Central diabetes insipidus is rare in. children and young adults, and central diabetes insipidus of the probable cause of central diabetes in …
The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
CENTRAL DIABETES INSIPIDUS IN CHILDREN AND YOUNG ADULTS
MOHAMAD MAGHNIE, MD, PHD, GIANLUCA COSI, MD, EUGENIO GENOVESE, MD, MARIA LUISA MANCA-BITTI, MD, AMNON COHEN, MD, SILVIA ZECCA, MD, CARMINE TINELLI, MD, MASSIMO GALLUCCI, MD, SERGIO BERNASCONI, MD, BRUNETTO BOSCHERINI, MD, FRANCESCA SEVERI, MD, AND MAURIZIO ARICÒ, MD
ABSTRACT
Background Central diabetes insipidus is rare in children and young adults, and up to 50 percent of cases are idiopathic The clinical presentation and the long-term course of this disorder are largely undefined Methods We studied all 79 patients with central diabetes insipidus who were seen at four pediatric endocrinology units between 1970 and 1996 There were 37 male and 42 female patients whose median age at diagnosis was 70 years range, 01 to 248 All patients underwent magnetic resonance imaging MRI and periodic studies of anterior pituitary function The median duration of follow-up was 76 years range, 16 to 262 Results The causes of the central diabetes insipidus were Langerhans-cell histiocytosis in 12 patients, an intracranial tumor in 18 patients, a skull fracture in 2 patients, and autoimmune
polyendocrinopathy in 1 patient; 5 patients had familial disease The cause was considered to be idiopathic in 41 patients 52 percent In 74 patients 94 percent the posterior pituitary was not hyperintense on the first MRI scan obtained, and 29 patients 37 percent had thickening of the pituitary stalk Eighteen patients had changes in the thickness of the pituitary stalk over time, ranging from normalization six patients or a decrease in thickness one patient to further thickening seven patients or thickening of a previously normal stalk four patients Anterior pituitary hormone deficiencies, primarily growth hormone deficiency, were documented in 48 patients 61 percent a median of 06 year range, 01 to 180 after the onset of central diabetes insipidus Conclusions Most children and young adults with acquired central diabetes insipidus have abnormal findings on MRI scans of the head, which may change over time, and at least half have anterior pituitary hormone deficiencies during follow-up N Engl J Med 2000;343:998-1007
2000, Massachusetts Medical Society
cell histiocytosis 2,3; inflammatory, autoimmune, and vascular diseases4,5; trauma resulting from surgery or an accident 6; and in
rare cases, genetic defects in the synthesis of vasopressin that are inherited as autosomal dominant or X-linked recessive traits7-9 However, 30 to 50 percent of cases are considered idiopathic10-15 Arginine vasopressin is transported from the hypothalamus through the neural component of the pituitary stalk and stored in nerve terminals in the posterior pituitary In many normal subjects, the posterior pituitary is hyperintense on sagittal T1-weighted magnetic resonance imaging MRI The absence of this finding serves as a nonspecific indicator of central diabetes insipidus,2-4,16-18 but the frequency of hyperintensity decreases with age in normal subjects19 The finding of a thickened infundibulum or pituitary stalk or both, although not specific, suggests the presence of an infiltrative disease1-5,14,16,17,20 The frequency of these radiologic abnormalities in patients with central diabetes insipidus is poorly defined We therefore investigated the clinical presentation, the morphologic characteristics of the pituitary region on MRI, and the size of the pituitary stalk over time in patients who had central diabetes insipidus with a variety of causes
METHODS
Patients We reviewed the
data bases of four pediatric endocrinology units to identify all patients who had documented cases of central diabetes insipidus between 1970 and 1996 and to record family histories, presenting features, hormone concentrations, and the results of imaging and genetic studies at diagnosis and during follow-up Seventy-nine patients 37 males and 42 females ranging in age from 01 to 248 years at diagnosis were seen at these units during this interval, 45 since 1991 The median age at diagnosis was 70 years The clinical protocols used at each center were approved by the appropriate review boards, and written informed consent for all invasive procedures was obtained from the patients or their parents or guardians The median duration of follow-up through May 1998 was 76 years range, 16 to 262 All patients had permanent central diabetes insipidus and were being treated with desmopressin acetate desamino-D-arginine-8-vasopressin, two or
C
998
ENTRAL diabetes insipidus is a heterogeneous condition characterized by polyuria and polydipsia due to a deficiency of arginine vasopressin In many patients, especially children and young adults, it is caused by the destruction or degeneration of the
neurons that originate in the supraoptic and paraventricular nuclei of the hypothalamus Known causes of these lesions include germinoma1 and craniopharyngioma2; LangerhansOc to b er 5 , 2 0 0 0
