There is a predominance of females in PV, demonstrated by the female: male ratio of 1.4 and 2.3, according to epidemiological studies carried out in China and Greece.3,11 A worse outcome of PV has also been observed in women; a recent analysis showed that the HLA alleles DRB1*04:02 and DQB1 *03:02 were associated with severe PV, and DQB1*03:02 was found more frequently in female patients.

A review involving 49 pregnant women with PV found that 37% had the disease for the first time during pregnancy, and adverse pregnancy outcomes (i.e., neonatal pemphigus and perinatal death) occurred in up to 10% of cases. PV tends to improve in pregnancy, after the third trimester, and reactivate in the postpartum period. Therefore, pregnancy can precipitate or worsen PV.12

The occurrence of neonatal PV is transitory, with its resolution occurring within three weeks; it is usually due to transplacental transmission of IgG antibodies from mother to fetus and is easily treatable.13 In addition to neonatal PV, other consequences for the newborn include intrauterine growth restriction, preterm newborn, and fetal death; however, it is difficult to differentiate the consequences of treatment from those of the disease itself.14 Moreover, differences in desmoglein expression in neonatal skin account for the diversity of clinical manifestations between mother and child.13 Adverse events seem to be more correlated with poor control of maternal disease, higher serum and umbilical cord level of maternal antibodies, than with specific medications used to treat maternal PV.15 Therefore, treatments that lead to decreased maternal serum antibody levels aiming to limit the transplacental passage of pathogenic IgG justify the immunosuppressive therapy, although this approach during pregnancy is not well established yet.15 It should be remembered that, before the introduction of systemic corticosteroid use for PV, the disease had high mortality rates.

PF has two forms: the classic (Cazenave's disease) and the endemic form. The latter includes Fogo Selvagem (FS), which corresponds to the South American endemic form of the disease.16 There is no report of gender predominance in PF or FS, but the endemic forms of PF observed in North Africa (Tunisia) show a characteristic female predominance, especially among young women (female-male ratio of 4:1). In a series of cases of 23 Tunisian patients with pemphigus, the triggering role of pregnancy in the development of this disease was suspected, as Tunisia is a country where young women have a high fertility rate.17

PF can have a variable course during pregnancy, with no effect on the fetus. The association with neonatal PF is extremely rare in the context of exceptionally high antibody titers.18 Desmoglein 3 is expressed throughout the epidermis, including the subcorneal layers in neonatal skin, and may be protective against disease-causing anti-desmoglein 1 antibodies.13,19

A study carried out in Brazil with 19 pregnant women with FS did not observe any clinical neonatal disease, direct immunofluorescence was negative in 12 of the 17 neonatal skin specimens, IgG autoantibodies were present in low titers (less than 1:40) and IgG4 was the predominant IgG subclass in 9 of 19 umbilical cord blood samples. The authors suggest that the placenta may serve as a biological immunosorbent for pathogenic autoantibodies.20

EBA is a rare, phenotypically heterogeneous subtype of AIBD, predominant in females and characterized by autoantibodies against type VII collagen. Two studies analyzed EBA and its associated data, with a total of 83 cases, showing a female-male ratio of 1.56.21,22

Linear immunoglobulin A (IgA) dermatosis (LAD) is a rare autoimmune bullous disease with subepidermal blisters, in which IgA autoantibodies recognize as antigens the collagen type XVII subepidermal ectodomains (COL17, BP180) in the basement membrane zone (BMZ). There is scarce epidemiological data on its occurrence by sex, as well as its incidence in pregnancy.23 Patients are categorized into two groups: one with prepubertal onset and the other with a later onset (adult LAD). Some patients with LAD are young women who have become pregnant. Little is known about the effect of LAD and its treatment during pregnancy, or about the effect of pregnancy on LAD. Its antigen is known to be expressed in the amniotic basement membrane and IgA deposition occurs in the area of ​​the amniotic basement membrane of the affected mother’s fetus.24

According to a case series of LAD in pregnancy, the disease usually improves by the end of the first trimester and usually recurs in the second trimester but it may relapse in the postpartum period. For this reason, it may be possible to discontinue dapsone use during pregnancy. There is evidence that IgA antibodies become glycosylated during pregnancy, which could explain this remission, and Treg cells would also play a role. However, patients should be counseled about the possibility of a postpartum relapse. The disease can also reactivate years later, even after a prolonged period of remission.25

DH is closely associated with celiac disease, with both conditions being mediated by IgA autoantibodies and targeting the transglutaminase autoantigen. In contrast to pemphigus and pemphigoid, the evidence for sex predominance in DH is conflicting, with no conclusion on its prevalence in either the female or male sex.26 As for its occurrence during pregnancy, there is scarcity of data, with only two articles found in the literature.

