Time from admission to randomization and the effect of empagliflozin in acute heart failure: A post‐hoc analysis from EMPULSE

• Published in: European journal of heart failure 2024-09, Vol.26 (9), p.1976-1983
• Main Authors: Ferreira, João Pedro, Blatchford, Jonathan P., Teerlink, John R., Kosiborod, Mikhail N., Angermann, Christiane E., Biegus, Jan, Collins, Sean P., Tromp, Jasper, Nassif, Michael E., Psotka, Mitchell A., Comin‐Colet, Josep, Mentz, Robert J., Brueckmann, Martina, Nordaby, Matias, Ponikowski, Piotr, Voors, Adriaan A.
• Format: Article
• Language: English
• Subjects: Acute Disease; Acute heart failure; Aged; Benzhydryl Compounds - therapeutic use; Empagliflozin; Female; Glucosides - therapeutic use; Heart Failure - drug therapy; Heart Failure - physiopathology; Hospitalization - statistics & numerical data; Humans; Life Sciences; Male; Middle Aged; Sodium-Glucose Transporter 2 Inhibitors - therapeutic use; Time Factors; Timing of randomization; Treatment Outcome
• ISSN: 1388-9842; 1879-0844; 1879-0844
• DOI: 10.1002/ejhf.3369

Aims Patients hospitalized for acute heart failure (HF) could be enrolled in EMPULSE (NCT04157751) upon haemodynamic stabilization and between 24 h and 5 days after hospital admission. The timing of treatment initiation may influence the efficacy and safety of drugs such as empagliflozin. The aim of this study was to evaluate patient characteristics, clinical events, and treatment effects according to time from admission to randomization. Methods and results The EMPULSE population was dichotomized by median time from hospital admission to randomization (1–2 days vs. 3–5 days). The primary outcome was a hierarchical composite endpoint of time to all‐cause death, number of HF events, time to first HF event, and a ≥5‐point difference in Kansas City Cardiomyopathy Questionnaire total symptom score change from baseline after 90 days, analysed using the win ratio (WR) method. Patients randomized later (3–5 days, average time 3.9 days; n = 312) had a higher risk of experiencing clinical events than patients randomized earlier (1–2 days, average time 1.7 days; n = 215). The treatment effect favoured empagliflozin versus placebo in patients randomized later (3–5 days: WR 1.69, 95% confidence interval [CI] 1.26–2.25) but was attenuated in patients randomized earlier (1–2 days: WR 1.04, 95% CI 0.74–1.44) (interaction p = 0.029). A similar pattern was observed for the composite of HF hospitalization or cardiovascular death and all‐cause hospitalizations (interaction p