Skip to main content
MedVellum
MCQsExamsAtlas
DashboardPricing
MBBS / Core medicine✳Dermatology✳ICU Fellowship (CICM)✳Anaesthesia✳Emergency Medicine✳Psychiatry Fellowship✳Paediatrics Fellowship✳Physician Medicine✳MCQs✳SAQs✳Vivas✳OSCE✳Evidence-first✳MBBS / Core medicine✳Dermatology✳ICU Fellowship (CICM)✳Anaesthesia✳Emergency Medicine✳Psychiatry Fellowship✳Paediatrics Fellowship✳Physician Medicine✳MCQs✳SAQs✳Vivas✳OSCE✳Evidence-first✳

MedVellum.

The folio

Exam-exhaustive medical education across every specialty — evidence-graded topics, engraved plates, and practice in every written and oral format. Educational content only — not medical advice.

llms.txt · psychiatry LLM catalog · sitemap

Atlas

  • Specialty atlas
  • MBBS / Core medicine
  • Dermatology
  • ICU Fellowship (CICM)
  • Anaesthesia
  • Emergency Medicine
  • Psychiatry Fellowship
  • Paediatrics Fellowship
  • Physician Medicine

Study & account

  • MCQ practice
  • Practice alias
  • Exam tools
  • Dashboard
  • Pricing
  • Sign in

© 2026 MedVellum. For education only — not a substitute for clinical judgement.

Folio edition · Set in Instrument Serif & Archivo

ICU TopicsGI/Nutrition

ICU · GI/Nutrition

Acute liver failure (hepatic failure)

Also known as Acute liver failure (ALF) · Fulminant hepatic failure · King's College Criteria · N-acetylcysteine (NAC) · Liver transplant for ALF · Cerebral oedema in ALF · O'Grady classification · Hyperacute / acute / subacute liver failure

Acute liver failure (ALF) is severe acute liver injury with encephalopathy and coagulopathy (INR >1.5) within 26 weeks, in a patient without pre-existing liver disease. Classification by tempo (O'Grady): hyperacute (<7 days), acute (8-28 days), subacute (29 days-12 weeks) — prognosis varies by interval. Causes: paracetamol (1 in UK/Aus), drug-induced (idiosyncratic), viral hepatitis (HBV 1 viral), Wilson disease, pregnancy-related (AFLP, HELLP), mushroom (Amanita phalloides), ischaemic, autoimmune, indeterminate. Pathophysiology of cerebral oedema: ammonia crosses BBB - astrocytic glutamine synthetase - glutamine accumulation - astrocyte swelling (cytotoxic oedema) - raised ICP - tonsillar herniation (1 cause of death). Management: N-acetylcysteine (NAC) for ALL causes (Lee 2009 showed transplant-free survival benefit beyond paracetamol), supportive care for complications, and urgent liver transplant assessment using King's College Criteria. Lactate is a key prognostic marker (early >3.5 mmol/L or persistent elevation predicts poor outcome). Do NOT routinely correct INR with FFP — it is a liver function marker, not a bleeding-risk indicator (rebalanced haemostasis).

high5 referencesUpdated 2 July 2026
On this page & tools

Your progress

Saved locally on this device.

Practise this topic

18 MCQs with explanations

Target exams

CICMFFICMEDIC

Red flags

NAC for ALL acute liver failure — not just paracetamol (improves transplant-free survival, Lee 2009)Cerebral oedema is the #1 cause of death in ALF — monitor ICP if Grade III-IV encephalopathyDo NOT routinely correct INR with FFP — INR is a liver function marker, not bleeding risk (rebalanced haemostasis)King's College Criteria determine transplant listing — urgent referral; do NOT wait for maximal deteriorationArterial lactate &gt;3.5 mmol/L early, or rising lactate, predicts poor outcome — flags transplant urgencyHyperacute ALF (paracetamol) has the BEST transplant-free survival; subacute (seronegative, idiosyncratic DILI) the WORSTHypoglycaemia is common and worsens brain injury — check glucose hourly, 10% dextrose infusionCRRT preferred over intermittent dialysis (continuous avoids solute shifts worsening cerebral oedema)

Your progress

Saved locally on this device.

Practise this topic

18 MCQs with explanations

Target exams

CICMFFICMEDIC

Red flags

NAC for ALL acute liver failure — not just paracetamol (improves transplant-free survival, Lee 2009)Cerebral oedema is the #1 cause of death in ALF — monitor ICP if Grade III-IV encephalopathyDo NOT routinely correct INR with FFP — INR is a liver function marker, not bleeding risk (rebalanced haemostasis)King's College Criteria determine transplant listing — urgent referral; do NOT wait for maximal deteriorationArterial lactate &gt;3.5 mmol/L early, or rising lactate, predicts poor outcome — flags transplant urgencyHyperacute ALF (paracetamol) has the BEST transplant-free survival; subacute (seronegative, idiosyncratic DILI) the WORSTHypoglycaemia is common and worsens brain injury — check glucose hourly, 10% dextrose infusionCRRT preferred over intermittent dialysis (continuous avoids solute shifts worsening cerebral oedema)

In one line

ALF = severe liver injury + encephalopathy + coagulopathy (INR >1.5) within 26 weeks, no pre-existing liver disease. Classified by tempo (O'Grady): hyperacute (<7 d, e.g. paracetamol — best prognosis), acute (8-28 d), subacute (29 d-12 wk, e.g. idiosyncratic DILI, seronegative — worst prognosis). Causes: paracetamol (#1 UK/Aus), drugs, viral (HBV), Wilson, pregnancy (AFLP/HELLP), mushroom (Amanita). Cerebral oedema pathophysiology: ammonia -> astrocyte glutamine synthetase -> glutamine -> astrocyte swelling -> cytotoxic oedema -> raised ICP -> herniation (#1 cause of death). NAC for ALL causes (Lee 2009 — transplant-free survival benefit beyond paracetamol). King's College Criteria determine transplant listing (paracetamol: pH <7.25 OR all of INR >6.5 + Cr >300 + grade III-IV HE; non-paracetamol: INR >6.5 OR >=3 of 5 criteria). Lactate is a prognostic marker (>3.5 mmol/L early = poor). Do NOT routinely correct INR (liver function marker). Support: hypoglycaemia (10% dextrose), AKI (CRRT), infection (surveillance), coagulopathy (vit K, FFP only if bleeding).

