Salicylates and emollients are drugs of choice for use during pregnancy. Systemic absorption of topical treatment depends on the concentration and duration of application and can be 10% to 25% of the amount applied to the skin3. Therefore, high concentrations (>3%), large quantities (>20 g/d) or use under occlusion should be avoided to prevent penetration4–6.

During breastfeeding, the application of topical treatments to the breast, large areas of skin, or for prolonged periods should be avoided in order to prevent direct absorption by infants, although significant amounts are unlikely to be found in breast milk7,8.

The retinoids adapalene, tazarotene, and tretinoin are potentially teratogenic during pregnancy9. Although at standard doses teratogenic, mutagenic, or carcinogenic effects have not been observed, teratogenicity may occur if applied to more than 20% of the body surface area4,5.

During breastfeeding, the transcutaneous absorption of tazarotene is very low and thus it appears to be a low-risk treatment.

Due to its larger molecular size, topical pimecrolimus has poor systemic absorption. Systemic accumulation is negligible after repeated applications, application to large body surface area (up to 69%), or after long-term use10. However, its use is not recommended during pregnancy7,8.

Ultra-violet B (UVB) therapy and narrowband UVB (NB-UVB) therapy appear to be safe during pregnancy. However, maternal folate depletion may occur and folic acid supplementation is recommended5. A study that included 107 patients under psoralen+UVA (PUVA) therapy reported cases of premature delivery and foetal abnormalities. The FDA has classified this drug within pregnancy category C. Oral psoralen is administered to patients to increase skin reactivity to UVA rays. There is a theoretical risk of teratogenic and mutagenic effects, which are due to DNA synthesis and the inhibition of cell division. It is recommended to avoid it during pregnancy4,5,9.

During pregnancy, paracetamol is the safest analgesic and antipyretic drug, but also during periconception and throughout pregnancy. Experimental studies in animals and prospective cohort studies in humans have shown that there are no increases in malformations at therapeutic doses. There is evidence suggesting a positive, although weak, association between prenatal paracetamol use and allergies and asthma in children, adolescents, and adults. This finding is probably related to a decrease in glutathione levels in the lungs, which causes oxidative damage and inflammation. Therefore, it is recommended to use intermittently at the lowest effective dose and for the shortest time possible11,12.

During the first trimester of pregnancy, the use of classic non-steroidal anti-inflammatory drugs (NSAIDs) appears to be free of the risk of teratogenicity or abortion. However, in the third trimester, they can cause premature closure of the ductus arteriosus, pulmonary hypertension, neonatal anuria, anemia, increased risk of pre- and postpartum hemorrhage, and oligohydramnios. They may also increase the risk of intracranial hemorrhage in preterm infants or low-weight neonates by causing decreased platelet aggregation in fetus.

High doses of acetylsalicylic acid (>3 g) inhibit uterine contractility and prolong gestation, whereas during pregnancy the use of low doses as an antiplatelet agent has been shown to be safe12.

Overall, NSAIDs can be used safely during breastfeeding, although some risk of jaundice and kernicterus has been described7,12. Ibuprofen, diclofenac, and naproxen are the most commonly used NSAIDs. Dose-dependent sperm abnormalities have been described in fathers under treatment with NSAIDS. In women, NSAIDs inhibit the production of prostaglandins, which play a role in ovulation and blastocyst implantation13–15.

Cyclooxygenase 2 (COX-2) inhibitors are associated with an increased risk of premature closure of the ductus arteriosus, especially during the third trimester of pregnancy. Therefore, their use is contraindicated in pregnancy14.

An observational study compared the effect of COX-2 inhibitors in 174 women treated with them during the first trimester of pregnancy and 521 women without COX-2 exposure. No significant differences were found in spontaneous abortions or congenital abnormalities, although more elective abortions were reported in the former group16.

In a cohort of pregnant women in Quebec, the risk of prematurity was 2.46-fold higher in those who were exposed to COX-2 inhibitors in the 3 months prior to delivery than in those who were not exposed to them. With celecoxib, the risk was 3.41-fold higher, suggesting that the use of celecoxib in advanced pregnancy may increase the risk of prematurity17.

Some celecoxib studies have shown that there is a minimal transfer of the drug into breast milk (<0.3%). Therefore, its use is considered to be safe during breastfeeding18, although some guidelines have contraindicated its use12.

Data on male fertility and COX-2 inhibitors are limited and make reference to data on NSAIDs, as described above. In women, the use of COX-2 inhibitors has been associated with reversible infertility due to alterations in ovarian follicle rupture. This effect also occurs with standard NSAIDs. Women trying to conceive are recommended to avoid using them12.

