Introduction
Macrolide antibiotics are commonly used to treat respiratory problems and certain sexually transmitted diseases.1-3 They are generally considered to have a good safety profile, but there have been conflicting findings about their potential to prolong the QT interval in ECG tests of the heart.1- 4 Azithromycin, an analogue of erythromycin, is considered to have a better safety profile than erythromycin and clarithromycin.1,4 It has a slow-acting anti-malarial property; and owing to growing resistance to sulphadoxine-pyrimethamine (SP), its combination with chloroquine has been proposed as an alternative to the current WHO recommended intermittent preventive treatment of malaria in pregnancy with SP.5-7 Currently, the potential use of azithromycin - chloroquine combination in COVID -19 treatment is being tested.8
Macrolides are widely used during pregnancy, but evidence of their teratogenic potential is limited and inconsistent. Animal studies have generally found limited evidence of teratogenicity of macrolides. However, ever since the thalidomide disaster in the 1960s, animal studies are not considered adequate to predict risk in humans.9 Pregnant women are excluded from human trials to determine drug efficacy and safety for ethical reasons.9 Safety information needed by pregnant women and their clinicians to guide risk and benefit analysis must, therefore, come from post-marketing pharmaco-epidemiological studies.
There has been a growing debate about the potential association of macrolides (especially erythromycin) with congenital heart defects (CHD). In 2003, a Swedish Birth Registry study reported that first trimester use of macrolides (mainly erythromycin) was associated with increased risk for CHD (OR 1.79; 95%CI 1.3 - 2.8).10Subsequent updates of this study in 2005 and 2013 found similar results.11,12 A recent UK study has also corroborated this result, reporting an increased risk of CHD (adjusted relative risk 1.62; 95%CI 1.05 - 2.51).13 Other studies have found no significant association between macrolides and CHD14 - 28 , including a 2019 meta-analysis.29 Most studies did not have an adequate sample size and power to investigate specific CHD. Most teratogens are specific to the defects they cause30 and analysing congenital anomalies (CA) in large groups may result in missed associations of the exposure with specific defects.
We investigate here the teratogenic potential of five macrolide antibiotics, as part of a wider study of antibiotics using the large European EUROmediCAT database (http://www.euromedicat.eu/currentresearchanddata/data), derived from registries belonging to the EUROCAT network for Surveillance of Congenital Anomalies (https://eu-rd-platform.jrc.ec.europa.eu/eurocat/eurocat-network_en). We also conduct a meta-analysis of the literature for CHD, including our own results.