0021-972X/97/$03.00/0
Journal of Clinical Endocrinology and Metabolism
Copyright © 1997 by The Endocrine Society
Vol. 82, No. 5
Printed in U.S.A.
Idiopathic Hypothalamic Diabetes Insipidus, Pituitary
Stalk Thickening, and the Occult Intracranial
Germinoma in Children and Adolescents*
SUDHA L. MOOTHA†, ANTHONY J. BARKOVICH, MELVIN M. GRUMBACH,
MICHAEL S. EDWARDS, STEPHEN E. GITELMAN, SELNA L. KAPLAN, AND
FELIX A. CONTE
Department of Pediatrics, Divisions of Neuroradiology (A.J.B.) and Pediatric Neurosurgery (M.S.E.),
University of California, San Francisco, California 94143
ABSTRACT
We report nine consecutive children and adolescents [five females
and four males; aged 2 yr 8 months (m) to 18 yr 1 m] studied over the
last 5 yr with idiopathic central diabetes insipidus. In addition to
vasopressin deficiency, anterior pituitary hormone deficiencies were
detected, either on evaluation at presentation or during follow-up
studies over the following 3 yr. Four patients had an increased concentration of plasma PRL. One patient had multiple pituitary hormone deficiencies at diagnosis, and two others developed the same by
21 m of follow-up. Brain magnestic resonance imaging scans, performed at presentation, were originally interpreted as normal in four
of nine patients, except for absence of the bright posterior pituitary
signal; after retrospective review, two of nine were considered normal.
All of the brain magnetic resonance imaging (MRI) scans showed
positive findings by 14 m of follow-up. The first abnormal finding in
all patients was isolated pituitary stalk thickening. Evaluation of
cerebrospinal fluid (CSF) for hCG was positive in three of eight evaluated patients; the three positive CSF values were found at presen-
C
ENTRAL diabetes insipidus (DI) is a rare disorder that
can result as a consequence of diverse etiologies, including malformations, autoimmune, infiltrative (e.g. neoplastic or histiocytsis) or traumatic processes, as well as mutations in the gene encoding arginine vasopressin. Idiopathic
central DI is a diagnosis of exclusion, and one that has been
made less frequently through the decades. Of 188 patients in
three series between 1955 and 1992, 15–30% with nonsurgical
central DI were classified as idiopathic (1–3). With the increasing sensitivity of diagnostic imaging techniques, a specific etiology for at least some patients with idiopathic central
DI can now be established. We report a study of nine consecutive patients referred with idiopathic central DI in whom
an isolated thickened pituitary stalk and anterior pituitary
hormone deficiencies were detected either at presentation or
during follow-up studies.
Received December 9, 1996. Revision received February 4, 1997.
Accepted February 12, 1997.
Address requests for reprints to: Dr. Felix Conte, Department of Pediatrics, Box 0136, University of California, San Francisco, California
94143-0136.
* This work was supported by a NIH grant from the NIDDK (5T32DK-07161) and the NIH Pediatric Clinical Research Center
(M01-RR-01271).
† Trainee in Pediatric Endocrinology under a program sponsored by
the NIDDK, NIH (Grant 5T32-DK-07161).
