Drug-Induced Liver Injury and Acute Liver Failure: Acute Liver Failure and the King's College Criteria

0:00Welcome to Board Pearls. This is episode two of two of the Drug-Induced Liver Injury and Acute Liver Failure chapter, in the Liver Disease module. This episode is acute liver failure: the definition that puts coagulopathy and encephalopathy together in under twenty-six weeks without prior cirrhosis, the etiology distribution led by acetaminophen, the two special causes with their own recognition cues, and the King's College Criteria that triage who needs a transplant.

0:28The prior episode treated the drug as the problem; this one treats the failing organ as the problem, which changes the questions from which agent did this to whether the liver can still recover on its own, how much time you have to decide, and at what point a different liver is the only answer. Start with the definition, because every downstream decision rests on it. Acute liver failure is rapid loss of hepatic synthetic and detoxifying function in a previously healthy liver, and the criteria are four: an INR of at least one and a half, hepatic encephalopathy of any grade, total illness duration under twenty-six weeks, and no pre-existing cirrhosis. Each is doing work. The INR threshold is the synthetic floor, because below it the liver is still making enough clotting factors that the patient isn't in true synthetic failure. Encephalopathy is the second pole, because coagulopathy without encephalopathy is severe acute hepatitis, not acute liver failure, and the two have to be present together because cerebral involvement is what signals the failing organ is now affecting the brain. The twenty-six-week boundary separates this from acute-on-chronic liver failure, which behaves differently, and the no-cirrhosis rule keeps the label honest, because a patient with established cirrhosis who decompensates isn't in acute liver failure no matter how high the INR climbs. Three exceptions get relabeled as acute liver failure when the other criteria are met: autoimmune hepatitis presenting fulminantly for the first time, Wilson disease, which by definition has chronic copper deposition but presents with sudden hepatocyte death and hemolysis, and hepatitis B reactivation during immunosuppression, because all three behave as acute liver failure and need the same urgent management.

2:12Why does the distinction from acute-on-chronic failure matter so much? Because cerebral edema with intracranial hypertension and herniation occurs in acute liver failure and not in the chronic form, since the chronic patient's brain has had time to adapt to ammonia while the acute patient's has not, so ammonia loading produces astrocyte swelling and globally swells the brain: same hyperammonemia, different cerebral consequence, and the triage of a confused jaundiced patient hinges on which entity you're looking at. Encephalopathy is graded one through four: grade one mild confusion with a reversed sleep-wake cycle, grade two lethargy, asterixis, and disorientation to time, grade three somnolence with marked asterixis and disorientation to place, and grade four coma unresponsive to pain, with cerebral edema risk rising sharply at grade three or four, a cutpoint that does real work in the King's College Criteria later, while grade two and below is still inside the window where transfer, optimization, and listing can happen with the patient awake.

3:15Now the etiology distribution, which is board content because etiology drives management. Acetaminophen is the leader at nearly half of cases, followed by indeterminate, meaning presumed but unproven, then non-acetaminophen drug injury, then viral hepatitis predominantly from hepatitis B, then autoimmune hepatitis, ischemic liver, and Wilson disease, with pregnancy-associated causes and Budd-Chiari contributing small fractions. The distribution matters because it maps to prognosis: acetaminophen has the best spontaneous recovery, most patients recovering without transplant because the injury stops as soon as the drug is metabolized and the antidote is on board, while indeterminate and idiosyncratic drug injury have the worst spontaneous recovery, only a minority, because the mechanism is ongoing immune injury that doesn't switch off cleanly. So etiology isn't a label, it's a probability of recovery without transplant, and the King's College Criteria are calibrated to that probability. In the developing world the distribution flips toward viral hepatitis, with hepatitis E in pregnant women producing high mortality, so the board move is to keep viral testing prominent in any workup regardless of the home epidemiology.

4:24Before the special causes, fix the non-cerebral management. Hypoglycemia is constant because hepatic gluconeogenesis is gone, so continuous dextrose is maintenance. Coagulopathy with a high INR isn't corrected with plasma unless the patient is bleeding or facing a procedure, and the reason is prognostic, because the INR is the synthetic readout the King's College Criteria and the transplant team track, so correcting it cosmetically removes the signal, with vitamin K given once to exclude reversible deficiency. Renal failure complicates a large fraction of cases and gets continuous renal replacement therapy, sepsis is common and often empirically covered because the systemic inflammatory response of the syndrome mimics infection, and the circulation is hyperdynamic so vasopressors are often needed. Cerebral edema is the syndrome-specific lethal complication, from hyperammonemic astrocyte swelling: ammonia crosses into the brain, astrocytes convert it to glutamine, glutamine is osmotically active, the astrocytes and brain swell, pressure rises, and herniation kills, with an arterial ammonia above roughly a hundred fifty to two hundred the threshold where the risk climbs sharply. Management is layered: head of bed at thirty degrees, sedation to drop cerebral demand, ventilation targeting mild hypocapnia, hypertonic saline to keep the sodium in the high-normal range to pull water out of swollen astrocytes, mannitol for breakthrough pressure, continuous renal replacement if ammonia stays very high because the circuit clears it faster than the failing liver, and therapeutic hypothermia as a bridge in refractory cases, while lactulose isn't very useful because the timeline is too short and invasive pressure monitoring is largely abandoned because the bleeding risk in a coagulopathic patient outweighs the information.

