A 20 year-old man was admitted to his local district hospital with a severe head injury following an assault. On arrival in the Emergency Department he was agitated with a reduced conscious level, with evidence of blunt trauma to the head and neck. Prior to intubation, his Glasgow Coma Score (GCS) was recorded as 7 (E1V2M4), and with cervical spine precautions he underwent intubation with subsequent mechanical ventilation and sedation.

An urgent CT brain and cervical spine revealed early evidence of intracerebral contusions with diffuse areas of petechial intracerebral haemorrhage identified. Nasal and maxillary fractures were also seen, with no cervical spine pathology identified. He was transferred to the regional neurological centre for assessment and ongoing management.

On arrival in the Neurosurgical Intensive Care unit the patient underwent insertion of an intracranial pressure monitor revealing an intracranial pressure (ICP) of between 30-35 mmHg. Pupil reactivity was sluggish bilaterally. Sedation was changed to infusions of propofol, fentanyl and midazolam, positioning was optimised with 20 degree head-up tilt, endotracheal tube ties were replaced and targeted mechanical ventilation to EtCO2 4- 4.5kPa. Central venous access was established and an infusion of Noradrenaline was used to target cerebral perfusion pressure to 70mmHg.

Initial medical management stabilised ICP below 25mmHg, but within the next 12 hours this began to rise despite neuromuscular blockade and infusion of hypertonic saline. Further CT imaging revealed progression of the intracerebral contusions with developing oedema. The patient was transferred to the operating theatre for insertion of an external ventricular drain. CSF drainage resulted in an immediate but small improvement in ICP but again over the next 12 hours it began to rise, and decision was made for bifrontal decompressive craniectomy.

Subsequent recovery was slow and was complicated by ventilator-associated pneumonia, a protracted tracheostomy wean and severe agitation. The patient underwent intensive neuro-rehabilitation and had been decannulated, but was left with persistent cognitive impairment, seizures and depression.

What is the rationale for performing decompressive craniotomy in TBI?

Nick Taylor

Traumatic brain injury is common, and the resultant disability can have significant social and economic consequences. Traumatic brain injury results in raised ICP and permanent neurological disability through a cycle of increasing intracerebral oedema, progressive increases in ICP, reduced cerebral oxygen delivery with resultant neuronal cell injury. The normal range for ICP lies between 5-10 mmHg, and is determined by the Monro-Kellie doctrine which states that the cranium is a rigid container filled with nearly impressible brain, with a constant total volume. The total volume comprises the additive component volumes of brain, blood and CSF.

The rationale for decompressive craniectomy is that decompression of this rigid compartment will improve compliance and therefore cerebral oxygen delivery. Following the primary insult, management of traumatic brain injury is focused on preventing secondary brain injury as outlined in the Brain Trauma Foundation Guidelines [1]. If medical management alone is unsuccessful in controlling ICP, it is unclear at what stage surgical decompression is most effective. A question mark remains over the most appropriate type of decompressive craniectomy (bifrontal, fronto-temporo-parietal, unilateral, bilateral or circumferential) and whether or not the dura remains intact, is breached with slits or is opened. A number of clinical trials have explored the potential benefit of decompressive craniectomy in the context of traumatic brain injury.

In 2003 a retrospective cohort study looked at the role of decompressive craniectomy in patients admitted with severe head trauma (GCS less than or equal to 8) [2]. 40 of 816 patients underwent decompressive craniectomy within or after 24 hours of primary brain injury. Early decompressive craniectomy (within 24 hours) was performed where GCS was less than 6 with clinical signs of cerebral herniation (absent pupillary reflexes) or with radiographic evidence of diffuse or unilateral cerebral oedema or herniation. ICP was not measured prior to decompressive craniectomy in this group.

Late decompressive craniectomy (after 24 hours) was performed in the presence of refractory intracranial hypertension (>35mmHg), absent pupillary reflexes and the same radiographic evidence in the early group. Social rehabilitation was achieved in 19% of patients in the early group versus 38% patients in the late group. 30% of patients in the early group remained in a persistent vegetative state or with severe disability, versus 38% patients in the late group. Death occurred in 52% patients in the early group versus 23% patients in the late group. A Cochrane review incorporating the results of this and other retrospective studies concluded that there is no evidence to support the routine use of decompressive craniectomy to reduce unfavourable outcome in adults with severe traumatic brain injury and refractory intracranial hypertension [3].