tation and 3 and 9 m after presentation. All eight patients assessed
were negative for CSF a-fetoprotein and cytology, and no patient had
serum tumor markers. Transsphenoidal biopsy of the lesion in seven
of nine patients showed a germinoma in six patients and inflammatory cells in one. The six patients with documented germinoma comprise 31% of the intracranial germinomas diagnosed in this age group
at the University of California-San Francisco during the last 5 yr. The
patient with mononuclear inflammatory cells on biopsy along with
one other patient have had spontaneous resolution of their stalk
thickening. So-called “idiopathic” central diabetes insipidus warrants
close follow-up to determine the etiology, especially if anterior pituitary hormone deficiencies are detected. Normal brain MRI scans or
scans that show isolated pituitary stalk thickening merit follow-up
with serial contrast enhanced brain MRI for the early detection of an
evolving occult hypothalamic-stalk lesion. CSF evaluation is recommended at presentation because elevated CSF hCG may precede MRI
abnormalities. (J Clin Endocrinol Metab 82: 1362–1367, 1997)
Subjects and Methods
Nine successive patients (five females and four males) were referred
to the pediatric endocrine unit at the University of California-San Francisco (UCSF) between September 1991 and September 1996 with polyuria and polydipsia of undetermined etiology (Table 1). Seven patients
presented without prior evaluation, and two were referred after initial
workup proved inconclusive. None had neurological deficits or a family
history of DI. Ages at presentation ranged from 2 yr 8 m months (m) to
18 yr 1 m (median age, 10.5 yr), and duration of symptoms at presentation was between 3 m to 4 yr (median, 17 m). All had polyuria and
polydipsia; patient 9 also had secondary amenorrhea of 17 m duration,
and patient 8 had arrested pubertal development. Although four of nine
patients complained of poor growth, only two of these four had a
demonstrable decrease in their growth rate. Patient 4 gave a family
history of goiter. No other history for autoimmune disease was elicited.
Past medical history was significant for patient 5 being small for gestational age and for in utero drug exposure in patient 3.
Methods
Polyuria and polydipsia were evaluated by either a modified water
deprivation test or measurement of simultaneous urinary and serum
osmolarity and antidiuretic hormone concentration, measured at SmithKline Laboratories (Dublin, CA) (4, 5). Insulin-like growth factor I (IGF-I)
and IGF-binding protein-3 (IGFBP-3) were measured by RIA at Nichols
Institute (San Juan Capistrano, CA). GH was measured by polyclonal
RIA. An ACTH test was performed in two of nine, and a GnRH test was
performed in four of seven pubertal aged patients. All stimulatory
testing was conducted according to previously described standard protocols (6, 7).
All subjects had brain magnetic resonance imaging (MRI) performed
1362
IDIOPATHIC HYPOTHALAMIC DIABETES
at presentation. T1-weighted sagittal and coronal 3-mm sections were
obtained without contrast enhancement in all patients. Eight MRIs were
also performed after gadolinium contrast enhancement (0.1 mg/kg);
four of the patients had postcontrast spinal MRI performed. Patients
who had normal brain MRI scans (with the exception of a posterior
pituitary bright spot) at presentation were followed with contrast-enhanced scans every 3– 6 m. All MRI scans were evaluated for this study
retrospectively by a pediatric neuroradiologist (A.J.B.).
Other diagnostic workup in all subjects included a skeletal survey to
rule out Langerhans’ histiocytosis. The tumor marker hCG was measured in the cerebrospinal fluid (CSF) and serum. This assay was performed using a UCSF clinical laboratory assay preparation specific for
the b-subunit of hCG (normal value, ,2 IU/L). Serum and CSF a-fetoprotein was determined by enzyme immunoassay; CSF cytology was
assessed on a concentrated sample.
Results
Endocrine profiles at presentation are summarized in Table 2. All nine patients had central DI. Six had inappropriately low arginine vasopressin concentrations (,1 ng/L) and
low urinary osmolarity for the concentration of serum sodium and osmolarity measured in a random sample. The
remaining three patients failed a water deprivation test. All
patients concentrated their urine appropriately in response
to exogenously administered 8-desamino-arginine vasopressin. Plasma PRL was elevated in four of eight patients. The
peak GH concentrations after arginine and l-DOPA stimulations were less than 7 ng/mL in seven of eight patients
tested; one of these seven patients had borderline GH deficiency and was growing at an appropriate rate. Plasma IGF-I
was below the expected value for age in three of seven
1363
patients. Patient 8 had mild hypothalamic hypothyroidism at
presentation. Patients 7, 8, and 9 required cortisol replacement 21 m after presentation, 3 m after presentation, and at
presentation, respectively. Two pubertal aged patients had
prepubertal LH response to GnRH; one of them presented
with secondary amenorrhea of 17 m duration.
