Human albumin in the management of complications of liver cirrhosis

Human serum albumin is the most abundant plasma protein, representing about 50% of the total protein content (3.5–5 g/l). Albumin is a protein of 585 amino acids and molecular weight 66 kDa encoded by a gene on chromosome 4 and is exclusively synthesized by liver cells, which release it directly into the blood stream without storage. Under physiological conditions, only 20–30% of hepatocytes are committed to the production of 9–12 g of albumin per day; therefore, the liver has a large functional reserve, so that it can increase the synthesis of this protein by 3–4 times, if necessary. The production of albumin is mainly regulated by the osmolarity and oncotic pressure of interstitial fluid in the liver extravascular space, but it is also induced by hormonal factors (insulin, cortisol and growth hormone) and inhibited by acute phase cytokines, such as interleukin (IL)-6 and tumor necrosis factor (TNF)-α [1].


Introduction
Human serum albumin is the most abundant plasma protein, representing about 50% of the total protein content (3.5-5 g/l). Albumin is a protein of 585 amino acids and molecular weight 66 kDa encoded by a gene on chromosome 4 and is exclusively synthesized by liver cells, which release it directly into the blood stream without storage. Under physiological conditions, only 20-30% of hepatocytes are committed to the production of 9-12 g of albumin per day; therefore, the liver has a large functional reserve, so that it can increase the synthesis of this protein by 3-4 times, if necessary. Th e production of albumin is mainly regulated by the osmolarity and oncotic pressure of interstitial fl uid in the liver extravascular space, but it is also induced by hormonal factors (insulin, cortisol and growth hormone) and inhibited by acute phase cytokines, such as interleukin (IL)-6 and tumor necrosis factor (TNF)-α [1].
Because serum albumin scarcely crosses the majority of capillaries, it remains in the blood stream and generates about 70% of the plasma oncotic pressure. Th is feature is two-thirds because of a direct osmotic eff ect and one third because of the Gibbs-Donnan eff ect, as a result of the strong negative charges attracting positively charged molecules into the intravascular compartment. Albumin is, therefore, a main modulator of fl uid distribution throughout body compartments [1].
Th e capacity of albumin to expand the plasma volume represents the pathophysiological background for the use of human albumin in many clinical conditions; nevertheless, its administration is often inappropriate, largely because of a common belief in its effi cacy, whereas many indications are still under debate or have been disproved by evidence-based medicine. Indeed, the high cost, the theoretical risk of viral disease transmission, and the availability of cheaper alternatives should be carefully weighed when considering prescription of albumin [2]. At present, it is generally accepted that the administration of non-protein colloids and crystalloids represents the fi rst-line treatment of resuscitation, and use of albumin in critically ill patients should be reserved for specifi c conditions, such as in patients with septic shock [3,4]. Albumin administration is not recommended to correct hypoalbuminemia per se (i.e., not associated with hypovolemia) or for nutritional intervention [2,5], but these indications are often disregarded in clinical practice. Albumin is also prescribed in certain specifi c conditions and diseases, such as kernicterus, plasmapheresis, and graft versus host disease [2], even though these indications are not supported by defi nite evidence.
Although the clinical use of albumin is mainly related to plasma volume expansion, albumin is more than a volume expander, having further biological properties: It binds and transports a variety of water-insoluble molecules, metals, and drugs, with implications for delivery and effi cacy of drugs, including antibiotics, and for detoxifi cation of endogenous and exogenous sub stan ces [5]. Furthermore, it constitutes the main circulating antioxidant in the body, being the major extracellular source of reduced sulfhydryl groups, potent scavengers of reactive oxygen species (ROS) (Figure 1) [6].
Among the 35 cysteine residues of albumin, 34 are employed for intramolecular disulfi de bonds, whereas cysteine 34 (Cys34) remains free. ROS react primarily with Cys34, leading to reversible, highly unstable derivatives: Th e sulfenic and sulfi nic acids. Th e subsequent, fi nal product of the reaction is the sulfonic acid-albumin complex. Oxidation alters the biological properties of albumin, which becomes more susceptible to proteinase digestion, undergoes faster degradation than the nonoxidized counterpart, and has reduced binding capacity to various substances, including bilirubin [6,7]. In healthy adults, about 70-80% of Cys34 contains a free sulfhydryl group (human mercaptalbumin); about 25% forms a disulfi de bond with small sulfhydryl compounds, such as cysteine, homocysteine or glutathione (human nonmercapt albumin 1); and a small fraction is highly oxidized to sulfonic acid (human nonmercaptalbumin 2). In contrast, in chronic diseases, such as diabetes, renal failure, ischemic heart disease, and cancer, the oxidized form of albumin greatly increases, leading to an impairment of its biological activities [7]. Ischemia-modifi ed albumin (IMA) has the capacity to bind cobalt and other metal ions, a property that contributes to the anti-oxidant activity of albumin by preventing these molecules from catalyzing pro-oxidative reactions. Increased IMA levels have been proposed as a mortality predictor in renal failure [8] and myocardial infarction [9].