No adverse effects have been described for the aminosalicylates mesalazine and sulfasalazine (SSZ), and are considered to be of low risk during pregnancy. SSZ has antifolic effects. Thus, folic acid supplements are recommended during the prenatal period and throughout pregnancy and do not exceed doses of 2 g/day. It can be used during breastfeeding because the amount excreted in breast milk is very small. Nevertheless, caution is recommended because cases have been described of blood in faeces20, and diarrhea in infants and jaundice in the newborn7,19.

Reversible oligospermia, asthenozoospermia, and teratozoospermia have been observed in men treated with SSZ, and thus it should be suspended before conception21,22.

Maternal treatment with azathioprine (AZA) or mercaptopurine (6-MP) during pregnancy is clearly teratogenic in animals at standard doses, but there is limited data on their effects in humans treatment with these drugs may interfere with DNA replication in germ cells by inhibiting nucleic acid synthesis23.

The Pregnancy Inflammatory bowel disease And Neonatal Outcomes (PIANO) registry investigated the effect of exposure to thiopurines and anti-tumor necrosis factor (TNF) in more than 1000 women from conception to delivery. No association was found between thiopurinehiopurine or anti-TNF use and congenital abnormalities or pregnancy complications. However, there was a significant increase in infections at 9 to 12 months of age among the 107 infants exposed to thiopurine andanti-TNF (RR = 1.50, 95% CI 1.08 to 2.09)24. Due to the potential for late infancy infections, the discontinuation of thiopurine after conception could be considered25.

The risk of congenital malformations was higher in women with irritable bowel syndrome (IBD) on thiopurine treatment than in healthy women (RR = 1.45; 95% CI 1.07 to 1.96). However, no increase was seen in comparison to disease controls (RR = 1.37; 95% CI 0.92 to 2.05).

Although most of the current literature does not recommend the administration of thiopurines to pregnant patients7,25,28–30, these results provide support for their use in patients at high risk of relapse27,28.

Mercaptopurine has been detected in the colostrum and breast milk of women receiving AZA treatment. It is recommended to wait 4 h after a dose before breastfeeding, as the peak of 6-MP in milk occurs within the first 4 h after ingestion31.

Although cases of congenital abnormalities have been reported in parents treated with AZA or 6-MP32, Akbari et al.33 did not find significant associations between paternal exposure to thiopurines at the time of conception and congenital abnormalities (combined OR: 1.87; 95% CI 0.67 to 5.25).

Fetus are at low risk from hydrocortisone, cortisone, prednisolone, and methylprednisolone because they are inactivated in the placenta. Although prednisone does not present teratogenic risk to humans at therapeutic doses, it is recommended to avoid high doses (1–2 mg/kg) in the first trimester due to the risk of cleft palate34. Doses of more than 20 mg/d can cause adrenal hypoplasia in fetus with transient adrenal insufficiency after delivery. Associations have been found between doses of prednisone of more than 5–10 mg/d and an increased risk of gestational diabetes, hypertension, oedema, premature membrane rupture, and osteoporosis. Chronic use causes suppression of adrenocorticotropic hormone (ACTH) production.

Corticosteroids can be administered during breastfeeding, but it is recommended to wait 4 h after a dose before breastfeeding to minimize infant exposure, and especially if the dose is more than 20 mg prednisone/day7,35.

Studies have shown that paternal exposure to prednisolone does not lead to increased rates of malformations15,29,35,36.

Methotrexate (MTX) has been classified by the FDA within pregnancy category X. It is contraindicated during pregnancy because of its teratogenic effect in humans25,29, especially in the first trimester of pregnancy, and even at doses of less than 30 mg/wk. In fact, its embryolethal effect is used in the treatment of ectopic pregnancy37. A prospective multicentre cohort study of 324 pregnant women exposed to MTX found that the incidence of spontaneous abortion and the risk of congenital defects were higher in this group than in the control group. No evidence of increased risk was found in the 136 pregnant women exposed to MTX before pregnancy38,39. It is excreted in breast milk and, because of its long half-life, it accumulates in the tissues of infants. Thus, its use in breastfeeding is contraindicated7.

Cyclosporine A (CsA) crosses the placental barrier and can reach fetal blood concentrations of up to 50% of the concentration in maternal blood40. The FDA has classified it within pregnancy category X. However, no teratogenic effects have been reported in humans.

