BACKGROUND
The indirect water-deprivation test is the current referencestandard for the diagnosis of diabetes insipidus. However, it is technicallycumbersome to administer, and the results are often inaccurate. The current study compared the indirectwater-deprivation test with direct detection of plasma copeptin, a precursor-derived surrogate of arginine vasopressin.
间接缺水试验是目前诊断尿崩症的参考标准。然而，管理在技术上很麻烦，结果往往是不准确的。目前的研究比较了间接水剥夺试验和血浆和肽素的直接检测，血浆和肽素是精氨酸加压素的前体衍生物替代物。
METHODS
From 2013 to 2017, we recruited 156 patients with hypotonicpolyuria at 11 medical centers to undergo both water-deprivation and hypertonicsaline infusion tests. In the latter test, plasma copeptin was measured whenthe plasma sodium level had increased to at least 150 mmol per liter afterinfusion of hypertonic saline. Theprimary outcome was the overall diagnostic accuracy of each test as compared with the finalreference diagnosis,which was determined on the basis of medical history, test results, andtreatment response, with copeptin levels masked.
从2013年到2017年，我们在11个医疗中心招募了156名低渗性多尿患者，进行缺水和高渗盐水输注试验。在后一种试验中，当输注高渗盐水后血浆钠水平增加至至少150mmol / L时，测量血浆和肽素。 主要结果是每个测试的总体诊断准确性与最终参考诊断相比，最终参考诊断是根据病史，测试结果和治疗反应确定的，其中隐蔽了copeptin水平。
RESULTS
A total of 144 patients underwent both tests. The final diagnosis was primary polydipsia in82 patients (57%), central diabetes insipidus in 59 (41%), and nephrogenicdiabetes insipidus in 3 (2%). Overall, among the 141 patients included in theanalysis, the indirect water-deprivation test determined the correct diagnosisin 108 patients (diagnostic accuracy, 76.6%; 95% confidence interval [CI], 68.9to 83.2), and the hypertonic saline infusion test (with a copeptin cutoff levelof >4.9 pmol per liter) determined the correct diagnosis in 136 patients(96.5%; 95% CI, 92.1 to 98.6; P<0.001). The indirect water-deprivation testcorrectly distinguished primary polydipsia from partial central diabetesinsipidus in 77 of 105 patients (73.3%; 95% CI, 63.9 to 81.2), and the hypertonic saline infusion test distinguished be-tween the twoconditions in 99 of 104 patients (95.2%; 95% CI, 89.4 to 98.1; adjusted P<0.001). One serious adverse event (desmopressin-induced hyponatremia thatresulted in hospitalization) occurred during the water-deprivation test.
共有144名患者接受了两项检查。最终诊断为82例患者（57％）的原发性多饮，59例中心性尿崩症（41％）和肾性尿崩症3例（2％）。总体而言，在分析中包括的141名患者中，间接水剥夺试验确定了108名患者的正确诊断（诊断准确性，76.6％; 95％置信区间[CI]，68.9至83.2）和高渗盐水输注试验（copeptin截断水平> 4.9 pmol / L）确定136名患者的正确诊断（96.5％; 95％CI，92.1至98.6; P <0.001）。间接缺水试验正确区分105例患者中77例（73.3％; 95％CI，63.9~81.2）的原发性多饮和部分中枢性尿崩症，高渗盐水输注试验在104例中的99例中区分患者（95.2％; 95％CI，89.4至98.1;调整后的P <0.001）。在缺水试验期间发生了一种严重的不良事件（去氨加压素诱导的低钠血症，导致住院治疗）。
CONCLUSIONS
The direct measurement of hypertonic saline–stimulatedplasma copeptin had greater diagnostic accuracy than the water-deprivation testin patients with hypotonic polyuria. (Funded by the Swiss National Foundationand others; ClinicalTrials.gov number, NCT01940614.)
高渗盐水刺激的血浆和肽素的直接测量比低渗多尿患者的水剥夺试验具有更高的诊断准确性。 （由瑞士国家基金会和其他人资助;ClinicalTrials.gov编号，NCT01940614。）
论文正文
The determination of a specific diagnosis in patients withpolyuria and low plasma osmolality (i.e., hypotonic polyuria) is a frequentproblem in clinical practice. In the absence of osmotic diuresis, polyuria canresult from one of three fundamentally different conditions1: insufficientproduction and secretion of the antidiuretic hormone arginine vasopressin(central diabetes insipidus), diminished renal sensitivity to the antidiureticactivity of arginine vasopressin (nephrogenic diabetes insipidus), or primaryexcessive fluid intake (primary polydipsia).
在多尿和低血浆渗透压（即低渗性多尿）患者中确定特定诊断是临床实践中的常见问题。 在没有渗透性利尿的情况下，多尿可以由三种根本不同的病症之一引起：抗利尿激素精氨酸加压素（中枢性尿崩症）的产生和分泌不足，肾脏对精氨酸加压素（肾性尿崩症）的抗利尿活性的敏感性降低，或原发性多余的液体摄入量（原发性多饮）。
It is important to differentiate these entities becausetreatments differ substantially, and incorrect strategies may lead to severecomplications. The indirect water-deprivation test measures the maximal urineconcentration during prolonged withholding of oral liquids and the renalresponse to administered desmopressin. It is conceptually simple, butdifficulties in interpretation are common, mainly because any water diuresismay compromise the renal medullary concentration gradient and promote adownregulation of kidney aquaporin-2 water channels, which could potentiallyaffect the diagnostic value of these urinary measures. Previous attempts toimprove the diagnosis of polyuric disorders with direct measurement ofcirculating arginine vasopressin failed to gain traction in clinical practice,largely because of the technical difficulties of measuring arginine vasopressin.
