PITUITARY DISORDERS

Diabetes insipidus Key points

Rupinder Singh Kochhar C Diabetes insipidus (DI) can be caused by lack of vasopressin
Stephen Ball (AVP) production (cranial DI) or a diminished response of the
kidneys to AVP (nephrogenic DI)

C Confirming polyuria and excluding metabolic causes are the

Abstract
first steps of management
Diabetes insipidus (DI) describes the excess production of dilute urine.
It is caused by the lack of production or action of the hormone vaso-
C Confirming the diagnosis involves a thorough history, physical
pressin (AVP). Diagnosis requires a targeted history and examination.
examination and specialized investigation such as a water
Confirmation requires referral to specialist services with expertise in
deprivation test or hypertonic saline test
diagnosis and management. Lack of AVP can be treated with synthetic
AVP analogues. Additional approaches can be needed for specific
C Treatment of these disorders requires careful attention to fluid
forms of DI. Combined defects in AVP production and thirst perception
and electrolyte balance, and should be done in consultation
require a structured programme of fluid intake and antidiuresis. Clini-
with an endocrinologist
cians should be aware of both systemic and locally progressive pa-
thologies that can present with DI.
Keywords Adipsia; diabetes insipidus; hypernatraemia; hypona-
 cranial or hypothalamic DI (HDI): a relative or absolute
traemia; MRCP; polydipsia; polyuria; vasopressin
lack of AVP
 nephrogenic DI (NDI): partial or total resistance to the
renal antidiuretic effects of AVP
Introduction  dipsogenic DI (DDI, primary polydipsia): inappropriate high
Diabetes insipidus (DI) is characterized by excess production of fluid intake in excess of the ability to excrete free water.
dilute urine e more than 40 ml/kg/24 hours in adults and more
than 100 ml/kg/24 hours in children. Hypothalamic diabetes insipidus
Urine concentration is regulated by vasopressin (AVP), a nine- Presentation with HDI occurs when 80% of hypothalamic mag-
amino-acid peptide produced by magnocellular neurones within nocellular neurone AVP production has been lost. Several path-
the supraoptic nucleus and paraventricular nucleus of the hy- ological processes can be involved. All lead to either the
pothalamus. AVP is released into the circulation from nerve destruction of AVP neurones, or the interruption of the transport
terminals in the posterior pituitary gland. Production is directly or processing of AVP as it moves along the axons of these neu-
related to plasma osmolality and inversely related to plasma rones for release at nerve terminals in the posterior pituitary
volume. The principal site of action of AVP is the kidney, where (Table 1). The condition can appear for the first time or become
it increases the synthesis and assembly of water channels worse in pregnancy as residual AVP is degraded by increased
(aquaporin 2 (AQP2)), leading to increased free water reab- placental enzyme (vasopressinase) activity. Up to 50% of
sorption. These effects are mediated through interaction with a
G-protein-coupled cell surface receptor (AVP-R2), found on Aetiology of HDI
target cells lining the luminal surface of the distal nephron.1
Primary Genetic Wolfram’s syndrome, autosomal
Classification and aetiology dominant, autosomal recessive
Developmental Septo-optic dysplasia
There are three subtypes of DI, each reflecting a specific element syndromes
of the physiology and/or pathophysiology of the AVP axis: Idiopathic
Secondary Trauma Head injury, post-surgery
(transcranial, trans-sphenoidal)
Tumour Craniopharyngioma, germinoma,
Rupinder Singh Kochhar MB BS MD MRCP is a Specialist Registrar metastases, pituitary macroadenoma
with North Western Deanery currently working at Manchester Royal Inflammatory Sarcoidosis, histiocytosis,
Infirmary, UK. Competing interests: none declared. meningitis, encephalitis,
Stephen Ball BSc PhD FRCP is Consultant and Honorary Professor of infundibuloneurohypophysitis,
Medicine. He studied Basic Science at Birmingham University before GuillaineBarre syndrome,
pursuing undergraduate and postgraduate medical studies in autoimmune
London. Middle-grade training in diabetes and endocrinology Vascular Aneurysm, infarction
included a period in the USA as a Medical Research Council (UK) and
Howard Hughes Fellow. He combines clinical medicine, research and Classification describes primary and secondary (acquired) causes. All result in
teaching. He is a member of the Council of the Society for decreased production of AVP through direct or indirect damage to hypothalam-
Endocrinology and a Senior Editor of Clinical Endocrinology. He is ic AVP magnocellular neurones.
also a European Endocrine Society representative within European
Guideline groups. Competing interests: none declared. Table 1

