Liver Failure Group, UCL Institute for Liver and Digestive Health, Royal Free Hospital, University College London, London, UK
Address reprint requests to: Rajiv Jalan, Liver Failure Group, UCL Institute for Liver and Digestive Health, Royal Free Hospital, Upper Third Floor UCL Medical School, Pond Street, London NW3 2PF, UK. E-mail: firstname.lastname@example.org
Potential conflict of interest: Rajiv Jalan: research funding and collaboration with Grifols (manufacturer of albumin) and GAMBRO (manufacturer of MARS machine). His university owns intellectual property regarding the development of a new liver device (UCL-ARSeNEL) using albumin dialysis and is developing a new biomarker (DASIMAR) which has as element albumin functionality. Mauro Bernardi: CSL Behring GmbH consultant and speaker and also a Plasma Protein Therapeutics Association (PPTA) Europe speaker. Paolo Caraceni: Grifols SA speaker. Pere Gines: Has received research funding from Grifols, and grants from Instituto de Salud Carlos III FIS PI12/00330 and EC07/90077. Vicente Arroyo: Has received grant and research support from Grifols. The EASL-CLIF Consortium is the first result of an initiative of several European and American investigators to potentiate research in Chronic Liver Failure (CLIF). An assembly of European hepatologists proposed that the European Association for the Study of the Liver (EASL) endorse a Consortium aimed to promote research on CLIF, to stimulate the formation of research groups in this field in Europe, and to identify potential areas of common interest with European industry. The Executive Committee of EASL accepted endorsing the consortium in June 2009 and elected Vicente Arroyo and Mauro Bernardi as Chairman and Vice-Chairman for a period of 5 years. Twelve other EASL members proposed by the assembly were elected to form the Steering Committee. During the period 2009-2012 the EASL-CLIF Consortium received unrestricted grants from Grifols and Gambro. Grifols has prolonged its unrestricted grant for an additional period of 4 years. There is no other support for the consortium. The Fundació Clinic, a foundation ruled by the Hospital Clinic and University of Barcelona, administers the EASL-CLIF Consortium grants. V. Arroyo, M. Bernardi, and members of the Steering Committee have no relationship with Grifols or Gambro other than conferences in international meetings (from which they may receive an honorarium) or as investigators on specific projects unrelated to the consortium. Up to now the CLIF Consortium has not performed any study promoted by pharmaceutical companies, although this is an important objective of the Consortium.
Rita Garcia-Martinez was supported by fellowship CM07/00109 from Fondo de Investigacion Sanitaria (Instituto de Salud Carlos III), Beca de Formaciò al Estranger (Societat Catalana de Digestologia), Beca Juan Rodés (Asociación Española para el Estudio del Hígado), and Sheila Sherlock EASL fellowship (European Association for the Study of the Liver).
Since the introduction of human serum albumin as a plasma expander in the 1940s, considerable research has allowed a better understanding of its biochemical properties and potential clinical benefits. Albumin has a complex structure, which is responsible for a variety of biological functions. In disease, the albumin molecule is susceptible to modifications that may alter its biological activity. During the last decades, different methods to measure albumin function have been developed. Recent studies have shown that not only albumin concentration but also albumin function is reduced in liver failure. This observation led to the concept of effective albumin concentration, which represents the fact that plasma albumin concentration does not reflect its function. Indeed, in liver disease albumin function is several times less than its concentration. In patients with cirrhosis, albumin infusion reduces mortality in patients with spontaneous bacterial peritonitis and improves outcome following large volume paracentesis. In combination with vasoconstrictors, albumin is useful in the management of patients with hepatorenal syndrome. Its role is being investigated in a large number of indications, which rely on its volume and nonvolume expansion functions such as stroke, severe sepsis, Alzheimer's disease, malaria, burns, and ovarian hyperstimulation syndrome. This review explores the above concepts, reviews the available evidence for the use of albumin in liver diseases, defines therapeutic limitations, and explores the challenges that should be addressed in future research. (Hepatology 2013;58:1836–1846)
Human serum albumin (HSA) is responsible for 75% of the plasma oncotic pressure and, therefore, intravascular HSA administration increases circulating blood volume. It was introduced as a plasma expander during World War II. Since then, it has been widely used essentially as a plasma volume replacement. Thereafter, other plasma expanders were developed mainly because they are not blood-derived and significantly less expensive. However, a better understanding of albumin structure and function has led to the concept that this molecule is complex and may subserve additional functions. Albumin has been shown to be a multifunctional protein with antioxidant, immunomodulatory, and detoxification functions. Impairment in albumin function during disease has been documented, leading to the notion that HSA administration could partially restore these biological functions. During the last decades the use of HSA has been a source of debate focusing on whether HSA is a plasma volume expander or whether HSA infusion can be beneficial regardless of its oncotic function.
