Neonatal Linear IgA Bullous Dermatosis Mediated by Breast Milk–Borne Maternal IgA

Key Points

Question

What is the origin of pathogenic IgA in neonatal linear IgA bullous dermatosis (LABD)?

Findings

This case study of a newborn with neonatal LABD and a healthy asymptomatic mother used split-skin indirect immunofluorescence testing to find that maternal breast milk contained immunoglobulin A (IgA) autoantibodies, which bound to the dermal side of the basement membrane zone. In addition, immunohistochemical staining demonstrated the deposition of secretory IgA in the neonatal skin by highlighting the presence of J chain—not seen in other cases of LABD—indicating that the autoantibodies producing blisters were derived from the maternal breast milk.

Meaning

In neonatal LABD cases, maternal breast milk should be examined for IgA autoantibodies, and if they are present, breast milk feeding should be discontinued.

This case report examines a case of linear immunoglobulin A bullous dermatosis in a breast-fed newborn with an asymptomatic mother.

Abstract

Importance

Neonatal linear immunoglobulin A (IgA) bullous dermatosis (LABD) is a rare disease that can be fatal when associated with respiratory failure. All previously reported cases of neonatal LABD have been in newborns with healthy asymptomatic mothers, and the pathogenic IgA was of unknown origin.

Objective

To clarify the origin of IgA associated with LABD in neonates born of healthy asymptomatic mothers.

Design, Setting, and Participants

This case study analyzed the laboratory findings of a single breast-fed newborn male with neonatal LABD admitted to the Keio University Hospital in Tokyo and his healthy asymptomatic mother. The healthy newborn developed life-threatening blisters and erosions of the skin and mucous membranes on day 4 after birth. Blood serum, skin, and maternal breast milk were examined for IgA autoantibodies.

Main Outcomes and Measures

Histopathologic and immunofluorescence analyses of specimens (serum, skin, and breast milk) from the patient and his mother.

Results

Histopathologic evaluation of the newborn’s skin revealed subepidermal blisters with neutrophil infiltrates, and immunofluorescence testing showed linear IgA deposition along the basement membrane zone (BMZ), which lead to the diagnosis of neonatal LABD. Indirect immunofluorescence using normal human skin after treatment with 1-mol/L sodium chloride showed the patient to have circulating IgA binding to the dermal side of BMZ. Immunohistochemical staining proved the deposition of secretory IgA in the neonatal skin by demonstrating the presence of J chain—not been seen in other LABD cases—indicating that the autoantibodies producing the blisters were derived from the maternal breast milk. Although no circulating IgA against the skin was detected in mother's sera, the breast milk contained IgA that reacted with the dermal side of the BMZ. No new blister formation was observed after cessation of breastfeeding.

Conclusions and Relevance

The results of this case study suggest a passive transfer of pathogenic IgA to a newborn from an asymptomatic mother via breast milk. In prior reports, no serum from asymptomatic mothers of newborns with LABD had IgA autoantibodies binding to skin components; however, in this case, we found that the maternal breast milk contained IgA autoantibodies associated with neonatal LABD. In neonatal LABD, maternal breast milk should be examined for IgA autoantibodies and breast milk feeding should be discontinued as soon as neonatal LABD is suspected.

Introduction

Blistering diseases in newborns have various causes, including viral and bacterial infections and genetic conditions. Autoimmune blistering disease (AIBD), a rare form of neonatal blistering, is associated with autoreactive immunoglobulins that target skin. Because the immune system of neonates is not fully developed, it produces insufficient immunoglobulins.¹ Therefore, most cases of AIBD in newborns are associated with maternal autoreactive antibodies.² In this Brief Report, we describe a case of neonatal linear immunoglobulin A (IgA) bullous dermatosis (LABD), an autoimmune blistering disease caused by autoreactive IgA antibodies that attack the basement membrane zone (BMZ). Although LABD occurs most often in adults older than 40 years, it can occur in children³ as chronic bullous disease of childhood, but rarely occurs in neonates.² In neonatal cases, blisters and erosions typically develop a few days after birth. To our knowledge, to date, all reported cases of neonatal LABD were from healthy mothers who had no symptoms.⁴ In this study, we demonstrated the presence of IgA autoantibodies targeting the BMZ in the asymptomatic mother’s breast milk but not her serum.

Methods

A healthy male born without complications at 36 weeks of gestation to a healthy mother developed multiple tense bullae on the neck, buttocks, and hands on day 4 after birth. Blisters soon spread to the oral mucosa, and excessive exudate from the oral erosion led to wheezing and difficulty breathing. Additional bullae and erosions were observed around the mouth, neck, perianal area, hands, and feet (Figure 1A-C). Fibroscopic findings revealed that the erosions extended to the hypopharynx and larynx (Figure 1D).

