Metformin Use and Clinical Outcomes among Patients with Diabetes Mellitus with or Without Heart Failure or Kidney Dysfunction: Observations from the SAVOR-TIMI 53 Trial

Background: Metformin is first-line therapy for type 2 diabetes mellitus, although its effects on the cardiovascular system are unproved. Methods: In this post hoc analysis, patients in SAVOR-TIMI 53 (Saxagliptin and Cardiovascular Outcomes in Patients With Type 2 Diabetes Mellitus) with baseline biomarker samples (n=12 156) were classified as ever versus never taking metformin during the trial period. Associations between metformin exposure and outcomes were estimated with inverse probability of treatment weighting Cox modeling for the composite end point of cardiovascular death, myocardial infarction, or ischemic stroke, as well as cardiovascular death and all-cause mortality, with biomarkers included as covariates. Additional sensitivity analyses included propensity score matching and Cox multivariable models. Results: Of the 12 156 patients with baseline biomarker samples, 8971 (74%) had metformin exposure, 1611 (13%) had prior heart failure, and 1332 (11%) had at least moderate chronic kidney disease (estimated glomerular filtration rate ≤45 mL·min^-1·1.73 m^-2). Metformin use was associated with no difference in risk for the composite end point (hazard ratio for inverse probability of treatment weighting, 0.92 [95% CI, 0.76-1.11]) but lower risk of all-cause mortality (hazard ratio for inverse probability of treatment weighting, 0.75 [95% CI, 0.59-0.95]). There was no significant relationship between metformin use and these end points in patients with prior heart failure or moderate to severe chronic kidney disease. Conclusions: In a cohort of 12 156 patients with type 2 diabetes mellitus and high cardiovascular risk, metformin use was associated with lower rates of all-cause mortality, including after adjustment for clinical variables and biomarkers, but not lower rates of the composite end point of cardiovascular death, myocardial infarction, or ischemic stroke. This association was most apparent in patients without prior heart failure or moderate to severe chronic kidney disease. Clinical Trial Registration: URL: https://www.clinicaltrials.gov. Unique identifier: NCT01107886.