6:07Now the two special causes plus the empiric antidote question. Wilsonian acute liver failure is the fulminant presentation of previously undiagnosed Wilson disease, almost always in young patients, and without transplant it's universally fatal. The mechanism explains the fingerprint: hepatocytes that have quietly accumulated copper for years suddenly die and dump copper into the circulation, producing massive oxidative hemolysis, so the patient has three things at once, synthetic failure, hemolysis from free copper, and cerebral involvement from both ammonia and copper. The fingerprint is a constellation to recognize as a unit: a Coombs-negative hemolytic anemia with high indirect bilirubin, because the hemolysis isn't autoimmune; an alkaline-phosphatase-to-bilirubin ratio under four, because the failing Wilson liver makes little alkaline phosphatase while hemolysis drives the bilirubin high so the ratio collapses; an AST-to-ALT ratio over two, because the release pattern differs from typical hepatitis; a low uric acid, because Wilson produces a Fanconi-like tubular dysfunction that wastes it; and Kayser-Fleischer rings, present in most patients with neurologic Wilson but only about half of hepatic-only Wilson, so their absence doesn't exclude it, though the slit-lamp exam is still done because a positive finding confirms quickly. The acute lab diagnosis is harder than the chronic one, because ceruloplasmin is unreliable acutely, an acute-phase reactant that can be falsely elevated by inflammation and also reduced by synthetic failure, so it can go either way, while serum free copper is the more reliable acute marker and the urinary copper is elevated, with genetic testing confirmatory but slow and hepatic copper on biopsy diagnostic but rarely safe in coagulopathic failure. Treatment is urgent transplant, with penicillamine, which is first-line for chronic Wilson, avoided in the fulminant setting because mobilizing copper too fast can worsen the acute phase, and plasmapheresis or albumin dialysis serving as bridges by removing free copper. So the vignette is the young patient, often a woman in her late teens or early twenties, with new jaundice, fatigue, and confusion over days to weeks, whose labs show a very low alkaline phosphatase relative to a high bilirubin, a wide AST-over-ALT ratio, Coombs-negative hemolysis, and low uric acid, with the rings on slit lamp as the closer, where the answer is to list urgently and bridge with plasmapheresis.

8:20Herpes acute liver failure is rare but high-stakes, producing fulminant hepatitis in two populations, pregnant women especially later in pregnancy and immunocompromised patients, with pregnancy accounting for a notable share of disseminated cases. The mechanism is primary or reactivated infection escaping immune control to produce direct cytopathic injury with the characteristic intranuclear inclusions. The fingerprint is the most distinctive in this section because it breaks the rule that liver failure equals jaundice: the enzymes are very high, often twenty-five to forty times normal with transaminases in the thousands, but the bilirubin is normal or low, the anicteric liver failure pattern, which is the textbook clue. Fever is almost universal, the vesicular rash is absent in most patients so a clean skin exam doesn't exclude it, and encephalitis can coexist. Diagnosis is by PCR on serum or tissue, but the board move is not to wait for it: empiric intravenous acyclovir at five to ten milligrams per kilogram every eight hours is started in any pregnant or immunocompromised patient with indeterminate acute liver failure, because antemortem diagnosis is missed in the majority and delay drives mortality from very high without acyclovir to dramatically better when it's started early. So the vignette is the pregnant woman or transplant recipient with very high transaminases, a normal bilirubin, fever, and altered mental status with no obvious cause, where the next step is empiric acyclovir while sending the PCR.

9:50The third empiric maneuver is N-acetylcysteine in non-acetaminophen acute liver failure, a different use than the antidote role. In the randomized data, it improved transplant-free survival in patients who presented with early-grade encephalopathy, grade one or two, but not in those with grade three or four, so it's most useful inside the early-grade window. The mechanism isn't the antidotal one, since NAPQI isn't in play, but likely anti-inflammatory and microcirculatory, improving hepatic blood flow and dampening systemic inflammation, so any patient with indeterminate or drug-induced acute liver failure and grade one or two encephalopathy gets empiric N-acetylcysteine at the same dosing as for acetaminophen, continued until improvement or transplant. Corticosteroids don't have that supporting role in drug-induced acute liver failure and may increase infection risk, so they're reserved for autoimmune hepatitis and severe checkpoint-inhibitor hepatitis, not routine drug-induced failure.

