Acute liver failure and liver transplantation

• Acute liver failure (ALF) is the sudden loss of liver function in a person with no previously known liver disease: a raised international normalised ratio (INR) of 1.5 or more, together with any degree of hepatic encephalopathy (HE), plus abnormal liver tests. ^[5]
• It is rare (about 1 to 6 cases per million people per year in developed countries, about 2,000 to 3,000 a year in North America) but carries a mortality of roughly 30% in Europe and North America and higher elsewhere ^[2]. ^[3]
• The causes are broad and depend on geography: in Western countries paracetamol overdose is the leading cause; in Asia and Africa, hepatitis viruses dominate ^[3]. ^[2]
• Most deaths come from brain swelling with herniation or from multi-organ failure and sepsis, so management is critical care in a transplant centre, with frequent reassessment of who needs emergency liver transplantation (ELT).
• Transplantation is the only definitive treatment for a liver that will not recover; current 1-year survival after transplant for ALF is above 80% ^[2][3].

By the end of this page you should be able to:

1. State the clinical and laboratory definition of ALF and distinguish it from acute liver injury and from acute-on-chronic liver failure.
2. Classify the tempo of ALF (hyperacute, acute, subacute) and link tempo to typical cause and outcome.
3. Explain how massive hepatocyte death leads to hepatic encephalopathy, raised INR and intracranial hypertension.
4. Work through the initial assessment of a patient with ALF, including the tests that establish cause and severity.
5. Apply the King's College (and UK revised) criteria and the US Status 1A listing criteria to decide on transplantation, and list the contraindications.
6. Outline the critical care management of hepatic encephalopathy and intracranial hypertension, cardiovascular, renal and coagulation support, and etiology-specific treatment.
7. Summarise post-transplant outcomes and the main early risks.

Two distinctions matter. First, acute liver injury (ALI) has the same liver damage and coagulopathy (INR greater than 1.5) but no hepatic encephalopathy ^[3]. Early treatment at the ALI stage can prevent progression. Second, acute-on-chronic liver failure is acute deterioration on top of established cirrhosis or chronic liver disease; it is a different entity and its management differs ^[3]. The definition of ALF excludes cirrhosis, except for a few conditions that can first present as ALF on top of silent disease, such as acute Wilson disease and acute autoimmune hepatitis ^[2].

The international societies EASL, AASLD and APASL all use the same four cardinal elements (raised transaminases and bilirubin, INR above 1.5, hepatic encephalopathy, and no pre-existing chronic liver disease) ^[4]. The older, US-based Acute Liver Failure Study Group (ALFSG) further classifies ALF by how quickly it evolves from the onset of jaundice or symptoms to the appearance of coagulopathy and encephalopathy ^[4]:

• Hyperacute: under 7 days. Typical causes are paracetamol (acetaminophen) overdose and ischaemic hepatitis. These cases show the highest aminotransferases, the lowest bilirubin, the commonest intracranial hypertension, and the best chance of spontaneous recovery ^[2][3]. ^[4]
• Acute: 7 to 28 days. ^[4]
• Subacute: 28 days to 6 months. Subacute disease is more often due to non-paracetamol drug-induced liver injury or autoimmune hepatitis, accumulates features of chronic disease (ascites, hepatorenal syndrome) and usually needs transplantation ^[4][3].

The West Haven criteria used to grade the severity of hepatic encephalopathy (grade 1 to 4) were developed for cirrhosis, but are the practical bedside scale used in ALF ^[2].

ALF is uncommon: roughly 1 to 6 cases per million population per year in developed countries, with 2,000 to 3,000 cases per year in North America ^[3][2]. Even in well-resourced settings, overall mortality sits at about 30%; it is considerably higher in underdeveloped countries ^[2]. Globally, hepatitis viruses account for the bulk of ALF, while in Europe and North America drug-induced liver injury is now the single most common group, paracetamol being the leading drug implicated ^[2][3]. In the US ALFSG registry over two decades, paracetamol overdoses accounted for about 46% of adult ALF, non-paracetamol drug injury about 11%, hepatitis B 8% and hepatitis A 3%; herbal and dietary supplements doubled within the drug group from 10 to 20% over that period ^[2]. ^[4]

Women are affected slightly more often than men (in one national cohort, 52% women versus 48%) and the median age is in the fifth decade for most causes, except paracetamol, where the median age is 37 with 75% women ^[2][4].