区分这些实体很重要，因为治疗方法存在很大差异，不正确的策略可能会导致严重的并发症。 间接禁水测试测量在长期停留口服液体期间的最大尿液浓度和对施用的去氨加压素的肾脏反应。它在概念上很简单，但解释上的困难是常见的，主要是因为任何水利尿都可能损害肾髓质浓度梯度并促进肾水通道蛋白-2水通道的下调，这可能潜在地影响这些排尿措施的诊断价值。先前通过直接测量循环精氨酸加压素来改善多尿症的诊断的尝试未能在临床实践中获得牵引力，主要是因为测量精氨酸加压素的技术困难。
Copeptin, the C-terminal segment of the arginine vasopressinprohormone, is an arginine vasopressin surrogate with high ex-vivo stabilitythat is easy to measure. In previous studies, we reported outcome data thatsuggested that measurement of osmotically stimulated copeptin might be usefulin differentiating the various causes of hypotonic polyuria. The current studyassessed the diagnostic performance of a test measuring copeptin that wasosmotically stimulated by water deprivation or by hypertonic saline infusion ascompared with the indirect water-deprivation test.
Copeptin是精氨酸加压素激素原的C末端片段，是精氨酸加压素替代物，具有高离体稳定性，易于测量。 在先前的研究中，我们报告的结果数据表明，渗透刺激的和肽素的测量可能有助于区分低渗多尿的各种原因。本研究评估了与间接缺水试验相比，通过水剥夺或高渗盐水输注渗透刺激的和肽素测试的诊断性能。
Study Design and Patients
This international, multicenter, prospective study wasconducted at 11 tertiary medical centers in Switzerland, Germany, and Brazilfrom July 2013 to June 2017; the 3-month follow-up visits were completed bySeptember 2017. We recruited 156 patients 16 years of age or older withhypotonic polyuria (a urine output of >50 ml per kilogram of body weightduring a 24-hour period, with a urine osmolality <800 mOsm per kilogram) orwith a confirmed diagnosis of central diabetes insipidus. Three patients wereexcluded from the analyses because they were found to have nephrogenic diabetesinsipidus, and 12 patients were excluded for other reasons (Fig. S1 in theSupplementary Appendix, available with the full text of this article atNEJM.org).
这项国际性，多中心，前瞻性研究于2013年7月至2017年6月在瑞士，德国和巴西的11个三级医疗中心进行;我们在2017年9月完成了为期3个月的随访。我们招募了156名16岁或以上的低渗性多尿患者（24小时内每公斤体重尿量> 50 ml，尿液渗透压<800 mOsm / kg）或确诊为中枢性尿崩症。3名患者被排除在分析之外，因为他们被发现患有肾性尿崩症，并且有12名患者因其他原因被排除在外（补充附录中的图S1，可在NEJM.org上获得本文全文）。
The local ethics committees at all centers approved thestudy protocol (available at NEJM.org). Written informed consent was obtainedfrom all patients or from a legal guardian, when applicable. Laboratorymeasurement of copeptin was funded by Thermo Fisher Scientific, which had noother role in the study; there was no other commercial support for the study.All the authors vouch for the accuracy and completeness of the data and for thefidelity of the study to the protocol.
所有中心的当地伦理委员会都批准了研究方案（可在NEJM.org上获得）。 在适用的情况下，从所有患者或法定监护人处获得书面知情同意书。 合成肽的实验室测量由赛默飞世尔科技公司资助，该研究在该研究中没有其他作用;这项研究没有其他商业支持。 所有作者都保证数据的准确性和完整性以及研究对协议的保真度。
The water-deprivation and hypertonic saline in-fusion testswere performed on separate days. After a detailed medical history was obtained,a standardized clinical and biochemical evaluation was performed. Magnetic resonance imaging (MRI) of the headwas performed at the discretion of the attending physician, although it wasrecommended in all patients if imaging had not been performed within 3 monthsbefore study enrollment. Diuretic or antidiuretic medications were discontinuedfor at least 24 hours before each test, and smoking and alcohol were prohibitedfor at least 12 hours before each test.
在不同的日子进行除水和高渗盐水融合试验。 在获得详细的病史后，进行标准化的临床和生化评估。头部的磁共振成像（MRI）由主治医师决定，尽管如果在研究登记前3个月内未进行成像，则建议所有患者使用。 在每次测试前至少24小时停用利尿剂或抗利尿药物，并且在每次测试前禁止吸烟和饮酒至少12小时。
Test protocols
Indirect water-deprivation test
As is standard for the water-deprivation test, a 17-hourfluid restriction started at midnight, or at 6 a.m. in patients with known orsuspected complete diabetes insipidus. Every 2 hours, vital signs and bodyweight were monitored, and urine was collected for measurement of volume andosmolality. Blood samples were obtained at 8 a.m. and immediately before the administrationof desmopressin (1 hour before the end of the test).
