Dysregulation of calcium homeostasis after severe burn injury in children: Possible role of magnesium depletion

of calcium homeostasis after severe :hildren: Possible role of magnesium Gordon L. K~inj MD, Marc Nicolai, P~, Craig B. Langman, MD, Bettina E Cuneo, MD, Dawn E. Sailo; Ms, and David N. Herndon, ~¢D

Objective: To determine the cause and extent of hypocalcemia observed in children after severe burns.

Design: We

in general and calcium regulation in particular.

studied 10 children with burns covering 57% * 17% (SD) body surface

area, ages 9.6 _+4.7 years, who were admitted consecutively during a 6-month period. Diet supplied a minimum of 2.7 grrdm2 of calcium, 0.3 gm/m 2 of magnesium, and 2.2 grrdm2 phosphate. Blood specimens were obtained daily for l0 e 5 days for the followhag tests: (1) simultaneous analysis for ionized calcium, magnesium, and intact parathyroid hormone (group A); (2) two of these children, randomly selected, had serial 2hour determinations on a single day (group B); (3) a modified EHsworth-Howard test, consisting of a 10-minute infusion of synthetic parathyroid hormone 18 + 10 days postburn and associated changes in urinary cyclic adenosine monophosphate excretion and renal threshold phosphate concentration (group C). Three of these children, when normomaguesemic, also received a standard magnesium infusion to determine magnesium retention (group D). Data were analyzed with chi-square, regression analysis, and nonparametric testing as appropriate. Continued on page 247

Burn injury exceeding 40% total body surface area leads to retardation of growth velocity, 1 reduced bone formation, 2'3 and a prolonged reduction in bone mineral density with a consequent increased risk for fracture and lower peak bone mass in children.4

The pathogenesis of these abnormalities is unclear. However, we have observed frequent and prolonged hypocalcemia in such patients, raising th e question whether the abnormalities seen in hone are a reflection of wider disturbances in mineral homeostasis

From the Departments of Pediatrics and Surgery, University of Texas Medical Branch and ShrilTers Burns Institute, Galveston, Texas; the Division of Nephrology and Mineral Metabolism, Children'sMemorial Hospital and Northwestern UniversityMedical Schoo~ Chicago;and the Department of Pediatr~, Rush Medical College, Chicago,Illinois.

Supported in part by grants from Shriners Hospitalsfor CrippledChildren No. 8870 (G.L.K.)and National Institutes of Health grant No. RR00048 (C.B.L.), and Children's Memorial Institute for Education and Research (C.B.L.). Presented in part at the annual meetingof the Societyfor Pediatric Research,Washington, DC, May 6-10, 1996. Submittedfor publicationMay 31, 1996; acceptedNov. 5, 1996. Reprint requests: Gordon L. Klein, MD, Pediatric GastroenterologyDivision, Children's Hospital, Room 3.240B, Universityof TexasMedicalBranch, Galveston,TX 77555-0352. J Pediatr 1997;131:246-51 Copyright© 1997by Mosby-YearBook, Inc. 0022-3476/97/$5.00 + 0 9/21/79176

246

We hypothesized that secondary hyperparathyroidism would develop in burn patients because of hypocaleemia. Accordingly, the aim of this study was to examine specific aspects of calcium regulation, including the parathyroid hormone response to hypocaleemia, as well as the renal and skeletal response to infusion of synthetic PTH, to determine whether seeondary hypoparathyroidism or P T H resistance develop after burn injury.

METHODS We prospectively studied 10 consecutive children (8 boys) with burn injuries of 57% _+ 17% total body surface area (range 30% to 82%). The children, aged 9.6 +_4.7 (SD) years (range 4 to 17years), were admitted to the Shriners Burns Institute in Galveston, Texas, during a 6month period between August 1995 and F e b r u a r y 1996 (Table I). In standard fashion, all patients underwent early wound excision and grafting procedures on the average of once per week. Within 24 hours of admission, all received enteral feedings of milk supplying a minimum intake of calcium, 2.7 gm/m2; magnesium, 0.3 gm/mY; and phosphate, 2.2 gm/mY.4

KLEIN ET AL.

