medwireNews: The arginine-stimulated copeptin test is less accurate at diagnosing arginine vasopressin (AVP) deficiency than the hypertonic saline-stimulated copeptin test, indicate findings from a noninferiority trial.
In this first head-to-head comparison of these tests, 158 adults with polydipsia and hypotonic polyuria or a confirmed diagnosis of AVR deficiency completed both tests on different days. A final diagnosis of AVP deficiency (44%) or primary polydipsia (56%) was made independently by two endocrinologists 3 months later based on the patients’ medical history, laboratory results, and the hypertonic saline-stimulated copeptin test results. However, the endocrinologists were blinded to the arginine-stimulated copeptin test results.
The arginine-stimulated copeptin test, based on a copeptin level below 2.4 pmol/L for AVP deficiency diagnosis and above 3.8 pmol/L for primary polydipsia, correctly differentiated patients with AVP deficiency from those with polydipsia with an accuracy, sensitivity, and specificity of 74.4%, 75.4%, and 73.6%, respectively.
However, the hypertonic saline-stimulated copeptin test showed better diagnostic accuracy. Based on copeptin cutoffs of 2.7 pmol/L and below for AVP deficiency and 4.9 pmol/L and above for polydipsia, the test correctly differentiated AVP deficiency with an accuracy, sensitivity, and specificity of 95.6%, 91.3%, and 98.9%, respectively.
This gave a significant mean difference for overall diagnostic accuracy of 21.2 percentage points in favor of the hypertonic saline-stimulated copeptin test, which was more than the prespecified 10 percentage points and meant that the arginine-stimulated copeptin test did not meet noninferiority for AVP deficiency diagnosis, say Julie Refardt (University Hospital Basel, Switzerland) and colleagues in The New England Journal of Medicine.
They comment that differentiating AVP deficiency and AVP resistance from primary polydipsia is crucial “because treatment of the three conditions differ and potential misdiagnosis carries the risk of severe complications.”
The researchers note that the arginine-stimulated test, based on a copeptin level of 2.4–3.8 pmol/L, was also inferior to the hypertonic saline-stimulated test, at a copeptin level of 2.7–4.9 pmol/L, for differentiating the 28 patients with partial AVP deficiency from those with primary polydipsia, doing so with an accuracy of 68.7% versus 95.2%.
Nevertheless, Refardt and team point out that the majority (72%) of patients preferred the arginine-stimulated test, and when they looked at different cutoff points for copeptin, they did find it had a high specificity of 90.9% for correctly diagnosing AVP deficiency at a copeptin level of 3.0 pmol/L or less, and 91.4% for diagnosing primary polydipsia at a copeptin level above 5.2 pmol/L, at respective sensitivities of 59.5% and 56.4%.
“It is possible that the diagnostic accuracy of arginine stimulation could be improved by raising serum osmolality by overnight water deprivation,” they suggest.
With both tests, thirst and headache were the most common adverse events, as reported by a respective 99% and 37% of patients after the arginine-stimulated test and by 98% and 59% after the hypertonic-saline stimulated test, respectively.
“The overall intensity of adverse effects was low with both tests but occurred with more frequency and higher intensity during hypertonic-saline stimulation,” say the investigators.
John Newell‑Price (University of Sheffield, UK) comments in a related editorial that “[g]iven the simplicity of the arginine-stimulation test, it is a shame that it was found to be inferior.”
The editorialist says that “hypertonic saline is clearly the superior test to discriminate between AVP deficiency and primary polydipsia,” but he adds that “among patients in whom the use of hypertonic saline is contraindicated, or because of patient preferences, the arginine-stimulation test is the next best option.”
Newell-Price stresses, however, that “regardless of which test is used, practitioners need to be aware that the cutoffs for appropriate copeptin levels are not generalizable among copeptin assays.”
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2023 Springer Healthcare Ltd, part of the Springer Nature Group
N Engl J Med 2023; 389: 1877–1887
N Engl J Med 2023; 389: 1920–1921