The list of causes is long but the broad groups recur everywhere ^[3][2].

Risk factors are etiology-dependent. Key points: the proportion of ALF without an identified cause varies widely by region and by how thoroughly it is investigated: about 15% of adults and 50% of children in some series ^[4], about 12% in unadjudicated North American registry cases (Tujios 2022) falling to roughly 5% after expert review and additional testing ^[2], and up to 35% in worldwide series (Fernandez 2024) ^[3]; HBV reactivation during immunosuppression (including biologics) is an increasing cause in developed countries ^[2]; and many cases of apparent indeterminate ALF prove, on careful adjudication, to be paracetamol or autoimmune when testing is thorough ^[2].

Fulminant Wilson disease deserves a special diagnostic tip. Ceruloplasmin is low in half of all ALF patients whatever the cause, so waiting for it does not help; instead, fulminant Wilson's can be suspected from a very high bilirubin level with a very low alkaline phosphatase, giving a bilirubin (in mg/dL) to alkaline phosphatase (in IU/L) ratio above 2.0, plus Coombs-negative haemolytic anaemia and an alkaline phosphatase to total bilirubin ratio below 2 ^[1][2].

Everything in ALF follows from the sudden death of a large mass of hepatocytes, whatever the trigger. The liver ordinarily detoxifies ammonia, synthesises clotting factors and albumin, and regulates glucose. When most of it fails at once:

• Hyperammonaemia and intracranial hypertension. The failing liver converts ammonia to urea poorly, so serum ammonia rises; it is neurotoxic, and above roughly 100 micromol/L it is an independent risk factor for high-grade hepatic encephalopathy, while above 200 micromol/L it predicts intracranial hypertension. Below 75 micromol/L, intracranial hypertension (ICH) is rare ^[1]. Astrocyte swelling and brain oedema follow, with the hyperacute forms most prone to cerebral oedema and herniation, which accounted for about 28% of deaths on the transplant waitlist in one recent cohort ^[3]. ^[2]
• Coagulopathy. Synthesis of procoagulant and anticoagulant factors falls. The resulting prolonged INR reflects severity of liver failure; it does not by itself predict bleeding risk because coagulation is rebalanced ^[1][2].
• Systemic inflammation and multi-organ failure. Dying hepatocytes release damage-associated molecular patterns that drive a systemic inflammatory response, producing vasodilatory shock, acute kidney injury, respiratory failure, infection susceptibility and relative adrenal insufficiency ^[2].
• Hypoglycaemia. Glycogen stores empty and gluconeogenesis and glycogenolysis fail ^[2][3].

The key clinical point is that the liver's regenerative capacity can restore full function within days if the patient can be kept alive. Whether a given patient has enough regenerative capacity is the whole question that prognostic scores try to answer.

Presentations depend on tempo. Hyperacute cases have abrupt onset over hours to days, very high aminotransferases (often above 3,000 IU/L), and relatively low bilirubin; jaundice can be shallow and HE may be the first obvious sign. Subacute cases have been ill for weeks, with deep jaundice, lower transaminases, and rising bilirubin and creatinine ^[2].