THE JOURNAL OF PEDIATRICS V o l u m e 13 I, N u m b e r 2

After informed consent was obtained, patients underwent blood and urine sampling for one or more of the following determinations between 1 and 3 A54 daily for 10 ± 5 days (range 4 to 17 days) (Table II), and appropriately timed urine collections were analyzed for serum ionized calcium, magnesium, and intact P T H concentrations (group A). In two randomly selected patients these determinations were performed at 2-hour intervals during an 8-hour period on a single day of study (group B). To test the renal and skeletal response to PTH, a modified Ellsworth-Howard test was performed s 18 ± 10 clays after the burn injury (range 7 to 33 days). The test consisted of four timed half-hour urine collections with one interspersed blood specimen before and two half-hour urine collections with two interspersed blood specimens after a 10-minute infusion of synthetic P T H 1-34 (teriparatide [Parathar, Rorer]) at a dose of 3 U/kg in children younger than 12 years of age 6 or 5 U/kg up to a maximum of 200 units in adolescents, s'6 Total duration of the study was 90 minutes before and 60 minutes after P T H administration for a total of 2.5 hours. We measured changes in urinary cyclic adenosine monophosphate excretion and renal threshold phosphate concentration, calculated according to the nomogram of Walton and Bijvoet, z as well as serum osteocalcin and bone-specific alkaline phosphatase, biochemical markers of bone formation, and interleukin-6 (group C). Finally, to more rigorously examine the possibility of magnesium depletion as a potential contributor to parathyroid abnormalities in these patients, three of the children who became normomagnesemic underwent magnesium infusion of 2.4 mg elemental iVig/kg as magnesium sulfate in 50 ml of a 5% dextrose-water solution administered intravenously over 4 hours (group D). A spot urine collection was obtained before the infusion for determination of magnesium and creatinine levels, and a 24-hour urine collection was initiated simultaneously with the magnesium infusion. Magnesium retention was calculated according to the formula:

Results: A l l patients showed sustained hypocalcemia and hypomagnesemia; intact parathyroid hormone response was inappropriately low and response to synthetic parathyroid hormone infusion was blunted. Lowest ionized calcium levels were associated with hypomagnesemia.

Conclusion: Hypopoa'athyroidism and blunted renal response to parathyroid hormone suggest that magnesium depletion may contribute to their pathogenesis. Magnesium repletion and monitoring are recommended. (J Pediatr 1997;131:246-51)

_

_

E 54-

5

3-

~.,~.,

•

2i

0.3

I

I

i

|

0.6

I

I

i

I

|

0.9

I

.j.P.; !

!

•

I

I

i

1.2

I

,

I

I

'

'

'

1.5

Ionized Ca (mmol/L) F ~ . 1. Values for iCa and iPTH in children with severe burns. Mean and 99% confidence interval lines are shown for a group of normal childreng; n = 41 observations for 7 children.

Table I.

Patient characteristics and studies

247

KLEIN ETAL.

Table 11.

Study group A: Serum ionized calcium, parathyroid hormone, and magnesium

%Mg retention = 1-[24-hr urine Mg (Spot urine Mg/Creatinine × 24-hr creatinine)]/[Total elemental Nlg infused] When this magnesium infusion protocol is used, magnesium depletion is considered definite when the urinary magnesium retention exceeds 50%, and is considered probable when urinary magnesium retention is between 25% and 50%. 8 Serum concentrations of iPTH were measured by a specific and well-characterized immunoradiometric assay (Incstar, Sfillwater, Minn.) to detect the intact 1-84 molecu]e; results have been validated over a wide range of PTH values. 9 The relationship between blood iCa and iPTH concentrations has been established in children with atrial septal defects after acute manipulations of blood iCa induced by carcliopulmonary bypass. There were no apparent abnormalities of calcium metabolism in this groupl and a nomogram was constructed that denotes the mean values and the 99% confidence intervals. 9 Serum concentrations of osteocalcin were measured by a noncompetitive binding enzyme-linked immunosorbent assay (Metra Biosystems, Mountain View, Calif.). 10 Bone-speclfic alkaline phosphatase concentrations were measured by

248

THE JOURNALOF PEDIATRICS AUGUST 1997

enzyme immunoassay (Metra Biosystems); serum levels of interleukin-6 were measured by a hlgh-sensitivity enzyme immunoassay2; urinary cyclic adenosine monophosphate was measured by radioimmunoassay.11 Phosphorus, magnesium, and creatinine clearances were measured in serum and urine by automated methods (Single Multiehannel Analyzer [Technicon] phosphomolybdate method or the CXS Autoanalyzer, Beckman Instruments, Brea, Calif.). Ionized calcium, uncorrected to p H 7.4, was measured in serum with an ion-specific electrode (NOVA STAT 5 Analyzer, NOVA Biomedical, Waltham, Mass.). This study was reviewed and approved by the institutional review board at the University of Texas Medical Branch and by Shriners Hospitals for Crippled Children, Tampa, Florida.

