Hepatic Encephalopathy, Hepatorenal Syndrome, Hepatopulmonary Syndrome, and Other Systemic Complications of Liver Disease

Pathophysiology

A number of factors, occurring alone or in combination, have been implicated in the development of HE. These factors may differ in acute and chronic liver disease and include the production of neurotoxins, altered permeability of the blood-brain barrier, and abnormal neurotransmission ( Fig. 94.1 ). The best-described neurotoxin involved in HE is ammonia, which is produced primarily in the colon, where bacteria metabolize proteins and other nitrogen-based products into ammonia. Enterocytes synthesize ammonia from glutamine. Once produced, ammonia enters the portal circulation and, under normal conditions, is metabolized and cleared by hepatocytes. In cirrhosis and portal hypertension, reduced hepatocyte function and portosystemic shunting contribute to increased circulating ammonia levels. Arterial hyperammonemia is observed in up to 90% of patients with HE, although serum levels are neither a sensitive nor specific indicator of its presence. Increased permeability of the blood-brain barrier increases the uptake and extraction of ammonia by the cerebellum and basal ganglia. Acute hyperammonemia appears to have a direct effect on brain edema, astrocyte swelling, and the transport of neuronally active compounds such as myoinositol, thereby contributing to HE. Ammonia may also directly induce an inflammatory response in astrocytes, thereby leading to swelling and cytotoxic brain edema.

Other alterations in HE affect neuronal membrane fluidity, CNS neurotransmitter expression, and neurotransmitter receptor expression and activation. ^, The γ-aminobutyric acid (GABA)-benzodiazepine system has been the best studied. Although CNS benzodiazepine levels and GABA receptor concentrations are unchanged in animal models of HE, increased sensitivity of the astrocyte (peripheral-type) benzodiazepine receptor enhances activation of the GABA-benzodiazepine system. ^, This activation occurs in part through a feed-forward system in which production of neurosteroids (allopregnanolone and tetrahydrodeoxycorticosterone) by astrocytes further activates the GABA _A -benzodiazepine receptor system. ^, Other factors that influence CNS neurotransmission, including serotonin (5-hydroxytryptamine [5-HT]), nitric oxide (NO), circulating opioid peptides, manganese, and increased oxygen-free radical production, have also been postulated to contribute to HE. ^, Manganese toxicity can cause dopaminergic dysfunction and is usually caused by oral, inhaled, or IV exposure to manganese in illegal designer narcotics. ^,

Distinct allelic mutations in the glutaminase gene increase the risk for overt HE, independent of hepatic synthetic function or the presence of minimal HE. This risk may be mediated by enhanced glutaminase transcriptional activity that results in increased levels of ammonia and glutamate. Other work has found differences in colonic mucosal microbiota in cirrhotic patients with and without HE that could influence the production of substances that lead to the development of HE. Finally, hyperammonemia, particularly in ALF, also increases astrocyte glutamine production via glutamine synthetase. The rise in astrocyte glutamine and glutamate concentrations contributes to factors associated with CNS dysfunction. ^{, ,}

Clinical Features and Classification

HE may present as a spectrum of reversible neurocognitive symptoms and signs that range from mild changes in cognition to profound coma in patients with acute or chronic liver disease. HE is often precipitated by an inciting event (e.g., GI bleeding, electrolyte abnormalities, infections, medications, dehydration). The diagnosis of HE, therefore, requires careful consideration in the appropriate clinical situation. Occasionally, HE may be the initial presentation of chronic liver disease. Subtle findings in overt HE may include forgetfulness, alterations in handwriting, difficulty with driving, and reversal of the sleep-wake cycle. ^, As HE worsens, findings may include asterixis, agitation, disinhibited behavior, seizures, and coma. Other causes of altered mental status—particularly hypoglycemia, hyponatremia, medication ingestion, and structural intracranial abnormalities resulting from coagulopathy or trauma, should be considered if focal neurologic deficits are present; otherwise, the likelihood of intracranial hemorrhage is low.