作为禁水测试的标准方法，对于已知或疑似完全性尿崩症的患者，在午夜或早上6点开始进行17小时的液体限制。 每2小时监测生命体征和体重，收集尿液用于测量体积和重量摩尔渗透压浓度。在上午8点和即将施用去氨加压素之前（测试结束前1小时）获得血液样品。
For safety reasons, the water-deprivation test was stoppedearly in patients who met one of the following criteria: a decrease in bodyweight of more than 3%, symptoms of orthostatic hypotension with an increase inheart rate or a decrease in mean arterial blood pressure of more than 15%, oran increase in plasma sodium level of 150 mmol or more per liter. At 4 p.m., orwhen the test was stopped, each patient received 2 μg of desmopressinintravenously, and a final urine specimen for osmolality measurement wasobtained at least 60 minutes thereafter.
出于安全原因，符合下列标准之一的患者早期停止缺水试验：体重下降超过3％，体位性低血压症状随着心率增加或平均值下降 动脉血压超过15％，或血浆钠水平升高150mmol或更高。 在下午4点，或当测试停止时，每个患者静脉内接受2μg去氨加压素，并且此后至少60分钟获得用于重量摩尔渗透压浓度测量的最终尿样。
Hypertonic SalineInfusion Test
Patients underwent the hypertonic saline infusion testbetween 8 a.m. and 11 a.m., as describedpreviously. An initial 250-ml bolus infusion of 3% saline was administered, andthe infusion was continued at a rate of 0.15 ml per kilogram per minute. Bloodsamples for the measurement of plasma osmolality and sodium, urea, and glucoselevels were obtained every 30 minutes, and sodium levels were monitored byvenous blood gas analysis until the target level of at least 150 mmol per literwas reached. Thereafter, a final blood sample for plasma copeptin measurementwas obtained, and patients were given water orally (30 ml of water perkilogram) within 30 minutes, followed by a 500-ml infusion of 5% glucose within40 to 60 minutes after the patients received water. For safety reasons, the plasma sodium levelwas measured again 1 hour after the start of the glucose infusion to ensurethat the level was within the normal range before the patient was discharged.
高渗盐水输注试验
如前所述，患者在上午8点至上午11点之间进行高渗盐水输注试验。最初250ml推注输注3％盐水，并以每公斤每分钟0.15ml的速度继续输注。每30分钟获得用于测量血浆渗透压和钠，尿素和葡萄糖水平的血液样品，并通过静脉血气分析监测钠水平，直至达到至少150mmol /L的目标水平。此后，获得用于血浆和肽素测量的最终血样，并且患者在30分钟内口服水（每公斤30ml水），然后在患者后40至60分钟内输注500ml的5％葡萄糖。出于安全原因，在葡萄糖输注开始后1小时再次测量血浆钠水平，以确保在患者出院前水平在正常范围内。
Adverse Events and Symptom Burden
Adverse events during both tests were strictly documented,and clinical symptoms were rated by patients according to a visual-analoguescale that ranged from 0 to 10, with 0 indicating no symptoms and 10 indicatingthe most severe symptoms imaginable. Additionaldetails are provided in the Supplementary Appendix.
不良事件和症状负担
严格记录两个测试期间的不良事件，并且患者根据视觉模拟量表评定临床症状，范围从0到10，其中0表示没有症状，10表示可以想象的最严重的症状。 补充附录中提供了其他详细信息。
Test Interpretation and Preliminary Diagnosis
After the patients had completed both tests, they weredischarged from the hospital with a preliminary diagnosis and treatmentrecommendation that were based on best current clinical practice. A follow-upvisit was scheduled for 3 months later to assess response to treatment andclinical outcome and to reevaluate the accuracy of the preliminary diagnosis.
测试解释和初步诊断
在患者完成两项测试后，他们从医院出院，并根据最新的临床实践进行初步诊断和治疗建议。计划在3个月后进行随访，以评估对治疗和临床结果的反应，并重新评估初步诊断的准确性。
Diagnostic Criteria
Indirect Water-Deprivation Test
In accordance with the original description of the indirectwater-deprivation test and the subsequent modification, complete centraldiabetes insipidus was diagnosed in patients who had a maximum urine osmolalityof less than 300 mOsm per kilogram and an increase in urine osmolality of morethan 50% after administration of desmopressin. Partial central diabetesinsipidus was diagnosed in patients who had a maximum urine osmolality of 300to 800 mOsm per kilogram and an increase in urine osmolality of 9 to 50% afteradministration of desmopressin. Primary polydipsia was diagnosed in patientswho had a maximum urine osmolality of 300 to 800 mOsm per kilogram and an increasein urine osmolality of less than 9% after administration of desmopressin.