There is no evidence of its negative impact on male and female fertility in patients treated with CsA4,15,41,42. Differences in sperm motility and shape were found in three cohorts of 19, 26, and 212 men who were treated with CsA. These differences were associated with a higher dose of the drug. They also showed normal spermatograms at 2 years after renal transplantation. Thus, it is recommended to wait for 2 years before conception15,43,44. There is no evidence of congenital malformations in children born to parents exposed to CsA44. Thus, doses of less than 2 mg/kg/d CsA do not seem to have a negative effect on male fertility, nor does paternal exposure lead to fetal harm15,43,44.

Regarding tacrolimus, there is no evidence that the risk of its use during pregnancy is higher than the risk of using other immunosuppressants.However, cases of spontaneous abortion have been reported, and so other safer alternatives may be considered25,27,30.

Mycophenolic acid (MPA) is a teratogenic agent and should not be used during pregnancy45. A prospective European multicentre cohort study of 57 pregnant women who had been exposed to mycophenolate mofetil (MMF) found a 45% cumulative incidence of spontaneous abortion and a 26% risk of severe congenital malformations46. It is recommended that treatment should be discontinued 6 weeks prior to conception7,27,47.

It is contraindicated during the breastfeeding period because of the potential risk of serious adverse reactions to MMF in nursing infants7,27,47.

Three studies on pregnancy after parental exposure to MMF derivatives found that the risk of congenital malformations in the children was no higher than in controls and the general population48. Nevertheless, it is recommended that treatment be discontinued 3 months prior to attempting conception7.

Hydroxychloroquine (HCQ) remains the antimalarial drug of choice in women with rheumatic disease who wish to conceive22. A prospective observational study found no association between its use and congenital malformations or neonatal infections in 40 infants born to mothers affected by rheumatic disease under treatment with HCQ. Preterm delivery was identified as the main complication (20.5%), which may be related to the rheumatic disease49.

The results of a double-blind placebo-controlled HCQ study found no congenital abnormalities or neuroophthalmological and auditory abnormalities in infants of 1.5 to 3 years of age50. No evidence of fetal ocular toxicity was found by a systematic review of 12 studies that investigated 588 live births to mothers treated with chloroquine or HCQ during pregnancy51.

Hydroxychloroquine crosses the placenta and its concentration in umbilical cord blood is approximately equal to the maternal concentration. The excretion of HCQ in breast milk is very low (<0.2 mg/kg/day), and so it is considered compatible with breastfeeding22.

Leflunomide and its active metabolite, teriflunomide, cause serious birth defects. Cases have been reported of malformations in children born to mothers exposed to leflunomide during pregnancy52.

In case of exposure, the elimination of teriflunomide can be accelerated by the administration of cholestyramine or activated charcoal35,41,53. A study on women with RA who became pregnant during leflunomide treatment and were treated with cholestyramine found that the percentage of malformations was comparable to the 3% to 4% found in the general population54,55.

Although it remains unknown whether it is excreted in human breast milk, its use is contraindicated during breastfeeding due to the high risk of adverse events in infants.

Acitretin is teratogenic and embryotoxic in animals, and thus its use is contraindicated in pregnancy and breastfeeding7,56. It does not appear to affect male fertility57. Post-authorization studies of the drug have found no evidence of reproductive risk56.

Its use during pregnancy and breastfeeding is contraindicated. Experimental animal studies have found dose-relatedembryofetal loss, reduced birth weight, and delayed ossification in mice at doses higher than the currently recommended maximum human dose. These effects have not been observed in animals exposed to 1.3 times the clinical dose4.

Data on their use in humans are sparse and unclear due to frequent concomitant use with MTX. Tofacitinib has been shown to be teratogenic in rats and rabbits, and affects parturition and peri- and postnatal development.

ofacitinib clinical safety databases, post-marketing studies in RA, and adverse event reports showed that out of 47 pregnancies, one baby was born with a congenital anomaly. There were also seven spontaneous abortions, of which three occurred when used in combination with MTX58.

It is not known whether tofacitinib is excreted in breast milk. A risk for infants cannot be excluded because the low molecular weight of the tofacitinib molecule could allow its passage into breast milk. Its use during breastfeeding is contraindicated as a precautionary measure47,56.

There are insufficient data on the use of baricitinib in pregnant women and whether it is excreted in breast milk. Animal studies have shown that it is toxic and excreted into breast milk. It is teratogenic in rats and rabbits because it affects in utero bone development at high doses. A risk to fetus or newborns cannot be excluded, and thus o it is generally contraindicated during pregnancy and breastfeeding59.

