It's 2am. A 22-year-old woman is brought in by her flatmate. She took "a lot of paracetamol" about 6 hours ago. She's awake, slightly nauseous. Her LFTs are normal. Her INR is normal. She doesn't look sick.
This is the most dangerous moment. Paracetamol poisoning has a deceptive latency — the injury happens silently for 24–72 hours before clinical signs emerge. By the time the liver is visibly failing, the window for the most effective treatment may have closed.
This module will teach you why the injury is silent at first, what is happening at a cellular level that you cannot yet measure, when each biomarker rises and falls and why, and how to make the right decisions at every stage — from the ED to the liver ICU.
Paracetamol is normally harmless because the liver converts it into safe waste. In overdose, the liver's detox system gets overwhelmed and starts producing a toxic byproduct (NAPQI) that's like acid burning through liver cells from the inside out. The only catch? The burning happens before any alarm goes off on your blood tests. By the time you see the damage on the numbers, the fire has already been raging for 24–48 hours.
Without liver transplantation, patients meeting King's College Criteria have <15% survival. Even with good ICU care, multi-organ failure — cerebral oedema, renal failure, sepsis — kills.
Most deaths are preventable with early recognition + NAC + timely specialist referral. This is a disease where knowing the physiology directly saves lives.
Paracetamol ALF has a better prognosis than any other cause of ALF in patients who receive early NAC. Even after liver transplant, long-term outcomes are excellent — the liver lesion doesn't recur.
Unlike other causes of ALF, the liver can fully regenerate if the patient is supported through the acute phase. Zone 3 hepatocytes can be replaced.
At therapeutic doses, 90–95% of paracetamol is safely conjugated by two high-capacity pathways — glucuronidation (~60%) and sulphation (~35%) — creating water-soluble, non-toxic metabolites excreted in urine and bile.
Only 5–10% is oxidised by CYP450 enzymes (predominantly CYP2E1, plus CYP1A2 and CYP3A4) into a reactive intermediate called NAPQI (N-acetyl-p-benzoquinone imine). Under normal conditions, this tiny amount is immediately neutralised by conjugation with glutathione (GSH) — forming a safe mercapturate that is excreted renally.
Step 1: NAPQI accumulates — no GSH left to mop it up.
Step 2: NAPQI covalently binds cysteine residues of mitochondrial proteins, forming NAPQI-protein adducts.
Step 3: Mitochondrial dysfunction → ROS/RNS (reactive oxygen/nitrogen species) generation → JNK (c-Jun N-terminal kinase) pathway activation.
Step 4: Mitochondrial Permeability Transition (MPT) pore opens → loss of membrane potential → ATP depletion.
Step 5: ATP depletion → oncotic necrosis (not apoptosis) — cell swells and bursts, releasing DAMPs (damage-associated molecular patterns) → sterile inflammatory cascade.
Chronic alcohol: Dual mechanism — CYP2E1 is upregulated (more NAPQI from any dose) AND chronic oxidative stress depletes GSH baseline stores.
Fasting/malnutrition: GSH requires cysteine — fasting reduces precursor availability. GSH stores can fall 50% after 24h fasting.
Enzyme inducers: Rifampicin, phenytoin, phenobarbital, carbamazepine, isoniazid — all induce CYP450, increasing NAPQI production.
Note on acute alcohol: Paradoxically, acute alcohol is somewhat protective — it competes with paracetamol for CYP2E1, reducing NAPQI generation.
Think of the liver like a factory floor with three sections. Section 1 (Zone 1) is nearest the loading dock — it gets the freshest oxygen and blood. Section 3 (Zone 3) is at the far end — it gets the depleted, oxygen-poor leftovers. Now here's the catch: Section 3 has the most of the dangerous machine (CYP2E1) that turns paracetamol into its toxic form. So Zone 3 gets the most poison AND the least oxygen to deal with it. It's the first to die.
Zone 1 hepatocytes (periportal) are oxygen-rich, CYP2E1-poor, and GSH-replete. They are the last to die and — critically — are the source of liver regeneration.
If Zone 1 is preserved, the liver can regenerate completely. This is why paracetamol ALF has a better prognosis than other causes: the regenerative infrastructure is often intact.
