Hepatic Encephalopathy 

Updated: Jan 30, 2017
  • Author: David C Wolf, MD, FACP, FACG, AGAF, FAASLD; Chief Editor: BS Anand, MD  more...
  • Print


Hepatic encephalopathy is a syndrome observed in patients with cirrhosis. Hepatic encephalopathy is defined as a spectrum of neuropsychiatric abnormalities in patients with liver dysfunction, after exclusion of brain disease. [1, 2, 3]  Hepatic encephalopathy is characterized by personality changes, intellectual impairment, and a depressed level of consciousness. [4] An important prerequisite for the syndrome is diversion of portal blood into the systemic circulation through portosystemic collateral vessels. [5] Hepatic encephalopathy is also described in patients without cirrhosis with either spontaneous or surgically created portosystemic shunts. The development of hepatic encephalopathy is explained, to some extent, by the effect of neurotoxic substances, which occurs in the setting of cirrhosis and portal hypertension.

Subtle signs of hepatic encephalopathy are observed in nearly 70% of patients with cirrhosis. Symptoms may be debilitating in a significant number of patients. Overt hepatic encephalopathy occurs in about 30-45% of patients with cirrhosis. [6] It is observed in 24-53% of patients who undergo portosystemic shunt surgery.

The development of hepatic encephalopathy negatively impacts patient survival. The occurrence of encephalopathy severe enough to lead to hospitalization is associated with a survival probability of 42% at 1 year of follow-up and 23% at 3 years. [7]

Approximately 30% of patients dying of end-stage liver disease experience significant encephalopathy, approaching coma. [8]

The economic burden of hepatic encephalopathy is substantial. After ascites, hepatic encephalopathy is the second most common reason for hospitalization of cirrhotic patients in the United States. [9] Hepatic encephalopathy is also the most common, possibly preventable, cause for readmission. [9] The US national expenditures related to hospitalizations for hepatic encephalopathy have been estimated to range from about $1 billion per year to upwards of $7 billion per year. [10, 11] These costs may underestimate the true economic burden of hepatic encephalopathy, in terms of the condition's negative impact on the employment and finances of patients and their caregivers. [12]

Hepatic encephalopathy, accompanying the acute onset of severe hepatic synthetic dysfunction, is the hallmark of acute liver failure (ALF). Symptoms of encephalopathy in ALF are graded using the same scale used to assess encephalopathy symptoms in cirrhosis. The encephalopathy of cirrhosis and ALF share many of the same pathogenic mechanisms. However, brain edema plays a much more prominent role in ALF than in cirrhosis. The brain edema of ALF is attributed to increased permeability of the blood-brain barrier, impaired osmoregulation within the brain, and increased cerebral blood flow. The resulting brain cell swelling and brain edema are potentially fatal. In contrast, brain edema is rarely reported in patients with cirrhosis. The encephalopathy of ALF is not covered in this article but is addressed in Acute Liver Failure.

A nomenclature has been proposed for categorizing hepatic encephalopathy. [13] Type A hepatic encephalopathy describes encephalopathy associated with A cute liver failure. Type B hepatic encephalopathy describes encephalopathy associated with portal-systemic B ypass and no intrinsic hepatocellular disease. Type C hepatic encephalopathy describes encephalopathy associated with C irrhosis and portal hypertension or portal-systemic shunts. Type C hepatic encephalopathy is, in turn, subcategorized as episodic, persistent, or minimal.

For patient education resources, see Digestive Disorders Center and Infections Center, as well as Cirrhosis.



A number of theories have been proposed to explain the development of hepatic encephalopathy in patients with cirrhosis. Some investigators contend that hepatic encephalopathy is a disorder of astrocyte function. Astrocytes account for about one third of the cortical volume. They play a key role in the regulation of the blood-brain barrier. They are involved in maintaining electrolyte homeostasis and in providing nutrients and neurotransmitter precursors to neurons. They also play a role in the detoxification of a number of chemicals, including ammonia. [14]

It is theorized that neurotoxic substances, including ammonia and manganese, may gain entry into the brain in the setting of liver failure. These neurotoxic substances may then contribute to morphologic changes in astrocytes. In cirrhosis, astrocytes may undergo Alzheimer type II astrocytosis. Here, astrocytes become swollen. They may develop a large pale nucleus, a prominent nucleolus, and margination of chromatin. In ALF, astrocytes may also become swollen. The other changes of Alzheimer type II astrocytosis are not seen in ALF. But, in contrast to cirrhosis, astrocyte swelling in ALF may be so marked as to produce brain edema. This may lead to increased intracranial pressure and, potentially, brain herniation.

