Fulminant type 1 diabetes is characterized by abrupt onset of loss of pancreas β-cell function with low initial HbA1c levels and absence of autoantibodies to islet antigens (1). Initial histological examinations revealed the absence of insulitis, and it was thought that the autoimmune process did not mediate this type of diabetes. However, with an accumulation of cases, possible involvement of an immune response to islet antigens has come to light. CD8+ lymphocytes infiltrating the pancreatic islets were seen in a patient who died of fulminant type 1 diabetes (2), and peripheral GAD-reactive interferon-γ-producing CD4+ lymphocytes were detected in another case of fulminant type 1 diabetes (3). Thus, T-cell-mediated autoimmunity may be involved. HLA analyses also indicate that an autoimmune process similar to that in autoimmune type 1 diabetes may be involved in the development of fulminant type 1 diabetes. In his editorial comment regarding the first report of fulminant type 1 diabetes, Lernmark (4) noted that in Japanese patients most had HLA class II antigens that confer type 1 diabetes. Furthermore, Tanaka et al. (5) recently reported that fulminant type 1 diabetes is associated with specific HLA class II haplotypes, which are also associated with autoimmune type 1 diabetes in Japanese patients. Most patients reported to have fulminant type 1 diabetes were Japanese. Occurrences in other ethnicities are rare (6,7), reflecting possible immunogenetic differences. We report herein a Filipino patient with fulminant type 1 diabetes and the results of HLA analysis.
A 32-year-old Filipino woman married to a Japanese man and living in Gunma, Japan, had a low-grade fever 2 weeks before admission and visited a general practitioner, but no specific diagnosis was made. Thirst, polydipsia, polyuria, and general malaise had developed. Then abdominal pain and vomiting occurred, and she visited a clinic where severe hyperglycemia was found. She was referred to Shiroyama Hospital and was admitted for diabetic ketoacidosis. On admission blood glucose was 800 mg/dl and HbA1c was 6.0%. Urinary ketone bodies were positive. Arterial blood pH was 7.11, serum amylase was 57 (IU/l), lipase was 32 (units/l), and elastase I was 344 ng/dl (all within normal range). Anti-GAD antibodies and islet cell antibodies were negative. After metabolic derangement was corrected, insulin secretion was evaluated. The fasting serum C-peptide concentration was 0.1 ng/ml; it was 0.2 ng/ml after glucagon injection. Urinary C-peptide was 1.6 μg/day. The severe abrupt-onset insulin deficiency, low HbA1c, and lack of antibodies to islet antigens were compatible with fulminant type 1 diabetes. DNA typing of HLA antigens showed homozygosity for the DRB1*0405-DQB1*05031 haplotype, which is unique to the Filipino population (8). She was also HLA A24-positive.
Most Japanese fulminant type 1 diabetes patients studied carried at least one of two haplotypes, DQA1*0303-DQB1*0401 and DQA1*0302-DQB1*0303 (5), which are known to confer susceptibility to autoimmune type 1 diabetes in the Japanese population. Only 2 of 22 did not have either haplotype. Furthermore, one-third of the patients were homozygous for the DQA1*0303-DQB1*0401 haplotype. Our Filipino patient had neither haplotype. However, she had two DRB1*0405 alleles. DRB1*0405 is a risk allele for type 1 diabetes in the Japanese population. The DQA1*0303-DQB1*0401 haplotype is in close linkage disequilibrium with DRB1*0405 in Japanese and Chinese individuals. These alleles form an extended haplotype DRB1*0405-DQA1*0303-DQB1*0401. Thus it is difficult to determine whether the DQ molecule, the DR molecule, or the haplotype is responsible for the susceptibility to fulminant type 1 diabetes. In Filipino autoimmune type 1 diabetes, the risk haplotypes are DRB1*0405-DQB1*0302 and DRB1*0405-DQB1*0201, followed by DRB1*0405-DQB1*0401 and DRB1*0405-DQB1*0402 (8). The DRB1*0405-DQB1*05031 haplotype is neutral to autoimmune type 1 diabetes. Our finding of a homozygous DRB1*0405 allele without a risk haplotype in a Filipino patient suggests that this allele rather than a haplotype is associated with fulminant type 1 diabetes. Continued worldwide collection of data and HLA analysis will further clarify the involvement of immunogenetics in fulminant type 1 diabetes.