Zone 1 also has high glucuronidation and sulphation capacity — the safe metabolic pathways — which is why these remain intact longer in overdose.
Centrilobular necrosis is not unique to paracetamol. Any cause of reduced Zone 3 oxygen delivery can mimic the pattern histologically:
Ischaemic hepatitis ("shock liver") — hypotension reduces portal and hepatic artery flow; Zone 3 dies first. AST/ALT can exceed 10,000 IU/L in severe ischaemia — similar to paracetamol.
Right heart failure / Budd-Chiari — venous congestion reduces hepatic venous outflow; Zone 3 is compressed. Always check for elevated CVP/JVP in unexplained transaminitis.
| Biomarker | Mechanism of Rise | Timing | Key Clinical Point |
|---|---|---|---|
| AST | Released from necrotic hepatocytes (cytosolic + mitochondrial isoforms). In liver, 80% of AST is mitochondrial. | Rises from ~12–24h. Peaks 48–72h. Falls fast (t½ ~15–18h) | Rises and falls faster than ALT. Rising AST on Day 4 is very poor prognosis. AST:ALT >2 in fulminant failure. |
| ALT | Predominantly cytosolic hepatocyte enzyme — more liver-specific than AST. Released by membrane permeability change and necrosis. | Rises parallel to AST. Peaks 48–72h. Falls slowly (t½ ~47h) | More liver-specific. Lingers ~24–48h after AST has normalised. Use ALT (not AST) to determine when liver has healed. |
| INR / PT | Liver synthesises clotting factors II, V, VII, IX, X. Factor VII has the shortest half-life (~4–6h). As hepatocytes are destroyed, synthesis falls. | Begins rising from 24–36h. Rising INR on Day 4 = very poor sign. | PT that exceeds hours since ingestion = extreme risk. INR >6.5 + AKI + Grade III/IV encephalopathy = KCC met. Do NOT give FFP prophylactically — it masks prognostic information. |
| Ammonia | Liver normally converts ammonia (from gut bacteria/amino acid metabolism) to urea. Hepatocyte loss → ammonia accumulates. Kidneys can't compensate adequately. | Rises later — often from 48–72h onwards in ALF. | Ammonia >124 μmol/L predicts mortality with 77% accuracy. Ammonia >200 μmol/L strongly associated with cerebral oedema and herniation. Always sample on ice, process immediately. |
| Lactate | In ALF, liver cannot clear lactate (normally 50–70% of lactate clearance occurs in liver). Hypoperfusion + mitochondrial dysfunction add to production. | Early predictor. Arterial lactate >3.5 mmol/L at 4h post-resuscitation. | Part of modified KCC. pH <7.3 OR lactate >3 mmol/L after full fluid resuscitation predicts 97% specificity for death without transplant. |
| Bilirubin | Impaired conjugation and excretion by damaged hepatocytes + haemolysis from DIC. Bilirubin peaks later than transaminases. | Rises from 24–48h. Peaks later than AST/ALT. | Bilirubin >300 μmol/L in non-paracetamol ALF is part of KCC. In paracetamol ALF, level correlates with severity but is not in the primary KCC. |
| Phosphate | Regenerating hepatocytes consume phosphate rapidly. Falling phosphate at 48–96h = hepatocyte regeneration beginning. | 48–96h. | Phosphate >1.2 mmol/L (3.75 mg/dL) at 48–96h = poor prognosis (liver not regenerating). Phosphate normalising = good sign of regeneration. |
The elevated INR in acute liver failure reflects reduced synthesis of both pro- and anti-coagulant factors. Whole blood coagulation (as measured by ROTEM/TEG) is often surprisingly preserved or even hypercoagulable, because factor VIII (made outside the liver by endothelium) is paradoxically elevated.
Do NOT give FFP to correct INR in ALF unless there is active bleeding or before an invasive procedure. Prophylactic FFP obscures the prognostic INR signal, risks fluid overload, ARDS and transfusion reactions, and does not reduce mortality.
AST has a serum half-life of ~15–18 hours. ALT has a serum half-life of ~47 hours. They rise together (peak at 48–72h), but AST normalises roughly 24–48 hours earlier than ALT in recovery.
