Puzzler of the month
GEORGE K. SIBERRY, MD, MPH, SECTION EDITOR
Her visit to you is for a two-month well-child exam, but what you find in this infant girl is anything but routine. You're struck, during your inspection, by her decreased muscle tone. A look at her and her mother's prenatal and birth records remind you that this baby has gathered an extensive history to herself in a short time.
She was born at 36 1/2 weeks' gestation to a 24-year-old, gravida 1 woman whose prenatal laboratory tests showed that she was blood type A-negative; group-B strep-negative; RPR-nonreactive; rubella-immune; and hepatitis-B surface antigen-negative. The pregnancy was complicated by gestational diabetes that required insulin. The baby was delivered by emergency cesarean section after her mother came to the prenatal clinic with hypertension and a complaint of seeing spots in front of her eyes, and decreased fetal movements were noted on ultrasonography (US).
Your patient was vigorous at birth, you recall, with Apgar scores of 8 at one minute and 9 at five minutes. Birth weight was 3.67 kg (>90th percentile); length, 52 cm (also >90th percentile). The physical exam at birth was unremarkable except for relative microcephaly (occipitofrontal circumference, 32 cm [25th percentile]).
Laboratory studies obtained at birth included a complete blood count, C-reactive protein (CRP) (to rule out sepsis as a potential cause of maternal and fetal distress), and glucose. The initial glucose was 49 mg/dL, but a follow-up test, after breastfeeding, returned as 20 mg/dL. The baby received fluids and dextrose intravenously, and hypoglycemia resolved over the next few hours; the presumption was that her hypoglycemia was a result of her mother's diabetes. The hemoglobin level, hematocrit, and white blood cell count were within normal limits but the platelet count was low at 39 x 103/µL. Then, the CRP level rose rapidlyfrom 0.2 mg/dL initially to 1.9 mg/dL on the third day of life. The thrombocytopenia and elevated CRP raised enough concern for sepsis that she was started on ampicillin and gentamicin.
Over the next four days, the newborn's CRP level normalized. She remained thrombocytopenic, however, with the platelet count reaching a nadir of 29 x 103/µL.
The relative microcephaly led to a urine cytomegalovirus test and US of the head, both normal. She received a platelet transfusion on the fifth day of life, with normal platelet counts afterward. The balance of her hospitalization was uneventful; she completed the seven-day course of antibiotics and was discharged home.
Perusing the record, you note that the baby's physical exam at 2 weeks of age was remarkable for firm "bumps" underneath the skin of her upper arms and back. At that time, her mother told you that these nodules appeared soon after discharge. At about 1 month of age, she was seen by a pediatric dermatologist who concluded that, although these could be nodules of sclerema neonatorum, they most likely represented subcutaneous fat necrosis (SFN): an uncommon but widely recognized disorder of unclear etiology in which skin lesions develop, often over bony prominences and extremities.1 At a follow-up visit with the dermatologist just before the baby's current visit with you, he expressed certainty that the nodules were those of SFN and judged them to be improving "slightly."
Today, you find that, in addition to hypotonia, your patient is still microcephalic (occipitofrontal circumference below the curve for age) and her height and weight have fallen to the third percentile for age. Her mother reports that the baby drinks Enfamil with iron, four ounces every four hours; has frequent stools and wet diapers; and is not taking any other food or nutritional supplement.
You palpate multiple erythematous, blanching, well-circumscribed subcutaneous nodules on the infant's upper chest, shoulders, and back. You agree with the dermatologist's diagnosis and his assessment that the size of the nodules is decreasing. Because of her history of thrombocytopenia at birth, however, you order a follow-up CBC.
When results return, you discover that the thrombocytopenia has been replaced by thrombocytosis: the platelet count is now 754 x 103/µL; in addition, the WBC count is 31.9 x 103/µL and the differential count shows 44% neutrophils, 47% lymphocytes, 4% monocytes, 2% bands, 2% eosinophils, and 1% atypical lymphocytes. The hematocrit is normal. Time to refer, you decide, to a hematologist-oncologist for further evaluation of the thrombocytosis and leukocytosis.
