• Change in plasma concentration of triglycerides, total cholesterol, HDL-cholesterol, free fatty acids, apolipoproteins, lipoprotein (a) [ Time Frame: within 5 days ] [ Designated as safety issue: Yes ]
  

• Incidence of hyperglycaemia [ Time Frame: within 5 days ] [ Designated as safety issue: Yes ]
  
• Alteration of liver function [ Time Frame: within 5 and 28 days ] [ Designated as safety issue: Yes ]
  
• Gastrointestinal intolerance [ Time Frame: within 5 days ] [ Designated as safety issue: No ]
  
• Gastrointestinal bleeding [ Time Frame: within 5 and 28 days ] [ Designated as safety issue: Yes ]
  
• Septic complications [ Time Frame: within 5 and 28 days ] [ Designated as safety issue: Yes ]
  
• Occurence of new organ dysfunction [ Time Frame: within 5 and 28 days ] [ Designated as safety issue: No ]
  
• Length of stay in the intensive care unit (ICU) [ Time Frame: within 5 and 28 days ] [ Designated as safety issue: No ]
  
• Mortality [ Time Frame: within 5 and 28 days ] [ Designated as safety issue: No ]
  

An impaired lipid metabolism is often found in patients with liver disease and is assumed to influence the prognosis. The central role of lipid metabolism in the pathophysiology of fatty liver disease and steatohepatitis is well established. In cirrhotic patients, serum lipid levels are mostly decreased and related to the severity of liver failure; in addition, the structure and composition of lipoproteins differ from that of healthy individuals. A reduction in high-density lipoproteins has been associated with higher cytokines levels and a poorer clinical outcome in septic patients. Furthermore, the oxidative stress induced by septic complications in critically ill patients with liver failure may lead to further hepatocellular injury and activation of systemic inflammation cascade.

In this setting, the influence of nutrition on lipid metabolism may have an impact on the severity of liver failure and associated complications. Although plasma clearance and oxidation of lipids were considered to be normal in the majority of patients with cirrhosis, most previous studies only reported the effects of an oral ingestion or parenteral infusion of lipids during a few hours.

The present randomized controlled trial will be conducted in a subgroup of patients enrolled in the SELLIFA-01 prospective study (NCT00488917). The purpose of the nutritional trial is to determine the tolerance on lipid metabolism and the safety of isocaloric short-term parenteral nutrition as compared to enteral nutrition in critically ill septic and non septic patients with liver failure. The nutrition will be delivered continuously for 5 days and will provide a daily energy supply corresponding to current resting energy expenditure as determined by indirect calorimetry, with 35% of total energy requirements as lipids, 15% as proteins (maximum 1.2g/kg ideal body weight/day), and 50% as dextrose. A tight glucose control strategy will be implemented to avoid hyperglycemia.

The trial is designed to randomly assign 15 patients in each interventional group in order to detect more than 25% increase in plasma triglycerides levels with 80% statistical power for two-tailed type I error of 5%.

Inclusion Criteria:

• All consecutive patients with a diagnosis of chronic liver failure;
• Planned total nutritional support;
• Adult patient aged 18 years or above, and less than 85;
• Admission to the ICU for an expected period of > 24 hours;
• Informed consent of the patient or nearest relative.

Exclusion Criteria:

• Absolute contra-indications to enteral nutrition : total mechanical ileus, persistent vomiting, high output fistulas, uncertain gastrointestinal anastomosis;
• Absolute contra-indications to parenteral nutrition : severe hypertriglyceridemia > 6 mmol/l (> 545 mg/dL), severe diabetic ketoacidosis;
• Age less than 18 years or more than 85;
• Pregnancy, including HELLP syndrome;
• Active malignancy with metastases (localized hepatocellular carcinoma is not an exclusion criteria);
• Systemic chemotherapy in the last 4 weeks (trans-arterial chemo-embolisation for localized hepatocellular carcinoma is not an exclusion criteria);
• Acquired immunodeficiency syndrome and antiretroviral therapy;
• Refusal of the patient or nearest relative.