From the Departments of Pediatrics MM, GC, FS, MA, Radiology EG, and BiometryScientific Direction CT, University of Pavia and Istituto di Ricovero e Cura a Carattere Scientifico Policlinico San Matteo, Pavia; the Department of Pediatrics, University of Rome Tor Vergata, Rome MLM-B, BB; the Department of Pediatrics, Istituto di Ricovero e Cura a Carattere Scientifico G Gaslini Institute, Genoa AC, SZ; the Department of Radiology, University of LAquila, LAquila MG; and the Department of Pediatrics, University of Modena, Modena SB — all in Italy Address reprint requests to Dr Maghnie at the Department of Pediatrics, University of Pavia, IRCCS Policlinico S Matteo, 27100 Pavia, Italy, or at maghnie@smatteopvit
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
C E N T R A L D I A B ET E S I N S I P I D U S I N C H I L D R E N A N D YO U N G A D U LT S
three times daily,
either intranasally in the case of 78 patients or orally in the case of 1 patient Anterior pituitary hormone deficiencies were treated as appropriate Diagnosis and Classification of Central Diabetes Insipidus The diagnosis of central diabetes insipidus was based on a history of polyuria and polydipsia, the results of a physical examination, laboratory evidence of arginine vasopressin deficiency, and imaging studies of the brain and pituitary gland Cases were further classified on the basis of the probable cause of central diabetes insipidus: Langerhans-cell histiocytosis biopsy-proven, an intracranial tumor including those occurring after tumor resection, familial disease at least one additional family member had to be affected, skull fracture ie, post-traumatic disease, and autoimmune polyendocrinopathy Cases with no identifiable cause were considered idiopathic Posterior Pituitary Function In 17 patients the diagnosis of central diabetes insipidus was based on the clinical findings of polyuria and polydipsia, an osmolality of less than 300 mOsm per kilogram of water or a specific gravity of less than 1010 in a 24-hour urine specimen, and an increase in urinary osmolality in
response to desmopressin acetate In 62 patients, a water-deprivation test was performed in which water was withheld for 4 to 14 hours 21 A ratio of urinary osmolality to plasma osmolality of 10 or less was taken to indicate the presence of complete central diabetes insipidus, and a ratio of more than 10 but less than 14 was taken to indicate partial central diabetes insipidus Serum sodium, plasma and urinary osmolality or specific gravity, and plasma immunoreactive arginine vasopressin were measured at the beginning time 0 and the end of the test At the latter time, the plasma arginine vasopressin concentrations in normal subjects ranged from 2 to 5 pg per milliliter 18 to 46 pmol per liter Anterior Pituitary Function Anterior pituitary function was assessed in 64 patients In 58 of these 64 patients, serum growth hormone was measured before and 30, 60, 90, and 120 minutes after the administration of arginine 05 g per kilogram of body weight, given intravenously over a period of 30 minutes, insulin 01 U per kilogram, given intravenously, or levodopa 500 mg per square meter of body-surface area, given orally Patients with serum growth hormone concentrations of less than 10 ng per
milliliter and a deceleration in the rate of growth were considered to have growth hormone deficiency The pituitarythyroid axis was assessed every 6 to 12 months by measuring serum free thyroxine, free triiodothyronine, and thyrotropin Hypothyroidism was defined as a low or low-normal serum thyrotropin concentration and low serum free thyroxine and free triiodothyronine concentrations Plasma corticotropin and serum cortisol were measured in the morning at presentation in 57 patients and every 6 to 12 months thereafter in 41 patients Corticotropin deficiency was defined by either a serum cortisol concentration of less than 36 g per deciliter 100 nmol per liter in the morning or a peak serum cortisol concentration of less than 20 g per deciliter 550 nmol per liter during insulin-induced hypoglycemia Serum follicle-stimulating hormone and luteinizing hormone were measured before and 30, 60, and 120 minutes after the intravenous administration of 100 g of gonadotropinreleasing hormone per square meter in patients who were thought to have hypogonadotropic hypogonadism Hypogonadism was diagnosed in boys and girls who had no pubertal development and no increase in serum
follicle-stimulating hormone and luteinizing hormone in response to gonadotropin-releasing hormone Ultrasonography was used to identify female patients with a prepubertal uterus Serum prolactin was measured in 54 patients at presentation Standard radioimmunoassays were used to measure all hormones
