Abstract Body: Background
SGLT2 inhibitors (SGLT2i) reduce cardiovascular (CV) events in individuals with type 2 diabetes (T2D); however, no direct comparison of individual SGLT2i drugs has been done in patients at moderate CV risk although this subgroup comprises the majority of T2D adults.

Hypothesis
We hypothesized that empagliflozin, canagliflozin, and dapagliflozin will be comparable in their effects on CV outcomes in individuals with T2D at moderate CV risk.

Methods
We used claims data from the OptumLabs® Data Warehouse (OLDW) and 100% sample of Medicare fee-for-service between 2015-2022. We identified all adults age ≥21 years who first started canagliflozin, dapagliflozin, or empagliflozin between 2015-2020, excluding individuals with type 1 diabetes, pregnancy, or metastatic cancer, and restricted the cohort to those with an estimated 1-5% annualized risk of experiencing a major adverse cardiovascular event (MACE) of hospitalization for non-fatal acute myocardial infarction (MI), non-fatal stroke, or all-cause mortality. The primary outcome was time to any MACE and the secondary outcome was time to MACE components and hospitalization for heart failure (HHF). Primary analysis was under the intention-to-treat framework. Random treatment assignment was emulated using propensity scores modeled using the SuperLearner ensemble method, and outcomes were modeled using inverse probability of treatment weighted Cox proportional hazards models.

Results
The final (weighted) study population (N=137,233) had 42,877 individuals in the canagliflozin group (mean age 65.59 [SD, 8.19], 74.9% non-Hispanic White, 57.1% male), 17,871 in the dapagliflozin group (mean age 65.88 [SD, 7.91], 76.3% non-Hispanic White, 56.8% male), and 76,485 in the empagliflozin group (mean age 65.75 [SD, 8.14], 75.3% non-Hispanic White, 56.9% male). The primary outcome was lower with empagliflozin than canagliflozin (HR 0.92; 95% CI 0.87-0.97), driven mostly by the reduced risk of all-cause mortality (HR: 0.86; 95% CI 0.80-0.94) in the empagliflozin group. The primary outcome did not differ between empagliflozin and dapagliflozin, or dapagliflozin and canagliflozin. There was no difference in HHF between the three drugs.

Conclusion
Empagliflozin was superior to canagliflozin in time to MACE in adults with T2D at moderate CV risk starting a new SGLT2i. This suggests that individual SGLT2i may differ in their CV outcomes instead of a universal class effect.