Typical features:

• History: jaundice, confusion progressing to drowsiness and coma, nausea and vomiting, right upper quadrant discomfort, gastrointestinal bleeding, and symptoms of the underlying cause (prodromal viral illness, drug ingestion, pregnancy).
• Examination: HE ranging from irritability and asterixis through confusion, agitation and somnolence to coma; signs of chronic liver disease are absent initially; in subacute disease, ascites or hepatorenal syndrome may develop ^[3]. Cerebral oedema causes hypertonic posturing, unequal pupils, papilloedema and brainstem signs; clinical deterioration rather than an ICP number usually drives decisions.
• Blood pressure and perfusion: hyperdynamic circulation with peripheral vasodilatation; hypotension. A US registry reported AKI in up to 70% of patients, with 30% needing renal replacement therapy ^[3].
• Laboratory pattern: in paracetamol and ischaemic injury, ALT often above 3,000 IU/L with bilirubin around 4 mg per dL (70 micromol/L); in subacute causes, bilirubin is around 12 to 23 mg per dL with much lower ALT ^[2]. Ammonia greater than 100 micromol/L predicts high-grade HE and is itself a target of therapy ^[1][3].

• Acute liver injury (ALI) before encephalopathy has set in. It is the window for early intervention and referral.
• Acute-on-chronic liver failure (ACLF) and decompensation of previously unrecognised cirrhosis. Imaging of the liver for nodularity and signs of portal hypertension, plus history, help; even experienced radiology can over-call cirrhosis in a regenerating, massively necrotic liver ^[1].
• Other causes of coma when jaundice is mild (hypoglycaemia, sedatives, sepsis, intracranial events).
• Shock liver in heart failure: ALT rises dramatically but usually improves as perfusion is restored.
• Pre-eclampsia/HELLP versus acute fatty liver of pregnancy in the third trimester.

Baseline studies ^[2][3]:

• Core laboratory: full coagulation screen (prothrombin time, INR, fibrinogen, factor V where available), liver panel (AST, ALT, alkaline phosphatase, GGT, bilirubin), renal function, electrolytes, glucose, arterial blood gas, lactate, ammonia, lipase and amylase, CBC, LDH and creatine kinase.
• Aetiological screen (all cases): paracetamol level, urine toxicology, viral serology (HAV IgM, HBsAg with anti-HBc IgM and HBV DNA, anti-HCV and HCV RNA, anti-HEV; HSV, CMV, EBV, parvovirus PCR in the immunosuppressed), autoimmune markers (ANA, smooth muscle antibody, IgG), ultrasound with Doppler, pregnancy test, and imaging as required. In Wilson suspicion, ceruloplasmin and 24-hour urine copper (remember the ratio above); in Budd-Chiari, Doppler flows and prothrombotic screen.
• Baseline imaging: chest X-ray and abdominal ultrasound to assess hepatic and portal vein patency, liver morphology, spleen size and exclude infiltrative disease; echocardiography if unstable or being considered for transplant; abdominal CT occasionally for vascular anatomy or suspected pancreatitis ^[3].

Repeat ammonia, creatinine, INR, blood glucose and gas frequently; the decision points change by the hour.

Diagnostic criteria for ALF (all three): INR 1.5 or greater; any degree of hepatic encephalopathy; no pre-existing chronic liver disease ^[2][3].

King's College criteria (KCC) are the most widely used prognostic criteria and distinguish paracetamol from other causes ^[3][1]. Their sensitivity is reported as 68 to 69% and specificity as 82 to 92% for non-paracetamol ALF ^[1].

Two caveats are worth stating. First, the US literature generally quotes the paracetamol pH threshold as under 7.30 (Fernandez 2024 uses under 7.30) ^[3], whereas a German review using the King's criteria quotes arterial pH under 7.25 in its table (Lemmer 2023/25) ^[4]. The difference reflects source variation and revised local practice; treat the criteria as screening tools, not as arithmetic rules ^[3]. Second, the KCC were derived and validated on patients treated without transplantation; as transplantation practice changed, the criteria were revised in the UK. The UK revised emergency liver transplantation selection criteria expanded the original KCC into seven categories, including a category 4 that allows listing in paracetamol cases on clinical grounds (two of the three high-risk criteria plus clear deterioration, such as rising intracranial pressure, high oxygen needs, or escalating inotropes) even when the formal criteria are not met, and listing of some non-paracetamol cases with deep jaundice and significant coagulopathy before encephalopathy appears ^[3].