Imaging Studies Computed tomographic CT scans were obtained in 22 patients at presentation MRI scans were obtained in all patients, in 57 of them at presentation Contrast enhancement with gadolinium was used in 65 of the 79 patients A spinecho technique with either a 05-T system in the case of 20 patients or a 15-T superconductive system in the case of 59 patients was used Sagittal and coronal T1-weighted images were obtained in which sections were 30 mm thick In 41 patients, follow-up MRI scans were obtained every 4 to 12 months for a median of 4 years range, 02 to 103 On these imaging studies, a normal pituitary stalk was defined as one that was less than 30 mm thick6 Thickened stalks were graded as minimally thickened 30 to 45 mm, moderately thickened 46 to 65 mm, or severely thickened more than 65 mm The size of the anterior pituitary was assessed in serial imaging studies in the
sagittal plane Other Studies Twenty-three of the patients with idiopathic central diabetes insipidus underwent skeletal surveys to rule out Langerhans-cell histiocytosis as the cause Ten patients underwent transfrontal biopsy of the pituitary stalk Statistical Analysis The clinical and laboratory features of the patients were compared with the use of chi-square tests One-way analysis of variance was used to compare mean values between groups Odds ratios, with 95 percent confidence intervals, were calculated according to Woolf s method, and the significance of any deviation from unity was estimated with use of Fishers exact test We used the Kaplan Meier method to estimate the probability of a first anterior pituitary hormone deficiency among patients with idiopathic central diabetes insipidus or central diabetes insipidus associated with Langerhans-cell histiocytosis Among the patients who had an anterior pituitary hormone deficiency at presentation, we estimated the probability that additional hormonal deficiencies would develop during follow-up Comparisons between groups were performed with use of the log-rank test All statistical tests were two-sided The Statistica 51 program
Statsoft, Tulsa, Okla was used for all statistical computations
RESULTS
The characteristics of the 79 patients are reported in Table 1 according to the cause of central diabetes insipidus Most patients were children when the disease was diagnosed, and the numbers of male and female patients were similar There was a significant difference P0001 in the age at presentation among the groups Fig 1 All five patients with family histories of polyuria and polydipsia were tested, and three had a mutation of the arginine vasopressin neurophysin II gene22 In 10 patients, central diabetes insipidus developed during an infectious illness or less than two months afterward varicella in 5 patients, mumps in 2 patients, and measles, toxoplasmosis, and hepatitis B in 1 patient each Thirty-two patients had symptoms and signs other than polyuria and polydipsia: headache in 9 patients, a visual defect in 5, growth retardation in 13, fatigue in 4, and failure to thrive in 1 The patient with autoimmune polyendocrinopathy, which was diagnosed during the second year of life, did not have polyuria and polydipsia until the age of 248 years
Vol ume 343 Numb e r 14
999
Downloaded from wwwnejmorg on
November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
1000
TABLE 1 CHARACTERISTICS
OF AND WITH
79 CHILDREN
YOUNG ADULTS
CENTRAL DIABETES INSIPIDUS
CAUSE
AT PRESENTATION AT FIRST
AND
FINDINGS MRI
AT LAST
NO OF PATIENTS SEX F/M AGE YR MRI
NO OF PATIENTS WITH SYMPTOMS OTHER THAN POLYURIA AND POLYDIPSIA
NO OF PATIENTS WITH ENDOCRINE DEFICIENCIES AT LAST VISIT
SIZE OF ANTERIOR PITUITARY ON LAST MRI SCAN
Oc to b er 5 , 2 0 0 0
23/18 8/10 Growth retardation in 7; headache in 3; fatigue in 3 50 05149 39 0168 34 11196 100 11213 118 27227 109 01299 118 01299 93 11263 141 50263 64 01149 73 45118 80 01299 74 46118 124 01299 118 48131 12/7 3/1 7/5 Not applicable GH in 9; GH and TSH in 3; GH and FSH or LH in 3; GH, TSH, and FSH or LH in 2 GH in 2 GH in 1 GH in 6; GH and TSH in 1; GH and FSH or LH in 1; panhypopituitarism in 1 Growth retardation in 1; headache in 2; fatigue in 1 None 6/12 3/3 Growth retardation in 3; visual defect in 2 85 58138 101 73138 116 58227 101 81139 121 84257 114 90139 2/4 Visual defect in 3; headache in 3; growth retardation in 2 Not applicable 94 58173 75 58120 128 58227
150 84276 1/5 92 73203 145 99216 GH and TSH in 1; GH and ACTH in 1; GH, TSH, and FSH or LH in 2; GH, TSH, and ACTH in 2 Panhypopituitarism in 4; GH, TSH, and ACTH in 1; GH, ACTH, and FSH or LH in 1 Panhypopituitarism in 3; GH, TSH, and ACTH in 1; GH, ACTH, and FSH or LH in 1 45 11355 148 139157 248 248 81 11355 178 168188 248 None 5/0 1/1 0/1 Not applicable None Failure to thrive in 1; headache in 1 14 0870 119 100138 Not applicable Normal
Idiopathic Thickened pituitary stalk
41 52 18 23
Reduced in 17; normal in 1
Normal pituitary stalk
19 24
CNS malformations
4 5
Normal in 16; reduced in 3 Normal in 2; empty sella in 2 Normal in 9; reduced in 3
Langerhans-cell histiocytosis
12 15
Intracranial tumor Germinoma
18 23 6 8
Increased in 4; normal in 2
Craniopharyngioma
6 8
Reduced in 4; could not be evaluated in 2
The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