诊断标准
间接缺水测试
根据间接剥夺水试验和随后修改的原始描述，在最大尿渗透压低于300mOsm / kg并且尿渗透压增加超过50％的患者中诊断出完全中枢性尿崩症,给予去氨加压素后。最大尿渗透压为300至800mOsm / kg且在给予去氨加压素后，尿渗透压增加9至50％的患者中诊断出部分中枢性尿崩症。在给予去氨加压素后，最大尿渗透压为300至800mOsm / kg且尿渗透压增加小于9％的患者中诊断出原发性烦渴。
Plasma Copeptin Stimulated by Water Deprivation
Previous data suggested that the diagnosticaccuracy of the indirect water-deprivation test could be improved by theadditional measurement of baseline (at 8 a.m.) as well as stimulated (beforeadministration of desmopressin) plasma copeptin levels. According to thoseresults, prespecified cutoff levels were used: a ratio of stimulated copeptin(the change in copeptin level over 8 hours during water deprivation, inpicomoles per liter) to plasma sodium (measured at the end of the test inmillimoles per liter) of 0.02 pmol or more per liter indicated primarypolydipsia, and a basal plasma copeptin level of less than 2.6 pmol per literindicated complete central diabetes insipidus. A ratio of less than 0.02 pmolper liter indicated partial central diabetes insipidus.
水剥夺刺激血浆和肽素
先前的数据表明间接水剥夺试验的诊断准确性可以通过额外测量基线（上午8点）以及刺激（施用去氨加压素前）血浆和肽素水平来改善。根据这些结果，使用预定的截止水平：刺激的和肽素的比例（在剥夺水中8小时内和肽素水平的变化，以每升皮摩尔为单位）与血浆钠（在试验结束时测量，以毫摩尔/升为单位）每升0.02pmol或更多表明原发性烦渴，并且基础血浆和肽素水平低于2.6pmol /升表明完全中枢性尿崩症。每升小于0.02pmol的比率表明部分中枢性尿崩症。
Plasma CopeptinStimulated by Hypertonic Saline Infusion
The diagnostic criteriafor hypertonic saline– stimulated copeptin in distinguishing primary polydipsiafrom central diabetes insipidus were suggested previously by our group and wereused in this study. A plasma copeptin cutoff level of 4.9 pmol or less perliter indicated complete or partial central diabetes insipidus, and a levelgreater than 4.9 pmol per liter indicated primary polydipsia.
高渗盐水灌注刺激血浆和肽素
高血压盐水刺激的和肽素区分原发性多饮症和中枢性尿崩症的诊断标准以前由我们小组提出并用于本研究。血浆和肽素截止水平为每升4.9pmol或更低表明完全或部分中枢性尿崩症，并且每升大于4.9pmol的水平表明原发性烦渴。
Final Reference Diagnosis
In the absence of adiagnostic standard, the final reference diagnosis was determined after thestudy was completed by two independent boardcertified experts in endocrinology,who were unaware of the copeptin levels, after careful consideration of eachpatient’s medical history and clinical symptoms, the results of the water-deprivation test, the available laboratory and imaging data, and thetherapeutic response at the 3-month follow-up. In the event of discordantdiagnoses (which occurred in 4 of 144 patients), a third expert was consulted,and results were discussed until a consensus was reached.
最终参考诊断
在缺乏诊断标准的情况下，在仔细考虑每位患者的病史和临床症状后，由两名独立的内分泌专家组成的完整的参考诊断确定最终参考诊断，他们不知道和肽素水平，在3个月的随访中，水剥夺试验的结果，可用的实验室和成像数据以及治疗反应。 如果出现不一致的诊断（144例患者中有4例发生），则咨询第三位专家，讨论结果直至达成共识。
Laboratory Measurements
Blood samples wereobtained and processed for measurement of plasma copeptin and for routinelaboratory measurements (urine and plasma osmolality, hematocrit, and plasmasodium, potassium, creatinine, urea, calcium, albumin, glucose, and hemoglobinlevels). Plasma copeptin was measured centrally in one batch with the use of acommercial automated immunofluorescence assay (B.R.A.H.M.S KRYPTOR CopeptinproAVP, Thermo Scientific Biomarkers). Details are provided in theSupplementary Appendix.
实验室测量
获得血液样品并进行处理以测量血浆和肽素并用于常规实验室测量（尿液和血浆渗透压，血细胞比容和血浆钠，钾，肌酸酐，尿素，钙，白蛋白，葡萄糖和血红蛋白水平）。使用商业自动化免疫荧光测定法（B.R.A.H.M.S KRYPTOR Copeptin proAVP，ThermoScientific Biomarkers）在一批中集中测量血浆和肽素。 详情见补充附录。
Statistical Analysis
The primary end point wasthe overall diagnostic accuracy — the percentage of correctly diagnosedpatients — in the differentiation of central diabetes insipidus from primarypolydipsia. Only patients with a final diagnosis were included in the analysis;however, the three patients with nephrogenic diabetes insipidus were onlydescriptively assessed. Details concerning the fullanalysis population and theper-protocol population, as well as additional statistical details, areprovided in the Supplementary Appendix.
