Weight loss and clinical decongestion across heart failure phenotypes: a post-hoc analysis of a multicentre randomised trial (duel-hf)

Background

Heart failure (HF) remains a major cause of hospitalisation and mortality worldwide. Despite advances in disease-modifying therapy, loop diuretics remain essential for symptom relief in HF. Torasemide has pharmacokinetic properties that may differ from furosemide; however, comparative data across HF phenotypes remain limited.

Purpose

To assess baseline characteristics and short-term clinical changes in patients hospitalised with acute decompensated HF treated with oral torasemide or furosemide, according to left ventricular ejection fraction (LVEF) phenotype.

Methods

This post-hoc analysis of the randomised DUEL-HF trial comparing oral torasemide and furosemide in patients hospitalised with acute decompensated HF. Patients were randomised 1:2 to furosemide (80 mg/day) or torasemide (20 mg/day), with dose adjustment at the discretion of the treating physician. Assessments were performed at baseline and at discharge or day 14, whichever occurred first.

The analysis included 458 patients with available LVEF data, categorised as HFrEF (≤40%), HFmrEF (41–49%) or HFpEF (≥50%). Clinical outcomes included body weight, congestion severity assessed by the Clinical State Assessment Scale (SHOKS), functional capacity and clinical compensation. Changes over time were analysed using linear mixed-effects models including time, treatment, HF phenotype and their interactions, with patient as a random effect.

Results

Across HF phenotypes, body weight decreased during follow-up. When analysed irrespective of diuretic type, the greatest absolute weight reduction was observed in patients with HFrEF, while patients with HFpEF showed a smaller decrease (time × phenotype interaction p=0.005), consistent with a lower baseline congestion burden.

Improvement in congestion severity was substantial across all phenotypes despite differences in baseline clinical severity. Patients with HFrEF, who presented with the most severe baseline profile, demonstrated a larger absolute reduction in SHOKS, likely reflecting higher baseline congestion severity; however, final congestion scores and the proportion of clinically decongested patients were numerically more favourable in HFmrEF and HFpEF.

Baseline clinical severity was similar between furosemide and torasemide groups. Torasemide was associated with greater weight reduction than furosemide (additional −1.45 kg; p=0.034), with similar relative effects across HFrEF, HFmrEF and HFpEF (interaction p=0.81). Improvement in congestion severity was comparable between the two diuretics.

Conclusions

In patients hospitalised with acute decompensated HF, early clinical improvement was observed across all LVEF phenotypes. Although torasemide was associated with greater weight reduction, improvement in congestion severity was comparable between diuretics. These findings indicate that changes in body weight alone do not fully reflect clinical decongestion and support the use of multidimensional congestion assessment in acute HF.