Vasopressin
Levels
and Pediatric
Head
Trauma
Guadalupe
Padilla,
MD, John A. Leake,
Robert
Castro,
MD,
M. Gore Ervin, PhD, Michael
G. Ross, MD, and
Rosemary
D. Leake,
MD
From the Departments of Pediatrics and Obstetrics, University of California Los Angeles School of Medicine, Harbor-UCLA Medical Center, Torrance, California
ABSTRACT. The syndrome of inappropriate secretion of
antidiuretic hormone is associated with head trauma;
however, there are no reports concerning vasopressin
levels in pediatric patients with head trauma. Urine va-sopressin in eight children (mean ± SEM, age 7.5 ± 1.6
years, range 1 to 15 years) was measured by radio-immunoassay during their hospitalization for head
trauma. Urine vasopressin values for ten healthy children
(mean age 5.4 ± 1.3 years) and for eight children
hospi-talized for systemic antibiotic treatment ofinfections (age 5.9 ± 1.8 years) also were obtained. Urine vasopressin, urine and serum sodium concentration and osmolality, urea nitrogen, creatinine, and fluid intake were measured within 24 hours of admission and daily for the following two days. For the first three days following head trauma,
mean urine vasopressin levels in pediatric patients with
head trauma were increased (P < .05) compared with
those of healthy children. Despite fluid restriction to 85%
of maintenance level, 25% of patients with head trauma
exhibited the clinical syndrome ofinappropriate secretion
of antidiuretic hormone (hyponatremia, increased un-nary sodium, diminished serum osmolality, and urine
osmolality greater than serum osmolality). Urine osmo-lality greater than 800 mosm/kg was associated with
markedly increased urine vasopressin levels (200 to 1,650 pg/mL); children with this finding may be at particular
risk for the syndrome of inappropriate secretion of
anti-diuretic hormone without restrictive water intake.
Pedi-atrics 1989;83:700-705 vasopressin, head trauma, water balance, hyponatremia, inappropriate secretion of antidi-uretic hormone syndrome.
Antidiuretic hormone secretion is said to be ap-propriate when it occurs in response to
intravas-Received for publication Feb 9, 1988; accepted June 30, 1988. Presented, in part, at the Western Society for Pediatric Research meeting, Carmel, CA, February 1987.
Reprint requests to (R.D.L.) Department of Pediatrics, RB-i, Harbor-UCLA Medical Center, 1000 West Carson St, Torrance, CA 90509.
PEDIATRICS (ISSN 0031 4005). Copyright © 1989 by the
American Academy of Pediatrics.
cular hypovolemia or hyperosmolality. The syn-drome of inappropriate secretion of antidiuretic hormone has been described in the absence of these stimuli in adults with neoplasms, CNS or pulmo-nary conditions, and in association with a variety of drugs.”2 In few studies has antidiuretic hormone, or vasopressin, been measured in children with these conditions. Increased plasma vasopressin concentrations have been described in pediatric patients with bacterial meningitis,2 asphyxiated in-fants,3 following patent ductus ligation,4 and during episodes of respiratory distress with bronchopul-monary dyspbasia.5 Generally, however, vasopressin levels are not known and clinicians must rely on indirect evidence of increased vasopressin secretion producing the syndrome of inappropriate secretion of antidiuretic hormone, including the following: (1) a serum Na <135 mEqJL, (2) urine Na >25 mEqJL, (3) serum osmobality <280 mosm/kg H2O, and (4) urine osmolality > serum, in the absence of renal, adrenal, or thyroid disease or dehydration.
Since the first reports of electrolyte disturbances associated with head injury,6 head trauma has been thought to produce inappropriate vasopressin se-cretion. Clinical and (indirect) laboratory evidence of the syndrome of inappropriate secretion of an-tidiuretic hormone has been described in adult pa-tients with head trauma.7’2 In a review of 1,800 adult patients with head trauma, Doczi et al’2 de-scnibed a prevalence of the syndrome of inappro-priate secretion of antidiuretic hormone of 5% in adults with mild head trauma, 11% with moderate head trauma, and 5% with severe trauma. There have been no reports of vasopressin levels in pedi-atric patients with head trauma, however,
in pediatric patients with head trauma to quantify vasopressin secretion, to correlate vasopressin 1ev-els with the clinical syndrome of inappropriate
se-cretion of antidiuretic hormone in this patient pop-ulation, and to attempt to identify a readily avail-able laboratory marker for those children with head trauma who were at particular risk for the syn-drome of inappropriate secretion of antidiuretic hormone.
