Dronedarone treatment following cardioversion in patients with atrial fibrillation/flutter: A post hoc analysis of the EURIDIS and ADONIS trials

Abstract Introduction The phase 3 EURIDIS and ADONIS studies evaluated dronedarone for atrial fibrillation (AF)/atrial flutter (AFL) recurrence in patients with nonpermanent AF. Here we assessed whether patient characteristics and/or treatment outcomes in these studies differed based on the need for cardioversion before randomization. Methods Time to adjudicated first AF/AFL recurrence, symptomatic recurrence, cardiovascular hospitalization/death, and AF hospitalization, and safety were assessed by cardioversion status. Results Of 1237 patients randomized (2:1 dronedarone:placebo), 364 required baseline cardioversion (dronedarone 243, placebo 121). Patients requiring cardioversion had a greater prevalence of cardiovascular comorbidities and shorter times to first AF/AFL recurrence compared with those not requiring cardioversion. Dronedarone was associated with longer median time to first AF/AFL recurrence vs placebo regardless of cardioversion status (cardioversion: 50 vs 15 days, hazard ratio [HR] 0.76; 95% confidence interval [CI], 0.59‐0.97; P = .02; non‐cardioversion: 150 vs 77 days, HR 0.76; 95% CI, 0.64‐0.90; P < .01). Dronedarone was similarly associated with prolonged median time to symptomatic recurrence vs placebo in the cardioversion (347 vs 87 days, HR 0.65; 95% CI, 0.49‐0.87) and non‐cardioversion (288 vs 120 days, HR 0.74; 95% CI, 0.62‐0.90) populations. Risk of cardiovascular hospitalization/death and first AF hospitalization was lower with dronedarone vs placebo regardless of cardioversion status, but differences were not statistically significant. The safety of dronedarone was similar in both groups. Conclusion Patients requiring baseline cardioversion represent a distinct population, having more underlying cardiovascular disease and experiencing a shorter time to AF/AFL recurrences. Dronedarone was associated with improved efficacy vs placebo regardless of cardioversion status.


| INTRODUCTION
In clinical practice, antiarrhythmic drugs (AADs) are typically not recommended after the first diagnosis of atrial fibrillation (AF). Rather, the decision to use AADs is made based on AF tolerance (symptoms, hemodynamics), the likelihood of AF recurrence, and/or the need for cardioversion to manage recurrences. Clinical trials evaluating the safety and efficacy of AADs in paroxysmal/persistent AF often include a mix of patients who are in sinus rhythm, require cardioversion before treatment or are assessed for pharmacologic cardioversion. [1][2][3][4][5][6] Baseline characteristics, comorbidities, and disease burden of patients requiring cardioversion for AF may differ from those who do not require cardioversion, as may responses to AAD therapy. Thus, cardioversion status in patients being considered for treatment with AADs is an important consideration in clinical practice; however, little information is available on this subject.
Dronedarone is an AAD indicated to reduce risk of hospitalization for AF in patients in sinus rhythm with a history of paroxysmal/ persistent AF. 7 In the European Trial in AF or Flutter Patients Receiving Dronedarone for the Maintenance of Sinus Rhythm (EURIDIS; NCT00259428) and the American-Australian-African Trial with Dronedarone in AF Patients for the Maintenance of Sinus Rhythm (ADONIS; NCT00259376), dronedarone significantly increased the time to first documented AF/atrial flutter (AFL) recurrence and reduced ventricular rate during recurrence compared with placebo. 8 In these trials, patients in AF/AFL were included if they reverted to sinus rhythm or had successful cardioversion within 5 days before randomization. 8 In this pooled analysis of EURIDIS and ADONIS, we assessed whether baseline characteristics and treatment outcomes differed by baseline cardioversion status.

| Overview of the EURIDIS and ADONIS studies
EURIDIS and ADONIS were concurrent and identically designed doubleblind, randomized, multicenter, phase 3 studies conducted to evaluate the safety and efficacy of dronedarone in maintaining sinus rhythm in individuals with nonpermanent AF/AFL; the eligibility criteria and design of these trials have been described previously. 8 The studies enrolled patients who experienced at least one episode of AF/AFL observed on electrocardiogram (ECG) in the preceding 3 months. 8 All patients were required to have been in sinus rhythm for ≥1 hour to be eligible for randomization. 8 Eligible patients not in sinus rhythm during the 7-day screening period were permitted to participate in the studies if they underwent successful cardioversion (electrical or with ibutilide) within 5 days before randomization and remained in sinus rhythm for ≥1 hour. 8 Concomitant treatment with Vaughan-Williams class I or III AADs was not allowed.
Patients were randomized in a 2:1 ratio to oral dronedarone 400 mg twice daily or placebo for 12 months. 8 A dynamic allocation was performed to balance treatment groups with regard to prognostic factors, study center, need for baseline cardioversion, and chronic treatment with amiodarone before randomization. The primary endpoint of the EURIDIS/ADONIS studies was time to first documented recurrence of AF/AFL within 12 months. Secondary endpoints included symptomatic AF/AFL recurrence and mean ventricular rate during the first recurrence. AF/AFL recurrence was defined as an episode lasting for ≥10 minutes and confirmed by two consecutive recordings taken 10 minutes apart on 12-lead ECG or trans-telephonic ECG monitoring (TTEM). AF/AFL recurrence was evaluated centrally by scheduled TTEM (on days 2, 3, and 5; months 3, 5, 7, and 10; and at symptom recurrence), as well as 12-lead ECG, obtained during study visits on days 7, 14, and 21, and months 2, 4, 6, 9, and 12. 8 Safety was assessed via an adverse event (AE) reporting, vital signs, ECGs, and laboratory evaluations. AEs were categorized according to the Medical Dictionary for Regulatory Activities (versions 6.0 and 6.1) dictionary terms, consistent with regulatory agency guidance. Treatment-emergent AEs (TEAEs) were defined as AEs that occurred or worsened during study treatment or within 10 days following the last drug intake.