[1]
Pathophysiology of cerebral oedema in acute liver failure: ammonia crosses the blood-brain barrier into astrocytes, is converted to glutamine by glutamine synthetase, causing astrocyte swelling, cytotoxic oedema and raised intracranial pressure
FigureThe ammonia-glutamine axis in acute liver failure. Ammonia crosses the blood-brain barrier and is metabolised by astrocytic glutamine synthetase to glutamine. Osmotic glutamine accumulation swells astrocytes (cytotoxic oedema), which — amplified by systemic inflammation — generates intracranial hypertension and the uncal/tonsillar herniation that remains the leading cause of death in ALF.

Definition

Acute liver failure (ALF) is the onset of coagulopathy (INR >1.5) and any degree of hepatic encephalopathy within 26 weeks of the first symptoms of liver injury, in a patient without pre-existing cirrhosis or liver disease. The 26-week ceiling separates ALF from acute-on-chronic liver failure (ACLF), which has a distinct management pathway.[1]

The older term "fulminant hepatic failure" (Trey & Davidson, 1970) required encephalopathy within 8 weeks of the first symptom; "late-onset" or "subfulminant" referred to cases developing encephalopathy later. This binary has been superseded by the O'Grady tempo-based classification (below), which carries prognostic meaning. [1]

Practice point

Why aetiology, not the label, drives the plan

The diagnostic workup must run in parallel with resuscitation: a paracetamol level, viral serology (HBV surface antigen/anti-HBc IgM, HAV IgM, HEV IgM, HSV/EBV/CMV PCR in the immunocompromised), ceruloplasmin and slit-lamp (Wilson), autoimmune markers (ANA, ASMA, IgG), urine drug screen, and a careful history of herbal/medicinal products. Aetiology changes both the prognosis (paracetamol and ischaemic do comparatively well; seronegative and idiosyncratic DILI poorly) and the specific therapy (NAC, silibinin, chelation, delivery of the fetus).

[1]

Classification by tempo — the O'Grady system

The interval between jaundice and encephalopathy defines three syndromes with distinct aetiologies and outcomes.[3]

O'Grady tempo classification — interval from jaundice to encephalopathy

29 days - 12 weeks

Mortality Worst without transplant

Encephalopathy 5-12 weeks after jaundice. Typical causes: seronegative ("non-A non-B") hepatitis, idiosyncratic drug-induced liver injury (DILI), some Wilson presentations. Carries the WORST transplant-free survival (often <20%) — cerebral oedema is less common but multi-organ failure and portal hypertension dominate. Urgent transplant is frequently the only option.

[3]

Hyperacute (&lt;7 d)

Best prognosis

  • Paracetamol overdose — the archetype
  • Ischaemic hepatitis / "shock liver"
  • Amanita phalloides mushroom poisoning
  • Cerebral oedema: COMMON and early (the dominant threat)
  • Transplant-free survival ~50-60% (highest of the three)

Acute (8-28 d)

Intermediate

  • Viral hepatitis — HBV (most common viral cause), HAV, HEV
  • Some idiosyncratic drug reactions
  • Autoimmune hepatitis (acute presentation)
  • Cerebral oedema: moderate risk
  • Transplant-free survival ~30-40%

Subacute (29 d - 12 wk)

Worst prognosis

  • Seronegative ("non-A non-B") hepatitis — often indeterminate
  • Idiosyncratic drug-induced liver injury (DILI)
  • Wilson disease (acute decompensation)
  • Cerebral oedema: LESS common; portal hypertension and renal failure dominate
  • Transplant-free survival often <20% — transplant is frequently the only exit
[3]

The counter-intuitive prognostic paradox

The faster the onset of encephalopathy, the better the transplant-free survival. Hyperacute ALF (paracetamol) looks the sickest yet has the best chance of spontaneous recovery, because the injury is often a single, self-limited insult and the liver regenerates once the patient is bridged through the cerebral-oedema window. Subacute ALF looks chronically ill yet has the worst outcome, because ongoing injury outstrips regeneration. This is why aetiology and tempo, not the INR in isolation, frame the transplant discussion.

[1]

Classification at the bedside — apply on admission

1

Confirm ALF

Acute liver injury (AST/ALT markedly elevated) + coagulopathy INR >1.5 + encephalopathy (any grade), within 26 weeks, AND no known pre-existing liver disease. If chronic liver disease/cirrhosis is present, re-classify as ACLF — different pathway.

2

Record the tempo

Date of first symptom (often jaundice or nausea) -> date of encephalopathy. <7 d = hyperacute; 8-28 d = acute; 29 d-12 wk = subacute. The tempo refines prognosis and frames how aggressively to push transplant assessment.

3

Identify aetiology

Paracetamol level, viral serology, ceruloplasmin + slit-lamp, autoimmune panel, pregnancy test, drug/herbal history, Amanita exposure (delayed GI symptoms then liver failure). Aetiology determines specific therapy AND prognosis.

4

Stratify for transplant

Apply King's College Criteria on admission and reapply serially. Add lactate and arterial ammonia to the prognostic picture. Involve the transplant centre EARLY — the window between meeting criteria and brain death from cerebral oedema can be hours.

[1]

Causes

Common causes

West / UK / Aus

  • Paracetamol (acetaminophen) overdose — #1 cause in UK, Australia, US (~40-50% of ALF). Both deliberate and therapeutic misadventure (staggered dosing in pain or malnutrition).
  • Drug-induced (idiosyncratic) — isoniazid, valproate, halothane, nitrofurantoin, phenytoin, statins, herbal/dietary supplements (e.g. kava, green-tea extract, Polygonum multiflorum), recreational MDMA
  • Viral hepatitis — HBV (#1 viral cause worldwide; acute infection or reactivation with immunosuppression/chemotherapy), HAV, HEV (endemic and zoonotic), HSV (immunocompromised/pregnant), EBV, CMV
  • Ischaemic hepatitis ("shock liver") — profound hypotension, cardiac arrest, severe heart failure. Often recovers with haemodynamic support.
  • Autoimmune hepatitis (acute presentation)
  • Indeterminate (~20%) — "seronegative hepatitis"; common in subacute ALF and carries a poor prognosis

