Glucagon-like peptide-1 receptor agonist initiation and subsequent risk of gallbladder and biliary tract cancers in type 2 diabetes

Background Glucagon-like peptide-1 receptor agonists (GLP-1RAs) are increasingly used for type 2 diabetes mellitus and obesity. Randomized trials and meta-analyses have demonstrated increased risk of gallbladder and biliary disease with GLP-1RA therapy, raising concern about whether these non-malignant biliary events could translate into increased long-term risk of gallbladder or biliary tract cancers. Because biliary tract cancers are rare and randomized trials are underpowered for malignant endpoints, large real-world analyses are needed. Methods We conducted a retrospective, new-user, active-comparator cohort study using the TriNetX U.S. Collaborative Network. Adults with T2DM initiating GLP-1RAs (without prior insulin) were compared with insulin initiators (without prior GLP-1RA). Patients with prior biliary malignancy or cholecystectomy were excluded. A 1-year washout period preceded index. Propensity score matching (1:1) was performed across demographics, metabolic comorbidities, hepatobiliary conditions, and lifestyle factors. To mitigate reverse causation, an 181-day landmark analysis was applied. The primary outcome was incident gallbladder or biliary tract cancer. Risk ratios (RR), hazard ratios (HR), and Kaplan–Meier estimates were calculated. Results After matching, 323,829 patients were included in each cohort. Baseline measured covariates were well balanced, with standardized mean differences less than 0.10. During landmarked follow-up, incident gallbladder or biliary tract cancer occurred in 14 GLP-1RA initiators and 167 insulin initiators. GLP-1RA initiation was associated with lower observed cancer incidence compared with insulin initiation (risk ratio, 0.084; 95% CI, 0.049–0.145; hazard ratio, 0.116; 95% CI, 0.067–0.201; log-rank p < 0.001). However, important imbalances were present: pre-landmark outcome exclusions were markedly higher in the insulin cohort than in the GLP-1RA cohort, and follow-up duration remained longer in the insulin cohort after matching. Conclusions In this large real-world cohort, GLP-1RA initiation was associated with lower observed incidence of gallbladder and biliary tract cancers compared with insulin initiation. However, the magnitude of association should be interpreted cautiously because residual confounding by diabetes duration and disease severity, landmark-related selection bias, differential follow-up, low event counts, and potential cholecystectomy-related bias may have contributed to the findings. These results do not establish a causal protective effect and should be confirmed in studies using alternative comparators, longer follow-up, and validated oncologic outcomes.