MATERIALS AND METHODS
After obtaining the approval of the Human Sub-ject’s Protection Committee and parental consent,
we studied eight children hospitalized for head trauma in the Harbor-UCLA pediatric intensive
care unit from March 1985 to May 1987. The
mech-anism of injury producing the head trauma included automobile v automobile for three children, auto-mobile vpedestrian for four children, and a fall for one child. The patients ranged from 1 to 15 years of age (mean ± SEM = 7 ± 1 years). Patients with severe trauma, hypovobemia, multiple fractures, or evidence of renal trauma were considered ineligible for inclusion in the study.
Patients with head trauma were assessed by
mi-tial Glasgow Coma Scale, serial neurologic exami-nation, cerebral computed tomographic (CT) scan findings, days of hospitalization, and patient out-come. All patients survived except one who died of adult respiratory distress syndrome 2 weeks follow-ing injury. Mean (± SEM) days of hospitalization for the remaining patients was 6.7 (±1.5) days. Glasgow Coma Scores assigned in the emergency room ranged from 7 to 15 (mean ± SEM = 11 ± 1). Seven of eight patients had abnormal CT scan findings, including fractures in five patients, hem-orrhage in two, and contusions in one. Four CT scans exhibited evidence of cerebral edema.No patients required neurosurgical procedures. One patient required ventilatony support because of deteriorating neurobogic status. One child re-ceived mannitol, three dexamethazone, one diphe-nylhydantoin, one phenobarbital, three cimetidine, one pancuronium, and one morphine. The anticon-vulsant medications were begun for seizure activity and the remaining medications for clinical
indica-tions.’5
A single urine specimen was collected from a bladder catheter or from a spontaneous voiding for osmolality, vasopressin, and sodium concentrations within 24 hours of admission and daily for two additional days (n = 7). Because of early hospital discharge, one patient was studied for the first two days only. Vasopressin levels were also measured from single spot urine samples from ten healthy, nonhospitalized children of similar ages (5.4 ± 1.3
years) receiving fluid as desired. To quantify vaso-pressin levels in the presence of stress, eight well-hydrated children of similar ages hospitalized for treatment of infections (excluding pneumonia or meningitis) were also studied. Serum sodium, urea nitrogen, and creatinine levels were determined daily within eight hours of the urine sample for the subjects with head trauma. Serum osmolabity was measured directly (n = 2) or calculated (n = 6)
using the following formula: osmolality = 2[Na] +
BUN/2.8 + glucose/18.
Fluid management consisted of 0.45% NaC1 in
5% dextrose IV with enteral feeding begun as to!-erated at rates calculated to provide fluid restriction
(85% of calculated maintenance based on the ca-boric method) and approximately 4 mEcijkg/d of sodium.
Urine vasopressin levels were measured by radio-immunoassay with the following method: Urine samples were collected in plain glass tubes contain-ing iN HC1 and stored frozen at -20#{176}C. Extraction was performed by means of disposable SepPak
col-umns (C,8 SepPak). Radioimmunoassay was per-formed as previously described.’3 Each sample was assayed in duplicate. Standardized samples were assayed to determine the amount of vasopressin recovered per assay. Mean recoveries were 49% ± 3.5; values were corrected for recoveries.
Urinary sodium was measured by flame pho-tometry (Instrumentation Labs). Urine osmolality was determined by freezing point depression using an advanced digimatic osmometer (Advanced In-struments). Serum sodium, BUN, and creatinine concentrations were measured in the hospital lab-oratory by standard methods.
Data for the serial study days for subjects with head trauma were compared by paired t tests; Stu-dent’s t tests were used to compare values from head trauma and each of the control groups. All data are reported as the mean ± SEM.
