In a large observational study published in the BMJ, researchers from Great Ormond Street Institute of Child Health found that prescribing a macrolide during the first trimester of pregnancy was associated with a small increased risk of major malformations – particularly heart defects –compared with prescribing a penicillin.
The researchers say their findings suggest that alternative antibiotics should be used whenever feasible, until further research is available.
However, they emphasise that women should not stop taking antibiotics when needed, as untreated infections pose a greater risk to the unborn baby.
The retrospective study included 104,605 children born in the UK from 1990 to 2016 whose mothers were prescribed one macrolide monotherapy (erythromycin, clarithromycin, or azithromycin) or one penicillin monotherapy from the fourth gestational week to delivery.
A further 82,314 children whose mothers were prescribed macrolides or penicillins before pregnancy, and 53,735 children who were siblings of children in the study group, acted as negative control cohorts.
The researchers excluded children with known chromosomal abnormalities and those whose mothers were prescribed teratogenic drugs during pregnancy. Children were followed for a median of 5.8 years after birth.
Major malformations were recorded in 186 of 8632 children whose mothers were prescribed macrolides at any point during pregnancy and 1666 of 95,973 children whose mothers were prescribed penicillins during pregnancy.
Macrolide prescribing during the first trimester was associated with an increased risk of any major malformation compared with penicillin (27.65 vs 17.65 per 1000, adjusted risk ratio [ARR] 1.55, 95% CI 1.19–2.03) and of cardiovascular malformations in particular (10.60 vs 6.61 per 1000, ARR 1.62, 95% CI 1.05–2.51). The increased risks were not observed in children of mothers prescribed macrolides in the second or third trimester.
Macrolide prescribing in any trimester was also associated with an increased risk of genital malformations, mainly hypospadias (4.75 vs 3.07 per 1000, ARR 1.58, 95% CI 1.14–2.19).
Erythromycin in the first trimester was associated with an increased risk of any major malformation (27.39 vs 17.65 per 1000, ARR 1.50, 95% CI 1.13–1.99).
There were no significant associations between macrolide prescribing and other malformations or neurodevelopmental disorders (cerebral palsy, epilepsy, ADHD, or autism spectrum disorder).
If the associations the researchers identified are causal, they estimate that an additional 4 children with cardiovascular malformations would occur for every 1000 children exposed to macrolides instead of penicillins in the first trimester.
The researchers acknowledge a retrospective study cannot establish causality and highlight other limitations of their work, such its limited power to examine treatment exposure during known critical periods for specific malformations and neurodevelopmental disorders. However, they say their results were largely unchanged after further analyses, suggesting that the findings withstand scrutiny.
Dr Sarah Stock, Senior Clinical Lecturer Maternal and Fetal Medicine at the University of Edinburgh Usher Institute described the study as 'high-quality' and called the finding of a small increased risk of some birth defects 'pretty convincing'.
'Using an alternative antibiotic wherever possible in early pregnancy seems sensible,' said Dr Stock. 'However, if macrolides are the only treatment option, women can be reassured that the absolute risk of a problem is low. The highest risk was seen with early pregnancy prescriptions, where the additional risk of a heart defect associated with macrolide use was less than half a percent.'
Professor Stephen Evans, Professor of Pharmacoepidemiology at the London School of Hygiene & Tropical Medicine, also judged the paper to be 'well-conducted and analysed' but cautioned against over-interpreting the findings.
Professor Evans said: 'Even if the evidence for harm of macrolides in general is not as strong as these authors suggest, there is no real evidence of absence of harm. Caution is certainly reasonable. Prescribing for respiratory infections, where the benefits may be largely absent if the primary infection is viral, should be a particular area for caution.'
The MHRA says it is reviewing the findings of the study in the context of similar studies that have not found the same association. It advises that macrolides should continue to be used in pregnancy 'where there is a clinical need'.