Diagnostic Criteria for Liver Failure
The diagnostic criteria for liver failure depend critically on whether you are evaluating acute liver failure (ALF) in a patient with previously normal liver function versus acute-on-chronic liver failure (ACLF) in a patient with underlying chronic liver disease or cirrhosis.
Acute Liver Failure (ALF)
For patients with previously normal liver function, diagnose ALF when the following are present 1, 2:
- Prothrombin time ≤40% of standardized values OR INR ≥1.5
- Hepatic encephalopathy (any grade)
- Symptom onset within 8 weeks (some definitions use up to 26 weeks)
- No pre-existing chronic liver disease
Key Laboratory Findings
- Elevated transaminases (AST/ALT)
- Elevated bilirubin
- Coagulopathy as defined above
ALF Subtypes
ALF is further classified based on timing of encephalopathy development 2:
- Acute type: Grade II or higher encephalopathy within 10 days of symptom onset
- Subacute type: Grade II or higher encephalopathy between 11-56 days after onset
Prognostic Assessment
The AGA recommends using MELD score ≥30.5 rather than King's College Criteria for prognosis, with higher scores predicting need for liver transplantation 1.
Acute-on-Chronic Liver Failure (ACLF)
For patients with chronic liver disease or cirrhosis, the 2024 AASLD guidance provides the most current framework. ACLF requires ALL three of the following minimum critical components 3:
- Acute onset with rapid clinical deterioration
- Liver failure defined by:
- Elevated bilirubin (cutoffs vary: 5-12 mg/dL in different definitions)
- Elevated INR (≥1.5-2.5 depending on criteria)
- At least one extrahepatic organ failure:
- Neurologic (hepatic encephalopathy)
- Circulatory (shock requiring vasopressors)
- Respiratory (requiring mechanical ventilation)
- Renal (creatinine ≥2.0 mg/dL or requiring dialysis)
Important Nuances About ACLF Definitions
There are three major competing definitions that capture different disease stages 3, 4:
- APASL criteria: Most sensitive, captures early-stage disease (may include patients who won't progress)
- EASL-CLIF criteria: Intermediate stage, includes grading system (ACLF-1, ACLF-2, ACLF-3) based on number of organ failures; best validated for mortality prediction 4
- NACSELD criteria: Most specific, requires ≥2 extrahepatic organ failures; captures late/preterminal stage
The EASL-CLIF criteria demonstrated superior performance for predicting 28-day and 90-day mortality (AUC 0.710 and 0.760 respectively) compared to AARC and NACSELD criteria 4.
Practical Approach to ACLF Diagnosis
When evaluating a cirrhotic patient with acute decompensation:
- Document baseline liver function (Child-Pugh score ideally 5-9 for ACLF diagnosis)
- Identify acute insult (infection, alcohol, GI bleeding, drug toxicity, viral hepatitis exacerbation)
- Assess liver failure markers:
- Bilirubin ≥5.0 mg/dL
- INR ≥1.5
- Systematically evaluate for organ failures:
- Brain: Grade II-IV encephalopathy
- Kidney: Creatinine ≥2.0 mg/dL or dialysis requirement
- Circulation: Mean arterial pressure <70 mmHg or vasopressor need
- Lungs: PaO2/FiO2 <200 or mechanical ventilation
- Coagulation: INR >2.5 or platelets ≤20,000
Critical Diagnostic Testing Recommendations
Based on AGA guidelines for ALF workup 1:
- Test for HSV (HSV DNA) and treat empirically with acyclovir if positive (conditional recommendation)
- Test for hepatitis E in pregnant women presenting with ALF
- Consider Wilson's disease testing only with high clinical suspicion (low alkaline phosphatase, hemolysis, low ceruloplasmin), though positive predictive value is low 1, 5
- Autoantibody testing for autoimmune hepatitis should be performed
- Routine liver biopsy is NOT recommended in ALF 1
Common Pitfalls
- Do not confuse simple decompensated cirrhosis with ACLF: ACLF requires both hepatic dysfunction AND extrahepatic organ failure with acute deterioration
- Timing matters: The "acute" component must show rapid deterioration (typically within 28 days of precipitant)
- Different definitions capture different populations: Use EASL-CLIF criteria for best mortality prediction in clinical practice 4
- Laboratory cutoffs are not absolute: While specific thresholds exist, clinical judgment about rapidity of decline is essential 3