Incidence of Atrial Fibrillation in Patients with either Heart Failure or Acute Myocardial Infarction and Left Ventricular Dysfunction: A Cohort Study

BACKGROUND: We examined the incidence of new-onset atrial fibrillation in patients with left ventricular dysfunction. Patients either had a recent myocardial infarction (with or without clinical heart failure) or symptomatic heart failure (without a recent MI). Patients were with and without treatment with the class III antiarrhythmic drug dofetilide over 36 months. METHODS: The Danish Investigations of Arrhythmia and Mortality ON Dofetilide (DIAMOND) studies included 2627 patients without atrial fibrillation at baseline, who were randomised to treatment with either dofetilide or placebo. RESULTS: The competing risk analyses estimated the cumulative incidences of atrial fibrillation during the 42 months of follow-up to be 9.6% in the placebo-treated heart failure-group, and 2.9% in the placebo-treated myocardial infarction-group. Cox proportional hazard regression found a 42% significant reduction in the incidence of new-onset AF when assigned to dofetilide compared to placebo (hazard ratio 0.58, 95% confidence interval 0.40-0.82) and there was no interaction with study (p = 0.89). In the heart failure-group, the incidence of atrial fibrillation was significantly reduced to 5.6% in the dofetilide-treated patients (hazard ratio 0.57, 95% confidence interval 0.38-0.86). In the myocardial infarction-group the incidence of atrial fibrillation was reduced to 1.7% with the administration of dofetilide. This reduction was however not significant (hazard ratio 0.61, 95% confidence interval 0.30-1.24). CONCLUSION: In patients with left ventricular dysfunction the incidence of AF in 42 months was 9.6% in patients with heart failure and 2.9% in patients with a recent MI. Dofetilide significantly reduced the risk of developing atrial fibrillation compared to placebo in the entire study group and in the subgroup of patients with heart failure. The reduction in the subgroup with recent MI was not statistically significant, but the hazard ratio was similar to the hazard ratio for the heart failure patients, and there was no difference between the effect in the two studies (p = 0.89 for interaction).