This has a practical consequence: if NAC discontinuation criteria are based on AST dropping to <50% of peak, NAC may be stopped up to 24 hours too early. Use ALT, not AST, as your guide for when the liver has truly recovered and NAC can be stopped.
Acute Liver Failure = coagulopathy (INR >1.5) + hepatic encephalopathy (HE) in a patient without prior liver disease, occurring within 26 weeks of onset of liver disease. The absence of prior liver disease is the key distinguishing feature from acute-on-chronic liver failure.
Hyperacute ALF (jaundice to HE in <7 days) — typical of paracetamol. Paradoxically, better prognosis due to higher spontaneous recovery rate. Subacute ALF (jaundice to HE in 5–26 weeks) — worst prognosis, least likely to spontaneously recover.
| Grade | Clinical Features | Mechanism | Action |
|---|---|---|---|
| I | Altered sleep, mild confusion, irritability, poor concentration. Subtle asterixis. | Ammonia crossing BBB → mild astrocyte swelling + GABA-A potentiation | Monitor closely. HDU admission. Check ammonia. |
| II | Drowsy but rousable. Disoriented to time and place. Obvious asterixis. Ataxia. | Progressive glutamine accumulation in astrocytes → cytotoxic oedema begins | HDU/ICU admission. Lactulose. Urgent KCC assessment. Contact transplant centre. |
| III | Somnolent, rousable only to pain. Severe disorientation. Hyperreflexia. Focal neurology may appear. | Cerebral oedema developing. ICP beginning to rise. Cerebral autoregulation impaired. | Intubate for airway protection. ICU. Hypertonic saline to Na 145–150. Head 20–30°. Avoid fever/stimulation. |
| IV | Coma. Absent response to pain. Signs of herniation (dilated pupils, Cushing's triad). | Severe ICP. Cerebral blood flow critically impaired. Brain herniation risk. | Intubated. Mannitol or hypertonic saline bolus. CRRT for ammonia clearance. Emergency transplant if criteria met. |
Ammonia crosses the blood-brain barrier. In astrocytes — the only brain cells expressing glutamine synthetase — ammonia is converted to glutamine. Glutamine accumulates intracellularly, becoming osmotically active and causing astrocyte swelling (cytotoxic oedema).
Simultaneously, ammonia inhibits neuronal mitochondria, impairs α-ketoglutarate dehydrogenase, and potentiates GABA-A receptor activity (causing sedation/coma).
Arterial ammonia >124 μmol/L predicts mortality with 77.5% accuracy. >200 μmol/L = high risk of cerebral oedema and herniation. Always use arterial samples, immediately on ice.
Cytotoxic oedema (cellular swelling — astrocytes) — from glutamine accumulation. Vasogenic oedema (BBB disruption) — from inflammatory mediators. Both contribute in ALF.
Management targets: ICP <20 mmHg, CPP >50 mmHg. Strategies: hypertonic saline (Na 145–150 mmol/L), CRRT (ammonia clearance), head-up 20–30°, avoid stimulation, treat fever, light sedation (propofol preferred).
Invasive ICP monitoring is now largely reserved for grade III/IV encephalopathy where transplant is being actively considered — the risk-benefit in a coagulopathic patient is carefully weighed.
The King's College Criteria answer one question: "Is this patient's liver going to kill them without a new one?" It uses simple blood tests and clinical signs that predict death with >90% specificity. If the criteria are met, the patient has <15% chance of surviving without a transplant. The criteria are specific but not perfectly sensitive — meeting them means "definitely list"; not meeting them doesn't guarantee safety.
The original KCC was described by O'Grady et al. (Gastroenterology 1989) from 588 patients at King's College Hospital. The paracetamol arm used pH <7.3 OR all three of INR >6.5 + creatinine >300 + Grade III/IV HE. The non-paracetamol arm used INR >6.5 alone OR any 3 of 5 criteria. These original thresholds remain the foundation — the modified version adds lactate (Bernal et al. Lancet 2002) and phosphate (Schmidt et al. 2002) to improve sensitivity. All specialist liver centres use the modified criteria in current practice.
Low sensitivity (~58–69%): Up to 40% of patients who will die without transplant are missed by standard KCC. Modified criteria with lactate and phosphate improve sensitivity to ~76–80%.