The Clichy criteria are older and use hepatic encephalopathy grade 3 or 4 with factor V under 20% of normal in patients under 30 years, or under 30% in patients over 30 years ^[4].

US Status 1A listing criteria (all three must be met): onset of encephalopathy within 8 weeks of first symptoms of liver disease; admitted to the ICU; and (a) INR greater than 2, or (b) intubated, or (c) on CRRT ^[2].

Patients with grade 2 or higher encephalopathy belong in intensive care, where advances in care have driven much of the recent mortality improvement ^[2]. The system-by-system framework below comes from recent reviews and major position papers on ICU management ^[2][3].

Brain and intracranial pressure. The first priority is to prevent cerebral oedema. Standard measures: head of bed elevated at least 30 degrees, neck neutral, minimal stimulation, control of agitation with sedation when needed, avoid hyponatraemia (maintain serum sodium at high-normal levels; targets differ between sources, 145 to 155 mEq/L in the AASLD/Tujios approach and 140 to 145 mEq/L in Fernandez), avoid hypo- and hyper-glycaemia, avoid fever, avoid over-hydration and hypercarbia ^[2][3]. Allow spontaneous hyperventilation (Fernandez 2024 targets PaCO2 34 to 41 mmHg in routine ventilation, while Tujios 2022 describes controlled hyperventilation to 30 to 40 mmHg for intracranial hypertension) and do not warm a patient with spontaneous hypothermia ^[2]. Early continuous renal replacement therapy (CRRT) is used to keep ammonia below about 100 micromol/L; modern practice treats hyperammonaemia as an indication for dialysis even without AKI ^[2][3].

For established intracranial hypertension, osmotherapy is first line: mannitol 0.5 to 1.0 g/kg (0.25 to 1.0 g/kg by some protocols) intravenously, repeatable if serum osmolality is under 320 mOsm/L, or hypertonic saline (10 to 30%) boluses ^[2][3]. CRRT should be intensified if ammonia is above 150 to 200 micromol/L (thresholds vary by source, with some using 100 to 150 micromol/L for early RRT and 150 to 200 micromol/L for intensified measures). Refractory cases (CPP under 50 mmHg, ICP above 20 mmHg) may be considered for moderate hypothermia at 33 to 34 degrees C, barbiturate coma, or deepest propofol sedation; these are desperate measures ^[2][3].

Targets during active ICH are ICP under 20 to 25 mmHg and cerebral perfusion pressure greater than 60 mmHg, with a target range of 60 to 80 mmHg, raised with norepinephrine if needed ^[2][1]. Invasive ICP monitoring is now used only selectively, usually in transplant candidates with high-risk features (SIRS, high-grade HE, ammonia above 100 micromol/L, renal failure) and clinical evidence of intracranial hypertension, because of declining incidence and bleeding risk ^[2].

Cardiovascular. Hypotension is common because of SIRS-driven vasodilation. Fluid-responsiveness is poor (around 29% respond), so be cautious with fluids and use norepinephrine as the vasopressor of choice; start a saline load first to reach a mean arterial pressure of at least 65 mmHg (higher, above 75, where cerebral perfusion is at risk) ^[1][2][3]. Vasopressin can be added for refractory hypotension. Relative adrenal insufficiency is common, so hydrocortisone (a stress dose, most often 50 mg intravenously every 6 hours (Fernandez 2024), although some sources cite 300 mg intravenously (Tujios 2022)) is reasonable in vasopressor-resistant shock ^[2][3].

Respiratory. Intubate for grade 3 or 4 HE, acute respiratory failure, or agitation in a patient being transferred. Use protective ventilation: low tidal volumes (around 6 ml/kg ideal body weight), low-to-moderate positive end-expiratory pressure, avoid both hyper- and hypo-carbia; oxygen saturation 94 to 98% ^[2][3]. Refractory respiratory failure is a contraindication to transplantation ^[3].

