Hepatic encephalopathy: A central neuroinflammatory disorder?


  • Roger F. Butterworth

    Corresponding author
    1. Neuroscience Research Unit, Hôpital Saint-Luc, Centre Hospitalier de l'Université de Montréal, Montreal, Quebec, Canada
    • Neuroscience Research Unit, Hôpital Saint-Luc, Centre Hospitalier de l'Université de Montréal, 1058 St-Denis Street, Montreal, Quebec, Canada H2X 3J4
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Studies of the Neuroscience Research Unit are funded by the Canadian Institutes of Health Research.


Encephalopathy and brain edema are serious central nervous system complications of liver failure. Recent studies using molecular probes and antibodies to cell-specific marker proteins have demonstrated the activation of microglial cells in the brain during liver failure and confirmed a central neuroinflammatory response. In animal models of ischemic or toxic liver injury, microglial activation and concomitantly increased expression of genes coding for proinflammatory cytokines in the brain occur early in the progression of encephalopathy and brain edema. Moreover, the prevention of these complications with mild hypothermia or N-acetylcysteine (two treatments known to manifest both peripheral and central cytoprotective properties) averts central neuroinflammation due to liver failure. Recent studies using anti-inflammatory agents such as ibuprofen and indomethacin have shown promise for the treatment of mild encephalopathy in patients with cirrhosis, whereas treatment with minocycline, a potent inhibitor of microglial activation, attenuates the encephalopathy grade and prevents brain edema in experimental acute liver failure. The precise nature of the signaling mechanisms between the failing liver and central neuroinflammation has yet to be fully elucidated; mechanisms involving blood-brain cytokine transfer and receptor-mediated cytokine signal transduction as well as a role for liver-related toxic metabolites such as ammonia have been proposed. The prevention of central proinflammatory processes will undoubtedly herald a new chapter in the development of agents for the prevention and treatment of the central nervous system complications of liver failure. (HEPATOLOGY 2011;)

Hepatic encephalopathy (HE) is a serious neuropsychiatric complication of both acute liver failure (ALF) and chronic liver failure with the potential to affect heath-related quality of life, clinical management strategies, liver transplant priority, and patient survival. The neuropathological features of HE primarily include changes in the morphology and function of cells of the glial (rather than neuronal) lineage and have led to the suggestion that HE is a primary gliopathy. In particular, morphological changes in astroglial cells are characteristic of HE. Such changes include cell swelling, a characteristic cell phenotype known as Alzheimer type II astrocytosis, and concomitant alterations in the expression of genes coding for a wide range of astrocytic proteins with key roles in the control of cellular energy status, cell volume regulation, and neurotransmission.1 The causes of these alterations of astroglial integrity have generally been attributed to the toxic effects of ammonia. However, in recent years, attention has increasingly been focused on the role of proinflammatory mechanisms. There is now a substantial body of evidence from studies in patients and animal models demonstrating that systemic inflammation causes worsening of encephalopathy and that proinflammatory mechanisms may act synergistically with ammonia toxicity and result in the cerebral complications of ALF and chronic liver failure. Several review articles have been written on the role of systemic inflammation in the pathogenesis of HE.2-4


ac, anterior cingulate cortex; ALF, acute liver failure; CNS, central nervous system; HE, hepatic encephalopathy; IL, interleukin; PET, positron emission tomography; SIRS, systemic inflammatory response syndrome; TNF-α, tumor necrosis factor α.

Neuroinflammation During ALF

In a landmark study of 16 patients with ALF due primarily to acetaminophen hepatotoxicity, Wright et al.5 measured proinflammatory cytokines such as tumor necrosis factor α (TNF-α), interleukin-1β (IL-1β), and IL-6 in blood sampled from an artery and a reverse jugular catheter. A significant correlation was observed between arterial cytokine levels and intracranial hypertension, and brain cytokine efflux was noted that was consistent with brain cytokine production. Working with an animal model of ALF, Jiang et al.6 demonstrated that alterations of a second type of glial cell, the microglia, accompany the onset of HE and brain edema in ALF. Microglia are bone marrow–derived myeloid lineage cells that represent approximately 15% of the total central nervous system (CNS) cell population. In the absence of an inflammatory stimulus, microglia remain quiescent and are involved in surveillance (the resting phenotype). However, in the presence of an inflammatory stimulus, these cells acquire a reactive profile (the activated phenotype) that is aimed at the prevention and control of CNS damage due to altered homeostasis resulting from a wide range of insults (from impending cerebral energy failure and metabolic lesions to cell death). In the study by Jiang et al., increases in the expression of the major histocompatibility complex class II antigen marker CD11b/c (also called OX-42) were observed; this feature is characteristic of microglial activation (neuroinflammation; Fig. 1A). Microglial activation occurred early in the progression of ALF and was found to be increased further as encephalopathy and brain edema became manifest. Furthermore, the prevention of encephalopathy and brain edema by agents currently employed in clinical management, such as hypothermia and N-acetylcysteine, was accompanied by the prevention of microglial activation in all ALF animals, and this suggested that central mechanisms may contribute to the action of these treatments. Microglial activation occurs in human ALF, as shown by increased human leukocyte antigen DR (CR3/43) immunostaining (Fig. 2A).

