(265) Diazoxide Toxicity Leading to Variable Hyperglycemic Crisis States in a Child

Introduction: Diazoxide is a standard treatment for hyperinsulinism. However, in excess, it may cause a hyperosmolar hyperglycemic state (HHS) or diabetic ketoacidosis (DKA). Here we describe a child with both presentations, although drug levels were only known after the fact.

Description: An 8-year-old female with chromosome 7p glucokinase duplication leading to hyperinsulinism and failure to thrive presented with severe hyperglycemia. One week prior, her gastric tube was replaced, after which she had frequent emesis and poor enteral intake. She received water and apple juice. She developed increased thirst and wet diapers with subjective weight loss. Per family, finger stick blood glucoses (BG) were normal.

In the emergency department, she had depressed mental status and hypotension. She received 80 ml/kg fluids, started continuous fluids at twice maintenance rate, and began a norepinephrine infusion. Initial labs were notable for BG 2,105 mg/dL, venous blood gas pH 7.24, serum bicarbonate 20 mmol/L, serum osmolality 403 mOsm/kg, and trace ketones.

In the Pediatric Intensive Care Unit (PICU), she was tired with GCS 8. After 3 ml/kg of 3% saline, GCS improved to 10. Diazoxide was held. She continued twice maintenance fluids, titrating down as BG improved. Mental status returned to baseline. Diazoxide level 2 days after admission was 13000 ng/mL (ref 1000 ng/mL). Home diazoxide restarted day 3.

She returned to the PICU from the floor one week later with acutely worsening hyperglycemia. Labs showed BG 743, venous blood gas pH 7.46 with bicarbonate 36 mmol/L, and small ketones. She received 40 ml/kg saline and twice maintenance fluids. Four hours later, she had BG >600 mg/dL, bicarbonate 13 mmol/L, and anion gap 24 mmol/L. She transitioned to DKA management, with an insulin infusion and 1.5 times maintenance fluids titrated per unit protocol. Diazoxide level was 97000 ng/mL.

Discussion: We describe two hyperglycemic presentations resulting from diazoxide toxicity. HHS and DKA due to diazoxide toxicity are rarely described in the literature. Unfortunately, diazoxide levels are often delayed. Thus, in a patient with chronic diazoxide use presenting in a hyperglycemic crisis, whether HHS or DKA, diazoxide toxicity should be suspected and clinicians must be ready to treat this complex hyperglycemic state.