King's College Criteria for Liver Transplantation in Acute Liver Failure
The King's College Hospital criteria remain the most widely used prognostic tool for determining transplant candidacy in acute liver failure, though newer evidence suggests MELD score may offer superior sensitivity. 1
Acetaminophen-Induced ALF Criteria
The King's College criteria indicate transplant candidacy when any one of the following is met: 1, 2
OR
- All three of the following simultaneously: 1, 2
- Prothrombin time >100 seconds AND
- Serum creatinine >300 μmol/L (approximately 3.4 mg/dL) AND
- Grade III or IV hepatic encephalopathy
Enhanced Prognostication with Lactate
Arterial lactate levels significantly improve early identification of poor prognosis in acetaminophen-induced ALF: 3
- Lactate >3.5 mmol/L at 4 hours after admission predicts mortality with 67% sensitivity and 95% specificity 3
- Lactate >3.0 mmol/L at 12 hours after fluid resuscitation predicts mortality with 76% sensitivity and 97% specificity 3
- Combining lactate with King's College criteria increases sensitivity from 76% to 91%, substantially reducing missed transplant candidates 3
Non-Acetaminophen-Induced ALF Criteria
The King's College criteria indicate transplant candidacy when any one of the following is met: 1, 2
OR
- Any three of the following five variables (irrespective of encephalopathy grade): 1, 2
- Etiology: non-A, non-B hepatitis (indeterminate), halothane hepatitis, or idiosyncratic drug reactions
- Age <10 years or >40 years
- Jaundice to encephalopathy interval >7 days
- Prothrombin time >50 seconds
- Serum bilirubin >300 μmol/L (approximately 17.5 mg/dL)
Performance Characteristics and Limitations
The King's College criteria demonstrate high specificity (82-95%) but limited sensitivity (58-68%) for predicting mortality in non-acetaminophen ALF: 4, 5
- Positive predictive value is excellent (70-100%), meaning patients meeting criteria truly need transplantation 1, 4
- Negative predictive value is poor (25-94%), meaning many patients not meeting criteria may still die without transplant 1, 5
- Specificity is highest (93%) when applied to patients with grade III-IV encephalopathy 4
- Sensitivity has declined in studies published after 2005 (58% vs 85% pre-1995), suggesting modern medical management may alter performance 4
Current Guideline Recommendations
The 2017 AGA guidelines conditionally recommend using MELD score over King's College criteria as the primary prognostic tool, with a MELD cutoff of 30.5 predicting transplant need: 1
- MELD demonstrates superior sensitivity (77% vs 61%) compared to King's College criteria 1
- King's College criteria maintain higher specificity (86% vs 72%) 1
- The 2005 AASLD position paper notes that prognostic scoring systems do not adequately predict outcome and reliance entirely upon these guidelines is not recommended 1
Special Considerations
Wilson disease and mushroom poisoning warrant urgent transplant evaluation regardless of criteria, as these etiologies carry particularly poor prognosis: 1, 2
- Patients with grade III-IV encephalopathy at presentation have significantly worse outcomes (33% transplant-free survival vs 52% for grade I-II) 2
- Early referral to transplant centers is critical for all patients with non-acetaminophen ALF or progressive coagulopathy, even before criteria are fully met 1, 2
Critical Pitfalls to Avoid
- Do not wait for full criteria fulfillment before transplant center referral, as delayed referral worsens outcomes through development of malnutrition, hepatorenal syndrome, and advanced UNOS status 1
- Do not rely solely on King's College criteria to exclude transplant candidacy, given the poor negative predictive value—patients not meeting criteria may still require transplantation 1, 5
- Consider etiology-specific prognosis: acetaminophen, hepatitis A, shock liver, and pregnancy-related ALF show 50% transplant-free survival versus 25% for other etiologies 1
- Incorporate lactate measurements in acetaminophen cases to identify high-risk patients earlier than King's College criteria alone 3