The mechanism of urinary concentration in nephrogenic diabetes insipidus

Abstract

The authors have evaluated urinary adenosine 3',5'-monophosphate (cyclic AMP) excretion and renal function during Pitressin administration, hypertonic saline administration, and water deprivation in two siblings with vasopressin-resistant diabetes insipidus and in normal control subjects. After vasopressin administration normal subjects experienced a 2-fold rise in urinary cyclic AMP excretion from 3.2 +/- 0.7 to 5.6 +/- 1.3 nmol/min (P less than 0.001) whereas cyclic AMP excretion was unchanged in both patients (patient AC 4.4 +/- 0.9 to 4.3 +/- 2.1; patient TC 2.2 +/- 0.9 to 2.6 +/- 0.9 nmol/min) with nephrogenic diabetes insipidus (NDI). Urinary cyclic AMP excretion was measured during infusion of 2.5% saline, after vasopressim administration, and after water deprivation. Cyclic AMP excretion was not different from control values in the NDI patients during any of the experimental conditions. Furthermore, there was no difference in cyclic AMP excretion when periods of dilute urine excretion (patient AC 4.5 +/- 1.1; patient TC 2.1 +/- 0.8 nmol/min) were compared with periods when urine concentration was greater than that of plasma (AC 3.5 +/- 1.3; TC 1.8 +/- 0.9 nmol/min). Both subjects responded to parathyroid hormone infusion with a 2-fold increase in urinary cyclic AMP excretion. Excretion of concentrated urine was paralleled by a marked decrease in urine flow to less than 1 ml/min/m2. During periods of hypotonic urine excretion (Uosm/Posm less than 1.0) average glomerular filtration rate (GFR) in patient AC was 67.0 +/- 3.0 ml/minm2 whereas in patient TC it was 70.1 +/- 8.1 ml/min/m2. When each patient was excreting a hypertonic urine (Uosm/Posm greater than 1.0) after fluid deprivation their GFR had decreased significantly (P = 0.001) to 31.6 +/- 8.9 and 33.3 +/- 10.3 ml/min/m2, respectively. Ability of these two subjects with NDI to concentrate their urine to Uosm/Posm greater than 1.0 in the absence of an increase in urinary cyclic AMP but associated with a decrease in GFR to 50% normal indicates that urinary concentration was effected by a reduction in GFR rather than a partial response to antidiuretic hormone (ADH). Their ability to concentrate their urine during periods of modest volume depletion would protect them from progressing to more severe stages of dehydration and result in the relatively benign course of their disease. It is feasible that in patients previously reported to have had clinically "partial" NDI this mechanism may have been operative.