In the late 1990s, authors from the University of Nebraska, using epidural catheters to measure intracranial pressure (ICP), reported elevated ICP in 12 patients with advanced cirrhosis and grade 4 hepatic coma over a 6-year period. [15] Cerebral edema was reported on CT scan of the brain in 9 of the 12 patients. Several of the patients transiently responded to treatments that are typically associated with the management of cerebral edema in patients with ALF. Interventions included elevation of the head of the bed, hyperventilation, intravenous mannitol, and phenobarbital-induced coma.

In the author's opinion, patients with worsening encephalopathy should undergo head CT scan to rule out the possibility of an intracranial lesion, including hemorrhage. Certainly, cerebral edema, if discovered, should be aggressively managed. The true incidence of elevated ICP in patients with cirrhosis and severe hepatic encephalopathy remains to be determined.

Work focused on changes in gene expression in the brain has been conducted. [16] The genes coding for a wide array of transport proteins may be upregulated or downregulated in cirrhosis and ALF. As an example, the gene coding for the peripheral-type benzodiazepine receptor is upregulated in both cirrhosis and ALF. Such alterations in gene expression may ultimately result in impaired neurotransmission.

Hepatic encephalopathy may also be thought of as a disorder that is the end result of accumulated neurotoxic substances in the brain. Putative neurotoxins include short-chain fatty acids; mercaptans; false neurotransmitters, such as tyramine, octopamine, and beta-phenylethanolamines; manganese; ammonia; and gamma-aminobutyric acid (GABA).

Ammonia hypothesis

Ammonia is produced in the gastrointestinal tract by bacterial degradation of amines, amino acids, purines, and urea. Enterocytes also convert glutamine to glutamate and ammonia by the activity of glutaminase. [17]

Normally, ammonia is detoxified in the liver by conversion to urea by the Krebs-Henseleit cycle. Ammonia is also consumed in the conversion of glutamate to glutamine, a reaction that depends upon the activity of glutamine synthetase. Two factors contribute to the hyperammonemia that is seen in cirrhosis. First, there is a decrease in the mass of functioning hepatocytes, resulting in fewer opportunities for ammonia to be detoxified by the above processes. Secondly, portosystemic shunting may divert ammonia-containing blood away from the liver to the systemic circulation.

Normal skeletal muscle cells do not possess the enzymatic machinery of the urea cycle but do contain glutamine synthetase. Glutamine synthetase activity in muscle actually increases in the setting of cirrhosis and portosystemic shunting. Thus, the skeletal muscle is an important site for ammonia metabolism in cirrhosis. However, the muscle wasting that is observed in patients with advanced cirrhosis may potentiate hyperammonemia.

The kidneys express glutaminase and, to some extent, play a role in ammonia production. However, the kidneys also express glutamine synthetase and play a key role in ammonia metabolism and excretion. [17]

Brain astrocytes also possess glutamine synthetase. However, the brain is not able to increase glutamine synthetase activity in the setting of hyperammonemia. Thus, the brain remains vulnerable to the effects of hyperammonemia.

Ammonia has multiple neurotoxic effects. It can alter the transit of amino acids, water, and electrolytes across astrocytes and neurons. It can impair amino acid metabolism and energy utilization in the brain. Ammonia can also inhibit the generation of excitatory and inhibitory postsynaptic potentials. Inflammation (eg, systemic, neuroinflammation, endotoxemia) in conjunction with ammonia also appears to play a role in hepatic encephalopathy in patients with cirrhosis, which may indicate that different types of anti-inflammatory therapy be a potential therapeutic approach. [2]

Additional support for the ammonia hypothesis comes from the clinical observation that treatments that decrease blood ammonia levels can improve hepatic encephalopathy symptoms. [18]

One argument against the ammonia hypothesis is the observation that approximately 10% of patients with significant encephalopathy have normal serum ammonia levels. Furthermore, many patients with cirrhosis have elevated ammonia levels without evidence for encephalopathy. Also, ammonia does not induce the classic electroencephalographic (EEG) changes associated with hepatic encephalopathy when it is administered to patients with cirrhosis.

GABA hypothesis

GABA is a neuroinhibitory substance produced in the gastrointestinal tract. Of all brain nerve endings, 24-45% may be GABAergic. For over 20 years, it was postulated that hepatic encephalopathy was the result of increased GABAergic tone in the brain. [19] However, experimental work is changing perceptions regarding the activity of the GABA receptor complex in cirrhosis. [16, 20]

The GABA receptor complex contains binding sites for GABA, benzodiazepines, and barbiturates. It was believed that there were increased levels of GABA and endogenous benzodiazepines in plasma. These chemicals would then cross an extrapermeable blood-brain barrier. Binding of GABA and benzodiazepines to a supersensitive neuronal GABA receptor complex permitted the influx of chloride ions into the postsynaptic neurons, leading to the generation of an inhibitory postsynaptic potential.

