Article Text
Abstract
Introduction Acute liver failure (ALF) is a rare but devastating clinical condition, a common cause of which is paracetamol overdose. Advances in treatment options for ALF have been hampered by the lack of a representative large animal model.
Aim This study was designed to develop a clinically relevant, fully monitored and managed model of ALF that would develop the clinical, biochemical, haemodynamic and inflammatory characteristics of the human equivalent.
Method Initial studies were performed in 36 landrace pigs to define a dose, which resulted in 100% mortality. We developed the model in 3 female pigs that were subjected to full intensive care but without any paracetamol administration and 8 pigs that were treated with paracetamol. After an overnight fast, 11 pigs [weighing 35-45 kg] were intubated and ventilated under general anaesthetic for the duration of the experiment. Catheters were placed for blood pressure monitoring, haemofiltration, urine measurement and triple lumen central intravenous lines for fluids, drugs and sampling. An ICP bolt was inserted and a separate catheter for cerebral microdialysis. Placebo (n=3) or a loading dose (0.25 g Kg-1) then hourly bolus? of paracetamol (n=8) were given via the jejunostomy to keep the serum paracetamol concentration 350 - 450 mg dl-1. Paracetamol was stopped when Quick index of 30% was reached. Animals were supported with fluids, glucose, fresh frozen plasma, inotropes, renal haemofiltration and mechanical ventilation until time of death.
Results The paracetamol overdosed animals developed typical changes of ALF manifested by attenuated mean arterial pressure requiring large amounts of fluid resuscitation (2.5 litres per hour) and inotropic support (noradrenaline; 15-150 ?g kg-1 hr-1) and increased intracranial pressure (p<0.001 compared to sham operated pigs). There were increases in PT to >160 s, creatinine (0.58±0.17 vs 1.45±0.13 mg kg-1, p<0.01), ammonia (41.6±7.4 vs 120.6±48.1 ?M L-1), lactate (2.1±0.4 vs 7.1±1.7 mM L-1, p<0.05) together with decreases in albumin (23±2 vs 2±0.4 mg L-1, p<0.01), urine (91±14 vs 6±5.5 ml hr-1 p<0.01) compared to baseline. The animals developed progressive albumin dysfunction (IMAR 0.014±0.002 vs 0.45±0.17, p<0.01) and endotoxaemia (0.57±0.17 vs 2.0±0.18 EU ml-1). They required increased ventilatory support and death was by respiratory failure following raised ICP. The mean time from paracetamol administration to ALF was 32±4.4 and from ALF to death 15.8±2.4 hrs.
Conclusion We have developed a stable, fully monitored and managed model of paracetamol induced ALF which exhibits the clinical, haemodynamic, biochemical and inflammatory characteristics of ALF that is suitable for interventional studies of novel therapies for this devastating rare disease.