Hypothalamic Adipsia Without Demonstrable Structural Lesion

(1)

Hypothalamic

Adipsia

Without

Demonstrable

Structural

Lesion

Alberto Hayek,

MD, and Glenn

T. Peake,

MD

From the Departments of Pediatrics and Medicine, University of New Mexico Medical

School, Albuquerque

ABSTRACT. The clinical and laboratory data of a 5-year-old boy with the syndrome of essential hypernatremia are

presented. In a four-year follow-up, no demonstrable

hy-pothalamic structural lesion has been identified. Review of the literature has uncovered four similar cases,

sug-gesting a distinct syndrome of altered hypothalamic

func-tion. The syndrome is characterized by:

adipsia-hypodip-sia (5/5 patients), recurrent hypematremia (5/5), obesity (4/5), inability to excrete a water load (5/5), lack of

growth hormone release in response to provocative

stim-uli (4/4), blunted thyrotropin releasing hormone

re-sponses (3/4), hypothyroidism (2/4), and hyperlipemia

associated with hypernatremic crisis (1/1). In one of the

patients the syndrome has been attributed to a

distur-bance of the opioid-peptide system. Pediatrics

70:275-278, 1982; adipsia, hypernatremia, hypothalamic

dis-ease.

of their endocrine findings, suggests a distinct syn-drome of altered hypothalamic function, at present,

of unknown etiology. In one of the patients the

syndrome has been attributed to a disturbance of

the hypothalamic opioid peptide system.’

CASE REPORT

The patient, now a 9-year-old boy, was born by cesar-ean section because of twinning. The twins shared the same placenta, weighed 2.5 kg at birth, had the same blood group, and were identical until age 5 years, when the patient began to develop obesity. Seven months later, the first episode of hypernatremia was documented with a serum sodium level of 178 mEq/liter. The clinical and laboratory findings can be summarized as follows.

General Findings

Adipsia-hypodipsia and recurrent hypernatremia

are usually manifestations of structural abnorrnali-ties of the hypothalamic-pituitary area. In 1979, Schaad et al’ reviewed the literature on this rare entity and described 35 patients of all ages with

“hypodipsia-hypernatremia,” also called the

syn-drome of “essential hypernatremia.”2 In two further reports,”4 three more patients have been identified bringing the total to 39; approximately one third are children less than 15 years of age. If tumors are

found, they fall more commonly in the general

category of germinoma, a generic term encompass-ing all CNS germ cell tumors.5

The child reported here, who originally had

adip-sia and hypernatremia, resembles four other

pa-tients described in the literature”3’6’7 in whom no

demonstrable hypothalamic structural lesion has

been identified. This, in addition to the similarity

Received for publication May 22, 1981; accepted Aug 5, 1981.

Reprint requests to (A.H.) Department of Pediatrics, University

of New Mexico, School of Medicine, Albuquerque, NM 87131.

American Academy of Pediatrics.

Coincident with the onset of obesity, the parents noted this boy’s decreased activity, episodes of lethargy, in-creased perspiration; mood changes, and wide tempera-ture oscillations (from 34.5 to 39 C). During

hyperna-tremic crises, he is unable to walk but returns to full activity once the serum sodium level reaches the normal range.

Growth and Development

The patient’s developmental milestones were the same

as those of his twin brother. Both are now in the third

grade. Growth rate in the twins was equal until age 5

years. For the last three years, the patient has grown 5 cm/yr (25th percentile) while his brother has grown

7

cm/yr (50th percentile).

Adipsia and Hypernatremia

Both at home and during several periods of hospitali-zation, the patient did not manifest thirst in spite of serum sodium levels ranging between 155 and 191 mEq/ liter.

Hormonal Status

(2)

accompa-276 HYPOTHALAMIC ADIPSIA

nied by urine specific gravities of at least 1,030 and

osmolalities >1,100 mOsm/kg. There is no history of

polyuria, and on a minimal fluid intake of 1.5 liters/day

his daily urine output ranges between 1 and 1.3 liters.

Also, he has never shown the clinical signs of dehydration, which are associated with hypovolemia, during episodes of hyperosmolality. In the course of a 24-hour water

deprivation test the patient’s serum sodium level rose

from 144 to 155 mEq/liter with serum and urine

osmolal-ities of 294 and 726 mOsm/kg, respectively (Dr R. Gotlin,

University of Colorado Medical Center). Recently, while

on cortisone and thyroid replacement therapy the patient excreted only 10% of a water load.

Prolactin (PRL). From age 5/12 years to the present, random serum prolactin levels have ranged between 58 and 98 ng/ml (normal <15 ng/ml).

