physician.

What About a History of Papillary Thyroid Cancer?

This is a gray area. Because papillary thyroid cancer arises from follicular cells — not C-cells — the mechanistic concern does not apply. However, given the unresolved questions in the observational literature, many endocrinologists take a cautious approach:

• Patients with a remote history of low-risk papillary thyroid cancer (fully treated, no evidence of disease) may still be candidates for GLP-1 therapy after shared decision-making with their endocrinologist.
• Patients with active surveillance of thyroid nodules or recent DTC diagnosis require individualized assessment.

Family History of Non-Medullary Thyroid Cancer

A family history of papillary or follicular thyroid cancer is not a contraindication to GLP-1 use. The black box warning specifically references medullary thyroid carcinoma and MEN2.

Part 5: Should You Be Screened Before Starting a GLP-1?

There is currently no formal guideline from the American Thyroid Association or the Endocrine Society mandating universal thyroid cancer screening before initiating GLP-1 therapy. However, clinical practice is evolving.

Kelly and Sipos (2025) outlined a practical clinical approach in their JCEM review PMID: 39400117:

Recommended Pre-Treatment Assessment

1. Detailed history: Specifically ask about personal or family history of thyroid cancer, with emphasis on MTC and MEN2.
2. Neck examination: Palpation of the thyroid for nodules, asymmetry, or lymphadenopathy.
3. Consider thyroid ultrasound if:
   • Palpable nodule is detected
   • Patient has risk factors (neck irradiation history, family history of thyroid cancer)
   • Clinical judgment warrants further investigation
4. Serum calcitonin: Not routinely recommended but may be considered in patients with suspicious family history. (Calcitonin is the tumor marker for MTC.)

During Treatment

• Neck examination at annual follow-up visits
• Thyroid function tests (TSH) as clinically indicated — note that GLP-1RAs can affect TSH levels in patients on levothyroxine
• No routine surveillance imaging is recommended in the absence of symptoms or examination findings

Part 6: Bottom Line — Absolute Risk Assessment

What does all this mean for an individual patient considering or currently taking a GLP-1 medication?

The Numbers in Perspective

* Obesity is an established risk factor for at least 13 cancer types. Weight loss from GLP-1 therapy may reduce these risks over time, potentially offsetting any small thyroid cancer signal.

Risk-Benefit Calculus

For a patient with class II or III obesity and/or uncontrolled type 2 diabetes, the proven benefits of GLP-1 receptor agonists include:

• 15–22% average body weight reduction
• Significant HbA1c reduction
• 20% reduction in major adverse cardiovascular events (in high-risk populations)
• Improvements in blood pressure, lipids, and inflammatory markers
• Potential reduction in obesity-related cancer risk

Weighed against a theoretical and unconfirmed thyroid cancer risk measured in single-digit cases per 100,000 patient-years, the risk-benefit ratio strongly favors treatment for the vast majority of eligible patients — provided they do not have the specific contraindications listed above.

What We Still Don’t Know (2026)

Honest medicine requires acknowledging uncertainty. As of mid-2026, the key unknowns include:

1. Long latency: If GLP-1s do increase human thyroid cancer risk, the latency period could be decades. Most trial data covers only 2–5 years.
2. Real MTC risk: The rodent data specifically concerns medullary thyroid cancer, but MTC is too rare for even large meta-analyses to adequately assess.
3. Tirzepatide and dual agonists: The dual GIP/GLP-1 agonists are newer, with less long-term safety data than single-agonist GLP-1s.
4. Interaction with obesity-related cancer risk: Weight loss from GLP-1s may reduce other cancer risks. The net oncologic effect over a lifetime is unknown.

Practical Takeaways

1. If you have no personal or family history of MTC or MEN2: The absolute thyroid cancer risk from GLP-1 medications appears to be very low. The cardiometabolic benefits likely outweigh this risk.
2. If you have a history of MTC or MEN2 (personal or family): GLP-1 medications are contraindicated. Discuss alternative treatments with your doctor.
3. If you have or had papillary thyroid cancer: The decision requires individualized assessment. The mechanistic concern does not apply, but caution is warranted given conflicting observational data.
4. Get a neck check: Before starting a GLP-1, a simple neck examination by your physician and a thorough family history review is prudent.
5. Stay informed: This is an active area of research. Post-marketing surveillance studies with longer follow-up will provide more definitive answers in the coming years.

References

1. Bjerre Knudsen L, Madsen LW, Andersen S, et al. Glucagon-like peptide-1 receptor agonists activate rodent thyroid C-cells causing calcitonin release and C-cell proliferation. Endocrinology. 2010;151(4):1473-1486. PMID: 20203154

2. Rosol TJ. On-target effects of GLP-1 receptor agonists on thyroid C-cells in rats and mice. Toxicol Pathol. 2013;41(2):303-309. PMID: 23471186

3. Hollingshead BD, Radi ZA. Human relevance of pharmaceutical drug-induced thyroid tumors in rats, labeling implications, and carcinogenicity study requirements. J Appl Toxicol. 2025;45(8):1426-1439. PMID: 40083171

4. Ko A, Chang YC, Bahar F, et al. Risk for cancer with glucagon-like peptide-1 receptor agonists and dual agonists: a systematic review and meta-analysis. Ann Intern Med. 2026;179(2):216-229. PMID: 41359966

5. Silverii GA, Marinelli C, Bettarini C, et al. GLP-1 receptor agonists and the risk for cancer: a meta-analysis of randomized controlled trials. Diabetes Obes Metab. 2025;27(8):4454-4468. PMID: 40437949

6. Vilsbøll T, Stellfeld M, Aroda VR, et al. Assessment of thyroid cancer risk associated with glucagon-like peptide 1 receptor agonist use. Diabetes Obes Metab. 2026;28(2):1499-1507. PMID: 41287564

7. Kelly CA, Sipos JA. Approach to the patient with thyroid nodules: considering GLP-1 receptor agonists. J Clin Endocrinol Metab. 2025;110(6):e2080-e2087. PMID: 39400117

This article was last reviewed on July 2, 2026. Medical information evolves rapidly; always consult your healthcare provider for personalized medical advice.