10:51That sets up the King's College Criteria, which identify patients unlikely to recover without transplant by capturing the threshold of synthetic, acid-base, renal, and neurologic dysfunction beyond which spontaneous survival becomes rare, so meeting them triggers urgent transfer and top-priority listing, and they split into acetaminophen and non-acetaminophen versions because the prognostic factors differ. For acetaminophen, either of two paths is met: a single value, an arterial pH under seven point three after adequate volume resuscitation, which captures profound lactic acidosis with the resuscitation qualifier excluding acidosis merely from hypoperfusion; or all three of grade three or four encephalopathy, an INR over six and a half, and a creatinine over three point four, identifying the patient with severe coagulopathy, severe encephalopathy, and renal failure at once, with some references adding an elevated arterial lactate as an additional acidosis criterion. For non-acetaminophen, either of two paths: an INR over six and a half regardless of other features, high enough that meeting it alone predicts non-recovery; or any three of five features, age under ten or over forty, an indeterminate or drug-induced etiology, jaundice for more than seven days before encephalopathy onset which captures the subacute course, an INR over three and a half, and a bilirubin over seventeen and a half, which together encode the subacute, drug-driven, prolonged-jaundice phenotype with the worst spontaneous recovery. The non-acetaminophen criteria are more permissive on each feature but require three at once because that population is heterogeneous and the high-mortality phenotype is recognized by the combination, while the acetaminophen criteria can lean on a single pH or a single trio because acetaminophen mortality tracks acidosis and multi-organ failure tightly.

12:37Meeting the criteria triggers top-priority listing, the highest tier on the allocation system, superseding the calculated score, and it requires ICU admission, adult age, and a life expectancy without transplant under a week, plus at least one of ventilator dependence, renal replacement, or an INR over two, with the label time-limited and re-listed weekly, and that aggressive prioritization is what makes the transplant possible inside the narrow window. The transfer decision should be made earlier than the listing decision, with the goal of having the patient at a transplant center while there's still time to optimize, so any non-acetaminophen patient with early encephalopathy, an INR over one and a half, or rising bilirubin is transferred, and for acetaminophen the thresholds are even lower, any encephalopathy, persistent acidosis after resuscitation, an INR over two not responding to vitamin K, or any rise in creatinine, because the acetaminophen patient can deteriorate within hours. The contraditions to transplant shape the listing: uncontrolled sepsis with pressor-dependent shock disqualifies until controlled, severe pulmonary disease and extrahepatic malignancy are contraindications, and irreversible brain injury from prolonged intracranial hypertension is a contraindication, which is why aggressive pressure management matters not just for survival but for eligibility, while non-compliance, lack of support, and active substance use are relative. Other systems exist alongside King's College, a continuous severity score not used as a transplant trigger in this setting, the Clichy criteria using a clotting-factor level with encephalopathy, and a study-group index combining several variables, but the boards focus on King's College, which has good specificity but imperfect sensitivity, which is why it's used as a threshold rather than a strict rule and the transplant team's judgment decides borderline cases.

14:26So the triage algorithm comes together: confirm acute liver failure by the four-criterion definition, define etiology and apply the etiology-specific empiric maneuvers, N-acetylcysteine for acetaminophen and for non-acetaminophen with early-grade encephalopathy, acyclovir for any pregnant or immunocompromised patient with indeterminate failure, and recognize the Wilsonian fingerprint and move toward transplant. Transfer to a center before the criteria are met, manage the organ-specific complications especially cerebral edema in grade three or four encephalopathy, and apply the King's College Criteria as the listing trigger. The way to think about it is layered: acute liver failure is a defined entity with a clean diagnostic frame, the etiology maps directly to prognosis, two cause-specific recognition patterns demand empiric therapy before the workup is complete, and a triage tool tells you when a different liver is the only answer, so the candidate who moves through that sequence in order gets the question right whether the etiology is acetaminophen, Wilson, herpes, or indeterminate.

15:25That closes the drug injury and acute liver failure chapter. The next chapter shifts to the autoimmune and cholestatic liver diseases, sorting autoimmune hepatitis from primary biliary cholangitis from primary sclerosing cholangitis, treated respectively with steroids and azathioprine, with ursodeoxycholic acid and the new second-line agents, and with surveillance for cholangiocarcinoma, and the organizing question being how the pattern across antibodies, cholangiogram, and biochemistry sorts three diseases whose treatments and surveillance burdens diverge sharply.

15:56For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode two of two of chapter nineteen, and I'll see you in the next one.