Population derivation: KCC was derived from a highly selected King's cohort in the 1980s, before modern ICU organ support. Spontaneous survival rates have improved since, meaning some patients meeting KCC criteria may now survive without transplant with optimal modern care.
Not a binary trigger: Clinical experience, trajectory of markers, and MDT assessment always adjunct the criteria.
Phosphate falling at 48–96h = hepatocyte regeneration confirmed — excellent sign. Ammonia trending down before Grade III HE = better prognosis.
AST <1000 IU/L by Day 4 without encephalopathy = likely recovery. Paracetamol aetiology itself is the best prognostic aetiology — higher spontaneous recovery than all other causes except hepatitis A and ischaemic hepatitis.
Hyperacute presentation (<7 days jaundice to HE), young age without comorbidities, and early NAC all predict transplant-free survival.
SNAP = Scottish and Newcastle Anti-emetic Pre-treatment for Paracetamol Poisoning. Endorsed by RCEM (Nov 2021), NPIS & BASL (Joint Statement May 2023), and TOXBASE. Replaces the old 3-bag 21-hour regimen as UK ED standard.
Pre-treatment: IV Ondansetron 4mg before starting (reduces ADRs)
Bag 1: 100 mg/kg in 200mL 5% dextrose over 2 hours
Bag 2: 200 mg/kg in 1L 5% dextrose over 10 hours
Total: 300 mg/kg over 12 hours
Pettie et al. eClinicalMedicine 2019 (n=3,340)
Hepatotoxicity: 3.6% vs 4.3% — no difference
ADRs (antihistamine needed): 2% vs 11%
ED stay: 17.8h vs 25.9h shorter · Ward: 2.6 vs 4.4 days shorter
⚠️ Anaphylactoid reaction (~2% with SNAP): Stop infusion → chlorphenamine 10mg IV ± nebulised salbutamol ± ondansetron. Restart at half rate once settled. Prior reaction is NOT a contraindication — withholding NAC is far more dangerous.
After SNAP completes at 12h, mandatory clinical reassessment. If ALT rising, INR rising, any clinical concern, or hepatotoxicity established — seek specialist hepatology input and extend NAC. Most UK liver units recommend:
Continue as continuous infusion: 150 mg/kg over 24 hours (≈6.25 mg/kg/h) until INR normalises
Maximum duration: 5 days. Beyond 5 days, evidence does not support additional hepatoprotective benefit from NAC alone.
Stop criteria: INR <2.0 and downtrending, ALT falling from peak, patient clinically recovering, paracetamol undetectable — whichever comes first
⏱️ Why 5 days? The NACSTOP trial (Wong et al. Hepatology 2019) and supporting observational data show no meaningful survival benefit from extending NAC beyond 5 days in ALF. If the patient hasn't recovered by Day 5, the pathway should pivot to transplant listing assessment, not continuation of NAC as treatment.
Used to determine if NAC is indicated based on serum paracetamol level at known time post-ingestion. Only valid for acute single-time ingestion.
UK treatment line: 100 mg/L at 4 hours (falling to ~15 mg/L at 15h). If the level plots above this line, give NAC.
Staggered/unknown time ingestion: If uncertain, treat. If paracetamol >2g taken in unknown circumstances, treat with NAC. Check TOXBASE for current guidance.
Note: Previously a separate high-risk treatment line existed for patients on enzyme inducers or with malnutrition. Current MHRA and TOXBASE guidance uses a single treatment line for all adults — high-risk patients are managed with the standard line, not an earlier threshold.
