Anda di halaman 1dari 10

Journal of Endocrinological Investigation


Copeptin in the differential diagnosis of hypotonic polyuria

M. Christ‑Crain1,2 · W. K. Fenske1,2

Received: 10 May 2019 / Accepted: 17 July 2019

© Italian Society of Endocrinology (SIE) 2019

Copeptin  Copeptin is secreted in equimolar amount to Arginine Vasopressin (AVP) but can easily be measured with a sand-
wich immunoassay. Both peptides, copeptin and AVP, show a high correlation. Accordingly, copeptin mirrors the amount of
AVP in the circulation and its measurement provides an attractive marker in the differential diagnosis of diabetes insipidus.
The polyuria polydipsia syndrome  Diabetes insipidus—either central or nephrogenic—has to be differentiated from primary
polydipsia. Differentiation is crucial since wrong treatment can have deleterious consequences. Since many decades, the
“gold standard” for differential diagnosis has been the classical water deprivation test, which has several limitations leading
to an overall limited diagnostic accuracy. In addition, the test has a long duration of 17 hours and is cumbersome for patients.
Clinical signs and symptoms as well as MRI characteristics overlap between patients with diabetes insipidus and primary
polydipsia. Direct measurement of AVP upon osmotic stimulation was first shown to overcome these limitations, but failed
to enter clinical practice mainly due to technical limitations of the AVP assay.
Copeptin as diagnostic tool in the polyuria polydipsia syndrome  We have recently shown that copeptin, without prior water
deprivation, identifies patients with nephrogenic diabetes insipidus. On the other hand, for the more difficult differentiation
between central diabetes insipidus and primary polydipsia, a copeptin level of 4.9 pmol/L stimulated with hypertonic saline
infusion differentiates between these two entities with a high diagnostic accuracy, and is superior to the water deprivation
test. It is important to note that close sodium monitoring during the hypertonic saline test is a prerequisite.
Conclusion  Therefore, we propose that copeptin upon hypertonic saline infusion should become the new standard test in the
differential diagnosis of diabetes insipidus.

Keywords  Diabetes insipidus · Primary polydipsia · Differential diagnosis

Introduction in the differential diagnosis of diabetes insipidus has been

Copeptin derives from the same precursor protein (i.e., Diabetes insipidus is part of the so-called polyuria-poly-
pre-pro-vasopressin) as arginine vasopressin (AVP) and dipsia syndrome. This syndrome is defined by an output of
neurophysin II. Since copeptin and AVP show a high cor- more than 50 mL/kg body weight per 24 h of hypotonic
relation and since they increase upon similar physiologic urine (< 300  mOsm/kg  H2O), which is accompanied by
stimuli, namely osmotic stimulation, the value of copeptin polydipsia of more than 3 L a day [1]. The differential diag-
nosis of hypotonic polyuria includes central or nephrogenic
diabetes insipidus on one hand and primary polydipsia on
* M. Christ‑Crain the other hand. A correct differential diagnosis is important
mirjam.christ‑ since treatment differs and application of the wrong treat-
* W. K. Fenske ment may lead to dangerous clinical consequences for the
WiebkeKristin.Fenske@medizin.uni‑ patients.
The “gold standard” for differential diagnosis since dec-
Department of Endocrinology, Diabetology and Metabolism,
University Hospital of Basel, University of Basel,
ades has been the classical water-deprivation test. This test,
Petersgraben 4, 4031 Basel, Switzerland unfortunately, has several limitations leading to a diagnostic
Medical Department III‑Endocrinology, Nephrology,
accuracy of only around 70%. Therefore, it was proposed
Rheumatology, University of Leipzig Medical Center, by Zerbe and Robertson in the 1980s to directly measure
University Hospital Leipzig, Leipzig, Germany

Journal of Endocrinological Investigation

arginine vasopressin (AVP) upon osmotic stimulation. and the ineffective folding of the AVP precursor in the
Despite initially promising results, however, this test did absence of copeptin was proposed [13], but never vali-
not enter clinical practice, mainly due to technical limita- dated. To date, copeptin has its main role as stable sur-
tions of the AVP assay. In this review, we discuss a new test rogate marker for osmotically regulated AVP [5].
approach which includes direct measurement of copeptin Provided that circulating plasma AVP is quantified with
instead of AVP concentrations. one of the rare well-established radio-immunoassays [14,
15], it strongly correlates with copeptin levels in healthy
volunteers, with a correlation index of r = 0.8 [16]. Of
Copeptin note, the correlation of plasma copeptin with plasma
osmolality was even stronger than the correlation of AVP
Copeptin is a relatively new biomarker, which is on its way with plasma osmolality [17], most likely due to the com-
to penetrate the clinical chemistry laboratories. Originally plexity and methodological drawbacks of the AVP assay.
detected in 1972 in the posterior pituitary of pigs [2, 3], Processed from the same precursor peptide, the release
copeptin derives from the 164 amino acid precursor pro- of plasma copeptin and plasma AVP into circulation is
tein pre-pro-vasopressin together with arginine vasopressin regulated by the same physiological stimuli, which is a
(AVP) and neurophysin II and is a 39 amino acid-long gly- relative increase in systemic osmolality and a relative
cosylated peptide with a leucine-rich core region [3, 4] and decrease in arterial blood volume and pressure [16, 18],
a molecular mass of around 5 kDa [5] (Fig. 1). The surrogate properties of copeptin for physiological
As a bioactive hormone, AVP has important physiologi- AVP release secondary to osmotic regulation was first
cal functions, including the homeostasis of fluid balance, shown in a study including 24 healthy adults, where fluid
the vascular tonus as well as the regulation of the endo- deprivation as well as hypertonic saline infusion led to a
crine stress response [6, 7]. Notwithstanding its biological significant increase in plasma copeptin levels [18]. We
importance and despite ongoing improvements of the com- recently showed that copeptin is released in equimolar
petitive AVP radioimmunoassays, measurement of mature amounts to AVP in response to osmotic stimulation (with
AVP remains technically difficult and subject to several pre- 3% saline infusion), again confirming its high potential as
analytical errors due to its small molecular size [8]. an AVP surrogate for differentiation of osmotic disorders
Copeptin by nature of its chemical structure has no such [17].
limitations and represents an appropriate “shadow” frag- On the other hand, hypotonic saline infusion prompts
ment reflecting AVP release, although its physiological suppression of plasma copeptin levels in healthy volunteers
function remains largely unknown. Earlier publications [18], similar to an oral water load [16]. The rapid increase
proposed a role as prolactin-releasing factor [9, 10] as well of copeptin secretion upon volume depletion was shown in
as its involvement in the folding of the AVP precursor, as a baboon model of experimental hemorrhagic shock [19].
copeptin seems to interact with the calnexin–calreticulin Thereby, plasma copeptin concentration increased from
system in the endoplasmic reticulum [11, 12]. Along these median levels of 7.5–269 pmol/L, with a subsequent decline
lines, a relationship between some forms of familial DI to 27 pmol/L after reperfusion.

