PEDIATRIC PUZZLER

GEORGE K. SIBERRY, MD, MPH, SECTION EDITOR

Emesis in an ill-appearing boy:
When one plus one is three

By Matt Grady, MD, Kevin C. Osterhoudt, MD, and Neil Caplin, MBBS

On a busy winter night in the emergency department, your next patient is a 7-year-old African-American boy who, like a number of children before him this night, has been brought in with a principal concern of vomiting. His illness started yesterday in a typical fashion, his mother reports: fever, diminished appetite, and some nonspecific abdominal pain. The subsequent course, however, became anything but typical.

His mother's concern for the boy heightened when he told her he was having chest pain and a stiff neck. He also had two episodes of green emesis earlier this evening, and had little to eat or drink all day. No history of diarrhea, rash, wheezing, or upper respiratory symptoms is reported. The boy's medical history is notable for mild asthma, without recent albuterol use. There is no history of surgery or hospitalization. He lives with his mother and has had no ill contacts, she reports. Immunizations are up to date and development has been normal. He is not taking any medications.

The mother decided that her son needed emergency medical evaluation after he would not get out of bed all day. The transporting emergency medical technicians (called because the mother does not have a car) noted that oxygen saturation, measured by pulse oximetry, was 88% and blood dextrose, 45 mg/dL. He was given nebulized albuterol and 25 mL of 50% dextrose intravenously en route to the ED.

As you approach the patient, who is lying in a darkened room, you are immediately impressed by his ill appearance. The chart lists the following vital signs: temperature, 38.4° C; heart rate, 124/min; blood pressure, 102/55 mm Hg; respirations, 24/min; oxygen saturation by pulse oximetry, 100% (breathing room air); and weight, 28 kg. He complains when you turn on the room lights and when you shine a penlight in his eye. His equally sized pupils react, but he will not tolerate a funduscopic examination.

His lips are dry and the mucous membranes have a tacky appearance. The pharynx is normal. He cries and resists when you attempt to flex his neck. There is no lymphadenopathy. Lungs are clear. Except for tachycardia, the boy's cardiac exam is unremarkable. Palpation of the abdomen reveals epigastric tenderness without rebound tenderness, involuntary guarding, or organomegaly.

Inspection of the skin reveals multiple hyperpigmented macules, which, his mother explains, are old varicella scars. No petechiae are found. The boy is quiet and passive, with symmetric strength and intact cranial nerves.

So this child with vomiting is, after all, remarkably different from the score of children with viral gastroenteritis whom you have seen! Instead, the constellation of fever, meningismus, photophobia, and vomiting in an ill-appearing child offers a textbook presentation of meningitis. You figure that the hypoglycemia is likely secondary to the suspected bacterial infection, compounded by poor intake. You order laboratory tests, an IV bolus of normal saline, and antibiotics, and request a lumbar puncture kit.

No sweet surprise without more sugar

As you leave the child's room, the hospital laboratory returns "panic" values: Glucose is 47 mg/dL and bicarbonate, 7 mEq/L! You had already checked dextrose at the bedside; finding that it measured 56 mg/dL, you give a 5 mL/kg of body weight bolus of 10% dextrose in water and start an infusion of 5% dextrose in 0.5% normal saline. You review the remainder of the laboratory studies: sodium, 131 mEq/L; potassium, 5.4 mEq/L, with slight hemolysis; chloride, 102 mEq/L; blood urea nitrogen, 16 mg/dL; and creatinine, 0.5 mg/dL. Calculated anion gap is increased at 22. Calcium, phosphate, and alkaline phosphatase levels are within the normal range. The white blood cell count is only 7.8 x 10³/µL, with a normal differential count and no band forms. Hemoglobin is 13.9 g/dL; platelets, 292 x 10³/µL. The prothrombin and partial thromboplastin times are slightly prolonged and the D-dimer is weakly positive.

As for that "textbook" diagnosis of meningitis you had entertained—it appears to be incorrect. A bit of surprise furrows your brow when the microscopic cerebrospinal fluid examination is normal: 3 WBC and 2 RBC per high-powered field; glucose, 49 mg/dL; and protein, 22 g/dL. You need to go back, add up the clues again, and arrive at a new summation.

Vomiting, hypoglycemia, and an anion-gap metabolic acidosis deserve careful reconsideration. Pancreatitis, liver disease, intestinal obstruction with mesenteric ischemia, pneumonia, drug toxicity, and a metabolic disorder all cross your mind. You order a chest radiograph and an abdominal obstruction series, but the films are unrevealing. Stool tests negative for occult blood.

A venous blood gas demonstrating a pH of 7.14, PaCO₂ of 38 mm Hg, and base deficit of 15.7 mmol/L—consistent with uncompensated metabolic acidosis—further defines the extent of acidemia. A fatty oxidation defect in the face of intercurrent illness can cause acidosis; in such a case, lack of ketone production in a fasting state is often the initial clue. A bedside urine dipstick shows a specific gravity of 1.030 with large ketones—but no blood, protein, or nitrates. The large ketones eliminate a fatty oxidation defect from your differential diagnosis but, taking into account the dehydration, you may now have an explanation for the acidosis.