The DECRA study recruited 155 adult patients with severe diffuse traumatic brain injury with raised ICP refractory to first-line therapy [4]. Patients were excluded if thought to have sustained an unsurvivable injury, had dilated and non-reactive pupils, had associated mass lesion, spinal cord injury or cardiac arrest at scene. First line treatment for ICP greater than 20 mmHg included sedation, normal CO2 tension, hypertonic saline, mannitol, CSF drainage and neuromuscular blockade. Following first line treatment, patients with ICP greater than 20 mmHg for more than 15 minutes were randomised within 72 hours to either bifrontal decompressive craniectomy or standard care. Patients in the standard care group could proceed to decompressive craniectomy after 72 hours.

Beyond the first six hours after randomisation, there was a sustained benefit in reduced ICP in the decompressive craniectomy group. There was a shorter period of mechanical ventilation and intensive care stay in the decompressive craniectomy group. 23.2% of patients in standard care group ultimately proceeded to decompressive craniectomy. Post- surgical complications included increased risk of cerebral abscess, CSF leak, haematoma formation, hydrocephalus and need for subsequent cranioplasty. An unfavourable Extended Glasgow Outcome Scale Score was seen in 70% patients in the decompressive craniectomy group versus 51% patients in the standard care group.

The reasons underlying this poor long-term outcome are likely to be multifactorial. The timing of surgical intervention is considered early in relation to standard practice. Bifrontotemporoparietal decompressive craniectomy was associated with longer operative times and appeared to be prone to complication, in particular hydrocephalus and infection (which might have related to breach of the frontal air sinus). Comparison between the two groups was also blurred by the rate of nearly a quarter of patients receiving decompressive craniectomy in the standard care group. In comparison to common practice in the United Kingdom, the ICP threshold for surgical decompression of 20mmHg was relatively low, and for a short period of time.

The Rescue ICP study is further evaluating whether decompressive craniectomy is effective for the management of patients with raised and refractory ICP following traumatic brain injury [5]. The ICP threshold for randomisation is higher than in DECRA (25mmgHg versus 20mmHg), and for a longer duration (60 minutes versus 15 minutes). Surgical technique is less specified, and can include unilateral or bilateral frontotemporoparietal decompression. Recruitment for Rescue ICP is now closed and the results are expected imminently.

Lessons Learnt

Therapeutic goals in traumatic brain injury include control of ICP and preservation of cerebral oxygen delivery to prevent long-term neurological disability. Surgical decompression plays an important part in this management, but the evidence to date suggests that decompressive craniectomy should be reserved for selected patients where medical management has failed to control refractory intracranial hypertension.

Decompressive craniectomy is associated with post-operative morbidity, and long-term consequences including the impact on rehabilitation and need for subsequent corrective procedures should be considered on a case-by-case basis when selecting patients for this procedure.

References

1. Guidelines for the management of severe traumatic brain injury 3rd edition. Journal of Neurotrauma Volume 24, Supplement 1, 2007. DOI: 10.1089/neu.2007.9999
2. Decompressive craniotomy for severe traumatic brain injury: evaluation of the effects at one year. Albanèse J, Leone M, Alliez JR et al. Crit Care Med 2003 Oct;31(10):2535-38.
3. Decompressive craniotomy for the treatment of refractory high intracranial pressure in traumatic brain injury. Sahuquillo J, Arikan F. Cochrane Database Syst Rev. 2006 Jan 25;(1):CD003983.
4. Decompressive craniotomy in diffuse traumatic brain injury.CooperDJ,Rosenfeld JV, Murray L et al. N Engl J Med. 2011 Apr 21;364(16):1493-502. doi: 10.1056/NEJMoa1102077. Epub 2011 Mar 25.
5. Update on the RESCUEicp decompressive craniotomy trial.PJHutchinson,AG Kolias, I Timofeev et al. Crit Care. 2011; 15(Suppl 1): P312. Published online 2011 March 11. doi: 10.1186/cc9732