Glucagon-like peptide-1 receptor agonists and lung cancer risk in patients with type 2 diabetes and pulmonary nodules

Background: Glucagon-like peptide-1 receptor agonists (GLP-1RAs) are associated with lower risk of several cancers in observational studies, whereas randomized-trial meta-analyses indicate little or no effect; lung-cancer-specific evidence is limited. Methods: We emulated a target trial in the TriNetX federated network comparing new users of GLP-1RAs versus metformin among adults with type 2 diabetes and a coded pulmonary nodule. Eligibility was assessed on or before the first qualifying prescription, a 12-month washout applied in both arms, and cohorts were 1:1 propensity score matched on 51 covariates. Outcomes were ascertained from day 90 to day 1,095. Drug-target cis-Mendelian randomization (MR) tested genetically proxied GLP1R expression against lung cancer in three GWAS consortia, with colocalization. Results: Among 12,622 matched pairs, lung cancer incidence was 1.60% with GLP-1RAs versus 2.14% with metformin (HR 0.77, 95% CI 0.64-0.92); distant metastasis (HR 0.58, 0.47-0.71) and systemic anti-cancer therapy (0.58, 0.46-0.74) were also lower with GLP-1RAs. Versus metformin, the lung cancer hazard ratio was 0.72 (0.56-0.92) at a 180-day and 0.77 (0.54-1.09) at a 365-day landmark; it was 0.74 (0.59-0.93) versus DPP-4 and 0.85 (0.69-1.05) versus SGLT2 inhibitors. Negative controls were null. MR showed no association between GLP1R expression and lung cancer; colocalization showed little signal at the locus. Conclusions: GLP-1RA initiation was associated with lower short-term lung cancer incidence versus metformin, but the null SGLT2 comparison and null genetic findings do not support sustained receptor-mediated chemoprevention. Impact: These findings caution against interpreting early observational signals as chemoprevention and define what future GLP-1RA-cancer studies should report.