Figure 1. Clinical Presentation and Course of Neonatal Linear IgA Bullous Dermatosis.

A-C, Clinical presentation on hospital admission (day 11 after birth). Tense bullae accompanied by erythema were seen on the face, neck, and buttocks. Bullae and erosions were prominent in the oral mucosa. D, Nasal fibroscopic findings obtained on day 12 after birth show erosions spread widely down to the hypopharynx and larynx, with prominent edema of the epiglottitis. The airway was almost occluded by the mucosal edema and exudate. E, Clinical course from birth until discharge at 7 months after birth.

IV denotes intravenous, and PICU, pediatric intensive care unit.

The patient then developed acute respiratory distress syndrome and was treated with steroid and betastimulant inhalants to support the respiratory system, as well as antibiotics for aspiration pneumonia. Breastfeeding was stopped and feeding was performed parenterally. The patient was transferred to the pediatric intensive care unit, which included intensive respiratory care with intubation and tracheotomy during several months.

Histopathologic and immunofluorescence studies of serum and skin samples from the patient and his mother and of breast milk were performed. No new blister formation was observed after breastfeeding cessation (Figure 1E). The lesions gradually healed, and the exudate subsequently decreased. The patient was discharged after 6 months and no relapse has been reported (>1 year).

The study followed the case series reporting guideline. Diagnostic testing of specimens collected from the patient and mother was approved by the Institutional Ethics Committee of the Keio University School of Medicine. Written informed consent was obtained from the mother. Additional details on the methods can be found in the eMethods of the Supplement.

Results

Results of a hematoxylin-eosin–stained skin biopsy from the tense bullae on the upper chest revealed a subepidermal blister with neutrophilic, lymphocytic, and eosinophilic infiltration (Figure 2A). Findings of electron microscopy detected no BMZ element defects, ruling out congenital epidermolysis bullosa. Results of direct immunofluorescence microscopy depicted linear deposits of IgA and C3 along the BMZ (Figure 2B). Based on these results, we made a final diagnosis of neonatal LABD.

Figure 2. Results of Histologic and Immunofluorescence Studies.

A, Hematoxylin-eosin staining of the skin specimen from the patient’s bullae showed subepidermal blister with dense inflammatory cell infiltration in the base. B, DIF test of the patient’s skin specimen revealed linear deposition of IgA on the BMZ (arrowheads). C, Results of SS-IFF of the patient’s serum was positive, showing a linear deposit of IgA on the dermal side of the BMZ (arrowheads). D, Results of SS-IFF of the maternal breast milk were positive, showing linear deposits on the dermal side of the BMZ (arrowheads). E, Anti-J chain antibody staining to perilesional skin of the neonate’s skin paraffin section revealed deposition of J chain along BMZ. F, No J-chain deposition was observed in a case of adult LABD.

BMZ denotes basement membrane zone; DIF, direct immunofluorescence; IgA, immunoglobulin A; LABD, linear IgA bullous dermatosis; and SS-IIF, split-skin indirect immunofluorescence.

A split-skin indirect immunofluorescence (SS-IIF) test (using normal human skin after treatment with 1-mol/L sodium chloride) of the neonate’s serum showed linear deposition of IgA on the dermal side of the BMZ (Figure 2C). An SS-IIF test of the mother’s serum yielded a negative result, but breast milk from approximately 1 week after birth was positive for IgA on the dermal side of the BMZ (Figure 2D). Results of SS-IIF tests of breast milk from 3 other healthy mothers were negative. Immunohistochemical staining proved the deposition of secretory IgA in the neonatal skin by demonstrating the presence of J chain along the BMZ in the perilesional skin (Figure 2E)—not seen in other LABD cases (Figure 2F). Based on these findings, we concluded that the neonatal blisters and erosions were caused by secretory IgA in the maternal breast milk. Results of type VII collagen enzyme-linked immunosorbent assay of the neonate’s serum and mother’s breast milk were negative. Further identification of the responsible antigen was hampered by a lack of sufficient serum and breast milk samples.

Discussion

The incidence of neonatal AIBDs is unknown but is believed to be lower among newborns than among adults. Based on a systematic review of neonatal AIBD cases reported worldwide from 1946 to 2014,² only 51 original cases of neonatal AIBDs had been identified. The types of AIBD include pemphigus vulgaris, pemphigus foliaceus, and bullous pemphigoid, all which are caused by the transfer of IgG autoantibodies in maternal serum through the placenta. However, LABD is distinctive because it is mediated by IgA, which rarely crosses the placenta. None of the mothers of 11 previously reported neonates with LABD had any symptoms, unlike those with IgG-mediated AIBDs. Serum test results in 4 of these neonates with LABD were negative for autoantibodies.^4,5,6,7 We confirmed that the mother in the present case was asymptomatic, and the results of serum testing by IIF were negative; however, results of IIF testing of the maternal breast milk were positive. Therefore, we suspect autoreactive IgA transfer via intestinal absorption of breast milk. The results and review of previous reports (cases confirmed by IgA deposition by direct immunofluorescence and analyzed serum antibodies in patients) are summarized in the Table,^5,6,7,8,9,10 but none of these reports had tested the maternal breast milk for IgA autoantibodies. Interestingly, all cases of SS-IIF testing of the patient’s serum were stained on the dermal side, suggesting the presence of a common target antigen(s) in neonatal LABD.

