Introduction: In patients with T2D at high cardiovascular (CV) risk, several classes of medications have been found to reduce incidence of CV events. However, long-term CV outcomes of T2D medications in lower risk patients are unknown.

Methods: We conducted a causal inference study by emulating a target trial using observational data from 12 health systems and insurance plans across the US. Eligible individuals were patients with T2D and hyperglycemia (HbA1c 7.0-11.0% or equivalent glucose levels) at moderate CV risk (no CV disease) on metformin monotherapy who initiated a second T2D medication between 01/01/13 and 09/01/21, with eGFR ≥ 45 ml/min/1.73m2 and with no contraindications to study medications. We compared patients initiating one of the following medication classes: DPP4s, GLP1s, SGLT2s and sulfonylureas (SUs). In each treatment arm we estimated the 5-year risk of a composite outcome (MI, ischemic CVA, HF hospitalization or CV death) after adjustment for demographics, comorbidities and laboratory values at baseline.

Results: We studied 55,441 patients followed for a median of 32 (IQR 18-54) months, with a median age of 60 (IQR 52-68) years and median baseline HbA1c of 7.8% (IQR 7.3-8.6%). Of these, 11,195 patients initiated a DPP4, 6,920 a GLP1, 7,911 an SGLT2 and 29,415 an SU; 40,673 (73.4)% of patients stopped the original study drug class before the end of follow-up after a median of 11 (IQR 3-22) months. A total of 2,566 (4.6%) of patients experienced the primary outcome endpoint. The estimated 5-year risk ratios (95% CI), compared to DPP4s, were 0.93 (0.73 - 1.13) for GLP1s, 0.96 (0.83 - 1.14) for SGLT2s and 1.08 (0.98 - 1.19) for SUs.

Conclusions: Among patients with T2D at moderate CV risk on metformin monotherapy, incidence of CV events was similar among four classes of most commonly used non-insulin T2D medications. Further research is needed to evaluate the risk of non-CV events to help patients and clinicians to make informed decisions on T2D therapy.

Disclosure

A. Turchin: Research Support; Eli Lilly and Company, Novo Nordisk. Consultant; Novo Nordisk, Proteomics International. Research Support; AstraZeneca. L. Petito: Research Support; Omron Healthcare Co., Ltd. E. Hegermiller: None. R.M. Carnahan: None. H. Kitzman: Advisory Panel; Novo Nordisk. M. Lansang: Research Support; Abbott, Dexcom, Inc. Consultant; Glooko, Inc. Research Support; Xeris Pharmaceuticals, Inc. M.E. McDonnell: None. V. Nair: None. E.L. Priest: Research Support; Boehringer-Ingelheim, AstraZeneca, Owkin. V. Willey: Other Relationship; Carelon Research. S. Goel: None. A.F. Kaul: None. M. Hernan: Consultant; ProPublica, Cytel. Other Relationship; ADIA Lab, Flatiron, Foundation Medicine.

Funding

PCORI (DB-2020C2-20308)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.