Growth Hormone (GH). In spite of the patient’s normal

growth, repeated growth hormone provocation studies at 6 and 8#{189}years of age have failed to provoke release of

GH. His basal somatomedin-C (Sm-C) level, however, is in the normal range of 0.6 unit/ml (normal for age, 0.4 to

2 units/nd).

Adrenocorticotropic Hormone (ACTH) and Adrenal Gland. At age 5/12 years, metyrapone administration raised the concentrations of deoxycortisol to 13.5 .tg/100

ml. Renin and aldosterone levels were also normal. Three

years later, coincident with the most severe episode of

hypernatremia (serum sodium 191 mEq/liter), vigorous

adminsitration of intravenous fluids decreased his serum

sodium level to 120 mEq/liter. After 24 hours of fluid

restriction, the serum sodium level decreased to 1 16 mEq/

liter. Basal cortisol levels measured <4 jtg/100 ml with a

prompt response to 34 tg/100 ml 60 minutes after

syn-thetic ACTH. Solucortef, administered intravenously,

in-duced a profuse diuresis with normalization of the serum

sodium level. Thus, although the patient’s adrenal

func-tion was normal initially, three years later it became

abnormal; one manifestation of this abnormality was failure to excrete a water load.

Thyroid-Stimulating Hormone (TSH) and Thyroid

Gland. Early in the course of his disease thyroxine (T4), triiodothyronine (T3), and TSH levels were normal but

there was no measurable TSH release following

stimula-tion with thyrotropin-releasing hormone (TRH). Two

weeks prior to the episode of adrenal insufficiency de-scribed above, the patient was started on thyroid

replace-ment therapy because of clinical signs suggestive of

hy-pothyroidism although his T4, T3, and TSH levels were

in the low-normal range. Recently, when synthroid

ther-apy was discontinued, the T4, T3, and TSH measured 4.4

tg/100 ml, 80 ng/100 ml, and 1 tU/ml, respectively.

Other Findings

Hyperlipemia was first documented coincident with

the most severe hypernatremic episode (serum sodium

191 mEq/liter) with values for cholesterol of 282 mg/100 ml and triglycerides of 786 mg/100 ml. Six days later, the sodium level measured 135 mEq/liter, cholesterol was 169 mg/100 ml, and triglycerides measured 104 mg/100

ml. During two further episodes of hypernatremia,

hy-perlipemia has been again documented with elevations

mainly of the serum triglyceride level and while the patient is on cortisone and thyroid replacement therapy.

Findings from skull films and pneumoencephalogram

were normal. Two computed tomography (CT) brain

scans were negative at 58,/,2 and 8/12 years of age,

respec-tively.

Eftects of Naloxone and Bromocriptine

Prompted by a report3 suggesting that naloxone ther-apy in a child affected with a similar disorder had led to normalization of some of the observed endocrine abnor-malities, we admitted our patient to the Clinical Research Center on two occasions after informed consent was

ob-tamed. During these hospitalizations, each lasting three

days, the patient received an infusion of naloxone (1.6

mg/hr for four hours) while from the opposite arm, blood samples were obtained following TRH (200 tg intrave-nously) for measurements of TSH and PRL. This was followed by a standard arginine-L-dopa provocative test

for GH release. The same procedure had been performed

two days before, with sodium chloride infused as control. After a week of bromocriptine therapy (7.5 mg daily), the patient was rehospitalized to assess the effect of PRL suppression on somatomedin-C levels. A repeat TRH test was performed measuring TSH and PRL at ten-minute

intervals for the next 60 minutes.

TSH,8 PRL,9 and GH’#{176}were measured by standard radioimmunoassay. Somatomedin-C was measured by ra-dioimmunoassay at Nichols Institute, Los Angeles.

RESULTS

As seen in Table 1, except for an increase in the FSH level at the end of the four hours of

intrave-nous naloxone therapy, there was little difference

in the TSH and PRL response to TRH. Although,

three years earlier, this child had a blunted TSH response to TRH, at this latter time, his responses were normal while he was receiving thyroid

replace-ment therapy. Naloxone administered

intrave-nously failed to normalize basal PRL levels,

al-though the peak response was brisker than in the

control test. GH was unmeasurable during and after

naloxone therapy in response to the combined

ar-gmlne-L-dopa stimulation.

During the one-week trial of bromocriptine, there were no noticeable changes in the patient’s drinking

behavior, mood, or level of activity. Suppression of

PRL, however, towards more normal levels

de-creased the Sm-C to 0.27 units/mI, a level abnor-mally low for his age (normal 0.4 to 2 units/mI).