However, experimental work has demonstrated that there is no change in brain GABA or benzodiazepine levels. Similarly, there is no change in the sensitivity of the receptors of the GABA receptor complex. [20]

Previously, it was believed that the administration of flumazenil, a benzodiazepine receptor antagonist, could improve mental function in patients with hepatic encephalopathy. It now appears that flumazenil improves mental function in only a small percentage of patients with cirrhosis.

The neuronal GABA receptor complex contains a binding site for neurosteroids. Some investigators contend that neurosteroids play a key role in hepatic encephalopathy. [1]

In experimental models, neurotoxins, like ammonia and manganese, increase the production of the peripheral-type benzodiazepine receptor (PTBR) in astrocytes. [21] PTBR, in turn, stimulates the conversion of cholesterol to pregnenolone to neurosteroids. Neurosteroids are then released from the astrocytes. They are capable of binding to their receptor within the neuronal GABA receptor complex and can increase inhibitory neurotransmission.

One study compared the levels of various chemicals in the autopsied brain tissues from patients with cirrhosis who had either died in hepatic coma or died without evidence of hepatic encephalopathy. Elevated levels of allopregnanolone, the neuroactive metabolite of pregnenolone, were found in the brain tissue of patients who died in hepatic coma. [22] Brain levels of benzodiazepine receptor ligands were not significantly elevated in patients with or without coma. This work further bolsters the role of neurosteroids in hepatic encephalopathy.

Reversibility of hepatic encephalopathy

Classically, hepatic encephalopathy was regarded as a reversible condition. Patients appeared to improve with either drug therapy (eg, lactulose or rifaximin) or liver transplantation. However, a recent study assessed cirrhotic patients who had apparently recovered from an episode of overt hepatic encephalopathy. After careful psychometric testing, it was discovered that these clinically improved patients had residual cognitive impairment compared with cirrhotic patients with either minimal hepatic encephalopathy or no encephalopathy. [23]

In 2009, Sotil et al evaluated 39 patients who had undergone liver transplantation about 1.5 years before the study. The 25 patients who had hepatic encephalopathy prior to transplantation, on the whole, performed worse on psychometric testing than the 14 patients with no history of overt encephalopathy prior to transplantation. [24]

In 2011, Garcia-Martinez et al assessed the cognitive function in 52 patients who had undergone liver transplantation. Global cognitive function after transplantation was worse in patients with a history of alcohol-induced cirrhosis, patients with diabetes, and patients with a history of hepatic encephalopathy prior to transplantation. Furthermore, the brain volume (as assessed by MRI) after transplantation was smaller in patients with a history of hepatic encephalopathy prior to transplantation than in patients with no overt encephalopathy. [25] These are provocative findings that require additional investigation.


Clinical Features of Hepatic Encephalopathy

Two broad categories of hepatic encephalopathy are covert (CHE) and overt (OHE) hepatic encephalopathy [3] ; CHE is particularly associated with poor outcomes. [3, 4]

Grading of the symptoms of hepatic encephalopathy is performed according to the so-called West Haven classification system, as follows [26] :

  • Grade 0 - Minimal hepatic encephalopathy (also known as CHE [27] and previously known subclinical hepatic encephalopathy); lack of detectable changes in personality or behavior; minimal changes in memory, concentration, intellectual function, and coordination; asterixis is absent.
  • Grade 1 - Trivial lack of awareness; shortened attention span; impaired addition or subtraction; hypersomnia, insomnia, or inversion of sleep pattern; euphoria, depression, or irritability; mild confusion; slowing of ability to perform mental tasks
  • Grade 2 - Lethargy or apathy; disorientation; inappropriate behavior; slurred speech; obvious asterixis; drowsiness, lethargy, gross deficits in ability to perform mental tasks, obvious personality changes, inappropriate behavior, and intermittent disorientation, usually regarding time
  • Grade 3 - Somnolent but can be aroused; unable to perform mental tasks; disorientation about time and place; marked confusion; amnesia; occasional fits of rage; present but incomprehensible speech
  • Grade 4 - Coma with or without response to painful stimuli

With minimal hepatic encephalopathy, patients may have normal abilities in the areas of memory, language, construction, and pure motor skills. However, patients with minimal hepatic encephalopathy demonstrate impaired complex and sustained attention. They may have delay in the choice reaction time. They may even have impaired fitness to drive. [28, 29, 30] Typically, patients with minimal hepatic encephalopathy have normal function on standard mental status testing but abnormal psychometric testing. Neurophysiologic tests in common use are the number connection test, the digit symbol test, the block design test, and tests of reaction times to light or sound (eg, critical flicker test).