Forty-nine articles were included (Appendix A), which reported a total of 57 women with 59 pregnancies (Table 2); 31 patients (54.3%) were diagnosed with pemphigus before pregnancy and 26 (45.6%) during pregnancy. Fifty-one patients (89.4%) were diagnosed with pemphigus vulgaris and 6 (10.5%) with pemphigus foliaceus. The mean age at disease onset was 28.4 years and the mean gestational week at disease onset/worsening was 16 weeks. The evolution of the disease was reported in 50 pregnancies. In 33/50 (66%) pregnancies the disease worsened, in 14/50 (28%) the disease remained stable, and in 3/50 (6%) the disease improved. In twelve of the 50 pregnancies (24%), a postpartum reactivation was reported, and adverse pregnancy outcomes such as neonatal pemphigus, perinatal death, or miscarriage were reported in 33% of these pregnancies. As for the treatment of pemphigus during pregnancy, 51 patients had their treatments described. In 41 of these 51 patients (80.3%), monotherapy was the treatment of choice and in 10 of 51 (19.6%), combination therapy was reported. Rituximab was used in 4 of 51 cases (7.8%); systemic steroids in all 51 (100%); azathioprine in 5 of 51 (9.8%); mycophenolate mofetil in 3 of 51 (5.8%); dapsone in 1 of 51 (1.9%) and intravenous immunoglobulin in 4 of 51 (7.8%).

Three articles were found on LAD and pregnancy. In one of them, 12 patients had a total of 19 pregnancies. The disease remained active in 13 of 19 pregnancies, underwent remission in four of 19 pregnancies, and showed recurrence in the postpartum period in two of 19 pregnancies. The age at LAD onset ranged from 1.5 to 33 years; three patients had chronic bullous disease of childhood, and nine had adult-onset LAD. Overall, LAD tends to improve during pregnancy, especially around the 10th week of pregnancy. Twelve of 19 patients received systemic treatment (7/12 received dapsone), 4/19 were in remission, and 1/19 had minimal disease without therapy.25

Two case report articles described LAD and pregnancy: one refers to a 24-year-old woman diagnosed with LAD and hydatidiform mole, excised in her third pregnancy, and the other refers to a 29-year-old woman in her 38th week of gestation, with autoantibodies against the non-collagen 16A domain (NC16A) of COL17.23,27

Three articles on EBA and pregnancy were retrieved. In one, a 32-year-old woman was diagnosed with EBA and, at the time of delivery, both mother and newborn had superficial and deep erosions on the face, chest, abdomen, and extremities.28 The second article describes a patient who developed EBA on the 2nd postpartum day with resolution of the blisters at menopause.29 The third article reports on a 26-year-old woman who had a recurrence of EBA during the first month of pregnancy. Due to insufficient response to systemic corticosteroids, she had a miscarriage in the 2nd month of pregnancy and then had a significant reduction in the number of bullous lesions.30

Only one article was retrieved. A 33-year-old female patient with severe DH became pregnant, requiring oral dapsone (25–50 mg per day) to control the disease. She gave birth to a full-term, healthy newborn, and progressed with disease reactivation in the postpartum period.31

For all cases with clinical suspicion of AIBD, the diagnosis is confirmed through cytological examination, skin biopsy with histopathology, direct and indirect immunofluorescence, and ELISA. Therefore, a patient with a confirmed diagnosis of AIBD prior to pregnancy does not need to be resubmitted to investigative workup, unless the current lesions do not show typical characteristics of the previously diagnosed AIBD. However, diagnostic confirmation is necessary in cases with symptoms suggestive of AIBD for the first time during pregnancy. Regarding the skin biopsy, topical application of chlorhexidine for antisepsis can be safely used in pregnant women. There are no reports of teratogenic effects related to topical anesthetics. However, a skin biopsy should be performed with caution in pregnant women due to the risk of fetal methemoglobinemia associated with prilocaine and decreased uterine flow secondary to the use of local anesthetics.32 For these reasons, alternative methods may be useful and include serological investigation using indirect immunofluorescence, ELISA, and biochip immunofluorescence microscopy.33

Topical glucocorticoids are often the first line of treatment for AIBD and remain safe for use in pregnancy, with the advantage of not sharing the same risks as systemic steroids.36

Apgar scores, congenital abnormalities including orofacial clefts, preterm birth, type of delivery, and fetal death were not related to the use of topical glucocorticoids of whatever potency. One exception was the dose-response association between topical glucocorticoids and intrauterine growth restriction (IUGR) and low birth weight (LBW) newborn, specifically when applied in the third trimester at very high doses.36

Topical glucocorticoids are safe during lactation, as transmission through breast milk is negligible and glucocorticoids constitute a natural element of breast milk.37 However, high-potency topical glucocorticoids should not be applied to the nipple prior to breastfeeding, due to the risk of infant high blood pressure.38

Case reports and a double-blind, placebo-controlled clinical trial have demonstrated the efficacy of topical calcineurin inhibitors in the treatment of AIBD, particularly if the disease is mild and localized.39 These agents are often used to prevent the side effects of long-term topical steroid use, especially in areas such as the face, neck, skin folds, and genitals.10

