Macrolides Can Kill

No Comments

altDid you know that the macrolide group of antibiotics (erythromycin and all its cousins) could kill you? Drop dead – sudden death. Well there is a growing body of literature saying just that.

There is a growing body of literature suggesting that the macrolide group of antibiotics (erythromycin and all its cousins) could cause patients to drop dead.



Did you know that the macrolide group of antibiotics (erythromycin and all its cousins) could kill you? Drop dead – sudden death. Well there is a growing body of literature saying just that. Here’s a 2004 paper that addresses the issues and found that:

Compared to the use of amoxicillin, there was a two fold increase in the risk of sudden death with oral erythromycin


When patients simultaneously took drugs that were potent inhibitors of the CYP450 3A4 pathway, the risk was five fold (and there is a huge list of other drugs that compete for this pathway).

Two important drugs whose levels increase in the setting of the macrolides are the calcium channel blockers verapamil and diltiazem (more on this later)
Although not specifically addressed in the study, physicians would be wise to use the macrolides cautiously in the subset of patients likely to have an increased risk of sudden death – in particular, patients with congestive failure given that serious arrhythmias are common in these patients

Ray, W.A., et al, N Engl J Med 351(11):1089, September 9, 2004

METHODS: The authors, from Vanderbilt University, examined data from the Tennessee Medicaid files (patients aged 15-84) for the period 1988 through 1993 to explore relationships between current and former use of erythromycin and sudden cardiac death. They also evaluated sudden cardiac death in patients taking amoxicillin, as well as the possible effect of concomitant use of erythromycin and selected medications that are potent inhibitors of CYP3A that might interfere with the metabolism of erythromycin (azole antifungal drugs, diltiazem, verapamil and troleandomycin).


RESULTS: Information was available for 1,249,943 person-years, including 5,305 person-years of current erythromycin use (10 sudden deaths), 111,779 person-years of past use (100 sudden deaths), and 6,846 person-years of current use of amoxicillin (8 sudden deaths). The study subjects had a mean age of 45 and were most often female (70%) and Caucasian (58%). When compared with persons who had never used erythromycin, the incidence-rate ratio for sudden cardiac death was increased in current users (2.01), but not in former users (0.89) or current users of amoxicillin (1.18). The risk was particularly high among current users of erythromycin together with one of the CYP3A inhibitors (incidence-rate ratio 5.35) (3 sudden deaths in 194 person-years). Previous use of these latter medications, or current use concomitantly with amoxicillin did not significantly influence the risk of sudden cardiac death.

CONCLUSIONS: These population-based findings support the implications of case reports suggesting an increased risk of sudden cardiac death in patients taking erythromycin, and particularly among those also taking one of the selected inhibitors of CYP3A. 45 references 1/05 – #28

To help put the magnitude of these interactions involving drugs using the P450 pathway into some focus, a study of multiple drug interactions (DRUG-DRUG INTERACTIONS AMONG ELDERLY PATIENTS HOSPITALIZED FOR DRUG TOXICITY Juurlink, D.N., et al, JAMA 289(13):1652, April 2, 2003) found that the odds ratio for an admission for digoxin toxicity was 11.7 within one week of starting clarithromycin and 8.5 within three weeks compared to controls. So these interactions aren’t particularly subtle. Two other studies citing odds ratios of 4.4 and 14.8 for clarithromycin and 3.7 for azithromycin for digoxin toxicity admissions.

Here’s a more recent study confirming that the macrolides can kill you. Operating on the theory that the chlamydial DNA repeatedly found in coronary plaques had something to do with the genesis of these plaques, a study was undertaken to determine if two weeks of clarithromycin would eradicate the Chlamydia and coronary outcomes would be better. 2,172 patients with stable coronary disease were randomized to clarithromycin or not for two weeks and followed for a mean of 2.6 years. The hazard ratio for sudden death increased 2.6 in the clarithyromycin group and the effect was noted even on follow-up at 6 years. The theory – clarithromycin sets up some sort of inflammatory process. Interestingly, patients on statins (know to have anti-inflammatory effects) had no increase in sudden death.
Winkel, P., et al, Cardiology 118:63, 2011

OBJECTIVES: To elucidate potential mechanisms for the clarithromycin-induced excess mortality observed in the CLARICOR trial during 2.6 year follow-up of patients with stable coronary artery disease. Methods: Cox analyses using out-of-hospital death as a proxy for sudden death compared to in-hospital (nonsudden) death.