Renal and metabolic. Avoid nephrotoxic drugs and intravenous contrast. Early CRRT is preferred to intermittent dialysis because it minimises cardiovascular instability and cerebral oedema. Tight glycaemic control with hypertonic glucose (10 to 20%) aims for blood glucose 110 to 180 mg/dL. Enteral nutrition can be started early but should be withheld if HE or hyperammonaemia worsens. Protein intake is deferred 24 to 48 hours while ammonia is high ^[2][3].

Infection. Screen with blood, urine and respiratory cultures, chest X-ray and urine sediment on admission and whenever the patient deteriorates; prophylactic antibiotics are controversial, but empirical broad-spectrum therapy is reasonable if sepsis, SIRS, worsening HE or deterioration occurs. Daily surveillance cultures are standard in many centres ^[2][3].

Coagulation. Do not chase the INR: correct coagulopathy and platelets only for active bleeding or before an invasive procedure, using viscoelastic tests (ROTEM or TEG) where available ^[2][3]. Practical transfusion guidance from US reviews is platelets when counts are under 60 x 10^9 per L, cryoprecipitate (or fibrinogen concentrate) when fibrinogen is under 100 mg/dL, RBCs when haemoglobin is under 7 g/dL, and prophylactic plasma for invasive procedures rather than correction of the INR alone ^[2].

Gastrointestinal. Stress ulcer prophylaxis with a proton pump inhibitor; early enteral nutrition where possible ^[2][3]. Spironolactone for ascites causing respiratory compromise ^[3].

A few treatment controversies are worth their own sentences. Corticosteroids in autoimmune-hepatitis ALF are only useful before high-grade HE develops, because once grade 3 or 4 encephalopathy is established they rarely salvage; the decision whether to transplant should be settled within 3 to 7 days of starting steroids, using the evolution of HE as the strongest signal ^[2]. Acyclovir rescue in herpes ALF is uncommon. Antiviral nucleoside analogues in severe acute HBV are standard to lower the chance of infecting a graft, even though robust evidence of improved survival is lacking ^[2]. Plasma exchange is also used as an aetiology-specific option: therapeutic plasma exchange (TPE) treats acute haemolysis in Wilson disease ^[3].

High-volume (about 3 plasma volumes) therapeutic plasma exchange has become mainstream therapy in many centres. It improves transplant-free survival, removes vasoactive cytokines, helps early restoration of fibrinogen and clotting factors (hence the INR cannot be used as a prognostic marker once TPE has started), and improves haemodynamic stability ^[2][3]. Standard- and low-volume protocols are also used with fewer adverse events. Newer extracorporeal devices (MARS, Prometheus, albumin dialysis, bioartificial livers) improve some physiological parameters but have not been shown to reduce mortality in ALF outside trials, so they cannot be recommended as routine therapy ^[1][3].

Indications. ELT is the only effective treatment when spontaneous recovery will not occur. The common practical sequence is: meet ALF criteria, start best care, and list when poor-prognosis criteria (KCC or UKRC) are met or the patient is deteriorating, provided there is no contraindication and the patient's multi-organ state is compatible with surgery ^[2][3]. Higher urgency exists for the hyperacute cases, especially paracetamol, which can progress to death within hours; for subacute (indeterminate, idiosyncratic drug) cases the need for transplant is more certain and should be acted on earlier ^[2].

Contraindications to ELT include active sepsis or fungal infection, irreversible brain injury from intracranial hypertension, and overwhelming multi-organ failure; potential recipients may also have underlying comorbidity or acute abdominal complications that make surgery impossible ^[3]. Thresholds for multi-organ failure must be individualised, taking age and baseline physical condition into account; respiratory failure or high-dose vasopressor dependence is a particular concern ^[3]. Malignant infiltration of the liver is another contraindication because transplantation will not cure it ^[3].