Human serum albumin and liver cirrhosis
Patients with advanced cirrhosis almost always have hypoalbuminemia caused both by decreased synthesis by the hepatocytes and water and sodium retention that dilutes the content of albumin in the extracellular space. Other factors likely contribute to the development of hypoalbuminemia, including an increased transcapillary transport rate [10].

Oncotic properties of albumin in cirrhosis
Th e therapeutic use of albumin in hepatology dates back to the 1960s when it was thought that hypoalbuminemia had a prominent role in the pathogenesis of ascites, because of alteration of the balance between the forces of Starling in the intrahepatic microcirculation. It was found that plasma albumin levels less than 3 g/l were almost constantly associated with the presence of ascites, which did not develop with plasma levels greater than 4 g/l [11]. Moreover, plasma oncotic pressure less than 20 mmHg signifi cantly increased the probability of developing ascites in the presence of portal hypertension [10]. Th ese fi ndings may support supplementation with exogenous albumin in patients with ascites and hypoalbuminemia. However, the net fl ow of fl uid from the vascular compartment to the interstitium is regulated by its transcapillary gradient rather than by its intravascular concentration alone. When the gradient between plasma and ascitic fl uid oncotic pressure was taken into account, no correlation was found with the rate of ascites formation [12], indicating that hypoalbuminemia simply refl ects a deterioration in liver synthetic function without playing a major, direct role in the pathogenesis of ascites formation.
From the 1990s better understanding of the cardiovascular alterations in patients with advanced cirrhosis led to a critical review of the clinical use of albumin. Th ese patients typically have a hyperdynamic circulatory syndrome, characterized by a decrease in peripheral vascular resistance and a compensatory increase in cardiac output, which manifest at the clinical level with arterial hypotension, tachycardia and hyperkinetic arterial pulses. Th e primary cause of the hyperdynamic circulatory syndrome is arterial vasodilation, mainly localized in the splanchnic circulatory area, of suffi cient magnitude to reduce the eff ective blood volume (i.e., the blood volume in the heart, lungs and central arterial tree that is sensed by arterial receptors). Arterial vasodilation in cirrhosis mainly results from the increased production of vasoactive substances, such as nitric oxide (NO), carbon monoxide, and endocannabinoids, which induce vasodilation and hamper the vascular response to   vasoconstrictors. Th e ensuing eff ective hypovolemia evokes the compensatory activation of neuro-humoral systems, including the renin-angiotensin-aldosterone axis, sympathetic nervous system, and arginine-vaso pressin, able to promote vasoconstriction and renal retention of sodium and water. As a result, from a functional point of view, patients with advanced cirrhosis are hypovolemic and exhibit cardiovascular hypo reac tivity, even though their cardiac output is usually signifi cantly elevated [13]. However, a decrease in cardiac output leading to an exacerbation of eff ective hypo volemia is observed in patients with severe disease, suggesting clinically relevant cardiac dysfunction in the more advanced stages of cirrhosis [13]. In addition to ascites, these cardiovascular alterations represent the pathophysiological background of several severe complications of cirrhosis, such as hepatorenal syndrome (HRS), precipitated or not by spontaneous bacterial peritonitis (SBP), and postparacentesis circulatory dysfunc tion (PPCD), which all share acute exacerbation of eff ective hypovolemia as the pathophysiological momentum.
Based on this background, preservation of central blood volume represents a major aim in the management of patients with advanced cirrhosis. Several controlled and/or randomized studies have shown that albumin administration is eff ective to prevent circulatory dysfunction after large-volume paracentesis [14] and renal failure after SBP [15], and to treat HRS when given together with vasocontrictors [16,17]. Furthermore, it is currently believed that the capacity of albumin to expand the central blood volume in cirrhosis is superior to that of several other plasma-expanders [18]. In contrast, the chronic use of albumin to treat ascites is still debated, due to the lack of defi nitive scientifi c evidence supporting its clinical benefi ts. Th us, albumin is now given in patients with cirrhosis to expand eff ective volemia regard less of its plasma level.

