Liver transplantation in patients with fulminant hepatitis

Published on 09/04/2015 by admin

Filed under Surgery

Last modified 09/04/2015

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 0 (0 votes)

This article have been viewed 2145 times

Chapter 97C Liver transplantation in patients with fulminant hepatitis


Fulminant hepatic failure (FHF) is a rare, cataclysmic clinical entity resulting from acute massive necrosis of liver cells and is characterized by the rapid onset of a potentially mortal condition in a previously healthy individual. Occurrence is estimated to range between 1% and 1.2% of all jaundice cases in acute hepatitis (Bernuau et al, 1999). Depending on the etiology and severity of liver injury, some patients may undergo rapid hepatic regeneration and spontaneously recover with ad integrum restitution of the liver; however, the majority will die without orthotopic liver transplantation (OLT).

Before the advent of liver transplantation (see Chapter 97A) and advances in the intensive care management of these patients (see Chapters 72, 73, and 97B), prognosis was very poor, with mortality rates exceeding 80% (Bernuau et al, 1986; Riordan et al, 2008). At the present time, mortality rates are approximately 30% to 40%; whereas 20% to 25% of patients require liver transplantation, only 40% to 45% of patients are able to survive spontaneously (Lee et al, 2008; Ostapowicz et al, 2002; Polson, 2008). Prompt recognition and early management decisions are critical and can affect outcome. Because of their medical complexity, patients with FHF should be managed in an intensive care unit (ICU) in centers with active liver transplant programs. Referral should be done as early as possible, as it is hazardous to transfer patients later in the disease course because of increased intracranial pressure (ICP) and clinical instability. It has been suggested that transfer to a specialized ICU should be arranged when the patient’s international normalized ratio (INR) exceeds 2.0 or grade 2 hepatic encephalopathy develops (Stravitz & Kramer, 2009). High-risk patients, including those at the extreme ends of the age range (>45 years or <10 years) or with etiologies of acute liver failure (ALF) that carry a poor prognosis should be considered for transfer to a liver transplant center whenever any degree of encephalopathy develops.


Originally defined by Trey and Davidson in 1970 as an acute hepatitis complicated by ALF with hepatic encephalopathy (HE) occurring less than 8 weeks after the onset of jaundice, it is nowadays accepted that FHF is a complex clinical syndrome characterized by the development of a severe acute liver injury with impaired synthetic function and encephalopathy in a patient without previous liver disease within 26 weeks from the onset of jaundice to the development of encephalopathy. The chronologic evolution of the clinical course of ALF, estimated by the interval between onset of jaundice and encephalopathy, is predictive of clinical features and outcome (O’Grady et al, 1993). According to the length of this presentation interval, Bernuau et al (1986) has described two distinct types of FHF: fulminant (interval <2 weeks) and subfulminant (interval between 2 weeks and 3 months).

The group at King’s College Hospital (KCH) in London has defined three groups of patients: those with hyperacute liver failure (development of encephalopathy within 7 days after jaundice), acute liver failure (within 7 to 28 days), and subacute liver failure (from 28 days to 3 months) (O’Grady et al, 1993). Hyperacute presentation is more frequently associated with hepatitis A virus (HAV) or paracetamol (acetaminophen) overdose, whereas idiosyncratic drug reactions usually present a delayed-onset clinical course. Cerebral edema, a major complication of ALF, is more frequent in hyperacute liver failure, but it is rare in subfulminant disease. In contrast, manifestations of portal hypertension with ascites and renal failure are far more common in patients with subfulminant liver failure.

The cessation of liver cell destruction, rather than the occurrence of liver regeneration, appears to be the most critical variable governing outcome (Bernuau et al, 1999). A protracted course in patients with a subfulminant presentation are associated with persistence of elevated transaminases, suggesting persistence of liver cell destruction. Not surprisingly, hyperacute disease has a relatively good likelihood of spontaneous recovery compared with subacute disease, which has a very poor prognosis without liver transplantation (O’Grady et al, 1993).

