Linear immunoglobulin A bullous dermatosis associated with herpes simplex virus infection and Kawasaki disease


Masahiro Hayashi, Ph.D., Division of Dermatology, Yamagata Prefectural Shinjo Hospital, Wakaba-Cho 12-55, Shinjo 996-0025, Japan. Email:

Dear Editor,

Linear immunoglobulin (Ig)A bullous dermatosis (LAD) is an autoimmune subepidermal bullous disease that involves IgA anti-basement membrane antibody and was suggested as an entity by Chorzelski and colleagues in 1979.1 Its clinical characteristics are vesicles distributed annularly around erythema and/or sparsely distributed tense bullae. Direct immunofluorescence (DIF) shows linear deposition of IgA at the basement membrane zone. The pathogenesis of LAD remains unknown. LAD may occur following varicella zoster virus infection, may accompany malignancy and may be drug induced.2,3 Kawasaki disease (KD) is an acute febrile eruptive disease that was described in 1967 by Kawasaki and preferentially affects younger children.4 It is characterized by various kinds of skin rash and systemic vasculitis. The pathogenesis of this disease has been suggested to be a viral infection, such as Epstein–Barr virus or herpes simplex virus (HSV), toxic shock syndrome toxin-1 (TSST-1) and endotoxin from Gram-negative cocci; however, it remains unknown.5 HSV infection could be causative for both KD and LAD, but their coexistence has not been reported. Recently, Rowley and coworkers6,7 reported that IgA plasma cell infiltration is observed at the involved vascular wall in KD patients, and such infiltrates are also seen in unaffected regions, such as the upper respiratory tract, pancreas and kidney. These are interesting findings when considering the coexistence of LAD and KD.

A 14-month-old boy was admitted to our division for pruritic erythematous plaques and blisters scattered across his body. His past medical history was not remarkable and neither did he have a history of atopic dermatitis, chickenpox and HSV infection. At first, he presented to the Division of Pediatrics due to fever of unknown origin, and he was admitted the next day. When cervical lymphadenopathy, reddening of the lips, skin rash, conjunctivitis and indurative edema of extremities appeared, he was diagnosed with KD 3 days after the onset of symptoms. i.v. immunoglobulin (IVIG) and urinastatin were administrated, and his condition including skin rash gradually improved. Acetylsalicylic acid was also administrated after 11 days to prevent cardiovascular complications. After 16 days, pruritic erythematous plaques and blisters appeared in the inguinal area, scrotum, perineum, face and neck (Fig. 1). Mucosal involvement was not seen. Blister was tense and large bullae without delle. Viral infection or autoimmune blister disease was suspected, and viral investigation was carried out. Blister fluid from a trunk blister was collected aseptically by inserting a sterile needle into the blister and aspirating the fluid. Spinal fluid was collected by a conventional technique. HSV DNA was detected in the spinal fluid taken after 9 and 20 days and also in the blister fluid after 20 days with the method described by Aurelius and colleages.8 The serum HSV IgG titer was elevated (Table 1). There was no suggestive symptom of herpetic meningitis. Administration of i.v. acyclovir was not effective. The patient was referred to our division for further examination and treatment.

Figure 1.

 Pruritic erythematous plaques and blisters are seen on the (a) face, neck, (b) inguinal area, scrotum, and perineum.

Table 1.   Serum viral titer
Days after KD onset0720
  1. Blister appeared 16 days after onset. The virus titer that comes off from the reference range is indicated by the bold. HSV, herpes simplex virus; Ig, immunoglobulin; KD, Kawasaki disease.