HE is classified according to 4 factors, including the underlying disease, severity of manifestations, time course, and existence of precipitating factors. There are 3 major types of HE related to the underlying disease: type A, associated with ALF; type B, associated with portosystemic shunts in the absence of liver disease; and type C, associated with chronic and end-stage liver disease and portal hypertension. Type C HE is the most common type and has historically been graded from 0 to 4 based on the West Haven criteria ( Table 94.1 ). The SONIC (spectrum of neurocognitive impairment in cirrhosis) nomenclature expands the delineation of severity of HE to reflect the wide spectrum of clinical findings and improve clinical and investigative classification. Based on the SONIC classification, cirrhotic patients are divided into (1) unimpaired, (2) covert (or minimal) HE, and (3) overt HE. Unimpaired patients have no clinical, neurophysiologic, or neuropsychometric abnormalities; patients with covert or minimal HE (clinically normal patients with abnormal cognition or neurophysiologic test results) align with grade 1 HE by the West Haven criteria; and patients with overt HE have grade 2 HE or higher by the West Haven criteria (see Table 94.1 ). This classification eliminates the need to distinguish minimal HE from grade 1 HE and takes advantage of the recognition that disorientation, specifically to time, is a distinct clinical feature that distinguishes grade 1 from grade 2 HE and covert from overt HE. ^, HE is also subdivided by time course into (1) episodic, recurrent HE (multiple episodes within a time interval of 6 months), and (2) persistent (altered behavior always present with relapses of overt HE). Finally, HE is divided into spontaneous or secondary based on the presence or absence of precipitating factors such as infections, GI bleeding, medications, and others. ^,

Diagnosis

No specific laboratory findings definitively indicate the presence of HE. Blood ammonia levels are commonly measured in patients with cirrhosis and portal hypertension but are not sensitive or specific for the presence of HE. Other factors such as GI bleeding, the ingestion of certain medications (e.g., diuretics, alcohol, narcotics, valproic acid), the use of a tourniquet when blood is drawn, and delayed processing and cooling of a blood sample may raise the blood ammonia level irrespective of the presence of HE. ^{, ,} Measurement of arterial ammonia offers no advantage over venous ammonia levels in patients with chronic liver disease. ^, Blood ammonia levels may be a useful indicator of HE in the absence of cirrhosis and portal hypertension, seen in patients with metabolic disorders that influence ammonia generation or metabolism, such as urea cycle disorders and disorders of proline metabolism ( Box 94.1 ). ^,

The development of standardized neuropsychometric and neurocognitive tests has led to the recognition that routine evaluation is insensitive for the diagnosis of clinically relevant HE. Simple tests such as the portosystemic encephalopathy syndrome test and the Stroop test (an evaluation of cognitive flexibility and psychomotor speed using either a paper-and-pencil or electronic format) evaluate the patient’s attention, concentration, fine motor skills, and orientation and have been shown to be highly specific for the diagnosis of HE. ^{, ,} With the use of these tests, covert HE has been found to be common and to negatively influence a patient’s quality of life and driving ability and increase the risk of overt HE. Moreover, treatment of minimal HE improves quality of life, cognitive test results, and driving ability. ^{, ,} A smartphone-based application (EncephalApp) is a streamlined version of the Stroop test that is validated for use in detection of covert/minimal HE.

A number of other novel imaging and functional tests for the diagnosis of HE have also been studied. Magnetic resonance (MR) spectroscopy and MR T1 mapping with partial inversion recovery (TAPIR) have been used to measure clinically relevant parameters quantitatively ^, The critical flicker frequency test, a simple light-based test that assesses cerebral cortex function, has been shown to be a reliable marker of minimal HE. Whether these functional tests will become useful in clinical practice is still unknown.

Treatment

The treatment of HE is directed primarily toward eliminating or correcting precipitating factors (e.g., bleeding, infection, hypokalemia, medications, dehydration), reducing blood ammonia levels, and avoiding the toxic effects of ammonia on the CNS. In the past, dietary protein restriction was considered an important component of the treatment of HE. Subsequent work has suggested that limiting protein-calorie intake is not beneficial in patients with HE. Vegetable and dairy proteins may be preferable to animal proteins because of a more favorable calorie-to-nitrogen ratio. Branched-chain amino acid supplementation may have a beneficial effect on HE but does not appear to affect mortality or quality of life.