Less common / specific

Selected populations

  • Mushroom poisoning (Amanita phalloides) — delayed GI symptoms (6-24 h) then liver failure at 48-72 h. Classically after foraging. Treat with silibinin + NAC + aggressive fluid.
  • Wilson disease — young patient, ALF + Coombs-negative haemolysis + low ceruloplasmin + Kayser-Fleischer rings + markedly low alkaline phosphatase relative to bilirubin. Always fatal without transplant.
  • Pregnancy-related: acute fatty liver of pregnancy (AFLP) and HELLP — third trimester; delivery is the definitive treatment.
  • Heat stroke / malignant hyperthermia
  • Malignant infiltration (lymphoma, melanoma, breast) — hepatomegaly, imaging
  • Budd-Chiari syndrome (hepatic venous outflow obstruction)
  • Veno-occlusive disease / sinusoidal obstruction syndrome (post-chemotherapy, haematopoietic stem cell transplant)
[1] [2]

Aetiology and prognosis

Transplant-free survival

  • Paracetamol — best (~50-60% spontaneous survival)
  • Ischaemic hepatitis — good with haemodynamic correction
  • Pregnancy-related (delivery) — good once delivered
  • HAV — relatively good
  • HBV — moderate (~25-35%)
  • Drug-induced (idiosyncratic) — poor (~20-30%)
  • Autoimmune — moderate with steroids, often needs transplant
  • Seronegative / indeterminate — poor (~15-25%)
  • Wilson disease — near-100% fatal without transplant; urgent listing
  • Amanita — variable; silibinin + NAC improve outcomes

Aetiology-specific therapy

Beyond supportive care

  • Paracetamol — NAC (regardless of time since overdose if any hepatotoxicity)
  • Amanita — silibinin (milk thistle) + NAC + aggressive crystalloid diuresis
  • HBV — nucleos(t)ide analogue (entecavir/tenofovir)
  • HSV — IV aciclovir
  • Autoimmune — corticosteroids (prednisolone 40-60 mg)
  • Wilson — chelation (penicillamine) usually futile in acute crisis; TRANSPLANT
  • Pregnancy (AFLP/HELLP) — URGENT DELIVERY of the fetus
  • Budd-Chiari — anticoagulation, TIPS, transplant
  • Drug-induced — STOP the offending agent immediately
[1]

Paracetamol — the leading cause in the West

Paracetamol (acetaminophen) hepatotoxicity is the single largest cause of ALF in the UK, Australia and the US. The toxic metabolite NAPQI (formed by CYP2E1) depletes glutathione and binds hepatocyte proteins causing centrilobular (zone 3) necrosis. Risk is amplified by fasting, malnutrition, alcohol, and concurrent CYP2E1 inducers. [1]

Paracetamol overdose — diagnostic and management cascade

1

Establish timing

Single ingestion time is essential for the Rumack-Matthew nomogram. If staggered or unknown ingestion time, the nomogram is NOT valid — treat with NAC if any evidence of hepatotoxicity (ALT rising or detectable paracetamol level).

2

Rumack-Matthew nomogram

Plots serum paracetamol level (y-axis) against time post-ingestion (x-axis, from 4 to 24 h). Level ABOVE the treatment line ("200" line at 4 h in some units, "150"/100 in others) -> give NAC. Use the local nomogram. NOT applicable after 24 h or in staggered overdose.

3

Give NAC

150 mg/kg IV over 1 h (loading), then 50 mg/kg over 4 h, then 100 mg/kg over 16 h. Newer 12 h and 20 h protocols exist. Continue beyond the standard course if ALT rising, INR >2, or paracetamol still detectable. Anaphylactoid reactions (rash, bronchospasm) are common in the loading dose — slow the infusion and treat with antihistamine; do NOT stop NAC.

4

Monitor for King's College Criteria

Check pH, INR, creatinine, lactate, and grade of encephalopathy at baseline and serially (4-12 h). Any deterioration toward KCC triggers urgent transplant referral.

[1]
Myth-buster

Myth: NAC only works if given within 8-24 hours

False. While earlier is better, NAC has mortality benefit even when started >24 h after paracetamol ingestion in patients who have already developed hepatotoxicity — it improves haemodynamics, tissue oxygen delivery, and cerebral oedema. Give NAC to any paracetamol patient with evidence of liver injury regardless of the time since ingestion. The 8-24 h nomogram window governs the decision to start in the asymptomatic patient, not whether to continue in the established hepatotoxicity patient.

[1]

Specific aetiologies — high-yield patterns

Wilson disease presenting as ALF

Young patient (typically <30) with Coombs-negative haemolytic anaemia + acute liver failure + low ceruloplasmin + Kayser-Fleischer rings on slit-lamp + disproportionately low alkaline phosphatase (ALP/bilirubin ratio <2) + high urinary copper. The haemolysis (from copper release) drives a high bilirubin with only modest transaminitis. Acute Wilsonian crisis is virtually always fatal without transplant — refer urgently. Chelation (penicillamine) is rarely effective once ALF is established; it is a bridge at best.

[1]

Amanita phalloides mushroom poisoning

Biphasic illness: (1) delayed GI phase 6-24 h after ingestion (severe vomiting, profuse watery diarrhoea, dehydration — leads to misdiagnosis as gastroenteritis); (2) hepatic phase at 48-72 h (massive transaminitis then ALF). The toxin alpha-amanitin inhibits RNA polymerase II. Treatment: silibinin (milk-thistle extract, the specific antagonist — IV where available), NAC, aggressive crystalloid resuscitation with forced diuresis to maintain urinary toxin clearance, and supportive ICU care. A history of foraging or a shared meal is the key clue.

[1]

Pregnancy-related ALF — delivery is the cure

Acute fatty liver of pregnancy (AFLP) and HELLP (Haemolysis, Elevated Liver enzymes, Low Platelets) occur in the third trimester or immediately postpartum. AFLP: microvesicular steatosis (long-chain 3-hydroxyacyl-CoA dehydrogenase [LCHAD] deficiency in fetus/mother); presents with malaise, nausea, abdominal pain, hypoglycaemia, coagulopathy. HELLP: part of the pre-eclampsia spectrum. Definitive treatment is URGENT DELIVERY of the fetus plus supportive care. Do not wait for transplant criteria if delivery is achievable — most recover postpartum. Distinguish from acute hepatitis and HSV hepatitis (which does NOT improve with delivery and needs aciclovir).

[1]

Drug-induced liver injury (DILI) — idiosyncratic

Unpredictable, dose-independent, with a latency of days to months. Common culprits: isoniazid (add pyridoxine; risk with alcohol/RIF), valproate (carnitine deficiency; treat with L-carnitine), nitrofurantoin, phenytoin, halothane, statins, amiodarone, herbal/dietary supplements (kava, comfrey, germander, green-tea extract, black cohosh). Management: STOP the offending drug immediately, support, and assess for transplant — idiosyncratic DILI carries a poor transplant-free survival.

