Apremilast
Apremilast (APR) is a drug from the group of phosphodiesterase inhibitors and is approved in Austria for the treatment of moderate to severe plaque psoriasis (PP) and psoriatic arthritis (PsA). Its anti-inflammatory effects are based on the inhibition of the intracellular enzyme phosphodiesterase-4, leading to an increase in cyclic adenosine monophosphate. The result is a decreased production of pro-inflammatory cytokines, including tumor necrosis factor alpha (TNF-α), interleukin (IL) 17 and IL 23. Animal studies have not indicated an increase in deformities under APR but a dose-dependent increase in miscarriages and reduced birth weight [
14]. No published human data are available; however, a pregnancy registry has been established. In view of the highly limited data, APR therapy should be discontinued at least 2 days before conception and APR has been detected in the milk of lactating mice [
14].
Azathioprine/6-mercaptopurine
As azathioprine (AZA) passes the placenta, metabolites, especially the pharmacologically active 6‑thioguanine nucleotide (6-TGN) but barely 6‑methyl mercaptopurine (6-MMP), have been detected in umbilical cord blood [
15]; however, children’s measured metabolite levels have shown to be clearly lower than those of their mothers. Moreover, a recent study has demonstrated statistically significant decreases in 6‑TGN levels and increases in 6‑MMP levels in expectant mothers, with no evidence of myelotoxicity or hepatotoxicity [
16]. Anemia has been identified at birth in approximately 60% of the neonates, yet the Apgar scores were normal and there was no evidence of congenital malformations (CM). These effects proved to be reversible and the anemia described has not yet been reproduced in subsequent studies. A rise in thiopurine S‑methyltransferase activity during pregnancy may explain the decrease in 6‑TGN and increase in 6‑MMP, as hormonal changes in pregnancy may influence enzymes regulating the breakdown of medications [
17,
18].
There have been two meta-analyses that substantiated that AZA and/or 6‑mercaptopurine (6-MP) only pose a minimal risk for fetuses [
19,
20]. No differences in CM or low birth weight (<2500 g) were identified in the first analysis in comparisons between pregnant subjects given AZA or 6‑MP and those not given thiopurine treatment. A significant difference in CM and low birth weight was only shown in comparison with the overall population. The second meta-analysis found no evidence of low birth weight or CM [
20]. In both publications, the increased risk of preterm delivery (prior to the 37th GW) during AZA and/or 6‑MP intake was primarily seen in connection with disease activity and not considered to be drug-related. Nevertheless, a slight risk of premature births cannot be excluded under thiopurines. Thus, a Swedish study has suggested that thiopurines elevate the risk of preterm delivery in patients with stable inflammatory bowel disease (IBD) (adjusted odds ratio, aOR, 2.41; 95% confidence interval, CI, 1.05–5.51) and active IBD (aOR, 4.90; 95% CI, 2.76–8.69) [
8]. In turn, the most recent investigation, a prospective cohort study with 309 IBD patients, showed no association between mothers’ thiopurine intake (35% of pregnancies) during gestation and increased risks of spontaneous abortion, “poor” birth results, or more frequent disorders among children within their first year of life [
21]. Thus, gastroenterological considerations serve to recommend continuing remission-maintaining thiopurine treatment during pregnancy [
22]. The European League Against Rheumatism (EULAR) also sees no reason not to continue thiopurine treatment but has limited doses to a maximum of 2 mg/kg/body weight [
23]. Due to the slow onset of effect and the risk of bone marrow suppression and pancreatitis, initiation of treatment with thiopurines during gestation is discouraged [
22].
Over an observation period of 3.8 years, a prospective study in 30 children whose mothers had taken thiopurines during pregnancy and/or breastfeeding found no evidence of physical or psychosocial developmental disorders, immunodeficiencies or increased risks of infection compared to a normative control group [
24]. The US Food and Drug Administration (FDA) adverse event reporting system has currently also received no reports indicating that thiopurines alone or in combination with TNF-α blockers are associated with an increased risk for mothers and their infants [
25]. Finally, having published the outcomes of 797 pregnancies in abstracts only, the Pregnancy in Inflammatory Bowel Disease and Neonatal Outcomes (PIANO) registry found no elevated risks of spontaneous miscarriages, CM, preterm deliveries, intrauterine growth disorders, abnormal development or cesarean sections under thiopurines [
26]; however, a higher infection rate was described in babies aged 9–12 months in the combination group (thiopurines and TNF-α blockers) compared to non-exposed children.
