Worsening renal function in heart failure is less harmful if associated with optimization of heart failure medication: insights from a large Registry
D Tomasoni, M Sapienza, L Benson, F Guidetti, C Basile, B N Beer, V Valente, D Stolfo, M Melin, M Metra, G SavareseAbstract
Background
Specific heart failure (HF) medications may temporarily worsen renal function after initiation. Worsening renal function (WRF) has been associated with worse outcomes; whether this association is modified by achieving better use of guideline-directed medical therapy (GDMT) remains unclear.
Purpose
To evaluate the association between estimated glomerular filtration rate (eGFR) trajectories over time and mortality according to changes in GDMT in HF with reduced or mildly reduced ejection fraction (HFrEF/HFmrEF).
Methods
Patients with HFrEF and HFmrEF from a large registry (2003-2023) with eGFR ≥20ml/min/1.73m2 and at least two consecutive eGFR measurements (one at baseline and one at 6 ± 3 months follow-up) were selected. Changes in renal function were quantified as ΔeGFR (%) = [(eGFRfollow-up - eGFRbaseline)/eGFRbaseline]*100. WRF was defined if ΔeGFR ≤ -20%. WRF was compared with stable/improved eGFR (i.e., non-WRF). Associations between WRF/non-WRF and all-cause death were analysed according to changes in prescription and doses of GDMT over time, i.e. stable/decrease in vs increase in GDMT from baseline to 6 months follow-up, and after adjustment for several potential confounders (i.e., patient characteristics, severity of HF, comorbidities, other medications). GDMT included renin–angiotensin system inhibitors/angiotensin receptor-neprilysin inhibitor (RASI/ARNI) and mineralocorticoid receptor antagonists (MRA) due to their direct effect on renal function. Sodium glucose co-transporter 2 inhibitors were not considered in the current analysis due to limited data availability.
Results
Among 8,912 patients (median age 72 [IQR 63-79] years; 72% male; 76% with HFrEF), 1,370 (15%) developed WRF at 6-month follow-up. Patients with WRF were older and presented with more severe HF (higher New York Hear Association class and higher N-terminal pro-brain natriuretic peptide at baseline). The restricted spline curve showed that ΔeGFR was negatively and nonlinearly associated with the risk of all-cause mortality (Figure 1a). Overall, WRF was associated with a crude 70% higher risk of all-cause death (unadjusted HR 1.70; 95% CI 1.55-1.86, p<0.001) (Figure 1b). After adjustments, WRF was associated with higher risk of all-cause death in patients with stable or decreasing RASI/ARNI use/doses (adjusted HR [adjHR] 1.07; 95%CI 1.02-1.11) but not in those with an increase in use/doses of RASI/ARNI (adjHR 0.98; 95%CI 0.91-1.04) (p-interaction=0.023). WRF was associated with higher mortality irrespective of changes in MRA therapy (adjHR 1.05; 95%CI 1.01-1.09 for stable or decreasing MRA use/doses vs adjHR 1.03; 95%CI 0.92-1.15 for increase in use/doses of MRA; p-interaction= 0.749)(Figure 2).
Conclusion
The association of WRF with all-cause death is attenuated in patients optimizing RASI/ARNI therapy.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.