[1]

Pathophysiology of cerebral oedema — the leading cause of death

Educational schematic of ammonia-glutamine cascade in acute liver failure: failed hepatic clearance, ammonia crosses BBB, astrocytic glutamine synthetase, cytotoxic oedema, raised ICP and herniation
FigureAmmonia–glutamine axis — failed urea-cycle clearance, ammonia enters astrocytes, glutamine accumulates as an osmolyte, cytotoxic oedema raises ICP, and herniation is the leading cause of death in ALF.

Cerebral oedema with intracranial hypertension is the #1 cause of death in ALF. Its incidence tracks encephalopathy grade: rare in Grade I-II, ~25-35% in Grade III, and >75% in Grade IV encephalopathy. The mechanism is cytotoxic (intracellular) oedema driven by the ammonia-glutamine axis.[2]

The ammonia-glutamine cascade

Ammonia -> glutamine -> astrocyte swelling -> herniation

1

Ammonia generation and failed clearance

Gut bacteria and enterocytes generate ammonia from dietary protein and glutamine. The failing liver cannot convert ammonia to urea (urea cycle), and portosystemic shunting bypasses residual hepatocytes. Serum ammonia rises.

2

Ammonia crosses the blood-brain barrier

Ammonia diffuses readily across the BBB (as NH3) and is also actively transported. In ALF the BBB itself becomes more permeable, accelerating influx.

3

Astrocytic glutamine synthetase

Within astrocytes, glutamine synthetase condenses ammonia with glutamate to form **glutamine**. This is the brain's only significant ammonia-detoxification route. The astrocyte is therefore the cellular target of ammonia toxicity.

4

Glutamine accumulation -> osmotic stress

Glutamine accumulates intracellularly and acts as an osmolyte, drawing water into the astrocyte. The astrocyte swells (**cytotoxic oedema**). Mitochondrial glutamine also enters and is deaminated back to ammonia by glutaminase, generating intramitochondrial ammonia and reactive oxygen species ("the Trojan horse" hypothesis).

5

Astrocyte swelling -> cerebral oedema

Swollen astrocytes (the foot processes regulate the blood-brain barrier and capillary diameter) compress microvasculature, reduce cerebral perfusion, and disrupt the extracellular matrix. Cytotoxic oedema generalises. Brain volume increases within a fixed skull.

6

Raised ICP -> herniation

Intracranial pressure rises; cerebral perfusion pressure (CPP = MAP - ICP) falls. The swollen brain shifts: uncal herniation (CN III palsy — fixed dilated pupil), then tonsillar herniation with brainstem compression and Cushing's triad (hypertension, bradycardia, irregular respiration), then brain death. This is why Grade III-IV encephalopathy mandates intubation and ICP-targeted therapy.

[2]

Ammonia as a bedside marker — but it is a surrogate

Arterial ammonia >150-200 umol/L correlates with intracranial hypertension and herniation risk, and a rising ammonia is ominous. However, the absolute level is less reliable than the trend and the grade of encephalopathy. Treat the encephalopathy and the ICP, not the number. Lactulose reduces ammonia in cirrhotic HE but has a more limited role in ALF (may precipitate ileus/volume depletion and has not shown a mortality benefit in ALF specifically).

[1]

Why cerebral oedema is uniquely dangerous in ALF (vs cirrhosis)

ALF cerebral oedema

Acute, cytotoxic

  • Rapid, massive ammonia rise -> acute astrocyte swelling
  • True cytotoxic cerebral oedema -> intracranial HYPERTENSION -> herniation
  • Common and lethal in Grade III-IV (~75% in Grade IV)
  • Astrocytes have no time to compensate (no osmotic adaptation)
  • Requires hyperosmolar therapy, head elevation, often ICP monitoring

Cirrhotic HE

Chronic, compensated

  • Chronic, lower-grade ammonia exposure -> astrocyte Alzheimer type II change
  • Astrocytes adapt osmotically (extrude myo-inositol) -> minimal oedema
  • Intracranial hypertension and herniation are RARE
  • Lactulose + rifaximin effective; management is precipitant-driven
  • Different disease, different management pathway
[2]
Practice point

Systemic inflammation is the multiplier

Ammonia is necessary but not sufficient. Systemic inflammation (infection, SIRS) synergises with ammonia to precipitate and worsen HE and cerebral oedema — a patient can deteriorate from Grade II to IV within hours of developing sepsis. This is why ALF protocols mandate infection surveillance and early empirical antibiotics at any sign of clinical deterioration: controlling inflammation is cerebral protection.

[1]

Management

Educational management infographic for acute liver failure: NAC for all causes, ICP-oriented care in grade III-IV encephalopathy, rebalanced haemostasis, King's College transplant referral
FigureALF ICU bundle — NAC for every aetiology, protect the brain in deep encephalopathy, support organs without chasing INR, and refer early against King's College Criteria.

General principles

ALF is a time-critical, multi-organ disease managed (ideally) in a transplant-capable centre. The four pillars are: (1) aetiology-directed therapy (NAC, silibinin, antivirals, steroids, delivery); (2) cerebral oedema prevention and control; (3) multi-organ supportive care (glucose, renal, haemodynamics, infection, coagulation); and (4) early transplant assessment against King's College Criteria. Refer to a transplant centre as soon as ALF is diagnosed — do not wait for deterioration.[1][2]

ALF ICU management protocol — the full bundle

1

NAC for ALL causes

N-acetylcysteine 150 mg/kg IV over 1 h, then 50 mg/kg over 4 h, then 100 mg/kg over 16 h. Lee et al (2009, Gastroenterology) showed improved transplant-free survival in non-paracetamol ALF, especially early-stage (Grade I-II). Give to ALL ALF patients regardless of cause. Low risk, potential benefit, cheap. Continue until INR improving, encephalopathy resolving, or transplant.

2

Cerebral oedema management

Grade III-IV encephalopathy = high risk of intracranial hypertension. INTUBATE for airway protection and to enable ICP-targeted care. Head of bed elevated 30 degrees, midline. Maintain normocapnia (PaCO2 35-40), normoxia, normothermia (avoid fever), normoglycaemia. Hyperosmolar therapy: hypertonic 3% saline (target Na 145-155 mmol/L) or mannitol 0.5-1 g/kg if not anuric. Consider ICP monitor in Grade IV. Avoid hypotension/hypoxia (each episode worsens secondary brain injury). Induced hypothermia (32-34 C) and hypernatraemia as rescue. Hyperventilation as a TEMPORARY bridge only (causes cerebral vasoconstriction lowering ICP but reduces cerebral perfusion if sustained).

