medwireNews: Study findings suggest that people with type 2 diabetes beginning glucagon-like peptide (GLP)-1 receptor agonist therapy have a short-term increased risk of a thyroid cancer diagnosis compared with those who initiate alternative diabetes medications.
However, the researchers emphasize that as the increased risk was “apparent only in the first year” of treatment, it is therefore “likely due to increased vigilance and case detection rather than de novo pathogenesis.”
Recognizing that there are conflicting reports on whether GLP-1 receptor agonist use is associated with thyroid cancer, Rozalina McCoy (University of Maryland School of Medicine, Baltimore, USA) and co-workers conducted a secondary analysis of a target trial emulation that compared the effectiveness of GLP-1 receptor agonists, sodium-glucose cotransporter (SGLT)2 inhibitors, dipeptidyl peptidase (DPP)-4 inhibitor, and sulfonylurea.
The study used data for 351,913 people (49.3% women, average age 65.3 years) with type 2 diabetes, a moderate risk for cardiovascular disease, and no history of thyroid cancer. Between 2014 and 2021, 41,112 of the cohort began treatment with a GLP-1 receptor agonist, 76,093 with a DPP4 inhibitor, 43,499 with a SGLT2 inhibitor, and 191,209 with sulfonylurea therapy.
Over a median 660 days of follow-up, 0.17% of the people given a GLP-1 receptor agonist were diagnosed with thyroid cancer, as were 0.23% of those given a DPP4 inhibitor over a median 1245 days, 0.17% of the patients given a SGLT2 inhibitor over a median 820 days, and 0.20% of the sulfonylurea-treated patients over a median 1207 days.
In modified intention-to-treat analysis, there was an overall 1.24-fold increased risk for thyroid cancer diagnosis among people who had initiated treatment with a GLP-1 receptor agonist versus those using other treatment but this was not statistically significant.
Further analysis indicated that GLP-1 receptor agonist users were a significant 1.85 times more likely to be diagnosed with thyroid cancer within 1 year of treatment initiation than those given other the other diabetes treatments, but that there was no longer a significantly increased risk after 1–2 years of treatment, or beyond 2 years of treatment.
“Similarly, GLP-1 [receptor agonist] use was associated with a higher risk of incident thyroid cancer diagnosis compared to other glucose-lowering drug classes in the as-treated analysis, which censored patients when they discontinued the study drug or added another glucose-lowering agent,” with a hazard ratio of 2.07, the researchers write in JAMA Otolaryngology–Head & Neck Surgery.
“This finding suggests that detection of thyroid cancer occurred both early after treatment initiation and only while receiving ongoing therapy.”
The team notes that thyroid cancer’s “typical latency period” of around 2.5 years and “generally slow progression” makes it “improbable that GLP-1 [receptor agonists] could induce the development of clinically noticeable tumors within just a few months of therapy.”
Instead, the authors suggest that “detection bias” may explain the increased risk within the first year of treatment, perhaps because of “heightened awareness of symptoms suggestive of thyroid nodularity due to existing drug warnings” leading to more comprehensive thyroid investigations than usual.
They found that people who began GLP-1 receptor agonist therapy were significantly more likely to have undergone thyroid ultrasound than those who received other treatments. Specifically, the rates of thyroid ultrasound were significantly higher in the GLP-1 receptor agonist users than other patients after 6 months (1.2 vs 0.8%) and 12 months (2.1 vs 1.5%) of treatment.
McCoy et al conclude: “Taken together, these results imply low confidence in a causal relationship between thyroid cancer and GLP-1[receptor agonist] use, particularly the timescale of the detected association.”
They add: “These results should be interpreted in light of methodological limitations, the relatively short follow-up period (particularly in the as-treated analysis), and potential unmeasured confounders.
“These findings highlight the need for further research to fully understand the nature of this association.”
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2025 Springer Healthcare Ltd, part of the Springer Nature Group
JAMA Otolaryngol Head Neck Surg 2025; doi:10.1001/jamaoto.2024.4852