To what extent these differences in survival are related to the rapidity of onset or to etiology itself is difficult to say. Likelihood of spontaneous recovery is not uniform and depends on the etiology of the insult. Previously, mortality was thought to be lower in patients with HAV and acetaminophen overdose (50% to 60%), compared with 80% to 95% mortality rate in patients with non–acetaminophen-induced ALF, halothane anesthesia, fulminant disease of indeterminate cause, and non-A, non-E fulminant hepatitis (O’Grady, 1988); however, more recent data used to evaluate prognosis of indeterminate liver failure showed no significant difference compared with other etiologies (Wei, et al, 2008).

Age younger than 10 years or older than 40 years is considered an adverse marker of outcome. Outcome in older patients was recently reevaluated by Schiodt and colleagues (2009), who observed superior overall survival in younger compared with older patients (67.9% vs. 48.2%; P < .001), yet no significant differences were demonstrated in reference to spontaneous survival among both groups, suggesting that hepatic regeneration is preserved in the elderly, a fact that is in accordance with the rise in α-fetoprotein (AFP) levels that is similar in both populations.

Other factors that have been associated with lower survival are a high body mass index (BMI), although data are still controversial (Kanda et al, 2005), and the presence of systemic inflammatory response syndrome (SIRS) (Miyake et al, 2000, 2007). The magnitude of SIRS in patients with ALF and infection correlate with mortality: 16.7%, 28.4%, 41.2%, and 64% in patients with 0, 1, 2, and 3 maximum concurrent SIRS components, respectively (Rolando et al, 2000).

ALF adversely affects most organ systems, and the majority of deaths result either from sepsis and subsequent multiorgan failure or from cerebral edema that evolves to intracranial hypertension and brainstem herniation. Patients with ALF are susceptible to a wide variety of additional complications, including renal failure, hypoglycemia, metabolic acidosis, coagulopathy, and cardiopulmonary distress. Advances in the intensive care management of patients with ALF and the early recognition of these complications has greatly contributed to a marked improvement in overall survival (see Chapters 72, 73, and 97B).


ALF is an infrequent entity that can affect patients of all ages, and it shows a predominance in females that remains unexplained. Etiologies and prognoses vary in adults compared with infants and children (Polson et al, 2005; Squires et al, 2006). ALF results from a wide variety of causes (Box 97C.1), yet relative frequency of etiologies shows geographic (Lee et al, 2008) and temporal variations. These variations depend mainly on epidemiologic, socioeconomic, and cultural factors. Worldwide, acute viral hepatitis is the predominant cause of ALF, particularly in developing countries and especially in pediatric series. Data from Sudan (Mudawi, 2007), Bangladesh (Alam et al, 2009), Mexico (Fernandez Hernandez et al, 2003), and India (Acharya et al, 1996; Dhiman et al, 1998; Jaiwal et al, 1996) report a prevalence of viral hepatitis among patients with ALF that varies from 27% to 94% and includes hepatitis A, B, D, and E and a combination of hepatitis A with hepatitis E.

In the United States and Western Europe, viral hepatitis prevalence in patients with ALF is decreasing (Taylor et al, 2006), and acetaminophen overdose has become the most common cause of fulminant hepatitis (Larson et al, 2005), resembling the experience in the United Kingdom, where it is responsible for more than 70% of ALF cases (Williams, 1996). Comparing two time periods, after 1993 in France, 28% of patients with ALF were affected with acute viral hepatitis (Feray et al, 1993) compared with 48% before 1993 (Wright et al, 1991). The United States Acute Liver Failure Study Group (USALFSG), a multicenter network established in 1997 to gather data prospectively on all forms of ALF, observed in a recent analysis of 1033 patients enrolled through 2007 that acetaminophen toxicity accounts for approximately 46% of ALF cases in the United States, followed by ALF of indeterminate origin (15%) and idiosyncratic drug reactions (12%).

The clinical pattern of presentation of acetaminophen toxicity is still different in the United States (Zimmerman et al, 1995) compared with the United Kingdom, because only half of the time does it stem from ingestion of large doses taken in an intent to commit suicide; in the remainder, ingestion results from unintentional overuse of acetaminophen-containing compounds for pain or fever relief at therapeutic doses, mainly by patients with ongoing alcohol abuse, starvation, or in those concurrently taking medications known to induce the cytochrome P450 system, such as anticonvulsants, a situation known as therapeutic misadventure (Licht et al, 1980; Seeff et al, 1986; Wootton & Lee, 1990). Although the mortality rates associated with this syndrome are lower than that observed with suicidal ingestion, they are still at least 20% in most reviews (Kumar et al, 1991; Wootton & Lee, 1990).