| System | Problem in ALF | Target / Intervention | Pitfalls |
|---|---|---|---|
| 🧠 Neurology | Cerebral oedema, ↑ICP, HE Grade I–IV | Na 145–150 (hypertonic saline) · Head 20–30° · Propofol sedation (preferred) · Avoid fever >37.5°C · Ammonia <100 μmol/L target via CRRT · Multimodal neuromonitoring (see below) | Avoid morphine (increases NH₃). Avoid benzodiazepines if possible (potentiate GABA). Don't sedate unnecessarily — monitor HE grade. |
| ❤️ Cardiovascular | Vasoplegic (↓SVR, ↑CO, ↓MAP) — similar to septic shock. Relative adrenal insufficiency common. | MAP >65 mmHg. Noradrenaline first-line vasopressor. Consider hydrocortisone if noradrenaline >0.3 μg/kg/min (relative adrenal insufficiency). Echo-guided haemodynamic assessment. | High CO may mask poor tissue perfusion — check lactate, CRT, urine output. Avoid large fluid boluses (aggravates cerebral oedema). |
| 🫘 Renal | AKI in ~50% — ATN, hepatorenal syndrome, or direct NAPQI nephrotoxicity | CRRT preferred over IHD (haemodynamic stability). For ammonia clearance, use high-flow CRRT: effluent rate ≥35–40 mL/kg/h (standard AKI rate of 20–25 mL/kg/h is insufficient for NH₃ clearance in ALF). Regional citrate anticoagulation where available. Target fluid balance neutral-to-negative once resuscitated. Avoid nephrotoxins. | Don't confuse HRS with ATN — response to fluids and vasoconstrictors differs. Urine sodium <20 mmol/L suggests HRS over ATN. |
| 🫁 Respiratory | Aspiration (HE), ARDS (sepsis/inflammation), pleural effusions (ascites) | Intubate early at Grade III HE. Lung-protective ventilation (6 ml/kg IBW, PEEP 5–8). Avoid hyperventilation (reduces CPP by causing cerebral vasoconstriction). | Hyperventilation temporarily lowers ICP but has rebound effect. Only use if acute herniation threatened. Keep PaCO₂ 35–40 mmHg. |
| 🩸 Haematology | Elevated INR, thrombocytopenia. Paradoxically re-balanced coagulation. | No FFP prophylactically. FFP only for active bleeding or invasive procedures. For procedures, consider ROTEM/TEG to guide specific factor replacement. PCC or recombinant FVIIa for refractory bleeding with high thromboembolic risk. | FFP obscures prognostic INR. Large FFP volumes cause fluid overload and ARDS. VTE prophylaxis is still needed despite elevated INR — these patients are not reliably anticoagulated. |
| 🍬 Metabolic | Hypoglycaemia (liver fails to maintain gluconeogenesis), hyponatraemia, hypokalaemia, hypophosphataemia | 10% dextrose infusion (target BM 5–10 mmol/L). Monitor BM 1–2 hourly in severe ALF. K, Mg, PO4 replacement. Na 145–150 for cerebral protection (hypertonic saline). | Severe hypoglycaemia can be catastrophic in already encephalopathic patients — treat aggressively. Avoid hypotonic fluids. |
| 🦠 Infection | Immune dysfunction → bacterial (60%) and fungal (30%) infections common and often covert | Low threshold for cultures and empirical antibiotics. Antifungal prophylaxis if prolonged ICU stay. SDD (selective decontamination) in some liver units. Daily surveillance cultures. | Infection can precipitate irreversible deterioration. Fever may be absent. WBC may be unreliable. CRP and procalcitonin less reliable in ALF. |
| 🧪 Nutrition | Catabolic state. But protein restriction (historically practised) worsens outcomes. | Early enteral nutrition (NG). Do NOT restrict protein — target 1.2–1.5 g/kg/day. Provide adequate calories (25–35 kcal/kg/day). Thiamine supplementation (Wernicke's risk). | Old teaching about protein restriction in HE is wrong and harmful. Muscle is the secondary ammonia detoxification organ — losing muscle mass worsens hyperammonemia. |
No single monitoring method is sufficient in ALF. Current practice at specialist liver centres uses a multimodal approach — combining non-invasive and (where indicated) invasive tools to detect rising ICP before clinical herniation.
