Lithium-induced nephrogenic diabetes insipidus: in vivo and in vitro studies

Abstract
The physiological basis for the polyuria and polydipsia occurring in some manic-depressive patients treated with lithium salts was studied in vivo and in vitro. Three lithium-treated polyuric patients, in whom other causes of a concentrating defect were excluded, had abnormal urinary concentrating abilities after a standard water depreviation test. Two of these patients failed to respond to exogenous vasopressin (ADH) and one had a subnormal response. The abilities of these patients to excrete solute-free water (CH2O) was comparable to normal subjects during steady-state water diuresis, suggesting no gross abnormalities in sodium transport. However, each of these patients demonstrated abnormally low capacities to reabsorb solute-free water (TCH2O) under hydropenic conditions after administration of hypertonic saline and vasopressin. These in vivo findings demonstrate at least a nephrogenic basis for the diabetes insipidus syndrome manifested by these three patients. The defect in water transport was further characterized in toad urinary bladders in vitro. Short-circuit current (I) and water flow (W) were studied under basal, ADH-stimulated, and cyclic adenosine 3′,5′-monophosphate (c-AMP)-stimulated conditions. Increasing mucosal [Li+] progressively inhibited basal I, and both I and W induced by ADH. Significant inhibition of basal and ADH-induced I was observed at mucosal [Li+] < 1.1 mEq/liter, and of ADH-induced W at mucosal [Li+] = 11 mEq/liter. On the other hand, at these lithium concentrations, neither c-AMP-stimulated W nor I was inhibited. Increasing serosal [Li+] produced significant inhibition of basal I only at [Li+] at least 50-fold greater than at the mucosal (urinary) surface. These in vitro studies confirm that mucosal lithium inhibits the action of ADH, but not c-AMP. Hence, lithium appears to be a significant inhibitor of ADH-stimulated water flow, probably acts from the urinary surface, and appears to exert its effect at a site biochemically proximal to c-AMP action.