Long-term weight loss and cardiorenal outcomes by baseline BMI in the VERTIS CV trial

Abstract

Aim: To assess weight loss and cardiorenal outcomes by baseline body mass index (BMI) in VERTIS CV.

Methods: Patients with type 2 diabetes and atherosclerotic cardiovascular (CV) disease were randomized to ertugliflozin or placebo. These post hoc analyses evaluated cardiometabolic and cardiorenal outcomes (a composite of death from CV causes or hospitalization for heart failure [HHF], CV death, HHF and an exploratory composite kidney outcome including ≥40% estimated glomerular filtration rate [eGFR] decrease) by baseline BMI, using conventional clinical categories and Cox proportional hazards models.

Results: In total, 8246 adults were randomized (mean age 64.4 years, diabetes duration 13.0 years, BMI 32.0 kg/m², 61% with BMI >30 kg/m²). Absolute body weight reduction was greater with ertugliflozin versus placebo at 3 and 5 years in the overall population (p < 0.001) and across BMI subgroups. Ertugliflozin increased the proportion of participants achieving ≥5% and ≥10% body weight reduction (ertugliflozin 34.9% and 13.6%, placebo 19.4% and 4.1%; odds ratio [95% confident interval, CI], 2.21 [1.76-2.77] and 3.65 [2.39-5.57], respectively) at 5 years. No significant difference was observed in the effect of ertugliflozin on HHF across BMI subgroups (P_interaction = 0.61). Similarly, no significant difference was observed in the effect of ertugliflozin on the kidney composite outcome across BMI subgroups (P_interaction = 0.39). Results were similar for other CV outcomes, and safety was consistent with the known ertugliflozin profile.

Conclusion: Weight loss was observed across baseline BMI and was sustained over 5 years of follow-up. The effects of ertugliflozin on HHF and kidney composite were consistent across baseline BMI.

Keywords: body mass index; cardiovascular and kidney outcomes; ertugliflozin; metabolic outcomes; type 2 diabetes.

FIGURE 1

Change in body weight in the overall population and according to baseline BMI for absolute change (kg) at Year 3 (A) and Year 5 (B) and percentage change at Year 3 (C) and Year 5 (D). LS mean change was evaluated using a constrained longitudinal data analysis model applied to each BMI subgroup separately; the model included fixed effects for treatment, visit (categorical), baseline estimated glomerular filtration rate (eGFR, continuous), baseline BMI (continuous) and treatment by visit interaction. BL, baseline; BMI, body mass index; CI, confidence interval; ERTU, ertugliflozin; LS, least squares; PBO, placebo.

FIGURE 2

Placebo‐adjusted LS mean difference in absolute change in body weight (kg) over time by baseline BMI for ertugliflozin (pooled doses). LS mean difference based on RMANCOVA model that adjusts for baseline, treatment, time, baseline A1C, baseline eGFR (continuous), subgroup, treatment‐by‐subgroup interaction and treatment‐by‐subgroup‐by‐time interaction. BL, baseline; BMI, body mass index; CI, confidence interval; ERTU, ertugliflozin; LS, least squares; PBO, placebo.

FIGURE 3

Proportion of participants with body weight reduction ≥5% in the overall population and according to baseline BMI at (A) Year 3 and (B) Year 5, and body weight reduction ≥10% at (C) Year 3 and (D) Year 5. Logistic regression was used to determine the odds ratio (95% CI) of achieving body weight reductions of ≥5% and ≥10% with ERTU relative to PBO; analysis was adjusted for baseline BMI (continuous) and baseline eGFR (continuous). If there were no body weight data at the specified timepoint, data were imputed as “not at goal”, that is, weight reduction <5% (or 10%). BMI, body mass index; CI, confidence interval; ERTU, ertugliflozin; PBO, placebo.

FIGURE 4

Forest plot of CV outcomes and kidney composite (sustained 40% decline in eGFR, chronic kidney dialysis/transplantation or death from kidney causes) in the overall population and by baseline BMI subgroup. ^aHazard ratio for ertugliflozin versus placebo, based on the stratified Cox proportional hazards model that included treatment, subgroup and treatment‐by‐subgroup interaction as explanatory factors and randomization cohort category as a stratification factor. ^bInteraction p‐value for the two level treatment group (all ertugliflozin versus placebo) for BMI subgroups. BMI, body mass index; CV, cardiovascular; eGFR, estimated glomerular filtration rate; HHF, hospitalization for heart failure; MACE, major adverse cardiovascular events.