3

Coagulopathy — do NOT chase the INR

INR is a liver FUNCTION marker (prognostic, central to King's College Criteria), NOT a bleeding-risk indicator. ALF is a state of "rebalanced haemostasis": both procoagulant factors AND anticoagulant factors (protein C/S, antithrombin) are reduced, and platelets are often dysfunctional but thrombopoietin-deficient. Viscoelastic testing (TEG/ROTEM) better reflects true haemostasis than INR. Give vitamin K 10 mg IV. Use FFP ONLY for active bleeding or before invasive procedures. Recombinant activated factor VII for emergencies. Do not transfuse platelets prophylactically unless bleeding.

4

Hypoglycaemia

Check glucose hourly. Maintain >4 mmol/L with a 10% dextrose infusion (and 50% boluses if needed). Impaired hepatic gluconeogenesis and depleted glycogen make hypoglycaemia both common and dangerous — it worsens the already-injured brain and can mimic or precipitate encephalopathy.

5

Renal support

AKI is common (~50-70%): hepatorenal syndrome, ATN from hypotension/paracetamol nephrotoxicity, or contrast. Avoid nephrotoxins. CRRT is PREFERRED over intermittent haemodialysis — continuous therapy avoids the rapid solute shifts and blood pressure swings that worsen cerebral oedema and raise ICP. Maintain adequate MAP for cerebral and renal perfusion.

6

Haemodynamics

ALF is a hyperdynamic, vasodilated state (NO-mediated). Target MAP >75 mmHg and CPP >60 mmHg. Norepinephrine is first-line vasopressor (alpha agonism preserves CPP). Add vasopressin/terlipressin for refractory vasodilation. Avoid pure beta-agonists. Fluid resuscitate with balanced crystalloid; albumin for hypoalbuminaemia. Invasive arterial and central monitoring essential.

7

Infection surveillance

Marked immune paralysis -> very high infection risk (bacterial ~60-80%, fungal ~30%). Surveillance cultures (blood, urine, sputum, line) daily. Empirical broad-spectrum antibiotics for any clinical deterioration (new fever, rising WCC/w lactate, worsening encephalopathy or haemodynamics) — infection precipitates multi-organ failure and amplifies cerebral oedema. Add antifungal cover for prolonged ICU stay. Prophylactic antibiotics/antifungals are used in many units.

8

Liver transplant assessment

Urgent referral to a transplant centre at diagnosis — transfer BEFORE deterioration. Apply King's College Criteria on admission AND serially (they are dynamic). Time is critical: brain death from cerebral oedema is the #1 cause of death. List EARLY; do not wait for maximal deterioration, because many die or develop contraindications (sepsis, brain death) while listed.

[1] [2]

N-acetylcysteine (NAC) — for ALL causes

NAC is the one therapy in ALF with the broadest evidence base. In paracetamol overdose it restores hepatic glutathione and detoxifies NAPQI. In non-paracetamol ALF it improves transplant-free survival through pleiotropic effects: antioxidant (replenishes glutathione systemically), improved microcirculation and oxygen delivery, anti-inflammatory, and attenuation of cerebral oedema.[4]

Lee 2009 — NAC for non-paracetamol ALF (Gastroenterology; PMID 19524577)

Source

Multicentre double-blind RCT, US Acute Liver Failure Study Group (ALFSG); 173 patients with non-acetaminophen ALF, stratified by grade I-II vs III-IV encephalopathy

Intervention

IV N-acetylcysteine vs placebo, standard regimen

Primary outcome

Overall survival at 3 weeks: 70% NAC vs 66% placebo (1-sided P=0.28) — NOT significant for the whole cohort

Subgroup (Grade I-II HE)

Transplant-free survival SIGNIFICANTLY higher with NAC (52% vs 30%, P=0.02) — the patients who matter, before they deteriorate

Subgroup (Grade III-IV HE)

No survival benefit (most were already too sick / destined for transplant)

Clinical bottom line

Give NAC to ALL ALF patients, especially early-stage (Grade I-II) disease of ANY aetiology — it improves transplant-free survival and is safe, cheap, and easy. The benefit extends well beyond paracetamol.

[1]

NAC administration — practical

1

Indication

ALL patients with ALF of any aetiology, AND any paracetamol patient with hepatotoxicity (rising ALT or detectable level) regardless of time since ingestion.

2

Dose (standard IV regimen)

Loading 150 mg/kg in 200 mL 5% dextrose over 1 h (60 min); then 50 mg/kg in 500 mL over 4 h; then 100 mg/kg in 1000 mL over 16 h. Total 21 h. Total dose 300 mg/kg. Modify volume for fluid restriction.

3

Adverse effects

Anaphylactoid (not IgE) reaction in ~10-20% during loading: flush, rash, bronchospasm, rarely hypotension. More common in asthmatics and with rapid infusion. Management: slow/stop infusion temporarily, antihistamine (chlorphenamine), bronchodilator if needed, then RESTART at slower rate. Do NOT abandon NAC for a rash.

4

When to stop

Continue beyond the 21 h course if INR still rising or >2, paracetamol still detectable, or encephalopathy present. Stop when INR <2 and falling, encephalopathy resolving, paracetamol undetectable — OR at transplant.

[4]

King's College Criteria — transplant listing

The King's College Criteria (KCC), derived by O'Grady et al from the King's College London ALF cohort, remain the most widely used and validated criteria for predicting death without transplant and therefore for listing for emergency liver transplantation. They are aetiology-specific (paracetamol vs non-paracetamol) and should be applied on admission and serially.[5]

King's College Criteria for liver transplant in ALF

PARACETAMOL ALF — list for transplant if EITHER:

  • Arterial pH < 7.25 after adequate fluid resuscitation (independent of the other three), OR
  • ALL THREE of: INR > 6.5 (PT >100 s) + creatinine > 300 umol/L (3.4 mg/dL) + Grade III-IV encephalopathy [1]

NON-PARACETAMOL ALF — list for transplant if EITHER:

  • INR > 6.5 (PT >100 s) alone, OR
  • ANY THREE of the following FIVE:
    1. INR > 3.5 (PT >50 s)
    2. Age <10 or >40 years
    3. Aetiology: non-A non-B (seronegative) hepatitis, halothane, or idiosyncratic drug reaction
    4. Duration of jaundice before encephalopathy >7 days
    5. Bilirubin >300 umol/L (17.5 mg/dL)
[5]
Practice point

KCC performance — sensitivity vs specificity

KCC have a high specificity (~90%) — meeting them is a strong signal of death without transplant — but a modest sensitivity (~70%), so a patient can still die of ALF without ever formally meeting KCC. This is why KCC are applied repeatedly and why adjunctive markers (lactate, phosphate, ammonia) and clinical judgement (deteriorating trajectory, organ failure) also drive the decision. Do not be falsely reassured by "not yet meeting KCC" in a clearly deteriorating patient — discuss early with the transplant centre.

