Recently, mutations in the gene KCNJ11 encoding the Kir 6.2 subunit of the ATP-sensitive K+ channel (KATP channel) have been described in patients with permanent neonatal diabetes (1). The KATP channel complex is an aggregate of four subunits of the Kir6.2 inward rectifier channel plus four regulatory units known as the sulfonylurea receptor (SUR1). In pancreatic β-cells, glucose metabolism leads to a rapid rise in intracellular ATP levels, leading to closure of the KATP channel. The resultant cell depolarization is critical for normal insulin secretion. Sulfonylureas are able to close the KATP channel by interacting with SUR1 via an ATP-independent mechanism. As a result, four young patients carrying mutations in the Kir6.2 channel have been treated with sulfonylureas (2,3) and withdrawn from insulin.
To test whether Kir6.2 mutations are present in Chile and to evaluate the response to sulfonylurea treatment, we tested five Chilean patients with diabetes diagnosed before age 6 months (range 0.25-4.1), requiring insulin treatment since diagnosis. The coding region and intron-exon boundaries of the KCNJ11 gene were sequenced as previously described (1).
A female patient had a novel heterozygous Kir6.2 mutation, R201L (CGT>CTT). She was underweight at birth (2,120 g at 37 weeks’ gestation) and presented with severe ketoacidosis at the age of 4.1 months (pH 6.9, plasma glucose 790 mg/dl). Her postnatal development was normal, without seizures or hypotonia. Her parents, who have normal oral glucose tolerance tests, do not carry the mutation.
At the age of 17 months, the sulfonylurea glibenclamide was slowly introduced. Initially, a once-daily dose of 0.1 mg · kg−1 · day−1 was given for a week and then changed to a twice-daily dose that increased weekly by 0.1 mg · kg−1 · day−1, allowing a simultaneous decrease in insulin doses. After 8 weeks with the glibenclamide dose of 0.8 mg · kg−1 · day−1, the insulin was stopped completely. HbA1c levels at the beginning and end of the transition period were 7.3 and 5.9%, and mean patient blood glucose readings dropped from a mean 196 to 155 mg/dl. The patient exhibited mild transient diarrhea (three to six episodes per day of soft stools) when the glibenclamide started and every time the dose was increased up to a dose of 0.6 mg · kg−1 · day−1. Each episode lasted for 4 or 5 days. After a month off insulin, the patient remained in stable metabolic control with no recurrence of gastrointestinal problems. Optimal glycemic control was achieved by giving the glibenclamide in three equal doses every 8 h.
Insulin concentration did not increase during an intravenous glucose tolerance test (0.3g glucose/kg) performed before glibenclamide was begun. After the 0.8 mg · kg−1 · day−1 dose was reached and insulin completely withdrawn, the test was repeated and insulin levels increased by 28 pmol/l.
In conclusion, we have described the first reported patient with a spontaneous mutation, R201L, in the Kir6.2 gene. This novel mutation affects a highly conserved arginine at position R201, which has been shown to be key for ATP binding (1). As seen in the previously reported R201 mutations (R201H and R201C), our patient did not have neurological abnormalities. This child is the fifth reported Kir6.2 patient to be able to discontinue insulin therapy and improve control, although a very high dose (0.8 mg/kg) was needed (equivalent to 60 mg in a 75-kg adult). We report the first case of diarrhea associated with sulfonylurea therapy in a patient with a Kir6.2 mutation, probably related to the action of glibenclamide on inwardly rectifying K+ channels in the human ileal mucosa (4). This unusual side effect is thought to be dose related and may be relatively common in patients with Kir6.2 mutations who require high doses of sulfonylureas. In our patient, this side effect was transitory, and given the potential benefits of sulfonylurea therapy, we recommend that diarrhea should not immediately lead to the discontinuation of treatment.