主要终点是中枢性尿崩症与原发性多饮症的鉴别诊断准确性 - 正确诊断患者的百分比。只有最终诊断的患者才被纳入分析;然而，仅对3例肾性尿崩症患者进行了描述性评估。有关全面分析人口和每个议定书人口的详细信息以及其他统计细节见补充附录。
The primary objectiveswere first to determine whether the measurement of copeptin during hypertonicsaline infusion and during water deprivation was superior to the indirect waterdeprivation test, and then to determine whether copeptin measurement duringhypertonic saline infusion was noninferior to copeptin measurement during waterdeprivation; the second objective would be tested only if superiority could beshown for the first objective. The primary hypothesis thus consisted of twocomponents, with a two-step statistical testing procedure. Sample size was estimated for the noninferiority test: assuming a diagnostic accuracy of 90%for water deprivation–stimulated copeptin and a noninferiority margin of 10%, atotal of 115 patients who could be evaluated would provide 90% power toestablish the noninferiority of hypertonic saline–stimulated copeptin measurement to water-deprivation–stimulated copeptin measurement. To assesswhether the diagnostic accuracy varied depending on the severity of centraldiabetes insipidus, a prespecified subgroup analysis was performed to assessthe diagnostic accuracy of hypertonic saline–stimulated copeptin andwater-deprivation–stimulated copeptin as compared with the indirectwater-deprivation test in specifically distinguishing primary polydipsia frompartial central diabetes insipidus.
主要目的是首先确定高渗盐水输注期间和水剥夺期间的和肽素的测量是否优于间接水驱除测试，然后确定高渗盐水输注期间的和肽素测量是否不劣于水剥夺期间的和肽素测量;只有在第一个目标可以表现出优势时才会测试第二个目标。因此，主要假设由两部分组成，具有两步统计测试程序。估计非劣效性测试的样本量：假设水剥夺刺激的和肽素的诊断准确度为90％，非劣效率为10％，总共115名可以评估的患者将提供90％的能力来确定高渗的非劣效性盐水刺激的和肽素测量对水剥夺刺激的和肽素测量。为了评估诊断准确性是否根据中枢性尿崩症的严重程度而变化，进行预先指定的亚组分析以评估高渗盐水刺激的和肽素和水剥夺刺激的和肽素的诊断准确性，与间接水剥夺试验相比较。特别区分原发性多饮和部分中枢性尿崩症。
Baseline Characteristics
Of the 141 patients (66%female) included in the analyses, 82 (58%) received a final diagnosis ofprimary polydipsia after all 3-month follow-ups were completed, and 59 (42%)received a diagnosis of central diabetes insipidus (Fig. S1 in theSupplementary Appendix). Among the 59 patients who received a diagnosis ofcentral diabetes insipidus, complete central diabetes insipidus was diagnosedin 36 patients (61%), and partial central diabetes insipidus in 23 (39%).There were significant differences between the groups in some baseline characteristics (Table 1).Results of MRI of the head were available for 97 patients. The hyperintensesignal in the posterior region on T1-weighted images, which is considered to bea physiological signal that indicates the pituitary arginine vasopressincontent, was absent in 70% of the patients with central diabetes insipidus, butit was also absent in 39% of the patients with primary polydipsia (Table 1).
基线特征
在分析中纳入的141名患者（66％女性）中，82名（58％）在完成所有3个月的随访后接受了原发性烦渴的最终诊断，59名（42％）接受了中枢性尿崩症的诊断（补充附录中的图S1）。在诊断为中枢性尿崩症的59名患者中，36名患者（61％）确诊为完全性中枢性尿崩症，23名患者中部分为中枢性尿崩症（39％）。
在一些基线特征中，各组之间存在显着差异（表1）。 97名患者可获得头部MRI检查结果。后部区域 T1加权像上的高信号，被认为是表明垂体精氨酸加压素含量的生理信号，在中枢性尿崩症患者中70％不存在，但也没有 39％的原发性烦渴患者（表1）。
Primary Outcome
The overall diagnosticaccuracy of the hypertonic saline infusion test was significantly higher thanthat of the indirect water-deprivation test (96.5% [95% confidence interval {CI}, 92.1 to 98.6] vs. 76.6% [95% CI,68.9 to 83.2]; P<0.001) (Table 2). The diagnostic accuracy of the hypertonicsaline infusion test was also clearly superior to that of the indirectwater-deprivation test when only patients with partial central diabetesinsipidus were compared with patients with primary polydipsia (95.2% [95% CI,89.4 to 98.1] vs. 73.3% [95% CI, 63.9 to 81.2]; adjusted P<0.001) (Table 2).Additional details on the test results are provided in Table S1 in theSupplementary Appendix.
