Back to research feed

Endocrine practice : official journal of the American College of Endocrinology and the American Association of Clinical Endocrinologists

May 15, 2026 · via Europe PMC

Divergent Risks of Endocrine and Nonendocrine Malignancies in GLP-1 Receptor Agonist Users: A Propensity-Matched Real-World Analysis of Over 140 000 Patients.

Hung CH, Yen FS, Huang JY, Cheng-Chung Wei J.

Journal ArticleAbstract only — full text under license

AI summary — for doctors, generated from the abstract only

In a large database analysis, GLP-1 receptor agonist users had lower mortality and cancer overall, with an early thyroid cancer signal that disappeared by 3 years.

What was studied

A TriNetX retrospective cohort of 144,410 propensity-matched GLP-1 RA users and nonusers with type 2 diabetes, examining mortality and site-specific cancers.

Key findings

  • All-cause mortality HR 0.693 and cancer incidence HR 0.905
  • Lower digestive (HR 0.710) and respiratory (HR 0.691) cancers
  • Thyroid cancer HR 1.411, no longer significant at 3 years, suggesting surveillance bias
  • Active-comparator analyses gave consistent findings

Clinical relevance

It provides reassurance on cancer safety of GLP-1 RAs in type 2 diabetes.

Who should care

Endocrinology

Limitations

A retrospective database study subject to residual confounding and healthy-user bias.

Cancer risk should not deter GLP-1 RA use in type 2 diabetes, though thyroid surveillance bias should be recognized.

This is an AI-generated summary of the published abstract, not a clinical recommendation. It may not reflect the full text or subsequent updates — verify against the original source before acting on it.

Abstract

Objective

Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are widely used for type 2 diabetes mellitus (T2DM). The long-term impact on cancer risk remains uncertain. This study aimed to evaluate the association between GLP-1 RA and the risk of cancer, including site-specific malignancies.

Methods

This retrospective cohort study utilized the TriNetX database. Adults with T2DM between 2015 and 2023 were identified. After 1:1 propensity score matching, 144 410 GLP-1 RA users were compared with nonusers. Primary outcomes were all-cause mortality and incident malignant neoplasms. Secondary outcomes included site-specific cancers.

Results

GLP-1 RA use was associated with lower all-cause mortality (hazard ratio [HR] 0.693) and a reduced overall incidence of cancer (HR 0.905). Regarding nonendocrine malignancies, significant risk reductions were observed for digestive cancers (HR 0.710) and respiratory cancers (HR 0.691), as well as for other sites including oral, female genital, and central nervous system cancers. In contrast, GLP-1 RA was linked to an increased risk of thyroid gland cancer (HR 1.411). Active-comparator analyses against dipeptidyl peptidase-4 inhibitor and sodium-glucose co-transporter 2 inhibitor yielded broadly consistent findings. Landmark analyses at 3 years showed attenuation of the thyroid cancer signal to nonsignificance, suggesting surveillance bias rather than a true carcinogenic effect.

Conclusion

This real-world analysis demonstrates divergent oncologic risks associated with GLP-1 RA use, while the therapy offers potential protective effects against nonendocrine malignancies (particularly digestive and respiratory). The initially elevated thyroid cancer risk was no longer significant at 3 years, suggesting surveillance bias. These findings support the overall oncologic safety of GLP-1 RA in T2DM management, while warranting further long-term studies to confirm these associations.

DOI

10.1016/j.eprac.2026.05.010

PMID

42142623

Research metadata (titles, abstracts, authors, publication dates) is sourced from Europe PMC, which aggregates content from PubMed/MEDLINE, PubMed Central, and other providers under the U.S. National Library of Medicine (NLM/NCBI). NLM does not claim copyright on abstracts; the original journal publishers or authors may hold rights to the full articles. AI-generated summaries are derived only from the abstract, never the full text, and are always labeled as AI-generated.

Want more like this?

Browse more from the feed. A free account unlocks the full feed, saving and communities.

Back to the research feed

Create a free account · Sign in