| Modality | How It Works | Threshold / Target | Limitations |
|---|---|---|---|
| Optic Nerve Sheath Diameter (ONSD) Non-invasive · Bedside ultrasound |
Raised ICP transmits to the subarachnoid space around the optic nerve. The nerve sheath dilates when ICP rises. Measured by ultrasound 3mm behind the globe, in 2 planes. | ONSD >5.0–5.7mm (varies by protocol) suggests ICP >20 mmHg. Trend is more useful than single value. | Operator dependent. Cannot provide continuous values. Less reliable after Grade IV HE. Good as screening and trend monitor. |
| Invasive ICP Monitoring Bolt / EVD — Grade III/IV HE only |
Intraparenchymal bolt or external ventricular drain provides continuous ICP values. Allows calculation of CPP (MAP − ICP). | Target ICP <20 mmHg · CPP >50–60 mmHg. Treat sustained ICP >20 mmHg with mannitol (0.5–1g/kg) or hypertonic saline bolus. | Risk of intracranial haemorrhage in coagulopathic patient (~10% for bolt). Now reserved for Grade III/IV HE in patients being actively assessed for transplant. Requires ROTEM/TEG guidance for insertion safety. |
| Jugular Bulb Oximetry (SjO₂) Retrograde jugular venous catheter |
Catheter placed retrograde into internal jugular vein to measure cerebral venous oxygen saturation. Reflects balance between cerebral O₂ delivery and consumption. | Normal SjO₂ 55–75%. SjO₂ <55% = cerebral ischaemia / reduced delivery (↑CPP needed). SjO₂ >80% = cerebral hyperaemia / luxury perfusion or shunting. | Risk in coagulopathic patients (IJV haematoma). Requires careful positioning. Values affected by systemic hypoxia. Most useful as adjunct to ICP monitoring in transplant candidates. |
| Transcranial Doppler (TCD) Non-invasive · Bedside |
Ultrasound of middle cerebral artery blood flow velocity. Pulsatility Index (PI) rises with ICP. Can detect hyperaemia or critically reduced flow. | PI >1.4 suggests raised ICP. Can assess autoregulation status (static rate of autoregulation). | Temporal window not available in ~10–15% of patients. Operator dependent. Less validated in liver failure specifically. |
Plasma exchange (PEx) in ALF removes circulating toxins, DAMPs, cytokines and ammonia while replenishing coagulation factors, albumin, and complement — effectively providing temporary partial liver support.
High-volume PEx (8–12L/day x 3 days) in non-paracetamol ALF
Transplant-free survival: 58% vs 26% (control) — p=0.001
Benefit driven by non-transplanted patients. In those who received OLT, survival was similar.
FULMEN excluded paracetamol patients. No large RCT data in paracetamol-induced ALF specifically.
Used in UK specialist centres on case-by-case basis — particularly as bridge to transplant or when deterioration is rapid and ammonia is very high.
| Aspect | Detail |
|---|---|
| Volume | High-volume: 8–12 litres per session (15% of body weight). Standard volume: 1–1.5 plasma volumes. Evidence favours high-volume. |
| Frequency | Daily for 3 days (FULMEN protocol), then as needed based on clinical trajectory. |
| Replacement fluid | FFP (provides coagulation factors). Albumin alone not adequate — misses coagulation factor replacement. |
| Current UK use | Increasingly used at specialist liver centres as bridge to transplant. Also used to reduce ammonia rapidly in refractory cerebral oedema. Not yet standard of care outside specialist units. |
| Key limitations | Requires specialised apheresis equipment and expertise. Hypocalcaemia (citrate), hypothermia, transfusion reactions. No evidence in paracetamol-specific ALF. INR correction post-PEx masks transplant listing signals — coordinate assessment timing carefully. |
ALF creates a haemodynamic state that superficially resembles septic shock — but the mechanisms are distinct and the cardiac complications are under-recognised. Accurate assessment and targeted management are essential.
The Core Haemodynamic Pattern: ALF produces a hyperdynamic, vasoplegic circulation: markedly reduced systemic vascular resistance (SVR), increased cardiac output (CO), high heart rate, and a normal or low MAP despite apparently good cardiac function. This state is driven by systemic nitric oxide (NO) overproduction (from iNOS upregulation in response to DAMPs and endotoxin), accumulated bile acids causing splanchnic vasodilation, and direct myocardial depressant factors.
