Surgery for Brain Edema

Peter Hutchinson, F.R.C.S.(SN); Ivan Timofeev, M.R.C.S.; Peter Kirkpatrick, F.R.C.S.(SN)


Neurosurg Focus. 2007;22(5):E14 

In This Article

Surgical Decompression for Stroke

Despite recent advances in the prevention and treatment of ischemic stroke, there is still a proportion of patients in whom large cerebral infarcts develop as a result of a major vessel occlusion. Rapidly progressing, malignant cerebral edema is commonly associated with this condition, and usually manifests as clinical and radiological deterioration on Days 2 to 5 (Fig. 3).[32] Escalating swelling and mass effect compromise cerebral perfusion, oxygenation, and metabolism and can eventually lead to refractory intracranial hypertension with increasing shifting of intracranial contents[27] and transtentorial herniation, the leading cause of death in these patients.[8,31] Despite optimal medical therapy and intensive care support, death from space-occupying cerebral infarction remains in the range of 70 to 80%,[7,32] with little change in recent years.[48] In survivors, severe residual disability is common.

Figure 3.

Not surprisingly, decompressive craniectomy has been performed[28,36,58] as an empiric lifesaving measure for a long time in an attempt to protect the brain from the damaging effects of propagating edema. More recently, the physiological rationale for this approach has been backed by evidence from animal studies. In experimental models of extensive ischemic stroke, decompressive craniectomy improved cerebral perfusion, survival, and neurological outcome.[19,20] It was also shown to reduce the volume of the infarction,[20,21,38] particularly when performed early after vessel occlusion. These beneficial effects are likely to be linked to increases in collateral circulation[25] and reductions in tissue edema,[32] together with improvements in oxygenation and energy metabolism in ischemic penumbra. Cerebral tissue monitoring in patients undergoing craniectomy for ischemic stroke has demonstrated that de compression leads to a reduction in ICP and an increase in brain tissue oxygen levels.[37,57] Imaging and clinical study data suggest that late restoration of function is possible in the affected cerebral hemisphere,[13,39] making decompression a preferred option over unguided "strokectomy."

The effect of decompressive craniectomy on neurological outcome is the crucial question in the assessment of its clinical utility. During the last three decades, more than 20 centers have published retrospective or nonrandomized prospective[26,59,64] case studies on outcomes in patients under going decompressive craniectomy for ischemic cerebral edema. Data from all groups revealed improved survival following surgical decompression,[41,70] with a mean mortality rate of 20 to 30% (range 8-50%) compared with 70 to 80% among patients treated conservatively. Nonetheless, as in other applications of decompressive craniectomy, concerns about the functional outcome in surviving patients remain. Because it is unrealistic to expect complete functional recovery following extensive ischemic stroke, most authors regard moderate or moderate-to-severe disability (mRS ≤ 4 or BI ≤ 60; that is, relative independence in daily activities) as a favorable outcome, and severe disability, vegetative state, or death as an unfavorable outcome. From this perspective, many authors have reported that a substantial proportion of surviving patients can attain a favorable outcome after decompression;[11,12,16,55,70] in other series, however, presented outcomes have been worse.[23,25,34,43] This discrepancy has fueled interest in isolating the criteria that help to identify patients who are most likely to benefit from decompressive craniectomy. An association between poor neurological outcome and patient age was found in several studies,[12,44,70,75] with some data demonstrating a correlation between postdecompression BI and age.[16,24] The timing of decompressive craniectomy has also been suggested as an important predicting factor, with more patients surviving with favorable outcomes when decompressive craniectomy was performed very early[14] or early[48,55] after the onset of symptoms and before the development of mass effect[27] and neurological features of herniation. Low initial Glasgow Coma Scale scores,[41,43,55] involvement of additional vascular territories[70] and the dominant hemisphere,[23] and failure to open the dura mater as part of the surgical technique[74] may also be independent predictors of a worse outcome.

Given the fact that most surviving patients have some residual neurological loss, the patient and family's perception of the craniectomy and resulting outcome is especially important. In study by Woertgen et al.,[74] 83% of patients or their relatives indicated that they would agree to surgery in the future. Erban and associates[22] found that patient approval of surgery was lower (61%) and depression in survivors was common.

The failure of case series and nonrandomized trials to provide a definitive answer on the benefits of decompressive craniectomy in ischemic brain edema was highlighted in recent systematic reviews[30,33,49] and triggered the initiation of several prospective, randomized, controlled trials in which investigators will compare decompressive craniectomy and medical treatment with the best conservative therapy alone. A summary of the protocol, progress, and findings from these trials is presented in Table 2 . Currently, only the combined results of the DECIMAL, DESTINY, and HAMLET trials have been published in the peer-reviewed literature.[71] These multicenter trials were started independently in France, Germany, and The Netherlands, respectively. Because of their similar design and outcome measures, however, the steering committees planned a pooled analysis of the data from these trials while they were still ongoing to increase the chances of obtaining sufficient data and to minimize the number of patients involved in individual trials. This pooled analysis included 93 patients younger than 55 to 60 years who underwent decompressive craniectomy within 48 hours from the onset of stroke. The primary outcome measure for the pooled analysis was the proportion of patients with an mRS score of 4 or lower in the surgical and medical treatment groups at 1 year posttherapy. Survival and an mRS score of 3 or lower were secondary end points. The results have suggested that significantly more patients in the surgical group than in the control group had an mRS score of 4 or lower (75% compared with 24%; pooled absolute risk reduction 51%, 95% CI 34-69), an mRS score ≤ 3 (43% compared with 21%; pooled absolute risk reduction 23%, 95% CI 5-41), and survived (78% compared with 29%; pooled absolute risk reduction 50%, 95% CI 33-67). This effect was highly consistent across the three trials and did not change on adjustment for baseline differences. This study represents the best evidence available to date (and possibly ever) on the role of surgical decompression in malignant brain swelling associated with extensive cerebral infarction. As acknowledged by the investigators, however, these data cannot be translated automatically in to the widespread use of decompressive surgery in all eligible patients: first, because the findings of an improved outcome cannot be extrapolated to older patients and those who undergo surgery within the first 48 hours after the onset of stroke; and second, although surgery doubles the number of patients who require minor support (mRS score ≤ 3), it increases 10-fold the number of patients with a moderately severe disability (mRS score ≤ 4) that requires almost continuous assistance. This fact must be carefully considered and discussed with the next of kin in each individual case before proceeding to surgery. More information is also required on the longer-term and neuropsychological effects of decompression, rate of surgical complications, and patient's viewpoint on the procedure.

Authors of several studies[18,46,51] have considered the role of decompressive craniectomy in the treatment of the hemorrhagic form of stroke, spontaneous intracranial hemorrhage, which is also associated with mass effect and elevated ICP. Data from all of these retrospective case series have revealed decreased mortality rates, and some suggest the possibility of an improved functional outcome[46,51] after decompressive craniectomy. Prospective, randomized evidence is required to support these findings.