Patients with grade 1 hepatic encephalopathy typically demonstrate decreased short-term memory and concentration on mental status testing. However, grade 1 hepatic encephalopathy may be difficult to diagnose. The presence of disorientation and asterixis are characteristic of grade 2 hepatic encephalopathy. [27, 31]

The borderline between covert and overt hepatic encephalopathy is being redrawn. Until recent years, the term "overt" hepatic encephalopathy was applied to patients with grades 1 through 4 encephalopathy. Now, patients with grades 0 and 1 hepatic encephalopathy are said to be "covert"; patients with grades 2 through 4 hepatic encephalopahty are said to be "overt." [27, 31]

In terms of the physical examination finding of asterixis, it must be emphasized that the flapping tremor of the extremities is also observed in patients with uremia, pulmonary insufficiency, and barbiturate toxicity.

Some patients with hepatic encephalopathy show evidence of fetor hepaticus, a sweet musty aroma of the breath believed to be secondary to the exhalation of mercaptans.

Other potential physical examination findings include hyperventilation and decreased body temperature.

Extrapyramidal symptoms—including tremor, bradykinesia, cog-wheel rigidity, and shuffling gait—have been described in patients with portosystemic shunting. [32, 33] These symptoms may or may not be associated with hyperammonemia. It is postulated that manganese deposition in the basal ganglia may predispose patients to develop these symptoms. [32] However, some patients with the "Parkinsonian phenotype of hepatic encephalopathy" may respond to treatment with rifaximin. [33]

Another neurologic condition that may be seen in the setting of portosystemic shunting is hepatic myelopathy. It is a rare condition that has been described in patients with cirrhosis of varying degrees of severity, patients who have undergone portosystemic shunt surgery or the creation of a transjugular intrahepatic portosystemic shunt (TIPS), and noncirrhotic patients with portosystemic shunts. Patients may present with lower extremity weakness, difficulty walking, spastic paraparesis, and hyperreflexia. [34] Although patients typically have concomitant hepatic encephalopathy, this is not invariable. [35] Symptoms may be rapidly progressive in some patients. Neurologic deficits do not typically respond to standard medical therapies for hepatic encephalopathy. Neurologic improvement has been described after TIPS closure [35] and after liver transplantation. [34]


Laboratory Abnormalities in Hepatic Encephalopathy

An elevated blood ammonia level is the classic laboratory abnormality reported in patients with hepatic encephalopathy. [18] This finding may aid in correctly diagnosing patients with cirrhosis who present with altered mental status. However, serial ammonia measurements are inferior to clinical assessment in gauging improvement or deterioration in a patient under therapy for hepatic encephalopathy. Checking the ammonia level in a patient with cirrhosis who does not have hepatic encephalopathy has no utility. Only arterial or free venous blood specimens must be assayed when checking the ammonia level. Blood drawn from an extremity to which a tourniquet has been applied may provide a falsely elevated ammonia level when analyzed.

Classic EEG changes associated with hepatic encephalopathy are high-amplitude low-frequency waves and triphasic waves. However, these findings are not specific for hepatic encephalopathy. When seizure activity must be ruled out, an EEG may be helpful in the initial workup of a patient with cirrhosis and altered mental status.

Visual evoked responses also demonstrate classic patterns associated with hepatic encephalopathy. However, this test is not in common clinical use.

Computed tomography (CT) and magnetic resonance imaging (MRI) studies of the brain may be important in ruling out intracranial lesions when the diagnosis of hepatic encephalopathy is in question. [36] MRI has the additional advantage of being able to demonstrate hyperintensity of the globus pallidus on T1-weighted images, a finding that is commonly described in hepatic encephalopathy. [37, 38] This finding may correlate with increased manganese deposition in this portion of the brain.


Common Precipitants of Hepatic Encephalopathy

Some patients with a history of hepatic encephalopathy may have normal mental status while under treatment. Others have chronic memory impairment in spite of medical management. Both groups of patients are subject to episodes of worsened encephalopathy. Common precipitating factors are as follows: [26]

Renal failure: Renal failure leads to decreased clearance of urea, ammonia, and other nitrogenous compounds.

Gastrointestinal bleeding: The presence of blood in the upper gastrointestinal tract results in increased ammonia and nitrogen absorption from the gut. Bleeding may predispose to kidney hypoperfusion and impaired renal function. Blood transfusions may result in mild hemolysis, with resulting elevated blood ammonia levels.

Infection: Infection may predispose to impaired renal function and to increased tissue catabolism, both of which increase blood ammonia levels.

Constipation: Constipation increases intestinal production and absorption of ammonia.

Medications: Drugs that act upon the central nervous system, such as opiates, benzodiazepines, antidepressants, and antipsychotic agents, may worsen hepatic encephalopathy.

Diuretic therapy: Decreased serum potassium levels and alkalosis may facilitate the conversion of NH4+ to NH3. At the author’s institution, diuretic-induced hypovolemia is the most common reason for patients with previously well-controlled hepatic encephalopathy to present to the emergency room with worsening mental function.