| Feature | ALF Pattern | Management Implication |
|---|---|---|
| SVR | ↓↓ (vasoplegia) | Noradrenaline first-line. Target MAP >65. Beware: high CO masks poor microcirculatory flow — check lactate, NIRS, CRT. |
| Cardiac Output | ↑ (hyperdynamic) | Echo to confirm. High CO despite vasopressor requirement = very poor prognosis. VTI and LVOT assessment critical. |
| Cirrhotic Cardiomyopathy (in ALF) | Diastolic dysfunction; blunted chronotropic response; QTc prolongation | Check 12-lead ECG; QTc >500ms increases arrhythmia risk. Avoid QT-prolonging drugs (haloperidol, azithromycin). |
| Relative Adrenal Insufficiency | Present in ~60% with severe ALF; cortisol axis blunted | Consider hydrocortisone 200mg/day if noradrenaline >0.3 μg/kg/min without clear infective cause. Short Synacthen test rarely practical in acute setting. |
| Pulmonary Hypertension (portopulmonary) | mPAP >25 mmHg — occurs in up to 6% of ALF/ALoCF; contraindication to OLT if severe | Bedside TTE/TOE to screen before transplant listing. mPAP >50mmHg = contraindication to OLT. Specialist pulmonary vasodilator management needed. |
| Stress Cardiomyopathy / Takotsubo | Under-recognised in ALF; apical ballooning pattern on echo; transient severe LV dysfunction | Usually reversible. Supportive management. May temporarily worsen shock state. Avoid inotropes unless truly needed — can increase afterload mismatch. |
Echo-Guided Haemodynamic Assessment — What to Look For: All ALF patients admitted to ICU should have a formal bedside TTE within 6 hours. Key questions: (1) Is the LV truly hyperdynamic (EF >65%, small end-systolic volume) or is EF preserved with occult diastolic dysfunction? (2) Is RV dilated — suggesting pulmonary hypertension or RV involvement from cytokine-mediated injury? (3) Is there pericardial effusion (common in ALF)? (4) What is LVOT VTI — guiding fluid and vasopressor responsiveness?
Vasopressor Choice: Noradrenaline remains first-line for vasoplegic shock in ALF. If noradrenaline >0.5 μg/kg/min, add vasopressin 0.03–0.04 U/min (acts on V1 receptors, independent of adrenergic pathway). Terlipressin has been used but carries risk of splanchnic ischaemia and worsening renal failure. Dopamine is avoided — associated with worse outcomes in cardiogenic and vasodilatory shock and worsens cardiac oxygen demand. Adrenaline is a last resort — its β1 effects increase metabolic rate and can exacerbate hyperlactataemia.
The "Macro-Micro Decoupling" Trap: A common and dangerous error in ALF haemodynamics is to be reassured by a normal or high CO on PiCCO or PA catheter while the patient is peripherally shut down and lactatemic. Cytokine-mediated arteriolar shunting and reduced mitochondrial O₂ utilisation (cytopathic hypoxia) mean that ScvO₂ can be paradoxically high (>80%) despite severe cellular hypoxia — the cells cannot extract oxygen. Always correlate with lactate trend, skin/CRT, urine output, and near-infrared spectroscopy (NIRS) if available.
Paracetamol is like a car that burns clean fuel. At normal doses, the exhaust is safe. In overdose, the engine overloads and starts producing toxic fumes (NAPQI) faster than the catalytic converter (glutathione) can handle. Once the converter is overwhelmed, the fumes burn the engine (liver cells) from the inside.
At therapeutic doses, 90–95% of paracetamol is safely conjugated via glucuronidation and sulphation. The remaining 5–10% is oxidised by CYP2E1 to NAPQI, which is immediately detoxified by glutathione (GSH). In overdose, conjugation pathways saturate, far more paracetamol is shunted to CYP2E1, NAPQI production exceeds GSH capacity, and free NAPQI covalently binds mitochondrial proteins in Zone 3 hepatocytes. This triggers ROS production, JNK activation, MPT pore opening, ATP depletion, and oncotic necrotic cell death. The key therapeutic window: GSH stores are not yet depleted for up to 8 hours post-ingestion — NAC replenishes GSH before this point to prevent necrosis.
Risk factors that lower the threshold: Chronic alcohol (↑CYP2E1 + ↓GSH), malnutrition/fasting (↓GSH precursors), enzyme inducers (rifampicin, phenytoin, carbamazepine — ↑CYP2E1).
The liver is like a factory floor — the workers nearest the supply dock (Zone 1) get the best resources and oxygen. The workers at the far end (Zone 3), near the rubbish chute (central vein), get the leftovers. Paracetamol is a poison that specifically targets the Zone 3 workers because they also have the most of the machine (CYP2E1) that converts paracetamol into its toxic form. Double jeopardy: most toxic production, least resources to survive it.