主要成果
高渗盐水输注试验的总体诊断准确性显着高于间接缺水试验（96.5％[95％可信区间{CI}，92.1至98.6]对76.6％[95％CI，68.9至83.2] ]; P <0.001）（表2）。当只有部分中枢性尿崩症患者与原发性多饮症患者进行比较时，高渗盐水输注试验的诊断准确性也明显优于间接缺水试验（95.2％[95％CI，89.4至98.1] vs 73.3％[95％CI，63.9至81.2];调整后的P <0.001）（表2）。有关测试结果的其他详细信息，请参见补充附录中的表S1。
The copeptin levelmeasured after hypertonic saline infusion more accurately distinguished primarypolydipsia from central diabetes insipidus than the water-deprivation test withor without copeptin measurement (Fig. 1). Additional details on the course ofcopeptin levels during hypertonic saline infusion are provided in Fig. S2 inthe Supplementary Appendix. The prespecified hypertonic saline–stimulatedcopeptin cutoff level of more than 4.9 pmol per liter19 had a 93.2% sensitivity(95% CI, 83.5 to 98.1) and 100% specificity (95% CI, 95.5 to 100.0) todiscriminate between primary polydipsia and central diabetes insipidus (Table2), with a receiver-operating characteristic area under the curve for thisdiscrimination of 0.97 (95% CI, 0.93 to 1.00). The most accurate copeptincutoff level was 6.5 pmol per liter (derived post hoc), which had a diagnosticaccuracy of 97.9% (95% CI, 93.9 to 99.6), sensitivity of 94.9% (95% CI, 85.9 to98.9), and specificity of 100% (95% CI, 95.5 to 100.0) (Fig. 2 and Table 2).
高渗盐水输注后测得的和肽素水平更准确地区分原发性多饮症和中枢性尿崩症，而不是使用或不使用和肽素测定的水剥夺试验（图1）。有关高渗盐水输注期间和肽素水平过程的其他详细信息，请参见补充附录中的图S2。预先指定的高渗盐水刺激的和肽素截止水平超过每升4.9 pmol19具有93.2％的敏感性（95％CI，83.5至98.1）和100％特异性（95％CI，95.5至100.0）以区分原发性多饮和中枢尿崩症（表2），曲线下接受者操作特征区域为0.97（95％CI，0.93至1.00）。最准确的和肽素截止水平为每升6.5 pmol（事后推断），其诊断准确度为97.9％（95％CI，93.9至99.6），灵敏度为94.9％（95％CI，85.9至98.9），以及特异性为100％（95％CI，95.5至100.0）（图2和表2）。
The overall diagnostic accuracy of water deprivation–stimulated copeptin(with use of the prespecified ratio of stimulated copeptin to plasma sodiumdescribed above) in distinguishing primary polydipsia from central diabetesinsipidus5 was significantly lower than that of the indirect water-deprivationtest (44.0% [95% CI, 35.7 to 52.5] vs. 76.6% [95% CI, 68.9 to 83.2]) (Table 2).When the prespecified morning copeptin cutoff level of less than 2.6 pmol perliter after overnight water deprivation was used to identify patients withcomplete central diabetes insipidus, the diagnostic accuracy was 78.4% (95% CI,70.6 to 84.9). Plasma copeptin values after overnight water deprivation and theassociated receiver-operating-characteristic area under the curve are providedin Figure S3 in the Supplementary Appendix.
水分剥夺刺激的和肽素的总体诊断准确性（使用上述刺激的和肽素与血浆钠的预定比例）区分原发性多饮症与中枢性尿崩症5的差异显着低于间接缺水试验（44.0％[95] ％CI，35.7至52.5]对76.6％[95％CI，68.9至83.2]）（表2）。当使用过夜水剥夺后预先指定的早上和肽素截止水平低于2.6pmol /升时，确定患有完全中枢性尿崩症的患者，诊断准确性为78.4％（95％CI，70.6-84.9）。过夜除水后的血浆和肽素值以及曲线下相关的接收器操作特征区域在补充附录中的图S3中提供。
图1.低渗性多尿症患者对高渗盐水输注和缺水试验的刺激和肽素水平。
显示刺激的和肽素水平响应高渗盐水输注测试和缺水多尿症患者的水剥夺试验，这是由中枢性尿崩症引起的与原发性烦渴相比（图A）和由完全引起的低渗性多尿症患者 与原发性烦渴相比，中枢性尿崩症或部分中枢性尿崩症（图B）。每个框中的水平线表示中间，框的下边界和上边界，四分位数范围，晶须的末端排列在四分位数范围的1.5倍内的最小值和最大值，以及点异常值。DI表示尿崩症。
Secondary Outcomes and Burden of Tests
Patients rated theoverall burden of the water deprivation test higher than that of the hypertonicsaline infusion test (median score on the visual-analogue scale, 6[interquartile range, 4 to 7] vs. 5 [interquartile range, 3 to 6]) and theoverall tolerability (i.e., convenience of the test and patients’ comfort levelduring the test) lower (38% of patients preferred the water-deprivation test,whereas 62% preferred the hypertonic saline infusion test) (Table 3). Theplasma sodium level increased to more than 155 mmol per liter in 12 patients duringhypertonic saline infusion (in 6 patients with a final diagnosis of primarypolydipsia, 5 with complete central diabetes insipidus, and 1 with partialcentral diabetes insipidus), as compared with 2 patients during waterdeprivation (both patients had complete central diabetes insipidus). All 12patients were female and had baseline plasma sodium levels of 140 to 144 mmolper liter. Additional information on the course of plasma sodium level duringhypertonic saline infusion is provided in Figure S4 in the SupplementaryAppendix.Nine adverse eventsoccurred during hypertonic saline infusion, and seven during water deprivation.One serious adverse event was reported: desmopressin-induced hyponatremia afterthe water-deprivation test, which resulted in hospitalization of the patient.