DISCUSSION

The clinical and laboratory data in this patient

indicate extensive hypothalamic derangement that

is unlikely to be associated with an expanding CNS lesion in view of the lack of other expected objective findings over the four-year period of observation.

at Viet Nam:AAP Sponsored on September 7, 2020

www.aappublications.org/news

(3)

Manifestations of hypothalamic dysfunction in this patient include loss of thirst, obesity, lethargy, decreased activity, and body temperature

oscilla-tions. Hormonal abnormalities are represented by

hyperprolactinemia, failure to release

GH

in

re-sponse to stimulation, and thyroid and adrenal in-sufficiency. The latter two were most likely also

hypothalamic in origin inasmuch as the patient

released TSH in response to TRH and failed to

increase his cortisol response during stress periods. The study ofADH dynamics has been the subject of most reports dealing with the

hypodipsia-hyper-natremia syndrome. Welt2 has proposed that such

a diagnosis be considered only when the hyperna-tremia is present in individuals adequately hydrated

and showing evidence of ADH secretion and renal

tubular responsiveness to ADH. Thus, relative

in-sensitivity or abnormal resetting of

ADH

osmore-ceptors have been advocated as factors leading to

hypernatremia. Laboratory data in our patient do

not clarify the status of ADH in this syndrome; it

is clear that diabetes insipidus cannot be implicated

as a cause of the hypernatremia. However, the

impaired ability to excrete a water load, in the

presence of adequate thyroid and glucocorticoid

replacement therapy, does suggest episodes of

in-appropriate ADH secretion as suggested in other

reports.3

The failure to identify an anatomic hypothalamic lesion in our patient led us to search the literature for similar occurrences. Four additional cases were identified and some of the similarities in their din-ical and laboratory data are summarized in Table 2. In spite of prolonged follow-up, ranging between two and nine years, no pathologic lesions have been visualized in any of the patients. Judicious fluid administration in all and hormonal replacement

when indicated have allowed all the children to

maintain normal activities.

Dunger et al3 suggested local overactivity of the

endogenous opioid peptide system as a possible

etiology for the syndrome, inasmuch as their patient clinically improved after naloxone administration.

In

that patient, the hyperprolactemia resolved

spontaneously, but in our case hyperprolactinemia

has been sustained and naloxone was ineffective in

normalizing

PRL

levels.

In both

cases also, naloxone administration failed to normalize the GH response to provocative stimuli. In all the patients, an

in-ability to excrete a water load has been

docu-TABLE 1. Hormonal Response s to Administratio n of Naloxone and Bromocriptine* Control Values IV Naloxone (1.6 After 7 Days of

mg/hr for 4 hr) Bromocriptine

LH (mIU/ml) 3 3

FSH (mIU/ml) <2 5

TSH post-TRH (tU/ml)

Basal <2.5 <2.5 <2.5

Peak 23 (at 45 mm) 23 (at 45 mm)

PRL post-TRH (ng/ml)

Basal 34 36 12

Peak 43 (at 45 mm) 56 (at 45 mm) 19 (at 20 mm)

GH post-argimne and L-dopa <2 <2

(ng/ml)

Sm-C (U/nd) .60 .27

* Abbreviations used are: LH, luteinizing hormone; FSH, follicle-stimulating hormones;

TSH, thyroid-stimulating hormone; TRH, thyrotropin-releasing hormone; PRL, prolactin; GH, growth hormone; Sm-C, somatomedmn-C.

TABLE 2.

Evidence of

Findings

Structural

in Five Children Reported with Adipsia-Hypernatremia Without

Hypothalamic Lesions* Case and Reference . Age at Onset

(yr)

Aclipsia- Obesity Rate of

Hyperna- Growth

tremia

.

Endocrine Studies CT Brain

Scans

GH ORL TSH

Release Post-TRH 1’ 2 34 47 Present case 3 4 7 5 5

+ +

Ni

+ +

+ + Ni

+ A Ni

+ + Ni

A

NR A

A I A

A Ni Nl

NR Nl NR

A A Ni Ni NR Ni Ni * Abbreviations and symbols used are: GH, growth hormone; PRL, prolactin; TSH, thyroid

stimulating hormone; TRH, thyrotropin releasing hormone; CT, computed tomography;

(4)

278 HYPOTHALAMIC ADIPSIA

mented, but in one case,3 naloxone apparently cor-rected the abnormalities in water and electrolyte excretion.

