PMID-40437949 – GLP-1RA Cancer Risk: Meta-Analysis of RCTs
Silverii GA et al. "Glucagon-like peptide-1 receptor agonists and risk of cancer: a meta-analysis of randomized controlled trials," Diabetes, Obesity and Metabolism, 2025.
Quick Reference
| Property | Value |
|---|---|
| PMID | 40437949 |
| DOI | — |
| Year | 2025 |
| Journal | Diabetes, Obesity and Metabolism |
| Study Type | Meta-analysis |
| Evidence Level | I |
| Sample | 50 RCTs pooled |
| Peptide(s) Studied | Semaglutide, Tirzepatide |
Key Findings
- Meta-analysis of 50 RCTs found no overall increase in cancer risk with GLP-1RA use
- A thyroid cancer signal was noted, consistent with the known rodent C-cell concern, but absolute risk remains very low
- No significant increase in pancreatic, breast, or colorectal cancer risk
- The thyroid signal was driven primarily by longer-duration trials and may reflect detection bias (increased thyroid monitoring in GLP-1RA-treated patients)
- Results were robust across sensitivity analyses excluding individual agents
Study Design
Meta-analysis of 50 randomized controlled trials reporting cancer-related adverse events in GLP-1RA-treated vs. control groups. Included all available GLP-1RA agents. Pre-specified subgroup analyses by cancer type, agent, and trial duration.
Limitations
- Cancer events reported as adverse events rather than primary endpoints in most trials
- Follow-up durations generally 1-3 years, insufficient for long-latency cancers
- Thyroid cancer signal may reflect ascertainment bias
- Individual cancer type analyses have limited statistical power due to rarity of events
Clinical Relevance
Provides further Level I evidence that GLP-1RAs do not meaningfully increase overall cancer risk. The thyroid signal is worth noting for patient counseling but must be contextualized with the very low absolute risk. Complements the larger Ko 2026 study (PMID-41359966) and addresses the same question from a slightly different trial pool. Supports continued use of semaglutide and tirzepatide with standard thyroid monitoring.
Related
#research #meta-analysis #evidence-level-I #cancer