Idiosyncratic drug reactions are the cause of about 15% of ALF cases in the United States and other countries. Antibiotics, nonsteroidal antiinflammatory drugs (NSAIDs), and anticonvulsants constitute some of the most commonly involved classes of prescription medications involved (Chalasani et al, 2008; Ostapowicz et al, 2002; Russo et al, 2004). It is expected that enhanced awareness of potential risks and drug mechanisms of liver injury may minimize in the future the frequency of serious hepatotoxicity (Murray et al, 2008).

Despite extensive evaluation of the etiology of ALF, in approximately 15% to 40% of cases, the cause remains elusive, even though the incidence of cryptogenic disease is declining (Larson et al, 2005). Available data suggest that in 20% of these patients, ALF may be attributable to undiagnosed acetaminophen overdose (Davern et al, 2006).

In Argentina, ALF represents 9% to 21% of all annual liver transplantations, even though its frequency has been decreasing in recent years. Data from the INCUCAI, the national organ procurement institution in Argentina, reflects a completely different etiologic profile. Indeterminate disease represents 29%; this is followed by 1) fulminant autoimmune hepatitis, which is rare in the United States and Western Europe and represents 19% of our total population with ALF; 2) hepatitis B virus (HBV), which is responsible for ALF in 15%; 3) drug-induced causes, which represent 12%; and 4) ALF secondary to HAV, which accouns for 11% in Argentina. Not a single transplantation was done for acetaminophen hepatotoxicity. The high prevalence of autoimmune hepatitis remains unexplained, even though it correlates with the high percentage of patients transplanted for autoimmune disease among those with cirrhosis, suggesting a more severe course of the disease. Similar data have been reported in other Latin American countries (Uruguay [Gerona, Cohen]; Chile [Clinica Alemana]; personal communication).

Clinical Presentation

The rapid identification of ALF in a patient with liver dysfunction and an altered mental status in the setting of an acute hepatic disease is of paramount importance, as a delay in diagnosis has relevant prognostic implications. French investigators have highlighted the importance of recognizing the syndrome even at earlier stages, as soon as prothrombin ratio falls below 50% during the course of an acute liver disease and before the development of encephalopathy, as it has been recognized that prompt referral to a specialized liver unit and rapid medical intervention with specific therapies in certain etiologies can prevent progression in the clinical picture (Bernuau et al, 1993).

HE is a major defining criterion of FHF in the adult population. Children, particularly the very young, do not demonstrate classic features of encephalopathy, resulting in a revision of the definition of ALF for this specific population of patients to include pediatric patients with advanced coagulopathy, regardless of mental status (Squires, 2008). Severity of encephalopathy can be classified into four stages according to the definition of Trey and Davidson (1970): stage 1, slow consciousness; stage 2, accentuation of stage 1 and presence of asterixis; stage 3, presence of deep confusion or reactivity only to vocal stimuli; and stage 4, presence of deep coma with minimal (4a) or no response (4b) to noxious stimulus.

The development of cerebral edema is a life-threatening complication. Although precise pathogenic mechanisms remain unclear, most widely accepted theories relate to the development of astrocyte swelling and cerebral hemodynamic derangements characterized by high cerebral blood flow and failure of cerebral autoregulation in response to changes in mean arterial pressure (Larsen, 1996; Larsen et al, 2000), resulting in cerebral edema (see Chapters 72 and 97B). Loss of cerebral autoregulation of blood flow has been related to the rapidity of occurrence of electrolyte disorders in ALF, in contrast to what happens in cirrhosis or even in subacute liver failure, in which the mechanisms of osmotic regulation are still active. Cerebral edema develops in 75% to 80% of patients with grade 4 encephalopathy (Williams, 1991), and incidence is variably reported, although it appears to be progressively less frequent.

