Review article: albumin in the treatment of liver diseases—new features of a classical treatment


Correspondence to: Dr V. Arroyo, Liver Unit, Hospital Clinic, Villarroel 170, O8036 Barcelona, Spain.


Albumin was introduced initially in the treatment of patients with cirrhosis and ascites to increase serum albumin concentration due to its oncotic effect. Although its administration declined some years later, at present it constitutes an essential treatment in clinical hepatology. Several studies have clearly demonstrated its efficacy in the prevention and treatment of circulatory dysfunction and hepatorenal syndrome in patients with cirrhosis. These effects can be due not only to its properties as a plasma expander but also to its capacity to bind numerous substances such as bile acids, nitric oxide and cytokines. Based on this capacity an albumin dialysis system (MARS) has recently been developed. The usefulness of this system in the management of patients with acute and chronic liver failure is, at present, under evaluation.


The traditional theory of ascites formation in cirrhosis was based on the classic experiments carried out by Starling in animals with hepatic vein ligation.1 It considered portal hypertension and hypoalbuminaemia as the main factors in the pathogenesis of this complication. According to this theory, the extravasation of fluid from the intravascular compartment to the interstitial space and to the peritoneal cavity would be a consequence of a rupture of the Starling equilibrium within the hepatic and splanchnic microcirculation.2 Albumin was introduced as a treatment in the 1950s and it was rapidly incorporated into the management of patients with decompensated cirrhosis. Since there was no medical treatment of portal hypertension, the only way to improve the haemodynamic imbalance in the hepatic and splanchnic microcirculation was to increase the serum albumin concentration and, consequently, the oncotic pressure. It is therefore not surprising that intravenous albumin solutions were widely used for the management of cirrhotic patients with hypoalbuminaemia and ascites during the 1950s and 1960s.3

The initial studies showing that hyperaldosteronism and renal sodium and water retention were the mechanisms of continuous formation of ascites was a further argument for the use of albumin in patients with decompensated cirrhosis. At that time, activation of the renin-aldosterone system in cirrhosis was thought to be due to an effective hypovolaemia. Although cirrhotic patients with ascites showed an increased plasma volume, most of this volume would be refilling the dilated splanchnic venous bed (splanchnic pooling); the effective blood volume (which is that sensed by the volume receptors, baroreceptors and the kidney) would actually be reduced (classic underfilling theory).4 The administration of albumin could therefore correct not only the hypoalbuminaemia but also the low effective circulating blood volume.

The demonstration by several studies that not only the plasma volume but also the cardiac output was greatly increased in patients with cirrhosis and ascites, and therefore that the hypervolaemia is not sequestered in any vascular territory—but is actually circulating through the intrathoracic venous system and the arterial system—was a demolishing argument against this traditional theory.5, 6 In view of these findings, in 1970 Lieberman et al. proposed an alternative hypothesis of renal dysfunction and ascites formation in cirrhosis. They suggested that advanced cirrhosis triggered a sodium-retaining signal in the renal tubules. The subsequent renal retention of sodium and water would result in an expansion of plasma volume and adaptative circulatory changes (high cardiac output and low systemic vascular resistance) to accommodate the excess intravascular volume. The interaction between portal hypertension and circulating hypervolaemia would lead to ‘overflow’ ascites formation (overflow theory).7 The decreased hepatic synthesis of a natriuretic substance, or the existence of a hepato-renal reflex leading to primary sodium retention (primary in the sense that it was unrelated to the disturbances in effective blood volume) were later proposed as a link between the diseased liver and the kidneys.8, 9

The overflow theory of ascites, which suggested that cirrhotic patients have hypervolaemia rather than hypovolaemia, a better understanding of the rational use of diuretics in the management of cirrhotic ascites, and the introduction of peritoneo-venous shunting10 for the management of the very few patients with ascites refractory to medical treatment, led to a decline in the use of albumin in the management of patients with cirrhosis and ascites.

The overflow theory did not satisfy many investigators in the field of ascites because it did not offer a rational explanation of the main clinical features of cirrhotic patients with ascites, namely, that these patients have a low arterial pressure despite an increased plasma volume and cardiac index and marked activation of the sympathetic nervous system and renin-angiotensin system.11 The plasma levels of renin in many patients with cirrhosis and ascites are similar to those seen in patients with malignant hypertension. Despite this observation, cirrhotic patients with ascites show a decreased arterial pressure.

In 1988 a new theory to explain the pathogenesis of renal dysfunction was proposed by Schrier et al.12 This theory reconsidered that sodium and water retention and ascites formation were secondary to the circulatory abnormalities. However, in contrast to the classic underfilling theory, this new hypothesis postulated that the circulatory abnormality causing renal dysfunction was in the arterial vascular compartment. There would be arterial vascular underfilling, not as the result of a reduction in arterial volume, but rather as a result of a disproportionate enlargement of the arterial tree secondary to arterial vasodilation (peripheral arterial vasodilation hypothesis). According to this theory, portal hypertension was the initial event, which resulted in splanchnic arterial vasodilation. This leads to high cardiac output and heart rate and reduced peripheral vascular resistance and arterial pressure. The central blood volume, that is the volume of blood contained in the heart, pulmonary circulation and aorta before the renal arteries are reduced.13 The right atrial pressure and pulmonary pressures are normal or reduced. Finally, the circulatory transient time from the right atria to the aorta is shortened.13 These features define what is called a hyperdynamic circulation, which is an increased blood volume circulating very rapidly within the central vascular compartment as a consequence of a high cardiac output. This high cardiac output arises because of a low cardiac preload and increased heart rate. Although the flow of blood throughout the central vascular compartment is increased, the central blood volume is decreased, and when the reduction in central blood volume is sufficient to stimulate the volume receptors (low-pressure receptors) in the right atria and pulmonary circulation and the pressure receptors in the aorta and carotid sinus, there is a reflex stimulation of the sympathetic nervous system and the renin-angiotensin system and vasopressin release which leads to continuous renal sodium and water retention and ascites formation.

