Trial of Decompressive Craniectomy for Traumatic Intracranial Hypertension
PubMed • Full text
[[Category:]]
Clinical Question
In patients with traumatic brain injury (TBI) and intracranial hypertension refractory to medical management, does decompressive craniectomy as a last-tier intervention improve outcomes as measured by the Extended Glasgow Outcome Scale (GOS-E)?
Bottom Line
Decompressive craniectomy in severe TBI patients with intracranial hypertension refractory to medical management improves mortality (by 22%) and functional outcomes compared to continued medical management. The survival benefit translates to both independent and dependent living.
Major Points
The Randomized Evaluation of Surgery with Craniectomy for Uncontrollable Elevation of Intracranial Pressure (RESCUEicp) trial was designed to assess the effectiveness of craniectomy as a last-tier intervention in patients with traumatic brain injury (TBI) and refractory intracranial hypertension. This trial showed that craniectomy increased the number of favorable outcomes compared to continued medical management and that for every 100 patients managed surgically vs medically there were 22 more survivors. Of these 22, 27% were in a vegetative state, 36% had lower severe disability (dependent on others for care) and 36% had upper severe disability (independent at home) or better. This informs the debate around historical concerns that decompressive craniectomy simply increases the number of patients who survive in a vegetative state. While surgical intervention did result in more vegetative patients than medical management, it also resulted in higher rates of upper severe disability, which is considered a favorable outcome. The rates of moderate disability and good recovery were similar to those who received medical management.
In contrast to the RESCUEicp trial, the DECRA trial showed that patients undergoing craniectomy had worse ratings on the GOS-E at 6 months than those receiving standard care (P = 0.03), although the rates of death were similar at 6 months (19% and 18%, respectively). The DECRA trial aimed to assess the effectiveness of early craniectomy — offered as a stage 2 treatment within 72 hours after injury — for moderate intracranial hypertension (intracranial pressure, >20 mm Hg for 15 minutes within a 1-hour period [continuous or cumulative]) in patients with diffuse TBI.
Guidelines
Brain Trauma Foundation TBI Guidelines 2016 (published prior to this trial)
- Level IIa: Bifrontal DC is not recommended to improve outcomes as measured by the GOS-E score at 6mo post-injury in severe TBI patients with diffuse injury (without mass lesions), and with ICP elevation to values >20mmHg for more than 15 min within a 1-h period that are refractory to first-tier therapies. However, this procedure has been demonstrated to reduce ICP and to minimize days in the ICU
- Level IIa: A large frontotemporoparietal DC (not less than 12 x 15 cm or 15 cm diameter) is recommended over a small frontotemporoparietal DC for reduced mortality and improved neurologic outcomes in patients with severe TBI.
Design
- Multicenter, parallel-group, randomized, superiority trial
- N = 398
- Surgical: 202
- Medical: 196
- Setting: 52 centers in 20 countries
- 71.1% of patients recruited in the United Kingdom
- Sites were hospitals that provided acute neurosciences care for patients with severe TBI and had 24-hr neurosurgical services
- Enrollment: January 2004 to March 2014
- Analysis: intention-to-treat
- 400 patients needed to meet power to detect 15% difference
- Primary outcome: Extended Glasgow Outcome Scale (GOS-E) at 6 months
Population
Inclusion Criteria
- Age 10 - 65 years
- TBI as assessed by abnormal CT scan of the brain
- Intracranial pressure monitor in place
- ICP > 25 mmHg for 1-12 hours despite stage 1 and 2 measures
- Stage 1: head elevation, ventilation, sedation, analgesia, paralysis (optional)
- Stage 2: continue stage 1 measures; addition of optional treatments - ventriculostomy, inotropes, mannitol, hypertonic saline, loop diuretics, hypothermia.
- Barbiturates not permitted.