Figure 1.

Microglial activation in animal models of ALF. (A) Increased intensity of CD11b/c (OX-42) immunolabeling that is characteristic of microglial activation in the brains of rats with ALF due to hepatic devascularization. (B) Effect of hypothermia (ALF-35) and minocycline (ALF-MINO) on the number of OX-42–positive cells [*P < 0.001 versus sham-operated controls, **P < 0.001 versus ALF rats in the coma stage of encephalopathy (ALF-COMA)].

Figure 2.

Microglial activation in patients with ALF or chronic liver failure. (A) Human leukocyte antigen DR (CR3/43) immunostaining in a patient with ALF (the patient material was provided by Dr Radhakrishan Dhiman, Post Graduate Institute of Medical Education and Research, Chandigarh, India) and in a control. (B) Neuroinflammation in a patient with HE by position emission tomography. The left image shows T1-weighted magnetic resonance imaging, and the right image shows an increased number of [11C]-R-PK11195 binding sites in the frontal lobe, particularly in the anterior cingulate cortex (ac), consistent with microglial activation and neuroinflammation in this patient (provided by Dr. Simon Taylor-Robinson, Imperial College, London, United Kingdom).

Neuroinflammation (microglial activation) has been described in a wide variety of neurological disorders, including Alzheimer's disease, multiple sclerosis, stroke, and the acquired immune deficiency syndrome–dementia complex.7 The new findings of microglial activation in HE suggest that this disorder be added to the growing list of conditions with a significant central neuroinflammatory component. ALF due to either hepatic devascularization in the rat6 or toxic liver injury in the mouse8 results in microglial activation and concomitantly increased brain concentrations of proinflammatory cytokines, including TNF-α, IL-1β, and IL-6. Care was taken by Jiang et al.6 to exclude peripheral sources of these cytokines (perfusion/fixation to remove residual blood from the brain and rigorous screening for infection/sepsis in all animals). Moreover, the expression of genes coding for TNF-α, IL-1β, and IL-6 was found to be significantly increased and to follow a comparable time course with respect to the increased brain concentrations of cytokines; this confirmed their synthesis in the brain in situ. Interestingly, microglial activation and proinflammatory cytokine synthesis in the brain during ALF occurred in the absence of neuronal cell death; this finding adds to a growing body of evidence demonstrating that neuroinflammation is not necessarily related only to neurodegeneration but may also result from potentially reversible cerebral metabolic compromise, as observed in ALF.9

Neuroinflammation in Chronic Liver Failure

Patients with cirrhosis are functionally immunosuppressed and are consequently prone to developing infections. Systemic inflammatory response syndrome (SIRS) results from the release of proinflammatory cytokines into the circulation due to liver damage and local or systemic infection.4 There is evidence that the nature and extent of both SIRS and neuroinflammation are dependent on the etiology and severity of liver injury.

A number of studies using animal models of minimal HE in the last 3 years have addressed the issue of the role of inflammation in the pathogenesis of CNS symptoms, and in some of these studies, central neuroinflammation was assessed. In a study by Cauli et al.,10 end-to-side portacaval anastomosis in the rat was found to result in increased brain concentrations of the proinflammatory cytokine IL-6 as well as increased activities of cyclooxygenase and inducible nitric oxide synthase. However, microglial activation was not assessed in these animals, and improvements in learning skills following ibuprofen administration occurred without a significant reduction in cytokine levels. In a more recent study by Brück et al.,11 locomotor activity deficits in rats with portal vein ligation were accompanied by increased expression of IL-6 messenger RNA without any evidence of microglial activation. The identity of the cell responsible for IL-6 expression was not established in that study.

In contrast to studies in animals after portal vein ligation, bile duct ligation/resection in both mice12 and rats13 results in microglial activation, which has been established with a range of cell-selective markers. Interestingly, in bile duct–ligated rats, microglial activation has been found to manifest brain regional selectivity; this finding contrasts with findings of more homogeneous localization in experimental ALF.6

Liver-Brain Proinflammatory Signaling

The nature of the signaling between the failing liver and central neuroinflammation is unknown. On the one hand, there is evidence suggesting that systemic proinflammatory mechanisms may initiate the signaling process. The onset of SIRS during ALF or chronic liver failure heralds a poor prognosis. Brain signaling in SIRS potentially occurs via one of several mechanisms: the direct transfer of cytokines by way of active transport, interactions with receptors on circumventricular organs lacking the blood-brain barrier, or the activation of afferent neurons of the vagus nerve. It has been suggested that systemic inflammatory signals have the potential to result in increased permeability of the blood-brain barrier to cytokines in those with liver disease.4 Direct evidence for this intriguing possibility, however, is not yet available.

More recently, using an animal model of biliary cirrhosis, D'Mello et al.12 demonstrated that the activation of cerebrovascular endothelial cells by peripherally administered TNF-α stimulated microglia to produce monocyte chemotactic protein 1, which mediated the recruitment of monocytes into the brain with subsequent in situ production of TNF-α. Whether these signaling mechanisms are modified by ALF or chronic liver failure has not yet been established.