RESULTS
Mean (± SEM) fluid intake, representing ap-proximately 85% of maintenance daily fluid re-quirements, was 63 ± 13 mL/kg/d on day 1, 59 ±
13 mL/kg/d on day 2, and 55 ± 14 mL/kg/d on day 3. Mean sodium intake was 3.6 ± 0.5 mEqJkg/d,
3.9 ± 0.6 mEqjkg/d, and 3.5 ± 1.5 mEqJkg/d on days 1, 2, and 3, respectively. Urine output was 1.7
± 0.3 mL/kg/h on day 1, 2.1 ± 0.3 mL/kg/h on day
2, and 1.6 ± 0.2 mL/kg/h on day 3.
The urine vasopressin levels for the patients with head trauma and healthy subjects are shown in Fig
900
800
700
::
600500
400
4 >
Li 300
cx:
200
* *
T
I
100
0 HT HT HT IC0NTR0LS1
DAYI DAY2 DAY3
Fig 1. Mean (± SEM) urine vasopressin levels (pg/mL) ofpatients with head trauma (HT) three days posttrauma and age-matched nonhospitalized healthy control sub-jects. Values for children hospitalized for other than head trauma were similar to those of healthy control subjects.
* P < .05, Student’s t test; head trauma vcontrol subjects.
160
A
155
.
150
Li #{149} .
E 145
S
2 S
a #{149} S
0140 s
C/) S
S
cx:
135
:
I 30
S
350
B
300
250 S
Li
!
200:D S
0
150
Li
z I
ix: S S
1 25 ii ‘I II
::
DAY I DAY 2 DAY 3 DAY I DAY 2 DAY 3
Fig 2. Serum (panel A) and urine (panel B) sodium levels of patients with head trauma three days posttrauma. Circles represent individual values; bars show mean ± SEM for each day.
pg/mL on day 3. Mean urine (± SEM) vasopressin bevels were 454 ± 248 pg/mL, 113 ± 36 pg/mL, and
108 ± 32 pg/mL on the three study days for these subjects; these values were not significantly differ-ent from each other (P > .05, paired t test). Urine vasopressin levels in the healthy children were 32
± 6 pg/mL and for hospitalized children 27 ± 10.7
pg/mL. Vasopressin values in head trauma were significantly greater on all three days of study than in the healthy control patients (P < .05, Student’s
t test).
Serum and urine sodium values for the study patients with head trauma are shown in Fig 2. Serum sodium ranged from 138 to 153 mEciJL on day 1, from 133 to 146 mEqJL on day 2, and from 128 to 144 mEqJL on day 3. Mean serum sodium values were 142 ± 1.6 mEQJL, 138 ± 2 mEqJL, and
136 ± 2.5 mEqJL, respectively.
Urine sodium values ranged from 67 to 251 mEuJ L on day 1, 31 to 208 mEqJL on day 2, and 28 to
176 mEqjL on day 3. Mean urine sodium levels were 122 ± 27 mEqjL, 75 ± 20 mEqJL, and 80 ± 19 mEqjL, respectively, on days 1, 2, and 3.
S
S S
S
S 55
.#{149}.. S. a
five to ten times those of plasma and provide an integrated value reflecting plasma bevels
through-I 500- out time.
All of the children included in the study group experienced head trauma associated with a cerebral
1300-concussion and/or a linear skull fracture. None experienced bong periods of unconsciousness or
de-I
‘ I I 00- pressed or compound skull fractures. Values from
all but one of these children were strikingly
in-z creased on the day of admission (mean ± SEM =
900- 454 ± 248 pg/mL) compared with those of
age-matched healthy children (32 ± 6 pg/mL) and Q.
700- children hospitalized for other causes (27 ± 10.7
pg/mL). The exact duration of increased
vasopres-‘U sin secretion remains unclear; by the third day of
500-D study, only two children with head trauma
exhib-ited urinary vasopressin bevels within the range of
300- #{174} the control patients. Several of the children with
markedly increased vasopressin levels were
dis-100- charged to home and directed to take water as
desired before their vasopressin values were known.
