Efficacy and safety of guideline-directed medical therapies according to background use of digoxin: Insights from RALES, PARADIGM-HF, and DAPA-HF
R Helseth, R O Ono, B L C Claggett, Y H Hamatani, M R Ruppert, A S D Desai, P H Jhund, B P Pitt, F Z Zannad, M P Packer, J J V M Mcmurray, K D Docherty, S S Solomon, M V VaduganathanAbstract
Background
Cardiac glycosides have a limited role in the current guidelines for heart failure with reduced ejection fraction (HFrEF), which may in part be due a lack of randomized evidence in the context of contemporary guideline-directed medical therapy (GDMT). The DIGIT-HF trial recently demonstrated benefit with digitoxin in well-treated patients with HFrEF, renewing the attention to cardiac glycosides as adjunctive therapy.
Purpose
We assessed the efficacy and safety of established HFrEF therapies according to background use of digoxin in pivotal randomized trials.
Methods
We leveraged data from RALES (spironolactone vs placebo; n=1,663), PARADIGM-HF (sacubitril/valsartan vs enalapril; n=8,399), and DAPA-HF (dapagliflozin vs placebo); n=4,744). Participants were categorized according to digoxin use at randomization, and treatment efficacy was evaluated across background digoxin therapy. Study endpoints were harmonized to a composite of cardiovascular (CV) death or HF hospitalization and all-cause mortality. For safety evaluation, rates of treatment discontinuation due to adverse events (AE) were determined. Time-to-event outcomes was analyzed using Cox proportional hazards models, applying stratification and fixed-effect factors consistent with the original protocol for each trial. Potential treatment heterogeneity by baseline digoxin use was tested using interactional analyses.
Results
Baseline digoxin use was reported in 1,216 of 1,663 patients (73%) in RALES, 2,539 of 8,399 (30%) in PARADIGM-HF, and 887 of 4,744 (19%) in DAPA-HF, corresponding to 4,642 (31%) across the 3 trials (n=14,806 participants). Patients using digoxin were younger (63±12 vs 66±11 years), had a higher resting heart rate (77±14 vs. 72 ±12 bpm), and were more likely to have a history of atrial fibrillation (55% vs. 31%); P values <0.01. Women were similarly represented in patients treated and not treated with digoxin (24% vs. 23%); P=0.13. Digoxin users had higher NYHA class (NYHA III/IV 49% vs. 29%), higher NT-proBNP levels (1,840 [1,029, 3,695] vs. 1,458 [836, 2,785] pg/mL), and lower left ventricular ejection fraction (29 [23, 33] vs. 31 [25, 35] %); P values <0.01. Digoxin users were less frequently treated with β-blockers (70% vs. 91%) and more often with diuretics (90% vs. 84%); P values <0.01. Spironolactone, sacubitril-valsartan, and dapagliflozin each reduced the risk of the composite endpoint of CV death or HF hospitalization and all-cause mortality, irrespective of background digoxin use; all Pinteraction ≥0.16 (Figure 1). Rates of treatment discontinuation due to AEs were not modified by background digoxin use; all Pinteraction ≥0.37.
Conclusion
These findings support the safe and effective use of contemporary GDMT alongside digoxin in patients with HFrEF.Efficacy of GDMT across digoxin useFor image description, please refer to the figure legend and surrounding text.