The initial brain MRIs were evaluated by various radiologists at UCSF and elsewhere (Table 3). These scans were
interpreted as normal in four of nine patients (2– 4, 8). Subsequently, all scans were reviewed by A.J.B., who found that
the scans in two of nine subjects (patients 2 and 4) were
normal and that seven of nine patients had diffuse thickening
of the pituitary infundibulum (Fig. 1, a and b). Later, some
exhibited lesion extension upward into the floor of the third
ventricle or downward into the pituitary fossa (Fig. 2). The
two patients with normal initial scans exhibited stalk enlargement at 3 m and 8 m follow-up, respectively. All patients had absence of the hyperintense signal normally seen
in the neurohypophysis on the T1-weighted images. The four
spinal MRI scans were interpreted as normal.
hCG and a-fetoprotein were not detected in the serum of
any of the subjects (Table 4). CSF was assessed in eight of nine
patients (Table 4). Two were evaluated at presentation, and
the others were examined between 3– 46 m after presentation, either when the MRI became positive or when the mass
lesion was biopsied. The CSF hCG was positive at low concentrations in three of eight subjects (patient 9 at presentation; patient 4, 3 m after presentation; patient 5, 9 m after
presentation). The CSF was negative for hCG in the other five
TABLE 1. Clinical findings
Patient
no.
Age at
presentation
Sex
Ht, cm (SD)
Signs/symptoms
Durationa
BA at CAb
1
2
3
4
5
6
7
8
9
2 yr 8 m
5 yr 10 m
8 yr 2 m
10 yr 5 m
10 yr 6 m
11 yr 6 m
11 yr 9 m
12 yr 11 m
18 yr 1 m
M
F
M
F
M
M
F
F
F
93.5 (20.5)
112.3 (21)
130.5 (mean)
116 (24.6)
127.5 (22.5)
127.9 (23.2)
129 (23.6)
152.1 (21.25)
161.7 (21)
PD/PU*
PD/PU
PD/PD/enuresis
PD/PU/poor growth
PD/PU/nocturia
PD/PU/poor growth
PD/PU/poor growth/enuresis
PD/PU/pubertal arrest
PD/PU, secondary amenorrhea
3m
5m
7m
4 yr
2m
3 yr
2 yr
25 m
17 m
3 yr 6 m at 2 yr 8 m
6 yr 6 m at 8 yr 3 m
8 yr 2 m at 8 yr 2 m
7 yr 10 m at 10 yr 3 m
10 yr at 10 yr 6 m
9 yr at 11 yr 10 m
10 yr at 11 yr 8 m
13 yr 5 m at 14 yr 9 m
Not done
PD/PU, Polydipsia and polyuria. BA and CA, Bone age and chronological age, respectively. m, Months.
Duration of symptoms at presentation.
b
Some studies were performed by referring physicians.
a
TABLE 2. Hormonal studies at presentation
Patient
no.
Age at
presentation
Sex
PRL
(ng/mL)
IGF-I
(ng/mL)
IGFBP-3
(mg/L)
ATT GH
(ng/mL)
1
2
3
4
5
6
7
8
9
2 yr 8 m
5 yr 10 m
8 yr 2 m
10 yr 5 m
10 yr 6 m
11 yr 6 m
11 yr 9 m
12 yr 11 m
18 yr 1 m
M
F
M
F
M
M
F
F
F
13
18
ND
17
30
8
19.7
61
57
54
67
71
,10
102
53
ND
89
80
2.6
1.6
ND
0.5
2.9
1.6
ND
2.1
2.3
20.4
1.8
1
4.1
6.1
,0.2
,1
ND
1.4
,19
b
b
.7
.7
Normal values
ND, Not done; m, months.
a
Early morning cortisol measured in patients 2, 7, and 8; others were random values.
b
Refer to normal values for age and sex.
FT4
(pmol/L)
Cortisol
(mg/dL)a
19.8
2.5
1.6
2
6.7
0.2
1.7
ND
1.7
16
24
18
15
14
15
ND
8
10
13
12
6
22
11
17
9
2
1
9 –24
5–20
L-DOPA
GH
(ng/mL)
1364
JCE & M • 1997
Vol 82 • No 5
MOOTHA ET AL.
