Use of clarithromycin and roxithromycin and risk of cardiac death: cohort study

This article by Svanström, Pasternak and Hviid is an important addition to the growing literature on the association between macrolides and their effect on cardiac mortality (1).

However, we believe their argument could be strengthened by the addition of data on both what the patient presented with and what they departed with (or rather, the mechanism by which they departed). By controlling for each drug on the basis of their prescription, a better comparison could be made regarding the relevant propensity to sustain cardiac complications (ultimately, in some cases, fatal). The paper states that “The indications for clarithromycin, roxithromycin, and penicillin V overlap”; however, this statement does not quantitatively address the question as to whether, for example, clarithromycin is simply more regularly prescribed to sicker patients, whom are more likely to have a rapid decline and demise, potentially from a cardiac origin, despite the antibiotic. Respiratory pathology and decreased lung function are associated with an increased frequency of cardiovascular disease and fatality in cardiac events, behaving in a disease dependent manner (2,3). Considering exact variations in patient presentation is not covered for in the exclusion criteria, controlling for the initial physiological compromise at prescription would be a useful addition to the study. However, we question whether the addition of more such variables to the propensity score would negatively affect the significance of the results: the confidence intervals are already dangerously wide for a study with such a large number of participants – 1.08 to 2.85 being close to ‘statistical insignificance’. Therefore, even greater sample sizes would perhaps be warranted.

Secondly, although the authors do add the ICD-10 disease classifications their results pertain to, they do not stratify according to the specific cardiac cause of mortality. A prolonged QT interval is generally described as the primary proposed mechanism of macrolide-induced cardiac death (from Torsades de pointes) (4). If the data showed that all 18 deaths associated with clarithromycin were arrhythmogenic in origin, it would add strength to the suspicion of a QT-associated origin of the excess mortality. Given that females on average have a longer QT interval, it may explain their apparent increased risk demonstrated in this study (5). Indeed, a genetic predisposition is proposed by the authors (1). It is likely that cardiac mortality arises via an additive effect on the QT, along with other factors, be they genetic or otherwise. While the authors importantly attempt to control for several drugs, it is done without relevance to their relative effect on the QT interval. For example, antipsychotics do not universally prolong the QT interval (e.g. haloperidol does, but not olanzapine), and the ones that do have differing potency – yet all antipsychotics are stratified together (6). Similarly, as another example, amiodarone has also been shown to prolong the QT and is not independently controlled for (7).

Overall, the authors present a comprehensive study. We believe it would, however, benefit from our suggested additions; while this more robust approach may affect the results, a true correlation would be better demonstrated and would add weight to the proposed underlying physiological mechanism.

The authors declare no competing interests.

References
1. Svanstrom H, Pasternak B, Hviid A. Use of clarithromycin and roxithromycin and risk of cardiac death: cohort study. Bmj 2014;349:g4930.
2. Arcari A, Magnacca S, Bracone F, et al. Relation between pulmonary function and 10-year risk for cardiovascular disease among healthy men and women in Italy: the Moli-sani Project. European journal of preventive cardiology 2013;20:862-71.
3. Engstrom G, Hedblad B, Janzon L. Reduced lung function predicts increased fatality in future cardiac events. A population-based study. Journal of internal medicine 2006;260:560-7.
4. Albert RK, Schuller JL, Network CCR. Macrolide antibiotics and the risk of cardiac arrhythmias. American journal of respiratory and critical care medicine 2014;189:1173-80.
5. Stramba-Badiale M, Locati EH, Martinelli A, Courville J, Schwartz PJ. Gender and the relationship between ventricular repolarization and cardiac cycle length during 24-h Holter recordings. European heart journal 1997;18:1000-6.
6. Glassman AH, Bigger JT, Jr. Antipsychotic drugs: prolonged QTc interval, torsade de pointes, and sudden death. The American journal of psychiatry 2001;158:1774-82.
7. Torres V, Tepper D, Flowers D, et al. QT prolongation and the antiarrhythmic efficacy of amiodarone. Journal of the American College of Cardiology 1986;7:142-7.