| Pooled analysis
We performed a post hoc analysis of the EURIDIS/ADONIS trials by  9 index, which were not assessed in the primary analysis of the EURIDIS/ADONIS studies) are summarized. Sinus rates at baseline were recorded using 12-lead ECG.
The EURIDIS/ADONIS primary endpoint of time to first documented recurrence of AF/AFL within 12 months was evaluated retrospectively in the cardioversion and non-cardioversion groups. The following outcomes were also analyzed: time to symptomatic first AF/AFL recurrence, symptoms at first recurrence, ventricular rates during adjudicated and symptomatic first recurrence, time to first cardiovascular hospitalization or death, and time to first AF hospitalization.

| Statistical analysis
The analysis was performed in all randomized patients who received at least one dose of study drug. Baseline data in the treatment arms of each group are descriptive.
The nonparametric Kaplan-Meier method was used to estimate cumulative incidence functions for time to events in the cardioversion and non-cardioversion groups. Within each cardioversion group, timeto-event endpoints were compared between dronedarone and placebo treatment groups using a 2-sided log-rank asymptotic test. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using a Cox model with study treatment as the only factor. Since the distribution of demographic characteristics was balanced in the groups, THIND ET AL.

| 1023
it was not considered necessary to adjust for baseline covariates in the model. Data were analyzed with SAS version 9.4 (Cary, NC).

| Baseline demographics and patient characteristics
In the EURIDIS and ADONIS studies, a total of 1237 patients were randomized to and received treatment with dronedarone (n = 828) or placebo (n = 409). Of these, 364 (29.4%) patients required cardioversion for study entry: 243 in the dronedarone arm (29.3%) and 121 in the placebo arm (29.6%).
Patients requiring cardioversion tended to have a higher prevalence of structural heart disease (driven by valvular and rheumatic heart disease), congestive heart failure, and greater left atrial diameter compared with patients not requiring cardioversion (

| Efficacy and clinical outcomes
Treatment with dronedarone vs placebo was associated with reduced risk of adjudicated first AF/AFL recurrence by 24% regardless of baseline cardioversion status ( Figure 1A

| Safety
In the primary analysis of the EURIDIS and ADONIS trials, hepatobiliary events were infrequent (<2%) and similar across treatment groups.
Elevations in liver enzymes were similar among patients treated with dronedarone (12.2%) vs placebo (13.6%). 8 Small elevations of serum    Table 3. Deaths due to any cause from first study drug intake up to 10 days after last study drug intake were reported in eight patients treated with dronedarone (two cardioverted and six non-cardioverted) and three patients treated with placebo (all non-cardioverted). Prior or concomitant digoxin use was reported in five of the 11 patients who died: four in patients treated with dronedarone (one cardioverted and three non-cardioverted) and one in a patient treated with placebo (non-cardioverted). Five of the patients who died also had a history of structural heart disease; four of these were treated with dronedarone (all non-cardioverted) and one with placebo (non-cardioverted). Time to symptomatic AF/AFL was longer in the cardioversion group than in the non-cardioversion group among patients treated with dronedarone (347 vs 288 days) but not placebo (87 vs 120 days). The frequency of symptoms at recurrence was lower among patients treated with dronedarone than placebo in both cardioversion groups. The rate-lowering properties of dronedarone in conjunction with the greater use of beta-blockers and digoxin in patients requiring cardioversion may explain these findings, which are also in line with reports that symptomatic AF is more likely to occur in patients with higher ventricular rates. 13 In addition, this observation suggests that the extent of AF-associated symptoms may not be directly associated with overall AF burden (and type of AF, ie, paroxysmal vs persistent). Although perhaps seemingly paradoxical, similar observations have been reported in previous AF studies. [14][15][16] The present analysis identified no new safety concerns with regard to dronedarone treatment in patients requiring cardio-

| CONCLUSIONS
Our analysis with dronedarone sought to better understand the characteristics and outcomes of patients with nonpermanent AF/AFL requiring cardioversion who are candidates for AAD therapy. Important differences were observed between patients who required cardioversion vs those who did not; these included a greater prevalence of underlying cardiovascular disease and structural heart disease among patients requiring cardioversion.
In addition, patients requiring cardioversion experienced substantially shorter times to first AF/AFL recurrence compared with patients not requiring cardioversion. Taken together, these results underscore that patients requiring cardioversion for management of AF/AFL constitute a higher-risk population; understanding the benefits and risks of AAD treatment in this patient population is of considerable clinical interest.
Dronedarone delayed adjudicated and symptomatic AF/AFL recurrence regardless of baseline cardioversion status. The safety profile of dronedarone was also similar in patients with and without baseline cardioversion despite baseline differences in comorbidities. Future prospective studies of dronedarone, as well as other AADs, are warranted to determine the efficacy and safety of AAD therapy post-cardioversion.