Radioimmunoassay of arginine-vasopressin in human urine and its use in physiological and pathological states.

A highly specific radioimmunoassay for arginine-vasopressin (AVP) in human urine has been developed, with a detection limit of 2.2 fmol/ml. The mean recovery of added AVP was 99.5 +/- 3.1 (S.D.)% when correction was made for the fact that an inverse relationship was observed between the recovery of AVP and the osmolality of the urine. The intra- and interassay coefficients of variation were 3.5--7 and 2.5--10% respectively. Arginine-vasopressin remains stable in urine after repeated freezing and thawing after storage at 4 or 20 degrees C for up to 7 days and at -20 degrees C for more than 3 months. During unrestricted fluid intake in normal people, the mean rate of renal excretion of AVP was 95 +/- 68 (S.D.) fmol/min. An isosmotic reduction of 9% in the plasma volume increased the excretion of AVP to 259 +/- 147 (S.D.) fmol/min. At the height of water-induced diuresis the rate of excretion fell to 70 +/- 28 (S.D.) fmol/min. Fluid deprivation for 18 h produced a moderate but significant increase in mean excretion of AVP, to a value of 116 +/- 67 (S.D.) fmol/min. Patients with compulsive water drinking showed a normal relationship between urine osmolality and the rate of excretion of AVP. In pituitary diabetes insipidus, AVP was undetectable, whereas in hereditary nephrogenic diabetes insipidus a progressive increase in the rate of excretion of AVP was observed in response to dehydration. There was a wide variation in the rate of excretion of AVP (range 126--8704 fmol/min) in patients with unexplained hyponatraemia, presumed to be due to an inappropriate secretion of antidiuretic hormone. Despite this variation, the relationship between urine osmolality and the rate of excretion of AVP clearly differed from that observed in normal people.