Prevention of post-paracentesis circulatory dysfunction
Large-volume paracentesis is the treatment of choice for the management of patients with massive or refractory ascites. Although large-volume paracentesis is a safe procedure and the risk of local complications, such as hemorrhage or bowel perforation, is very low, it can favor the development of PPCD, which is characterized by the exacerbation of eff ective hypovolemia and is defi ned by an increase of more than 50% in the basal plasma renin activity 4-6 days after paracentesis. PPCD predisposes to rapid re-accumulation of ascites, hyponatremia, renal failure, and shortened survival [19].
Th e occurrence of PPCD is signifi cantly reduced when large-volume paracentesis is combined with infusion of albumin (8 g/l of ascites removed). Albumin was more eff ective than other plasma expanders (dextran-70 or polygeline) for the prevention of PPCD when more than 5 l of ascites were removed; however, the incidence of PPCD was similar when less than 5 l of ascites were tapped [18,19]. Moreover, use of albumin after largevolume paracentesis compared with alternative but cheaper plasma expanders appears to be more costeff ective because albumin post-paracentesis is associated with fewer liver-related complications within the fi rst month [20].
Despite these fi ndings, the clinical value of albumin administration as an adjunct to large-volume paracentesis has been questioned because a survival benefi t could not be shown in individual randomized trials. However, a meta-analysis that included all available randomized clinical trials comparing albumin with alternative treatments showed that albumin infusion signifi cantly reduces the incidence of PPCD, hyponatremia and mortality, providing evidence-based support for the well-accepted clinical practice of infusing albumin as fi rst choice in patients requiring large-volume paracentesis (Bernardi M, unpublished results).

Hepatorenal syndrome
HRS is defi ned as the occurrence of renal failure in patients with advanced liver disease without another identifi able cause of renal failure, and is usually classifi ed into two types according to the defi nition of the International Ascites Club [21]. Type 1 HRS is a rapidly progres sive acute renal failure diagnosed when the serum creatinine increases more than 100% from baseline to a fi nal level greater than 2.5 mg/dl within two weeks. It is usually precipitated by an acute insult often represented by a bacterial infection, although it may occur without any clear triggering event. Type 2 HRS is a relatively steady but moderate degree of functional renal failure with serum creatinine levels above 1.5 mg/dl. It is usually associated with refractory ascites and often hypo natremia. Patients with type 2 HRS may eventually develop type 1 HRS either spontaneously or following a precipitat ing event, such as SBP.
HRS can be seen as the fi nal event of the hemodynamic disturbances characterizing advanced cirrhosis. Th e development of HRS is related to an extreme reduction in the eff ective arterial blood volume combined with a decrease in the mean arterial pressure (MAP). Th is condition results from either a marked vasodilatation mainly in the splanchnic arterial bed or an impairment of cardiac function as a result of cirrhotic cardiomyopathy. Th us, the cardiac output in patients with HRS may be low or normal (infrequently high), but always insuffi cient for the patient's needs [21]. Eff ective hypovolemia produces a very intense compensatory activation of the sympa thetic nervous and renin-angiotensin-aldosterone systems which causes renal vasoconstriction and a shift in the renal autoregulatory curve, rendering renal blood fl ow much more sensitive to the negative eff ects of low MAP [22].
Th e prognosis of type 1 HRS is dismal with almost 90% of patients dying without therapy within 2 weeks. In the last decade, combined therapy with vasoconstrictors and albumin has proved to be eff ective in reversing HRS in approximately half of the cases [22]. Th e rationale for this approach is to improve the markedly impaired circulatory function by causing vasoconstriction of the extremely dilated splanchnic vascular bed and by antagonizing eff ective hypovolemia so that MAP and renal perfusion can increase [22].
Terlipressin, a vasopressin analog, is the most studied vasoconstrictor and two multicenter randomized trials have demonstrated the effi cacy of increasing doses of terlipressin in combination with albumin (1 g/kg or 100 g on day 1 followed by 20-40 g/day) ( Figure 2) [16,17]. Albumin administration appears necessary to improve the effi cacy of the vasoconstrictor as the rate of HRS reversal is signifi cantly lower with terlipressin alone compared to combination therapy [23]. Th e most frequent side eff ects of treatment are cardiovascular or ischemic complications, which have been reported in more than 10% of patients treated [24].
Vasoconstrictors other than vasopressin analogs that have been used in the management of type 1 HRS include norepinephrine and midodrine plus octreotide, both in combination with albumin. Although the number of patients enrolled in the studies evaluating these vasoconstrictors is quite small, their use appeared to improve renal function in a consistent number of patients with HRS [25,26]. Finally, a systematic review of randomized studies using terlipressin as well as other vasoconstrictors, always in combination with albumin, has shown that treatment is associated with an improved short-term survival [24]. Th is result may be of paramount importance to keep patients waiting for liver transplantation alive until an organ becomes available [27].