This review aims to highlight HSA function derived from in vitro, in vivo, and human studies that generates the hypothesis that albumin is more than just a fluid.
Biological Properties of Albumin
Albumin is a constituent of the blood in mammalian and avian species, and its structure across these vastly different organisms shows great similarity. In healthy individuals HSA constitutes 50% of the plasma proteins. It is synthesized in the liver (10-15 g/d) and released into the intravascular space. From the whole body albumin pool, 30%-40% stays in the intravascular compartment, whereas the remaining goes to the interstitial space through capillaries and back to systemic circulation by way of the lymphatic system. Although albumin is thought to be an extracellular molecule, it is taken up by many cell types by endocytosis and catabolized through lysosomal degradation (Fig. 1). The half-life of albumin in health is 12-19 days but this is altered in disease. The mechanism governing this is not fully understood. It has been reported that the half-life of albumin is prolonged in cirrhosis but the observed hypoalbuminemia is likely to be multifactorial. Although more than 75 human isoforms have been described, albumin is a very stable protein.
Human albumin is a small globular protein (67 kDa) comprised of 609 amino acids. It has a higher content of acidic amino acids, which gives HSA its negative charge at pH 7. It contains a free cysteine (Cys-34) residue which accounts for the single free redox-active thiol (-SH) moiety of the molecule that is capable of thiolation, nitrosylation, and oxidation. In health, albumin exists predominantly (70%-80%) in reduced form, known as mercaptoalbumin (HMA). A small fraction (20%-30%) exists as mixed disulfide compounds (cysteine, homocysteine, and glutathione), known as nonmercaptalbumin1 (HNA1). HSA also can be found in a highly oxidized form (5%) with the Cys-34 as sulfinic or sulfonic acid known as nonmercaptalbumin 2 (HMA2) (Fig. 2).
A detailed three-dimensional structure remained unknown until the 1990s, when crystals of HSA at low resolution were reported. Since then, several reports on crystals of albumin alone or in combination with different ligands enhanced knowledge of the structural features of the protein and its interactions with different molecules. According to these investigations, albumin has a heart-shaped tertiary structure with high α-helical content and three homologous domains (I-III) (Fig. 2). This structure is stable but flexible, which allows the molecule to accommodate many protein-bound compounds. These studies also identified specific binding sites for a large number of substances. Albumin structure is susceptible to modification, following both enzymatic and nonenzymatic reactions. Whether structural changes correlate with its biological functions is not completely clarified. It has been shown that glycated albumin participates in the pathogenesis of diabetes-related vascular complications. Moreover, a genetic variant of HSA (Liprizzi) has a potent antibacterial activity in vitro. These data support the hypothesis that there is a link between albumin structure and function.
The multifunctional properties of albumin are shown in Fig. 3.
Plasma oncotic pressure. This is the most well-known property of albumin, which relies on its high plasma concentration and its net negative charge. HSA constitutes more than half of total plasma proteins. Its net negative charge facilitates the attraction of molecules of sodium, and secondarily, water. Thus, albumin represents 75% of the plasma oncotic pressure.
Solubilization, transport, and metabolism. Human albumin has the capacity to bind an extraordinarily diverse range of molecules. This is possible because the negative charge of HSA facilitates electrostatic binding of many substances, acting as a storage and vehicle for many compounds. Human albumin also performs transport functions through binding sites formed within its tertiary structure. Among the substances transported by albumin there are a large number of drugs, bilirubin, bile acids, hormones, metals, anions, long-chain fatty acids, L-thyroxine, nitric oxide, endotoxin,[6, 9] and other bacterial products such as the protein G-like albumin binding molecule.
