Diagnosis of central diabetes insipidus using a vasopressin radioimmunoassay during hypertonic saline infusion.

Central diabetes insipidus (CDI) is characterized by polyuria and polydipsia caused by impairment of arginine vasopressin (AVP) secretion. In this study, we evaluated plasma AVP concentrations during a hypertonic saline infusion test using a new AVP radioimmunoassay (RIA) which is now available in Japan. Thirteen control subjects, mostly with hypothalamo-pituitary disease but without CDI, and 13 patients with CDI were enrolled in the study. Whether or not subjects had CDI was determined based on the totality of clinical data, which included urine volumes and osmolality. Regression analysis of plasma AVP and serum Na concentrations revealed that the gradient was significantly lower in the CDI group than in the control group. The area under the receiver-operating-characteristic (ROC) curve was 0.99, and the <0.1 gradient cut-off values for the simple regression line to distinguish CDI from control had a 100% sensitivity and a 77% specificity. The ROC analysis with estimated plasma AVP concentrations at a serum Na concentration of 149 mEq/L showed that the area under the ROC curve was 1.0 and the <1.0 pg/mL cut-off values of plasma AVP had a 99% sensitivity and a 95% specificity. We conclude that measurement of AVP by RIA during a hypertonic saline infusion test can differentiate patients with CDI from those without CDI with a high degree of accuracy. Further investigation is required to confirm whether the cut-off values shown in this study are also applicable to a diagnosis of partial CDI or a differential diagnosis between CDI and primary polydipsia.