Several risk factors for the development of cerebral edema have been identified. It is more prevalent among patients with hyperacute liver failure compared with subacute disease, probably as a result of the rapid accumulation of glutamine in hyperacute presentation that overwhelms the mechanisms of compensatory expulsion of organic osmolytes from the astrocytes. High serum ammonia concentrations (>150 to 200 µmol/L) also increase the risk of cerebral edema, although the relationship between ammonia and ICP is not linear (Bernal et al, 2007; Bhatia et al, 2006; Clemmesen et al, 1999a).

Recent work suggests a central role of proinflammatory cytokines—such as tumor necrosis factor (TNF)-α and interleukin (IL)-6 and IL-1β—in patients with uncontrolled ICP, indicating activation of the inflammatory cascade in the brain (Wright, 2007). Other variables predictive of progression of intracranial hypertension and HE are the Model for End-Stage Liver Disease (MELD) score alone and in combination with ammonia levels, young age, requirement for vasopressors and renal replacement therapy, and the presence of infection and/or SIRS (Bernal et al, 2007; Vaquero et al, 2003). Evidence also emphasizes the importance of free-radical formation occurring at a mitochondrial level as the potential mediator of cellular dysfunction induced by ammonia neurotoxocity (Norenberg et al, 2009).

In severe cases, consequences of cerebral edema within the confinement of the cranial vault are the development of intracranial hypertension with potential risk of brainstem herniation and a decrease in intracerebral perfusion (Blei, 1991; see Chapter 97B). Severity of encephalopathy has prognostic implications, and overall prognosis for patients with only grade 1 or 2 encephalopathy is good, whereas for patients with grade 3 or 4 encephalopathy, prognosis is much poorer. Full recovery of cerebral function is the rule in patients whose liver function recovers, but permanent brain damage has been observed in patients making an otherwise complete recovery of liver function (O’Brien et al, 1987).


Along with neurologic involvement, coagulopathy is also a major criterion of ALF, resulting from the inadequate hepatic synthetic capacity of clotting factors. This is associated with increased consumption, particularly of factors II, V, VII, and X; reduced clearance of both activated factors and/or factor inhibitor complexes; and quantitative and qualitative platelet dysfunction (see Chapters 72 and 97B). Patients are seen with low fibrinogen and low levels of factors II, V, VII, IX, and X that result in the prolongation of prothrombin time (PT) and partial thromboplastin time (PTT) (Gazzard et al, 1975). Platelet counts are less than 100,000/mm3 in more than two thirds of patients, and platelet function is altered (O’Grady et al, 1986); however, serum concentrations of thrombopoietin do not correlate with platelet counts (Schiodt et al, 2003), and the pathogenesis of thrombocytopenia remains unclear.

Although low-grade fibrinolysis and intravascular coagulation may occur, these syndromes are hard to distinguish from the changes that result from the failure of hepatic synthesis alone (Preston, 1991). Among patients with ALF, it has been observed that coagulopathy was moderate in 81% (INR 1.5 to 5), severe in 14% (INR 5 to 10), and very severe (INR >10) in 5%. Certain etiologies are associated with greater severity of coagulopathy, and ALF in fatty liver of pregnancy represents the least severe coagulopathy (Muñoz et al, 2008).

In spite of the evidence of a bleeding diathesis in patients with ALF, spontaneous bleeding is unusual, being most common in the gastrointestinal (GI) tract (8%) or brain (<1%); yet spontaneous intracranial bleeding is exceedingly rare in the absence of an ICP monitor. The incidence of upper GI bleeding in ALF patients has been decreased by gastric acid suppression with intravenous H2 receptor antagonists (McDougall et al, 1978), therefore intravenous H2 blockers or proton pump inhibitors are recommended.