Dietary protein overload: This is an infrequent cause of hepatic encephalopathy.


Differential Diagnosis for Hepatic Encephalopathy

Distinguishing hepatic encephalopathy from other acute and chronic causes of altered mental status may be difficult in patients with cirrhosis. A decision to perform additional neurologic studies should be based on the severity of the patient's mental dysfunction, the presence of focal neurologic findings (observed infrequently in patients with hepatic encephalopathy), and the patient's responsiveness to an empiric trial with cathartic agents. Even patients with severe hepatic encephalopathy should demonstrate steady improvement in mental dysfunction after initiation of treatment with lactulose.

Differential diagnoses of encephalopathy are as follows [39] :

  • Intracranial lesions, such as subdural hematoma, intracranial bleeding, stroke, tumor, and abscess
  • Infections, such as meningitis, encephalitis, and intracranial abscess
  • Metabolic encephalopathy, such as hypoglycemia, electrolyte imbalance, anoxia, hypercarbia, and uremia
  • Hyperammonemia from other causes, such as ureterosigmoidostomy and inherited urea cycle disorders
  • Toxic encephalopathy from alcohol intake, such as acute intoxication, alcohol withdrawal, and Wernicke encephalopathy
  • Toxic encephalopathy from drugs, such as sedative hypnotics, antidepressants, antipsychotic agents, and salicylates
  • Organic brain syndrome
  • Postseizure encephalopathy

Management of Hepatic Encephalopathy

Approach Considerations

The approach to a patient with hepatic encephalopathy depends upon the severity of mental status changes and upon the certainty of the diagnosis. As an example, a patient with known cirrhosis and mild complaints of decreased concentration might be served best by an empiric trial of rifaximin or lactulose and a follow-up office visit to check its effect. However, a patient presenting to the emergency department with severe hepatic encephalopathy requires a different approach. General management recommendations include the following:

  • Exclude nonhepatic causes of altered mental function.
  • Consider checking an arterial ammonia level in the initial assessment of a hospitalized patient with cirrhosis and with impaired mental function. Ammonia levels have less use in a stable outpatient.
  • Precipitants of hepatic encephalopathy, such as hypovolemia, metabolic disturbances, gastrointestinal bleeding, infection, and constipation, should be corrected.
  • Avoid medications that depress central nervous system function, especially benzodiazepines. Patients with severe agitation and hepatic encephalopathy may receive haloperidol as a sedative. Treating patients who present with coexisting alcohol withdrawal and hepatic encephalopathy is particularly challenging. These patients may require therapy with benzodiazepines in conjunction with lactulose and other medical therapies for hepatic encephalopathy.
  • Patients with severe encephalopathy (ie, grade 3 or 4) who are at risk for aspiration should undergo prophylactic endotracheal intubation. They are optimally managed in the intensive care unit.

Most current therapies are designed to treat hyperammonemia that is a hallmark of most cases of hepatic encephalopathy.

Treatments to Decrease Intestinal Ammonia Production


In the late 19th century, it was recognized that the feeding of a high-protein diet to dogs that had undergone portosystemic shunt surgery could produce symptoms of abnormal coordination and stupor in the treated animals.

In the 20th century, low-protein diets were routinely recommended for patients with cirrhosis, in hopes of decreasing intestinal ammonia production and in preventing exacerbations of hepatic encephalopathy. An obvious consequence was the worsening of preexisting protein-energy malnutrition. Protein restriction may be appropriate in some patients immediately following a severe flare of symptoms (ie, episodic hepatic encephalopathy). However, protein restriction is rarely justified in patients with cirrhosis and persistent hepatic encephalopathy. Indeed, malnutrition is a more serious clinical problem than hepatic encephalopathy for many of these patients.

In the author's experience, it is the infrequent patient who is intolerant of a diet high in protein. Most patients with mild chronic hepatic encephalopathy tolerate more than 60-80 g of protein per day. Furthermore, one study administered a protein-rich diet (>1.2 g/kg/d) to patients with advanced disease awaiting liver transplantation, without inducing a flare of encephalopathy symptoms. [40] Another study randomized patients with severe episodic encephalopathy to either a low-protein diet or a high-protein diet, administered via nasogastric tube. [41] All patients received the same regimen of neomycin per nasogastric tube. Mental function improved at the same rate in both treatment groups. Importantly, patients receiving the low-protein diet had evidence of increased protein breakdown during the duration of the study.

Diets containing vegetable proteins appear to be better tolerated than diets rich in animal protein, especially proteins derived from red meats. This may be because of increased content of dietary fiber, a natural cathartic, and decreased levels of aromatic amino acids. Aromatic amino acids, as precursors of the false neurotransmitters tyramine and octopamine, are thought to inhibit dopaminergic neurotransmission and worsen hepatic encephalopathy.