TABLE 3. Imaging profile
Patient
no.
Sex
Age
1
2
3
4
5
M
F
M
F
M
2 yr 8 m
5 yr 8 m
8 yr 2 m
10 yr 5 m
10 yr 8 m
Thick stalk
Normal
Normal
Normal
Thick stalk
Thick stalk
Normal
Thick midstalk
Normal
Thick stalk
6
M
11 yr 4 m
7
8
F
F
11 yr 8 m
13 yr
Thick stalk and small
pituitary
Thick stalk (no contrast)
Normal
Thick stalk and small
pituitary
Thick stalk
Thick stalk
9
F
18 yr
Thick stalk and abnormal
enhancement
Thick stalk
Original head MRI
interpretation
Retrospective evaluation
Intervala
Follow-up
6m
8m
9m
3m
12 m
Increased stalk thickening
Thick stalk, full pineal
Normal stalk and pineal
Thick, nodular stalk
Suprasellar and
infundibular mass
Reduced stalk thickening
Pending
Germinoma
Not done
Germinoma
Germinoma
Extensive enhancement
Enhancement of 3rd
ventricular floor
Normal scan status post-bx
and tx
Germinomac
Germinoma
3m
11 m
14 m
8m
Biopsy results
Inflammationb
Germinoma
m, Months; bx, biopsy; tx, therapy.
Interval from presentation to first follow-up MRI with a notable change.
b
Mononuclear inflammatory cells.
c
First biopsy was of inflammatory cells; second biopsy was germinoma (see text).
a
FIG. 1. a and b, Saggital view brain MRI scans with contrast enhancement in patients 2 and 6. Pituitary stalk thickening is designated by the
arrows.
patients, and it was negative for a-fetoprotein and cytology
in all eight patients.
Biopsy of the stalk lesion was performed in seven of nine
patients when their brain MRI scans showed extension of the
lesion (to the hypothalamus, including the floor of the third
ventricle) or if the CSF was positive for hCG. The biopsy was
performed transsphenoidally in the six patients who had
extension of the lesion to the pituitary fossa or to the hypothalamus. Patient 6 did not have extension of the lesion to the
pituitary and had a craniotomy for biopsy. The transsphenoidal approach was preferred because of its lower morbidity and mortality compared to a craniotomy. All the patients
were biopsied by the pediatric neurosurgeon (M.S.E.), and
none had any operative complications. Six of the seven biopsied lesions were positive for germinoma (see Fig. 2). The
biopsy of patient 6 showed mononuclear inflammatory infiltrates with no evidence of neoplastic cells. Patient 7 had a
biopsy 2 yr after presentation; it was a small nondiagnostic
sample, showing chronic inflammation and a granulomatous
process. Over the ensuing 2 yr she had further progression
of her lesion, and a second biopsy was consistent with
germinoma.
Focal brain irradiation therapy (5000 cGy) was given to
five of the patients with biopsy-proven germinoma, and they
now have normal follow-up brain MRI scans. One patient
(patient 5) with a germinoma is receiving chemotherapy with
irradiation at another institution. Patient 6, whose biopsy
showed an inflammatory lesion, had a decrease in stalk thickening at 3 m follow-up, but he continues to have evidence of
multiple pituitary hormone deficiencies. He is Tanner stage
IDIOPATHIC HYPOTHALAMIC DIABETES
1365
FIG. 2. Serial coronal views of the contrast-enhanced brain MRI scans in patient 8, with arrows pointing to the pituitary stalk. The first scan
was performed at presentation and is normal. The second scan was performed after the patient failed follow-up for 14 m. She had pituitary
stalk thickening, with extension of the lesion to the floor of the third ventricle. The third scan is a normal scan performed after transsphenoidal
biopsy and irradiation treatment.
TABLE 4. Tumor markers
Patient
no.