[1]

Paracetamol KCC

Single or combined

  • Arterial pH <7.25 after resuscitation (any grade of HE), OR
  • All three combined: INR >6.5 AND creatinine >300 umol/L AND Grade III-IV HE
  • Lactate adjunct: early arterial lactate >3.5 mmol/L (pre-resuscitation) or >3.0 after resuscitation predicts mortality
  • Phosphate adjunct: hypophosphataemia (Phosphate <0.55 mmol/L day 2-3) reflects hepatic regeneration failure

Non-paracetamol KCC

INR or 3 of 5

  • INR >6.5 (PT >100 s) alone, OR
  • Any 3 of 5: INR >3.5 / age <10 or >40 / unfavourable aetiology (seronegative, halothane, idiosyncratic) / jaundice-to-encephalopathy >7 d / bilirubin >300 umol/L
  • Aetiology-specific Wilson and seronegative cases almost always meet criteria
  • Less sensitive than paracetamol KCC; combine with lactate and clinical trajectory
[5]

Prognostic markers — beyond King's College

ALF prognostication is multi-modal. No single marker is sufficient; the trajectory matters as much as any single value. [1]

Static markers

On / near admission

  • King's College Criteria — aetiology-specific (above)
  • Arterial lactate — >3.5 mmol/L early (or >3.0 after resuscitation) predicts poor outcome
  • Arterial ammonia — >150-200 umol/L predicts intracranial hypertension
  • Phosphate — low phosphate (day 2-3) = failed regeneration = poor
  • Aetiology — paracetamol/ischaemic/pregnancy good; seronegative/DILI/Wilson poor
  • Tempo — hyperacute best; subacute worst
  • Grade of encephalopathy at presentation

Dynamic markers

Trajectory over hours

  • Rising INR despite therapy = worsening synthetic function
  • Rising arterial lactate despite resuscitation = worsening perfusion/clearance
  • Progression of encephalopathy grade (II -> III -> IV) = cerebral oedema risk
  • Rising ammonia = intracranial hypertension risk
  • Development of renal failure / vasopressor dependence / new infection
  • Falling CPP on ICP monitoring
[2]

Lactate — the neglected prognostic champion

Arterial lactate reflects both impaired hepatic clearance (the failing liver is a major lactate-clearing organ) and tissue hypoperfusion. An early arterial lactate >3.5 mmol/L (before resuscitation) or >3.0 mmol/L after resuscitation, or a persistently rising lactate despite adequate resuscitation, is a strong independent predictor of mortality in paracetamol and non-paracetamol ALF. Lactate is now incorporated into modified King's College algorithms and CLIF-C/ALF scores. Check it early and serially, alongside pH, INR, creatinine, and ammonia — a rising lactate should escalate transplant discussion even before formal KCC are met.

[1]

Alternative and composite scores

Other prognostic tools (know they exist)

1

Clichy criteria (France)

Historically used for HBV-related ALF: factor V level <20% of normal in patients <30 years, or <30% in patients >30 years, combined with Grade III-IV encephalopathy. Less used outside France.

2

MELD

Model for End-stage Liver Disease — validated in cirrhosis; used in some ALF settings but less discriminating than KCC for ALF. Rising MELD reflects worsening.

3

CLIF-C ACLF / ALF scores

Derived from the CANONIC ACLF cohort and extended to ALF; combine organ-failure score with age and white-cell count. Useful for mortality prediction and goals-of-care discussions.

4

APACHE II

Generic ICU score; reasonable for overall mortality prediction but not designed for ALF transplant decisions. Not a substitute for KCC.

[2]

Complications of ALF

Cerebral

#1 cause of death

  • Cerebral oedema / intracranial hypertension -> tonsillar herniation
  • Seizures (often non-convulsive — low threshold for cEEG)
  • Cerebral hypoperfusion (low CPP) -> secondary injury

Metabolic / haematologic

  • Hypoglycaemia (impaired gluconeogenesis) — check hourly
  • Coagulopathy (low factors + low anticoagulants = rebalanced) — bleeding LESS common than INR suggests
  • Thrombocytopenia (consumption, splenomegaly, sepsis)
  • Metabolic acidosis (lactate) — prognostic
  • Hypophosphataemia, hypokalaemia, hypomagnesaemia

Renal / circulatory

  • AKI (~50-70%) — HRS, ATN, paracetamol nephrotoxicity
  • Hyperdynamic vasodilatory shock (NO-mediated) — norepinephrine first-line
  • Pancreatitis (paracetamol association)

Infective

  • Bacterial infection (~60-80%) — Gram-positive and Gram-negative; immune paralysis
  • Fungal infection (~30%) — Candida; prolonged ICU stay
  • Sepsis precipitates multi-organ failure and amplifies cerebral oedema
[2]

Key trials and evidence

O'Grady 1989 — derivation of the King's College Criteria (Gastroenterology; PMID 2490426)

Source

Retrospective analysis of 588 ALF patients (King's College London, 1973-1985), validated prospectively in 175 patients (1986-1987)

Objective

Identify early, objective indicators of prognosis to select patients for liver transplantation

Paracetamol criteria

Arterial pH <7.30 (revised to <7.25), OR all three of (INR/PT >100 s + creatinine >300 umol/L + grade III-IV encephalopathy)

Non-paracetamol criteria

INR >6.5 alone, OR any 3 of 5 (INR >3.5, age <10/>40, unfavourable aetiology, jaundice-to-encephalopathy >7 d, bilirubin >300 umol/L)

Performance

Sensitivity ~70%, specificity ~90% for death without transplant — high specificity (meeting criteria = very likely to die), modest sensitivity (not meeting criteria does NOT guarantee survival)

Clinical bottom line

Apply KCC on admission AND serially; meeting them triggers urgent transplant referral. They remain the global standard for ALF transplant listing 35 years on.