Like other anti-TNFs, adalimumab is classified within pregnancy category B. A registry of 500 pregnant women found no evidence of embryotoxicity, teratogenicity, and spontaneous abortions60. Subsequent studies on AR and CD found that the relative risk of birth defects and spontaneous abortion was no higher in the adalimumab-exposed group than in controls. However, the risk of preterm delivery was higher in this group than in healthy women, regardless of adalimumab-exposure61,62.

Treatment with anti-TNFs during pregnancy could lead to neonatal immunosuppression and an increased risk of infections. Therefore, vaccination with live vaccines should be avoided during the first 6 months at least63 or adalimumab treatment should be suspended between week 20 and 24 of pregnancy22,47.

It is excreted through breast milk in very low concentrations. This IgG antibody undergoes intestinal proteolysis and has low bioavailability when taken orally. It can be used during breastfeeding, given that no adverse effects have been reported in children8.

A recent study analyzed the semen of 14 patients exposed to the drug and found no evidence of an effect on male fertility or harm to their offspring15 and no relevant or negative effects on fertility64.

Several cohort and case control studies including more than 300 pregnant women have shown no evidence of an increased risk of major birth defects47, although a case has been reported of a child born with Vater's disease after etanercept exposition during pregnancy65.

In 138 women exposed to etanercept, there were six spontaneous abortions, two terminations (one for cardiac anomaly and one Turner syndrome) and 130 births with no malformation patterns66.

Etanercept has been reported to be excreted in breast milk after subcutaneous administration. The aforementioned guidelines suggest that it is compatible with breastfeeding22,47. According to a recent review of anti-TNFs, etanercept does not appear to affect male fertility or harm offspring67.

Infliximab is an anti-TNF IgG1 monoclonal antibody. A systematic review reported that the PIANO registry found no differences in congenital abnormalities in the offspring of 260 women exposed to infliximab during pregnancy and in those of the control group68. Starting at week 30 of pregnancy, mean infliximab levels reach concentrations of 160% of the mother's and remain detectable for up to 6 months69.

It is excreted in milk and it is recommended that breastfeeding should be avoided for at least 6 months after treatment. The PIANO registry found no risk of infection or growth retardation in children whose mothers were treated with thiopurines or anti-TNFs70. Some guidelines suggest that it is compatible.

A discrete case series found that infliximab does not affect male or female fertility and is neither genotoxic nor embryotoxic71. However, in vitro studies have found evidence of decreased sperm motility and integrity72,73.

Certolizumab is the biologic drug of choice for pregnant women. A registry of 1137 patients found no evidence of teratogenicity or an increased risk of fetal death74. Its elimination half-life is 14 days. However, it does not have an Fc region and does not cross the placenta. Thus, guidelines recommend maintaining treatment throughout pregnancy22,47.

Its transfer from maternal plasma to milk is minimal and can therefore be administered during breastfeeding. According to the results of a clinical trial in 20 healthy men, semen quality and other semen properties were similar under certolizumab or placebo15.

Anakinra is an IL 1α and IL 1β inhibitor. Animal studies have shown that it does not have negative effects on fetus despite being detected in amniotic fluid, even at doses 100 times the therapeutic dose76. Previous case series have reported spontaneous abortion and two babies with congenital abnormalities77–79. Due to the lack of data, it is not generally recommended to continue treatment during pregnancy or during breastfeeding.

Animal studies have shown that abatacept crosses the placenta, and can reach fetal concentrations 1.7 to 2.4 times lower than those in maternal serum. Eratogenicity has not been demonstrated in mice at doses of up to 30 times higher than those used in humans80.

A recent study of 151 pregnant women exposed to abatacept reported 86 live births, 40 spontaneous abortions, and 19 elective abortions. In total, 51.3% of the participants had received concomitant MTX.Seven of the babies presented congenital abnormalities and of these, three were exposed to other concomitant teratogenic medications, such as mycophenolate and leflunomide81.

Tocilizumab is a monoclonal antibody against the IL-6 receptor. A study found a high rate of miscarriages, but no congenital malformations, in the offspring of women under concomitant treatment with MTX47. Due to a lack of data, the use of tocilizumab is not recommended during pregnancy or breastfeeding82.

Nonclinical data suggest that female fertility is unaffected by treatment with tocilizumab83. Animal studies using doses of up to 10 times higher than those used in humans have found no evidence of abnormalities in male fertility.No congenital malformations were recorded in 22 cases of pregnancies involving parents exposed to Tocilizumab15.