RESULTS Serum iCa and Magnesium

Concentrations Hypocalcemia was detected in all nine patients studied in groups A and B. Fortynine of 70 determinations (70%)were

below the lower limits of normal. Hypomagnesemia was also detected in all patients and in 54 (770/0) of the 70 determinations (Table II). None of the patients studied had clinical symptoms including tetany, seizures, or cardiac arrhythmias. In three of the nine children, serum pH values were also determined. Mean pH values were within normal range and there was no relationship between serum iCa levels and pH in may of the patients studied.

Parathyroid Response to

Hypocalcemia Serum PTH concentrations were below the lower limits of normal in 64 (91%) of the 70 specimens and 45 (92%) of the 49 low concentrations of iCa in the serum. Low serum PTH levels were associated with hypocalcemia (chi square = 5.52; p < 0.025), but not with hypomagnesemia (chi square = 0.005; difference not significant). The nomogram of normal P T H responses for iCa concentrations in the serum for the 10 children, ages 1.5 to 10 years (6 boys, 4 girls) with atrial septal defects was used for comparison with the P T H responses of the burned children. When plotted on this nomogram, approximately 85% of the values of our patients

KLEIN ET AL.

THE JOURNAL OF PEDIATRICS

Volume 13 I, Number 2 fell below the 99% confidence interval for the above-mentioned children (Fig. 1). Furthermore, for the two patients who had the 2-hour determinations of iCa and PTH, nearly all the points fell b d o w the normal ranges for both P T H and iCa (Fig. 2). These combined data provide evidence for diminished P T H secretory response to hypocalcemia.

Effects of Parathyroid Infusion Table III shows the marked attenuation in the increase in urinary cyclic adenosine monophosphate and in the decrease in T m P / G F R expected in response to synthetic P T H infusion, s The first two patients listed in the table received furosemide within 1 hour of beginning the preinfusion urine collections. The determinations of urinary parameters are given in the table with these patients included and are listed in the footnote with the patients excluded. No significant difference was observed between the two Sets of values. Serum magnesium concentration at the time of PTH infusion was 0.76 _+0.10 mmol/L (range 0.61 to 0.95 rnmol/L). Serum concentrations of osteocalcin and bone-specific alkaline phosphatase were reduced below normal in the preadolescents and did not change after P T H infusion. Osteocalcin levels were 8 -+ 1 ng/ird in the preadolescents (n = 3; normal range 18 to 42 ng/ml), and 13 _+5 ng/ml in the adolescents (n = 3; normal range 8 to 18 ng/ml). These values are consistent with previous findings. 2'4 Similarly, bonespecific alkaline phosphatase activity was 28 U/L in the preadolescents (n = 2; norreal 50 to 150 U/L) and 27 + 18 U/L in the adolescents (n = 3; normal 10 to 50 U/L). Serum concentrations of interleukin-6 were uniformly high, as previously reported 2 and did not increase further despite P T H administration.

Assoeiation Between Serum iCa and Magnesium Depletion Although there was a weak correlation between serum concentrations of iCa and magnesium (n = 70, r = 0.215,p < 0.05), only 4.6% of the variability in serum iCa could be explained by serum magnesium concentrations. However, the lowest serum iCa

Patient 6

q

16 14

O) o.

Z IQ.

1.14 1.12

E 12 V

1.16

10 8

1.10 3 3 1.08 o

6

1.06

4"

1.04

r'-

2

I

I

I

I

I

8A

10

Noon

2P

4

1.02

Time

Patient 9

'~

16

1.17

14

1.15

12

.

.

.

.

1.13

.

B ,

1.11 O

" 10 "" -r 8 po.

1.09 3 3 1.07 O m

6

1.05 v"

4

~1.03

2

v

. 8A

.

. 10

. . Noon

1.01 2P

4

Time F ~ . 2. Relationship between iPTH (solid circles) and iCa (open circles) in two of the burned patients in our study who had blood sampling every 2 hours on a particular day.The solid line represents the lower limit of normal for iPTH and the dashed line represents the lower limit of normal for iCa.

levels (<_1.02 mmol/L, n = 17) were associated with low serum magnesium concentrations (chi square = 4.90, p < 0.05. Furthermore, two of the three normomagnesemic patients who received a 4-hour magnesium infusion had increased magnesium retention of 45% and 49.3%, respectively. The third patient had a magnesium retention value of only 9.2% and was thus magnesium replete. Serum iCa concentrations were 1.09 _*0.07 mmol/L (range 1.02 to 1.15 mmol/L) at the start of the infusion,

whereas the P T H level was available in only one of the three and was low.

