A 28-year-old lady presented to the medical team jaundiced with cramping upper abdominal pain and multiple episodes of vomiting over the previous day. She admitted ingesting Paracetamol 8 grams 3 days previously (staggered throughout the day) ostensibly to treat a frontal headache. She had been commenced on Citalopram 1 week previously for depression but denied taking any intentional overdose. On examination, she was slightly drowsy but GCS 15. HR 109, BP 136/92. Sats 98%on air. Her chest was clear, she was warm peripherally but jaundiced with some epigastric and RUQ tenderness on palpation. Her urine output was 10-20ml/hr.
Full Blood Count revealed Hb 152, WCC 24.7, Plats 301. She was in acute liver failure with Bil 189, AST 22970, ALT 13040, ALP 426 and coagulopathic with PT 82, APTT 72, Fib 0.7 Urea 5.7, Cr 193. Paracetamol and Salicylate were not detected. She was not acidotic with H+ 35, OCI2 3.7, pO2 17, Bic 20, BE –3. Lactate 7.1.
She was commenced on N-acetylcysteine and transferred to Critical Care. She was reviewed by the Hepatobiliary surgical team and placed on the super-urgent list for liver transplant.
On Day 2, she became encephalopathic with GCS E3M5V5 and she was intubated and ventilated.Her PT had increased to 168 (INR >15) and she became anuric. She commenced FFP and Cryoprecipitate transfusions that improved her PT to 17, APTT 34 and Fibrinogen 1.5. An Intracranial Pressure (ICP) monitor was inserted and an opening pressure of 19mmHg was found. 2 hours post-insertion, it was noticed that her right pupil had increased in size from 2mm to 4mm and was poorly reactive. ICP remained at 16 and pCO2 4.1.
A brain CT showed a large haematoma in the right frontal region around the ICP bolt (which was not in the brain parenchyma but sitting in the skull) and mass effect with 5mm midline shift. There was also some lack of grey-white matter differentiation and sulcal effacement in keeping with diffuse oedema and mass effect.
INR was 1.7 and so further FFP was given. She was discussed with the neurosurgical registrar (in a separate hospital) who advised they would not drain at present but he would discuss with his Consultant and call back.
Soon after, her right pupil increased to 8mm and the left to 7mm. Repeat CT brain showed slightly increased right frontal haematoma with 6mm midline shift and global oedematous cortical changes but no herniation. The ICP readings were thought to be inaccurate due to proximal placement and she was medically treated for raised ICP with hypertonic saline, mannitol and then therapeutic hypothermia. Despite this treatment, her pupils were fixed and dilated and so a thiopentone infusion was commenced.
The neurosurgeons advised that they would insert a further ICP monitor when INR <1.3 and so further FFP was given. An ICP bolt was inserted and the opening pressure was >120.
Discussions between the ICU, hepatobiliary and neurosurgical teams confirmed that she had a non-survivable injury and so this was discussed with her family. She was rewarmed, paralysis and then sedation were discontinued, brain stem death testing took place and she was extubated in the presence of her family. She died on Day 3 and was referred to the Coroner for further investigation.
What is the rationale for measuring ICP in acute liver failure?
Acute Liver Failure (ALF) is defined by the occurrence of acute hepatitis (an elevation in AST/ALT) with elevation of INR to >1.5, a degree of mental alteration (hepatic encephalopathy; HE) and an illness of less than 26 weeks duration in the absence of pre-existing liver disease (1).
It is subdivided based upon the time taken to develop HE after the first appearance of jaundice: “hyperacute” ALF is HE within 0 to 7 days of jaundice, “acute” is from 8 to 28 days and “subacute” is between 29 and 84 days (2).
The heterogenous nature of (and lack of an International Classification of Disease code for) ALF means that the incidence is likely under-reported. However, it is thought to affect approximately one in every 2500 patients in the United States each year; amounting to 2800 cases per annum and 3.5 deaths per million population (3). In Scotland, paracetamol-induced ALF alone is thought to result in 8.4 deaths per million population per year (4).
Aetiology also varies with geographical location, with paracetamol poisoning and idiosyncratic drug reactions dominating in the UK and US, whereas viral aetiologies such as Hepatitis A and B are more common in developing countries. Other relatively common causes include autoimmune hepatitis, Wilson disease, Budd-Chiari and HELLP syndromes (5). Prognosis varies considerably with causation: whilst Hepatitis A and B, paracetamol and ischaemic injuries have spontaneous survival rates of around 36%, idiosyncratic reactions and indeterminate causes have only a 14% survival rate without transplantation (6).
The pathogenesis of raised intracranial pressure (ICP) in ALF is multifactorial and not entirely understood. High levels of ammonia enter the systemic circulation, being incompletely cleared by the failing liver. Astrocytes attempt detoxification by converting it to glutamine. Glutamine, however, causes astrocyte swelling by increasing intracellular osmolarity, enhancing the passive movement of water from the intravascular space and resulting in brain swelling (7). Glutamine also alters astrocyte mitochondrial function, increasing oxidative stress and causing cerebrovascular dilatation and hyperaemia; increasing intracranial blood volume and further raising ICP (7).
Despite the common occurrence of raised ICP in ALF, ICP monitoring remains a contentious issue. This is largely due to a lack of randomised trials demonstrating improved survival, risks of bleeding and infection, and the lack of a consensus regarding treatment goals (5). However, invasive monitoring remains the only means to detect raised ICP and groups such as the US Acute Liver Failure study group advocate their use in patients listed for transplant with Grades III-IV encephalopathy (8). In a prospective study by said group, of 332 patients with ALF and severe encephalopathy, 92 had ICP monitoring (28%). Only 58 of these patients were “surveyed”, and intracranial haemorrhages were found in 10.3% (roughly half of these were thought to be incidental radiological findings) (9). Other studies also support this figure of 8-10% incidence on intracranial bleeding in ICP-monitored patients (10). 30-day survival for patients listed for liver transplant was roughly equal in both monitored and non-monitored groups (85% each) but it was not studied as to whether this influenced long-term neurological outcome.
When ICP monitoring is used, the aim is to keep ICP <20mmHg and CPP >70mmHg. Manoeuvres include elevation of the head of the bed to 30 degrees, minimising invasive procedures such as tracheal suctioning, maintaining pCO2 of 4-4.5kPa. Established treatments include the use of mannitol and barbiturates such as thiopentone and there is less robust evidence supporting the use of hypertonic saline, indomethacin and propofol (5). Some centres also use therapeutic hypothermia (32-33oC) in the perioperative period surrounding transplantation as a means to control ICP.
Raised intracranial pressure in acute liver failure is common and may contribute to overall morbidity and mortality. Despite this, there is currently a lack of robust evidence to advise upon the use intracranial pressure monitors in such patients, primarily due to the persistent risks of haemorrhage and the absence of a clear survival benefit at this time. Further research is needed in this area.