GLP-1 Receptor Agonists Are Safe to Use in Hurthle Cell Thyroid Cancer
GLP-1 receptor agonists can be safely used in patients with Hurthle cell (oncocytic) thyroid carcinoma, as the FDA contraindication applies specifically and exclusively to medullary thyroid carcinoma and multiple endocrine neoplasia type 2, not to differentiated thyroid cancers like Hurthle cell carcinoma.
Understanding the Contraindication
The FDA black box warning for GLP-1 receptor agonists (particularly longer-acting agents: semaglutide, liraglutide, exenatide ER, and dulaglutide) stems from C-cell tumors observed in rodent studies 1, 2. This contraindication is limited to:
- Personal or family history of medullary thyroid carcinoma
- Multiple endocrine neoplasia syndrome type 2
The FDA has explicitly stated that patients taking GLP-1 RAs do not need monitoring for medullary thyroid carcinoma (e.g., with calcitonin levels) 3. Importantly, the human relevance of rodent C-cell findings remains unknown, and medullary thyroid carcinoma has not been shown to be caused by GLP-1 RAs in humans 3.
Hurthle Cell Carcinoma Is Not a Contraindication
Hurthle cell (oncocytic) carcinoma is a differentiated thyroid cancer, not a medullary thyroid cancer. The two are fundamentally different:
- Medullary thyroid cancer arises from parafollicular C-cells that produce calcitonin
- Hurthle cell carcinoma arises from follicular epithelial cells and is classified as a variant of follicular thyroid cancer
Recent high-quality evidence directly addresses this distinction. A 2026 retrospective cohort study of 1,072 patients with differentiated thyroid cancer (including oncocytic carcinoma) found that GLP-1RA exposure was not significantly associated with recurrence or progression (HR 0.87,95% CI 0.62-1.21, P=0.39) over a median follow-up of 69 months 4. The study specifically included one patient with oncocytic carcinoma among those with RET fusion-positive thyroid cancers 5.
Recent Safety Data
Multiple 2025-2026 studies provide reassuring evidence:
A multisite international cohort of 98,147 GLP-1RA users found no increased risk of thyroid cancer overall (pooled HR 0.81,95% CI 0.59-1.12) 6
An active surveillance study of 18 patients with papillary thyroid carcinoma exposed to GLP-1RA showed no effect on tumor growth kinetics over median 5.5 years follow-up 7
A large real-world analysis of 144,410 GLP-1RA users demonstrated reduced overall cancer incidence (HR 0.905), though an initially elevated thyroid cancer signal (HR 1.411) became nonsignificant at 3 years, suggesting surveillance bias rather than true carcinogenic effect 8
Critical Caveat About Early Detection
One 2025 study found increased thyroid cancer diagnoses within the first year of GLP-1RA initiation (HR 1.85,95% CI 1.11-3.08) 9. However, this likely represents enhanced early detection rather than causation, as the risk attenuated with longer follow-up. This finding should not preclude GLP-1RA use in patients with established Hurthle cell carcinoma, but clinicians should be aware that baseline thyroid evaluation may be prudent before initiation.
Clinical Recommendation
For patients with Hurthle cell thyroid carcinoma who have indications for GLP-1RA therapy (type 2 diabetes, obesity, cardiovascular disease prevention), these medications should not be withheld. The cardiometabolic benefits—including MACE reduction with liraglutide, semaglutide, and dulaglutide 3—are substantial and evidence-based. Eligible patients with differentiated thyroid cancers, including Hurthle cell carcinoma, should not be denied these benefits 4.
The only absolute contraindications remain medullary thyroid carcinoma and MEN-2 syndrome 1, 2, 3.