Sex
Age at
presentation
Time
intervala
CSF hCGb
(IU/L)b
CSF a-fetoprotein
(mg/L)b
1
2
3
4
5
6
7
8
9
M
F
M
F
M
M
F
F
F
2 yr 8 m
5 yr 10 m
8 yr 2 m
10 yr 5 m
10 yr 6 m
11 yr 6 m
11 yr 9 m
12 yr 11 m
18 yr 1 m
6m
14 m
ND
3m
9m
P
46 m
19 m
P
,2
,2
ND
3
9
2c
,1.6
,2
6
,0.5
1.3
ND
,0.5
,8.2
,0.5
,0.5
,0.5
0.5
a
Time interval from presentation (P).
Normal values: CSF hCGb, less than 2 IU/L; a-fetoprotein, less
than 8.2 mg/L.
c
Repeat assays on same sample gave a result of less than 2 IU/L.
b
I for genital development and pubic hair at the age of 12 yr.
Patient 3 (not biopsied) has had resolution of his lesion on
MRI, but continues to have central DI and GH deficiency
without a decrease in the growth rate. Both of these patients
are under observation with serial scans. Patient 1 is the
youngest and most recent patient. He has isolated stalk thickening and no anterior pituitary hormone deficiencies, and is
negative for serum and CSF tumor markers; he has not had
a biopsy and is being followed with serial scans.
Discussion
All nine patients in this series presented with central DI
and no clear etiology; however, each patient had a variable
course with respect to duration of symptoms, presence of
pituitary hormone deficiencies, and rate of progression. Anterior pituitary deficiencies were unmasked either after investigation at presentation or with follow-up studies. Four
patients (no. 6 –9) had multiple pituitary hormone deficiencies after having had symptoms of central DI for 18 m to 3
yr. In contrast, patient 4 presented with a 4-yr history of
central DI and has only GH deficiency, with other anterior
pituitary functions intact. The finding of anterior pituitary
deficiencies in the presence of central DI greatly increased
our suspicion of an occult hypothalamic lesion (8).
Our experience indicates that close clinical follow-up accompanied by serial brain MRIs are essential to determine a
diagnosis in patients with so-called idiopathic DI. Germinoma was diagnosed in six of the nine patients because brain
MRIs were performed on a frequent (every 3– 6 m) basis, and
a biopsy was performed when the lesion progressed or if the
CSF markers were positive. Contrast enhancement was useful in the identification of subtle extension of the lesion. A
time lag of up to 14 m between clinical presentation and
development of brain MRI abnormalities was seen in our
series. Reports of patients followed by computerized tomography of the head indicate that imaging follow-up for many
years may be needed to establish a diagnosis (9 –11). In our
series it took between 1– 46 m from presentation to establish
a biopsy-proven diagnosis. Scans at presentation were interpreted as normal in four patients (no. 2, 3, 4, and 8), and
even when reviewed retrospectively, scans of patients 2 and
4 were read as normal without evidence of a mass lesion.
With follow-up brain MRIs initially performed every 3– 6 m,
we found that the rate of tumor progression was variable; it
took only 3 m for patient 4 to progress from a normal scan
to stalk thickening, whereas others in our series have taken
up to 15 m for MRI changes to evolve. Discrepancies in MRI
interpretation exist because stalk thickening is a nonspecific
finding, and there are no useful published standards for
pituitary stalk size in children and adolescents.
The importance of isolated pituitary stalk thickening became apparent in the evaluation of this consecutive group of
patients. In all nine of our patients, stalk thickening was the
first abnormal finding on brain MRI. Six of these patients had
biopsy-proven germinoma, and two patients had spontaneous resolution of the stalk thickening, one is being followed
at this time. The underlying etiology for the two patients with
spontaneous resolution of the stalk thickening is unclear.
Patient 6, one of these two patients, has multiple pituitary
hormone deficiencies and a pituitary stalk biopsy that
showed mononuclear inflammatory cells. Patient 3 was followed without biopsy and had spontaneous resolution of his
stalk lesion 9 m after presentation.
