albumin. Albumin induced centralisation of the circulation results in increased diuresis and sodium excretion and in normalisation of the activated vasoconstrictor and sodium retaining

(PRA) and plasma concentration of arginin vasopressin (AVP) in this group. No responses to albumin infusion were noted in six of 15 patients (group B). These patients had high ANF and low PRA, suggesting a relative volume overload. It is well known that patients with liver cirrhosis do not represent a homogeneous group.5 There are at least two mechanisms leading to sodium and water retention and thus ascites.' Patients with decreased effective intravascular volume have an activated sympathicoadrenal and renin-angiotensin-aldosterone axis, raised vasopressin level, and decreased plasma concentration of ANF (underfilling theory). In contrast, volume overloaded patients have decreased vasoconstrictor and increased vasodilator hormone activity (overflow theory); the renal sodium retention in these cases represents a primary event due probably to intrarenal microcirculatory changes. The patients presented here show characteristics ofboth causes of liver cirrhosis. It is of interest, however, that only patients with signs of underfilled effective intravascular volume provided an adequate response to intravenous albumin. Albumin induced centralisation of the circulation results in increased diuresis and sodium excretion and in normalisation of the activated vasoconstrictor and sodium retaining hormone systems. These data may give a better insight into the pathogenesis of sodium and water retention and also into the mechanism of the effect of albumin infusion in liver disease.