Table. Summary of Neonatal LABD Cases for Which Serum Analysis Was Performed.

Sourcea	Patient					Mother
	Onset, d	Mucosal involvement	Respiratory failure	In vivo IgA deposition (DIF)	Circulating serum IgA (SS-IIF)	Circulating serum IgA (ss-IIF)	Secreted IgA in breast milk (SS-IIF)
Hruza et al5	2	Positive	Positive	Positive	Negative	Negative	NP
Gluth et al8	7	Positive	Positive	Positive	Positive dermal side	NP	NP
Kishida et al9	1	Positive	Positive	Positive	Positive dermal side	NP	NP
Romani et al6	3	Positive	Positive	Positive	Negative	Negative	NP
Diociaiuti et al10	5	Positive	Positive	Positive	Negative	NP	NP
Giraud et al7	2	Negative	Negative	Positive	Negative	Negative	NP
Raiber et al4	14	Positive	Positive	Positive	Negative	Negative	NP
Present case	4	Positive	Positive	Positive	Positive dermal side	Negative	Positive dermal side

Abbreviations: DIF, direct immunofluorescence; IgA, immunoglobulin A; LABD, linear IgA bullous dermatosis; NP, not performed; SS-IIF, split-skin indirect immunofluorescence.

^aTo our knowledge, all previous neonatal LABD cases are listed in the Supplement.

In this present case, the presence of J-chain–confirmed secretory IgA deposition in the neonatal skin indicates that the autoantibodies producing blisters were derived from maternal breast milk, while conventional LABD is caused by J chain-negative serum IgA.¹⁰ The mucosal surface is coated by secretory IgA, and the striking aerodigestive involvement in neonatal LABD may be associated with the direct secretory IgA access to the sites, in addition to the circulating secretory IgA.

Almost all reported neonatal LABD cases (listed in the eReferences of the Supplement) were in newborns who were healthy at birth and then developed multiple bullae after a few days (Table). Most cases required intubation to address severe respiratory failure. However, the disease durations were short, with symptoms resolving within a few weeks or months, except in 2 newborns who died of respiratory distress.^11,12

Such a short disease course supports our hypothesis that autoreactive IgA is transferred to newborns through maternal breast milk. Because secretory IgA in breast milk originates in the mother’s intestinal immune system, it is assumed to reflect the mother’s intestinal microflora during pregnancy.¹ The epitope specificity of breast milk secretory IgA significantly differed from that of the maternal serum in an antibody repertoire study with overlapping peptides.¹³ Therefore, IgA in breast milk is newly created during late pregnancy by a mechanism that differs from that for serum IgA, making the 2 types of IgA completely discrete.

Human breast milk contains large amounts of immunoglobulins, 90% of which are subclasses of IgA and whose primary role is to protect against gastrointestinal infections in newborns. The amount of secretory IgA in breast milk varies widely among individuals, and is highest in early colostrum, then gradually decreases during the next 6 months. Mickleson and Moriarty¹⁴ reported a mean colostral IgA value of 3200 mg/dL (range, 150 ~ 8370 mg/dL; to convert to g/L, multiply by 0.01). Considering the feeding volume of the early newborn (150 ~ 500 mL/d), total IgA intake is 5000 mg to 16 000 mg per day. The mean (SD) blood IgA concentration in neonates is 6.7 (0.4) mg/dL,¹⁵ and the total IgA in the blood, calculated from the circulating blood volume (approximately 10% of total body weight), is approximately 200 mg. Therefore, it is surmised that the IgA in maternal breast milk may be sufficient to cause disease in the newborn.

Limitations

A limitation of this study is that it is an observation of single case of neonatal LABD. The accumulation of similar cases with testing of maternal breast milk is necessary.

Conclusions

Almost all cases of neonatal LABD present with severe mucosal involvement that progresses to respiratory compromise requiring aggressive therapy. Based on these study findings, maternal breast milk can transmit pathogenic IgA related with autoimmune disease from mother to child. Therefore, despite its low incidence, the severe course of neonatal LABD portends that clinicians know of this potential cause and ensure that breastfeeding is discontinued if LABD is suspected.

Supplement.

eMethods

eReferences