MEDICINE 45:8 488 Crown Copyright Ó 2017 Published by Elsevier Ltd. All rights reserved.
 PITUITARY DISORDERS
children and young adults with HDI have an underlying tumour
or central nervous system malformation. Familial HDI comprises
5% of cases.2
Nephrogenic diabetes insipidus
Renal resistance to AVP can be the result of the adverse effects of
specific drugs (e.g. lithium toxicity). These effects can be tem-
porary but sometimes persist. NDI can also occur after obstructive
nephropathy, acute kidney injury and electrolyte disturbances
(e.g. hypokalaemia, hypercalcaemia). Prolonged polyuria of any
cause can result in partial NDI through disruption of the intra-
renal solute gradient that is an obligatory requirement for the
production of concentrated urine.
NDI presenting in childhood is most commonly caused by
inherited defects in AVP action. X-linked familial NDI results
from loss-of-function mutations in the renal AVP-R2 receptor.
Autosomal recessive or dominant NDI can be caused by loss-of-
function mutations in the AVP-dependent water channel AQP2.3
Dipsogenic diabetes insipidus (primary polydipsia)
Persistent inappropriate fluid intake leads to appropriate poly-
uria. If it exceeds the limit of renal free water excretion, it can
result in hyponatraemia. DDI can be associated with the
following abnormalities of thirst perception:
 low threshold for thirst
 exaggerated thirst response to osmotic challenge
 inability to suppress thirst at low plasma osmolalities
Structural lesions can be present, but neuroimaging is normal
in most cases. DDI is associated with affective disorders.
Epidemiology
DI is rare. The prevalence of HDI is estimated at 1/25,000 with
equal gender distribution. The prevalence of the other forms is
unclear. Although most cases present in adulthood, familial HDI
and NDI characteristically present in childhood.
Diagnosis and investigations
DI presents with polyuria and polydipsia.
 Polyuria must be distinguished from simple frequency
without excess urine volume, commonly seen in urinary
infection or prostatic enlargement.
 Nocturnal symptoms are usually the first clue, as nocturia
is often the first manifestation of the loss of urinary con-
centration capability.
 Characterizing the onset of polyuria can be helpful in
differentiating the underlying cause. Relatively abrupt
onset is commonly a feature in HDI, as opposed to a more
gradual onset in NDI and DDI.
History and examination can reveal important features sug-
gestive of:
 systemic disease
 associated endocrinopathy
 an associated neurological problem suggestive of struc-
tural disease
 drug toxicity.
The initial approach should be to confirm excess urine volume
and exclude simple metabolic causes such as hyperglycaemia
and hypercalcaemia.
MEDICINE 45:8
Measuring plasma sodium concentration and urinary osmo-
lality (preferably early morning) is helpful in the early work-up
of disorders of urinary concentrating ability:
 A low plasma sodium concentration with a low urine
osmolality is suggestive of DDI.
 A normal plasma sodium concentration associated with urine
osmolality >600 mosmol/kg excludes a diagnosis of DI.
Definitive diagnosis requires referral to an endocrine service
for testing AVP production and action in response to osmolar
stress. The water deprivation test measures renal concentrating
capacity in response to dehydration, and is an indirect assess-
ment of the AVP axis. It is usually followed by assessment of
renal response to the synthetic AVP analogue desmopressin
(DDAVP). Characteristic findings are:
 in DDI, urine concentration is normal in response to
dehydration.
 in HDI, urine concentration fails in response to dehydra-
tion but not to DDAVP.
 in NDI, urine concentration fails in response to both
manoeuvres.
In practice, many results are indeterminate.4
Direct measurement of AVP production during graded hyper-
osmolar stimulation is the gold standard method for diagnosing
and classifying DI (Figure 1). Recent data suggest that measure-
ment of co-peptin, an inactive fragment of the AVP precursor that
is released in proportion to AVP, may prove to be an alternative.5
Confirmation of HDI should lead to further pituitary function
testing and cranial magnetic resonance imaging. NDI requires
renal tract imaging and additional renal function studies.
Diabetes insipidus and hypernatraemia
There is a close neuroanatomical relationship between the hy-
pothalamic structures responsible for osmoregulation of thirst
Diagnosis and classification of diabetes insipidus
by measurement of plasma vasopressin in
response to graded hyperosmolar stimulation
20
Plasma vasopressin
Nephrogenic
Dipsogenic
15
(pmol/litre)
Hypothalamic
10
5
0
280 290 300 310 320
Plasma osmolality (mOsmol/kg)
The shaded area depicts the normal-range response of plasma
vasopressin as plasma osmolality is raised. Patients with
hypothalamic diabetes insipidus have a response below the
normal range. Those with dipsogenic diabetes insipidus have a
normal response. Patients with nephrogenic diabetes insipidus
have a response at the top of the reference range but coincident
urine osmolalities that are dilute, consistent with vasopressin
resistance.