Fig. 1  AVP and its protein products. The prohormone is packaged terior pituitary, enzymatic cleavage of the prohormone generates the
into neurosecretory granules of magnocellular neurons. During final products: AVP, neurophysin and the COOH-terminal glycopro-
axonal transport of the granules from the hypothalamus to the pos- tein copeptin. Adapted from [8]

Journal of Endocrinological Investigation

Importantly, also unspecific somatic stress is a determi- cleared by the kidneys [29] and that in patients with chronic
nant of copeptin regulation. Several observational studies kidney disease, plasma copeptin levels inversely correlate
demonstrated the predictive character of plasma copeptin with decreasing glomerular filtration rate [30].
as an unspecific marker of stress in different settings of The normal range of plasma copeptin levels has been
acute disease, mainly ischaemic stroke, myocardial infarc- established in two large clinical trials evaluating healthy
tion and lower respiratory tract infections [20–22]. Our volunteers under normal conditions. In the first study includ-
group showed that copeptin even more sensitively reflects ing more than 300 subjects, plasma copeptin levels ranged
individual stress level than plasma cortisol in patients with from 1.0 to 13.8 pmol/L with median levels of 4.2 pmol/L
increasing disease-related stress [23]. In contrast to somatic [5]. The second evaluation with over 700 randomly selected
stress, the impact of psychological stress on copeptin release volunteers reported similar results with plasma copeptin
is rather limited, as shown by recent studies in medical stu- levels ranging from 1.0 to 13.0 pmol/L [31]. Both studies
dents tested before and after the written exam and upon a showed higher median plasma copeptin levels in men than in
psychological stress test [24, 25]. Increased levels of plasma women, without a correlation with age [5, 31]. Of note, the
copeptin were also reported in response to physical exercise, difference between men and women seems not to be present
but levels remained in the upper normal range [26, 27]. in the hyperosmolar range [17].
The half life of AVP is reported between 10 and 44 min, Small studies suggest that plasma copeptin levels do not
depending on the state of hydration, the chosen test method, have a significant variation in response to circadian rhythm
species and the specificity of immunoassays. Our group [32, 33] or menstrual cycle [34]. However, copeptin sensi-
recently showed that copeptin has a two-times-longer half- tively responds to even small amounts of fluid intake [35],
life in relation to AVP [17] (Fig. 2a, b). which has to be taken into account for interpretation of val-
This difference in half-lives of copeptin and AVP most ues in clinical practice.
likely reflects the distinct metabolic clearance rates. In con- In contrast to AVP, copeptin can be measured in clini-
trast to AVP, which is inactivated by plasma and tissue endo- cal routine with commercially available assays with a high-
peptidases in the kidney and in the liver [28], the catabolism standard technical performance. Two assays are currently
of copeptin remains largely unknown. The fact that copeptin available and validated, on one side the original manual
does not accumulate as a junk protein in the circulation, sandwich immunoluminometric assay (LIA) [5] and on the
and that its elimination is stopped once removed from the other side the automated immunofluorescent successor (on
circulation, argues against the role of circulating proteases. the KRYPTOR platform). Main advantages of measuring
Due to its small size, it could theoretically be cleared by the copeptin as compared to AVP are that it requires only a small
kidneys. Indeed, data show that copeptin is at least partly sample volume (50 μL of serum or plasma), no extraction

Fig. 2  Decay kinetics of AVP and copeptin. Percent change from and directly before initiation of the oral water load test, peptide lev-
baseline of AVP (open circle and dashed line) and copeptin (closed els reach maximum concentrations, with a median copeptin and AVP
square and solid line) over time is shown for hypertonic saline infu- of 33.8 (22.9; 48.2) pmol/L and 10 (5.8; 16) pg/mL, respectively. b
sion (a). Change from start of oral water load is shown in b. Hyper- The different decay dynamics of copeptin and AVP upon a water load
tonic saline infusion, oral water load and glucose infusion were and upon glucose infusion: after the oral water load test and before
administered in a continuos sequence without time interruption 5% glucose infusion copeptin levels already distinctively drop to
in between. Adapted from [17]. a The parallel increase of copeptin 18.1 (13.4; 25.8) pmol/L and AVP even more pronouncedly drop to a
and AVP upon hypertonic saline infusion: at the end of this phase median plasma AVP of 1.9 (1.2; 4.1) pg/mL