Your consultant examines the girl and obtains a repeat CBC and a metabolic profile and liver function tests. Several results are notable: calcium, 15.3 mg/dL; blood urea nitrogen, 31 mg/dL; creatinine, 0.9 mg/dL; WBC count, 27 x 103/µL; and platelet count, 703 x 103/µL. Tests of liver function are normal. She doesn't believe that the child has a primary hematologic problem right now; rather, she thinks that the leukocytosis and thrombocytosis may be a response to the metabolic stress of ongoing hypercalcemia. When you find out about the calcium level, you decide to admit the child for management and workup of an alarmingly high level.
Hypercalcemia in childhood is rare, you recall. Among adults, hyperparathyroidism and cancer of the breast, lung, kidney, head, and neck are the most common disorders that cause hypercalcemia, but these entities are uncommon in children.2 You also recall that the differential diagnosis for hypercalcemia in an infant and child is quite broad:
While contemplating this wide-ranging differential, you act on the high calcium level by obtaining an electrocardiogram, hydrating the baby with IV fluids, and giving furosemide to increase calcium excretion. The ECG is reassuring: no shortened QT interval, no heart block, no other arrhythmia associated with hypercalcemia. Then it's time to draw blood, obtain urine, and call for consults from both endocrinology and nephrology. You also change the patient's formula to a low-calcium, vitamin D-free product (Calcilo XD, Ross).
Lab results begin to become available. The parathyroid hormone (PTH) level is undetectable, which seems to eliminate hyperparathyroidism as the culprit. Of interest, her PTH-related protein level (PTH-rp) is slightly elevated at 2.4 pmol/L (normal, <2 pmol/L), but because no clinical findings support hypercalcemia of malignancy, you are comfortable eliminating that possibility. You use the urine calcium:creatinine ratio, rather than a 24-hour urine collection, to assess calcium excretion quickly; the ratio is elevated at 1.8, which helps eliminate the rare but benign condition of familial hypocalciuric hypercalcemia. Excess vitamin D administration can cause hypercalcemia; your patient's 1,25-dihydroxycholecalciferol is, however, normal. Levels of thyroid-stimulating hormone and of free T4 are also normal, eliminating hyperthyroidism from the list. Because a grade II/VI systolic murmur was heard on admission in the setting of hypercalcemia and hypercalciuria, you consider Williams syndrome; an echocardiogram reveals a structurally normal heart but fluorescent in-situ hybridization (FISH) analysis is ordered nonetheless to rule out this diagnosis.
As you narrow the differential, issues of concern arise. The child's initial phosphorus level was normal, but the calcium-phosphorus product is 70; a value higher than 60 can be associated with calcifications in the body. Given the elevated calcium-phosphorus product and elevated creatinine level, you order a renal US scan, which reveals extensive bilateral medullary nephrocalcinosis.
Meanwhile, your patient's calcium level is behaving oddly: At first, it falls to 13.6 mg/dL (from the 15.3 mg/dL seen on admission) with hydration and furosemide, but then it rises over the following two daysto 14.2 and then to 15.0 mg/dL. You are reluctant to stop the furosemide because of its effect on calcium excretion, but you are also somewhat nervous giving the drug because it can exacerbate nephrocalcinosis.
Because hydration isn't having the desired effect, the consulting endocrinologist recommends pamidronate to lower the calcium level. This drug, used more often in adults, is a bisphosphonate that has dual activity: it inhibits osteoclast function and diminishes osteoclast viability.3 After IV administration of pamidronate over the next two days, the calcium level declines steadilyto 10.2 mg/dLbringing the important calcium-phosphorus product down to 30, out of range of concern. The leukocytosis and thrombocytosis, although slower to respond, soon normalize as wellas does the PTH-rp level.
You have now narrowed the differential considerably with testing, and medication has succeeded in bringing the calcium level out of the danger range. But is there a unifying diagnosis left to uncover that explains the overall presentation and the clinical course?