DiscussioN In this study we found a high prevalence of low serum concentrations of iCa, magnesium, and i P T H in children with burn injuries covering >50% total body surface area. The hypocalcemia wa s unaccompanied by elevated serum concentra-

249

KLEIN ET AL.

TIlE JOURNAL OF PEDIATRICS

AUGUST 1997

Table IlL Study group C: Cyclic adenosine monophosphate and phosphaturic responsesto PTH

tions of iPTH. Furthermore, infusion of synthetic PTH into hypocaleemic patients resulted in blunted renal response to the infusion. In addition, inasmuch as the production of interleukin-6 by murine 12 and human marrow stromal cells 13is sfA-nulated by PTH, failure of interleukin-6 to respond to PTH administration constitutes additional evidence supporting tissue resistance to PTH after burn injury. The presence of both hypoparathyroidism and end-organ PTH resistance in the same patients suggests that magnesium depletion may play a critical role in the pathogenesis of these abnormalities. 14'15 Magnesium depletion can occur after burn injury. The magnesium deficiency is not en6rely explained. Although an intracellular shift of magnesium is possible after the burn injury, recent preliminary data suggest that at least some of the magnesium loss occurs through the burn wound exudate. 16 The significant association of hypomagnesemia with the lowest serum iCa concentrations is consistent with this. Moreover, the finding of increased magnesium retention in two of three patients with normomagnesemia suggests that normal serum magnesium concentrations are not necessarily indicative of magnesium repletion. Although these data are suggestive of generalized abnormalities in calcium regulation, regardless of their relation to magnesium

250

status, the data obtained are subject to certain limitations that might temper the interpretation of our findings. With regard to the renal and skeletal responses to PTH, the only normal values published for the modified EllsworthHoward test are in young adults. 5 Some adolescent patients with hypoparathyroidism or pseudohypoparathyroidism were also included in this study, providing some overlap between our patient population and that of Mallette et al., 5 because 4 of the 10 patients in our study were adolescents. Inasmuch as our conditions of P T H administration followed those of the published reports as closely as possible, s'17 and the data obtained from the PTH infusion in the adolescents were not different from those obtained from the preadolescents, we believe that the blunted renal response is indeed valid. Moreover, the high baseline values of T m P / G F R in our study are consistent with the hypoparathyroid state. A population of children with an atrial septal defect acted as control subjects 9 for the relationship between serum iCa and iPTH. Although the presence of an isolated atrial septal defect raises the concern that other abnormalities may exist, including those of the musculoskeletal system, the latter have not been demonstrated in this or other groups of such patients. Therefore for purposes of this study, we

can make the assumption that calcmm homeostasis is normally regulated in this group of children, because they demonstrated an increase in iPTH during hypoca]cemia. Another potential problem in the interpretation of the clata obtained from the P T H infusion data was the therapeutic administration of furosemide to two of the patients immediately before the basal (pre-PTH infusion) urine collection. Data from these patients were included in the table because existing data on the subject of the influence of furosemide on the response to PTH as studied by Fujita et al.18 suggest that the diuretic itself induces an endogenous PTH response, and thereby blunts the renal effects of subsequently infused parathyroid extract. We are certain that this mechanism is not operative in our patients after burn injury, because serum levels of iPTH did not increase after furosemide administration (data not shown). Furthermore, in these patients furosemicle was given only occasionally and would not likely be responsible for the depletion of calcium stores. With regard to the failure of skeletal response to PTH infusion, no firm conclusions can be drawn because the time period of the study may have been too short to detect any changes in osteocalcin or bonespecific alkaline phosphatase. Linclsay et al. 19 experienced a similar failure of re-

THE JOURNALOF PEDIATRICS Volume 13 I, Number 2 sponse 4 hours after administration of synthetic PTH to postrnenopausal women. The exact time of osteocalein response to PTH administration in children is not certain. Although the relative paucity of data regarding the response to P T H infusion limits our ability to more accurately interpret the data we have obtained, it is apparent that the hypoparathyroidism detected is valid. The hypoparathyroidism could possibly play a role in the reduced bone formation previously reported in these patients, 3 inasmuch as markers of bone formation such as osteocalcin are lower under hypoparathyroid conditions.20 • Whether magnesium depletion is a contributing factor to the pathogenesis of postburn hypocalcemic hypoparathyroidism a n d P T H resistance can only be determined by a prospective study that examines the effects of more aggressive magnesium repletion on calcium homeostasis in these patients. Our new observations in burn patients should alert the clinician caring for similar patients to observe their patients closely for symptomatic hypocalcemia. Furthermore, although we remain unsure whether magnesium repletion will correct the observed abnormalities in calcium homeostasis, we encourage consideration of the appropriate use of calcium infusions in symptomatic patients.