Journal of Endocrinological Investigation

step or other pre-analytical procedures, and that results are and release are suppressed. The consequence is excretion
normally available in less than 2 h. Moreover, copeptin is of free water. If polydipsia is happening over an extended
much more stable in plasma or serum ex vivo with less than period of time, it can result in reno-physiological adapta-
20% loss of recovery for at least 7 days at room temperature tions, including downregulation of the AQP2 channels in the
and at 14 days at 4 °C making the handling of patient blood kidneys, which compromises the renal medullary concentra-
samples less complicated. tion gradient. For this reason, diagnostic measures based
on urinary concentration capacity are sometimes difficult
to interpret [42].
Diabetes insipidus and primary polydipsia

Diabetes insipidus belongs to the so-called polyuria-polydip- Clinical manifestation, radiological

sia syndrome. This syndrome is not uncommon in clinical findings, and differential diagnosis
practice with increasing prevalence especially since many of polyuria‑polydipsia syndrome
life-style programs suggest that consuming several litres of
fluids a day is generally healthy. The polyuria-polydipsia Clear diagnostic distinction between the different forms
syndrome includes central or nephrogenic diabetes insipidus of diabetes insipidus and primary polydipsia is critical as
on one hand and primary polydipsia on the other hand [8]. causal treatment obviously varies and application of the
Diabetes insipidus leads to hypotonic polyuria which wrong treatment can be clinically harmful and potentially
is usually accompanied by subsequent polydipsia. In cen- life-threatening [43]. Still, diagnostic differentiation is often
tral diabetes insipidus, there is an insufficient secretion of difficult to achieve [44] especially as available functional
AVP from the posterior pituitary [36, 37]. In contrast, in test criteria lack high enough sensitivity and/or specificity
nephrogenic diabetes insipidus AVP levels are normally under certain circumstances [45], leading to frequent false
secreted, but there is a resistance towards the action of AVP diagnoses, especially in patients with primary polydipsia and
at the level of the kidneys [38]. Central diabetes insipidus is mild forms of diabetes insipidus [1, 46].
most often induced by lesions of the posterior pituitary or
the hypothalamic median eminence. The most commonly Clinical signs and symptoms
acquired causes are trauma, surgery of the pituitary, neo-
plastic, vascular, autoimmune, infectious or granulomatous The characteristic and all forms of diabetes insipidus under-
diseases. Pituitary surgery leads to central diabetes insipi- lying clinical symptoms are the polyuria and polydipsia that
dus in up to 30% of cases, which is most often of transient result from the impairment of urinary concentrating mecha-
nature. Permanent postsurgical diabetes insipidus is much nisms. Additional specific symptoms may be diagnostically
less common and only occurs in 2–10% [39]. Central diabe- supportive in individual cases. Patients with severe forms
tes insipidus can also be inherited, however, inherited forms of diabetes insipidus, especially those with osmoreceptor
are rather rare [13, 40]. As mentioned above, in most cases, defect, can show various degrees of dehydration and hyper-
thirst mechanisms are intact, therefore leading to a subse- osmolality, if water losses are not properly compensated by
quent polydipsia. If thirst mechanisms are also impaired, fluid intake. Therefore, clinical manifestations of hyperos-
as happens in the so-called osmoreceptor dysfunction, lack molality should be considered as well. These can be divided
of polydipsia may lead to hyperosmolality and dehydration into symptoms resulting from dehydration itself, which are
which can have clinically serious complications [8, 41]. mainly cardiovascular signs, and those rather caused by the
Nephrogenic diabetes insipidus is also most often hyperosmolality, which are predominantly neurologic and
acquired, with the best-known cause being drug-induced reflect brain dehydration as a result of water shifts out of the
nephrogenic diabetes insipidus. Mainly lithium intake is central nervous system (CNS) into the extracellular space.
known to induce nephrogenic diabetes insipidus [38]. Fur- Neurologic manifestations may include non-specific symp-
thermore, it can also be inherited, with mutations the key toms such as irritability and cognitive dysfunction but also
proteins of AVPR2 and AQP2. more severe clinical signs such as disorientation, reduced
In contrast, in primary polydipsia, AVP secretion and consciousness, seizure, coma, focal neurologic deficits
renal action is not affected. The primary problem is an and cerebral infarction [47]. In clinical practice, however,
excessive fluid intake over a longer period of time. Very most patients with diabetes insipidus have an intact thirst
rarely, it can result from an abnormality in the thirst center perception.
(in which case it is called dipsogenic diabetes insipidus). Patients with central diabetes insipidus typically report
However, much more often, it is seen in different psychi- a sudden onset of signs and symptoms and a more frequent
atric disorders (called psychognic polydipsia). Subsequent nocturia, which is remarkably constant by nature. The sud-
to the excessive fluid intake osmolality and AVP synthesis den onset of clinical manifestation results from the fact that

