Inflammatory bowel disease (IBD) commonly develops during young age and this period belongs to the reproductive age [1-3]. Especially, women IBD patients would experience pregnancy and delivery along with the disease. In addition, as IBD is a chronic disease and patients should receive long-term medical therapies, IBD women would also have to continue medication even during their pregnancies.

In previous studies, pregnant IBD women’s disease activity seemed to affect the pregnancy-related outcomes such as preterm delivery, miscarriage and normal delivery rate [4-8]. For example, if pregnant IBD women maintained the remission state of the disease, perinatal complications and normal delivery rate were similar with healthy subjects [4-7,9,10]. However, if pregnant IBD women’s clinical status were active, the rate of spontaneous abortion and preterm delivery was higher than healthy subjects and their disease itself also seemed to worsen [4-8,11]. Therefore, the continuation of maintenance therapy seems necessary for reproductive-aged women to maintain remission state, and also, additional medical treatment seems inevitable when the disease activity becomes worse [8,12-17]. As a result, recently updated treatment protocols have recommended the recognition of the influence of medication to pregnant IBD women and the use of safe medication with appropriate duration to maintain the remission state [8].

Recently, anti-TNF-α agents have shown effectiveness in cases of refractory or severe IBD and its use is rapidly increasing [18,19]. These drugs consist of IgG1 monoclonal antibody (Ab) for TNF-α, and infliximab and adalimumab are the most commonly used. The Food and Drug Administration (FDA) divides drugs into categories A, B, C, D, and X in pregnancy. U.S. and European guidelines classifies infliximab and adalimumab as category B (no evidence of risk in humans) [8,20]. However, Australian Drug Evaluation Committee (ADEC) pregnancy category is different from FDA categories. Australia classifies them as category C (Drugs that have caused, or may be suspected of causing, harmful effects on the human fetus or neonate without causing malformations. These effects may be reversible) [21]. Anti-TNF-α agents may cross the placenta after 20 weeks gestational age, and can remain in the blood of infants until 3 to 6 months and in maximum up to 1 year after the last administration [22-24]. In the United States and Europe, the 2011 London Position Statement on Biological Therapy for IBD describes that the risk of pregnancy and breastfeeding of mother and fetus for anti-TNF-α therapy is due to active IBD, but not medication, and that the use in the first 2 trimesters is low risk [8]. Many studies on pregnant IBD women treated with anti-TNF-α did not show significant increase of perinatal complication. However, in the 2 largest safety studies on the infliximab and pregnancy in the United States, the TREAT registry and the Infliximab Safety Database, some fetal malformations and perinatal complications such as preterm births and stillbirth have been reported [5,25]. In addition, its influence on their children’s future growth also remain inconclusive [25-33].

Therefore, this study aimed to evaluate the pregnancy-related outcomes in IBD women who had received anti-TNF-α therapy during pregnancy, and the immunity of their children such as serologic responses to vaccine and serious infection rate.

Among the 18 female IBD patients who had delivered their child while receiving anti-TNF-α treatments, all subjects-maintained remission state without any significant complication including perinatal complications. Their average delivery age was 31.9 ± 3.3 years old, which was similar with Korean women population (32.6 years old). However, the average birth weight of infants born from IBD women was 2.95 ± 0.30 kg, which was lower than 3.2 kg of the Korean newborn population [34]. The method of delivery was vaginal delivery in 4 patients and cesarean section in 12 patients (75%), which was higher than Korean women population (42.6%) [34].

Past studies have shown conflicting results on safety of anti-TNF-α therapy to IBD women during pregnancy. The PIANO study performed in the United States had shown that the use of anti-TNF-α agents was not associated with any increase of the complications even when adjusted for disease type or disease activity of IBD [32]. Similarly, data from the TREAT registry also have shown that the rates of miscarriage and neonatal complications were not significantly different between infliximab-treated and infliximab-naive IBD patients [25]. In contrast, U.S. FDA database had demonstrated 61 anomalies, such as vertebral and cardiac anomalies in IBD mother’s children, who had been exposed to anti-TNF-α in utero [29]. There were several studies which demonstrated that disease activity when patients become pregnant affect the pregnancy-related outcomes and normal delivery rate [4-7]. In this study, all subjects maintained remission state during pregnancy continuing the medical treatment, and showed no perinatal complication. Therefore, even though the disease duration is long, pregnancy outcome would not be affected by the disease if patients could maintain the remission state during pregnancy.