REFERENCES 1. Rutan RL, Herndon DN. Growth delay in postburn patients. Arch Surg 1990;125: 392-5.

KLEIN El" AL.

2. Klein GL, Herndon DN, Rutan TC, Sherrard DJ, Coburn JW, Langman CB, et al. Bone disease in burn patients.J Bone Miner Res 1993;8:337-45. 3. KleinGL, Herndon DN, Goodman WG, LangmanCB, PhillipsWA, Dickson IR, et al. Histomorphometric and biochemical characterizationof bone followingacute severe burns in children. Bone 1995;17:45560. 4. Klein GL, Herndon DN, Langman CB, Rutan TC, YoungWE, PembletonG, et al. Long term reduction in bone mass following severe burn injuryin children.J Pediatr

13.

14.

1995;126:252-6. 5. 2vlalletteLE, Kirkland JL, Gagel RF, Law WM Jr, Heath H III. Synthetic human parathyroid hormone(I-34)for the study of pseudohypoparathyroidism. J Clln EndoerinolMetab 1988;67:964-72. 6. United States Pharmacopeia. Drug informarion for the health care professional.6th ed. Rockville(MD): United States Pharmacopeial Convention;1996.p. 2814-5. 7. WaltonRJ, Bijvoet OLM. Nomogram for the derivationof renal threshold phosphate concentration.Lancet 1975;2:309-10. 8. Rude RK. Magnesium deficiencyand hypermagnesemia. In: Favus IV[J, editor. Primer on the metabolicbone diseases and disorders of mineral metabolism. 2nd ecl. New York: Raven Press; 1993.p. 210-2. 9. Cuneo BE Langman CB, Ilbawi MN, RamakrishnanV, CutillettaA, DriscollDA. Latent hypoparathyroidism in Children with eonotruneal cardiac defects. Circulation 1996;95:1702-8. 10. Reed A, Haugen M, Pachman LM, Langman CB. Abnormalitiesof serum osteoealtin in childrenwith chronic rheumaticdiseases. J Pediatr 1990;t 16:574-80. 11. Law WH dr, Heath HIII. Increased renal responses to exogenous parathyroid hormone in postsurgieal hypoparathyroidism. d Clin EndocrinolMetab 1.984;59:394-7. 12. SakagamlY, Girasole G, Yu X-P, Boswell HS, Manolagas SC. Stimulation of interleukin-6 production by either calcitonin

15.

16.

17.

18.

19.

20.

gene-related peptide or parathyroid hormone in two phenotypicallydistinct bone marrow-derivedmurinestromalcelllines.J Bone Miner Res 1993;8:811-6. Noordln S, Cheluitte D, Yfishra N, Bleiberg I, GlowacldJ. Effects of parathyroid hormone, 1,25-dihydrox~tamin D, and interleukin-1on the secretionof interleukin-6 and -11 by human bone marrow (abstract). J Bone lVilner Res 1996; 1l(Supp 1):$416. AnastCS, WinnackerJL, Forte LR, Burns TW. Impaired release of parathyroid hormone in magnesium deficency, d Clin EndocrinolMetab 1976;42:707-17. Rude RK, Oldham SB, Singer FR. Functional hypoparathyroidismand parathyroid hormone end-organ resistance in human magnesiumdeficiency. Clin Endocfinol 1976;5:209-24. Berger MM, Rothen C, Cavadini C, Chiolero R. Magnesiumdeficiencyin major burns: exudativelossescontributeto the depletion (abstract). JPEN J Parenter Enteral Nutr 1996;20(suppl):20S. Furlong TJ, Cornish CJ, Seshadri M, Luttrell B, Wilkinson MR, Posen S. Clinical experience with human parathyroid hormone 1-34. Aust N Z d Med 1986; 16:794-8. Fujita T, Delea CS, Bartter FC. The effects of oral furosemide on the response of urinary excretion of cyclic adenosine monophosphate and phosphate to parathyroid extract in normal subjects. Nephron 1985;41:333-6. LindsayR, Nieves J, Henneman E, Shen V, Cosman F. Subcutaneousadministration of the amlno-terminalfragment of human parathyroid hormone(l-34): kinetics and biochemicalresponsein estrogenizedosteoporofic patients. J Clin Endocrinol Metab 1993;77:1535-9. Brown JE Delmas PD, Malaval L, Meunier PJ. Serum bone Gin-protein: a specific marker for bone formationin postmenopausal osteoporosis. Lancet 1984;1: 1091-3.

251