Prophylactic administration of fresh frozen plasma (FFP) is usually not recommended, because it has not been proven to influence mortality rate, it can interfere with assessment of liver function, and it may worsen cerebral edema (Caraceni et al, 1995); therefore it is only indicated in the setting of active hemorrhage or prior to invasive procedures. The goal for correction of the INR to a value of no higher than 1.5 to minimize bleeding risk arises from common practice, although it has not been adequately tested. Use of recombinant human factor VIIa has been evaluated in small pilot studies (Chuansumrit et al, 2000; Negrier et al, 2000; Shami et al, 2003), and it was associated with improvement or normalization of PT and bleeding control. It has also been suggested that the combination of recombinant human factor VIIa with FFP is superior to FFP alone in correcting coagulopathy, with a beneficial effect on morbidity and mortality (Shami et al, 2001). Larger controlled studies are still required, but it has become acceptable to administer recombinant factor VIIa (rFVIIa) when FFP has failed to correct PT-INR to an acceptable level, or when the patient has become volume overloaded, before performing invasive procedures with a high risk of bleeding, such as transjugular liver biopsy or placement of an ICP monitor. Serial determinations of INR and factor V provide useful prognostic information in the follow-up of patients with ALF.


Infections are common in patients with ALF and constitute a leading cause of mortality. Patients with ALF are at increased risk of infection and sepsis as a result of deficient opsonization resulting from reduction of factors such as fibronectin, opsonins, and chemoattractants, including components of the complement system. Additionally, infections in ALF patients present an increase in bacterial translocation, white blood cell dysfunction, Kupffer cell dysfunction, and a high requirement for invasive maneuvers. Bacteriologically proven infection was recorded in up to 80% of patients, and fungal infection was found in 32%, with a predominance of Candida. Most common sites of infection are respiratory tract (47%) and urinary tract (22%), but bacteremia with (12%) or without (15%) a defined source was present, as was bacteremia of indwelling catheters (4%).

In one prospective study of 50 patients, 80% had culture-proven infection, and in half of the remaining patients, infection was suspected, but cultures were negative (Rolando et al, 1990). gram-positive organisms, mainly streptococci and Staphylococcus aureus, predominate in ALF, suggesting that entry through the skin is more important than intestinal entry, the pathway of gram-negative organisms (Wyke et al, 1982). Leukocytosis and fever are frequently absent (<50%), but a high index of suspicion of active infection should be raised at the worsening of encephalopathy, renal function, or clinical status. Daily bacterial surveillance cultures (blood and urine) and chest radiographs are recommended, as early diagnosis of infection may improve outcome.

Aggressive treatment of presumed infection is essential, because prophylactic antibiotic regimens have shown little benefit (Rolando et al, 1993); however, a low threshold for starting appropriate antibiotics and antifungals should be maintained. Administration of antibiotics is recommended when infection or the likelihood of impending sepsis is high, surveillance cultures reveal significant isolates, clinical progression of HE or renal failure is apparent, hypotension is refractory, or SIRS is evident, individual components of which include temperature above 38° C or below 36° C, white blood count above 12,000/mm3 or below 4,000/mm3, and pulse greater than 90 beats/min (Stravitz et al, 2007).

Broad-spectrum coverage for gram-positive and gram-negative bacteria, such as with a third-generation cephalosporin, should be chosen with consideration of patient-specific isolates from surveillance cultures and historic hospital-specific isolates (Stravitz et al, 2007). Vancomycin is specifically recommended for all patients with possible intravenous catheter–related sepsis and/or risk factors for infection with methicillin-resistant Staphylococcus aureus (MRSA). An antifungal agent also is recommended in any patient who does not display prompt improvement in signs of infection after institution of antibacterial agents (Stravitz et al, 2007).

The occurrence of infection may aggravate all instances of the inflammatory mechanisms that are activated by ALF and may precipitate the occurrence of multiorgan failure and death. A finding of disseminated fungemia is predictive of a poor prognosis (Walsh et al, 1983).

Renal Disorders

Renal compromise is frequent in patients with fulminant and subfulminant hepatic failure, and oliguria is a common finding (Caraceni et al, 1995). Acute renal failure complicates 30% to 50% of cases, taking an increase in creatinine concentrations as a diagnostic criterion; yet normal creatinine concentrations do not exclude the presence of acute renal failure. Therefore the true incidence of renal compromise is probably underestimated.