Within the first 4 h following intake, 6‑MP is particularly detectable in breast milk, whereas the quantity assessed corresponds to less than 1% of the maternal doses [
27]. The authors therefore recommend expressing and discarding milk within 4 h after intake of the substance; however, it should be noted that this is an additional safety measure, as neither 6‑MP, 6‑TGN nor 6‑MMP have so far been detected in the blood of breastfed infants, nor is there clinical or hematological evidence of immunosuppression [
28,
29]. No increase in risk of infections was observed in 15 children whose mothers had taken AZA in pregnancy and the lactation period compared to non-exposed subjects [
30]. These children also showed normal courses of physical and mental development.
Cyclophosphamide
Cyclophosphamide (CP) is a potent immunosuppressive with embryotoxic and teratogenic properties and should be discontinued 3 months prior to planned pregnancy. Accordingly, available data is limited to two cohort studies and case reports, including oncological patients and yielding a significantly increased risk for both miscarriages (OR 25 5) [
31] and CM (>26%) [
32,
33]. Administration of CP may only be considered in the second and third trimesters in the presence of severe (life-threatening or organ-threatening) disorders in mothers and the absence of treatment alternatives, always following strict indications. It crosses into the milk and, as an alkylating agent, may lead to bone marrow depression in neonates. It is therefore contraindicated in the lactation period.
Cyclosporine A, tacrolimus
Cyclosporine A (CsA) and tacrolimus (TAC) are second-line agents used to treat acute and severe courses of ulcerative colitis (UC), systemic lupus erythematosus, pyoderma gangrenosum and psoriasis; however, most data pertaining to gestation and lactation derive from patients treated with these substances on account of organ transplantations. Observations have become available with respect to more than 1100 pregnancies under CsA and more than 500 under TAC [
23]. Increased rates of preterm deliveries and decreased birth weight were identified with both agents, which, however, may have been facilitated by the patients’ primary disorder. No elevated rates of deformities were observed under these treatments [
23]. Higher rates of eclampsia and hypertension were seen in patients under CsA and of gestational diabetes in those under TAC. Cases of temporary renal failure and hyperkalemia were identified among neonates whose mothers had been given TAC during pregnancy. In line with the recommendations issued by EULAR and the Embryotox center in Berlin, and in appropriate indications, CsA and TAC may be prescribed during gestation. Patients stably medicated with CsA or TAC should not switch and level measurements are recommended.
Data on children breastfed under these treatments are available for 76 cases under CsA and 154 cases under TAC [
23]. In the majority of infants, no or merely subtherapeutic levels have been detected and no side effects have been observed. The EULAR and Embryotox recommendations thus consider breastfeeding under these medications as feasible. Breastfeeding is fully possible under topical TAC but the substance must not be topically applied to the mamilla.
Leflunomide
The effect of leflunomide (LEF) is based on the inhibition of dihydroorotate dehydrogenase, pyrimidine nucleotide synthesis and protein tyrosine kinases. Animal experiments with rodents have shown the drug to be embryotoxic and teratogenic but with no corresponding CM pattern in humans. Its half-life is long and LEF can be detected in tissues up to 2 years following discontinuation, where it has an abortion-inducing effect.
A total of three studies have investigated pregnancy outcomes under LEF, both in the preconceptional setting and into the first trimester, and found no increased risk of CM. In most cases, however, wash-out procedures with cholestyramine had been carried out [
34‐
36]. For this reason, these definitely promising data have not yet resulted in a modified assessment of LEF with respect to its application in pregnancy.
The data are similar in neurology, in the framework of which teriflunomide, the active metabolite of LEF, is applied in multiple sclerosis therapy and in which no typical CM patterns have been seen in relevant pregnancy reports [
37]. Even though the substance has so far not been proven to be a major human teratogen, its application in gestation is contraindicated due to insufficient data. It is recommended to discontinue the medication 2 years prior to planned pregnancy. In women desiring to have children or in cases of unintended pregnancy, wash-out procedures with cholestyramine (8 g 3 × daily for 11 days) are recommended within 2 years after discontinuation. The objective is a negative LEF level (<0.02 g/l) in 2 level determinations at an interval of 14 days.
Due to insufficient data and the long half-life of LEF, its application is not recommended in the lactation period [
23,
38].
