Diuretic therapy as prognostic enrichment factor for clinical trials in patients with heart failure with reduced ejection fraction

Background Loop diuretics are the mainstay of congestion treatment in patients with heart failure (HF). We assessed the association between baseline loop diuretic use and outcome. We also compared the increment in risk related to diuretic dose with conventional prognostic enrichment criteria used in the EMPHASIS-HF trial, which included patients with systolic HF and mild symptoms, such as prior hospitalization and elevated natriuretic peptides. Methods Individual analyses were performed according to baseline loop diuretic usage (furosemide-equivalent dose > 40 mg, 1-40 mg, and no furosemide), and according to enrichment criteria adopted in the trial [i.e. recent hospitalization (< 30 days or 30 to 180 days prior to randomization) due to HF or other cardiovascular cause, or elevated natriuretic peptides]. The primary endpoint was a composite of cardiovascular death or HF hospitalization. Results Loop diuretic usage at baseline (HR for > 40 mg furosemide-equivalent dose = 3.16, 95% CI 2.43-4.11; HR for 1-40 mg = 2.06, 95% CI 1.60-2.65) was significantly associated with a higher risk for the primary endpoint in a stepwise manner when compared to no baseline loop diuretic usage. In contrast, the differences in outcome rates were more modest when using history of hospitalization and/or BNP: all HR ranged from 1 (reference, non-HF related CV hospitalization > 30 days) to 2.04 (HF hospitalization < 30 days). The effect of eplerenone on the primary endpoint was consistent across subgroups in both analyses (P for interaction >= 0.2 for all). Conclusions Loop diuretic usage (especially at doses > 40 mg) identified patients at higher risk than history of HF hospitalization and/or high BNP blood concentrations. Graphic abstract