Figure 1 Hypertonic saline infusion test.
489 Crown Copyright Ó 2017 Published by Elsevier Ltd. All rights reserved.
 PITUITARY DISORDERS
and AVP production. Certain structural, neurovascular and
neurodevelopmental lesions are thus associated with combined
defects in both processes. Absent or reduced thirst (adipsia) in
association with HDI predisposes to hypernatraemic dehydra-
tion. Diagnosis follows the protocol for HDI, but benefits from a
parallel assessment of thirst perception.
Management
Mild forms of HDI may not require treatment. Significant polyuria
and polydipsia are treated effectively with oral or intranasal
DDAVP. In NDI, causal drugs should be withdrawn where
possible and causative electrolyte disturbances corrected. NDI can
partly respond to high-dose DDAVP. Thiazide diuretics and non-
steroidal anti-inflammatory drugs can be of additional benefit
but may not correct abnormal urine production. The only rational
approach to DDI is reduction in fluid intake. DDAVP treatment
must be avoided in DDI, because of the risk of hyponatraemia.
Follow-up
Following initiation of DDAVP, patients can require frequent re-
view by an endocrine service for dose titration. Thereafter, they
can be seen annually to assess symptom control and check serum
sodium concentration to avoid overtreatment. The combination of
HDI and adipsia requires meticulous follow-up in a specialist ser-
vice, with a structured approach to fluid intake and antidiuresis.
TEST YOURSELF
To test your knowledge based on the article you have just read, please complete the questions below. The answers can be found at the
end of the issue or online here.
Question 1
A 23-year-old woman presented with polyuria and polydipsia.
She had also lost weight. She had a history of bipolar disorder
that was stable on lithium carbonate, and she did not take any
other medications. Clinical examination was unremarkable.
What is the next best management step?
A. Perform a water deprivation test
B. Start a trial of DDAVP (desmopressin)
C. Check lithium concentrations
D. Check blood glucose
E. Check serum calcium
Question 2
A72-year-old man was noted to have had hypernatraemia 2
weeks following drainage of a brain abscess. He had a history
of hypertension and ischaemic heart disease. On examination,
he was slightly dehydrated and confused but able to comply
with instructions. The remainder of the examination was
unremarkable. His urine output had been 2000e2500 ml per
day for the previous 4 days, and he had been given 5%
MEDICINE 45:8
Prognosis
Isolated HDI has an excellent prognosis and patients should
anticipate a normal quality of life. When HDI is associated with
pituitary hormone deficiency or structural or systemic disease,
prognosis is determined by the natural history of the underlying
pathology and co-morbidities. HDI resulting from trauma may
resolve spontaneously. NDI is difficult to treat, and the symptoms
seldom respond completely. Future developments may further
help early diagnosis and guide appropriate intervention. A
KEY REFERENCES
1 Ball SG. Vasopressin and disorders of water balance: the physi-
ology and pathophysiology of vasopressin. Ann Clin Biochem
2007; 44: 417e31.
2 Babey M, Kopp P, Robertson GA. Familial forms of diabetes
insipidus: clinical and molecular characteristics. Nat Rev Endocrinol
2011; 7: 701e14.
3 Moeller HB, Rittig S, Fenton RA. Nephrogenic diabetes insipidus:
essential insights into the molecular background and potential
therapies for treatment. Endocr Rev 2013; 34: 278e301.
4 Ball SG, Barber T, Baylis PH. Tests of posterior pituitary function.
J Endocrinol Invest 2003; 26(suppl): 15e24.
5 Fenske W, Allolio B. Current state and future perspectives in the
diagnosis of diabetes insipidus: a clinical review. J Clin Endocrinol
Metab 2012; 97: 3426e37.
dextrose intravenously (2000 ml per day) for the previous 2
days.
Investigation
 Haemoglobin 164 g/litre (130e180)
 White cell count 8.9  109/litre (4.0e11.0)
 Platelets 243  109/litre (150e400)
 Serum sodium 155 mmol/litre (137e144)
 Serum potassium 4.5 mmol/litre (3.5e4.9)
 Serum urea 8.7 mmol/litre (2.5e7.0)
 Serum creatinine 98 mmol/litre (60e110)
 C-reactive protein 24 mg/litre (0e5.0)
 Serum calcium (adjusted) 2.54 mmol/litre (2.20e2.60)
 Plasma glucose 5.6 mmol (3.0e6.0)
 Liver function tests were normal
What investigation will best confirm the most likely diagnosis?
A. Urine sodium
B. Plasma NT-proBNP
C. Plasma vasopressin
D. Urine osmolality
E. Plasma aldosterone
490 Crown Copyright Ó 2017 Published by Elsevier Ltd. All rights reserved.
 PITUITARY DISORDERS

Question 3 What is the most likely diagnosis?

A. Hypothalamic diabetes insipidus
A 24-year-old woman presented with a 6-month history of
B. Nephrogenic diabetes insipidus
increasing polyuria of up to 8 liters per day. The results of a hy-
C. Dipsogenic diabetes insipidus
pertonic saline test were equivocal. She was admitted overnight for
D. Syndrome of inappropriate antidiuretic hormone secretion
a supervised trial of low-dose desmopressin. Early next morning,
E. Primary adrenal failure
serum sodium was 128 mmol/litre (normal range 137e144).

MEDICINE 45:8 491 Crown Copyright Ó 2017 Published by Elsevier Ltd. All rights reserved.