Methotrexate
The folic acid antagonist methotrexate (MTX) may have a teratogenic effect, particularly in high doses, if applied during pregnancy. Given in the first trimester, high-dose MTX may result in typical embryopathy with craniofacial anomalies, anomalies of the central nervous system or extremity malformations. The MTX is applied in various doses in various indications. Thus, individual high doses of MTX, such as 1 mg/kg bw or 50 mg/m2, may already lead to miscarriages due to inhibition of DNA synthesis. Historically, this property was used for elective, medically induced abortion. Low doses (5–25 mg/week) are commonly applied to treat chronic inflammatory arthropathies, systemic rheumatic disorders and IBD. Only a few cases of MTX embryopathies following low-dose treatment have been reported in the literature.
A prospective multicenter cohort study [
39] investigated the outcomes of 324 pregnancies in patients treated with low-dose MTX (≤30 mg/week). MTX was given to 136 patients within 12 weeks pre-conception and to 188 post-conception [
39]. The pregnancies were compared with a disease-matched and a healthy cohort. Neither the risk of spontaneous abortion nor of severe CM (3.5%) was aggravated in the pre-conception cohort under low-dose MTX. The cumulative incidence of spontaneous abortion was significantly higher (42%) in the post-conception cohort than in either control group, as was the risk of severe CM (6.6% vs. 3.6% and 2.9%, respectively). There were seven children with large CM, yet none showed typical MTX-induced embryopathies [
39].
A study by Martin et al. explored the outcomes of eight pregnancies in patients receiving low-dose MTX. This case series yielded one typical MTX embryopathy [
40]. Other studies have found no increase in CM rates following low-dose MTX exposure in the first trimester [
41]. Therefore, it is generally recommended to discontinue MTX treatment 3 months prior to planned conception [
23]. Treatment discontinuation is to be accompanied by high-level folic acid substitution (5 mg/day), sustained until the end of the first trimester. The necessary minimum period between discontinuation and conception remains a matter of controversy. Based on current data, and under certain circumstances, an MTX-free interval of only 1 month may be acceptable [
39]. MTX crosses into breast milk, at least in traces, while the concentration of MTX in the milk is less than 10% of that in plasma [
42]. Nevertheless, breastfeeding should be refrained from under MTX [
43].
Mycophenolate mofetil
Mycophenolic acid is a potent, selective and reversible inhibitor of inosine monophosphate dehydrogenase, which blocks the synthesis of guanine-containing nucleotides. Its half-life is 12 h following oral intake of mycophenolate sodium and 16–18 h following mycophenolate mofetil (MMF) administration. Its molecular weight is low and it thus may pass the placental barrier. In conclusion, MMF intake during pregnancy is associated with an increased teratogenic risk [
44]. Hoeltzenbein et al. published the results of a prospective European multicenter cohort study with 57 pregnancies under MMF and identified a 45% cumulative incidence of spontaneous abortions and 26% risk of severe CM [
45]. In addition, pregnancies were electively terminated in 12 patients, in 2 of them termination was based on evidence of multiple fetal malformations. Among the 29 live births, 6 severe malformations were detected, e. g. auditory canal atresia, tracheoesophageal atresia, hydronephrosis and atrial septal defect. A nationwide cohort study in the UK described the course of pregnancies in patients following organ transplantation [
46]. Of these patients nine received MMF, amongst whom “poor fetal outcomes” were reported in seven patients, including spontaneous abortions, birth weights below 1500 g and CM. The European Medicines Agency (EMA) thus had a press release issued to point to the aggravated teratogenic risk associated with mycophenolic acid, especially when taken by men prior to conception [
47].
Tofacitinib
Tofacitinib (TOF) is an oral Janus kinase inhibitor approved for the treatment of rheumatoid arthritis (RA). Data on the influence of TOF in expectant mothers or its crossing into breast milk are scarce. Pregnant women have been excluded from randomized controlled approval studies, as the small-molecule substance may possibly pass to their fetuses. In spite of the compulsory application of effective contraceptive methods, pregnancies did occur in those investigations and required discontinuation of treatment and follow-up. An assessment of 47 pregnancies (33 under TOF monotherapy and 13 under TOF and MTX combination therapy) yielded no increased risk of deformity compared to the overall population. In addition, no elevated risk potential was identified among the 25 pregnancies documented in the approval studies in UC (OCTAVE 1, 2 and open-label extension) compared to non-exposed patients [
48,
49]. Teratogenic effects (cardiac, skeletal and cranial malformations), increased abortion rates and lower birth weights have been observed at substantially elevated doses in preclinical animal studies. Rat models have shown TOF to cross into maternal milk [
50].
In summary, no clear recommendation regarding the administration of TOF in pregnancy and lactation is currently possible on account of the low level of experience [
23]. Due to hematological side effects, it is recommended to discontinue treatment 6 weeks prior to planned pregnancy in spite of its short half-life of 3 h.