Antioxidant. The thiol group of albumin accounts for about 80% of the extracellular thiols, making it the most important extracellular antioxidant. This group is an avid scavenger for different oxidative and nitrosative reactive species. In addition, albumin also restricts oxidative stress damage by neutralizing free Cu2+ and iron, ions that catalyze reactions in which free radicals are released. Moreover, albumin shows high affinity for heme groups. This albumin-heme complex has been shown to provide a lipid antioxidant effect.
Immunomodulation. Human albumin is known to have endotoxin-binding capacity and in vitro studies showed that endotoxin activity decreases in the presence of physiological albumin concentrations. In cirrhotic patients with acute alcoholic hepatitis, neutrophil dysfunction was associated with higher risk of infection, organ failure, and mortality. In vitro studies suggested that this could be prevented by albumin.[14, 15] Whether this effect is related to the inactivation of endotoxin or to additional effects is unclear. Ex vivo studies performed in human pulmonary epithelial cells (replicated in vivo) showed that albumin increases intracellular glutathione and regulates nuclear factor-kappa B (NF-κB) activation. Similarly, albumin inhibits TNF (tumor necrosis factor) α-induced upregulation of vascular cell adhesion molecule-1 (VCAM-1) and NF-κB activation in human aortic endothelial cells, suggesting that albumin enhances intracellular protection against inflammatory and oxidative stress damage. Some clinical studies found a positive effect of HSA infusion in patients suffering from infection, either with or without liver diseases,[18, 19] but definitive studies are needed.
Capillary permeability. More than 50% of total body albumin is present in the extravascular compartment and may directly influence vascular integrity and permeability by way of interactions with the extracellular matrix.
Hemostatic effects. Albumin is able to bind nitric oxide (NO) in position Cys-34. This ability may underlie the mechanism by which albumin dialysis using MARS (molecular absorbent recirculatory systems) increases mean arterial pressure. Potential clinical effects of nitroalbumin (HSA-NO) include vasodilatation and inhibition of platelet aggregation. Clinical studies suggest that hypoalbuminemia is linked to hyperaggregation of platelets and that albumin modifications can impact platelet aggregation. This action may be modulated through nitrosoalbumin.
Endothelial stabilization. Vascular endothelium produces substances that maintain vascular homeostasis, modulating vascular tone, thrombogenesis, fibrinogenesis, and maintenance tissue integrity. Inflammation and oxidative stress are factors that are known to alter this balance and induce endothelial dysfunction. The ability of albumin to modulate inflammation, reduce oxidative damage, and interfere in neutrophil adhesion could therefore potentially impact endothelial function. Ex-vivo studies support the beneficial effect of albumin stabilizing the endothelium.[17, 22] In a randomized clinical study among patients with spontaneous bacterial peritonitis (SBP), those who received HSA showed an improvement in systemic hemodynamics and markers of endothelial dysfunction, whereas those who were treated with colloid did not, supporting a relevant role of albumin on endothelial stabilization.
Methods of Assessment of Albumin Function and Their Prognostic Role
Albumin function is closely related to the tertiary structure of the protein, since both posttranslational modifications and saturation of the binding sites can impact the biological actions of albumin. Consequently, albumin function depends on the total plasma albumin concentration and on its functional capacity. Since serum albumin concentration does not provide information regarding the functional state, it is not an accurate index of its biological function. Several methods have been developed in order to estimate the biological activity of albumin that have some usefulness as markers of disease severity in liver disease.
