Acute conversion of persistent atrial fibrillation during dofetilide loading does not predict long-term atrial fibrillation-free survival

Abstract

Purpose: Pharmacologic conversion of atrial fibrillation and flutter (AF/AFl) is common with dofetilide. We determined whether pharmacologic conversion with dofetilide predicts long-term arrhythmia-free survival.

Methods: We retrospectively studied a cohort of 264 consecutive patients who initiated dofetilide during persistent AF/AFl between 2008 and 2013. Patients were excluded if dofetilide was discontinued prior to five doses or electrical cardioversion was performed prior to four doses. Incidence of and characteristics associated with pharmacologic conversion were determined. Patients were followed for long-term AF/AFl recurrence. Predictors of recurrence were identified using multivariate Cox modeling.

Results: Of 205 patients meeting study criteria, 92 (44.9%) converted to sinus rhythm during dofetilide loading. Female gender, history of AFl, greater number of prior catheter ablations, shorter duration of current AF/AFl, and presentation in AFl were all associated with acute pharmacologic conversion (p = 0.001, 0.05, 0.001, 0.003, and 0.003, respectively). In multivariate modeling, longer time since first AF/AFl diagnosis (hazard ratio (HR) = 1.07 per 1-year increase, 95% confidence interval (CI) 1.03-1.10, p < 0.001), longer duration of current AF/AFl episode (HR = 1.01 per 1-month increase, 95% CI 1.00-1.01, p = 0.003) and greater number of failed antiarrhythmic drugs (HR = 1.43 per one drug increase, 95% CI 1.20-1.70, p < 0.001) were independently associated with shorter time to AF/AFl recurrence. Pharmacologic conversion was not significantly associated with time to AF/AFl recurrence (HR = 0.79, 95% CI 0.57-1.10, p = 0.2).

Conclusions: Acute pharmacologic conversion of persistent AF/AFl to sinus rhythm frequently occurs during dofetilide loading. Nevertheless, acute pharmacologic conversion does not predict long-term arrhythmia control, which was moderate at best.