Journal of Endocrinological Investigation

urinary concentration usually can be maintained until the has its place is on diagnosing the underlying pathology once
residual capacity of the AVP producing neurons in the hypo- central diabetes insipidus is established. Here, an MRI of
thalamus falls below 10–15% of normal, after which urine the sellar and suprasellar regions with gadolinium needs to
output increases significantly. It has also been reported that be obtained to evaluate for any lesions of the pituitary or
patients with diabetes insipidus, in contrast to patients with hypothalamic anatomy (including macroadenomas, empty
primary polydipsia, prefer cold water as best quenching bev- sella, infiltrative or inflammatory diseases).
erage and have persistent symptoms, in contrast to more fluc-
tuating symptoms in patients with primary polydipsia [43]. Tests for differential diagnosis
Though, having only recently critically evaluated clinical
signs and symptoms of both patient groups in a prospective For the last decades the standard diagnostic test for the eval-
design, we can say that although the majority of patients uation of polyuria-polydipsia syndrome was the classical
with diabetes insipidus indeed are reporting a sudden onset water-deprivation test [37]. With this test, insufficient AVP
of symptoms, still more than one-third tend to experience secretion or effect is diagnosed upon insufficient concentra-
a slow process. Also, the majority of patients with primary tion capacity of the kidneys over osmotic stimulation which
polydipsia reported nightly drinking and preferred cold is reached with a prolonged period of thirsting (usually 16 h)
beverages, and even a higher percentage indicated a sus- and its response to exogenous AVP administration [14, 15,
tained character of symptoms. Interestingly, less than 30% 46]. Interpretation of the test varies, but the “classical”
of patients with primary polydipsia had a diagnosis of psy- interpretation is based on results on data from Miller et al.
chiatric disease, which was the same prevalence as found in [37]. In this study, 29 patients with central diabetes insipi-
patients with diabetes insipidus. In summary, in the majority dus (11 with a partial diabetes insipidus), 2 patients with
of patients, clinical signs and symptoms are neither specific nephrogenic diabetes insipidus and 5 patients with primary
nor sensitive enough to reliably differentiate between the polydipsia were evaluated. Patients with a urinary osmo-
different entities of the polyuria-polydipsia syndrome [14]. lality below 300 mOsm/kg during the water-deprivation
test were diagnosed to have complete diabetes insipidus. If
Radiological characteristics these patients increased with their urinary osmolality > 50%
after exogenous AVP administration, the final diagnosis was
Assessment of the posterior pituitary hyperintensity on T1 complete central diabetes insipidus, and conversely, if their
images and the pituitary stalk in MRI has historically been urinary osmolality after exogenous AVP administration did
reported to provide diagnostically important information in increase less than 50%, the final diagnosis was complete
the differential diagnosis of polyuria-polydipsia syndrome. nephrogenic diabetes insipidus. In those patients in whom
This idea is based on the fact that the normal posterior pitui- urinary osmolalities increased to values between 300 and
tary demonstrates the so-called “bright spot”, which is char- 800 mOsm/kg upon water deprivation, partial central diabe-
acterized by an area of hyperintensity in the posterior part tes insipidus or primary polydipsia is present. Partial central
of the sella turcica in sagittal views on T1-weighted images diabetes insipidus patients increased upon exogenous AVP
[48], and which is thought to result from the T1-shortening administration > 9% whereas patients with primary polydip-
effects of AVP, stored in neurosecretory granules of the sia increased less than 9%.
posterior lobe of the pituitary [49]. Although earlier, small However, the these cutoff values derive from only one
studies indeed observed an absence of this bright-spot phe- single post-hoc analysis of this small cohort of patients
nomenon in patients with central diabetes insipidus [50], and as evident in the raw data show a quite wide overlap in
larger subsequent studies reported an age-related absence of urinary osmolality levels [37]. Recent data in fact, aiming
the bright spot in more than 50% of normal, healthy subjects to validate these findings, showed a diagnostic accuracy of
[51] rather than a disease-specific phenomenon. On the other using these criteria in the classical water-deprivation test of
hand, individual cases with persistent bright spot, despite only around 70%, with an especially low diagnostic accuracy
the presence of central diabetes insipidus, have also been in patients with primary polydipsia [14, 46].
reported [52]. In our recent prospective imaging analysis Zerbe et al. [53] in 1981 aimed to improve the differential
including a larger population of patients with polyuria-poly- diagnosis of the polyuria-polydipsia syndrome by proposing
dipsia syndrome, we found a persistent bright spot in 36% of the so-called “direct” test. Hereby, plasma AVP is meas-
patients with central diabetes insipidus, and a missing equiv- ured upon osmotic stimulation not only by thirsting, but by
alent in 36% of patients diagnosed with primary polydipsia stimulation with hypertonic saline infusion. AVP levels are
[14]. Based on these data, the presence or absence of the then interpreted in relation to the area of normality describ-
bright spot is neither sensitive nor specific enough as a dif- ing the physiological relationship between AVP release
ferential diagnostic test in patients with unknown polyuria- and plasma osmolality. Patients with osmotically stimu-
polydipsia syndrome. Where diagnostic imaging definitely lated plasma AVP levels above the area of normality are