RESULT: In 100 of 189 (53%) cardiovascular (CV) deaths in which it was possible to examine the question, there was a strong association between place of death and the classification of CV death as sudden or not-sudden. The excess mortality in the clarithromycin group was confined to sudden CV death in patients not on statins at trial entry (HR: 2.61, 95% CI: 1.69–4.05, p < 0.0005). Other categories of deaths showed no marked drug-placebo difference.

CONCLUSIONS: Short-term clarithromycin administration was significantly associated with increased risk of sudden CV death in stable coronary heart disease patients not using statins.

The Italians seem to be up in arms about the increased risks of ventricular arrhythmias and sudden death due to antibiotics. Using three large databases, after adjusting for numerous variables, they concluded that the odds ratio for ventricular arrhythmias or sudden death were 2.13 for recent macrolide use and 3.58 for recent fluoroquinolone use (more on these drugs next time).

OK, enough about sudden death with the macrolides. Anything else of which we should be aware? Remember when it was noted that the calcium channel blockers also competed for the same metabolic pathway as the macrolides? What could be anticipated if blood levels of the calcium channel blockers increased as a result of this competition? Just as you would expect – more admissions to the hospital for hypotension and shock. Of 7,100 patients admitted for hypotension or shock who were on CCBs, the odds ratio for admission was 5.8 when on erythromycin, 3.7 when on clarithromycin and 1.5 (NS) when on azithromycin.

Wright, A.J., et al, Can Med Assoc J 183(3):303, February 22, 2011

BACKGROUND: In 2008, 66 million prescriptions were written for macrolide antibiotics in the USA and nearly 90 million prescriptions were written for calcium channel blockers (CCBs). The metabolism of CCBs is dependent upon the cytochrome P450 3A4 enzyme, which is inhibited by clarithromycin and erythromycin but not by azithromycin.

METHODS: This case-crossover study, from Sunnybrook Health Sciences Center in Ontario, examined the relationship between use of a macrolide antibiotic and hospital admission for hypotension in persons aged 66 or older who were treated with a single CCB (verapamil, diltiazem, nifedipine, amlodipine or felodipine) in 1994-2009. Prescription of a macrolide antibiotic was compared during the seven days prior to hospital admission for hypotension or shock and during a seven-day control period in the same patient one month prior to the hospitalization.

RESULTS: Of 7100 patients admitted for hypotension or shock, prescription of a macrolide was documented for 2.5% (176). When the use of macrolides during the week prior to the hospitalization was compared with the one-week control period, odds ratios for hospital admission for hypotension were 5.8 (95% CI 2.3-15.0) with use of erythromycin and 3.7 (95% CI 2.3-6.1) with use of clarithromycin (p<0.001 for both), but 1.5 (95% CI 0.8-2.8) with use of azithromycin (p=NS). Similar findings were observed in an analysis of patients treated with a dihydropyridine CCB (nifedipine, amlodipine or felodipine).

CONCLUSIONS: These findings confirm the underappreciated potential for a drug-drug interaction in patients treated with a calcium channel blocker and a macrolide antibiotic. The authors suggest the use of azithromycin when a macrolide is indicated in these patients. 20 references ( – no reprints) 9/11 – #24

The drug-drug reactions noted here are not flukes or rare events but are predictable consequences of the pharmacology of these drugs and, as such,, can be expected to routinely occur when these drugs are given concomitantly. Although these interactions are likely common, it is unlikely that they will necessarily result in clinically apparent signs and symptoms and, as such, these drug-drug interaction are generally underappreciated. Clinicians need to be aware of these more atypical reactions so as to limit patient risk when optimize care.

A review at notes 110 major drug reactions involving the macrolides (267 moderate ones and 104 minor ones). Most of these reactions center around the competition for catabolism via the CYP450 3A4 pathway resulting in increased blood levels of the involved drugs. Check out for the full list and an explanation of each interaction.

Richard Bukata, MD Editor of Emergency Medical Abstracts (

Leave A Reply