Can the Cardiovascular Risk Reductions Observed with Empagliflozin in the EMPA-REG OUTCOME Trial be Explained by Concomitant Changes Seen in Conventional Cardiovascular Risk Factor Levels?
Coleman RL., Gray AM., Broedl UC., Fitchett D., George JT., Woerle HJ., Zinman B., Holman RR.
AIM: The EMPA-REG OUTCOME trial demonstrated a significant reduction in cardiovascular (CV) outcomes in patients with type 2 diabetes given empagliflozin, a selective sodium-glucose cotransporter-2 inhibitor. We performed post-hoc analyses of this trial examining the degree to which empagliflozin-induced changes in conventional CV risk factors might explain the observed CV benefits. MATERIALS AND METHODS: We estimated three-year EMPA-REG OUTCOME CV event rates using a type 2 diabetes-specific clinical outcomes simulation model applied to annual patient-level data. Variables included were atrial fibrillation, smoking, albuminuria, HDL-cholesterol, LDL-cholesterol, systolic blood pressure, HbA1c , heart rate, white cell count, haemoglobin, estimated GFR, and prior histories of ischaemic heart disease, heart failure, amputation, blindness, renal failure, stroke, myocardial infarction or diabetic ulcer. Multiple simulations were performed for each participant to minimize uncertainty and optimize confidence interval precision around CV risk point estimates. Observed and simulated cardiovascular relative risk-reductions were compared. RESULTS: Model-predicted relative risk reductions were smaller than those observed in the trial, with empagliflozin-associated changes in conventional CV risk factor values appearing to explain only 12% of the observed relative risk reduction for all-cause death (4% of 32%), 7% for CV death (3% of 39%) and 15% for heart failure (4% of 29%). CONCLUSIONS: Empagliflozin-associated changes in conventional CV risk factors in EMPA-REG OUTCOME appear to explain only a small proportion of the CV and all-cause death reductions observed. Alternative risk-reduction mechanisms need to be explored to determine if the observed CV risk changes can be explained by other factors, or possibly by a direct drug-specific effect. This article is protected by copyright. All rights reserved.