The author recommends that patients consume well-cooked chicken and fish in addition to vegetable proteins. Malnourished patients are encouraged to add commercially available liquid nutritional supplements to their diet. Patients rarely require specialized treatment with oral or enteral supplements rich in branched-chain amino acids.

To evaluate the beneficial and harmful effects of branched-chain amino acids (BCAA) versus any control intervention for people with hepatic encephalopathy, Gluud and colleagues conducted a systematic review involving 16 randomized clinical trials that included 827 participants with hepatic encephalopathy. Primary outcomes included mortality (all cause), hepatic encephalopathy (number of people without improved manifestations of hepatic encephalopathy), and adverse events. The control groups received placebo/no intervention, diets, lactulose, or neomycin. In 15 trials, all participants had cirrhosis. Analyses showed that BCAA had a beneficial effect on hepatic encephalopathy. The authors found no effect of BCAA on mortality, quality of life, or nutritional parameters, but they recommended additional trials to evaluate these outcomes. [42]


Lactulose (beta-galactosidofructose) and lactilol (beta-galactosidosorbitol) are nonabsorbable disaccharides that have been in common clinical use since the early 1970s (the latter is not available in the United States). They are degraded by intestinal bacteria to lactic acid and other organic acids.

Lactulose appears to inhibit intestinal ammonia production by a number of mechanisms. The conversion of lactulose to lactic acid and acetic acid results in acidification of the gut lumen. [43, 44] This favors conversion of ammonia (NH3) to ammonium (NH4+); owing to the resultant relative impermeability of the membrane, the NH4+ ions are not easily absorbed, thereby remaining trapped in the colonic lumen, and there is a reduction in plasma NH3. [43, 44, 45]  Gut acidification inhibits ammoniagenic coliform bacteria, leading to increased levels of nonammoniagenic lactobacilli. [43] Lactulose also works as a cathartic, reducing colonic bacterial load.

Initial lactulose dosing is 30 mL orally, daily or twice daily. The dose may be increased as tolerated. Patients should be instructed to reduce lactulose dosing in the event of diarrhea, abdominal cramping, or bloating. Patients should take sufficient lactulose as to have 2-4 loose stools per day.

Great care must be taken when prescribing lactulose. Overdosage can result in ileus, severe diarrhea, electrolyte disturbances, and hypovolemia. Hypovolemia may be sufficiently severe as to actually induce a flare of encephalopathy symptoms.

High doses of lactulose (eg, 30 mL q2-4h) may be administered orally or by nasogastric tube to patients hospitalized with severe hepatic encephalopathy. Lactulose may be administered as an enema to patients who are comatose and unable to take the medication by mouth. The recommended dosing is 300 mL lactulose plus 700 mL water, administered as a retention enema every 4 hours as needed.

Lactulose has been the subject of dozens of clinical trials over almost 4 decades. Many small trials demonstrated the medication's efficacy in the treatment of hepatic encephalopathy. A controversial meta-analysis published in 2004 contradicted these trials and most physicians' clinical experiences. [46] When assessing high-quality randomized trials, lactulose was no more effective than placebo at improving encephalopathy symptoms. In trials comparing lactulose to an antibiotic (eg, neomycin, rifaximin), lactulose was actually inferior to antibiotic therapy.

In subsequent years, multiple randomized trials have reinvestigated the efficacy of lactulose.

In patients with minimal hepatic encephalopathy, lactulose was more effective than placebo in terms of improving patient performance on psychometric testing. [47, 48]

Lactulose was studied in large randomized trials as secondary prevention against recurrent overt encephalopathy. [49, 50] In the study by Sharma et al, patients who were recovering from hepatic encephalopahty were randomized to receive lactulose (n = 61) or placebo (n = 64). Over a median follow-up of 14 months, 12 patients (19.6%) in the lactulose group developed overt hepatic encephalopathy as compared with 30 patients (46.8%) in the placebo group (P = .001). The authors concluded that lactulose effectively prevented the recurrence of overt hepatic encephalopathy in patients with cirrhosis. [49]

Lactulose also appeared to be effective as primary prophylaxis against the development of overt hepatic encephalopathy, [51] although few physicians in the United States would advocate the use of lactulose for this indication.

An updated meta-analysis published in 2013 included these studies and affirmed the utility of lactulose in the management of hepatic encephalopathy. [52]


Neomycin and other antibiotics, such as metronidazole, oral vancomycin, paromomycin, and oral quinolones, are administered in an effort to decrease the colonic concentration of ammoniagenic bacteria. Initial neomycin dosing is 250 mg orally 2-4 times a day. Doses as high as 4000 mg/d may be administered. Neomycin is usually reserved as a second-line agent, after initiation of treatment with lactulose. Long-term treatment with this oral aminoglycoside runs the risks of inducing ototoxicity and nephrotoxicity because of some systemic absorption.