Once again, you review the entire case, birth to current admission. Your patient was brought to you two weeks ago with subcutaneous fat necrosis, first described clinically by Cruse in 1815,4 when it was called "scleroderma of the newborn." The disorder was first characterized histopathologically in the literature in 1907 as nonsuppurative skin nodules on the cheeks, forearm, legs, and buttocks.5
Today, the most important diagnoses to consider in a newborn with subcutaneous firm nodules are SFN and sclerema neonatorum. The two can be differentiated by findings on the history and physical, with sclerema neonatorum having a much higher mortality rate (75%) than SFN (15%, in association with hypercalcemia).
An association between SFN and hypercalcemia was first reported in 1956 by Clay; 19 such cases were found when the world literature was reviewed up to 1990.6 Since that review, 29 more cases of SFN have been reported; in 17 of them, hypercalcemia was detected.7
The pathogenesis of SFN has been ascribed to several possible mechanisms, including:
But where does the hypercalcemia arise? Development of hypercalcemia in SFN has been related to increased vitamin D sensitivity, the parathyroid hormone level, prostaglandin E2 activity, and calcium release from necrotic adipose tissue. The most likely explanation? Unregulated production of a relatively high level of 1,25- dihydroxycholecalciferol by macrophages participating in the granulomatous inflammatory process associated with SFN. Hypothetically, a high level of this vitamin D metabolite increases intestinal absorption of calcium, bone mineralization, and soft tissue calcification.5 Of interest, this patient's 1,25 dihydroxycholecalciferol level is at the upper limit of normal (90 pg/mL).
The neonatal history revealed a hypoglycemic, thrombocytopenic babyearly clues that your patient had been under significant perinatal distress, and signs that are usually associated with sepsis. Once subcutaneous fat nodules appeared a few weeks later and the clinical picture of failure to thrive was documented, the picture becomes clearer: You're dealing with a relatively common complication (hypercalcemia) of a rare disorder (SFN).8
The infant was treated as described and discharged. SFN resolved completely by 3 months of age. Her calcium level remained normal. At 5 months of age, the low-calcium, vitamin D-free formula was stopped; Enfamil with iron was reintroduced without recurrence of hypercalcemia. At 8 months of age, she was at the 25th percentile for height, weight, and occipitofrontal circumference and was meeting developmental milestones.
You may never see a case of SFN. But this child's course is a reminder that, although thrombocytopenia and hypoglycemia in the neonatal period are often caused by sepsis, they may represent a response to significant fetal distress of another cause. Fetal distress can be associated with damage to major organs, and this patient reminds us that skin is a major organ.
If you do have occasion to note SFN when examining a distressed infant after the neonatal period, consider secondary hypercalcemia and check that calcium level! And always keep an eye on the road when an infant with violaceous, indurated nodules fails to thrive!
1. Thomsen RJ: Subcutaneous fat necrosis of the newborn and idiopathic hypercalcemia. Arch Dermatol 1980;118:1155
2. Bilezikian JP: Management of acute hypercalcemia. N Engl J Med 1992;326:1196
3. Carano A, Teitelbaum SL, Konsek JD, et al: Bisphosphonates directly inhibit the bone resorption activity of isolated avian osteoclasts, in vitro. J Clin Invest 1990;85:456
4. Cruse P: Ein fall von sclerodermie (Sogenannt Sclerodermis Adulotorum bei Saighing, St. Petersburg). Zeitschrift 1875:5:306
5. Fabyan M: Disseminated sclerema neonatorum occurring in an infant with other lesions. Bull Johns Hopkins Hosp 1907;18:349
6. Hicks MJ, Levy ML, Alexander J, et al: Subcutaneous fat necrosis of the newborn and hypercalcemia: Case report and review of the literature. Ped Derm 1993;10:271
7. Burden AD, Krafchik BR: Subcutaneous fat necrosis of the newborn: A review of 11 cases. Ped Derm 1999; 16:384
8. Rice AM, Rivkees SA: Etidronate therapy for hypercalcemia in subcutaneous fat necrosis of the newborn. J Ped 1999;134:349
George Siberry, ed. Brian Youth, Todd Green. Pediatric Puzzler: Hypotonia.