Certain metabolic disorders elevate the serum lactate concentration, you recall, but lactate is a normal 1.2 mmol/L. The abnormal electrolytes are retested, and sodium remains low at 131 mEq/L, potassium is 4.6 mEq/L, chloride is 105 mEq/L, and bicarbonate has improved to 13 mEq/L. The anion gap acidosis is improving with IV fluids, glucose, and antibiotics, but, being a well-read physician (see the Pediatric Puzzler in the July 2001 issue), you wonder whether the low sodium might warrant more careful consideration than first given. Hepatic transaminase, amylase, and lipase levels are normal, which excludes hepatitis or pancreatitis. After dextrose supplementation, serum glucose at last measures a respectable 147 mg/dL.

One hour later, however, you are notified that the boy is hypotensive and that the glucose level by fingerstick measurement has fallen to 57 mg/dL despite the infusion of 5% dextrose in saline. Ten percent dextrose in water and a second normal saline bolus are ordered and you review the common causes of hypoglycemia (see the table.). Toxic ingestion is unlikely by history, but you confirm that the child did not get into any alcohol or aspirin and that there are no b blockers or sulfonylurea medications in the home. A serious amino acidemia should have manifested before 7 years of age, but a mild metabolic defect unmasked by concurrent illness remains a plausible explanation. Hyperinsulinemic states, such as an insulinoma, are rare but not unheard of in this age group. He has not been given steroids recently, eliminating the possibility of secondary adrenal suppression. The low BP and that low sodium level are consistent with Addison disease, but other clues—elevated potassium level, skin tanning—are absent.

Perusing the list again, a mild metabolic defect or adrenal insufficiency seem most likely; how convenient that the initial treatment for both is glucose and fluids! You decide to order a test of adrenocorticotropic hormone (ACTH) and a morning cortisol level to check adrenal function. The metabolic work up includes levels of serum amino acids and urine organic acids, and you also order viral studies to identify the cause of the precipitating illness. Insulin and growth hormone levels are sent for to complete the evaluation of hypoglycemia.

The boy is admitted to the pediatric intensive care unit to receive IV fluids and glucose and to have his BP monitored. No steroids are administered. Sodium and glucose levels normalize within 12 hours. Antigen testing of a nasopharyngeal swab is positive for influenza B virus. The morning cortisol level is low at 8 µg/dL (normal range, 10–25 µg/dL)—suspicious, yes, but not diagnostic of adrenal insufficiency. An adrenocorticotropic stimulation test is performed and, because he looks well, the boy is transferred from the pediatric ICU to a general pediatric ward. But that evening, BP again drops—to 91/33 mm Hg. He is now given hydrocortisone, and BP normalizes.

Adding up the son's clues

The next morning, definitive laboratory results are available. Insulin and growth hormone levels are normal, but the ACTH stimulation test demonstrates little response (7.2 µg/dL and 7.57 µg/dL at 30 and 60 minutes, respectively, after infusion). The measured ACTH is 1,285 pg/mL (normal for his age, <28 pg/mL), confirming a diagnosis of primary adrenal insufficiency. The stress of the influenza B infection likely unmasked his adrenal insufficiency and accounted for much of his presenting display of symptoms. You now have a diagnosis and a starting point in the search for an underlying cause of the boy's condition.

Addison disease, described by Thomas Addison in 1885,¹ was first observed as a consequence of adrenal destruction by tuberculosis. Destruction of the adrenal gland causes deficiency of both mineralocorticoid and glucocorticoid hormones. Nonspecific symptoms include weakness, anorexia, weight loss, vomiting, and postural hypotension. Salt craving and abdominal pain that mimics an acute abdomen are less common. Deficiency of aldosterone, the primary mineralocorticoid, causes renal sodium wasting, hyperkalemia, and bicarbonate loss, in turn prompting metabolic acidosis. A compensatory increase in secretion of antidiuretic hormone (ADH) helps maintain intravascular volume but contributes to the hyponatremia. The serum renin level is usually elevated. A low level of cortisol, the primary glucocorticoid, results in hypoglycemia, low BP, and malaise, especially during a time of stress such as surgery or, as in this case, infection.²

Mucocutaneous hyperpigmentation is a classic physical finding in Addison disease. In response to a low cortisol level, the pituitary increases production of ACTH and melanocyte-stimulating hormone. Increased melanin deposits can give the skin a tan appearance. In darker skinned persons, such tanning is often difficult to appreciate. A subtler finding is hyperpigmentation at sites of friction, such as the buccal mucosa and flexor creases of the palms and soles. Notably, old scars may become more prominent. This patient was not noticeably hyperpigmented, but the fact that scars of an old varicella infection were pronounced offered a small clue to his underlying disease.

