Senior Research Fellow National Drug and Alcohol Research Centre, University of New South Wales, Sydney, Australia, Australia
Background: The pharmacotherapies varenicline, bupropion and nicotine replacement therapy (NRT) are the most effective strategies for smoking cessation. Prior studies of these medicines in pregnancy have not been adequately powered to generate clear evidence on the risk of congenital malformations.
Objectives: To assess the risk of major congenital malformations (MCM) associated with exposure to varenicline, NRT and bupropion (for smoking cessation) in the first trimester.
Methods: This multi-national cohort study applies a common protocol to high-quality population-based data on all pregnancies that resulted in a birth in two Australian states (New South Wales and Western Australia), New Zealand, Norway, and Sweden between 2005 and 2019. Routinely collected birth notifications or medical birth registrations in each jurisdiction are linked to records of prescription medicine dispensings, congenital malformations, inpatient and outpatient care, and mortality. First trimester exposure is defined as days’ supply of NRT (Anatomical Therapeutic Chemical code N07BA01), varenicline (N07BA03), or bupropion (N06AX12, where indicated for smoking cessation) overlapping with the first 13 weeks of gestation. Any MCM, common body-system and specific malformations are identified from children’s records within 18 months after birth. Children having MCMs due to chromosomal anomalies, teratogenic viral infections and exposure to potentially and known teratogenic medicines are excluded. Propensity score matching with up to 10 matched controls is used to compare outcomes in exposed children to those in children born to mothers who smoked and did not use any smoking cessation pharmacotherapy during pregnancy. Conditional Poisson regression modelling with a robust variance estimator is used to calculate relative risk (RR) and 95% confidence interval in each jurisdiction, followed by a meta-analysis.
Results: Data analysis is underway. Estimated number of births exposed to NRT in the first trimester is 13,830, varenicline 3529 and bupropion 830. These cohort sizes will yield minimum detectable RRs (power 0.80, alpha 0.05) for any MCM (population prevalence 3.1%) of 1.14, 1.27 and 1.59, respectively. The respective minimum detectable RRs are 1.34, 1.68 and 2.51 for cardiac system malformations (prevalence 0.57%); 1.64, 2.34 and 4.08 for nervous system (prevalence 0.17%); and 1.67, 2.83 and 4.20 for orofacial clefts (prevalence 0.16%).
Conclusions: This is the largest investigation of the possible teratogenic effects of NRT, varenicline, and bupropion (for smoking cessation) to date. This evidence is urgently needed to support informed decision-making and optimise outcomes for maternal smokers and their children.