Nonabsorbable disaccharides have been the cornerstone of the treatment of HE. Oral lactulose or lactitol (the latter is not available in the USA) are metabolized by colonic bacteria to by-products that appear to have beneficial effects by causing catharsis and reducing intestinal pH, thereby inhibiting ammonia absorption. These agents improve symptoms in patients with acute and chronic HE when compared with placebo but do not improve psychometric test performance or mortality. Side effects are common and include abdominal cramping, flatulence, diarrhea, and electrolyte imbalance. Lactulose may be administered per rectum (as an enema) to patients who are at increased risk of aspiration, although the efficacy of administration by enema has not been evaluated.

Oral antibiotics have also been used to treat HE, with the aim of modifying the intestinal flora and lowering stool pH to enhance the excretion of ammonia. Antibiotics are generally used as second-line agents after lactulose or in patients who are intolerant of nonabsorbable disaccharides. Rifaximin given orally in a dose of 550 mg twice daily was approved in 2010 for the treatment of chronic HE and reduction in the risk of recurrence of overt HE in patients with advanced liver disease. ^, The tolerability and side-effect profile of rifaximin are superior to those of lactulose, albeit at greater financial cost. Other antibiotics, including neomycin, metronidazole, and vancomycin, have been studied in small trials and case series, but their effectiveness in patients with chronic HE is not established.

Several other agents that may modify intestinal flora and modulate the generation or intestinal absorption of ammonia have been evaluated as potential treatments for HE. Acarbose, an intestinal α-glucosidase inhibitor used to treat type 2 diabetes mellitus, inhibits the intestinal absorption of carbohydrates and glucose and results in their enhanced delivery to the colon. As a result, the ratio of saccharolytic to proteolytic bacterial flora is increased, and blood ammonia levels are decreased. A randomized controlled double-blind crossover trial has demonstrated that acarbose improves mild HE in patients with cirrhosis and adult-onset diabetes mellitus. Similarly, probiotic regimens have been used to modify intestinal flora and diminish ammonia generation. Several studies have suggested that these agents may be beneficial in humans with mild HE. A Cochrane database review showed that probiotics probably improve recovery, overt HE, quality of life, and plasma ammonia levels, but not mortality.

Strategies to enhance ammonia clearance may also be useful in the treatment of HE. Sodium benzoate, sodium phenylbutyrate, and sodium phenylacetate, all of which increase ammonia excretion in urine, are approved by the FDA for the treatment of hyperammonemia resulting from urea cycle enzyme defects and may improve HE in patients with cirrhosis. Administration of sodium benzoate, however, results in a high sodium load, and the efficacy of this agent is not clearly established. Administration of zinc, which has been used because zinc deficiency is common in patients with cirrhosis and because zinc increases the activity of ornithine transcarbamylase, an enzyme in the urea cycle, may also improve HE; however, clear efficacy has not been established. Extracorporeal albumin dialysis using the molecular adsorbent recirculating system (MARS) has resulted in a reduction in blood ammonia levels and improvement in severe HE in patients with acute-on-chronic liver failure (see Chapters 74 and 95 ). ^, Finally, l -ornithine– l -aspartate, a salt of the amino acids ornithine and aspartic acid that activates the urea cycle and enhances ammonia clearance, has been shown in a Cochrane database review to have a possible beneficial effect on mortality, HE, and serious adverse events compared with placebo or no intervention; however, this agent is not available in the USA.

Pathophysiology

The pathophysiology of HRS is complex and incompletely characterized. Three important components contribute to the initiation and perpetuation of altered renal perfusion ( Fig. 94.2 ): (1) arterial vasodilatation in the splanchnic and systemic circulation; (2) renal vasoconstriction; and (3) cardiac dysfunction. These components influence renal function in concert and form the basis for current therapies and preventative strategies.

Clinical Features and Diagnosis

HRS is a functional disorder and, therefore, laboratory and imaging studies alone are not sufficient for making the diagnosis. A high index of clinical suspicion and exclusion of other potential causes of kidney injury are required. The majority of patients with HRS are asymptomatic, although some may report decreased urine output. AKI decreases GFR and increases the blood urea nitrogen level and may result in HE as the initial clinical presentation of HRS.