I I F I I -I As seen in the Table, there was no clear
correla-200 400 600 800 1000 1200
URINE OSMOLALITY (m0sm/kg H20) tion between vasopressin levels and Glasgow Coma
Fig 3. Mean (± SEM) urine vasopressin levels corn- Score or presence of cerebral edema in this limited
pared with urine osmolality for patients with head series of patients. Similarly, Feigin and Kaplan’6 trauma. Black circles within circles represent urine values reported no correlation ofplasma vasopressin bevels
from patients in whom clinical syndrome ofinappropriate and the degree of meningeal inflammation (as
man-secretion of antidiuretic hormone developed. ifested by CSF cell count, CSF protein or glucose
concentration, or CSF to blood glucose ratios) in
urine osmolality on these days was 821 ± 83 mosm/ children with meningitis.’6
kg, 610 ± 122 mosm/kg, and 563 ± 90, mosm/kg, Both of the patients with head trauma and the respectively. As seen in Fig 3, a urine osmolality clinical syndrome of inappropriate secretion of an-greater than 800 mosm/kg correlated with a mark- tidiuretic hormone had cerebral edema according edly increased urine vasopressin level (200 to 1,650 CT scan. In the rhesus monkey brain, water pg/mL). BUN level was less than 16 mg/dL and permeability is affected by intracerebroventricular serum creatinine level less than 1.0 mg/dL through- injection of vasopressin; cerebral edema results.’7 out the study for all study subjects. All were cmi- Brain water content also is increased by
adminis-cally euvolemic. tration of IV infusions of water and
vasopressin-Despite markedly increased vasopressin levels, prOducing hypotonic plasma.’8 The pathogenesis of the syndrome of inappropriate secretion of antidi- cerebral edema in children with head trauma re-uretic hormone did not develop in any subjects with mains unclear. Cerebral compression may produce head trauma on day 1, perhaps because of water vasopressin secretion. We and others’9’2#{176}have de-restriction. Two patients exhibited hyponatremia scribed increased plasma vasopressin bevels in in-with hypertonic urine relative to serum consistent delivered vaginally, contrasted with mildly with the clinical syndrome of inappropriate secre- increased bevels in infants delivered by cesarean tion of antidiuretic hormone, one on both days 2 section. In rhesus monkeys increased intracranial and 3 following trauma, and the other on day 3. pressure produced by a subdural balloon inflation These two patients exhibited the highest urine Va- results in increased u#{241}nar vasopressin bevels.2’ sopressin bevels (1,653, 326, and 275 pg/mL) of all The increased vasopressin levels in the patients
of the study subjects. with head trauma appear to be due principally to
head trauma. Urine vasopressin results from
gb-RESULTS merular filtration because there is essentially no
tubular secretion22 and blood falsely increases Va-These are the first reported values for urinary sopressin levels. Patients with anuria, oligouria, vasopressin levels in pediatric patients with head and/or urinary blood secondary to renal trauma trauma. Urinary vasopressin levels are generally were excluded from study. One patient required
I
TABLE. Demographic Data of Head Trauma Patients* Patient
No.
Glasgow
Score CT Scan
Argi nine Vasopressin
(pg/mL)
Cerebral Other Findings Day 1 Day 2 Day 3
Edema
1 9 + Contusion,
tern-poral fracture
1,653 195
2 7 + Intraventricular
hemorrhage
28 69 144
3 9 + Parietal fracture 283 326 275
4 14 - Temporal
frac-ture
214 38 95
5 13 + Frontal fracture 65 132 26
6 - Subarachnoid
hemorrhage
481 38 95
7 12 - Occipital fracture 77 20
8 15 - None 32 104
aSymbols: +, present; -, absent. Means ± SEM were 454 ± 248 pg/mL for day 1, 113 ±
36 pg/mL for day 2, and 108 ± 32 pg/mL for day 3.