The hepatic acinus is divided into three zones by oxygen and nutrient gradient from the periportal triad (Zone 1) to the centrilobular central vein (Zone 3). Zone 3 has the highest CYP2E1 concentration, lowest oxygen tension, and greatest vulnerability to mitochondrial failure. Histologically, paracetamol toxicity produces centrilobular (Zone 3) necrosis — haemorrhagic necrosis spreading outward from the central vein. Zone 1 is preserved until very advanced disease — and Zone 1 hepatocytes are the source of liver regeneration, which is why paracetamol ALF has better regenerative potential than other causes.
AST and ALT are like smoke detectors — they go off when cells are burning. But AST is battery-powered (runs out fast, t½ ~18h), while ALT is hardwired (stays on much longer, t½ ~47h). The INR is more like the building's structural integrity — when the liver can't make clotting factors, the whole structure is weakening. Ammonia is the gas that leaks into the brain when the liver can't process waste. And phosphate falling after 48h means the firefighters (new hepatocytes) have arrived and are rebuilding.
AST (t½ ~18h): rises and falls faster than ALT. Do not use AST to determine when NAC can safely be stopped — it normalises before the liver is fully healed. ALT (t½ ~47h): more liver-specific, slower clearance — use this to determine recovery. INR: reflects coagulation factor synthesis (particularly Factor VII, t½ 4–6h) — the most sensitive early marker of synthetic failure. Rising INR on Day 4 is ominous. Do not correct prophylactically. Arterial ammonia: sample on ice, process immediately. >124 μmol/L predicts mortality. >200 = cerebral oedema risk. Arterial lactate: reflects both impaired lactate clearance by liver and systemic hypoperfusion. Part of modified KCC. Phosphate falling at 48–96h = hepatocytes regenerating — excellent prognostic sign. Phosphate remaining elevated = no regeneration.
The King's College Criteria are your threshold for saying "this liver cannot repair itself — this patient needs a new one." Think of it as checking whether the building is beyond repair. If the acid level in the blood is too low (pH <7.3), the liver's not processing waste. If all three — the clotting system (INR >6.5), the kidneys (creatinine >300), and the brain (Grade III/IV coma) — are failing simultaneously, the rest of the body is going with the liver.
Paracetamol KCC: pH <7.3 after resuscitation (sensitivity 67%, specificity 95%) OR all three of: INR >6.5, creatinine >300 μmol/L, Grade III/IV encephalopathy. Modified KCC adds: arterial lactate >3.5 mmol/L at 4h post-resuscitation OR >3.0 mmol/L at 12h (sensitivity 76%, specificity 97%); phosphate >1.2 mmol/L at 48–96h. Limitations: sensitivity only ~58–69% — not meeting criteria doesn't guarantee safety. Practical rule: Contact the transplant centre when Grade II HE develops or INR is rising — don't wait for full criteria to be met, because finding and allocating an organ takes time. Post-transplant 1-year survival >80%.
The liver normally disposes of ammonia by turning it into urea. When the liver fails, ammonia builds up in the blood and crosses into the brain. Brain cells called astrocytes try to deal with it by converting it to glutamine. But glutamine accumulates and acts like a sponge — it draws water into the astrocytes, making them swell. This swells the brain inside the skull (which can't expand), raising pressure inside, and eventually cutting off blood flow to the brain. This is why patients go from confused to comatose to dying.
The mechanism: ammonia → astrocyte glutamine synthesis (via glutamine synthetase) → glutamine accumulation → osmotic astrocyte swelling → cytotoxic cerebral oedema → ↑ICP → ↓CPP → cerebral ischaemia → herniation. Simultaneously, systemic inflammation disrupts the BBB (vasogenic oedema), and loss of cerebral autoregulation means MAP fluctuations directly impact ICP. ICU management: target Na 145–150 mmol/L (hypertonic saline) to draw water out of astrocytes osmotically; head 20–30°; avoid fever (>37.5°C doubles glutamine production rate); CRRT to clear ammonia; propofol sedation. Aim ICP <20 mmHg, CPP >50 mmHg. Invasive ICP monitoring is now reserved for Grade III/IV encephalopathy in transplant candidates only.