Acute renal failure in ALF can result from two different situations: First, direct nephrotoxic effects result from some causes of ALF, as has been observed with toxic quantities of acetaminophen (Mazer et al, 2008) and NSAIDs and with Amanita mushroom intoxication (Escudié et al, 2007; Joshi et al, 2007), usually resulting in acute tubular necrosis. Second, patients may present with a certain degree of dehydration or may develop functional renal failure that resembles the hepatorenal syndrome of cirrhosis. Pathogenesis remains unclear, even though marked reduction in renal blood flow and glomerular filtration rate is similar to renal perfusion changes in cirrhosis, with marked renal arteriolar vasoconstriction as a result of loss of systemic vascular resistance, activation of compensatory vasoconstrictor systems, and reduced renal prostaglandin excretion; however, ALF has distinctive hemodynamic features, and only rarely is the degree of portal hypertension in ALF comparable to that of hepatorenal syndrome in cirrhotics (Navasa et al, 1992). The mean reported hepatic venous pressure gradients for patients with ALF and renal dysfunction is significantly lower compared with those in cirrhotic patients with hepatorenal syndrome (14 vs. 21 mm Hg) (Navasa et al, 1992; Ruiz-del-Arbol et al, 2005).

Recently, Leithead and colleagues (2009) demonstrated that ALF is associated with SIRS, and this association is independent of the presence of infection and severity of liver injury. Recognition of renal failure also has important clinical implications, and once established, it is usually progressive and associated with increased mortality (Jain et al, 2000; O’Grady et al, 1989; Ring-Larsen et al, 1981). Treatment options without transplantation are limited, and preventive measures are restricted to maintain adequate hemodynamics, early identification and adequate treatment of infection, and avoidance of nephrotoxic agents such as NSAIDs and aminoglycosides.

Renal replacement therapy (RRT) is usually recommended early to correct intravascular fluid overload, electrolyte disorders, or acidosis. Potential triggers for initiation of RRT might be low urine output, a rise in creatinine levels over 0.3 mg/dL over baseline, or a rise in serum ammonia above 150 µmol/L (Stravitz & Kramer, 2009).

Continuous rather than intermittent modes of renal replacement, such as continuous venovenous hemofiltration, are usually preferred to avoid abrupt shifts in solute concentrations that can result in increased ICP (Davenport et al, 1993).

Hemodynamic Complications

ALF is characterized by a hyperdynamic circulation with low systemic vascular resistance (SVR), high cardiac output, and low mean arterial pressure (MAP). Features of systemic inflammation, hemodynamic patterns, and progression of organ dysfunction resemble sepsis. Generalized systemic vasodilation is not limited to the splanchnic circulation, as it occurs in cirrhotics (Maroto et al, 1993; Ytrebø et al, 2006). Volume status in patients with ALF is difficult to assess, so placement of a pulmonary artery catheter may be of help, as management of hemodynamic balance can be complicated by the occurrence of elevated ICP and renal dysfunction. In most hypotensive patients, initial volume resuscitation is required, and it has been recommended that fluid replacement with colloids (albumin) should be preferred, rather than crystalloids (saline), and all solutions should include dextrose to maintain euglycemia (Polson et al, 2005).

Even after adequate volume repletion, vasopressors may be required to maintain a MAP of at least 50 to 60 mm Hg and a cerebral perfusion pressure (CPP) of 60 to 80 mm Hg (Polson et al, 2005). Vasopressor support with epinephrine, norepinephrine, or dopamine is usually instituted to maintain adequate perfusion of vital organs.

Hepatectomy has been advocated based on anecdotal references as a terminal salvage therapy in patients with severe refractory circulatory dysfunction. It has also been suggested that adrenal dysfunction might be an additional contributing factor to circulatory collapse; however, the benefits of supplemental corticosteroids in these patients are still unknown (Harry et al, 2002).

Pulmonary Complications

Pulmonary compromise is usually present in 30% of patients with ALF. Impaired gas exchange leading to hypoxemia and mixed acidosis may favor progression of HE (Kramer et al, 1991; Larsen et al, 2000), and mechanical ventilation may be required to ensure oxygenation and to secure the airway in patients who progress to grade 3 encephalopathy. Caution should be maintained, as it has been observed that positive end-expiratory pressure (PEEP) can worsen cerebral edema. In patients with acetaminophen intoxication, acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) can occur in one third of patients with ALF and can cause intractable hypoxemia that may contribute to death (Baudouin et al, 1995).