The differential diagnosis of pituitary stalk thickening
includes germinoma; however, germinomas commonly
present with a hypothalamic/pineal mass and in the past
only infrequently have been noted to have isolated stalk
thickening (8, 12). Other considerations in the evaluation of
pituitary stalk thickening in children or adolescents include
Langerhans’ histiocytosis and putative antivasopressin cell
antibody-induced central DI (10, 12–15). A thickened pituitary stalk was also common in a series of 17 adult patients
with lymphocytic infundibuloneurohypophysitis that resolved by the 2-yr follow-up visit in all patients, which sug-
1366
JCE & M • 1997
Vol 82 • No 5
MOOTHA ET AL.
FIG. 3. Algorithm for the evaluation of
central DI with no clear etiology.
gested a self-limited process (16). Only isolated cases of this
disorder have been described in children; those suspected of
having such a lesion, which was aggressive in one child (17),
need continued monitoring to confirm the diagnosis and to
follow its progress (17, 18).
Intracranial germ cell tumors comprise 7.8% of primary
pediatric brain tumors (19). Our series of 6 patients with a
documented intracranial germinoma make up 6 of 19 (31%)
of all the children and adolescents with intracranial germinomas who were studied at UCSF during the same 5-yr
period. Although serum hCG and a-fetoprotein were undetectable in all of our patients, evaluation of the CSF hCG was
useful. Three of 7 patients (no. 4, 5, and 9) had positive, but
low, hCG concentrations (normal, ,2 IU/L) in their CSF, a
finding consistent with some pure germ cell tumors. Patient
4 presented with a normal scan, but progressed to stalk
nodularity and thickening on MRI at the 4 m follow-up, at
which time the CSF hCG level was 3 IU/L. Patient 5 presented with a large stalk and on the 9 m follow-up scan had
extension of the lesion to the pineal gland and floor of the
third ventricle; the CSF hCG level was 9 IU/L at that time.
Patient 9 had a CSF hCG value of 6 IU/L when she presented
with mild stalk enlargement. Others have reported positive
CSF cytology 1 yr before radiological evidence of germinoma
(20). We recommend CSF evaluation at presentation, and
biopsy when the lesion extends beyond the stalk or if tumor
markers are positive. Irradiation treatment should be given
only after a biopsy-proven diagnosis is established.
We note a decrease in the frequency of the diagnosis of
idiopathic central DI. An underlying etiology must be pursued especially in the presence of anterior pituitary hormone
deficiencies that increase the suspicion of an occult hypothalamic lesion (8). A normal brain MRI or an MRI with only
pituitary stalk enlargement calls for serial scanning. An approach to a diagnosis in patients with idiopathic CDI is given
in the algorithm in Fig. 3. Even though a brain MRI without
contrast better detects the absence of the bright posterior
pituitary signal, this observation only confirms the known
clinical diagnosis of (nongenetic) central DI (12). More important in our view is the usefulness of a contrast-enhanced
brain MRI in the detection of subtle abnormalities in the
hypothalamic-pituitary region. For example, an MRI without
contrast can miss subtle changes in the floor of the third
ventricle that are crucial to establishing an early diagnosis.
CSF evaluation for cytology and markers is recommended at
presentation, given the possibility of a time lag before evidence of a lesion appears on the brain MRI (20). In cases with
a normal brain MRI, our experience has shown that it may
take up to 1 yr for the scans to become positive and up to 4
yr for a diagnosis to be established by biopsy. Accordingly,
we recommend that a contrast-enhanced MRI brain scan be
performed every 3– 6 m for the first 1–2 yr after presentation,
followed by a brain MRI every 6 m for at least 5 yr if a lesion
is not detected. This approach will assure early detection of
an expanding lesion that may be readily treatable.
References
1. Pomarede R, Czernichow P, Rappaport R, et al. 1980 Le diabète pitressosensible de l’enfant. Arch Fr Pediatr. 37:37– 44.
2. Greger NG, Kirkland RT, Clayton GW, Kirkland JL. 1986 Central diabetes
insipidus. Am J Dis Child. 140:551–554.