- Immediate operation for evacuation of intracranial hematoma (as long as the bone flap was replaced at the end of the procedure)
Exclusion Criteria
- Craniectomy performed for evacuation of intracranial hematoma
- Bilateral fixed and dilated pupils
- Bleeding diathesis
- Injury deemed to be unsurvivable
Baseline Characteristics
No significant differences between groups except for history of drug or alcohol abuse
Comparisons are listed as surgical group vs medical group
- Age (years): 32.3 vs 34.8
- Male: 81.7% vs 80%
- GCS motor score 1 or 2: 53% vs 50%
- GCS motor score 3-6: 47% vs 50%
- Pupillary abnormality: 29.2% vs 29.1%
- Hypotension: 19.8% vs 21.4%
- Hypoxemia: 24.3% vs 26.5%
- History of drug or alcohol abuse: 24.8% vs 35.2%
- Extracranial injury: 37.1% vs 42.3%
- Diffuse injury on CT: 81.3% vs 75.8%
- Mass lesion on CT: 18.7% vs 24.2%
- Craniotomy for evacuation of hematoma: 12.9% vs 15.3%
- Ventriculostomy: 16.8% vs 21.9%
- Neuromuscular paralysis: 50% vs 52.6%
- BP augmentation: 55.4% vs 59.2%
- Osmotherapy: 72.3% vs 73.5%
- Therapeutic hypothermia: 23.3% vs 27%
Interventions
All patients were managed in ICUs and treated according to the Stage 1 protocol aimed at maintaining ICP < 25 mmHg, CPP > 60 mmHg, normothermia, normoglycemia, mild hypocapnia, O2 sat >97%
- head elevation
- ventilation
- sedation
- analgesia
- paralysis (optional)
Stage 2 initiated if ICP remained > 25 mmHg
- ventriculostomy (if an external ventricular drain had not already been inserted for ICP monitoring)
- inotropes
- mannitol or hypertonic saline
- loop diuretics
- hypothermia (not <34°C)
- moderate hypocapnia
- barbiturates not permitted
Stage 3 initiated if ICP remained > 25 mmHg for 1 - 12 hours despite Stage 1 and 2 measures
- Patients randomized 1:1 to
- continued Stage 1 and 2 management with the option to add barbiturates OR
- continued Stage 1 and 2 management with decompressive craniectomy
- large unilateral frontotemporoparietal craniectomy (hemicraniectomy) for patients with unilateral hemispheric swelling OR bifrontal craniectomy for patients with diffuse brain swelling that affected both hemispheres
- 63% of surgical patients underwent bifrontal craniectomy and 37% underwent unilateral craniectomy
- large unilateral frontotemporoparietal craniectomy (hemicraniectomy) for patients with unilateral hemispheric swelling OR bifrontal craniectomy for patients with diffuse brain swelling that affected both hemispheres
- Patients assigned to medical treatment alone could undergo craniectomy later in case their condition deteriorated
- 73 medical patients (37.2%) underwent craniectomy
- Patients who were assigned to undergo craniectomy could have barbiturate infusion in case of further deterioration
- 19 surgical patients (9.4%) received barbiturates
Outcomes
Comparisons are surgical therapy vs. medical therapy.
Primary Outcomes
- Extended Glasgow Outcome Scale (GOS-E) at 6 months
- Favorable Outcomes: 42.8% vs 34.6% (P=0.12)
Secondary Outcomes
- GOS-E Results at 12 and 24 months
- Favorable Outcomes at 12 months: 45.4% vs 32.4% (P=0.01)
- Favorable Outcomes at 24 months: data not yet analyzed
- Mortality at 6, 12 and 24 months
- 6 months: 26.9% vs 48.9% (P<0.001)
- 12 months: 30.4% vs 52% (P<0.001)
- 24 months: data not yet analyzed
- Quality of life at 6, 12 and 24 months
- Data not yet analyzed
- Glasgow Coma Scale (GCS) score at discharge
- Death: 22.7% vs 48.5% (Absolute difference -25.8; 95% CI -35.5 to -16.2)
- 3-5: 7% vs 6.4% (Absolute difference 0.6; 95% CI -4.6 to 5.8)
- 6-8: 11.9% vs 5.8% (Absolute difference 6; 95% CI 0.2 to 11.9)
- 9-12: 36.2% vs 21.6% (Absolute difference 14.6; 95% CI 5.3 to 23.9)
- 13-15: 22.2% vs 17.5% (Absolute difference 4.6; 95% CI -3.6 to 12.9)
- Assessment of intracranial pressure control
- Surgical patients had better intracranial pressure control vs medical patients with 5 hrs vs 17 hrs spent with ICP > 25 mmHg after randomization (P<0.001)
- Time in the ICU
- No significant difference between groups
- Time to discharge from hospital
- 15 days vs 20.8 days (P=0.01)
- Economic evaluation
- Data not yet analyzed
Subgroup Analysis
A sensitivity analysis was performed for the primary-outcome measure in the per-protocol population. The per-protocol population was defined as the patients in the intention-to-treat population who did not have a severe breach of protocol. This analysis did not alter the results that were observed in the intention-to-treat analysis.
6 a priori subgroups were analyzed. The surgical intervention was more beneficial (in terms of favorable outcome) for patients ≤ 40 years old vs those > 40 years old (P=0.025).
- Diffuse injury versus non‐diffuse
- Initial GCS total score 3‐8 versus 9‐15
- Age ≤ 40 versus Age > 40 years
- Craniotomy for evacuation of mass lesions before randomization
- Time from injury to randomization ≤72 versus >72 hours
- Patients randomized in the UK versus rest of the world
Adverse Events
Adverse events occurred in 33 (16.3%) surgical patients vs 18 (9.2%) medical patients (P=0.03). This difference is driven mainly by the 5 post-op hematomas and 5 surgical site infections that occurred in the surgical group.
Criticisms
- Clinical teams caring for the patients were aware of the trial-group assignments, however outcome adjudication on the basis of the GOS-E questionnaires was done by personnel who were unaware of the group assignments.
- A relatively large portion of patients in the medical group underwent decompressive craniectomy (37.2%)
- Long-term data on cranial reconstruction was not systematically obtained.
- This study did not examine the effectiveness of primary decompressive craniectomy.
Funding
Supported by the Medical Research Council (MRC) and managed by the National Institute for Health Research (NIHR).
Further Reading
NEJM 2016;375(12):1119-30.