[1]

Lee 2009 — NAC for non-paracetamol ALF (Gastroenterology; PMID 19524577)

Source

Multicentre double-blind RCT, US ALFSG; 173 patients with non-acetaminophen ALF

Intervention

IV N-acetylcysteine vs placebo

Whole-cohort survival

70% NAC vs 66% placebo at 3 weeks — not significant

Grade I-II subgroup

Transplant-free survival 52% NAC vs 30% placebo (P=0.02) — significant

Clinical bottom line

NAC improves transplant-free survival in early-stage non-paracetamol ALF -> give NAC to ALL ALF, especially early-grade disease. Cheap, safe, broad benefit.

[1]

O'Grady 1993 — redefining the syndromes of ALF (Lancet; PMID 8491116)

Source

Conceptual reclassification of ALF by the interval between jaundice and encephalopathy, grounded in the King's College cohort

Contribution

Defined hyperacute (<7 d), acute (8-28 d), and subacute (5-12+ weeks) sub-syndromes with distinct aetiologies, complication profiles, and prognoses

Key insight

The faster the encephalopathy, paradoxically the BETTER the transplant-free survival (hyperacute/paracetamol regenerates; subacute/seronegative does not)

Clinical bottom line

The classification that frames every modern ALF discussion — tempo plus aetiology determines prognosis and transplant urgency.

[1]

ALF evidence and outcomes — the synthesis

NAC

Mortality benefit in paracetamol ALF even when started >24 h after ingestion; transplant-free survival benefit in non-paracetamol ALF (Lee 2009). Give to ALL ALF.

King's College Criteria

Specificity ~90%, sensitivity ~70% for death without transplant. Apply serially; adjunct lactate/ammonia/phosphate refine.

Transplantation

1-year survival ~80-90% for ALF — the definitive therapy for those who meet criteria.

MARS / albumin dialysis

RELIEF and subsequent trials: no mortality benefit; may improve encephalopathy and bridge to transplant. Not standard of care for survival.

Spontaneous survival

Overall ~40-60% without transplant; aetiology-dependent (paracetamol/pregnancy/ischaemic best; seronegative/DILI/Wilson worst).

Cerebral oedema

#1 cause of death; hypertonic saline + mannitol + ICP-directed care in Grade III-IV. Induced hypothermia and indometacin as rescue.

[1]
King's College Criteria decision algorithm for liver transplant in acute liver failure, branching for paracetamol vs non-paracetamol aetiology
FigureDecision algorithm. A patient meeting ANY arm of the King's College Criteria (with or without supportive lactate/ammonia) triggers urgent transplant-centre referral. Because specificity exceeds sensitivity, a deteriorating patient who has not yet met KCC still warrants early discussion — apply the criteria repeatedly.

SAQ — Paracetamol-induced acute liver failure: King's College Criteria and the transplant decision

10 minutes · 10 marks

A 24-year-old woman is brought to the emergency department 48 hours after a staggered paracetamol overdose (estimated 25 g). She is drowsy (GCS 13), jaundiced and has been vomiting. Bloods: ALT 8,500 U/L, INR 6.8, creatinine 312 umol/L, arterial pH 7.18, lactate 5.2 mmol/L, glucose 2.6 mmol/L, paracetamol level 220 mg/L at 48 h. She has received 30 mL/kg crystalloid. The liver transplant unit is on the phone asking whether to accept her for listing.

[1]

SAQ — Cerebral oedema and intracranial hypertension in acute liver failure

10 minutes · 10 marks

A 30-year-old man with paracetamol-induced acute liver failure (INR 7.2, ALT 11,000 U/L) has progressed from Grade II to Grade IV encephalopathy over 6 hours. He is intubated and ventilated. He becomes hypertensive (BP 188/96) with bradycardia (HR 44), his pupils become asymmetric (right 6 mm and fixed; left 3 mm) and he develops decerebrate posturing to stimuli. Serum ammonia is 210 umol/L. You are called to manage suspected intracranial hypertension.

[1]

Clinical pearls

High-yield ALF points for the CICM/FFICM exam

  1. NAC for ALL ALF — not just paracetamol (Lee 2009 — transplant-free survival benefit in early-stage non-paracetamol ALF).[4] }
  2. Cerebral oedema = #1 cause of death in ALF (~75% in Grade IV HE). Intubate Grade III-IV; head up 30 degrees; hypertonic saline/mannitol; consider ICP monitor.[2] }
  3. King's College Criteria determine transplant listing — paracetamol: pH <7.25 OR all of (INR >6.5 + Cr >300 + grade III-IV HE); non-paracetamol: INR >6.5 OR >=3 of 5 (INR >3.5, age <10/>40, unfavourable aetiology, jaundice-to-HE >7 d, bilirubin >300). Apply serially.[5] }
  4. Do NOT routinely correct INR with FFP — it is a liver function marker (rebalanced haemostasis); FFP only for bleeding/procedures. Use TEG/ROTEM to assess real bleeding risk.[2] }
  5. Paracetamol is #1 cause in UK/Aus/US. Viral hepatitis (HBV) in the developing world.[1] }
  6. Hypoglycaemia is common and dangerous — check hourly, give 10% dextrose.[2] }
  7. CRRT preferred over intermittent dialysis (avoids solute/pressure shifts worsening cerebral oedema).[2] }
  8. Infection precipitates multi-organ failure and amplifies cerebral oedema — surveillance cultures, low threshold for empirical antibiotics/antifungals.[2] }
  9. Rumack-Matthew nomogram for paracetamol determines NAC need — NOT applicable in staggered overdose or >24 h (treat if any hepatotoxicity).[1] }
  10. Grade III-IV HE: intubate, hyperosmolar therapy, consider ICP monitoring.[2] }
  11. Wilson disease: young + Coombs-negative haemolysis + low ceruloplasmin + low ALP + Kayser-Fleischer rings = Wilsonian crisis. Near-100% fatal without transplant.[2] }
  12. Mushroom poisoning (Amanita phalloides): delayed GI symptoms (6-24 h) then liver failure at 48-72 h. Treat with silibinin + NAC + forced diuresis.[2] }
  13. Pregnancy (AFLP, HELLP): third trimester — delivery is the definitive treatment.[2] }
  14. Spontaneous survival: ~40-60% without transplant — paracetamol/ischaemic/pregnancy best; seronegative/idiosyncratic/Wilson worst.[1] }
  15. O'Grady tempo classification: hyperacute (<7 d, best prognosis), acute (8-28 d), subacute (29 d-12 wk, worst). Faster onset paradoxically = better recovery.[3] }
  16. Ammonia-glutamine axis: ammonia -> astrocyte glutamine synthetase -> glutamine -> astrocyte swelling (cytotoxic oedema) -> raised ICP -> herniation.[2] }
  17. Lactate is a prognostic champion — early arterial lactate >3.5 mmol/L (or rising despite resuscitation) predicts poor outcome; escalate transplant discussion.[2] }
  18. Norepinephrine is first-line vasopressor — preserves cerebral perfusion pressure (MAP - ICP).[2] }
  19. Anaphylactoid reaction to NAC (rash, bronchospasm) is common during loading — slow the infusion, antihistamine, then RESTART; do NOT abandon NAC.[4] }
  20. KCC sensitivity (~70%) < specificity (~90%) — a deteriorating patient can die without ever formally meeting criteria; use trajectory and adjunct markers.[5] }
  21. Non-convulsive seizures are common in deep HE — low threshold for continuous EEG.[2] }
  22. Hypertonic saline (target Na 145-155) is increasingly favoured over mannitol for cerebral oedema in ALF; mannitol reserved if not anuric.[2] }
Myth-buster

