Hydroxychloroquine early in pregnancy and risk of birth defects: don’t throw out the baby with the bathwater

To the Editors:

We read with interest the article by Huybrechts et al¹ that reported increased major congenital anomalies among first-trimester hydroxychloroquine-exposed pregnancies. The methods used by this excellent research group are sound. Using claims data, they compared hydroxychloroquine-exposed pregnancies with matched unexposed pregnancies and found a 1.26 adjusted relative risk of all major malformations among exposed pregnancies. No particular pattern of anomalies was found. The authors concluded that first-trimester hydroxychloroquine exposure leads to a small increased risk of congenital anomalies. We worry that the findings presented in this paper will be interpreted by practicing clinicians as causally related. This information could tip the balance in clinical decision-making toward anticipated risk rather than the established benefit of hydroxychloroquine during pregnancy.

The authors’ conclusions imply that hydroxychloroquine is potentially teratogenic. Establishment of teratogenicity requires several criteria.² This study fulfills the tenet of identifying exposure to an agent at a critical time in gestation. However, the study does not satisfy other criteria, such as careful delineation of the clinical cases, exposure associated with a specific pattern of defects, and the association making biologic sense. The study’s large sample size provided sufficient statistical power to demonstrate a small but significant increased risk of congenital anomalies with hydroxychloroquine exposure. Although the authors used state-of-the art methods to adjust for confounders, small effect sizes can sometimes be explained by unmeasured confounding. As in all claims data, there is insufficient information to control for exposures to tobacco, alcohol, other drugs, folic acid supplements, and over-the-counter medications that may have impacted the risk of malformations. Most importantly, given the lack of any specific pattern of congenital anomalies, these results could be spurious. Evaluation of increased risks for specific congenital anomalies with hydroxychloroquine exposure in case-control studies could help clarify this issue.

In contrast, several studies have now clearly established the critical role of hydroxychloroquine in controlling systemic lupus erythematosus disease activity in pregnancy and improving outcomes. Currently, several professional societies recommend continuing hydroxychloroquine during lupus pregnancies.³ Moreover, this medication reduces the risk of congenital heart block in offspring of women who are anti-Ro and anti-La positive.⁴

We would caution readers in their interpretation of the results presented. This study has not proven that hydroxychloroquine is teratogenic, whereas data supporting the benefits of hydroxychloroquine during pregnancy for malarial prophylaxis, lupus pregnancy outcome, and prevention of congenital complete heart block are sound.