Among patients with atrial fibrillation, amiodarone (Cordarone, Pacerone) is the most useful drug for the preservation of sinus rhythm.^1-3 Although amiodarone carries a well-known risk of potentially serious noncardiac toxicities, perhaps less well recognized is its major cardiovascular side effect of bradycardia.

Many patients with atrial fibrillation may be predisposed to bradyarrhythmias because of atrioventricular conduction disturbances or sinoatrial node dysfunction (sick sinus syndrome), either of which can be exacerbated by the electrophysiologic effects of antiarrhythmic drugs. A large retrospective study has shown that there is an association between amiodarone use for atrial fibrillation and permanent pacemaker insertion among patients older than 65 years of age after myocardial infarction.⁴ Whether this association differs by sex has not been previously investigated.

Patients and methods

The current investigation was performed using data from the Fibrillation Registry Assessing Costs, Therapies, Adverse Events, and Lifestyle (FRACTAL). This 17-center, prospective, observational cohort study included 1,005 patients who were enrolled after their first documented episode of atrial fibrillation. Extensive baseline information was collected on standardized case report forms at the time of enrollment, and detailed data regarding atrial fibrillation therapies (both pharmacologic and nonpharmacologic) and clinical events were collected at 3- to 6-month intervals for 36 months.

The association between new pacemaker implantation and amiodarone use was evaluated using multivariable Cox proportional hazards modeling. Amiodarone exposure was used as a time-varying covariate in these models, which were also adjusted for age and the presence of coronary artery disease, congestive heart failure, hypertension, and atrial flutter. We specifically examined potential sex differences in the association between amiodarone use and pacemaker implantation by stratified analysis and by introducing a formal interaction term for sex and amiodarone into the Cox model. The effects of body mass index (BMI), weight, amiodarone dosage, and use of other antiarrhythmic or rate-control drugs on these outcomes were also explored.

Results

Of the initial 1,005 patients in the registry, 32 were excluded from this analysis because a pacemaker or implantable cardioverter defibrillator had been placed prior to enrollment. The remaining 973 patients were followed for a mean of 2.0 ± 0.9 years until death, implantation of a new pacemaker, or end of follow-up. The mean age of these patients at enrollment was 66 ± 14 years, and 40% were women. Many patients had a history of hypertension (49%), congestive heart failure (30%), or coronary artery disease (24%). The initial treatment plan for atrial fibrillation included amiodarone in 5.1% of women and 6.5% of men. The proportion of patients treated with amiodarone at any time during follow-up increased to 19.5% for women and to 26.2% for men. At baseline, 71% of women and 65% of men were also treated with one or more rate-controlling agents.

During follow-up, a total of 60 patients (25 women and 35 men) had permanent pacemakers implanted because of bradyarrhythmias (78%), most commonly sick sinus syndrome. An additional 17 patients underwent pacemaker implantation in conjunction with atrioventricular junction ablation and were excluded from further analysis.

Overall, amiodarone exposure was significantly associated with pacemaker implantation for bradyarrhythmia (hazard ratio [HR], 2.01; 95% confidence interval [CI], 1.08-3.76), after adjustment for age, sex, atrial flutter, coronary artery disease, congestive heart failure, and hypertension. In contrast, no association was found between exposure to sotalol (Betapace, Sorine) or Vaughan-Williams class I antiarrhythmic drugs and pacemaker insertion.

The relationship between amiodarone use and pacemaker implantation was significantly different for women than for men. In a stratified analysis, the HR for pacemaker insertion following amiodarone use was 4.69 (95% CI, 1.99-11.05; P < .001) in women, but only 1.05 (95% CI, 1.99-11.05; P = .92) in men. The interaction term between sex and amiodarone exposure in the Cox model was statistically significant (P = .02) and remained so after adjustment for weight, BMI, weight-adjusted amiodarone dose (amiodarone dose divided by body weight), and use of other antiarrhythmic or rate-control drugs.

Discussion

Our study findings confirm those of a prior report, which documented an association between amiodarone use in atrial fibrillation patients and subsequent pacemaker implantation.⁴ Because this association has been identified for amiodarone, but not for other antiarrhythmic drugs used for atrial fibrillation (eg, sotalol or Vaughan-Williams class I agents), it is almost certainly related to amiodarone’s unique pharmacologic and pharmacodynamic properties. Amiodarone’s complex electrophysiologic effects are mediated by interaction with multiple cardiac ion channels, blockade of cardiac β-adrenergic receptors, and possibly changes in thyroid metabolism. As with other antiarrhythmic drugs, these effects can be associated with clinically significant bradyarrhythmias. Amiodarone appears to be particularly potent at inhibiting sinus node automaticity. Amiodarone’s pharmacodynamic profile—marked by lipophility, an unusually large volume of distribution, and a long serum half-life—also decreases the likelihood that drug-induced bradyarrhythmias can be practically managed with dose adjustment or drug discontinuation. This study extends previous data by finding a marked difference between the sexes in the relationship between amiodarone exposure and pacemaker implantation, with the effect appearing much stronger in women than in men.

The explanation for this finding is not immediately clear, but it may relate to differences in the characteristics of men compared with women at the time atrial fibrillation is diagnosed (eg, mean patient age); to sex-specific differences in cardiac physiology or drug metabolism; or to a combination of these factors. Women tend to be older than men when they first develop atrial fibrillation and have a higher prevalence of hypertension and valvular heart disease, but a lower prevalence of coronary artery disease.^5-7 Women are also known to have longer QT intervals than men. These factors, particularly age, could increase the likelihood of bradyarrhythmias, but they were controlled for in our analysis.

Another potential explanation for a greater tendency toward bradyarrhythmia when using amiodarone in women is body size. Women are generally smaller than men, yet drug dosing is typically the same for both sexes (the usual target maintenance dosage is 200 mg/day), and the drug has a massive volume of distribution. As might be expected, we did find a trend toward a greater need for permanent pacing with higher doses of amiodarone; however, adjusting for weight and BMI did not alter our main findings.

Finally, the bioavailability of amiodarone could be greater in women than in men because of sex-specific differences in drug transporter function, drug elimination, hepatic cytochrome P450 metabolism, protein binding, absorption, or other pharmacokinetic factors.⁷ We are not aware, however, of a definite mechanism that would explain the marked differences in the effects of amiodarone by sex seen in this study. Sex differences in the incidence of adverse effects with other medications have been reported.

Conclusion

We think the findings of our study have practical implications for atrial fibrillation management. When an antiarrhythmic drug is required to control atrial fibrillation-related symptoms, amiodarone remains an important option because of its superior efficacy relative to other agents. Clinicians who prescribe amiodarone need to be aware of its potential to cause significant bradyarrhythmias and should manage this issue proactively by carefully monitoring patients during drug loading, adjusting the dosages of other rate-slowing agents, if necessary, and ensuring appropriate follow-up. Amiodarone-associated bradycardia requiring pacemaker placement appears to be more common in women; thus, extra caution should be used when prescribing this drug to women, particularly the elderly, given that the need for pacing is also age-related. Less aggressive loading regimens and maintenance doses may be appropriate in such cases.

References