Post-resection
6 8
Reduced in 6
Familial
5 6
Normal in 4; reduced in 1 Panhypopituitarism in 1; Reduced in 1; normal ACTH in 1 in 1
Post-traumatic skull fracture
2 3
Autoimmune polyendocrinopathy
1 1
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without
permission Copyright 2000 Massachusetts Medical Society All rights reserved
MRI denotes magnetic resonance imaging, GH growth hormone, TSH thyrotropin, FSH follicle-stimulating hormone, LH luteinizing hormone, CNS central nervous system, and ACTH corticotropin Values for age are medians and ranges
One patient had isolated Langerhans-cell histiocytosis in the central nervous system
C E N T R A L D I A B ET E S I N S I P I D U S I N C H I L D R E N A N D YO U N G A D U LT S
22 20 18
Age at Diagnosis yr
16 14 12 10 8 6 4 2 0 Intracranial tumor 18 patients Familial 5 patients Idiopathic 41 patients Langerhans-cell histiocytosis 12 patients
Figure 1 Age at Diagnosis According to the Cause of Central Diabetes Insipidus The patients who did not have an intracranial tumor were significantly younger at diagnosis than those who did P0001 for all comparisons The horizontal lines indicate the medians
Posterior Pituitary Function
Responses to the water-deprivation test were compatible with the presence of complete central diabetes insipidus in 61 patients and partial central diabetes insipidus in 1 patient Plasma vasopressin values after the test ranged from 05 to 15 pg per milliliter
05 to 14 pmol per liter in the 21 patients in whom it was evaluated Urinary osmolality increased to 470 to 780 mOsm per kilogram in response to exogenous desmopressin acetate in the 62 patients tested
MRI Findings
The posterior pituitary was not hyperintense on the first MRI scan obtained in 74 of 79 patients 94 percent Fig 2 The pituitary stalk was thickened in 29 of the 79 patients 37 percent, 18 of whom had idiopathic diabetes insipidus Fig 2C, 5 of whom had Langerhans-cell histiocytosis Fig 3, 5 of whom had a germinoma, and 1 of whom had autoimmune polyendocrinopathy In 22 patients, the entire pituitary stalk was thickened, whereas in 7 patients, only the proximal portion was thickened Twelve of 79 patients had MRI findings suggestive of hypothalamic pituitary tumors: 11 patients had a craniopharyngioma post-resection in 5 and 1 patient had a post-resection germinoma The pituitary stalk was normal in 36 patients 46 percent, 19 of whom had idiopathic diabetes insipidus, 7 of whom had Langerhans-cell histiocytosis, 5 of whom had familial diabetes insipidus, 4 of whom had central nervous system malfor-
mations the empty-sella syndrome in 2, holoprosencephaly in 1, and cerebral
aneurysm in 1, and 1 of whom had a germinoma The pituitary stalk was transected in the two patients with post-traumatic central diabetes insipidus After a median follow-up of 15 years range, 02 to 75, posterior pituitary hyperintensity was no longer present on the MRI scans of the five patients 6 percent with hyperintensity on the initial scan, three of whom had idiopathic diabetes insipidus, one of whom had Langerhans-cell histiocytosis,23 and one of whom had familial diabetes insipidus Eighteen patients had changes in their pituitary stalks over time Table 2 Four patients whose stalk size was normal on the first MRI scan two with idiopathic diabetes insipidus, one with Langerhans-cell histiocytosis, and one with a germinoma had thickening of the stalk after a median of 08 year range, 02 to 30 Seven patients who had a thickened stalk at presentation five with idiopathic diabetes insipidus and two with a germinoma had increases in the thickness of the stalk after a median of 16 years range, 08 to 103 Six other patients five with idiopathic diabetes insipidus and one with Langerhans-cell histiocytosis with a thickened stalk at presentation had a normal stalk after a median of 13
years range, 10 to 57 Fig 3 One patient with idiopathic diabetes insipidus had a decrease in his initially thick stalk after 21 years of follow-up, with no additional changes over a period of 14 years Persistently thickened stalks
Vol ume 343 Numb e r 14
1001
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
A
B
Figure 2 MRI Scans of the HypothalamicPituitary Region in a Normal Subject Panels A and B and Two Patients with Central Diabetes Insipidus Panels C, D, E, and F Coronal Panel A and sagittal Panel B T1-weighted scans in a normal subject show a normal anterior pituitary arrowheads and pituitary stalk thick arrows and hyperintensity of the posterior pituitary thin arrow in Panel B In a patient with idiopathic diabetes insipidus, coronal T1-weighted scans with gadolinium enhancement show diffuse cerebral inflammatory disease black arrows in Panel C and thickened pituitary stalk open arrow in Panel C and spontaneous normalization of brain signals and the size of the pituitary stalk open arrow in Panel D