The splanchnic arterial vasodilation also plays an important role in the increased splanchnic production of lymph and therefore in the leakage of fluid from the intravascular compartment to the peritoneal cavity.14 The increased arterial blood inflow into the splanchnic microcirculation that follows arterial vasodilation is a major factor in the increased hydrostatic pressure in the splanchnic capillaries and is associated with an increased vascular permeability. Therefore, renal dysfunction and ascites formation in cirrhosis are mainly related to events occurring in the arterial vascular compartment.

The peripheral arterial vasodilation hypothesis stimulated research on circulatory function in cirrhosis, and this has been important in the design of new treatments for these patients. Several studies have shown that the hyperdynamic circulation is solely due to a vasodilation in the splanchnic circulation, related to a local release of vasodilators such as nitric oxide.15–18 In fact there is a compensatory vasoconstriction in the remaining major vascular organs such as the kidneys, brain and muscle, and skin.15–17 Hepatorenal syndrome in this theory is considered to be the extreme expression of this impairment in circulatory function, secondary to splanchnic arterial vasodilation. However, studies have also suggested that a reduction in cardiac output secondary to an impaired venous return could be a contributory mechanism in hepatorenal syndrome. Hepatorenal syndrome would develop as a result of a combination of splanchnic arterial vasodilation and decreased cardiac output.19

Therapeutic paracentesis has been the main therapeutic improvement in patients with cirrhosis and ascites over the last few years.20–22 Interestingly enough, the mobilization of ascites by paracentesis is associated with an aggravation of the arterial vasodilation present in these patients, an intense stimulation of the endogenous vasoactive systems, and an increased probability of deterioration in renal function in a significant number patients.23–25 By an unknown mechanism, this paracentesis-induced circulatory dysfunction is not spontaneously reversible and significantly reduces the probability of survival.25 The most interesting feature is that this complication of paracentesis can be almost totally prevented if the procedure is performed in association with an expansion of plasma volume using albumin.25

Type-1 hepatorenal syndrome is the most severe renal complication in patients with decompensated cirrhosis.26 It is a rapidly progressive functional renal failure that occurs in association with a deterioration of circulatory function. Spontaneous bacterial peritonitis (SBP) is frequently associated with type-1 hepatorenal syndrome.27–29 It has been proposed that cytokines released in the abdominal cavity by the infection activates nitric oxide and impairs circulatory function.29 A decrease in cardiac output has also been observed in these patients, suggesting that the rapid deterioration in renal function is due to both arterial vasodilation and reduced cardiac function due to a decreased venous return.19 The most interesting observation is that the incidence of type-1 hepatorenal syndrome in patients with SBP and its associated hospital mortality can be reduced by more than 60% if plasma volume is energically expanded with albumin at the time of diagnosis.30

Since hepatorenal syndrome is probably the result of a combination of an arterial vasodilation and a decrease in cardiac output due to a reduction in venous return,19 the most rational therapy of this complication is the combination of plasma volume expansion and vasoconstrictors of preferential effect in the splanchnic circulation. Several studies have demonstrated that this hypothesis is correct.31–36 Long-term (at least 1–2 weeks) administration of albumin, together with arterial vasoconstrictors (terlipressin, amidodrin or noradrenalin) reverses hepatorenal syndrome in most patients. This does not occur in patients treated with vasoconstrictors alone,37 indicating the importance of volume expansion with albumin in the treatment of hepatorenal syndrome. Interestingly enough, hepatorenal syndrome does not recur following the withdrawal of treatment, and a significant number of patients have an increased survival and may be suitable for liver transplantation.

In addition to its effect as plasma expander, albumin has a great capacity to bind numerous substances, including bile acids, nitric oxide and cytokines.38, 39 Therefore, its biological effects may also be related to this effect. The recent development of the Molecular Absorbents Recirculating System (MARS), an albumin dialysis that removes albumin-bound substances as well as water soluble substances in patients with acute and chronic liver failure, is a clinical application for albumin, based on its capacity to remove water-unsoluble substances.40 The MARS system has been shown to be very effective in the treatment of hepatic hepatic encephalopathy41 and intractable pruritus (A. Parés, unpublished observations). Furthermore, it markedly improves circulatory and renal function in patients with cirrhosis and ascites.42,43 Several randomized controlled trials are currently being performed to assess the use of MARS in acute and chronic liver failure.

The use of albumin in patients with liver diseases has therefore followed a seesaw evolution. Initially it was widely used to increase serum albumin concentration and for the treatment of ascites. Subsequently, albumin was rarely prescribed in patients with liver diseases. At this moment, with several randomized trials and pilot studies indicating that albumin is extremely effective in the prevention and treatment of circulatory dysfunction and hepatorenal syndrome in patients with cirrhosis, this molecule is again becoming an essential treatment in clinical hepatology. Finally, the MARS system, which has opened the haemodyalisis world to patients with acute and chronic liver failure, will increase our knowledge on the mechanisms of action of albumin and will probably expand the therapeutic indications of this molecule. Without any doubt, albumin will be an exciting topic for research in the near future, as it was many years ago.