3. Wang LC, Cohen ME, Duffner PK. 1994 Etiologies of central diabetes insipidus in children. Pediatr Neurol. 11:273–277.
4. Frasier SD, Kutnik LA, Schmidt RT, Smith FG. 1967 A water deprivation test
for the diagnosis of diabetes insipidus in children. Am J Dis Child. 114:157–160.
5. Richman RA, Post EM, Notman DD, Hochberg Z, Moses AM. 1981 Simplifying the diagnosis of diabetes insipidus in children. Am J Dis Child.
135:839 – 841.
6. Goodman HG, Grumbach MM, Kaplan SL. 1968 Growth and growth hormone. II. A comparison of isolated growth hormone deficiency and multiple
hormone deficiencies in 35 patients with hypopituitary dwarfism. N Engl
J Med. 278:57.
7. Grumbach MM, Roth JC, Kaplan SL, Kelch RP. 1974 Hypothalamic-pituitary
regulation of puberty in man: evidence and concepts derived from clinical
IDIOPATHIC HYPOTHALAMIC DIABETES
8.
9.
10.
11.
12.
13.
14.
research. In: Grumbach MM, Grave GD, Mayer FE, eds. The control of the onset
of puberty. New York: Wiley and Sons; 115–166.
Sklar CA, Grumbach MM, Kaplan SL, Conte FA. 1981 Hormonal and metabolic abnormalities associated with central nervous system germinoma in
children and adolescents and the effect of therapy: report of 10 patients. J Clin
Endocrinol Metab. 52:9 –16.
Cianfarani S,, Nicholl RM, Medbach S, Charlesworth MC, Savage MO. 1993
Idiopathic hypothalamus-pituitary dysfunction: review of five cases. Horm
Res. 39:47–50.
Czernichow P, Pomarede R, Basmaciogullari A, Brauner R, Rappaport R.
1985 Diabetes insipidus in children. III. Anterior pituitary dysfunction in
idiopathic types. J Pediatr. 106:41– 44.
Sherwood MC, Stanhope R, Preece MA, Grant DB. 1986 Diabetes insipidus
and occult intracranial tumours. Arch Dis Child. 61:1222–1235.
Tien R, Kucharczyk J, Kucharczyk W. 1991 MR imaging of the brain in
patients with diabetes insipidus. Am J Neuroradiol. 12:533–542.
Egeler RM, D’Angio GJ. 1995 Langerhans cell histiocytosis. J Pediatr. 127:1–11.
Tien RD, Newton TH, McDermott MW, Dillon WP, Kucharczyk J. 1990
15.
16.
17.
18.
19.
20.
1367
Thickened pituitary stalk on MR images in patients with diabetes insipidus and
Langerhans cell histiocytosis. Am J Neuroradiol. 11:703–708.
Scherbaum WA, Czernichow P, Bottazzo GF, Doniach D. 1985 Diabetes
insipidus in children. IV. A possible autoimmune type with vasopressin cell
antibodies. J Pediatr. 107:922–925.
Imura H, Nakao K, Shimatsu A, et al. 1993 Lymphocytic infundibuloneurohypophysitis as a cause of central diabetes insipidus. N Engl J Med.
329:683– 689.
Bettendorf M, Frank S, Grulich-Henn J, Engenhart R, Heinrich U. Lymphocytic
hypophysitis: malignant clinical course in prepubertal girl (Abstract no. P1-434;
P243). Proc of the 10th Annual Meet of the Int Congr of Endocrinol. 1996.
Thodou E, Asa SL, Kontogeorgos G, Kovacs K, Horvath E, Ezzat S. 1995
Clinical case seminar: lymphocytic hypophysitis: clinicopathological findings.
J Clin Endocrinol Metab. 80:2302–2311.
Horowitz MB, Hall WA. 1991 Central nervous system germinomas: a review.
Arch Neurol. 48:652– 657.
Pomarede R, Czernichow P, Finidori J, et al. 1982 Endocrine aspects of
tumoral markers in intracranial germinomas. J Pediatr. 101:374 –378.