Use of copeptin in interpretation of the water deprivation test

Abstract Introduction Currently, the water deprivation test remains the standard method for distinguishing primary polydipsia (PP) from cranial diabetes insipidus (cDI) and nephrogenic diabetes insipidus (nDI). There is increasing interest in a direct estimate of antidiuretic hormone using plasma copeptin as a stable and reliable surrogate marker. We present our experience of measuring copeptin during the water deprivation test. Methods Forty‐seven people (17 men) underwent a standard water deprivation test between 2013 and 2021. Plasma copeptin was measured at the start of the test and at the end of the period of water deprivation (maximum osmotic stimulation). Results were classified using prespecified diagnostic criteria. As it is known that a significant proportion of tests will reveal indeterminate results, a final diagnosis was obtained by including relevant pre‐ and post‐test clinical criteria. This diagnosis was then used to plan individual treatment. Results Basal and stimulated copeptin were significantly higher in the nephrogenic DI group than other categories (p < .001). There was no significant difference in basal or stimulated copeptin between PP, cDI or partial DI. Nine results were indeterminate where the serum and urine osmolality did not give a unified diagnosis. Stimulated copeptin was helpful in reclassifying these patients into the final diagnostic groups. Conclusion Plasma copeptin has additional clinical utility in interpretation of the water deprivation test and may continue to have a place alongside newer stimulation tests.


| INTRODUC TI ON
For many years, the water deprivation test has been the standard method for differentiating between the causes of the polyuria/ polydipsia syndrome. 1 The test will distinguish between diabetes insipidus and the syndrome of primary polydipsia (PP). With cranial diabetes insipidus (cDI), there is a deficiency of antidiuretic hormone (ADH), also known as arginine vasopressin (AVP) following damage to the hypothalamus or posterior pituitary, resulting in an inability to retain water and concentrate urine. With PP, there is an urge to drink excessively that results in high urine output. It is recognized that there may be partial deficiency of ADH-partial cranial diabetes insipidus (pDI) resulting in abnormal but not total inability to concentrate urine. There are a smaller group of people who develop resistance to the action of antidiuretic hormone at the level of the kidney (nephrogenic DI). It is important to distinguish between PP

| ME THODS
This was a retrospective single centre-based study involving 47 people who underwent a standardized water deprivation test between 2013 and 2021. All patients had been referred for investigation of polyuria/ polydipsia with the differential diagnosis including diabetes insipidus or primary polydipsia. Prior to the test, other causes such as diabetes mellitus, hypercalcaemia, renal insufficiency or other electrolyte disturbance had been excluded.
The patients were admitted to the endocrine investigation unit at 08.00 on the morning of the test. The protocol is summarized below: fluid was not restricted before arrival in the unit. Consumption of any liquids during the test was prohibited, and patients were not allowed to leave the ward. Body weight, serum and urine osmolality, electrolytes and copeptin were measured at the start of the test.
Serum and urine osmolality and urine volume were recorded every 90 minutes. Copeptin was repeated at the end of 6 h of water restriction, at which point a decision was made, based on the serum and urine osmolality and urine volumes, whether to proceed with desmopressin. Weight was measured every 2 h. If desmopressin was administered, serum and urine osmolality were followed hourly for the next 4 h. The test was concluded if, at any point, body weight fell by >5% or the urine osmolality exceeded 700 mmol/kg (at which point serum and urine osmolality, weight and electrolytes were measured).
The criteria developed for the interpretation of the water deprivation test are shown in Table 1. The results for each test were reviewed by the multidisciplinary endocrine team. Based on the presenting clinical picture and results of water deprivation, a provisional diagnosis was determined. There were four potential diagnoses: cranial diabetes insipidus (cDI), partial diabetes insipidus (pDI), primary polydipsia (PP), which is essentially an ability to normally concentrate urine in the face of water restriction, and nephrogenic diabetes insipidus (nephrogenic DI). Previous data have suggested that the diagnosis of diabetes insipidus could be improved by the addition of copeptin measurement. 8 Suggested copeptin diagnostic cut-offs derived from previous studies are also included in Table 1.
It is accepted that the water deprivation test can produce equivocal results particularly in distinguishing pDI and PP. It is usual practice in these situations to combine the results of the water deprivation test with other clinical information to conclude a presumptive diagnosis. This might then involve a trial of desmopressin or advice about fluid consumption and continued clinical observation. Following a variable period of follow-up, a substantive diagnosis is then reached.
The patient may continue to be followed or discharged depending upon the clinical course. Although urine volumes and weight were recorded as part of the test, these were kept as paper records and were not available for analysis in this cohort.
Copeptin was measured using the Brahms CT-proAVP Kryptor assay. According to the manufacturer's instruction for use, the limit of detection was assessed as being 1.2 pmoL/L, the interassay coefficient of variation as being 5.5% for mean copeptin concentrations of an internal quality control material at a concentration of 5.4 pmoL/L, 5.1% for a serum pool with a copeptin concentrations of 16.9 pmoL/L and 3.5% for a quality control material at a concentration of 102.6 pmoL/L.