Journal of Endocrinological Investigation

diagnosed as nephrogenic diabetes insipidus, patients with accuracy of 94% in differentiating patients with central dia-
levels below the area of normality as central diabetes insipi- betes insipidus from patients with primary polydipsia [46].
dus, and patients with levels within the normal area as pri- Timper et al. then aimed to further improve the efficacy
mary polydipsia [40, 53]. Reassuringly, the results showed of osmotic stimulation using a combined water-deprivation
that direct measurement of plasma AVP has the potential test followed by 3% saline infusion. With this procedure,
to improve the diagnostic accuracy of the classical “indi- the aim was to increase plasma sodium levels to above
rect” water-deprivation test with interpretation of urinary 147 mmol/L which is only rarely reached by water depriva-
osmolality levels. However, despite these promising data, tion alone, especially in patients with primary polydipsia or
this direct test based on AVP measurement did not enter mild forms of diabetes insipidus. In this study, 55 patients
everyday clinical routine. Unfortunately, recent studies failed with nephrogenic or central diabetes insipidus or primary
to confirm these promising data, when using commercially polydipsia were included [15]. The first finding was that in
available AVP assays. Specifically, with these commercially a larger number of patients with nephrogenic diabetes insipi-
available assays, a correct diagnosis was only reached in dus it was confirmed that using a single baseline copep-
38% of patients and diagnostic accuracy was especially low tin level of > 21.4 pmol/L without prior thirsting, patients
in the differentiation between partial central diabetes insipi- with nephrogenic diabetes insipidus can be easily identi-
dus and primary polydipsia [46]. The problem is that first, an fied. As expected, however, baseline copeptin values in all
accurate definition of the normal physiological relationship other diagnoses (i.e., central diabetes insipidus and primary
describing plasma AVP as a function of osmotic activity polydipsia) largely overlapped. The second finding was that
has long been missed [54], but is an important condition for upon osmotic stimulation, a copeptin level of > 4.9 pmol/L
the use of direct AVP measurement [46]. Second, the AVP differentiated patients with central diabetes insipidus from
assay per se has several technical limitations, resulting in a patients with primary polydipsia with a high diagnostic
high pre-analytical instability [1, 55, 56]. Of note, the few accuracy of 96%. AVP was also measured in this study,
reliable assays are not commercially available. with a validated assay, showing a slightly lower diagnostic
accuracy [15].
Copeptin as new test method in the differential In the most recent study, we validated this copeptin cutoff
diagnosis of hypotonic polyuria of 4.9 pmol/L in an international multicenter study including
156 patients with the polyuria-polydipsia syndrome [14]. We
Within the last years, the direct test has returned, due to the further aimed to simplify the test protocol by omitting the
availability of the copeptin assay [5, 16, 57]. thirsting period and using only the hypertonic saline infu-
In a proof-of-concept study in 2007 it was shown that sion, with the aim to increase plasma sodium levels to at
in patients after pituitary surgery, an insulin tolerance test least 150 mmol/L. Towards this aim, hypertonic saline was
inducing hypoglycemia and therefore a non-osmotic stress, initially given as a bolus dose (250 mL over 10–15 min),
led to a significant induction of copeptin levels in those followed by a continuos infusion rate of 0.15 ml/kg/min until
patients with an intact posterior pituitary function, but the serum sodium concentration exceeded ≥ 150 mmol/L. At
copeptin levels remained low in patients with postsurgical this time point, copeptin was measured, and serum osmolal-
diabetes insipidus [58]. Clearly, however, induction of hypo- ity was normalized by glucose infusion and standardized
glycemia is not an attractive test for differential diagnosis of fluid intake [14]. The results showed that 97% of the patients
polyuria-polydipsia syndrome, and is not feasible as a rou- were correctly diagnosed with the pre-defined copeptin cut-
tine test since it can be associated with severe hypoglycemia off level of > 4.9 pmol/L (Fig. 3). Post hoc, an even slightly
and is contraindicated in patients with cardiovascular disease better cutoff of 6.5 pmol/L was found. The diagnostic accu-
or seizure history. racy was similar in the differential diagnosis of patients with
We therefore aimed to develop a routine test for differ- partial diabetes insipidus and patients with primary polydip-
ential diagnosis of polyuria-polydipsia syndrome based on sia with a correct diagnosis in 95%. Nephrogenic diabetes
copeptin. First, Fenske et al. evaluated the diagnostic accu- insipidus patients were only included in the beginning of
racy of the classical water-deprivation test plus measuring the study and all three patients again exceeded the cutoff
plasma copeptin [46]. In this cohort of 50 patients with dia- level of 21.4 pmol/L. Of note, and in contrast to the study
betes insipidus or primary polydipsia, plasma copeptin levels of Fenske et al., the copeptin–sodium ratio did not improve
after water deprivation (from 24:00 to 08:00) > 20 pmol/L the diagnostic accuracy of the classical water-deprivation
diagnosed patients with nephrogenic diabetes insipidus, test [46], and had a low diagnostic accuracy of only 44%
while levels < 2.6 pmol/L indicated central diabetes insipi- [14]. Therefore, copeptin measurement after osmotic stimu-
dus. A ratio of the Δplasma copeptin levels (before and after lation with thirsting alone cannot be recommended. This
the water-deprivation phase) to the plasma sodium level at is most probably because water deprivation alone does not
the end of the deprivation phase showed a high diagnostic lead to a high enough osmotic stimulation to increase AVP

Journal of Endocrinological Investigation

These data underline the fact that sufficient osmotic stim-

ulation as reached by hypertonic saline infusion is impor-
tant to obtain reliable diagnostic results of plasma copeptin.
However, it is important to note that hypertonic saline infu-
sion requires close monitoring of sodium levels to ensure
increase of plasma sodium levels into the hyperosmotic
range [36, 59] while preventing osmotic overstimulation
[14]. Also rapid normalization of sodium levels after the
osmotic stimulation is crucial to guarantee the safety of the
test [14]. It is therefore crucial to measure sodium levels
every 30 min during the hypertonic saline test.
Based on these results, it was concluded that the hyper-
tonic saline test plus copeptin measurement might replace
the classical water-deprivation test in the future differential
diagnosis of hypotonic polyuria [60]. A new algorithm with
and without availability of copeptin measurement for the
differential diagnosis of polyuria-polydipsia syndrome is
displayed in Fig. 4.
Importantly, although side effects like headache, vertigo
and malaise were more common during the hypertonic saline
infusion test than with the classical water-deprivation test,
the majority of patients preferred the hypertonic saline stim-
ulation with copeptin measurement over the standard water-
deprivation test. Most probably, the reason was the clearly
shorter test duration (approximately 2 h for the hypertonic
saline test versus 17 h for the water-deprivation test) [14].


