Risk of Sudden Cardiac Death with Azithromycin
Abstract & Commentary
By John P. DiMarco, MD, PhD
Professor of Medicine, Division of Cardiology, University of Virginia, Charlottesville
Source: Ray WA, et al. Azithromycin and the risk of cardiovascular death. N Engl J Med 2012;366:1881-1890.
In this paper, the authors performed a pharmacoepidemiologic study on the relationship between azithromycin, a frequently used broad-spectrum macrolide antibiotic, and cardiovascular death. The authors analyzed data from the Tennessee Medicaid program. This database provides information on medical care encounters and dates and causes of death and is linked to death certificates and hospital discharge data. Data on antibiotic use were taken from Medicaid pharmacy files. The cohort included patients who had been prescribed azithromycin between 1992 and 2006. Patients who had a concomitant life-threatening non-cardiovascular illness, had a diagnosis of drug abuse, resided in a nursing home, or had been hospitalized within the prior 30 days were excluded. Several control groups were analyzed. These included patients who had received no antibiotics and patients who had received courses of therapy with three other commonly used antibiotics: amoxicillin, ciprofloxacin, and levofloxacin. The study outcomes were cardiovascular death, death from any cause, and sudden cardiac death. Deaths during the typical 5-day course of azithromycin therapy and the succeeding 5 days were compared to deaths during the usual 10-day course of the other study antibiotics. Each study comparison was adjusted for an extensive set of covariates using a propensity score that also included a risk for cardiovascular disease.
The study cohort included 348,000 patients with prescriptions for azithromycin who were compared to 1.4 million patients with no antibiotic prescriptions, and to 1.35 million, 265,000, and 194,000 patients with prescriptions for amoxicillin, ciprofloxacin, and levofloxacin, respectively. The mean ages for the groups were between 48 and 51 years of age. Approximately 75% in each group were female. The prevalence of heart failure, chronic obstructive pulmonary disease, diabetes, and measures of disability were roughly similar between groups. The most common indications for use of azithromycin and amoxicillin were respiratory tract infections. The most common indication for ciprofloxacin was genitourinary tract infection. Levofloxacin was used for a variety of infections, including both respiratory and genitourinary tract infections.
The cardiovascular death rate during a 5-day course of treatment for azithromycin was 85.2 per 1 million courses with an estimated 64.6 sudden cardiac deaths per 1 million courses. For patients who did not take antibiotics, there were 29.8 cardiovascular and 24.0 sudden cardiac deaths per 1 million matched 5-day intervals. For amoxicillin, the cardiovascular death rate was 31.5 and 29 sudden deaths per 1 million courses of therapy. The hazard ratio for death during a 5-day course of azithromycin therapy compared to no antibiotic treatment was 2.88. Amoxicillin use was not associated with an increased risk of death. Patients who took ciprofloxacin did not have an increased risk of either cardiovascular death or death from any cause, but there was a nonsignificant trend toward increased risk of cardiovascular death with the use of levofloxacin (hazard ratio 1.50). The absolute excess risk of cardiovascular death among patients who took azithromycin was related to the baseline risk for cardiovascular disease. In patients with the highest risk score, there were an estimated 245 additional cardiovascular deaths per 1 million 5-day courses of azithromycin therapy.
The authors conclude that azithromycin results in a small absolute increase in cardiovascular deaths that is particularly noted in patients with the highest cardiovascular risk.
Azithromycin is a macrolide antibiotic. Two other agents in the same class, erythromycin and clarithromycin, have been shown to block IKr and prolong the QT interval, and have been associated with reports of drug-induced polymorphic ventricular tachycardia (torsades de pointes). Limited animal data have suggested that azithromycin has a much lower proarrhythmic potential.1 The current report, however, suggests that the risk for proarrhythmia with azithromycin, although low, can be quantitated.
Azithromycin was developed and released before the current rigorous protocols for testing new drugs for QT prolongation were standardized. Recently, the FDA revised the azithromycin product label to include a warning about a low risk for proarrhythmia. They did not add a black-box warning since the risk is felt to be low.
Most proarrhythmia associated with antibiotics involve the combination of more than one factor that favor development of the problem. These might possibly include genetic factors, electrolyte imbalance, cardiac status, bradycardia, and other drug therapy. Physicians should be aware of how these factors might interact in a given patient and consider the use of alternative therapy when several risk factors are present.
- Milberg P, et al. Divergent proarrhythmic potential of macrolide antibiotics despite similar QT prolongation: Fast phase 3 repolarization prevents early afterdepolarizations and torsade de pointes. J Pharmacol Exp Ther 2002;303:218-225.