Amiodarone and survival in heart failure patients with atrial fibrillation: insights from the Icelandic Heart Failure registry

Introduction Heart failure (HF) is frequently accompanied by atrial fibrillation (AF) and often treated with amiodarone despite potential serious adverse effects. However, its efficacy and safety in HFpEF remain poorly characterized. Aim To investigate whether amiodarone treatment is associated with survival in HF patients with AF. Design: We analyzed HF patients with AF from the Icelandic HF Registry, which includes all individuals admitted with HF to the Department of Cardiology in a university hospital, between January 1, 2020, and December 31, 2024. HF was categorized by left ventricular ejection fraction (LVEF) and defined as HFpEF if LVEF ≥50% and LVEF <50% otherwise. HFrEF and HFmrEF were combined for analytical purposes. Survival was assessed using Cox regression adjusted for age and sex and further evaluated using propensity score matching (PSM), inverse probability weighting (IPW), and augmented IPW (AIPW). Results The final study population (n=1158) included 488 patients with HFpEF and 670 with LVEF <50%. In HFpEF, 72 (15%) received amiodarone. Treated patients were younger (77 vs 83 years), had fewer patients in NYHA class III-IV (67 vs 79%) or with dementia (14 vs 25%), but more patients with COPD (29 vs 22%) and CKD stage 3-5 (67 vs 61%). The HF etiology was more commonly ischemic heart disease (IHD, 17 vs 12 %) and AF (36 vs 31%), but less commonly diastolic dysfunction (7 vs 13%). NT-proBNP was lower (2,138 vs 3,497 ng/L). Amiodarone was not associated with survival (adjusted HR 1.00, 95% CI 0.58–1.43) (Figure 1). Results were consistent after PSM (71 matched pairs; HR = 0.92, 95% CI 0.55-1.56), IPW (0.83, 95% CI 0.56-1.24), and AIPW (0.88, 95% CI 0.59-1.33). In patients with LVEF <50%, 144 (21%) received amiodarone. Treated patients were younger (74 vs 80 years), less often female (20 vs 28%), had a lower prevalence of dementia (15 vs 26%) and NYHA class III-IV (70 vs 82%). HF etiology was more often IHD (38 vs 31%) and dilated cardiomyopathy (13 vs 3%), but less commonly AF (20 vs 31%) and hypertension (2 vs 9%). They more often had a dilated left ventricle (47 vs 27%), lower NT-proBNP (3745 vs 5105 ng/L), and lower eGFR (46 vs 54 mL/min/1.73m²). Amiodarone was not associated with survival (adjusted HR 1.03, 95% CI 0.70–1.41) (Figure 2), with similar findings after PSM (131 matched pairs; HR 0.87, 95% CI 0.60–1.28), IPW (HR 1.02, 95% CI 0.76–1.39), and AIPW (HR 1.07, 95% CI 0.79–1.45). Conclusions Amiodarone treatment was not associated with survival in either HFpEF or LVEF <50% despite baseline differences. Although not statistically significant, point estimates in the HFpEF group were consistently below 1.0 across analytical methods, suggesting a possible modest survival benefit that the study may have been underpowered to detect. Lower NYHA class and NT-proBNP levels in amiodarone treated patients may reflect differences in baseline disease severity or symptomatic status.AF and HFpEFAF and LVEF<50%