We read with interest the recent letter by Yu and Howard (1) describing recurrent episodes of hepatomegaly and pronounced elevation of liver enzymes (aspartate aminotransferase [AST] and alanine aminotransferase [ALT]) in type 1 diabetic patients with poor metabolic control and treated with high daily doses of insulin (1.3–2.2 units · kg−1 · day−1), considering the extra doses are to correct frequent hyperglycemia and/or diabetic ketoacidosis.
Here we describe similar experiences in our patients affected by autoimmune and nonautoimmune diabetes in the pediatric age range. In particular, we have observed a child with diabetes onset at the age of 66 days that was probably not of autoimmune origin (2) because he did not present with type 1 diabetes–susceptible non–Asp/Arg HLA DQ heterodimers (3) and showed the absence of islet cell antibodies, GAD, insulin autoantibodies, and insulinoma-associated protein 2. These last dosages were carried out on frozen serum because they were not all available at the time of diagnosis.
At age 3 years, he presented with the first episode of hepatomegaly and elevated liver enzymes. Other liver function tests, such as alkaline phosphatase, protrombin/partial prothombin time, and total bilirubin, were normal. Two other episodes were reported in the following 2 years. Although his metabolic control was acceptable on all three occasions, the insulin dose was maintained at a high level for the best possible HbA1c (Table 1). This dose was almost twice the usual dose used at this age and is comparable to the 2.2 units · kg−1 · day−1 described by Yu and Howard.
Hepatic biopsy during the second episode showed abundant glycogen deposits in the hepatocytes and normal mitochondria. On all three occasions, continuous low-dose intravenous insulin normalized the liver enzymes in just a few days.
Now at age 15 years, aside from diabetes, he is in good health. During adolescence the maximum dosage of insulin was 1 unit · kg−1 · day−1, which was lower than the dose described by Yu and Howard, yet we found no other episodes of liver damage.
In three other patients with type 1 autoimmune diabetes, we observed a pattern of liver dysfunction during puberty similar to Yu and Howard. In all patients, diabetes was very poorly controlled despite high insulin dosage (Table 1). Liver size and enzymes in these three patients also returned to normal after intravenous insulin. No liver biopsy was performed in any of these cases.
On the basis of our experience, we agree with Yu and Howard when they indicate the pathogenesis of this liver disease in chronic overtreatment with insulin and self-induced hyperinsulinism rather than low metabolic control or diabetes etiology.