Prevention of hepatorenal syndrome after spontaneous bacterial peritonitis
All patients with cirrhosis and ascites are at risk of developing a bacterial infection of the ascitic fl uid, called SBP, which is diagnosed when the neutrophil count in ascitic fl uid exceeds 250/mm 3 as determined by micro scopy [19]. SBP, even without septic shock, may precipitate circulatory dysfunction with severe liver failure, hepatic encephalopathy and type 1 HRS, causing the death of the patient in approximately 20% of cases despite infection resolution with antibiotic treatment [19].
Administration of albumin at a dose of 1.5 g/kg body weight at diagnosis followed by 1 g/kg on day 3 signifi cantly decreased the incidence of type 1 HRS from 30% to 10% and reduced mortality from 29% to 10% compared with cefotaxime alone in a randomized, controlled study in patients with SBP ( Figure 3) [15]. If the effi cacy of albumin administration in patients with advanced disease, as wit nessed by a serum bilirubin greater than 4 mg/dl and a serum creatinine greater than 1 mg/dl, appears to be undisputed, this is not the case for patients with moderate liver failure and without renal dysfunction at diagnosis as they show a very low incidence of HRS precipitated by SBP [15]. At the same time,  it is not established whether crystalloids or artifi cial colloids could replace albumin, although equivalent doses of hydroxyethyl starch (HES) have not been found to be as eff ective as albumin in preventing the circulatory dysfunction caused by SBP [28]. Interestingly, albumin infusion was able to prevent the increase in nitrate levels and to reduce the elevated plasma levels of Von Willebrand-related antigen observed in patients with SBP treated with HES [29]. Th is fi nding suggests an eff ect of albumin on endothelial function that appears to be related to the non-oncotic properties of this molecule rather than to its activity as plasma-expander.

Ascites
Th e chronic use of albumin to treat ascites is still debated, due to the lack of defi nitive scientifi c evidence supporting its clinical benefi ts. When human albumin became available, studies performed several decades ago failed to show clear usefulness of this substance in relieving ascites and preventing its recurrence; however, these investigations were uncontrolled and/or included a small number of patients [30]. Th us, treatment of cirrhotic ascites with diuretics and albumin has been practiced on anecdote and experience for many years. More than a decade ago, a controlled clinical trial in hospitalized patients with cirrhosis and ascites randomized to receive diuretics associated or not with low-doses of albumin (12.5 g/day) showed that treatment with diuretics plus albumin was overall more eff ective than diuretics alone in determining the disappearance of ascites and reducing the length of the hospital stay [31]. However, this result was only achieved in patients who received the initial dose of diuretics, whereas the diff erence in the treatment response was not statistically signifi cant in those patients who required higher doses of diuretics or were refractory to diuretic treatment. Albumin administration at home also reduced the rate of ascites recurrence and readmission to the hospital due to ascites, but its prescription was very expensive and no eff ect on survival was observed [31]. More recently, the same research group showed, in an unblinded randomized trial, that longterm albumin administration is also able to increase patient survival and to reduce the risk of ascites recurrence [32].
No other controlled clinical trials have so far been performed to evaluate the eff ectiveness of prolonged albumin administration in the treatment of cirrhosis and ascites. Th us, the absence of confi rmatory multicenter random ized studies, together with the high cost of human albumin, explains why albumin infusion is not usually included among the therapeutic options for diffi cult-to-treat ascites. A multicenter randomized clinical trial (NCT 01288794, www.clinicaltrials.gov) planning to enroll more than 400 patients is ongoing in Italy to cover this lack of information.