Age. Age is a strong determinant: donor and recipient age both limit outcome, and most programmes use an upper age threshold around 40 to 50 years for transplantation in ALF, varying by centre and organ quality ^[3]. Paracetamol-intentional overdose recipients in one European registry had a 10-times higher risk of graft loss and death from non-compliance and repeated suicide compared with other causes ^[2].

Organ availability and the purpose of prognostic scores. Listing is not the same as receiving an organ. A positive prognostic score identifies a patient who will probably die without transplant, but does not identify those who will tolerate surgery or enjoy long-term graft survival. That distinction is why criteria have been revised to add clinical judgement categories, such as deterioration markers ^[3].

After ELT, the highest risks cluster in the first three months and are dominated by infection (including fungal infection), neurological complications and multisystem organ failure ^[2]. Impaired early graft function is poorly tolerated by a critically ill recipient and predisposes to sepsis ^[3]. One-year post-transplant survival in ALF is slightly lower than in chronic liver disease but comparable to critically ill, high-MELD patients, at over 80%; five-year survival is over 70% ^[3][4].

Determinants of recipient survival include recipient age (the same 40 to 50 year threshold), donor age, severity of multi-organ failure at transplant, use of a suboptimal or ABO-incompatible graft, and small graft size ^[3]. Practical post-transplant care is standard transplant ICU care: continue protective ventilation, renal support (CRRT often continued), immunosuppression as per the transplant team, infection surveillance and prophylaxis, and neurological monitoring until encephalopathy and cerebral oedema resolve; enteral nutrition is reintroduced early; glucose and electrolytes are kept in range. Living-donor liver transplantation is a possible answer to scarcity, but ethical concerns about expedited donor evaluation and coercion are particularly acute in ALF, and it accounts for only about 1% of adult ALF transplants ^[2][3]. Auxiliary (partial, removable) transplantation for ALF exists but remains controversial and is not standard ^[1].

Complications are the main cause of death and should be anticipated rather than await them.

• Cerebral oedema and intracranial hypertension (see brain section): formerly a primary cause of death; incidence has fallen, with ICH dropping from 52 to 30% and cerebral-oedema death from 12 to 4.5% between early and recent US cohorts, due largely to early CRRT and maintenance of high-normal sodium ^[2].
• Infection and sepsis: the commonest cause of early post-transplant and late pre-transplant death. Surveillance cultures, early antimicrobials, and attention to fungal infection during prolonged admission ^[2][3].
• Acute kidney injury: very common (up to 70%); usually multifactorial (SIRS, hypoperfusion, nephrotoxins). Reversible in many survivors ^[2][3].
• Cardiovascular collapse and adrenal insufficiency from SIRS and relative adrenal failure.
• Respiratory failure and ARDS: low incidence but a serious concern.
• Coagulopathy and bleeding: bleeding complications are now uncommon (11% in a large US cohort, proximate cause of death in only 2%), because of rebalanced haemostasis; over-treatment with plasma can increase thrombosis risk ^[2].
• Hypoglycaemia and metabolic disturbances.
• Gastrointestinal bleeding and ascites.

Outcome without transplantation varies sharply by cause: in the US ALFSG registry, transplant-free survival at 21 days was 65% for paracetamol, 57% for ischaemia, 51% for hepatitis A, 24% for idiosyncratic DILI, 19% for hepatitis B and 14% for autoimmune ALF ^[2]. ALF associated with pregnancy (HELLP or acute fatty liver) has a much better spontaneous outcome (around 78% transplant-free survival) ^[2]. Where cerebral oedema or multi-organ failure have developed, mortality without transplant is far higher. Outcomes have improved over two decades: inclusion of high-volume TPE, early CRRT, better ICU care, and improved post-transplant management have all contributed ^[2][3].

King's College and related criteria are good at identifying the patient who will die without a transplant (high specificity) but less good at predicting who will improve (only moderate sensitivity). MELD and newer markers such as the ALFSG index (incorporating apoptosis markers) and cell-death markers have been proposed; none has replaced King's College criteria in routine practice ^[1][3].