ventilatory support throughout the hospitalization until he expired, but he was not hypoxic during the study period. In adults, positive pressure ventilation increases plasma vasopressin bevels to a limited extent; it is unclear whether there is a similar increase in vasopressin secretion in children.23’24
None of the patients exhibited renal, adrenal, or thyroid disease or dehydration. These patients did receive either single or multiple doses of dexameth-azone, cimetidine, pancuronium, mannitol, and phenobarbital. All of these drugs were given prior to obtaining the first urine sample. The urine was collected at least six hours after the medication was administered, with the exception of cimetidine, decadron, and phenytoin. Moreover, none of the drugs given or treatments used has a clear-cut effect of increased vasopressin release. Morphine has been reported to have varying effects on vasopres-sin release, probably reflecting the effect of the underlying pain for which the morphine is admin-istered.25 Pain per se affects vasopressin secretion, especially in adults with intraabdominal surgery.229 We showed previously, however, that infants undergoing circumcision did not demon-strate increased vasopressin levels.30 Moreover, pain control was provided as needed for our patients and none of the patients included as study subjects underwent surgery during the study days. Mannitol might have an indirect effect if hypovolemia results from its administration, but none of our patients was clinically hypovolemic. Therapy with decadron has been previously reported to increase sodium excretion and produce a diuresis or to have no effect.3’ Diphenylhydantoin may actually inhibit vasopressin secretion.32’33 Physical and emotional stress may increase vasopressin secretion, perhaps secondary to catecholamine release.33 Stress as a cause cannot be excluded and may have contributed
to the increased vasopressin levels34; vasopressin values in the children hospitalized for treatment of infections were similar to values of nonhospitalized healthy children, however.
In our study it was shown that nearly all of the children with head trauma experience a two- to three-day period of increased argrnine vasopressin secretion. Laboratory criteria for the clinical syn-drome of inappropriate antidiuretic hormone secre-tion were met by 25% of the children with head trauma. Thus, despite fluid restriction to 85% maintenance, two of eight children became hypo-natremic, exhibited urinary sodium excretion in-appropriate for plasma sodium, and a relatively hyperosmolar urine developed concurrently with hypoosmobar plasma. These findings may be cmi-cally important in that the degree and duration of hyponatremia appears to correlate with the extent of neurobogic sequebae.’5 The clinical practice of providing water restriction appears appropriate for the first three days after significant head trauma even if there is clinical improvement. It is clear that some children would profit by an even greater water restriction. Studies are in progress to deter-mine whether urine osmolality can be used to pre-dict which patients these are.
ACKNOWLEDGMENT
This work was supported by US Department of Health
and Human Services, Public Health Service, grant
HD-06335 from the National Institute of Child Health and
Human Development and by grant RE-00425 (General Clinical Research Center) from the National Institutes of Health.
REFERENCES
i. Kaplan SL, Feigin RD: Syndromes of inappropriate
secre-tion of antidiuretic hormone in children. Adv Pediatr
2. Kaplan SL, Feigin RD: The syndrome of inappropriate
secretion of antidiuretic hormone in children with bacterial meningitis. J Pediatr 1978;92:758-76i
3. Kaplan SL, Feigin RD: Inappropriate secretion of
antidi-uretic hormone complicating neonatal hypoxic-ischemic
en-cephabopathy. J Pediatr i978;92:431-433
4. Weinberg JA, Weitzman R, Zakauddin 5, et al:
Inappro-priate secretion of antidiuretic hormone in a premature infant. J Pediatr 1977;90:iii-i14
5. Rao M, Eid N, Herrod L, et al: Antidiuretic hormone
re-sponse in children with bronchopulmonary dysplasia during episodes of acute respiratory distress. Am J Dis Child
1986;140:825-828
6. Higgins G, O’Brien JRP, Lewis W, et al: Metabolic disorders
in head injury. Lancet i954;i:61-67
7. Born JD, Hans P, Smitz 5, et al: Syndrome of inappropriate secretion of antidiuretic hormone after severe head trauma.
Surg Neurol i985;23:383-387
8. Maroon JC, Campbell RI: Subdural hematoma with map-propriate antidiuretic hormone secretion. Arch Neurol
1970;22:234-239
9. Carter NW, Rector FC, Seldin DW: Hyponatremia in cere-bral disease resulting from the inappropriate secretion of antidiuretic hormone. N Ertgl J Med i961;264:67-72
iO. Richards DE, White BA, Yashaw D: Inappropriate release
of ADH on subdural hematoma. J Trauma 197i;ii:758-762
ii. Shenkin HA, Bezier HS, Bouzarth WF: Restricted fluid
intake: Rational management of the neurosurgical patient.