HSV IgM (<0.80)
HSV IgG (<2.0)<2.0≥128.055

Blood tests showed increased white blood cell count (20 980/μL) and C-reactive protein of 0.52 mg/dL (normal <0.24). Liver function, renal function and serum Ig were all within normal limits. Biopsy from a blister on his back showed subepidermal blister formation and mild inflammatory infiltration that consisted of lymphocytes, neutrophils and partially of eosinophils in the upper dermis (Fig. 2). An intraepidermal neutrophilic microabscess was also observed. Vasculitis, balloon cell or intranuclear inclusion body were not seen. Tzanck test was not performed. DIF revealed linear deposition of IgA at the basement membrane zone (Fig. 3a). Indirect immunofluorescence of 1 mol/L NaCl-split human skin demonstrated that the epidermal side was positive for IgA antibodies at a titer of 1/40 (Fig. 3b). Immunoblotting studies identified that the patient’s serum reacted with the 120-kDa antigen (linear IgA bullous dermatosis antigen, LAD-1) (Fig. 4). Finally we diagnosed this case as LAD, lamina lucida type. Acetylsalicylic acid was discontinued due to suspected drug-induced LAD; however, the symptoms did not improve. Oral prednisolone (PSL) 10 mg/day and 4,4-diamino-diphenyl-sulfone (DDS) 10 mg/day were administrated, and blisters and erosions rapidly epithelialized. Then, both were gradually tapered, PSL to 2.5 mg/day and DDS to 2.5 mg/day, and the boy showed persistent high grade fever for more than 5 days, cervical lymphadenopathy, reddening of the lips and tongue, skin rash and conjunctivitis. He was diagnosed as KD again. PSL and DDS were discontinued immediately. IVIG was effective for KD as well as at first onset. Acetylsalicylic acid was also administrated again and resurgence of LAD was not seen. To date, 32 months after KD settled, the patient has been well, and there have been no additional symptoms of KD and LAD.

Figure 2.

 Subepidermal blister formation and mild inflammatory infiltration consisting of lymphocytes and partially of eosinophils in the upper dermis (hematoxylin–eosin stain, original magnification ×200).

Figure 3.

 Direct immunofluorescence revealed linear deposition of immunoglobulin (Ig)A at the basement membrane zone (a). Indirect immunofluorescence on 1 mol/L NaCl-split human skin demonstrated that IgA antibodies were positive for the epidermal side at a titer of 1/40 (b).

Figure 4.

 Immunoblotting studies identified that the patient’s serum reacted with the 120-kDa antigen. LAD, linear immunoglobulin A bullous dermatosis.

The trigger of LAD is unknown. Post-chicken pox and drug-induced LAD have been reported.2 In our case, HSV DNA was detected in both the spinal fluid and the blister fluid, and the HSV IgG titer was elevated, suggesting systemic HSV infection. Blister was clinically considered due to LAD but not HSV because they were tense and large bullae without delle. In addition, we could not find the unique histological features of herpes simplex from skin biopsy, hence there is a possibility that detection of HSV DNA from the blister fluid may have been by contamination of systemic infection of HSV. Although serum HSV IgG titer may also have been influenced by the administration of IVIG, we considered that HSV infection actually existed because HSV DNA was detected from spinal fluid. To our knowledge, this is the first case of LAD following HSV infection. Acetaminophen-induced LAD has been reported.3 We ruled out this case from drug-induced LAD, because urinastatin was discontinued before the onset of LAD and symptoms did not recur when IVIG and acetylsalicylic acid were reinstituted.

Kawasaki disease involves systemic vasculitis, especially in the coronary arteries, resulting in aneurysm formation that is often fatal.4 Although multiple infectious agents and toxins including HSV have been considered as the implication of the etiology of KD and many studies were investigated, none have been identified so far.5 Various kinds of skin rash (so-called undefined skin rash) may be observed in patients with KD, even pustule and chicken pox-like eruptions.9,10 LAD-like eruptions in KD have not been described so far. Recently, IgA plasma cell infiltration has been observed at the vascular wall,6 and also in the upper respiratory tract, pancreas and kidney.7 These findings suggest that an IgA immune response may spread to various organs, even in patients without vasculitis, by entry of the etiological agent through the upper respiratory tract.7,11,12 Other investigators have reported that patients with KD show increasing serum levels of IgA class antibodies to several antigens.13,14 In our case, there is the possibility that antecedent KD and HSV infection may have triggered the onset of LAD, which involves anti-basement membrane IgA antibody. Because the onset of KD was earlier than the appearance of HSV infection, we considered that KD and HSV were independent events. LAD did not relapse upon the resurgence of KD. Our hypothesis is that in this patient, LAD was triggered by two components, KD and HSV infection. We believe that this is a unique and interesting case to consider the etiology of LAD.


We are grateful to Mr. Katsumi Mizuta, Department of Microbiology, Yamagata Prefectural Institute of Public Health for the genetic analyses and Dr. Norito Ishii and Dr. Takashi Hashimoto, Department of Dermatology, Kurume University for the serum immunoblotting analysis.