Rituximab is a monoclonal antibody medication that binds specifically to the CD20 antigen of B lymphocytes. B-lymphocyte levels have not been determined in the newborns of rituximab-exposed women, but transient B-cell depletion and lymphocytopaenia have been reported in some infants born to women exposed to rituximab during pregnancy47.

Exposure prior to pregnancy and during the first trimester does not appear to put fetus at excessive risk84, although transient lymphopaenia or neutropaenia have been reported. It suppresses neonatal B-cell development in fetus during the 2nd and 3rd trimester of pregnancy, but B-cell counts recover during the first 6 months of life76. Studies have found an increase in miscarriages, although remains unknown if they are directly related to exposure to the drug or to the disease. No increase has been found in congenital malformations47,84.

It is unknown whether rituximab is excreted in breast milk47. However, maternal IgG is excreted in milk and thus it is recommended that women should not breastfeed their children during treatment with rituximab or during the following 12 months8.

Studies have found that children exposed to maternal rituximab treatment during pregnancy have a normal response to vaccination. However, due to the limited data, caution is advised regarding its administration during pregnancy76,85.

Ustekinumab is an IgG1 monoclonal antibody that inhibits cytokines IL-12 and IL-23. Animal studies have shown that it has no adverse effects on pregnancy, embryo/fetal development, parturition, or postnatal development. Nevertheless, there are insufficient data on the use of ustekinumab in pregnant women and thus its use should be avoided7,86. One review referenced pregnancy outcomes for 26 of 29 pregnancies, of which 5 involved spontaneous abortions (19%). This rate is similar to that of the general population87. Other reviews have suggested that ustekinumab may be safe and have recommended its use even up to 8 to 10 weeks before delivery25.

Vedolizumab is a monoclonal antibody that targets the a4ß7 integrin, which is mainly expressed on T-helper lymphocytes that are recruited to the intestine

A case series of 24 women treated with vedolizumab reported 12 live births, four spontaneous abortions, five selective abortions, and one congenital anomaly88. Although its half-life is three times that of infliximab (25 days), it was found that levels were lower in newborns than in the 7 mothers who were exposed to treatment during pregnancy89.

The European CONCEIVE study was a retrospective observational case–control study of pregnant women with IBD. It included 79 women exposed to vedolizumab, a control group of 186 women exposed to anti-TNFs, and 184 unexposed women. No significant differences were found in the rate of spontaneous abortion between the vedolizumab and anti-TNF groups, or between the vedolizumab and unexposed groups. The mean gestational age, weight, and prematurity of newborns were similar between groups, even after correcting for disease activity during pregnancy. There were no differences in congenital abnormalities, nor were malignancies detected at 1 year after birth. Further prospective studies may be needed to confirm these data90.

Vedolizumab has been detected in breast milk with a peak of less than 1% of maternal serum concentrations91,92. Thus, although it is considered safe, further studies are required to evaluate its effect on the infant immune system25.

Secukinumab is a fully human IgG1/κ monoclonal antibody that inhibits IL-17A. There is a lack of data on the use of secukinumab in pregnant women. Animal studies have found no evidence of negative effects on pregnancy, embryonic or fetal development, parturition, or postnatal development.

One study analyzed a safety database that included 238 cases of maternal exposure and 54 cases of paternal exposure. Most patients discontinued secukinumab treatment in the first trimester of pregnancy. The miscarriage rate was 10.3% and there were three cases of congenital abnormalities. These outcomes are similar to those found in the general population93.

Ixekizumab is an IgG4-type monoclonal antibody that inhibits IL-17A.

Animal studies have found no evidence of negative effects on pregnancy, embryonic or fetal development, parturition, or postnatal development. There are insufficient data on the use of ixekizumab in pregnant women and it is recommended that its use should be avoided4. Given that it is an IgG4 antibody, it is thought that its rate of transport across the placenta is less than that of other monoclonal antibodies with an IgG1 structure94.

It remains unknown whether ixekizumab is excreted in breast milk or absorbed after ingestion. However, as it is a high molecular weight protein, it is thought that it is not absorbed during the first few weeks postpartum4,87.

Tildrakizumab is a humanized IgG1/κ-type monoclonal antibody that inhibits IL-23.

Animal studies have found no evidence of harmful effects on reproduction, but data on pregnant women are limited. Clinical trials with tildrakizumab in women who became pregnant did not find increased rates of spontaneous abortions or congenital abnormalities95.

It remains unknown if tildrakizumab is excreted in breast milk. No data are available on the response to live microorganism vaccines, and thus it is recommended that they are avoided during treatment and for at least 17 weeks thereafter86.