Beyond KCH selection and options in acute liver failure.

Area of Science:

• Hepatology
• Medical Informatics
• Prognostic Modeling

Background:

• The King's prognostic criteria (1989) for acute liver failure (ALF) distinguished hyper-acute cases but had limited sensitivity.
• Subsequent models like MELD and composite scores improved sensitivity but lacked high specificity.
• Advancements in computing enable sophisticated algorithms for outcome prediction in ALF.

Purpose of the Study:

• To review the evolution of prognostic criteria for acute liver failure (ALF).
• To describe recent dynamic and etiological models for predicting ALF outcomes.
• To discuss the impact of improved medical management and novel non-transplant therapies on liver transplant decisions.

Main Methods:

• Review of historical and contemporary prognostic models for ALF.
• Description of a dynamic model for paracetamol-induced ALF using sequential data.
• Presentation of the US Acute Liver Failure group model for predicting transplant-free survival across all etiologies.

Main Results:

• Recent models demonstrate improved accuracy in predicting spontaneous survival in ALF.
• Medical management, especially in hyper-acute ALF, now rivals transplant survival rates.
• New non-transplant treatments like plasma exchange show promise.

Conclusions:

• Evolving prognostic models offer better prediction of ALF outcomes.
• Improved medical and regenerative therapies challenge traditional reliance on liver transplantation.
• Further research into novel treatments is crucial for ALF management.