during long-term follow-up Sagittal T1-weighted scans obtained in a patient with diabetes insipidus caused by Langerhans-cell histiocytosis show a thickened proximal pituitary stalk thick arrow, the absence of posterior pituitary hyperintensity thin arrow, and a smaller-than-normal anterior pituitary arrowhead at presentation Panel E and a normal-sized pituitary stalk thick arrow, the absence of 
posterior pituitary hyperintensity thin arrow, and a smaller-than-normal anterior pituitary arrowhead after three years of follow-up
were noted in five patients four with idiopathic diabetes insipidus and one with Langerhans-cell histiocytosis over a median follow-up of 40 years range, 30 to 55 Fifteen patients 12 with idiopathic diabetes insipidus and 3 with Langerhans-cell histiocytosis who were followed for a median of 38 years range, 27 to 65 had persistently normal stalks Changes in the size of the anterior pituitary region at the time of the last MRI scan are summarized in Tables 1 and 2
Anterior Pituitary Function
Deficits in anterior pituitary hormones were documented in 48 patients 61 percent a median of 06 year range, 01 to 180 after the onset of diabetes insipidus Table 1 The
most frequent abnormality was growth hormone deficiency 47 patients [59 percent], followed by hypothyroidism 22 patients [28 percent], hypogonadism 19 patients [24 percent], and adrenal insufficiency 17 patients [22 percent] All patients with growth retardation had a serum growth hormone response of less than 5 ng per milliliter after any of the stimulation tests Serum prolactin concentrations were high 20 ng per milliliter at the time of the diagnosis of diabetes insipidus in 16 of 54 patients 30 percent Twenty of the 41
1002
Oc to b er 5 , 2 0 0 0
patients with idiopathic diabetes insipidus 49 percent had one or more anterior pituitary hormone deficiencies Table 1 The median times from the onset of diabetes insipidus to the appearance of the first and final anterior pituitary defects in these 20 patients were 06 year range, 01 to 180 and 77 years range, 01 to 180, respectively Fig 4 Nine of the 12 patients 75 percent with Langerhans-cells histiocytosis had an anterior pituitary hormone deficiency Table 1 that was first detected a median of 35 years range, 01 to 60 after the onset of diabetes insipidus The mean estimated probability that an anterior pituitary hormone defect
would develop within six years after the diagnosis of diabetes insipidus was 81 percent for patients with Langerhans-cell histiocytosis and 49 percent for those with idiopathic diabetes insipidus Fig 4 The probability of developing additional defects did not differ significantly between these two groups
Prognostic Factors
Among the 29 patients with a thickened pituitary stalk, 27 93 percent had anterior pituitary hormone deficits, as compared with 6 of the 36 patients 17 percent with a normal pituitary stalk odds ratio, 18;
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
C E N T R A L D I A B ET E S I N S I P I D U S I N C H I L D R E N A N D YO U N G A D U LT S
C
D
E
F
95 percent confidence interval, 4 to 115; P0001 Similarly, among the 18 patients with idiopathic diabetes insipidus who had a thickened pituitary stalk, 17 94 percent had anterior pituitary hormone deficits, as compared with 2 of the 19 11 percent with idiopathic disease who had a normal pituitary stalk odds ratio, 113; 95 percent confidence interval, 9 to 4956; P0001 By contrast, in
patients with Langerhans-cell histiocytosis, the risk of an anterior pituitary hormone abnormality was independent of the size of the pituitary stalk Among patients with idiopathic diabetes insipidus, anterior pituitary hormone deficits were strongly associated with a smaller-thannormal anterior pituitary: 20 of 22 patients with deficits 91 percent had a smaller-than-normal pituitary, as compared with 3 of 19 patients 16 percent with a normal-sized pituitary odds ratio, 54; 95 percent confidence interval, 6 to 606; P0001
DISCUSSION
In our study, most of the 79 patients with central diabetes insipidus who were identified in four pediatric endocrinology units were children Tumor-associated diabetes insipidus was not diagnosed in children who were younger than five years of age, whereas the familial form of the disorder was diagnosed by the age
of seven in all five affected patients Forty percent of the patients had symptoms other than polyuria and polydipsia at presentation We did not find that growth retardation was significantly more common in a single group of patients, in contrast to previous reports indicating that such delays are strongly suggestive of an intracranial tumor
as the cause of central diabetes insipidus13,24 In about one fourth of the patients with idiopathic diabetes insipidus, there was a temporal association between a viral infection and the onset of diabetes insipidus, a finding previously reported in isolated cases in neonates10,25 Many patients did not have posterior pituitary hyperintensity on serial MRI scans, indicating that the absence of hyperintensity, although nonspecific, is a cardinal feature of central diabetes insipidus18 In the five patients who did have posterior pituitary hyperintensity at diagnosis,2 this feature invariably disappeared during follow-up23 Thickening of either the entire pituitary stalk or just the proximal portion was the second most common abnormality on MRI scans It was helpful in diagnosing idiopathic diabetes insipidus and that associated with germinomas or Langerhans-cell histiocytosis, but it was not specific for any single subtype Other studies that have reported a thickened pituitary stalk in association with autoimVol ume 343 Numb e r 14