Myth: A high INR in ALF means a high bleeding risk

False — and acting on it causes harm. ALF is a state of rebalanced haemostasis: the failing liver underproduces BOTH procoagulant factors (II, V, VII, IX, X) AND anticoagulant factors (protein C, protein S, antithrombin), and platelets are thrombopoietin-deficient. The INR measures only the procoagulant side and therefore overstates bleeding risk. Viscoelastic tests (TEG/ROTEM) typically show near-normal or even procoagulant traces. Prophylactic FFP raises volume, transfusion burden, and risk, and obscures the INR as a prognostic marker. Reserve factor replacement for active bleeding or before invasive procedures.

[1]
Practice point

The transfer window is measured in hours

The single biggest system-level failure in ALF is late referral. Patients are managed optimistically in a non-transplant centre until they meet KCC or deteriorate — by which point cerebral oedema, sepsis, or multi-organ failure may have closed the window. Refer at diagnosis, not at deterioration. A telephone discussion with the transplant centre costs nothing; the cost of a delayed transfer is frequently death.

[1]

Red flags

Critical ALF points

  • NAC for ALL ALF — not just paracetamol (Lee 2009). Give immediately upon diagnosis.[4] }
  • Cerebral oedema is #1 cause of death — monitor ICP if Grade III-IV. Head up 30 degrees, hypertonic saline/mannitol, avoid hypotension/hypoxia/hypoglycaemia.[2] }
  • Do NOT routinely correct INR with FFP — it is a liver function marker (rebalanced haemostasis), not bleeding risk. FFP only for bleeding/procedures.[2] }
  • King's College Criteria — urgent transplant referral. Apply on admission AND serially. Do NOT delay for deterioration.[5] }
  • Hypoglycaemia is common and worsens brain injury — check hourly. 10% dextrose infusion.[2] }
  • Lactate >3.5 mmol/L early, or rising — predicts poor outcome; escalate transplant discussion even before formal KCC.[2] }
  • Wilson disease + ALF — near-100% fatal without transplant; urgent listing. Look for Coombs-negative haemolysis, low ceruloplasmin, low ALP.[2] }
  • Pregnancy-related ALF (AFLP/HELLP) — urgent delivery of the fetus is definitive treatment.[2] }
  • Amanita mushroom poisoning — delayed GI then liver failure; silibinin + NAC + forced diuresis.[2] }
  • CRRT over intermittent dialysis — continuous avoids solute shifts worsening cerebral oedema.[2] }
  • Refer to transplant centre at diagnosis — the transfer window is measured in hours, not days.[1] }

Exam summary — the ten things to take to the exam

ALF at a glance — viva-ready

1

Define

Acute liver injury + INR >1.5 + encephalopathy, within 26 weeks, no pre-existing liver disease.

2

Classify (O'Grady)

Hyperacute <7 d (paracetamol, best); acute 8-28 d (viral); subacute 29 d-12 wk (seronegative/DILI, worst). Faster = better.

3

Causes

Paracetamol #1 (UK/Aus), idiosyncratic drugs, viral (HBV), ischaemic, autoimmune, Wilson, pregnancy (AFLP/HELLP), Amanita, indeterminate.

4

Mechanism of death

Ammonia -> astrocyte glutamine -> astrocyte swelling -> cytotoxic cerebral oedema -> raised ICP -> herniation (#1 cause of death).

5

Drug all patients get

N-acetylcysteine (NAC) for ALL causes — Lee 2009 showed transplant-free survival benefit beyond paracetamol. Also vitamin K 10 mg.

6

Transplant trigger

King's College Criteria (paracetamol: pH <7.25 OR INR >6.5 + Cr >300 + grade III-IV HE; non-paracetamol: INR >6.5 OR >=3 of 5). Apply serially.

7

Prognostic markers

KCC + arterial lactate (>3.5 early = poor) + ammonia (>150-200 = intracranial HTN risk) + phosphate (low = failed regeneration) + aetiology + tempo.

8

Do NOT do

Do NOT routinely correct INR with FFP. Do NOT use intermittent dialysis (use CRRT). Do NOT delay transplant referral. Do NOT use the nomogram in staggered paracetamol overdose.

9

Supportive bundle

Hypoglycaemia (10% dextrose, hourly glucose), cerebral oedema (intubate Grade III-IV, head up, hypertonic saline/mannitol, ICP monitor), AKI (CRRT), infection (surveillance + empirical antibiotics), haemodynamics (norepinephrine, MAP >75, CPP >60).

10

Outcomes

Spontaneous survival ~40-60% (aetiology-dependent); transplant 1-year survival ~80-90%. Refer early.

[1] [2]

References

  1. [1]Lee WM, Stravitz RT, Larsen RA. Introduction to the revised American Association for the Study of Liver Diseases Position Paper on acute liver failure 2011 Hepatology, 2012.PMID 22213561
  2. [2]Bernal W, Wendon J. Acute liver failure N Engl J Med, 2013.PMID 24369077
  3. [3]O'Grady JG, Schalm SW, Williams R. Acute liver failure: redefining the syndromes Lancet, 1993.PMID 8101303
  4. [4]Lee WM, Hynan LS, Rossaro L, Fontana RJ, Stravitz RT, Schiodt FV, et al. Intravenous N-acetylcysteine improves transplant-free survival in early stage non-acetaminophen acute liver failure Gastroenterology, 2009.PMID 19524577
  5. [5]O'Grady JG, Alexander GJ, Hayllar KM, Williams R. Early indicators of prognosis in fulminant hepatic failure Gastroenterology, 1989.PMID 2490426