The rate of growth of 5 cm/yr documented in our patient in the last three years, in the absence of

immunoassayable GH but normal Sm-C levels, is in

accordance with a recent report’ ‘ suggesting that

hyperprolactinemia is associated with increased

Sm-C immunoreactivity in GH-deficient individu-als. Further evidence in favor of this observation is provided here by the demonstration that normali-zation of PRL levels by bromocriptine was followed by a decrease to an abnormally low Sm-C level.

The associated hyperlipemia noted by some

au-thors in the adipsia-hypernatremia syndrome has

been attributed, both in humans4 and laboratory

12to destructive lesions of the ventromedial

hypothalamus. Absence of a demonstrable lesion in

our patient makes any such explanation purely

speculative. It is clear that adrenal and/or thyroid insufficiency does not explain the lipid abnormali-ties because while the patient was receiving

corti-sone and thyroid replacement therapy, two more

documented hypernatremic episodes were

accom-panied by hyperlipemia. Inasmuch as these changes in lipid levels were transient and appeared to

coin-cide with episodes of hypernatremia, it may seem

more reasonable to characterize them as secondary effects of hypernatremia. If this is so, it raises an interesting question about mechanisms.

In conclusion, the data on this patient and the

four others reported in the literature (Table 2)

suggest a distinct syndrome of disordered hypotha-lamic function. This entity appears not to be genet-ically acquired as, in our case, the identical twin is

normal. At some time during their illnesses, the

following manifestations have been documented: adipsia-hypodipsia (5/5 patients), recurrent hyper-natremia (5/5), obesity (4/5), hyperprolactinemia (2/4), hypothyroidism (2/4), and hyperlipemia

as-sociated with hypernatremic crisis (1/1). Only

fur-ther follow-up will reveal whether an anatomic

ab-normality may explain the above abnormalities. To

determine whether the opioid system is implicated

in the syndrome, longer clinical trials with oral

endogenous opioid antagonists appear indicated if

they become available.

ACKNOWLEDGMENTS

This work was supported by grants 5 R01-HD05794-08 and General Clinical Research Centers Program DRR, NIH 5 M01-PR00997-81.

Dr Myra T. Buckman, Senior Investigator, Veterans Administration Hospital, Albuquerque, NM, performed the proiactin measurements.

REFERENCES

1. Schaad V, Vaselta F, Zuppinger K, et at:

Hypodipsia-hyper-natremia syndrome. Heir’ Paediatr Acta 34:63, 1979

2. Welt LG: Hypo- and hypernatremia. Ann Intern Med 56:101, 1962

3. Dunger DB, Leonard JV, Wolff OH, et at: Effect of naloxone

in a previously undescribed hypothalamic syndrome. Lancet

1:1277, 1980

4. SkIar CA, Grumbach MM, Kaplan SL, et at: Hormonal and

metabolic abnormalities associated with central nervous sys-tern germinoma in children and adolescents and the effect of

therapy: Report of 10 patients. J Clin Endocrinol Metab

52:9, 1981

5. Rubinstein LI: Tumors of the central nerves, in Atlas of

Tumor Pathology. Washington, DC, Armed Forces Institute of Pathology, 1972, p 269

6. Blank MS, Farnsworth PB: Idiopathic symptomatic

hyper-natremia in a 9-year-old boy: A clinical and physiological

evaluation. J Pediatr 85:215, 1974

7. Conley SB, Brocklebank JT, Taylor IT, et at.: Recurrent

hypernatremia: A proposed mechanism in a patient with

absence of thirst and abnormal water excretion. J Pediatr

89:898, 1976

8. Utiger R: Thyrotropin, in Jaffe B, Behrman HR (eds):

Methods of Hormone Radioimmunoassay, ed 2. New York, Academic Press, 1979, pp 315-324

9. Sinha YN, Selby FW, Lewis UJ, et at: A homologous

ra-dioimmunoassay for human prolactin. J Clin Endocrinol

Metab 36:590, 1973

10. Peake GT, Morris J, Buckman MT: Growth hormone, in

Jaffe B, Behrman HR (eds): Methods ofHormone Radioim-munoassay, ed 2. New York, Academic Press, 1979 pp

223-243

11. Ciemmons DR, Underwood LE, Ridgway EC, et al:

Hyper-prolactinemia is associated with increased immunoreactive

somatornedin-C in hypopituitarism. J Clin Endocrinol

Me-tab 52:731, 1981

12. Frohman LA, Bernardis LL, Schnatz JD, et at: Plasma

insulin and triglyceride levels after hypothalamic lesions in

weaniing rats. Am J Physiol 216:1496, 1969

(5)

1982;70;275

Pediatrics

Alberto Hayek and Glenn T. Peake