J Neurosurg i976;45:432-436
12. Doczi T, Tarjanyi J, Huszka E, et al: Syndrome of
inappro-priate secretion ofantidiuretic hormone (SIADH) after head
trauma. Neurosurgery 1982;iO:685-688
13. Tausch A, Stegner H, Leake RD, et al: Radioimmunoassay of arginine vasopressin in urine: Development and
applica-tion. J Clin Endocrinol Metal, i983;57:777-781
i4. Matsui K!, Share L, Wang BC, et al: Effects of changes in
steady state plasma vasopressin levels on renal and urinary
vasopressin clearance in the dog. Endocrinology i983;112:2107-21i3
15. Raphaely RC, Swedbow DB, Downes JJ, et al: Management
of severe pediatric head trauma. Pediatr Clin North Am
i980;27:7i5-727
16. Feigin RD, Kaplan 5: Inappropriate secretion of antidiuretic
hormone in children with bacterial meningitis. Am J Clin Nutr i977;30:1482-1484
i7. Raichle ME, Grubb RL: Regulation of brain water
permea-bility by centrally released vasopressin. Brain Res
1978;143:191-194
18. Dila CJ, Pappius HM: Cerebral water and electrolytes-An
experimental model of inappropriate secretion of
antidi-uretic hormone. Arch Neurol i972;26:85-90
19. Hoppenstein JM, Miltenberger FW, Moran WH Jr: The
increase in blood levels of vasopressin in infants during
birth and surgical procedures. Surg Gynecol Obstet 1968;127:966-974
20. Hadeed AJ, Leake RD, Weitzman RE, et al: Possible
mech-anisms of high blood levels of vasopressin during the
neo-natal period. J Pediatr i979;94:805-808
2i. Gaufin L, Skowsky WR, Goodman SJ: Release of
antidi-uretic hormone during mass-induced elevation of intracra-nial pressure. J Neurosurg 1977;46:627-637
22. Lindheimer MD, Reinharz A, Grandchamp A, et al: Rate of vasopressin perfused into nephrons of Wistar and
Brattle-boro (diabetes insipidus) rats. Clin Sci 1980;58:139-144
23. Leslie GI, Philips JB, Work J, et al: The effect of assisted
ventilation on creatinine clearance and hormonal control of electrolyte balance in very low birth weight infants. Ped Res
1986;20:447-45i
24. Sladen A, Layer MB, Pontoppidan H: Pulmonary
compli-cations and water retention in prolonged mechanical venti-lation. N EngI J Med 1968;279:448-452
25. Moses AM, Miller M, Streeten DHP: Pathophysiologic and
pharmacologic alterations in the release and action of ADH. Metabolism i976;25:697-721
26. Kendler KS, Weitzman RE, Fisher DA: The effect of pain
on plasma arginine vasopressin concentrations in man. Clin Endocrinol i978;8:89-94
27. Moran WH, Zimmerman B: Mechanisms of antidiuretic hormone (ADH) control of importance to the surgical pa-tient. Surgery 1967;62:639-644
28. Simpson PJ, Forsling ML: The effect of halothane anes-thesia on plasma vasopressin during cardiopulmonary by-pass. Clin Endocrinol 1977;7:33-39
29. Sinnatamby C, Edwards CRW, Kitau M, et al: Antidiuretic hormone response to high and conservative fluid regimes in
patients undergoing operation. Surg Gynecol Obstet i974;
139:715-719
30. Waters CB, Weinberg JE, Leake RD, et al: Arginine vaso-pressin levels during a painful stimulus in infancy. Pediatr Res i982;i6:569
3i. Shenkin HA, Gulterman P: The analysis of body water compartments in postoperative craniotomy patients. Part 3:
The effects of dexamethasome. J Neurosurg
i969;31:400-407
32. Landolt AM: Treatment of acute post-operative inappro-priate antidiuretic hormone secretion with
diphenylhydan-tom. Acta Endocrinol i974;76:625-629
33. Fichman MP, Kleeman CR, Bethune JE: Inhibition of an-tidiuretic hormone secretion with diphenylhydantoin. Arch Neurol 1970;22:45-53
34. Schrier RW, Berl T: Nonosmolar factors affecting renal
