Association of Empagliflozin Treatment With Albuminuria Levels in Patients With Heart Failure: A Secondary Analysis of EMPEROR-Pooled

Abstract

Importance: Albuminuria, routinely assessed as spot urine albumin-to-creatinine ratio (UACR), indicates structural damage of the glomerular filtration barrier and is associated with poor kidney and cardiovascular outcomes. Sodium-glucose cotransporter-2 (SGLT2) inhibitors have been found to reduce UACR in patients with type 2 diabetes, but its use in patients with heart failure (HF) is less well studied.

Objective: To analyze the association of empagliflozin with study outcomes across baseline levels of albuminuria and change in albuminuria in patients with HF across a wide range of ejection fraction levels.

Design, setting, and participants: This post hoc analysis included all patients with HF from the EMPEROR-Pooled analysis using combined individual patient data from the international multicenter randomized double-blind parallel-group, placebo-controlled EMPEROR-Reduced and EMPEROR-Preserved trials. Participants in the original trials were excluded from this analysis if they were missing baseline UACR data. EMPEROR-Preserved was conducted from March 27, 2017, to April 26, 2021, and EMPEROR-Reduced was conducted from April 6, 2017, to May 28, 2020. Data were analyzed from January to June 2022.

Interventions: Randomization to empagliflozin or placebo.

Main outcomes and measures: New-onset macroalbuminuria and regression to normoalbuminuria and microalbuminuria.

Results: A total of 9673 patients were included (mean [SD] age, 69.9 [10.4] years; 3551 [36.7%] female and 6122 [63.3%] male). Of these, 5552 patients had normoalbuminuria (UACR <30 mg/g) and 1025 had macroalbuminuria (UACR >300 mg/g). Compared with normoalbuminuria, macroalbuminuria was associated with younger age, races other than White, obesity, male sex, site region other than Europe, higher levels of N-terminal pro-hormone brain natriuretic peptide and high-sensitivity troponin T, higher blood pressure, higher New York Heart Association class, greater HF duration, more frequent previous HF hospitalizations, diabetes, hypertension, lower eGFR, and less frequent use of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers and mineralocorticoid receptor antagonists. An increase in events was observed in individuals with higher UACR levels. The association of empagliflozin with cardiovascular mortality or HF hospitalization was consistent across UACR categories (hazard ratio [HR], 0.80; 95% CI, 0.69-0.92 for normoalbuminuria; HR, 0.74; 95% CI, 0.63-0.86 for microalbuminuria; HR, 0.78; 95% CI, 0.63-0.98 for macroalbuminuria; interaction P trend = .71). Treatment with empagliflozin was associated with lower incidence of new macroalbuminuria (HR, 0.81; 95% CI, 0.70-0.94; P = .005) and an increase in rate of remission to sustained normoalbuminuria or microalbuminuria (HR, 1.31; 95% CI, 1.07-1.59; P = .009) but not with a reduction in UACR in the overall population; however, UACR was reduced in patients with diabetes, who had higher UACR levels than patients without diabetes (geometric mean for diabetes at baseline, 0.91; 95% CI, 0.85-0.98 and for no diabetes at baseline, 1.08; 95% CI, 1.01-1.16; interaction P = .008).

Conclusions and relevance: In this post hoc analysis of a randomized clinical trial, compared with placebo, empagliflozin was associated with reduced HF hospitalizations or cardiovascular death irrespective of albuminuria levels at baseline, reduced progression to macroalbuminuria, and reversion of macroalbuminuria.

Trial registration: ClinicalTrials.gov Identifiers: NCT03057977 and NCT03057951.

Figure 1.. Association of Empagliflozin With Baseline Urinary Albumin-to-Creatinine Ratio (UACR) Categories

Cox proportional hazard model adjusted for age (continuous), baseline estimated glomerular filtration rate (eGFR; continuous), baseline left ventricular ejection fraction (continuous), study, region, baseline diabetes status, sex, UACR category, treatment, and treatment-by-UACR category. Composite kidney end point defined by sustained decline in eGFR ≥40% from baseline and sustained eGFR <15 or <10 mL/min/1.73 m² for patients with baseline eGFR ≥ or <30 mL/min/1.73 m², respectively; long-term dialysis; or kidney transplant. Alternative kidney end point defined by sustained decline in eGFR ≥50% from baseline and sustained eGFR <15 or <10 mL/min/1.73 m² for patients with baseline eGFR ≥ or <30 mL/min/1.73 m², respectively; long-term dialysis; or kidney transplant. Total hospitalizations for heart failure were analyzed using a joint frailty model accounting for cardiovascular death and adjusting for the same covariates as the Cox model. Hazard ratios (HRs) for the composite kidney end point should be interpreted with caution due to significant heterogeneity across the 2 trials. The subgroup analysis by trial estimated an HR of 0.51 (95% CI, 0.33 to 0.79) in the EMPEROR-Reduced trial and 0.95 (95% CI, 0.73 to 1.24) in the EMPEROR-Preserved trial; P value for interaction between trials = .02. NA indicates not applicable; NNH, number needed to harm; NNT, number needed to treat; py, person-year.

Figure 2.. Cumulative Incidence Curves Considering All-Cause Mortality as Competing Risk for Time to Incidence of Macroalbuminuria

A, Time to incidence of macroalbuminuria in patients with normoalbuminuria or microalbuminuria (urinary albumin-to-creatinine ratio [UACR] ≤300 mg/dL) at baseline. B, Time to remission to normoalbuminuria or microalbuminuria in patients with macroalbuminuria (UACR >300 mg/dL) at baseline.

Figure 3.. Association of Empagliflozin With Incidence of Macroalbuminuria in Subgroups of Interest

A, Association of empagliflozin with macroalbuminuria in patients with normoalbuminuria or microalbuminuria (urinary albumin-to-creatinine ratio [UACR] ≤300 mg/dl) at baseline. B, Remission to microalbuminuria or normoalbuminuria in patients with macroalbuminuria (UACR >300 mg/dL) at baseline. Cox proportional hazard model adjusted for age (continuous), baseline estimated glomerular filtration rate (eGFR; continuous), baseline left ventricular ejection fraction (LVEF; continuous), study, region, baseline diabetes status, sex, subgroup of interest, treatment, and treatment-by-subgroup category.

Figure 4.. Association of Empagliflozin With Urinary Albumin-to-Creatinine Ratio (UACR) Over Time

A, Overall population. B, Patients with normoalbuminuria (UACR <30 mg/g) at baseline. C, Patients with microalbuminuria (UACR 30-300 mg/g) at baseline. D, Patients with macroalbuminuria (UACR >300 mg/g) at baseline. Models analyzed using a mixed model for repeated measures, including age, baseline estimated glomerular filtration rate, and baseline left ventricular ejection fraction as linear covariates and study, region, baseline diabetes status, sex, log (baseline UACR) by visit, visit by treatment, and week reachable as fixed effects (overall population). For the subgroup analysis, visit by treatment-by-UACR subgroup was used. Gmean indicates geometric mean.