Read the Complete EBM Focus and Earn CME.
Reference: Ann Fam Med 2014 Mar-Apr;12(2):121 full-text (level 2 [mid-level] evidence)
Several macrolide antibiotics, particularly erythromycin and clarithromycin, have been shown to cause QT interval prolongation and increase the risk of cardiac arrhythmias such as torsades de pointes (Curr Drug Saf 2010 Jan;5(1):85). Azithromycin was initially thought to have minimal cardiotoxicity, but was later shown to increase the risk of cardiovascular death compared to amoxicillin in a large Medicaid cohort in the United States (N Engl J Med 2012 May 17;366(20):1881 full-text). The U.S. Food and Drug Administration subsequently revised the prescribing information for azithromycin to strengthen the Warnings and Precautions section with information related to the risk of QT interval prolongation and torsades de pointes (FDA Press Release 2013 Mar 12). In addition, QT interval prolongation is considered a class effect of fluoroquinolone antibiotics, and quinolone-related torsades de pointes has been described for sparfloxacin, levofloxacin, and grepafloxacin (J Antimicrob Chemother 2000 May;45(5):557 full-text). A recent large retrospective cohort study of Veterans Affairs medical centers in the United States evaluated the risk of all-cause mortality and serious cardiac arrhythmia in persons receiving levofloxacin, azithromycin, or amoxicillin.
A total of 1,757,689 treatments with levofloxacin, azithromycin, or amoxicillin to United States veterans (mean age 57 years, 89% male) between September 1999 and April 2012 were evaluated. Azithromycin was dispensed mostly for 5 days, whereas amoxicillin and levofloxacin were each dispensed mostly for ≥ 10 days. Compared to other antibiotics, azithromycin was more often prescribed for COPD, and levofloxacin was more often prescribed for genitourinary infection (no p values reported). Crude mortality and serious cardiac arrhythmia rates were assessed, with analyses adjusted for multiple clinical and demographic characteristics at baseline, including indication for antibiotic prescription.
At 5 days, the adjusted mortality per million antibiotics dispensed was 384 with levofloxacin (p < 0.001 vs. amoxicillin), 228 with azithromycin (p < 0.003 vs. amoxicillin), and 154 with amoxicillin. At 10 days, the adjusted mortality per million antibiotics dispensed was 714 with levofloxacin (p < 0.001 vs. amoxicillin), 422 with azithromycin (not significant vs. amoxicillin), and 324 with amoxicillin. Levofloxacin was also associated with an increased risk of serious cardiac arrhythmia vs. amoxicillin at 5 days and 10 days, while azithromycin was associated with an increased risk of serious cardiac arrhythmia vs. amoxicillin at 5 days but not 10 days.
These data show that in a U.S. population, predominantly older men, both levofloxacin and azithromycin were associated with an increased risk of all-cause mortality and serious cardiac arrhythmia during the typical dosing cycle for each drug (10 days for levofloxacin and 5 days for azithromycin). This conclusion is strengthened by the large sample size and the use of actual pharmacy dispensing data, rather than prescriptions. However, it is limited by its observational design, particularly due to imbalanced patient characteristics within the cohort. Although the study used a multivariable model to adjust for a wide variety of baseline characteristics, including antibiotic indication, these differences may still have biased the results. Nonetheless, these findings suggest that physicians should consider prescribing medications other than levofloxacin and azithromycin for older patients when multiple antibiotic choices are available, particularly for those with cardiac comorbidities.
For more information see the Levofloxacin (Systemic) and Azithromycin topics in DynaMed.