Additionally, evidence suggests that toxins generated by the failing liver (other than cytokines) may also play a role in the pathogenesis of neuroinflammation. A wide range of molecules with the potential to threaten the functional integrity of the brain have the capacity to trigger the transformation of microglia from the resting state to the activated state. Such molecules include ammonia, lactate, glutamate, manganese, and neurosteroids,14 all of which have been reported to be increased in concentration in the brain during liver failure. In favor of a role for ammonia toxicity, a recent study clearly demonstrated that hyperammonemia in the absence of liver disease resulted in microglial activation that was comparable to that observed in the bile duct–ligated rat with respect to the magnitude and the regional distribution in the brain, and both hyperammonemia and bile duct ligation led to cognitive and motor impairment.13 However, studies using cultured microglial cells exposed to ammonia did not reveal any significant effect on the synthesis or release of proinflammatory cytokines,15 and this suggested that the ammonia molecule per se may not have been the entity responsible for the neuroinflammatory consequences of hyperammonemia. The exposure of cultured cells to lactate in concentrations equivalent to those described in the brain during liver failure led to several-fold increases in the release of TNF-α and IL-1β. An increased brain lactate concentration during liver failure has been attributed to an inhibitory effect of ammonia on cellular oxidative metabolism,16 and increased brain lactate synthesis significantly correlates with the severity of encephalopathy, with the presence of brain edema, and with microglial activation and cytokine production in the brain during ALF.6, 17 A single report suggests that manganese toxicity also has the potential to lead to microglial activation18: manganese deposition is a consistent feature of cirrhosis, and the deposition is greatest in basal ganglia structures of the brain.19

Diagnostic and Therapeutic Implications of Neuroinflammation

Whatever mechanism is responsible, the consistent finding of induction of central neuroinflammatory processes in patients with acute and chronic liver diseases has the potential to significantly affect diagnostic, management, and treatment options in the future. For example, the demonstration of microglial activation could stimulate the use of diagnostic neuroimaging techniques such as positron emission tomography (PET). Activated microglia express transcripts for the so-called translocator protein (previously known as the peripheral-type benzodiazepine receptor), and the extent of neuroinflammation is currently assessed in a wide range of neurological disorders, such as multiple sclerosis and the acquired immune deficiency syndrome–dementia complex, by PET with the translocator protein ligand [11C]-PK11195. Increased binding sites for this PET ligand have been reported in patients with cirrhosis and HE,20 with particularly intense signals observed in the anterior cingulate cortex, a structure known to be associated with the control of attention (Fig. 2B).

The discovery of brain inflammation and central neuroinflammatory mechanisms in patients with liver failure will undoubtedly provide new therapeutic targets. Several recent studies have assessed the beneficial effects of known anti-inflammatory agents with respect to the cerebral complications of liver failure. Significant improvements in locomotor impairment after the administration of indomethacin in portal vein–ligated rats were accompanied by the prevention of a rise in IL-6 messenger RNA.11 Ibuprofen was reported to improve learning ability10 and locomotor deficits21 in rats with portacaval shunts, but in this case, the protective effect was independent of an action on increased brain cytokine levels. Ibuprofen significantly reduced neuroinflammation in bile duct–ligated rats, inhibited microglial activation, and restored cognitive and motor function in these animals.13 In the latter study, ibuprofen was also found to normalize blood and brain ammonia levels, and this suggested that effects on systemic inflammation and improvements of hepatic function may also have contributed to the beneficial effects of ibuprofen. These disparate findings for the effects of anti-inflammatory drugs in different experimental models likely reflect differences in systemic inflammation versus neuroinflammation in these models.

Therapies directly targeting neuroinflammatory processes include those aimed at the inhibition of microglial activation as well as the inhibition of the effects of the proinflammatory cytokines resulting from microglial activation. One such example is mild hypothermia, which is increasingly being employed in the management of the cerebral complications of ALF before liver transplantation.22, 23 Hypothermia delays the onset of encephalopathy, prevents brain edema, and impairs both microglial activation (Fig. 1B) and proinflammatory cytokine production in the brain.6 A more recent study has demonstrated that TNF-α or IL-1 receptor gene deletion delays the onset of encephalopathy and attenuates brain edema in mice with ALF resulting from toxic liver injury,8 and treatment with the TNF-α receptor antagonist etanercept likewise attenuates encephalopathy severity and prevents brain edema during ALF.24 An interesting new dimension pertinent to novel therapeutics for ALF is provided by the report that minocycline, an agent with established and potent inhibitory properties25 with respect to microglial activation that are independent of its antimicrobial properties, inhibits proinflammatory cytokine production, delays the progression of encephalopathy, and attenuates brain edema in experimental ALF26 (Fig. 1B). Another interesting agent that has potent inhibitory action on microglial activation and has been found to improve cognitive function in those with neuroinflammatory disorders is the acetylcholinesterase inhibitor rivastigmine.27 The translation of these promising leads into the clinic has the potential to stimulate further research on the role of neuroinflammation and to provide novel alternative (or additional) strategies for the management and treatment of the neurological complications of liver failure in the future.