In conclusion, copeptin is a stable surrogate marker of AVP

and provides a valuable and reliable diagnostic marker in
the differential diagnosis of the polyuria-polydipsia syn-
drome. In patients with unclear hypotonic polyuria and
polydipsia, measurement of basal copeptin levels identifies
nephrogenic diabetes insipidus. In patients with high sus-
picion of complete central diabetes insipidus, an overnight
water-deprivation test (i.e., thirsting from midnight until
8 am of the following day) might confirm diagnosis pro-
Fig. 3  Receiver operating curve analysis for the hypertonic saline vided urine osmolality remains below 300 mOsm/kg and
infusion test (a) as compared to the water-deprivation test (b) Modi- plasma sodium levels increase above 147 mmol/L. In all
fied from [14].
other patients, copeptin measurement after osmotic stimula-
tion with 3% saline solution with the aim to increase plasma
or copeptin levels. In fact, most patients in our study did not sodium level ≥ 150 mmol/L is recommended. Importantly,
reach hyperosmotic plasma sodium levels during the classi- close monitoring of plasma sodium levels is needed to ascer-
cal water-deprivation test. Also, the proposed copeptin cutoff tain a diagnostically meaningful increase in plasma sodium
level of < 2.6 pmol/L after an overnight water-deprivation within the hyperosmotic range while preventing a marked
test to diagnose complete central diabetes insipidus had a increase [14].
diagnostic accuracy of 78%.

Journal of Endocrinological Investigation

Suspected hypotonic polyuria

Confirm the presence of polyuria Urinary Volume <50ml/kg/24h


Urine osmolality <800mOsm/kg GU evaluaon

Serum Sodium, Plasma osmolality

Low serum sodium (<135) High serum sodium (>147)

normal serum sodium

Primary polydipsia Central or nephrogenic DI

Water deprivaon test Copepn

Copepn Copepn
Urine osmolality Urine osmolality Urine osmolality >21.4pmol/L <21.4pmol/L
>800mOsm/kg 300-800mOsm/kg <300mOsm/kg
Nephrogenic DI Hypertonic saline test
Mild Desmopressin Desmopressin
Primary test test Smulated copepn Smulated copepn
polydipsia (at plasma sodium (at plasma sodium
>150mmol/L >4.9pmol/L >150mmol/L <4.9pmol/L
>9% <9% >50% <50%
increase increase increase increase

Primary Paral Complete Nephrogenic Complete or Paral

polydipsia central central DI DI Primary polydipsia
central DI

Fig. 4  Proposed algorithm for the differential diagnosis of polyuria-polydipsia syndrome Modified from [61]

the third domain of the vasopressin precursor. Int J Peptide Protein

Res 27(3):320–324
Funding  M. Christ-Crain was supported by a grant from the Swiss 4. Land H, Schutz G, Schmale H, Richter D (1982) Nucleotide
National Science Foundation (SNF-162608) and the University Hos- sequence of cloned cDNA encoding bovine arginine vasopressin-
pital Basel, Switzerland. neurophysin II precursor. Nature 295(5847):299–303
5. Morgenthaler NG, Struck J, Alonso C, Bergmann A (2006) Assay
for the measurement of copeptin, a stable peptide derived from the
Compliance with ethical standards  precursor of vasopressin. Clin Chem 52(1):112–119
6. Nathanson MH, Moyer MS, Burgstahler AD, O’Carroll AM,
Conflict of interest MCC and WF received speaker honoraria from Brownstein MJ, Lolait SJ (1992) Mechanisms of subcellular cyto-
Thermofisher AG, the manufacturer of the Copeptin assay. solic ­Ca2+ signaling evoked by stimulation of the vasopressin V1a
receptor. J Biol Chem 267(32):23282–23289
Ethical approval  All procedures involving human participants were in 7. Baylis PH (1987) Osmoregulation and control of vasopres-
accordance with ethical standards of the institutional and/or national sin secretion in healthy humans. Am J Physiol 253(5 Pt
research committee with the Helsinki declaration and its later amend- 2):R671–R678
ments or comparable ethical standards. 8. Christ-Crain M, Fenske W (2016) Copeptin in the diagnosis of
vasopressin-dependent disorders of fluid homeostasis. Nat Rev
Informed consent  Informed consent was obtained from all individuals Endocrinol 12(3):168–176
included in one of the cited studies. 9. Nagy G, Mulchahey JJ, Smyth DG, Neill JD (1988) The glycopep-
tide moiety of vasopressin-neurophysin precursor is neurohypo-
physial prolactin releasing factor. Biochem Biophys Res Commun
10. Hyde JF, North WG, Ben-Jonathan N (1989) The vasopressin-
References associated glycopeptide is not a prolactin-releasing factor: studies
with lactating Brattleboro rats. Endocrinology 125(1):35–40
11. Barat C, Simpson L, Breslow E (2004) Properties of human vaso-
1. Robertson GL (1995) Diabetes insipidus. Endocrinol Metab Clin
pressin precursor constructs: inefficient monomer folding in the
N Am 24(3):549–572
absence of copeptin as a potential contributor to diabetes insipi-
2. Holwerda DA (1972) A glycopeptide from the posterior lobe of
dus. Biochemistry 43(25):8191–8203
pig pituitaries I Isolation and characterization. Eur J Biochem
12. Parodi AJ (2000) Protein glucosylation and its role in protein fold-
FEBS 28(3):334–339
ing. Annu Rev Biochem 69:69–93
3. Levy B, Chauvet MT, Chauvet J, Acher R (1986) Ontogeny of
bovine neurohypophysial hormone precursors. II. Foetal copeptin,

