OBJECTIVE

To assess the association between glucagon-like peptide 1 receptor agonist (GLP-1RA) use and risk of incident thyroid tumors.

RESEARCH DESIGN AND METHODS

The retrospective, active-comparator new-user cohort study used international administrative claims and electronic health record databases. Participants included patients with type 2 diabetes mellitus (T2DM) with prior metformin therapy initiating a GLP-1RA versus new users of sodium–glucose cotransporter 2 inhibitors (SGLT2is), dipeptidyl peptidase 4 inhibitors (DPP-4is), and sulfonylureas (SUs). The outcome was incident thyroid tumor and thyroid malignancy. Propensity score matching and stratification were used to adjust for confounders with an intention-to-treat and on-treatment strategy. Cox regression was used to estimate hazard ratios (HRs) pooled using a random-effects meta-analysis. Unmeasured confounding was evaluated using negative outcomes, with calibration of the HR.

RESULTS

A total of 460,032 users of GLP-1RAs, 717,792 users of SGLT2is, 2,055,583 users of DPP-4is, and 1,119,868 users of SUs were included. Only U.S. cohorts passed study diagnostics. Thyroid tumor incidence ranged from 0.88 to 1.03 per 1,000 person-years in GLP-1RA cohorts. GLP-1RA exposure was not associated with an increased risk of thyroid tumors compared with SGLT2is, DPP-4is, or SUs (meta-analysis: GLP-1RA vs. SGLT2i HR range from 0.83 [95% CI 0.57–1.27] to 0.95 [0.85–1.06]; GLP-1RA vs. SU HR range from 0.95 [0.75–1.20] to 1.03 [0.87–1.23]; GLP-1RA vs. DPP-4i HR range from 0.78 [0.60–1.01] to 0.93 [0.83–1.04]). Analysis using thyroid malignancy and including a 1-year lag period produced similar conclusions.

CONCLUSIONS

In patients with T2DM initiating second-line treatments, we observed no increased risk of thyroid tumors with GLP-1RA exposure.

This article contains supplementary material online at https://doi.org/10.2337/figshare.28952591.

This content is only available via PDF.
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.