No significant evidences of developmental problems were found in the children. Their average percentile weight when they had pediatric examination was 42.6% ± 28.5% which was similar to average Korean age-matched children. Approximately 16.7% of the children’s weight were under the 10 percentile range, which was slightly higher than the average, but the development of all children were within the normal range.

The positive rate of anti-HBs titer from IBD mother’s children who had taken regular hepatitis B vaccination was 66.7%, which was similar to 62.7%–78.6% resulted in general Korean age-matched children. Seroconversion rate after 1 booster vaccination was 100%, and this was also not lower than the general population, which was 88.8% [35]. Therefore, immunological memory for hepatitis B vaccination of these children seemed intact. Recently, the PIANO study have reported that adequate response to both tetanus and Haemophilus influenzae type b (Hib) vaccines were found in 92% infants [36], and another study performed in year 2017 had demonstrated that the Ab titer for Hib and tetanus toxin in children over 7 months old did not show significant difference between the biologic agent exposed group and the unexposed group during pregnancy [37]. However, there was a significant increase in infant infections from 9 to 11 months of age in the combination therapy group relative to the unexposed group [32]. In this study, 3 children whose anti-HBs were negative were admitted due to infections before 12 months of age, but they were discharged without any specific complication. It is reported that the drugs are detected in the blood of the children until 6 months after delivery, up to maximum of 12 months, when anti-TNF-α has been used during pregnancy. During that period, the possibility of infection may be high, and it can be assumed that antibodies were not produced sufficiently in children with weak immunity. However, they were well recovered without significant complications, and immunologic memory was intact after hepatitis B booster vaccination. Therefore, the correlation between anti-TNF-α during pregnancy and immunity in children is not clear. Since the number of children in this study was small, further research on more subjects is necessary to estimate their correlation.

Previous studies have demonstrated that if patients had undergone anti-TNF-α therapy during pregnancy, the monoclonal Ab may cross the placenta and be delivered to the fetus, and it can remain until 6 months, up to 12 months after birth [23,26-28,33]. However, there are also growing evidences demonstrating that following the anti-TNF-α medication even until the third trimester does not affect the fetus significantly [8]. In this study, there was no significant difference between the patients who had their last anti-TNF-α medication at the second trimester and the third trimester regarding their pregnancy outcomes and immunity of the children. Therefore, it would be more beneficial to continue anti-TNF-α treatment until 32 weeks of gestational age in the third trimester to prevent aggravation of disease activity rather than choosing the early discontinuation.

In this study, 7 children have had live vaccination, BCG and/or rotavirus, before 6 months old, but there were no significant side effects. Current guidelines suggest that children born from IBD patients who had been treated with anti-TNF-α medication during pregnancy can take inactive vaccination according to the regular vaccination schedule, because there are no evidences that intra-uterine exposure to TNF-α monoclonal Ab caused inactivated vaccine related side effects. However, it has been suggested that live vaccination should be performed after 6 months old or when the drug which had been treated to mother is not detected from the children’s blood [8]. However, studies on whether this is sufficient for disease prevention or its consequence on immunity acquisition with these live vaccination schedule are rare [17,31,38]. One study showed that there was no significant perinatal complication in 15 children from patients who received anti-TNF-α therapy during pregnancy who had BCG vaccination before 6 months old in the previous study [39]. Recent data show that the IgG, IgA levels were adequate, but IgM level was low in 50% infants who did not show higher infection rate [36]. However, one study showed that of the 40 children born from mothers who had biologic therapy, 7 of them had mild reaction after rotavirus vaccination [37]. In addition, there was a case report of a child who had disseminated BCG infection after BCG vaccination at 3 months old [40]. Therefore, further additional tracking research on a large number of children’s safety of live vaccination and other blood test indexes that can determine immunity is necessary.

This study has several limitations. First, the number of subjects was relatively small, and second, all of subjects had CD and there was no UC patient. Third, there was no control group who had taken the therapy only using the nonbiologic agent such as 5-aminosalicylic acid during pregnancy. To overcome these limitations, further prospective study subjecting a bigger group of reproductive-aged IBD woman who might keep their follow-ups during the pregnancy and also subjecting their children should be necessary.

In summary, this was the first study on Korean IBD women who gave birth after anti-TNF-α therapy during pregnancy even including their children’s immunity. IBD women who had anti-TNF-α during pregnancy did not show significant perinatal complications when they maintained remission state even in immunosuppressive combination therapy. From the clinical history about vaccination and anti-HBs titers, immunity of children born from IBD women who had been treated with anti-TNF-α medication during pregnancy seemed to be intact.