1003
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society
All rights reserved
The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
A
B
C
Figure 3 Sagittal T1-Weighted MRI Scans with Gadolinium Enhancement Obtained in a Patient with Central Diabetes Insipidus Caused by Langerhans-Cell Histiocytosis At presentation Panel A, the posterior pituitary is not hyperintense small black arrow, the pituitary stalk is thickened large black arrow, and the size of the anterior pituitary is normal white arrowhead Ten months later Panel B, there is further thickening of the pituitary stalk arrow and a reduction in the size of the anterior pituitary arrowhead After approximately two years Panel C, the pituitary stalk is not as thick arrow and there has been no change in the size of the anterior pituitary arrowhead
mune or inflammatory disease — termed lymphocytic infundibuloneurohypophysitis,4 lymphocytic hypophysitis,26 or necrotizing infundibulohypophysitis27 — have focused on adults with histologic features of lymphocyte and plasma-cell infiltration, fibrosis, and necrosis Any of these diagnoses seems unlikely in our patients with idiopathic diabetes insipidus, not only because of their young age, but also because 17 of 18 patients with
idiopathic disease and a thickened pituitary stalk also had an anterior pituitary hormone deficiency, and some had progressive disease Follow-up MRI scans of patients with idiopathic central diabetes insipidus and a thickened pituitary stalk showed a range of changes, from a spontaneous resolution of the abnormality to further enlargement; some had no change Nonetheless, our findings suggest that a germinoma can be recognized early during follow-up on the basis of an increase in the size of the stalk In our study, such follow-up led to a bi1004
Octo b er 5 , 2 0 0 0
opsy-proven diagnosis of germinoma within a median of 1 year after the diagnosis of central diabetes insipidus, in contrast to the 25 years reported for patients evaluated by MRI performed a mean of 13 years after the diagnosis of diabetes insipidus28 In most of the patients with diabetes insipidus who had a normal pituitary stalk on the first MRI examination, there was no change in the size of the stalk during followup However, the protracted interval median, 15 years between the onset of diabetes insipidus and the first MRI scan may have hidden a rapid change, as suggested by the finding that the size of the stalk
returned to normal within a median of 12 years in other patients The incidence of anterior pituitary hormone deficits, particularly of growth hormone deficiency in patients with idiopathic diabetes insipidus 49 percent, was higher than that reported in other studies13,29 Furthermore, the observation that an anterior pituitary hormone deficit may indicate the presence of a
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
C E N T R A L D I A B ET E S I N S I P I D U S I N C H I L D R E N A N D YO U N G A D U LT S
TABLE 2 COURSE
OF
CENTRAL DIABETES INSIPIDUS IN 18 PATIENTS WITH A NORMAL OR THICKENED PITUITARY STALK ON MAGNETIC RESONANCE IMAGING AT PRESENTATION
ENDOCRINE HORMONE DEFICIENCIES IDENTIFIED AT ANY TIME
ON LAST
PATIENT NO
AGE YR AT PRESENTATION/ SEX
CAUSE
OF
DISEASE
ON FIRST
STALK SIZE MRI
SCAN
SIZE OF ANTERIOR PITUITARY ON LAST MRI SCAN
AT
ON FOLLOW-UP
PRESENTATION
MRI
SCAN
MRI
SCAN
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
52/M 55/M 56/M 74/F 75/F 72/M 85/F 95/M 60/M 75/F 80/F 80/F 91/F 71/F 64/M 118/M 111/M 198/M
Idiopathic
Idiopathic Idiopathic Idiopathic Idiopathic Idiopathic Idiopathic Idiopathic Idiopathic inflammatory Idiopathic inflammatory Idiopathic inflammatory Germinoma Germinoma Langerhans-cell histiocytosis Idiopathic Idiopathic Germinoma Langerhans-cell histiocytosis
Normal Normal Normal Normal
Normal 15 15 21 103 Normal 10 20 Normal 10 Normal 10 Normal 57 16 50 08 15 Normal 51 30 10 05 02
Normal 23 25 14 0 Normal 36 42 Normal 29 Normal 30 Normal 0 10 04 Could not be evaluated Could not be evaluated Normal 25 11 03 05 26
GH GH and TSH GH GH and FSH or LH GH GH, TSH, and FSH or LH GH GH GH GH and TSH GH and FSH or LH GH and ACTH GH and TSH GH None None None None
Reduced Reduced Reduced Reduced Reduced Reduced Reduced Reduced Reduced Reduced Reduced Normal Increased Reduced Reduced Normal Increased Normal
MRI denotes magnetic resonance imaging, GH growth hormone, TSH thyrotropin, FSH follicle-stimulating hormone, LH luteinizing hormone, and ACTH corticotropin A normal stalk was defined as one that was less than 30 mm thick A single plus sign indicates a thickness of 30 to 45 mm, two plus signs a thickness of 46 to 65 mm, three plus signs a thickness
