Potential interactions between erythromycin and protease inhibitors

BACKGROUND: The antibiotic erythromycin causes prolongation of the QT interval and has been associated with rare instances of cardiac arrhythmia (notably torsade de pointes) and sudden cardiac death. Concurrently administered medications that inhibit the hepatic metabolism of erythromycin via the cytochrome P450 3A(CYP3A) system may increase serum levels of erythromycin and thereby increase the risk of cardiac arrhythmia.

SUMMARY: This retrospective evaluation of the association of erythromycin and sudden cardiac death in an outpatient Medicaid population in Tennessee found that individuals who used erythromycin had double the risk of sudden death from cardiac causes compared with control groups (patients not currently taking this antibiotic). Patients taking both erythromycin and a "strong inhibitor" of CYP3A had five times the risk of cardiac death compared to the control groups (patients not on erythromycin, and patients not concurrently on CYP3A inhibitors with antibiotics). The CYP3A inhibitors identified in this study were predominantly the calcium channel blockers diltiazem or verapamil. No patients were taking protease inhibitors. In fact, patients with HIV infection were excluded from the study.

The absolute number of sudden cardiac deaths in patients taking erythromycin and CYP3A inhibitors was small (3 deaths in 194 person-years of follow-up); however, the difference between the incidence-rate ratio in this group (5.35) and the incidence-rate ratios in the groups not taking this combination was statistically significant (p=0.004). (Incidence rate ratios were normalized to the incidence rate in the control group of subjects who took neither antibiotics nor CYP3A inhibitors.) Patients who took erythromycin without CYP3A inhibitors had an incidence-rate ratio of 1.79 for sudden cardiac death, compared with 1.48 for users of amoxicillin alone; these differences were not statistically significant. Thus, the increased risk of sudden death from cardiac causes seen in the overall erythromycin group occurred primarily in those concurrently taking the study-defined CYP3A inhibitors.

While this study raises concerns about the risk of erythromycin when used with CYP3A inhibitors, its conclusions are limited by potential confounding. Nearly all of the patients on the study-defined CYP3A inhibitors were taking diltazem or verapamil rather than other CYP3A inhibitors. Of the deaths that occurred in the group concurrently taking erythromycin and CYP3A inhibitors, all occurred in patients on the diltiazem or verapamil. Both erythromycin and these calcium channel blockers are CYP3A inhibitors, and each may increase the serum levels of the other. Diltiazem and verapamil are themselves known to be potential causes of cardiac arrhythmia and death. This study does not elucidate whether the deaths in this population were due erythromycin, to the calcium channel blockers, or to the combination of the two.

Although this study did not evaluate interactions between protease inhibitors and erythromycin, it is important to question whether the combination of PIs and erythromycin could increase the risk of cardiac arrhythmia and death. It also is reasonable to consider the possible cardiac toxicity of other macrolide antibiotics.

Macrolides may exert cardiotoxic effects through both direct and indirect mechanisms. Macrolides exert a direct effect on cardiac repolarization (as do numerous other medications, including some fluoroquinolones, tricyclic antidepressants, azole antifungals, and antipsychotics), and may thereby increase the risk of arrhythmia.(1-4). Combination of these agents can cause additive toxicity. Some macrolides (notably erythromycin and clarithromycin) also inhibit CYP3A and thus may increase serum levels of other proarrhythmic drugs. PIs are potent inhibitors of CYP3A4 and are known to increase serum levels of erythromycin, as well as of clarithromycin, potentially prolonging cardiac repolarization. Thus, combining a macrolide with a CYP3A inhibitor (such as a PI) or with another medication that prolongs cardiac repolarization may increase the risk of arrhythmia. It appears that azithromycin has far less effect on the QT interval, and is less affected by CYP3A inhibition, than erythromycin or clarithromycin; it appears to convey less risk of arrhythmia. (1, 3-7)

There is little data on the risk of arrhythmia in patients treated with erythromycin or other macrolides in conjunction with PIs. While further studies are needed to define the risk of macrolide use in HIV-infected patients who are treated with protease inhibitors or other potent CYP3A4 inhibitors, caution may be advisable when using macrolides in conjunction with other potentially proarrhythmic agents or with agents (such as PIs) that increase serum macrolide levels.