Rifaximin (Xifaxan), a nonabsorbable derivative of rifampin, has been used in Europe for more than 20 years for a wide variety of gastrointestinal indications. Multiple clinical trials have demonstrated that rifaximin at a dose of 400 mg taken orally 3 times a day was as effective as lactulose or lactitol at improving hepatic encephalopathy symptoms. [46, 53, 54] Similarly, rifaximin was as effective as neomycin and paromomycin. Rifaximin had a tolerability profile comparable to placebo. It was better tolerated than both the cathartics and the other nonabsorbable antibiotics. A potential mechanism for rifaximin's clinical activity is its effects on the metabolic function of the gut microbiota, rather than a change in the relative bacterial abundance. [55]

In 2004, rifaximin received approval by the FDA in the United States for the treatment of travelers' diarrhea. In 2005, it received orphan drug status as a treatment for hepatic encephalopathy. In March 2010, rifaximin was approved by the FDA to reduce recurrence of hepatic encephalopathy. The approval was based on a phase 3 clinical trial conducted by Bass et al. [56]

Bass et al evaluated rifaximin’s ability to reduce the risk of recurrent hepatic encephalopathy (HE). [56] In this double-blind, placebo-controlled, multinational, phase 3 clinical trial, 299 patients received either rifaximin 550 mg or placebo BID. Each group also received lactulose. Breakthrough episodes of HE occurred in 22% of patients treated with rifaximin and 46% of patients with placebo (P < 0.001). HE-related hospitalization occurred in 14% of patients treated with rifaximin and 23% of patients treated with placebo (P = 0.01).

Peripheral edema and nausea are described in some rifaximin-treated patients. There are also questions whether long-term treatment with rifaximin can induce microbial resistance. Thus far, microbial resistance has not been reported in patients using the medication. It remains unclear whether diarrhea caused by Clostridium difficile occurs at a higher rate in rifaximin-treated patients than untreated patients. In the study by Bass et al, 2 rifaximin-treated patients and no placebo-treated patients developed C difficile infection. [56]

Rifaximin was also examined in patients with minimal hepatic encephalopathy. In a large study by Sidhu et al, rifaximin was more effective than placebo in terms of improving patient performance on psychometric testing and in terms of improving health-related quality of life. [57]

Treatments to Increase Ammonia Clearance

L-ornithine L-aspartate (LOLA)

LOLA (Hepa-Merz) is available in Europe in both intravenous formulations and oral formulations. It is not available in the United States. LOLA is a stable salt of the 2 constituent amino acids. L-ornithine stimulates the urea cycle, with resulting loss of ammonia. Both l-ornithine and l-aspartate are substrates for glutamate transaminase. Their administration results in increased glutamate levels. Ammonia is subsequently used in the conversion of glutamate to glutamine by glutamine synthetase. LOLA was found to be effective in treating hepatic encephalopathy in a number of European trials. [58, 59]


Zinc deficiency is common in cirrhosis. Even in patients who are not zinc deficient, zinc administration has the potential to improve hyperammonemia by increasing the activity of ornithine transcarbamylase, an enzyme in the urea cycle. The subsequent increase in ureagenesis results in the loss of ammonia ions.

Zinc sulfate and zinc acetate have been used at a dose of 600 mg orally every day in clinical trials. Hepatic encephalopathy improved in 2 studies [60] ; there was no improvement in mental function in 2 other studies. [61]

Sodium benzoate, sodium phenylbutyrate, sodium phenylacetate, glycerol phenylbutyrate

Sodium benzoate interacts with glycine to form hippurate. The subsequent renal excretion of hippurate results in the loss of ammonia ions. Dosing of sodium benzoate at 5 g orally twice a day can effectively control hepatic encephalopathy. [62] Use of the medication is limited by the risk of salt overload and by its unpleasant taste. The medication, also used as a food preservative, is available through many specialty chemical manufacturers throughout the United States. The author has limited its use to patients with severe encephalopathy symptoms. However, in the author’s opinion, doses of sodium benzoate as low as 2.5 g orally three times per week significantly improved mental function in outpatients who had persistent encephalopathy symptoms despite cotherapy with lactulose and rifaximin.

