Pediatric Hypoglycemia Workup

Hyperinsulinism

Persistent hyperinsulinemic hypoglycemia of infancy can be focal or diffuse. Routine abdominal ultrasonography, computed tomography (CT) scanning, and MRI are of little use in distinguishing between the forms. Positron emission tomography (PET) scanning with [¹⁸ F] dihydroxyphenylalanine (DOPA) has been shown to effectively distinguish focal from diffuse disease. This study is easier to perform than invasive radiologic techniques such as transhepatic venous sampling or intra-arterial calcium stimulation with hepatic venous sampling. PET scanning is also helpful in locating insulin-producing tumors in older children with acquired hyperinsulinism, which is rare.

Hypopituitarism

Perform head MRI to identify pituitary or hypothalamic neoplasms or congenital abnormalities.

The presence of hypoglycemia should prompt a thorough investigation for counterregulation abnormalities or a lack of alternative substrate. Plasma glucose concentrations should be measured in all neonates and children with the symptoms listed above (see History), with due consideration given to the temporal relationships of the test samples.^[6]

The exact glucose level that constitutes hypoglycemia is debatable, particularly in neonates. Older literature suggests levels of more than 1.7 mmol/L are acceptable in this age group. Newer publications suggest levels of less than 2.5 mmol/L are inappropriate. The Whipple triad is used to support a diagnosis of hypoglycemia and its symptomatic consequences. The triad consists of (1) the presence of symptoms likely or known to be caused by hypoglycemia, (2) a low plasma glucose concentration when symptoms are present, and (3) subsequent relief of symptoms when the hypoglycemia is corrected.

The plasma glucose concentration should ideally be measured with a laboratory-based glucose analyzer. If this is unavailable, home blood-glucose monitors may be used; however, their accuracy in the low range is questionable, and they have been shown to provide false-positive and false-negative results.

Screening for hypoglycemia in the asymptomatic neonate is controversial. Studies suggest that screening is appropriate in infants of mothers who are diabetic, infants who are large or small for their gestational age, and infants who are premature. Screening should begin within the first 2-3 hours of life and continue through the first 24 hours of life.

The ability to properly sort through the differential diagnoses of hypoglycemia depends on obtaining the critical sample at the time of hypoglycemia. This sample is used to measure the various metabolic precursors and hormones involved in glucose counterregulation, including glucose, insulin, growth hormone, cortisol, lactate, pyruvate, beta-hydroxybutyrate, free fatty acid, carnitine, branched-chain amino acid, and insulinlike growth factor-binding protein-1 (IGFB-1) levels. (A urine sample for organic acid analysis is also critical.)

If the critical-sample measurements are not available at the time of initial presentation, the hypoglycemia must be reproduced. This is usually achieved using a closely monitored fast. This fast should be conducted in a center that can respond quickly and appropriately if significant hypoglycemic consequences develop. When the plasma glucose concentration falls to less than 2.5 mmol/L, the fast has ended, and the critical sample is drawn. The maximum length of the fast depends on the age of the child. Conservative recommendations for maximum lengths of fasting are as follows:

• Younger than age 6 months - 8 hours
• Aged 6-8 months - 12 hours
• Aged 8-12 months - 16 hours
• Aged 1-2 years - 18 hours
• Aged 2-7 years - 20 hours
• Older than age 7 years - 24 hours

Interpretation of the critical sample

Metabolically, plasma free fatty acid levels should increase to more than 0.5 mmol/L, and beta-hydroxybutyrate levels should increase to more than 1 mmol/L to provide alternative fuel. A failure of both to increase suggests hyperinsulinemic lipolytic suppression. An increase in free fatty acid levels to more than 3 mmol/L without an increase in beta-hydroxybutyrate levels suggests a defect in fatty acid metabolism. (See the diagram below.)

High plasma lactate levels suggest gluconeogenesis, glycolysis, or respiratory-chain defects. Plasma insulin levels should be suppressed, and cortisol levels should be increased (>550 nmol/L [20 mcg/dL]). Growth hormone levels should also be increased (>6 mcg/L).

Some authors have suggested measuring free carnitine, total carnitine, and acyl carnitine levels before performing a fasting study in order to detect medium-chain acyl-CoA dehydrogenase deficiency; this may prevent life-threatening hypoglycemia and hyperammonemia during the fast. Many states now conduct neonatal screening for medium-chain acyl-CoA dehydrogenase deficiency.

Measuring IGFBP-1 before and after the fast may also be useful. IGFBP-1 levels are suppressed by insulin and, therefore, increase during fasting in healthy individuals but decrease or remain stable in individuals who are hyperinsulinemic.

A glucagon stimulation test at the end of the fast may be useful as well. In most individuals, the glucose level does not increase following hypoglycemia because the glycogen stores are significantly depleted before hypoglycemia develops. However, in hyperinsulinemia, endogenous glucagon secretion and glycogenolysis are suppressed, and the plasma glucose concentration increases more than 1.9 mmol/L (35 mg/dL) following glucagon administration. Glucagon does not increase the blood glucose concentration in patients with glycogen-storage disease type I even in the fed state. Cortisol and growth hormone levels can also be drawn 30 and 60 minutes into the test to determine if levels rise following hypoglycemia.

After the fast is completed and the patient has been fed, glucose and lactate levels should be measured for evidence of glycogen synthase deficiency.

Measuring sulfonylurea, ethanol, or salicylate levels is appropriate if hypoglycemia is believed to be secondary to their ingestion. The presence of a low C-peptide level with a high insulin level suggests exogenous insulin administration.

Oral glucose tolerance tests do not aid in the diagnosis of hypoglycemia, because many healthy patients have low plasma glucose concentrations following a large glucose bolus. In addition, a low plasma glucose concentration during an oral glucose tolerance test does not prove that the patient is hypoglycemic when symptoms occur.

Commercially available genetic analysis is now available to identify many of the genetic disorders associated with hyperinsulinism.