The effects on the fetus, which can be extrapolated from studies on the systemic administration of tacrolimus in transplant patients, include preterm delivery and LBW newborn. Although formal studies have not been carried out in pregnant women, experts recommend using a maximum of 5 g per day, for 2 to 3 weeks, in small areas.40

Safety recommendation data for topical tacrolimus during lactation is limited; therefore, it should be used sparingly, avoiding its use on the nipple area.41

Systemic steroids are still considered an important and safe treatment option for AIBD in pregnancy. Non-fluorinated steroids, such as prednisone and hydrocortisone, are generally the preferred forms because the inactivation by 11-beta-hydroxysteroid dehydrogenase limits fetal exposure.10 The risks for the fetus are orofacial clefts, IUGR, LBW newborn, premature rupture of the membranes, delayed neuropsychomotor development, and preterm birth, whereas for pregnant women the main side effects include pre-eclampsia, eclampsia, high blood pressure, gestational diabetes, osteoporosis, poor wound healing, striae formation, and postpartum and postoperative infections.10,40,41 The recommended dose should not exceed 20 mg/day of prednisone, and long-term use should be avoided.41 If higher doses are required, the option is to switch to a steroid-sparing agent for disease control.

Maternal breastfeeding should take place at least 4 hours after prednisone administration, as its serum half-life is 60 min, with levels in breast milk peaking after 2 hours of ingestion and declining rapidly thereafter.10,40

Azathioprine is an off-label synthetic purine-based analog used in the treatment of AIBD.10,40,41 Several clinical studies and reviews recommend azathioprine as a relatively safe immunosuppressant to be used during pregnancy and lactation.42 The placenta can limit the entry of azathioprine and its metabolite, 6-mercaptopurine. Transient lymphopenia and immunodeficiency can be prevented if the daily dose does not exceed 2 mg/kg/day.43

The European Crohn's and Colitis Organisation classifies azathioprine as “probably safe” for use during breastfeeding.42 Breastfeeding should take place at least 4 hours after the last dose.44

Mycophenolate mofetil (MPM) is an excellent steroid-sparing agent in autoimmune diseases, including AIBD. However, it is a known teratogenic component and should not be used during pregnancy.45 Like methotrexate, MPM inhibits DNA synthesis and stops cell division, causing an increased risk of miscarriage. MPM has been associated with miscarriages (especially in the first trimester), limb and face abnormalities, midline defects, and other congenital malformations.45

Despite the scarcity of data on the use of MPM during breastfeeding, experts currently classify it as contraindicated. The small molecular weight of MPM allows it to pass in significant amounts into breast milk, which could cause developmental problems and risk of infection in the newborn.38

Dapsone (diaminodiphenylsulfone) is often used as adjuvant therapy for AIBD. Case reports have shown it is safe to use it during pregnancy, with no evidence of teratogenicity, as well as in studies of pregnant women being treated for malaria, leprosy, or pneumonia caused by Pneumocystis jiroveci.41 Dapsone, especially at a dose > 50 mg/day, should be avoided in mothers or fetuses with glucose-6-phosphate dehydrogenase (G6PD) deficiency, due to the risk of hemolytic anemia, and all patients should be tested before use.46

The American Academy of Pediatrics considers dapsone to be compatible with breastfeeding, as only small amounts of its metabolites are excreted in breast milk. Jaundice in the baby should be monitored if breastfeeding occurs during dapsone therapy, especially in case of G6PD deficiency diagnosed in the newborn.

Doxycycline is an antibiotic with anti-inflammatory effects and can be useful in the treatment of AIBD.47 Doxycycline has long been combined with other tetracyclines and classified as unsafe during pregnancy, but this combination may not be correct, as doxycycline was developed after the toxicity of the other tetracyclines was well established. Doxycycline is likely to have a different safety profile from other tetracyclines, with no evidence to support overlapping toxicities. Tetracyclines carry an increased risk of teratogenicity, tooth discoloration, bone growth arrest, and maternal hepatotoxicity, but doxycycline does not seem to be associated with these risks. The data that differentiates doxycycline from tetracyclines is still limited, so clinicians tend to avoid doxycycline when there is a choice for other drugs with a known safety profile.48

Short-term doxycycline use during lactation is not contraindicated, but the drug is excreted in breast milk and studies are still scarce.35

Rituximab (RTX) is a chimeric anti-CD20 monoclonal antibody that targets B cells, involved in the pathogenesis of AIBD. RTX is successful in providing disease remission, but it is contraindicated during pregnancy, as it crosses the placenta, depletes fetal B cells, and potentially increases the risk of maternal-fetal and newborn infections. The transplacental passage of RTX is minimal in the first trimester, moderate in the second, and higher in the third. RTX remains in serum for 3 to 6 months and is able to cross the placenta; therefore pregnancy planning is essential and patients are advised to wait 12 months after their last infusion before attempting to conceive.10,35 In case reports of patients who accidentally received RTX during pregnancy, no definitive correlation was observed with adverse events, but data on the subject remain scarce.49

RTX is not recommended during lactation due to a lack of evidence on the subject.50