Journal of Endocrinological Investigation

13. Birk J, Friberg MA, Prescianotto-Baschong C, Spiess M, Rut- 32. Darzy KH, Dixit KC, Shalet SM, Morgenthaler NG, Brabant G
ishauser J (2009) Dominant pro-vasopressin mutants that cause (2010) Circadian secretion pattern of copeptin, the C-terminal
diabetes insipidus form disulfide-linked fibrillar aggregates in the vasopressin precursor fragment. Clin Chem 56(7):1190–1191
endoplasmic reticulum. J Cell Sci 122(Pt 21):3994–4002 33. Beglinger S, Drewe J, Christ-Crain M (2016) The circadian
14. Fenske W, Refardt J, Chifu I et  al (2018) A copeptin-based rhythm of copeptin, the C-terminal portion of Arginin Vasopres-
approach in the diagnosis of diabetes insipidus. N Engl J Med sin. In: Poster presentation, SGED Congress Nov. 17–18, 2016,
379(5):428–439 Bern, Switzerland
15. Timper K, Fenske W, Kuhn F et al (2015) Diagnostic accuracy of 34. Puder JJ, Blum CA, Mueller B, De Geyter C, Dye L, Keller U
copeptin in the differential diagnosis of the polyuria-polydipsia (2006) Menstrual cycle symptoms are associated with changes in
syndrome: a prospective multicenter study. J Clin Endocrinol low-grade inflammation. Eur J Clin Investig 36(1):58–64
Metab 100(6):2268–2274 35. Walti C, Siegenthaler J, Christ-Crain M (2014) Copeptin levels
16. Balanescu S, Kopp P, Gaskill MB, Morgenthaler NG, Schindler are independent of ingested nutrient type after standardised meal
C, Rutishauser J (2011) Correlation of plasma copeptin and administration—the CoMEAL study. Biomarkers 19(7):557–562
vasopressin concentrations in hypo-, iso-, and hyperosmolar 36. Robertson GL (1976) The regulation of vasopressin function in
States. J Clin Endocrinol Metab 96(4):1046–1052 health and disease. Recent Prog Horm Res 33:333–385
17. Fenske WK, Schnyder I, Koch G et al (2018) Release and decay 37. Miller M, Dalakos T, Moses AM, Fellerman H, Streeten DH
kinetics of copeptin vs AVP in response to osmotic alterations (1970) Recognition of partial defects in antidiuretic hormone
in healthy volunteers. J Clin Endocrinol Metab 103(2):505–513 secretion. Ann Intern Med 73(5):721–729
18. Szinnai G, Morgenthaler NG, Berneis K et al (2007) Changes in 38. Bockenhauer D, Bichet DG (2015) Pathophysiology, diagnosis
plasma copeptin, the c-terminal portion of arginine vasopressin and management of nephrogenic diabetes insipidus. Nat Rev
during water deprivation and excess in healthy subjects. J Clin Nephrol 11(10):576–588
Endocrinol Metab 92(10):3973–3978 39. Nemergut EC, Zuo Z, Jane JA Jr, Laws ER Jr (2005) Predictors of
19. Morgenthaler NG, Muller B, Struck J, Bergmann A, Redl H, diabetes insipidus after transsphenoidal surgery: a review of 881
Christ-Crain M (2007) Copeptin, a stable peptide of the arginine patients. J Neurosurg 103(3):448–454
vasopressin precursor, is elevated in hemorrhagic and septic 40. Babey M, Kopp P, Robertson GL (2011) Familial forms of dia-
shock. Shock. 28(2):219–226 betes insipidus: clinical and molecular characteristics. Nat Rev
20. Katan M, Fluri F, Morgenthaler NG et al (2009) Copeptin: a Endocrinol 7(12):701–714
novel, independent prognostic marker in patients with ischemic 41. Thompson CJ, Baylis PH (1987) Thirst in diabetes insipi-
stroke. Ann Neurol 66(6):799–808 dus: clinical relevance of quantitative assessment. Q J Med
21. Reichlin T, Hochholzer W, Stelzig C et al (2009) Incremental 65(246):853–862
value of copeptin for rapid rule out of acute myocardial infarc- 42. Epstein FH, Kleeman CR, Hendrikx A (1957) The influence of
tion. J Am Coll Cardiol 54(1):60–68 bodily hydration on the renal concentrating process. J Clin Inves-
22. Katan M, Christ-Crain M (2010) The stress hormone copeptin: tig 36(5):629–634
a new prognostic biomarker in acute illness. Swiss Med Wkly. 43. Fenske W, Allolio B (2012) Clinical review: current state and
140:w13101 future perspectives in the diagnosis of diabetes insipidus: a clini-
23. Katan M, Morgenthaler N, Widmer I et al (2008) Copeptin, cal review. J Clin Endocrinol Metab 97(10):3426–3437
a stable peptide derived from the vasopressin precursor, cor- 44. Carter AC, Robbins J (1947) The use of hypertonic saline infu-
relates with the individual stress level. Neuro Endocrinol Lett sions in the differential diagnosis of diabetes insipidus and psy-
29(3):341–346 chogenic polydipsia. J Clin Endocrinol Metab 7(11):753–766
24. Siegenthaler J, Walti C, Urwyler SA, Schuetz P, Christ-Crain M 45. Barlow ED, De Wardener HE (1959) Compulsive water drinking.
(2014) Copeptin concentrations during psychological stress: the Q J Med 28(110):235–258
PsyCo study. Eur J Endocrinol 171(6):737–742 46. Fenske W, Quinkler M, Lorenz D et al (2011) Copeptin in the
25. Urwyler SA, Schuetz P, Sailer C, Christ-Crain M (2015) Copep- differential diagnosis of the polydipsia-polyuria syndrome—revis-
tin as a stress marker prior and after a written examination—the iting the direct and indirect water deprivation tests. J Clin Endo-
CoEXAM study. Stress 18(1):134–137 crinol Metab 96(5):1506–1515
26. Maeder MT, Staub D, Brutsche MH et al (2010) Copeptin response 47. Palevsky PM, Bhagrath R, Greenberg A (1996) Hypernatremia in
to clinical maximal exercise tests. Clin Chem 56(4):674–676 hospitalized patients. Ann Intern Med 124(2):197–203
27. Hew-Butler T, Hoffman MD, Stuempfle KJ, Rogers IR, Mor- 48. Fujisawa I, Nishimura K, Asato R et al (1987) Posterior lobe of
genthaler NG, Verbalis JG (2011) Changes in copeptin and bio- the pituitary in diabetes insipidus: MR findings. J Comput Assist
active vasopressin in runners with and without hyponatremia. Clin Tomogr 11(2):221–225
J Sport Med 21(3):211–217 49. Arslan A, Karaarslan E, Dincer A (1999) High intensity signal
28. Robertson GL (2001) Antidiuretic hormone. Normal and disor- of the posterior pituitary. A study with horizontal direction of
dered function. Endocrinol Metab Clin N Am 30(3):671–694 (vii) frequency-encoding and fat suppression MR techniques. Acta
29. Nakajima A, Lu Y, Kawano H, Horie S, Muto S (2015) Associa- Radiol 40(2):142–145
tion of arginine vasopressin surrogate marker urinary copeptin 50. Moses AM, Clayton B, Hochhauser L (1992) Use of T1-weighted
with severity of autosomal dominant polycystic kidney disease MR imaging to differentiate between primary polydipsia and cen-
(ADPKD). Clin Exp Nephrol 19(6):1199–1205 tral diabetes insipidus. AJNR Am J Neuroradiol 13(5):1273–1277
30. Roussel R, Fezeu L, Marre M et al (2014) Comparison between 51. Cote M, Salzman KL, Sorour M, Couldwell WT (2014) Normal
copeptin and vasopressin in a population from the community and dimensions of the posterior pituitary bright spot on magnetic reso-
in people with chronic kidney disease. J Clin Endocrinol Metab nance imaging. J Neurosurg 120(2):357–362
99(12):4656–4663 52. Hannon MJ, Orr C, Moran C et al (2012) Anterior hypopituita-
31. Bhandari SS, Loke I, Davies JE, Squire IB, Struck J, Ng LL (2009) rism is rare and autoimmune disease is common in adults with
Gender and renal function influence plasma levels of copeptin in idiopathic central diabetes insipidus. Clin Endocrinol (Oxf)
healthy individuals. Clin Sci (Lond) 116(3):257–263 76(5):725–728
53. Zerbe RL, Robertson GL (1981) A comparison of plasma
vasopressin measurements with a standard indirect test