次要结果和测试负担
患者认为水剥夺试验的总负担高于高渗盐水输注试验（视觉模拟评分中位数，6 [四分位数间距4至7]对5 [四分位数范围，3至6]）和总体耐受性（即，试验的方便性和患者在试验期间的舒适水平）较低（38％的患者优选缺水试验，而62％的患者优选高渗盐水输注试验）（表3）。在高渗盐水输注期间，12名患者的血浆钠水平增加至超过155 mmol / L（最终诊断为原发性烦渴的6名患者，5名患有完全性中枢性尿崩症，1名患有部分中枢性尿崩症），与2例患者在缺水期间（两名患者均患有完全性中枢性尿崩症）。所有12名患者均为女性，基线血浆钠水平为每升140至144 mmol。有关高渗盐水输注过程中血浆钠水平过程的更多信息，请参见补充附录中的图S4。
在高渗盐水输注期间发生了9个不良事件，在缺水期间发生了7个不良事件。报告了一个严重不良事件：去水加压试验后去氨加压素诱导的低钠血症导致患者住院治疗.
Discussion
This prospective, multicenterstudy showed that measurement of hypertonic saline–stimulated copeptin wassuperior to the indirect water-deprivation test in distinguishing polyuria dueto primary polydipsia from polyuria due to central diabetes insipidus. However,the postulated superiority of water-deprivation–stimulated copeptin to theindirect water-deprivation test could not be confirmed in this study. Thediagnostic accuracy of the indirect water-deprivation test of approximately 70%in our study, which is consistent with previous findings in smaller studies,resulted in approximately 30% of patients with primary polydipsia incorrectlyreceiving a diagnosis of central diabetes insipidus. All the patients in thecurrent study whose diagnosis was misclassified according to the indirect waterdeprivation test received the correct diagnosis according to the results of thehypertonic saline infusion test with the prespecified copeptin cutoff of 4.9pmol per liter.
这项前瞻性多中心研究表明，高血压盐水刺激的和肽素的测量优于间接水剥夺试验，以区分由于中枢性尿崩症引起的多尿引起的多尿症。然而，在本研究中无法证实水剥夺刺激的和肽素对间接缺水试验的假定优势。在我们的研究中，间接缺水试验的诊断准确率约为70％，这与先前在较小研究中的发现一致，导致大约30％的原发性烦渴患者错误地接受了中枢性尿崩症的诊断。根据高渗盐水输注试验的结果，当前研究中根据间接水剥夺试验将诊断错误分类的所有患者接受了正确的诊断，其中预定的和肽素截留值为每升4.9pmol。
As reported previously,indirect measures of renal arginine vasopressin activity do not accuratelydiscriminate primary polydipsia from central diabetes insipidus, which haschallenged the use of the indirect water-deprivation test as the diagnosticstandard. An essential limitation of urinary measures is the variably reduced maximalurinary concentration capacity inherent in all forms of chronic polyuria.Moreover, enhanced renal sensitivity to even low levels of circulating argininevasopressin in patients with central diabetes insipidus may complicate theinterpretation of indirect tests.
如先前所报道的，肾精氨酸加压素活性的间接测量不能准确地区分原发性多饮症与中枢性尿崩症，这对使用间接缺水试验作为诊断标准提出了挑战。尿液测量的一个基本限制是所有形式的慢性多尿症中固有的最大尿液浓度能力的可变降低。此外，对中枢性尿崩症患者甚至低水平的循环精氨酸加压素的肾脏敏感性增强可能使间接测试的解释复杂化。
Our data confirm thatadditional measurements such as the basal plasma sodium level or theurine-to-plasma osmolality ratio after fluid restriction are of limiteddiagnostic value. In addition, in our study, the measurement of water deprivation–stimulatedcopeptin levels did not improve diagnostic discrimination (73% of patients didnot achieve hyperosmolality after 16 hours of fluid deprivation). Finally,although some clinical criteria (e.g., the presence of certain diseases,additional clinical presentations, and findings on MRI of the head) aresometimes recommended to help in making specific diagnoses in patients withpolyuria, evidence to support their diagnostic value is insufficient and wasnot supported by our results.
我们的数据证实，额外的测量如流体限制后的基础血浆钠水平或尿液与血浆渗透压比率的诊断价值有限。此外，在我们的研究中，水剥夺刺激的和肽素水平的测量没有改善诊断辨别力（73％的患者在16小时的液体剥夺后没有达到高渗性）。最后，尽管有时建议某些临床标准（例如某些疾病的存在，其他临床表现和头部MRI检查结果）有助于对多尿患者进行特异性诊断，但支持其诊断价值的证据不足，我们的结果不支持。
Consequently, a testmethod that provides high diagnostic specificity, particularly for the criticaldistinction of primary polydipsia from central diabetes insipidus, is needed. Afterthe early report on hypertonic saline administration by Hickey and Hare in1944, in which indirect measures of renal function were used to detect therelease of arginine vasopressin, Zerbe and Robertson further developed themethod by introducing direct measurement of plasma arginine vasopressin todifferentiate primary polydipsia from diabetes insipidus. Given the technicalconstraints of arginine vasopressin quantitation and the highly variablefunctional sensitivity and specificity of the few assays approved for clinicaluse, we designed our study to determine whether the measurement of copeptin ismore reliable and easier to process and whether it can be standardized as atest that would detect the osmotic arginine vasopressin reserve.
