Cardiovascular outcomes associated with use of clarithromycin: population based study

Thank you for the response of Dr. Blake. We have read through your comments and we hope the following response will address the issues you raised about our study.

In our paper, we attempted to use three study designs to conduct studies with two cohorts of patients to investigate the association between clarithromycin and cardiovascular outcomes. The first was a propensity score adjusted cohort design comparing exposure to clarithromycin with exposure to amoxicillin/ amoxicillin with clavulanate potassium. The second was a self-controlled case series design comparing the rate of cardiovascular events after exposure to Helicobacter pylori therapy containing-clarithromycin with baseline of follow up. The third study design was a case-crossover study design comparing odds of exposure to H. pylori therapy containing-clarithromycin in 14 days before a cardiovascular event with the odds of exposure during control periods. The latter two designs are both case-only designs using the same cohort, but the designs are considerably different.1-2

Dr. Blake raises an important concern regarding the self-controlled study designs. An indication for testing and treating for H. pylori infection is epigastric pain. If angina were misdiagnosed as epigastric pain due to gastritis, this could create a spurious temporal association between treatment for H. pylori infection and myocardial infarction. While this non-causal mechanism cannot be entirely ruled out the fact that an elevated risk was only seen for the duration of treatment would be unusual for this mechanism. In addition, the elevated short term risk of arrhythmia is congruent with our hypothesis that the risk is due to the clarithromycin prolonging the QT interval. Our study findings also supported the current literature that there was an increased short-term risk of cardiac death including myocardial infarction and arrhythmia associated with clarithromycin.3
Concerning the cardiovascular safety of proton pump inhibitors, a recent meta-analysis suggested that patients receiving both clopidogrel and proton pump inhibitors may be at higher cardiovascular risk but that this did not represent a direct cause of ischaemic events.4 Other studies also showed no evidence of an increased cardiovascular events including myocardial infarction after adjustments among patients using proton pump inhibitors.5-8 Therefore there is some evidence that PPIs are a marker of ill health rather than a direct cause of events such as myocardial infarction. Given the conflicting results in recent studies, it is warranted to conduct future studies to validate the findings of the association between the risk of myocardial infarction and proton pump inhibitors.

We chose amoxicillin/amoxicillin with clavulanate potassium as the comparator as it has not been reported to be associated with cardiovascular events and it has similar indications to clarithromycin. Choosing a comparator with similar indications serves as a proxy for selecting groups of patients with similar baseline characteristics. With additional use of propensity score modelling, comparison was made between two groups of patients with similar characteristics.9 Many other observational studies investigating the association between cardiovascular events and macrolides also chose amoxicillin/ amoxicillin with clavulanate potassium as comparator.10-12 We are aware that doxycycline is also given for patients with community acquired pneumonia in Hong Kong. However, recent literature has explored and seems to support its cardio-protective effect, and so we decided it would not have been an ideal comparator in this context.13-14

With regard to the implications of our findings on current practice, we feel that it is prudent to exercise caution in prescribing clarithromycin to patients at high risk of cardiovascular events and especially in those with a history of cardiac arrhythmia. Our estimate of the risk of cardiovascular events during current use of clarithromycin needs to be balanced against the often large benefits of treatment. Whether the risks identified by our study should be discussed with the patient will ultimately be a clinical decision taken on a case by case basis and will depend on the absolute risk of harm from treatment, the availability of suitable alternatives and the risk of non-treatment. It should be noted that our recommendations are broadly in line with the Physicians' Desk Reference (PDR), British National Formulary and summary of product characteristics of clarithromycin.15-17

References:
1. Whitaker HJ, Farrington CP, Spiessens B, Musonda P. Tutorial in biostatistics: the self-controlled case series method. Stat Med 2006;25: 1768-97.
2. Maclure M, Mittleman MA. Should we use a case-crossover design? Annu Rev Public Health 2000;21: 193-221.
3. Svanstrom H, Pasternak B, Hviid A. Use of clarithromycin and roxithromycin and risk of cardiac death: cohort study. BMJ. 2014;349:g4930.
4. Cardoso RN, Benjo AM, DiNicolantonio JJ, et al. Incidence of cardiovascular events and gastrointestinal bleeding in patients receiving clopidogrel with and without proton pump inhibitors: an updated meta-analysis. Open Heart. 2015;2(1):e000248.
5. Douglas IJ, Evans SJ, Hingorani AD, et al. Clopidogrel and interaction with proton pump inhibitors: comparison between cohort and within person study designs. BMJ. 2012;345:e4388.
6. Blackburn DF, Lamb DA, McLeod MM, Eurich DT. Increased use of acid-suppressing drugs before the occurrence of ischemic events: a potential source of confounding in recent observational studies. Pharmacotherapy. 2010;30(10):985-993.
7. Arana A, Johannes CB, McQuay LJ, Varas-Lorenzo C, Fife D, Rothman KJ. Risk of Out-of-Hospital Sudden Cardiac Death in Users of Domperidone, Proton Pump Inhibitors, or Metoclopramide: A Population-Based Nested Case-Control Study. Drug Saf. 2015;38(12):1187-1199.
8. Turkiewicz A, Vicente RP, Ohlsson H, Tyden P, Merlo J. Revising the link between proton-pump inhibitors and risk of acute myocardial infarction-a case-crossover analysis. Eur J Clin Pharmacol. 2015;71(1):125-129.
9. Stürmer T, Rothman KJ, Avorn J, Glynn RJ. Treatment effects in the presence of unmeasured confounding: dealing with observations in the tails of the propensity score distribution--a simulation study. Am J Epidemiol 2010;172: 843-54.
10. Ray WA, Murray KT, Hall K, Arbogast PG, Stein CM. Azithromycin and the risk of cardiovascular death. N Engl J Med. 2012;366(20):1881-1890.
11. Rao GA, Mann JR, Shoaibi A, et al. Azithromycin and levofloxacin use and increased risk of cardiac arrhythmia and death. Ann Fam Med. 2014;12(2):121-127.
12. Chou HW, Wang JL, Chang CH, Lai CL, Lai MS, Chan KA. Risks of cardiac arrhythmia and mortality among patients using new-generation macrolides, fluoroquinolones, and beta-lactam/beta-lactamase inhibitors: a Taiwanese nationwide study. Clin Infect Dis. 2015;60(4):566-577.
13. Cerisano G, Buonamici P, Valenti R, et al. Early short-term doxycycline therapy in patients with acute myocardial infarction and left ventricular dysfunction to prevent the ominous progression to adverse remodelling: the TIPTOP trial. Eur Heart J. 2014;35(3):184-191.
14. Cerisano G, Buonamici P, Gori AM, et al. Matrix metalloproteinases and their tissue inhibitor after reperfused ST-elevation myocardial infarction treated with doxycycline. Insights from the TIPTOP trial. Int J Cardiol. 2015;197:147-153.
15. PDR Network. Biaxin (clarithromycin) - Full Prescribing Information. 2016; http://www.pdr.net/full-prescribing-information/Biaxin-clarithromycin-6. [Accessed 24 February 2016.]
16. Joint formulary committee, 2016. Macrolides. In: joint formulary committee. British National Formulary. [online]. London: BMJ Group and Pharmaceutical Press. https://www-medicinescomplete-com.eproxy2.lib.hku.hk/mc/bnflegacy/curren... [Accessed 24 February 2016].
17. Clarithromycin 250mg and 500mg Film-Coated Tablets. Summary of Product Characteristics (SPC), Wockhardt UK Ltd. Updated 19 June 2014. https://www.medicines.org.uk/emc/medicine/25086 [Accessed 24 February 2016].