Birth defects reporting and the use of dydrogesterone: a disproportionality analysis from the World Health Organization pharmacovigilance database (VigiBase).

Study question: Is there an association between dydrogesterone exposure during early pregnancy and the reporting of birth defects?

Summary answer: This observational analysis based on global safety data showed an increased reporting of birth defects, mainly hypospadias and congenital heart defects (CHD), in pregnancies exposed to dydrogesterone, especially when comparing to progesterone.

What is known already: Intravaginal administration of progesterone is the standard of care to overcome luteal phase progesterone deficiency induced by ovarian stimulation in ART. In recent years, randomized controlled clinical trials demonstrated that oral dydrogesterone was non-inferior for pregnancy rate at 12 weeks of gestation and could be an alternative to micronized vaginal progesterone. Safety profiles in both mother and child were similar. However, concerns have been raised regarding an association between dydrogesterone usage during early pregnancy and CHD in offspring.

Study design size duration: We performed a disproportionality analysis, also called case-non-case study, similar in concept to case-control studies, using the WHO global safety database, VigiBase. The study cohort consisted of individual pregnancy-related safety reports, using the ad hoc standardized query (SMQ 'Pregnancy and neonatal topics'). Cases of birth defects consisted of safety reports containing terms related to the 'congenital, familial and genetic disorders' System Organ Class from the Medical Dictionary for Regulatory Activities. Non-cases consisted of safety reports containing any other adverse event, in pregnancy-related safety reports.

Participants/materials setting methods: Considering reports since database inception to 31 December 2021, we first compared the reporting of birth defects with dydrogesterone to that of any other drug on the database, then to any other drug used for ART. Secondly, we performed a comparison on the reporting of birth defects for dydrogesterone with progesterone. Results are presented as reporting odds ratio (ROR) and their 95% CI. For each comparison, two sensitivity analyses were performed. Finally, a case-by-case review was performed to further characterize major birth defects and sort anomalies according to classification of EUROCAT.

Main results and the role of chance: Study cohort consisted of 362 183 safety reports in pregnant women, among which 50 653 reports were related to the use of drugs for ART, including 145 with dydrogesterone and 1222 with progesterone. Of these, 374 (0.7%) were cases of birth defects: 60 with dydrogesterone and 141 with progesterone, including 48 and 92 cases compatible with major birth defect cases according to EUROCAT classification, respectively. Major birth defects reported with dydrogesterone were mainly genital defects such as hypospadias and CHD. A significantly higher disproportionate reporting of birth defects was found with dydrogesterone when compared to any other drug (ROR 5.4, 95% CI [3.9-7.5]), to any other ART drug (ROR 6.0, 95% CI [4.2-8.5]), and to progesterone (ROR 5.4, 95% CI [3.7-7.9]). Sensitivity analyses found consistent results.

Limitations reasons for caution: First, under-reporting, being inherent to pharmacovigilance systems, impedes the measurement of the incidence of adverse drug reactions and can limit the sensitivity of signal detection. Second, drug causality, not being the same for all cases, is challenging for such events and requires further assessment. However, sensitivity analyses showed consistent results.

Wider implications of the findings: This possible safety signal emphasizes the need for further investigation regarding the fetal safety profile of dydrogesterone.

Study funding/competing interests: No funding was received for this study. None of the authors have any financial and personal relationships with other people or organizations that could influence the design, conductor or reporting of this work.

Trial registration number: N/A.