A Case of Fulminant Type 1 Diabetes With Strong Evidence of Autoimmunity

We have recently reported (1) that T-cell autoimmunity may be involved in so-called “fulminant” type 1 diabetes (characterized by diabetic ketoacidosis [DKA], low HbA_1c level at onset, insulin deficiency, and negative islet-associated autoantibodies), which was originally proposed as a novel subtype of type 1B diabetes (2). In our previous report (1), we found a high serum level of interferon-inducible protein-10 (IP-10), an important chemokine inducing migration of activated T-cells to local lesions (3), and GAD-reactive CD4+ cells in the periphery of a patient, even though no islet-associated antibody was detected.

Here, we report another case of fulminant type 1 diabetes, a 48-year-old man who was proven to have not only a high serum IP-10 level and GAD-reactive CD4+ cells in the periphery but also a high anti-GAD antibody level 1 year after the onset of diabetes, even though no islet-associated autoantibody was detected at onset. The patient developed fatigue, fever (38.5–39.5°C), and abdominal discomfort 2 weeks before being seen at our hospital. Because he subsequently developed thirst, polyuria, and weight loss (8-kg reduction in 10 days), he presented at our hospital. Based on the presence of DKA (blood glucose level of 728 mg/dl, marked ketonuria [4+], pH 7.282), relatively low HbA_1c level (7.3%) at onset, absence of GAD 65 antibody (detection limit <0.4 units/ml; 100% sensitivity and 100% specificity of the assay in the GAD antibody proficiency test [Immunology of Diabetes Workshop], lab ID no. 305), IA-2 antibody (detection limit <0.75 units/ml; M. Powell, S. Chen, H. Tanaka, M. Masuda, C. Beer, B. Rees Smith, J. Farmaniak, unpublished observations), islet cell antibody and insulin autoantibody, low serum C-peptide level (0.4 ng/ml at 6 min after intravenous injection of 1 mg glucagon) and 24-h urine C-peptide level (<3.0 μg/day), he was diagnosed as having fulminant type 1 diabetes, and intensive insulin therapy was started (total 50 units/day at discharge).

Regarding HLA typing, A24, which is considered to be related to total β-cell destruction (4), was detected (other HLA types: A26, B54, B60, Cw1, Cw4, and DR4). Moreover, a high level of serum IP-10 (285 pg/ml, mean 38.2 pg/ml in healthy subjects) was observed, and GAD-reactive γ-interferon–producing CD4+ cells were detected in peripheral blood (10 of 50,000 CD4+ cells). Thereafter, the titer of serum GAD 65 antibody was followed, and a significant increase was found (1.7 units/ml at 1 month, 39.4 units/ml at 12 months, and 48.9 units/ml at 18 months after the onset of diabetes), indicating that autoimmunity was definitely involved. The HbA_1c level of this patient is now controlled at 7.2% by 49 units/day insulin, although his serum C-peptide level is below the detectable limit (<0.2 ng/ml).

Although it has previously been reported that ∼15% of cases of “classical” type 1 diabetes without islet-associated autoantibody had become positive for islet-associated autoantibody at 1 year after the onset of diabetes (5), it is unknown whether cases of so-called “fulminant” type 1 diabetes, which was originally reported as nonautoimmune type (no islet-associated autoantibody at onset of diabetes) (2), also later become positive for islet-associated autoantibody. This case of fulminant type 1 diabetes showed not only cellular immunity against islets but also clear “seroconversion” of GAD 65 antibody during the disease course. Therefore, we propose that fulminant type 1 diabetes should not be diagnosed as idiopathic type simply as a result of islet-associated autoantibody negativity at onset. Careful periodic measurement of islet-associated autoantibodies such as GAD 65 autoantibody should be performed in fulminant type 1 diabetes as well as assessment of cellular immunity against islet-associated antigen for proper classification of type 1 diabetes.