因此，需要一种提供高诊断特异性的测试方法，特别是对于中枢性尿崩症的原发性多饮症的关键区分。在1944年Hickey和Hare对高渗盐水给药的早期报告（其中使用肾功能的间接测量来检测精氨酸加压素的释放）后，Zerbe和Robertson通过引入血浆精氨酸加压素的直接测量来进一步发展该方法以区分原发性来自尿崩症的多饮症。鉴于精氨酸加压素定量的技术限制以及批准用于临床的少量检测的高度可变的功能灵敏度和特异性，我们设计了我们的研究以确定和肽素的测量是否更可靠和更容易处理以及是否可以标准化为检测渗透性精氨酸加压素储备的试验。
The copeptin assay isdesigned to overcome the technical and functional caveats inherent in thearginine vasopressin assay, and it appears to have the diagnostic potential notonly to identify nephrogenic diabetes insipidus, but also to distinguish centraldiabetes insipidus from primary polydipsia. Building on our previous work, inwhich hypertonic saline infusion was initiated after fluid deprivation, thecurrent prospective validation study used a simplified protocol: the teststarted with a saline bolus, which was followed by an infusion (at a rateaccording to each patient’s body weight), thereby providing a more potent andprompter osmotic stimulus. This modified protocol attained a better outcomethat validated the prespecified copeptin cutoff of 4.9 pmol per liter and,excluding post hoc analysis, yielded the highest diagnostic accuracy for theentire population (96.5%), as well as for the critical distinction between mildforms of arginine vasopressin deficiency and primary polydipsia (95.2%).
和肽素测定旨在克服精氨酸加压素测定中固有的技术和功能警告，并且它似乎具有诊断潜力，不仅可以识别肾性尿崩症，还可以区分中枢性尿崩症和原发性烦渴。基于我们之前的工作，其中在液体剥夺后开始高渗盐水输注，目前的前瞻性验证研究使用简化的方案：测试以盐水推注开始，然后输注（按照每个患者身体的速率）重量），从而提供更有效和促进渗透刺激。这一修改后的方案获得了更好的结果，证实了预先指定的和肽素截止值为每升4.9 pmol，并且不包括事后分析，对整个人群产生了最高的诊断准确率（96.5％），以及温和形式之间的关键区别精氨酸加压素缺乏和原发性烦渴（95.2％）。
We note caveats withrespect to the use of hypertonic saline infusion in the clinical evaluation ofpatients with polyuria. More adverse effects were reported with the hypertonicsaline infusion test than with the water-deprivation test. The hypertonicsaline infusion test required close monitoring of sodium levels to ascertain adiagnostically meaningful increase in plasma sodium within the hyperosmoticrange while preventing a marked increase, to which female patients appearedmore vulnerable than male patients in this study.
我们注意到在多尿症患者的临床评估中使用高渗盐水输注的注意事项。高渗盐水输注试验报告的不良反应比禁水试验更多。 高渗盐水输注测试需要密切监测钠水平以确定在高渗范围内血浆钠的诊断上有意义的增加，同时防止显着增加，在该研究中女性患者似乎比男性患者更容易受到伤害。
Our study has limitationsand strengths. One limitation is that there is no diagnostic standard forhypotonic polyuria. Here, we constructed criteria for reference diagnoses thatwere based on the full set of clinical data, the results of the indirectwater-deprivation test, and the response of each patient to individual therapyat a 3-month follow-up visit, in accordance with clinical practice. Thesimultaneous evaluation of the diagnostic accuracy of the indirectwater-deprivation test and the use of those results in final decision makingmay have resulted in an incorporation bias. However, if this bias happened atall, it may have resulted in an overestimation of the diagnostic performance ofthe water-deprivation test. The strengths of the study involve theinternational multicenter design, the prospective validation of prespecifiedcutoff levels for hypertonic saline– stimulated copeptin release, and arelatively large sample size of patients with diabetes insipidus and primarypolydipsia.
我们的研究有局限性和优势。一个限制是没有低渗多尿的诊断标准。在这里，我们构建了基于全套临床数据的参考诊断标准，间接缺水测试的结果，以及每个患者在3个月随访时对个体治疗的反应，与临床实践。同时评估间接缺水试验的诊断准确性以及在最终决策中使用这些结果可能导致纳入偏差。但是，如果发生这种偏差，可能会导致对剥夺水分测试的诊断性能的过高估计。该研究的优势包括国际多中心设计，高渗盐水刺激的和肽素释放的预定截止水平的前瞻性验证，以及相对较大的尿崩症和原发性烦渴患者的样本量。
In conclusion, thisprospective evaluation of patients with hypotonic polyuria validated hypertonicsaline–stimulated copeptin measurement as a diagnostic method that appeared tobe superior to the indirect water-deprivation test in distinguishing centraldiabetes insipidus from primary polydipsia.
总之，这项对低渗性多尿患者的前瞻性评估验证了高渗盐水刺激的和肽素测定作为一种诊断方法，在区分中枢性尿崩症和原发性多饮症方面似乎优于间接水剥夺试验。