Diuretic resistance and the efficacy of hydrochlorothiazide in acute decompensated heart failure: A post-hoc analysis of the CLOROTIC trial

Abstract

Aims Diuretic resistance (DR) in acute decompensated heart failure (ADHF) is associated with poor outcomes. The CLOROTIC trial demonstrated hydrochlorothiazide (HCTZ) with intravenous furosemide improved weight loss in ADHF. We assessed performance of the BAN-ADHF DR score and association between DR risk and HCTZ treatment response.

Methods and results We included participants from CLOROTIC and assessed BAN-ADHF performance in identifying lowest net diuretic efficiency phenogroup, as identified with a random forest classifier. Participants were stratified by the BAN-ADHF score (≤12: low DR risk and >12: high DR risk) to compare treatment effect of HCTZ versus placebo across DR risk. Outcomes were weight change and area under the curve (AUC) of dyspnoea visual analogue scale (VAS) at 72/96 h, fluid loss at 72 h, and 30/90-day mortality. Among 220 participants (50.9% male, mean age 83 years), the BAN-ADHF score demonstrated good performance in identifying low net diuretic efficiency phenogroup (area under the receiver operating characteristic curve: 0.81). Weight change and VAS AUC at 72 h was similar across DR risk strata, however, participants with high versus low DR risk had lower fluid loss at 72 h (686 vs. 889 ml, p = 0.003), higher 30-day (11.1% vs. 2.4%, p = 0.035) and 90-day mortality (23.0% vs. 10.6%, p = 0.033). The treatment effect of HCTZ (vs. placebo) on weight change was similar across DR risk (pint = 0.75 at 72 h; pint = 0.50 at 96 h). DR risk modified efficacy of HCTZ versus placebo on dyspnoea VAS, with high (vs. low) DR risk having attenuated improvement at 72 h (−236 vs. 476 mm; pint = 0.03) and 96 h (−400 vs. 777 mm; pint = 0.001).

Conclusions The BAN-ADHF score demonstrated good performance in identifying DR, and high DR risk was associated with worse outcomes. HCTZ treatment response on weight loss was similar across DR risk, but improvement in dyspnoea with HCTZ was attenuated in high DR risk.