Can GLP-1 Receptor Agonists Cause Changes to Thyroid Levels?
GLP-1 receptor agonists do not cause clinically significant changes to thyroid function tests (TSH, T3, T4) in humans, but they carry a black-box warning for medullary thyroid carcinoma (MTC) risk based on rodent studies, making them absolutely contraindicated in patients with personal or family history of MTC or multiple endocrine neoplasia type 2 (MEN 2). 1
Understanding the Thyroid Safety Profile
What Actually Happens in Humans
Thyroid function remains normal – Clinical trials spanning up to 2 years show that mean calcitonin levels in patients treated with liraglutide remained at the lower end of the normal range, with no difference in the proportion of patients exceeding the clinically relevant cutoff of 20 pg/mL 2
No evidence of thyroid hormone disruption – GLP-1 receptor agonists do not alter TSH, free T4, or T3 levels in clinical practice; they are not thyroid medications and do not interfere with thyroid hormone synthesis or secretion 3
GLP-1 receptors are present in human thyroid tissue – These receptors are detected in approximately 18% of papillary thyroid carcinoma specimens and roughly 35% of normal C-cells, but this expression is far lower than in rodents 3
The Rodent Data That Led to the Black-Box Warning
Rodent C-cell tumors are well-documented – In male and female rats, dulaglutide causes dose-related and treatment-duration-dependent increases in thyroid C-cell adenomas and carcinomas after lifetime exposure 1
Mechanism is GLP-1 receptor–mediated in rodents – GLP-1 receptor agonists stimulate calcitonin release, up-regulation of calcitonin gene expression, and subsequent C-cell hyperplasia in rats and mice through direct GLP-1 receptor activation on C-cells 2, 4
This mechanism does not translate to primates – Humans and cynomolgus monkeys have low GLP-1 receptor expression in thyroid C-cells, and GLP-1 receptor agonists do not activate adenylate cyclase or generate calcitonin release in primates; 20 months of liraglutide treatment at >60 times human exposure levels did not lead to C-cell hyperplasia in monkeys 2
Clinical Evidence in Humans
Randomized Controlled Trial Data
Thyroid cancer is a rare event in trials – Meta-analyses of randomized controlled trials show thyroid cancer occurs infrequently in individuals exposed to GLP-1 receptor agonists, with effect estimates that are imprecise but without conclusive evidence of increased risk 5
One MTC case reported in trials – A single case of MTC was reported in a patient treated with dulaglutide in clinical trials, but this patient had pretreatment calcitonin levels approximately 8 times the upper limit of normal, suggesting pre-existing disease 1
Pharmacovigilance and Observational Data
Disproportionate reporting exists – Analysis of the European pharmacovigilance database (EudraVigilance) showed disproportionality for thyroid cancer reports with GLP-1 analogues, with liraglutide having a proportional reporting ratio of 27.5 (95% CI, 22.7–33.3) and exenatide 22.5 (95% CI, 17.9–28.3) 6
Pharmacovigilance does not prove causation – These studies consistently show a signal of increased reporting but do not address causality; they reflect postmarketing surveillance patterns rather than definitive risk 5
Observational studies are inconsistent – Real-world observational studies show low event rates for thyroid cancer with inconsistent effect estimates among different studies, and these studies are at higher risk of bias 5
Absolute Contraindications
Never prescribe GLP-1 receptor agonists to:
- Patients with personal history of medullary thyroid carcinoma 1, 3
- Patients with family history of medullary thyroid carcinoma 1, 3
- Patients with multiple endocrine neoplasia type 2 (MEN 2) 1, 3
Monitoring and Patient Counseling
What to Tell Patients
Counsel about MTC symptoms – Inform patients of thyroid tumor symptoms including a mass in the neck, dysphagia, dyspnea, and persistent hoarseness 1
Routine calcitonin monitoring is not recommended – Routine monitoring of serum calcitonin or thyroid ultrasound is of uncertain value for early detection of MTC and may increase the risk of unnecessary procedures due to low test specificity and high background incidence of thyroid disease 1
When to Measure Calcitonin
If calcitonin is measured and elevated – Significantly elevated serum calcitonin values may indicate MTC; patients with MTC usually have calcitonin values >50 ng/L, and such patients should be further evaluated 1
Thyroid nodules require evaluation – Patients with thyroid nodules noted on physical examination or neck imaging should be further evaluated regardless of GLP-1 receptor agonist use 1
Common Pitfalls to Avoid
Do not confuse thyroid cancer risk with thyroid function changes – GLP-1 receptor agonists do not alter TSH or thyroid hormone levels; the concern is about a theoretical cancer risk, not hypothyroidism or hyperthyroidism 2, 3
Do not screen all patients with calcitonin – Routine calcitonin screening before or during GLP-1 receptor agonist therapy is not recommended and may lead to unnecessary procedures 1
Do not withhold therapy based on unproven risk – Excessive worry about unproven thyroid cancer risk in humans should not lead to underutilizing GLP-1 receptor agonists in patients who could experience substantial cardiovascular and metabolic benefits 5
Do not ignore absolute contraindications – Always screen for personal or family history of MTC or MEN 2 before prescribing; this is a black-box warning and must be respected 1, 3
Bottom Line for Clinical Practice
GLP-1 receptor agonists do not cause changes to thyroid function tests (TSH, T3, T4) in humans. The black-box warning exists because of rodent C-cell tumors, but the mechanism does not translate to primates, and clinical trial data spanning 2 years show no evidence of increased MTC risk in humans without pre-existing risk factors. Screen for MTC/MEN 2 history before prescribing, counsel patients about neck mass symptoms, but do not routinely monitor calcitonin or withhold these highly effective medications based on theoretical concerns. 1, 2, 5