Older diagnostic criteria for HRS included an increase in the serum creatinine level by 50% above baseline to a level higher than 1.5 mg/dL (133 μmol/L). Although this definition was standardized, a subset of patients with cirrhosis and end-stage liver disease have a profound decrease in muscle mass and urea synthesis that may, in turn, result in reduced serum creatinine and blood urea nitrogen levels, thereby potentially delaying recognition of HRS. ^, Therefore, in a 2015 position paper, the International Club of Ascites (ICA) developed an updated diagnostic criteria in which new definitions of AKI were incorporated (1): cirrhosis with ascites; (2) diagnosis of AKI according to ICA-AKI criteria ( Table 94.2 ); (3) lack of response after at least 2 days of diuretic withdrawal and volume expansion with albumin (1 g/kg of body weight/day, to a maximum of 100 g/day); (4) absence of shock; (5) lack of current or recent treatment with nephrotoxic drugs; and (6) absence of parenchymal kidney disease as indicated by proteinuria of more than 500 mg/day, microhematuria (>50 red blood cells/high-power field), or abnormal renal US findings ( Box 94.2 ). Type 1 HRS, more recently designated AKI-HRS, is characterized by stage 2 or 3 AKI (see Table 94.2 ), and type 2 HRS, designated CKD (chronic kidney disease)–HRS is considered a form of CKD, with a longer, more gradual course than type 1 HRS. ^,

Several aspects of the diagnosis of HRS deserve emphasis. First, the consensus guidelines were revised to facilitate utilization of the current criteria for AKI. A key difference in cirrhotic patients with AKI, in comparison with noncirrhotic patients, is that a reduction in urine output is not considered a feature owing to the fact that patients with cirrhosis and ascites are frequently oliguric with increased sodium retention, although they may maintain a relatively normal GFR. Therefore, early recognition of even a small increase in serum creatinine is important. Second, many medications, most notably diuretics, lactulose, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and NSAIDs, may influence intravascular volume status and renal perfusion and should be discontinued expeditiously in the setting of acute renal dysfunction. Third, even though SBP may not be accompanied by obvious symptoms and signs, HRS may develop in as many as 20% of affected patients. ^, Therefore, a low threshold for evaluating cirrhotic patients with ascites for the presence of SBP is required (see Chapter 93 ). Finally, urinary biomarkers (e.g., interleukin [IL]-18, urinary neutrophil gelatinase-associated lipocalin, the Modification of Diet in Renal Disease equation [6], EGF, fatty acid binding protein 2) may be able to distinguish ATN from other etiologies of renal failure in patients with cirrhosis, but their role in clinical practice is not well defined.

Prevention and Treatment

The high mortality rate of HRS underscores the importance of prevention. Intravascular volume depletion (resulting from overdiuresis, diarrhea caused by lactulose, GI bleeding from gastroesophageal varices, or large-volume paracentesis without colloid administration) should be avoided and corrected. Similarly, nephrotoxic drugs (e.g., NSAIDs, certain antibiotics) should be avoided, and infections (SBP, bacteremia) should be treated. Specific guidelines for the primary and secondary prophylaxis of variceal bleeding, administration of colloid (albumin) to patients with a rising serum creatinine level after a large-volume paracentesis or with SBP, prophylactic administration of antibiotics to patients at high risk of SBP or other infections, and those hospitalized for GI bleeding have been published (see Chapter 92, Chapter 93 ). ^,

Routine invasive hemodynamic monitoring of cirrhotic patients with a rising serum creatinine level does not have a clear benefit and is not recommended. The concept that specific treatment of HRS is possible and may improve survival has emerged since 2000. Current options include medical therapies, TIPS placement, and LT. Medical therapies for HRS are directed toward reversing the underlying splanchnic and systemic vasodilatation with vasoconstricting agents and increasing the effective circulatory volume with the use of colloid. Such treatment is used as a temporizing measure until definitive treatment for liver disease (LT) or portal hypertension (TIPS) is undertaken, or until an acute process (SBP, GI bleeding) has been reversed ( Box 94.3 ). ^,