Metabolic Disturbances

Common metabolic abnormalities in FHF include acid-base and electrolyte disorders, hypophosphatemia, and hypoglycemia. Among acid-base disorders, hypocapnia with mixed alkylosis is frequent in the early stages of FHF in more than 50% of patients (O’Grady et al, 1993), yet as ALF progresses, patients typically evolve to metabolic acidosis with respiratory alkalosis. At the same time, ALF is a catabolic state characterized by negative nitrogen balance and increased resting energy expenditure. It is recommended that metabolic homeostasis be carefully maintained, and overall nutritional status—as well as phosphate, glucose, potassium, and magnesium levels—must be frequently monitored (Polson et al, 2005). Enteral nutrition should be administered early whenever possible, avoiding excessive free-water provision that can result in hypoosmolality, which aggravates cerebral edema. No firm evidence proves that branched-chain amino acids are superior to other enteral preparations (O’Grady et al, 1989; Naylor et al, 1989). If enteral feeding is contraindicated, parenteral nutrition remains an option, but severe protein restrictions should be avoided.


As a result of its high mortality, heterogeneity, and low frequency, FHF is extremely difficult to study, randomized controlled trials are limited (Kaidu, 2008), and optimal management remains poorly defined and center specific. Management should be centered in three main areas: 1) definition of etiology so that specific treatment can be provided in selected cases; 2) prevention, recognition, and early treatment of complications that could potentially lead to multiorgan failure and death; and 3) assessment of prognosis for distinguishing patients who have the potential to improve spontaneously versus those who will require liver transplantation and who should be urgently included on the waiting list (see Chapters 72, 73, and 97B).

Specific treatments are limited and are reserved exclusively for a few etiologies, and any available treatment should be delivered as early as possible (Box 97C.2). Some treatments have been systematically studied only for acetaminophen overdose. The role of N-acetylcysteine (NAC) in limiting liver injury through repletion of hepatic glutathione (Mitchell et al, 1974), improving prognosis, has been shown to be effective and safe in large case series (Harrison et al, 1990; Smilkstein et al, 1988, 1991) and in a small, controlled trial (Keays et al, 1991); firm evidence supports the utility of NAC administration, even when there is doubt concerning timing, dose ingested, or plasma concentration of acetaminophen, and NAC should not be withheld, even if ingestion was 48 to 72 hours before the patient sought medical attention. In acetaminophen overdose, oral NAC is recommended as first-line therapy in patients with mild or no encephalopathy, whereas intravenous NAC should be reserved for patients with advanced encephalopathy, hypotension, or some other reason to suggest that oral dosing might not be tolerated, such as vomiting or ileus. The length of NAC administration should be determined by clinical improvement—resolution of HE, improving coagulopathy, and declining transaminases—or outcome (death or liver transplantation) rather than by time or serum acetaminophen levels.

In addition to its role in limiting liver injury, NAC also has beneficial effects on systemic hemodynamic parameters and oxygen delivery to peripheral tissues (Harrison et al, 1991, 1996). Consequently, it has also been evaluated as a treatment for other forms of ALF unrelated to acetaminophen overdose. A recent study evaluated 173 patients with non-acetaminophen ALF, who were randomly allocated to receive either intravenous NAC or placebo; no benefit was demonstrated on 21-day survival, although transplant-free survival and 1-year posttransplant survival were improved. Furthermore, in the subgroup of patients with grade 1 or 2 HE, survival at 1 year was significantly better in those who received NAC than in all other individuals (Lee et al, 2009).

Even though there is still insufficient data to recommend specific therapies for ALF resulting from other etiologies, with the exception of prompt delivery in women with acute fatty liver of pregnancy or HELLP (hemolysis, elevated liver enzymes, and low platelet count) syndrome, several etiology-specific measures are recommended based upon experience, relative innocuousness of the measures, and the high mortality of the clinical syndrome. Specific measures have been suggested in ALF secondary to autoimmune hepatitis, HBV, and herpesvirus infection; ingestion of Amanita species; ischemia; and fatty liver of pregnancy.