Non-oncotic properties of albumin in cirrhosis
In recent years, scientifi c interest into the non-oncotic properties of human serum albumin has progressively increased in the fi eld of hepatology. Initial fi ndings showed that patients with chronic liver disease have decreased amounts of circulating human nonmercaptalbumin (the 'good guy') and elevated levels of the oxidized form human nonmercaptalbumin 26 (the 'bad guy') [33,34]. More recently, binding, transport, and detoxifi cation capacities of albumin have been found to be severely impaired in cirrhotic patients, particularly in those with acute-on-chronic liver failure precipitated by bacterial infections [35]. As a result, endogenous and exogenous substances that are normally bound to albumin are free to react arbitrarily. Th e reduced detoxifi cation capacity of albumin is especially important, as it suggests that the molecule cannot bind and remove waste products. Furthermore, IMA is signifi cantly increased, suggesting an impaired metal chelation capacity that contributes to generate an environment of sustained oxidative stress by favoring radical formation through Fenton chemistry processes. Most important, the loss of albumin function and the increase in the IMA level were associated with poor survival [8,9,35]. From a clinical context, the non-oncotic properties of albumin have so far been exploited in extracorporeal albumin dialysis systems in patients with acute or acuteon-chronic liver failure. Th e two commercially available detoxifi cation systems are the "Molecular Adsorbent Recirculating System" (MARS) and the "fractionated plasma separation and adsorption" (Prometheus), which are based on the association of a conventional dialysis membrane with a second dialysis circuit fi lled with a circulating 20% albumin solution in the case of MARS or with a second fi lter containing albumin in the case of Prometheus. In addition to the positive consequences associated with the removal of toxic molecules, early studies of these systems showed a favorable impact on systemic hemodynamics likely due to the purifi cation of substances involved in the pathogenesis of the hyperdynamic circulatory syndrome, such as NO, and the proinfl ammatory cytokines, TNF-α and IL-6 [36,37]. Such systems, currently available only in specialized and transplant centers, have a potential therapeutic role in the following medical conditions: Acute fulminant liver failure, acute liver failure on chronic liver disease, primary dysfunction of the transplanted liver, liver failure after resection and intractable pruritus [37]. It has to be stressed, however, that the results of the fi rst large-scale controlled trials that have employed either MARS or Prometheus in patients with cirrhosis and acute-onchronic liver failure did not show a signifi cant impact on major outcomes, such as survival [38,39]. Likely, subgroups of patients who would actually benefi t from these detoxifi cation systems need to be identifi ed, as suggested by the survival benefi t observed in patients with severe cirrhosis (model for end-stage liver disease [MELD] score > 30) and HRS type 1 treated with the Prometheus system [39].

Conclusion
Th e use of human albumin in the setting of liver cirrhosis is supported by evidence arising from prospective randomized trials and meta-analyses. Albumin administration is, therefore widely accepted and recommended by current international guidelines for the prevention of PPCD and acute renal failure in patients with SBP, and the treatment of HRS as an adjunct to vasoconstrictors.
All these complications share the common patho physiological background of reduced eff ective volemia, mainly because of peripheral arterial vasodilation, and the rationale underlying the use of albumin is related to its activity as a plasma expander. However, it is likely that the benefi cial eff ects of albumin are also linked to its non-oncotic properties, including binding capacity, antioxidant activity, and eff ects on capillary integrity. Th ese eff ects represent the main fi elds of current and future research, from which will likely arise further and more appropriate indications for albumin administration in patients with liver disease.