| RE SULTS
Between 2013 and 2021, 47 consecutive water deprivation tests were performed at University Hospital Plymouth. Copeptin was measured at baseline and at the end of a 6-h period of water deprivation for each subject. Details of the clinical cohort are displayed in Table 2.
All the subjects described had presented with a syndrome of polyuria and polydipsia. All had demonstrated high volumes of dilute urine production. The differential diagnosis for this presentation includes cDI, pDI, PP or nephrogenic DI. It is accepted that the water deprivation test alone can struggle to accurately define a specific diagnosis in all cases. There is often a discrepancy in the diagnostic criteria with, for example, an individual maintaining a normal serum osmolality with water deprivation but not achieving the urine concentration required to make a diagnosis of PP or alternatively showing an ability to adequately concentrate the urine but also showing a serum osmolality that climbs above the normal range. The water deprivation test therefore requires interpretation in the clinical context for that individual considering their medical history and subsequent clinical course. The four diagnostic groups shown in Table 2 are based on the tests results together with 1-8 years of clinical follow-up. The largest group were those with a final diagnosis of PP (n = 24), and pDI was the next largest group (n = 11). There were relatively few cases of cDI (n = 8) tested this may reflect the fact that in the correct clinical context (for example following pituitary surgery) patients may be directly started on desmopressin therapy based on close observation of post-operative fluid balance and electrolytes and do not proceed to formal water deprivation. There were no significant age or gender differences between the groups.  there were no significant differences between groups. Figure 2B shows the individual copeptin values following a period of water deprivation. As the values of copeptin are so much higher for the nephrogenic DI group, they have been omitted from the figure to emphasize the differences between the other three diagnoses. Again, there is no significant difference between groups. The groups that do get more dehydrated (cDI and pDI) are those with defects in copeptin release and also do not show a significant rise.
There are two individuals (one in the cDI group and one in the pDI group) who show an unexpected rise in copeptin despite having a confirmed diagnosis of presumptive antidiuretic hormone deficiency. These two cases are discussed further below. Figure 3  anterior pituitary hormone replacement and desmopressin as an inpatient. Following water deprivation, there was no change in the serum osmolality but also no change in the urine osmolality with a peak urine osmolality of 115 mmol/Kg. Following desmopressin, the urine concentrated to 402 mmol/Kg. His stimulated copeptin was comfortably normal at 6.5 pmol/L despite not seeing a significant change in his serum concentration. He was discharged on a continuing dose of desmopressin and unfortunately readmitted to hospital 3 months later with a dilutional hyponatraemia that responded to a reduction in desmopressin dose. He had responded to desmopressin treatment, but had normal copeptin suggesting relative deficiency together with some resistance to hormone action.

| CON CLUS ION
We have studied the utility of adding copeptin measurement at baseline and following water deprivation as an addition.
Nephrogenic diabetes has significantly higher copeptin measurements on both baseline and stimulated samples. This diagnosis can be confirmed on a baseline blood test, and water deprivation may not be required. To distinguish between cDI, pDI and PP, the water deprivation test showed significant differences in both serum and urine osmolality between groups and specifically between pDI and PP. Stimulated or baseline copeptin, however, did not show a significant difference between groups (excluding nephrogenic DI).
This does not imply that adding copeptin does not have some clinical utility when added to the water deprivation test. It has previously been established that a significant proportion of test results will be indeterminate 8 and this was true for this cohort. For the indeterminate group in this cohort, the final clinical diagnosis for the majority was primary polydipsia with a smaller number being diagnosed with partial DI. The copeptin measurements for all but one person were below the suggested cut-off for diagnosing pDI.
The copeptin measurement was therefore helpful in these cases in There were also some differences in the osmolality cut-offs defining the diagnoses. The conclusions of the study were broadly similar; adding copeptin produces a modest improvement in diagnostic accuracy. Their cohort did not produce any examples of unexpectedly high copeptin measurement in individuals with a diagnosis of diabetes insipidus or partial diabetes insipidus. Subsequent studies have shown that the test could be improved by the addition of water deprivation with hypertonic saline. 10 Since then, this has developed into hypertonic saline stimulation alone or, more recently, arginine stimulation. 8,11,12 Both new tests look very promising with significantly improved diagnostic accuracy although the tests may not be suitable for all patients.
One of the strengths of these data is that they represent realworld data with a relatively long period of follow-up for most subjects after the test. The test of time helps in ensuring that the final diagnosis is secure.
A limitation of this study is the lack of information about urine volumes or change in weight over the length of the test. This information would have been available to the clinicians when interpreting the results and have been factored into the presumptive diagnosis but are not available to include in this paper. A further difficulty with using any test to diagnose ADH deficiency is that it appears to be a hormone axis where we see more variation in function with time. It is common to see transient diabetes insipidus after surgery that recovers in a few days. 13,14 Similarly with pituitary haemorrhage or infarction, DI may be transient over a longer period. 15 With inflammatory pituitary or hypothalamic lesions, DI may fluctuate with periods of recovery and then recurrence. 16 This can make the final diagnosis more elusive as it possible that it may change after the diagnostic test has been performed.
Our data would also perhaps suggest that there will be individuals where the copeptin appears normal on stimulation, but the person demonstrates an inability to concentrate urine that is reversed with desmopressin therapy. It is important to be aware of this if clinical practice switches primarily to direct stimulation test. The future paradigm may require a water deprivation test with copeptin measurement for these individuals where there is a very strong clinical suspicion despite normal copeptin stimulation.