This target trial emulation compared adults with type 2 diabetes and a pulmonary nodule starting GLP-1 receptor agonists or metformin. Among 12,622 matched pairs, lung cancer incidence was 1.60% with GLP-1 receptor agonists. With metformin it was 2.14% (HR 0.77, 95% CI 0.64-0.92). The hazard ratio was 0.77 (0.54-1.09) at a 365-day landmark and 0.85 (0.69-1.05) against SGLT2 inhibitors. Cis-Mendelian randomisation, using gene variants that alter GLP1R expression, found no link with lung cancer. The design is observational and open to residual confounding, and outcomes cover only days 90 to 1,095 after starting treatment.
Why it is interesting: The study pairs a trial emulation with genetic evidence to test an observational drug–cancer signal. Neither the SGLT2 inhibitor comparison nor the genetic analysis supported the lower lung cancer incidence seen against metformin.
Cite this item
BibTeX
Free sign-up. A digest of the new papers twice a week, grouped by domain.