of more than 65 mm, and a plusminus sign a stalk size that was smaller than it had been on the previous measurement but that was still more than 35 mm Values in parentheses are the intervals in years since presentation, in the case of the follow-up MRI scan, or since the follow-up scan, in the case of the last MRI scan This patient had MRI evidence of diffuse inflammatory infiltrates in the brain In this patient, the findings on pituitary-stalk biopsy were compatible with the presence of an inflammatory process This patient was treated with prednisolone during follow-up This patient was treated with prednisolone and vinblastine during follow-up
tumor in patients with diabetes insipidus13,24 was not confirmed by our data In idiopathic diabetes insipidus, growth hormone deficiency can develop soon after the onset of disease, with additional pituitary defects becoming evident later on Our finding of a longer median interval between the diagnosis of diabetes insipidus and the detection of the first anterior pituitary hormone deficit in patients with Langerhans-cell histiocytosis than in patients with idiopathic diabetes insipidus 35 years vs 06 year confirms our recent findings30 and
could be related to a reactivation of the disease, the effects of treatment, or perhaps both
MRI evidence of a progressive reduction in the size of the anterior pituitary was associated with a higher risk of an additional endocrine defect This decrease in size, as suggested by a recent report,28 may be a consequence of vascular damage8,31 or a deficiency of hypothalamic hormones MRI evidence of an increase in the size of the anterior pituitary with thickening of the stalk was strongly associated with the presence of a germinoma In conclusion, children and young adults with acquired central diabetes insipidus, especially those with a thickened pituitary stalk and a reduction in the size
Vol ume 343 Numb e r 14
1005
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
The Ne w E n g l a nd Jo u r n a l o f Me d ic i ne
10
Proportion with a First Anterior Pituitary Hormone Deficiency
09
Langerhans-cell histiocytosis 12 patients
08 07 06 05
Idiopathic 41 patients
04 03 02 01 00 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Years after Diagnosis
Figure 4
Cumulative Probability of the Development of an Anterior Pituitary Hormone Deficiency during Follow-up in 41 Patients with Idiopathic Central Diabetes Insipidus and 12 Patients with Central Diabetes Insipidus Associated with Langerhans-Cell Histiocytosis
of the anterior pituitary, are prone to the development of a growth hormone deficiency or a deficiency of another anterior pituitary hormone The high frequency 52 percent of cases of idiopathic central diabetes insipidus in our study most likely reflects our incomplete knowledge of the causes of this disease The term lymphocytic infundibuloneurohypophysitis, which is used to describe cases of central diabetes insipidus in adults with a thickened pituitary stalk, is applicable to childhood cases only when the pituitary stalk is transiently or persistently thickened, the posterior pituitary is not hyperintense on MRI, and the size of the anterior pituitary is normal The term lymphocytic infundibulohypophysitis would be more appropriate to describe patients with anterior pituitary hormone deficiencies A progressive increase in the size of the anterior pituitary should alert physicians to the possibility that a germinoma is present,
whereas a decrease can suggest the presence of an inflammatory or autoimmune process Biopsy should be reserved for patients with progressive thickening of the pituitary stalk, since spontaneous recovery may occur in patients without this feature
Supported in part by funds from Telethon, Italy project E755, to Dr Aricò, and by grants 390RCR97 /01, to Dr Aricò, and 3180/GEN/98, to Dr Maghnie from the Istituto di Ricovero e Cura a Carattere Scientifico Policlinico San Matteo, Pavia, Italy
REFERENCES
1 Mootha SL, Barkovich AJ, Grumbach MM, et al Idiopathic hypothalamic diabetes insipidus, pituitary stalk thickening, and the occult intracranial germinoma in children and adolescents J Clin Endocrinol Metab 1997;82:1362-7 2 Maghnie M, Villa A, Aricò M, et al Correlation between magnetic resonance imaging of posterior pituitary and neurohypophyseal function in children with diabetes insipidus J Clin Endocrinol Metab 1992;74:795-800 3 Maghnie M, Aricò M, Villa A, Genovese E, Beluffi G, Severi F MR of the hypothalamic-pituitary axis in Langerhans cell histiocytosis AJNR Am J Neuroradiol 1993;13:1365-71 4 Imura H, Nakao K, Shimatsu A, et al Lymphocytic infundibuloneurohypophysitis as a cause
