Pregnancy outcome following gestational exposure to TNF-alpha-inhibitors: A prospective, comparative, observational study

Highlights

• Pregnancy use of anti-TNF-α medications was not associated with a teratogenic risk in 65 live-born infants.

• There were no cases of VATER/VACTERL association.

Abstract

Objective

To evaluate pregnancy safety of anti-TNF-α medications.

Design

Prospective, comparative, observational study done at the Israeli Teratology Information Service between 2002 and 2011.

Results

83 anti-TNF-α-exposed-pregnancies (97.6% in the first trimester, T1) were followed-up and compared with 86 disease-matched (DM) and 341 non-teratogenic-exposed (NTE) pregnancies. The anti-TNF-α group consisted of 35 infliximab-, 25 etanercept-, and 23 adalimumab-exposed pregnancies. The rate of major congenital anomalies did not significantly differ between the three groups [3/65 (4.6%) (anti-TNF-α, T1), 5/79 (6.3%) (DM), 8/336 (2.4%) (NTE)], even after excluding genetic or cytogenetic anomalies [3/65 (4.6%) (anti-TNF-α, T1), 4/79 (5.1%) (DM), 6/336 (1.8%) (NTE)]. There were no cases of VATER/VACTERL association.

Conclusion

The present study suggests that anti-TNF-α treatment does not pose a major teratogenic risk in humans. This conclusion is based on relatively small numbers of exposed pregnancies and should be interpreted with caution. Larger studies are needed to establish anti-TNF-α pregnancy safety.

Introduction

Infliximab, etanercept, and adalimumab are tumor necrosis factor-α inhibitors (anti-TNF-α) produced by genetic engineering. Infliximab is a mouse-human chimeric monoclonal IgG1 antibody, etanercept is a TNF receptor-IgG1 fusion human protein, and adalimumab is a fully human monoclonal antibody. Anti-TNF-α medications have led to a vast improvement in the treatment of inflammatory conditions. These new biopharmaceuticals target inflammatory cytokines, and are used in the treatment of autoimmune diseases including rheumatoid arthritis, psoriatic arthritis, juvenile idiopathic arthritis, ankylosing spondylitis, and inflammatory bowel diseases. Many of these illnesses are more common in women and occur during reproductive age. When anti mouse TNF-α monoclonal antibody was administered to pregnant mice, there were no teratogenic effects [1]. The three anti-TNF-α antibodies, infliximab, etanercept, and adalimumab, have a large molecular weight of approximately 149 kDa, and are, therefore, not expected to cross the human placenta in significant concentration in the first trimester of pregnancy. It is hypothesized, however, that the Fab’ fragment may passively cross the placenta in low levels in the first trimester during the period of organogenesis. Placental transfer of infliximab, etanercept, and adalimumab close to term has been documented [2], [3], [4], [5]. The terminal elimination half-lives of these molecules are relatively long: 4 days for etanercept [6], 19 days for infliximab [7], and 15–19 days for adalimumab [8]. Human pregnancy experience with these medications is rather limited. It is based mostly on case reports, small case series, regulatory agency records, manufacturer's registries, and preliminary data presented at scientific meetings published as abstracts. Human pregnancy data on infliximab specifically [9] and on anti-TNF-α medications in general [10] has been summarized in two recent review articles.

Carter et al. [11] published a case report describing VATER association in a neonate born to a mother who was treated with etanercept during pregnancy. They reviewed the TNF-α-antagonists use in pregnancy report to the US Food and Drug Administration. They reported that 24 of the 41 children with congenital anomalies born to mothers who took infliximab or etanercept during pregnancy had one or more anomalies that are part of the VACTERL association [12]. They concluded that maternal treatment with TNF antagonists pose an increased risk for VATER or VACTERL association. A critical editorial [13] and a letter to the editor [14] published in the same journal raised questions and pointed out numerous methodological problems in this study. The analysis suffers from selection bias, is based on voluntary reports lacking a denominator of total number or pregnancies with maternal anti-TNF-α treatment and, therefore, cannot be used for risk calculation. Sporadic reports of isolated anomalies that could be a part of the VACTERL association do not establish this diagnosis. The diagnosis requires a constellation of anomalies in the same patient. The study did not address potential confounders. In addition, placental transmission of these molecules during the first trimester of pregnancy, covering the period of organ formation, is not expected in significant amounts, and makes the suggested causation between VACTERL association and anti-TNF-α medications, unlikely.

Contrary to the reports by Carter et al. [11], [12], human pregnancy experience with TNF-α antagonists has been mostly reassuring [5], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. A summary of the reports of pregnancies with exposure to anti-TNF-α medications is presented in Table 1.

The primary objective of the present study was to prospectively evaluate the rate of major congenital anomalies after maternal treatment with TNF-α-antagonists (infliximab, etanercept, or adalimumab) during pregnancy compared with two comparison groups: (1) pregnant women with autoimmune diseases not treated with anti-TNF-α medications during pregnancy (disease matched), (2) pregnant women counseled for non-teratogenic exposure. Secondary endpoints of interest were pregnancy outcome, birth weight, and gestational age at delivery.