Sodium phenylbutyrate is converted to phenylacetate. Phenylacetate, in turn, reacts with glutamine to form phenylacetylglutamine. This chemical is subsequently excreted in the urine, with the loss of ammonia ions. Sodium phenylbutyrate (Buphenyl), intravenous sodium phenylacetate in combination with sodium benzoate (Ammonul), and glycerol phenylbutyrate (Ravicti) are approved by the FDA for the treatment of hyperammonemia associated with urea cycle disorders. [63] The latter is currently in clinical trials in cirrhotic patients with hepatic encephalopathy. [64]

In a phase II trial involving 178 patients with cirrhosis (including 59 already taking rifaximin) who had experienced two or more hepatic encephalopathy (HE) events in the previous 6 months, glycerol phenylbutyrate (GPB), at a dose of 6 mL orally twice-daily, significantly reduced the proportion of patients who experienced an HE event, time to first event, and total events. [65] In addition, GPB was associated with fewer HE hospitalizations. For patients not on rifaximin at enrollment, GPB reduced the proportion of patients with an HE event, time to first event, and total events. Plasma ammonia was significantly lower in patients on GPB than in patients on placebo. Adverse events occurred in a similar proportion of patients in the GPB and placebo groups. The authors concluded that the results implicated ammonia in the pathogenesis of HE and suggested that GPB had therapeutic potential in this patient population. [65]


L-carnitine improved hepatic encephalopathy symptoms in several small studies of patients with cirrhosis. [66] Whether the medication works by improving blood ammonia levels or whether it works centrally perhaps by decreasing brain ammonia uptake remains unclear. [67]

Treatments to Improve Sleep Disturbances

Sleep disturbances are more common in patients with cirrhosis than in control subjects. Whether or not this relates to hepatic encephalopathy is unclear. A trial compared the histamine H1 blocker hydroxyzine to placebo in patients with cirrhosis and minimal hepatic encephalopathy. [68] Sleep efficiency and the patients' subjective quality of sleep improved in patients receiving hydroxyzine (25 mg) at bedtime. However, there was no accompanying improvement in cognition, as measured by neurophysiologic tests. The authors urged caution when prescribing hydroxyzine, on account of the risk of worsening encephalopathy in some patients.

Post-TIPS Hepatic Encephalopathy

Hepatic encephalopathy is seen in about 1 in 3 patients who undergo the creation of a transjugular intrahepatic portosystemic shunt (TIPS). Typically, post-TIPS encephalopathy symptoms are well controlled with the use of rifaximin or lactulose. However, post-TIPS encephalopathy symptoms can be profound in some instances. In a study by Fanelli et al, 12 of 189 patients undergoing TIPS developed encephalopathy that was refractory to conventional therapy with lactulose. These patients subsequently underwent placement of an hourglass-shaped balloon-expandable polytetrafluoroethylene (ePTFE) stent-graft inside the original shunt. Encephalopathy symptoms resolved in all of the patients over the next 18-26 hours. [69] Of course, such a procedure is not expected to improve a patient's overall condition. At the end of a mean of 74 weeks of follow-up, only 5 of the 12 patients remained alive and in good clinical condition.


Minimal Hepatic Encephalopathy

The subject of minimal hepatic encephalopathy—also known as covert hepatic encephalopathy—has attracted increasing attention. Minimal hepatic encephalopathy describes a state of low-level cognitive dysfunction that is present in as many as 70% of patients with cirrhosis. [70] It may be marked by decreased attention and executive function, as well as depressed psychomotor speed and visuomotor activity. Typically, the patient and those around the patient, including physicians, are not aware that the condition is present. Minimal hepatic encephalopathy is detected through psychometric testing (eg, the number connection test, the digit symbol test, the block design test, reaction times to light or sound, and the reaction time to interference in a task). [71, 72, 73]

Minimal hepatic encephalopathy is most likely the result of hyperammonemia. Elevated ammonia levels are detected in most patients. Similarly, the subtle neurological changes of minimal hepatic encephalopathy can be improved by the administration of lactulose. [74] Minimal hepatic encephalopathy is an important diagnostic consideration in patients with cirrhosis. It is associated with a diminished quality of life, [75] an increased risk of falls, [76] and impaired ability to operate a motor vehicle. [77, 78] For this reason, a number of authors have raised concern that the term "minimal" may trivialize the condition and have proposed that this disease stage be renamed covert encephalopathy. [27]

Two articles have noted that rifaximin can be helpful in patients with minimal hepatic encephalopathy. Sidhu et al found that rifaximin treatment led to improved health-related quality of life. [57] Bajaj et al noted that rifaximin led to decreased driving errors compared with placebo when patients with minimal hepatic encephalopathy operated a driving simulator. [79]

At this time, no guidelines address the testing and treatment of minimal hepatic encephalopathy in patients with cirrhosis. The neuropsychometric tests typically used to diagnose minimal hepatic encephalopathy can be time-consuming and cumbersome to perform in a busy physician’s office. Similarly, no consensus has been reached on how physicians should approach the issue of the fitness of patients with minimal hepatic encephalopathy to drive automobiles. [80]