of central diabetes insipidus N Engl J Med 1993; 329:683-9 5 Maghnie M, Genovese E, Sommaruga MG, et al Evolution of childhood central diabetes insipidus into panhypopituitarism with a large hypothalamic mass: is lymphocytic infundibuloneurohypophysitis in children a different entity? Eur J Endocrinol 1998;139:635-40 6 Verbalis JG, Robinson AG, Moses AM Postoperative and post-traumatic diabetes insipidus Front Horm Res 1985;13:247-65 7 Robinson AG, Fitzsimmons MD Diabetes insipidus Adv Endocrinol Metab 1994;5:261-96 8 Robertson GL Diabetes insipidus Endocrinol Metab Clin North Am 1995;24:549-72 9 Hansen LK, Rittig S, Robertson GL Genetic basis of familial neurohypophyseal diabetes insipidus Trends Endocrinol Metab 1997;8:363-72 10 Blotner H Primary or idiopathic diabetes insipidus: a system disease Metabolism 1958;7:191-200 11 Bode HH Disorders of posterior pituitary In: Kaplan SA, ed Clinical pediatric endocrinology Philadelphia: WB Saunders, 1990:63-86 12 Pomarède R, Czernichow P, Rappaport R, et al Le diabète insipide pitresso-sensible de lenfant II Etude de 93 cas observés entre 1955 et 1978 Arch Fr Pediatr 1980;37:37-44 13 Czernichow P, Pomarede R, Brauner R, Rappaport R
Neurogenic diabetes insipidus in children Front Horm Res 1985;13:190-209 14 Greger NG, Kirkland RT, Clayton GW, Kirkland JL Central diabetes insipidus: 22 years experience Am J Dis Child 1986;140:551-4 15 Wang LC, Cohen ME, Duffner PK Etiologies of central diabetes insipidus in children Pediatr Neurol 1994;11:273-7 16 Tien R , Kucharczyk J, Kucharczyk W MR imaging of the brain in patients with diabetes insipidus AJNR Am J Neuroradiol 1991;12:533-42 17 Moses AM, Clayton B, Hochhauser L Use of T1-weighted MR imaging to differentiate between primary polydipsia and central diabetes insipidus AJNR Am J Neuroradiol 1992;13:1273-7
We are indebted to Professor Melvin M Grumbach, for reviewing the manuscript; to Dr Alessandra Di Cesare Merlone and Dr Simona Piccinini, for their help in data collection; to Dr Mauro Pocecco, for referring one patient to the study; and to Mr John Gilbert, for critical comments and editorial assistance
1006
Octo b er 5 , 2 0 0 0
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
C E N T R A L D I A B ET E S I N S I P I D U S I N C H I L D
R E N A N D YO U N G A D U LT S
18 Maghnie M, Sommaruga MG, Beluffi G, Severi F Role of MR imaging in the evaluation of the functional status of the posterior pituitary gland: the view of a pediatric endocrinologist AJNR Am J Neuroradiol 1993;14:1443-5 19 Brooks BS, el Gammal T, Allison JD, Hoffmann WH Frequency and variation of the posterior pituitary bright signal on MR images AJNR Am J Neuroradiol 1989;153:1033-8 20 Fujisawa I, Asato R , Okumura R , et al Magnetic resonance imaging of neurohypophyseal germinomas Cancer 1991;68:1009-14 21 Frasier SD, Kutnik LA, Schmidt RT, Smith FG Jr A water deprivation test for the diagnosis of diabetes insipidus in children Am J Dis Child 1967;114:157-60 22 Gagliardi PC, Bernasconi S, Repaske DR Autosomal dominant neurohypophyseal diabetes insipidus associated with a missense mutation encoding Gly23Val in neurophysin II J Clin Endocrinol Metab 1997;82: 3643-6 23 Maghnie M, Genovese E, Bernasconi S, Binda S, Aricò M Persistent high MR signal of the posterior pituitary gland in central diabetes insipidus AJNR Am J Neuroradiol 1997;18:1749-52 24 Sklar CA, Grumbach MM, Kaplan SL, Conte FA 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 1981;52:9-16
25 Mena W, Royal S, Pass RF, Whitley RJ, Philips JB III Diabetes insipidus associated with symptomatic congenital cytomegalovirus infection J Pediatr 1993;122:911-3 26 Thodou E, Asa SL, Kontogeorgos G, Kovacs K, Horvath E, Ezzat S Lymphocytic hypophysitis: clinicopathological findings J Clin Endocrinol Metab 1995;80:2302-11 27 Ahmed SR , Aiello DP, Page R , Hopper K, Towfighi J, Santen RJ Necrotizing infundibulo-hypophysitis: a unique syndrome of diabetes insipidus and hypopituitarism J Clin Endocrinol Metab 1993;76:1499-504 28 Leger J, Velasquez A, Garel C, Hassan M, Czernichow P Thickened pituitary stalk on magnetic resonance imaging in children with central diabetes insipidus J Clin Endocrinol Metab 1999;84:1954-60 29 Czernichow P, Pomarede R , Basmaciogullari A, Brauner R , Rappaport R Diabetes insipidus in children III Anterior pituitary dysfunction in idiopathic types J Pediatr 1985;106:41-4 30 Maghnie M, Bossi G, Klersy C, Cosi G, Genovese E, Aricò M Dynamic endocrine testing and magnetic resonance imaging in the long term
follow-up of childhood Langerhans cell histiocytosis J Clin Endocrinol Metab 1998;83:3089-94 31 Maghnie M, Genovese E, Aricò M, et al Evolving pituitary hormone deficiency is associated with pituitary vasculopathy: dynamic MR study in children with hypopituitarism, diabetes insipidus, and Langerhans cell histiocytosis Radiology 1994;193:493-9
IMAGES IN CLINICAL MEDICINE The Journal has a large backlog of Images in Clinical Medicine that have been accepted for publication Therefore, we will not consider new submissions in 2000 This decision will be reevaluated in December
Vol ume 343
Numb e r 14
1007
Downloaded from wwwnejmorg on November 16, 2008 For personal use only No other uses without permission Copyright 2000 Massachusetts Medical Society All rights reserved