• Pregnancy. HELLP and acute fatty liver of pregnancy usually in the third trimester. Many respond to expedited delivery; a subset progresses to ALF and may need transplantation. Maternal outcomes have improved with delivery and ICU support ^[2][3].
• Children. Causes differ (metabolic, infectious in infants; paracetamol and autoimmune disease in older children). Paediatric ALF results require paediatric-specific criteria and teams; the West Haven scale is unreliable in young children and special thresholds apply. Surgery and, later, the need for post-transplant immunosuppression are managed by paediatric services. Standard adult texts and this page concentrate on adult ALF and adult listing criteria.
• Older adults. Age is a strong modifier of transplant benefit; adult programmes effectively cap elective transplant for ALF around 40 to 50 years, and older patients rely on meticulous critical care and etiology-directed treatment.
• Immunosuppressed hosts. HBV reactivation, herpes simplex and adenovirus can be catastrophic; screen and prophylactise HBV in anyone starting biologics or chemotherapy ^[2].

• The diagnosis is clinical plus basic testing: keep ALF in mind in any young, previously well person with jaundice and a rising INR who becomes confused. A transthoracic ultrasound and simple labs, repeated frequently, are more useful than delayed specialist tests.
• Hepatitis viruses, toxins and infections dominate causes in low-resource regions; paracetamol is far less common than in Europe and North America ^[2][3].
• The priorities are to keep the brain safe (avoid hypoglycaemia, hyponatraemia and fluid overload; nurse with head elevated), maintain haemodynamics, prevent and treat sepsis, and identify the cause early.
• Therapeutic options that still help without a transplant centre include early CRRT or simple dialysis for hyperammonaemia, osmotherapy for cerebral oedema, antimicrobial and specific antiviral therapy when available.
• Transplantation is unavailable in many settings; early referral to the nearest centre before irreversible brain injury develops is the only bridge. Outcomes are consequently worse, and honest shared decision-making with families is essential.
• Viral hepatitis vaccination programmes and safe medication storage reduce the incidence in the first place.

• ALF is a clinical syndrome, not a cause: sudden liver failure in a person without known chronic liver disease, INR 1.5 or greater plus hepatic encephalopathy.
• It is rare but deadly (about 30% mortality in well-resourced settings); transfer early to a transplant-capable ICU.
• Cause determines both tempo and outcome: paracetamol and ischaemic injury are hyperacute and more likely to recover without transplant; subacute causes usually need ELT.
• Massive hepatocyte death leads to hyperammonaemia, intracranial hypertension, coagulopathy and multi-organ SIRS; clinical deterioration matters more than any single laboratory value.
• Ammonia under 75 micromol/L makes intracranial hypertension unlikely; over 100 micromol/L predicts high-grade HE and over 200 micromol/L predicts ICH.
• Prevent cerebral oedema with head-up nursing, gentle fluids, high-normal sodium and early CRRT; first-line osmotherapy is mannitol or hypertonic saline.
• Etiology-specific treatment is urgent and, in paracetamol overdose, time-critical (NAC).
• King's College criteria are the standard screen for transplant need; US Status 1A criteria are used for listing. UK revised criteria add clinical-drift categories.
• Transplantation is the only definitive option; 1-year survival is above 80%. Contraindications include active sepsis, irreversible brain injury and overwhelming multi-organ failure.
• Post-transplant, infection, neurological complications and multi-organ failure dominate the first three months.
• Non-transplant ICU care has improved enough that no artificial liver device is yet standard.

• Hepatic encephalopathy and cerebral oedema
• Paracetamol and drug-induced liver injury
• Autoimmune hepatitis and Wilson disease (fulminant presentation)
• Liver transplantation: indications, contraindications and post-transplant care
• Coagulopathy in liver disease and rebalanced haemostasis
• Acute-on-chronic liver failure and cirrhosis decompensation