Albumin binding capacity (ABiC). This method is based on the estimation of the unbound fraction of a specific fluorescent marker (Dansylsarcosine) on the albumin binding site II. This semiquantitative method has been tested in patients with cirrhosis and acute-on-chronic liver failure (ACLF) showing that the more advanced the disease (higher Child-Pugh or Model for Endstage Liver Disease [MELD] score), the worse the ABiC. In addition, treatment with albumin dialysis was associated with a decrease in plasma levels of bilirubin and bile acids (albumin-bound substances) and an improvement in ABiC. These data suggest a competitive interaction in albumin binding sites by endogenous molecules which is intensified with the severity of the disease and potentially reversible. Progressive albumin functional impairment has also been demonstrated in patients with chronic kidney disease using this test. Although ABiC shows an association with severity of liver disease, its prognostic value and its correlation with other functional methods have not been studied.
Albumin fatty acid binding capacity. The principle of this technique relies on the use of a spin label detectable by electron paramagnetic resonance (EPR). The spin label used for albumin is 16-doxyl stearic acid, which is a specific fatty acid for this protein. EPR generates information regarding binding affinity, strength of the binding, and transport of the fatty acid together with data on the metabolism of toxic substances. It has been shown that albumin function is decreased in cirrhosis and further decreased in ACLF. In addition, the detoxification efficiency of albumin was significantly reduced in nonsurvivors, although its prognostic role was not completely clarified in that study. Albumin fatty acid binding capacity has been explored in cancer and sepsis, showing promising diagnostic and prognostic utility.
Ischemia modified albumin (IMA). This technique is based on the ability of HSA in binding cobalt. It has been shown that this binding affinity is altered in different pathological conditions. Initially, it was found that the avidity of cobalt for albumin is decreased in myocardial ischemia. Following these observations, similar alterations were found in other ischemic conditions such as stroke or mesenteric ischemia, revealing its lack of tissue specificity. In addition, the same phenomenon was also observed in nonischemic diseases with recognized oxidative stress such as ketoacidosis, endstage kidney disease, and liver disease. Therefore, IMA became a marker of oxidative stress. In liver failure it has been shown that the normalized ischemia modified albumin ratio (IMA/total albumin [IMAR]) correlated with the severity of the disease (MELD) and was significantly higher in nonsurvivors. In addition, IMAR correlated with other functional parameters assessed by EPR. Interestingly, different investigators reported its reversibility after a short period of time, suggesting that transient conformational changes in the structure of albumin can lead to this functional disturbance.
Albumin oxidative state. It is possible to measure the redox status of albumin. The three major albumin species according to the Cys34 oxidative state can be separated and quantified by high-performance liquid chromatography (HPLC). An increase in the HNA1 and HNA2 has been documented in end stage kidney disease and in liver diseases. In both cases, the degree of oxidative damage to albumin correlated with the progression of the illness. The functional consequences of this modification have been assessed using different oxidative methods and functional approaches. Albumin oxidation is associated with decreased antioxidant capacity. This is associated with a reduced binding ability at binding site II and also occasionally at binding site I.
Some other functional methods have been used in order to measure albumin function such as quantification of unbound ligands by HPLC, circular dichroism, or nuclear magnetic resonance. A potential advantage of IMA is its possible prognostic value, although this needs further validation. Nevertheless, it is possible that many of these tests exhibit independent information that can be complementary.
Recently, the functional capacity of albumin in cirrhosis was shown to be impaired. In this study, albumin function was assessed using two techniques: electron paramagnetic resonance and ischemia-modified albumin. Albumin dysfunction closely correlated with the degree of liver failure. Moreover, the ratio of ischemia-modified albumin / total albumin was significantly higher in nonsurviving patients with ACLF. It is worth noting that albumin dialysis (MARS) did not restore the native HSA dysfunction, suggesting that in liver failure albumin may be irreversibly destroyed.
The mechanism of this functional derangement remains unclear. The fact that poorer liver function correlates with albumin dysfunction suggests that toxin accumulation may physically impair its structure and prevent normal function. Alternatively, it is possible that HSA structure has been modified (posttranslational modification) such that either the binding sites are altered and/or there is a change in the tertiary structure. It is also not known whether the cirrhotic liver is capable of producing normal, functional HSA.
