Sulfonylurea Treatment in Young Children With Neonatal Diabetes: Dealing with hyperglycemia, hypoglycemia, and sick days

Recently, heterozygous activating mutations in the genes forming the ATP-sensitive K⁺ channel (K_ATP channel), KCNJ11 and ABCC8, have been shown to cause neonatal diabetes (1–4). Sulfonylurea treatment restores insulin secretion in these patients (3,5,6), but information on the practical management of children with mutated K_ATP channels taking this medication is limited.

We report clinical aspects of the successful transfer to oral treatment in three cases of young children with KCNJ11 and ABCC8 mutations (Table 1). All parents gave written consent.

In case 1, a girl was transferred from insulin to glibenclamide at 17 months (7) and had been on this treatment for 2 years. During this period, blood glucose testing decreased from 5–6 to 2–3 tests/day. As blood glucose levels were not affected by the ingestion of different amounts of carbohydrates, a free diet was initiated. Unexplained episodes of hyperglycemia were occasionally observed, and an appropriate decrease in the blood glucose level was observed with the usual dose of glibenclamide, even in cases of hyperglycemia (blood glucose 350 mg/dl). When the parents missed one dose, blood glucose was 455 mg/dl without ketosis, which was treated at home with lispro insulin dose and administration of the missed sulfonylurea dose. Only one episode of symptomatic hypoglycemia (30 mg/dl) occurred and was successfully treated with fruit juice and a temporary decrease in glibenclamide.

Minor episodes of viral respiratory disease were managed by decreasing the sulfonylurea dose to avoid hypoglycemia. One episode of rotavirus diarrhea was managed in a hospital using insulin and by stopping glibenclamide. Upon discharge, the sulfonylurea was restarted at the previous dose. Ketones were not detected on any of these acute illnesses.

In case 2, a boy with a KCNJ11 mutation was successfully transferred from insulin to glibenclamide at 38 months. This patient also had some episodes of unexpected hyperglycemia, which responded to taking the normal glibenclamide dose. An episode of a febrile upper respiratory tract viral illness was managed with a decrease of the glibenclamide dose, and ketones were not detected and insulin was not required.

In case 3, a girl was treated with insulin until the confirmation of a novel mutation in ABCC8 when aged 3 years. Unexpectedly, a low dose of glibenclamide (0.1 mg · kg⁻¹ · day⁻¹) not only allowed the stopping of insulin but also resulted in episodes of asymptomatic hypoglycemia. The dose was reduced and then discontinued completely for 12 days. However, as hyperglycemia recurred tolbutamide was started, which resulted in good control without hypoglycemia.

These cases show that the use of sulfonylureas in children with K_ATP mutations differ from adults with type 2 diabetes. In two cases, glibenclamide was best given three times a day. These children also required a higher nighttime dose to lower morning glucose, possibly as sulfonylureas act through facilitating the response to incretins in this type of diabetes (5). Sulfonylurea treatment was well tolerated; however, the risk of hypo- and hyperglycemia persists, and education in their prevention and treatment should be given. Hyperglycemia, even at blood glucose level 350 mg/dl, responded to the usual dose of sulfonylureas, but if these patients consistently miss medications they risk ketoacidosis.