Propylthiouracil Versus Methimazole During Pregnancy: An Evolving Tale of Difficult Choices

Abstract

Antithyroid drugs (ATDs) were introduced for clinical care more than six decades ago and revolutionized the treatment of hyperthyroidism (1). These drugs, however, were introduced well before current rigorous preclinical and clinical testing standards that include assessment of teratogenic potential. Early case reports described aplasia cutis associated with the use of methimazole (MMI) during pregnancy (2), whereas there was a paucity of reports of birth defects in infants born to mothers taking propylthiouracil (PTU). Thus, the notion evolved that MMI (or the parent compound carbimazole) was teratogenic, whereas PTU was safe for use during pregnancy. Reports describing choanal atresia and common facial features in children born to mothers taking MMI in early pregnancy further crystallized the belief that there is a specific methimazole embryopathy (3, 4). Considering that up to 4000 women per year in the United States will have Graves’ disease (GD) (5), determining whether PTU or MMI should be preferentially used during pregnancy became the subject of expert panel reviews. As recently as 2007, an Endocrine Society Clinical Practice Guideline (6) affirmed the notion of potential MMI teratogenicity and recommended that PTU be used exclusively during pregnancy (6). However, when the clarion call was raised in 2009 about potential hepatic toxicity and hepatic failure associated with PTU use (1 in 2000 children and 1 in 10 000 adults) (7, 8), women taking ATDs during pregnancy were faced with a new dilemma. Do I take MMI to spare hepatotoxicity risks to myself and put my child at harm, or do I take PTU to spare my child harm and place myself at risk? Grappling with this issue, in 2009 the American Thyroid Association and the American Association of Clinical Endocrinologists Taskforce on Hyperthyroidism and Other Causes of Thyrotoxicosis proposed that PTU be used during the first trimester and MMI be used over the remainder of pregnancy, a strategy aimed at minimizing risks to both mother and fetus (9). This cautionary recommendation came with the disclaimer that data pertaining to the relative birth defect risks of MMI and PTU were modest and contradictory (9). The panel also called for additional basic research and epidemiological and clinical studies (9). The report by Andersen et al (10) in this issue of the JCEM is the latest in a series of studies on this topic over the past few years. Their compelling data dispute the notion that PTU is not teratogenic and put the relative birth defect risks of PTU and MMI in clinical perspective. Examining a Danish nationwide cohort of 817 093 children, the authors identified 564 infants exposed to PTU during early pregnancy, 1097 infants exposed to MMI (or carbimazole), 159 children exposed to both PTU and MMI, and a large nonexposed control population (10). Thanks to a centralized healthcare reporting system, the authors could assess birth defect risks in a manner that minimized data gaps, reporting bias errors, and patient dropout. The authors discovered a relatively high prevalence of birth defects in the children exposed to ATDs in early pregnancy. Comparable birth defect rates were observed among fetuses exposed to PTU (8.0%), MMI (9.1%), and MMI and PTU 10.1%, vs 5% in nonexposed children (10). PTU was most commonly associated with malformations in the face and neck region, and some PTU-exposed children were diagnosed with cardiovascular system defects including heart septal defects, pulmonary valve stenosis, and pulmonary artery stenosis (10). In com-