Journal of Endocrinological Investigation

in the differential diagnosis of polyuria. N Engl J Med 58. Katan M, Morgenthaler NG, Dixit KC et al (2007) Anterior and
305(26):1539–1546 posterior pituitary function testing with simultaneous insulin tol-
54. Baylis PH, Gaskill MB, Robertson GL (1981) Vasopressin secre- erance test and a novel copeptin assay. J Clin Endocrinol Metab
tion in primary polydipsia and cranial diabetes insipidus. Q J Med 92(7):2640–2643
50(199):345–358 59. Robertson GL, Shelton RL, Athar S (1976) The osmoregulation
55. Milles JJ, Spruce B, Baylis PH (1983) A comparison of diag- of vasopressin. Kidney Int 10(1):25–37
nostic methods to differentiate diabetes insipidus from primary 60. Rosen CJ, Ingelfinger JR (2018) A reliable diagnostic test for
polyuria: a review of 21 patients. Acta Endocrinol (Copenh) hypotonic polyuria. N Engl J Med 379(5):483–484
104(4):410–416 61. Gubbi S, Hannah-Shmouni F, Koch CA, Verbalis JG (2000) Diag-
56. Kluge M, Riedl S, Erhart-Hofmann B, Hartmann J, Waldhauser nostic Testing for Diabetes Insipidus. In: Feingold KR, Anawalt
F (1999) Improved extraction procedure and RIA for determina- B, Boyce A et  al (eds) Endotext., Inc., South
tion of arginine 8-vasopressin in plasma: role of premeasurement Dartmouth
sample treatment and reference values in children. Clin Chem
45(1):98–103 Publisher’s Note Springer Nature remains neutral with regard to
57. Fenske WK, Schnyder I, Koch G et al (2017) Release and decay jurisdictional claims in published maps and institutional affiliations.
kinetics of copeptin versus AVP in response to osmotic altera-
tions in healthy volunteers. J Clin Endocrinol Metab. https​://doi.