The fact that patients with liver failure are hypoalbuminemic and also have albumin dysfunction generates the novel concept of “effective albumin concentration.” This concept makes reference to the actual functionality of circulating albumin assessed by EPR as a product of total concentration of albumin (g/L) and percent detoxification efficiency (0 to 1). It has been shown that effective albumin concentration in cirrhosis patients is several fold less than the actual HSA concentration and further reduced in ACLF. This disturbance leads to alterations in the transport and metabolism of many endogenous and exogenous substances and cause conflicts in other systems habitually influenced by a normal albumin function such as redox balance, coagulation, and inflammation.
Human albumin infusion could potentially compensate these disturbances and improve HSA function. The benefit of HSA described in clinical trials may well be explained on this basis, acting at different levels (Fig. 4), but further studies are needed to determine the function/effect relationship. A randomized unblinded trial showed that long-term HSA administration was able to increase patient survival and reduce the risk of ascites recurrence. Whether long-term treatment with HSA for preventing clinical decompensation in cirrhosis is a valid concept is currently being investigated (NCT00839358 http://clinicaltrials.gov/ct2/home).
Evidence for Albumin Administration in Liver Diseases
Since the introduction of HSA as a plasma expander, it has been broadly used in patients with or without liver diseases. In fact, it has been widely utilized in critically ill patients for circulatory support, showing a favorable safety profile. Similarly, it has been used in women with ovarian hyperstimulation syndrome and in patients with malaria and burns, showing positive effects. More recently, some studies have disclosed a potential protective role in patients with ischemic stroke[45, 46] and Alzheimer's disease.
Table 1. Clinical Studies Evaluating the Effect of Albumin Infusion in Patients with Spontaneous Bacterial Peritonitis in Cirrhosis
Patients with cirrhosis develop portal hypertension, which is associated with splanchnic arterial vasodilatation, consequently leading to high cardiac output, increased heart rate, and reduced peripheral vascular resistance culminating in a reduction in mean arterial pressure. This cardiovascular scenario, known as hyperdynamic circulation, deteriorates with the advance of liver failure and leads to a progressive reduction in individual organ perfusion. Human albumin was introduced as a treatment in the management of cirrhosis patients with hypoalbuminemia and ascites in the 1950s. Initially it was thought that its benefits relied on its facility to expand intravascular volume. In subsequent years, a better understanding of the pathophysiology of liver failure, together with a better knowledge of HSA biology, have changed this view. Recently, improvements in the management of cirrhosis patients have led to increases in survival. The current clinical approach includes well-established indications for HSA infusion[50, 51]:
Spontaneous bacterial peritonitis is a frequent infection in cirrhosis patients with ascites and usually complicated with systemic inflammatory response. This trigger can cause deterioration in liver and hemodynamic functions and subsequently progress to multiorgan failure, despite successful sterilization of the ascitic fluid. Several studies[18, 23, 52] and a recent meta-analyses evaluated the effect of HSA infusion during SBP (Table 1). They found that HSA administration, in addition to antibiotics, reduces the incidence of renal failure and decreases mortality. Additionally, two studies pointed out that those patients at low risk of developing renal failure have a favorable outcome without receiving HSA.[55, 57] This suggests that HSA infusion would have a greater benefit in sicker patients. Indeed, those studies showed that patients at high risk certainly benefited from receiving HSA infusion.
Hepatorenal syndrome (HRS) develops from an extreme form of circulatory dysfunction. The splanchnic arterial vasodilatation together with cirrhotic cardiomyopathy (characterized by diastolic/systolic dysfunction) result in severe underfilling of the systemic vascular territory with renal hypoperfusion, ultimately leading to renal failure. Several studies have demonstrated that a combination strategy of HSA infusion with vasoconstrictors has a significant survival benefit in these patients. One of the potential mechanisms of the albumin benefit could be the improvement in cirrhotic cardiomyopathy, as has been shown in experimental cirrhosis.
Paracentesis-induced circulatory dysfunction (PICD) consists of an exacerbation of arteriolar vasodilatation following a large volume paracentesis. Although PICD can cause acute renal failure with a high associated mortality rate, the incidence of this complication is relatively low. The infusion of HSA has been widely evaluated in the prevention of this syndrome, showing a positive effect preventing renal failure and improving the outcome. A recent meta-analysis showed that HSA is the most effective agent in the prevention of humoral, hemodynamic, and clinical effects associated with this condition (Fig. 5). The potential benefits of HSA infusion have been explored in other scenarios in cirrhosis.