Today, autoimmune disease accounts for most cases of primary adrenal insufficiency. In infants, congenital adrenal hyperplasia and, among boys, adrenal hypoplasia congenita and continuous gene deletion syndrome constitute the most common diagnoses. After 2 years of age, autoimmune adrenalitis accounts for 70% to 90% of cases of Addison disease, as an isolated condition or as part of autoimmune polyglandular syndrome (APS) types 1 or 2. APS-1 includes Addison disease, acquired hypoparathyroidism, and chronic mucocutaneous candidiasis. Fungal infection of the mouth and, later, the nails with Candida albicans is often the presenting symptom; neither was evident on physical examination of this patient. APS-2 is defined as autoimmune adrenalitis, autoimmune thyroid disease, and insulin-dependent diabetes. Associated autoimmune disorders, including vitiligo, alopecia, hepatitis, pernicious anemia, and celiac disease may be present, but are not defining features of either syndrome.

In this child, subsequent investigation was undertaken, as is necessary, to determine the underlying cause of the adrenal insufficiency: A test of adrenal antibodies was negative, thyroid studies were normal, and serum levels of very long-chain fatty acids were markedly elevated. This laboratory picture led to an ultimate, and unfortunate, diagnosis of adrenal leukodystrophy, an X-linked genetic disorder that affects one of every 20,000 males and involves impaired degradation of very long-chain fatty acids in the peroxisomes. Demyelination of the central nervous system and adrenocortical deficiency develop over time. Multiple mutations of the peroxisomal membrane protein have been described. There is some clinical variability of expression, but most patients experience a progressive degenerating neurologic process, with death a few years after neurologic presentation. Neurologic symptoms of dementia and deteriorating hearing, speech, vision, and gait often lag years behind declining adrenal function. Approximately 3% of female carriers develop neurologic symptoms. In boys younger than 15 years, serum testing for very long-chain fatty acids is recommended, especially if a test of adrenal antibody is negative, as was the case here.

Therapy for adrenal insufficiency is steroid replacement. Daily corticosteroid replacement is started with hydrocortisone, at 10–15 mg/m² of body surface area. During time of stress, the dosage is generally tripled. Mineralocorticoid, in the form of daily fludrocortisone acetate, is started at 0.05–0.3 mg/d, titrated to keep the serum renin level normal.

Third time's a charm

As is often the case, new diagnostic impressions must be formulated after the findings of an initial evaluation surprise us. This is the third puzzling case of adrenal hormone dysregulation presented in Pediatric Puzzler since the July 2001 issue (see also the May 2002 issue), each with distinguishing features. Clinicians often play the role of medical detective: The differential diagnosis becomes a list of suspects, and clinical clues are considered until the diagnosis reveals itself. Even though functional adrenal insufficiency is uncommon, this malady should be rounded up as one of the usual suspects when you are faced with a patient with hyponatremia, hypoglycemia, and a septic appearance. Then, those clues may all add up!

REFERENCES

1. Addison T: On the constitutional and local effects of disease of the supra-renal capsules. Highly, London, England, 1855

2. Ten S, New M, Maclaren N: Clinical Review 130: Addison's Disease 2001. J Clin Endocrinol and Metab 2001; 86:2909

3. Berman R, Kliegman R, Jenson H (eds): Nelson Textbook of Pediatrics, ed 16. Philadelphia, Pa., WB Saunders, 2000, pp 725–1728

DR. GRADY is a pediatric hospitalist at Christiana Hospital, Newark, Del.

DR. OSTERHOUDT is an emergency department attending physician in the department of pediatrics, The Children's Hospital of Philadelphia and the University of Pennsylvania School of Medicine, Philadelphia.

DR. CAPLIN is a pediatric endocrine fellow at the Children's Hospital of Philadelphia.

DR. SIBERRY is a fellow in pediatric infectious disease at The Johns Hopkins Hospital, Baltimore, Md.

Hypoglycemia: A list of suspects

Adrenal insufficiency

Adrenal suppression
Hypopituitarism
Primary adrenal failure

Hepatic failure

Hyperinsulinemic states

Injected insulin
Insulinoma

Malnutrition

Metabolic defects

Amino acidemias
Fatty acid oxidation defects
Organic acidurias
Other enzyme deficiencies

Sepsis

Starvation

Toxic Ingestion

ß-adrenergic antagonists (propranolol)
Ethanol
Salicylates
Sulfonylurea oral hypoglycemic agents

Adapted from Berman R et al (2000)³

Some of our most satisfying moments in pediatrics—and some of our best clinical pearls—result from solving perplexing cases. If you have a favorite, send it to

The Editors Contemporary Pediatrics
5 Paragon Drive
Montvale, NJ 07645

If it is used, you will be cited as the contributor.

 

George Siberry, ed. Kevin Osterhoudt, Matt Grady, Neil Caplin. Pediatric Puzzler: Emesis. Contemporary Pediatrics 2003;2:23.