Hepatic encephalopathy (HE) is a common complication of cirrhosis. Ammonia, inflammation, and circulatory disturbances are involved in its pathogenesis. A clinical study comparing 4.5% HSA or colloid in patients with diuretic-induced HE showed an improvement in hemodynamics and in plasma ammonia levels in both groups. However, a more marked improvement in HE together with a reduction in oxidative stress markers was found in HSA-treated patients. Another randomized trial performed in patients with severe HE showed that a neurological improvement was reached faster and more frequently in patients receiving standard medical therapy plus albumin dialysis than in those who received standard medical therapy alone. These data indicate a potential benefit of albumin dialysis in HE through its detoxification properties.
Infections are common complications of cirrhosis frequently associated with the development of renal dysfunction. A randomized study in cirrhosis patients with infection showed an improvement in circulatory and renal function in patients who received HSA. A 3-month survival benefit was not observed in the intention to treat analysis, which could be due to the fact that relatively “low-risk” patients were enrolled. An adequately powered study is due to start under the auspices of the CLIF-consortium.
Albumin Dialysis: What We Have Learned
Acute and severe liver failure in patients without preexisting liver disease (acute liver failure [ALF]) or among patients with previously stable chronic liver disease (ACLF) is a condition that carries high mortality. Liver transplantation (LT) is the only effective treatment for these patients. However, the current organ shortage limits its use. Liver support systems are being developed as a temporary therapy, which can be used on an emergency basis as a bridge to liver regeneration or LT. Since toxin accumulation during liver failure is involved in the persistence of liver disease and in the subsequent organ failure, a detoxification approach seems a promising therapeutic option. Many of the toxic substances such as bilirubin, endotoxin, and cytokines are mostly albumin-bound. For this reason, the current artificial liver devices are based on detoxifying albumin. Most of the data regarding the efficacy of these devices are from studies that used MARS or Prometheus. Both systems combine removal of albumin-bound and water-soluble substances.
Both systems have been tested in ALF and ACLF. Improvement in systemic and liver hemodynamics, in biochemistry, and in organ function (intracranial pressure in ALF and HE) has been documented in randomized studies.[69, 72, 73] Despite its safety and effectiveness in removing toxins and clinical benefits, no survival improvement has been shown in large clinical trials in ALF or ACLF.[75, 76] In the FULMAR study, 102 patients with fulminant/subfulminant liver failure (38% with paracetamol overdose) were randomized to receive standard medical therapy alone or in association with MARS. No significant differences in 6-month survival were observed. Two large randomized trials evaluated these devices in ACLF: the RELIEF study randomized 189 patients and the HELIOS study included 145, both of them over a period of more than 7 years. Neither device showed improvement in survival.
One of the many possible reasons why these devices have not shown improvement in outcome is because the treatment did not alter the impaired albumin function, suggesting that HSA may be irreversibly modified in liver failure. In contrast, a recent study using high-volume plasma exchange has shown improved transplant-free survival in patients with ALF, suggesting that the principle of detoxification and HSA exchange may have beneficial effects. Based on this hypothesis, the group at University College London has been testing a new device based on the principle of albumin exchange and endotoxin removal (UCL-ARSeNEL). Early data in a large animal model of ALF showed a survival advantage.
It is clear from the above discussion that albumin is a multifunctional protein that has a diversity of biological functions and effects. Emerging data have started to build the concept that albumin is substantially more than just a fluid and its administration should be guided by functional rather than quantitative endpoints. Thus, recent investigations have led to the development of the concept “effective albumin concentration.” Several research questions remain to be answered, but the most urgent and important is to further establish a structure-function-biological activity relationship and validate the concept of “effective albumin concentration.” In addition, it is also a priority to perform long-term clinical trials of albumin use to prevent future complications in cirrhosis and further develop the concept of albumin dialysis based on our better understanding of albumin biology.
The authors thank Dr. Nathan Davies and Dr. Raj Mookerjee for helpful advice.