Craniectomy Rationale: Outcomes Data and Surgical Techniques
Pearls
Space-occupying edema leading to transtentorial herniation is the leading cause of death immediately following massive cerebral infarction.
Although maximal cerebral edema is thought to occur around day 2 to day 4, fatal cerebral herniation can occur as early as within 24 hours of ischemia.
According to the best class I data currently available, decompressive craniectomy reduces the risk of mortality by 50%, and the risk of severe disability or death by 42%, compared with the best medical therapy, for patients <60 years old undergoing surgery <48 hours after stroke ictus.
The data do not support exclusion of patients from surgical consideration based on laterality of infarct or presence/absence of aphasia
Decompressive craniectomy should include a large craniectomy (≥12 cm anterior-posterior) and duraplasty, with the possible additional inclusion of an anterior temporal lobectomy
♦ Epidemiology and Natural History
Massive strokes producing life-threatening space-occupying edema represent one of the most challenging and fatal neurologic diseases, and comprise 1 to 10% of all ischemic strokes.1 These infarcts typically involve the majority of the middle cerebral artery (MCA) distribution, with the occasional addition of the anterior (ACA) or posterior (PCA) cerebral artery territories. The malignant edema associated with these events usually reaches a peak 2 to 4 days following the ictus, but can manifest as early as within the first 24 hours2 ( Fig. 16.1 ). Despite optimal medical management in intensive care settings, the mortality rate from strokes of this size is approximately 80%.3 The distribution of mortality rates is bimodal, with an early peak within the first 3 to 6 days, followed by a second peak during the 2nd and 3rd weeks after stroke.4 Mortality during the first peak is primarily due to transtentorial herniation from edema-related increased intracranial pressure (ICP) within a fixed-volume skull. Delayed mortality is often a result of complications related to both hospitalization, such as pneumonia, as well as medical comorbidities, such as myocardial infarction and heart failure.
Given the severity of this disease and the benefit of early intervention, several predictors of progression to malignant edema have been identified, the most important of which is the size of the stroke. An infarct volume greater than 145 cc measured by diffusion-weighted magnetic resonance imaging (MRI) within 14 hours of stroke onset has high sensitivity (100%) and specificity (94%) for predicting progression to life-threatening edema. Combining diffusion-weighted imaging (DWI) with apparent diffusion coefficient (ADC) imaging can increase specificity to almost 100%.5 Other radiographic and clinical predictors include computed tomography (CT) showing stroke volume >50% of the MCA territory, National Institutes of Health Stroke Scale (NIHSS) >20 on admission, development of nausea/vomiting within 24 hours of onset of infarction, systolic blood pressure ≥180 mm Hg 12 hours after the onset, and a history of hypertension or heart failure.6 , 7
♦ Medical Management of Massive Cerebral Infarction
Massive cerebral infarction is a devastating disease, and patients with this condition require intensive monitoring on an inpatient unit. The guidelines that apply to the care of all patients with ischemic stroke with respect to blood pressure and glucose control, nutrition, and pulmonary embolus prophylaxis are still relevant in these patients. Although medical management is rarely therapeutic for massive cerebral infarction, several measures can be taken to treat increased ICP through minimally invasive means ( Table 16.1 ).
The mainstays of medical therapy include control of agitation and pain, and osmotic therapy with mannitol and hypertonic saline.8 Although mannitol is widely used, class I data on its efficacy in reducing morbidity and mortality after stroke is not available.9 Hypertonic saline, however, has been shown to reduce ICP, and carries the additional benefit of counteracting mannitol-induced hyponatremia, as often results from multiple administrations of the osmotic diuretic. Glycerol has also been used as an osmotic agent in the treatment of elevated ICP. Barbiturates reduce cerebral metabolism, and the resultant decrease in cerebral blood flow may theoretically decrease edema. The decrease in ICP is usually shortlived, however, and often carries with it a deleterious decrease in cerebral perfusion pressure (CPP). Thus, barbiturates should be used with caution in stroke patients.
Therapies with some evidence of benefit |
Therapies with little evidence of benefit |
Additional complementary measures have been tested to aid in the treatment of raised ICP, although data on these techniques are varied. Hyperventilation is a commonly used procedure aimed at decreasing ICP by inducing cerebral vasoconstriction to reduce cerebral blood flow. This method has been successful in reducing ICP acutely, but its effects are not typically long-lasting and have not resulted in improvements in neurologic outcome or mortality for stroke patients, possibly due to the attendant decrease in CPP. Other medical interventions include alkalinizing cerebrospinal fluid with tromethamine and reducing vasogenic edema with corticosteroids. Although these therapies have shown some benefit in select patient populations, they lack strong clinical evidence and so cannot be advocated for all stroke patients. Mild to moderate hypothermia has shown some early positive results, and although further characterization of the technique and its protocol are required, this maneuver may represent a viable therapy in the near future.9
♦ Evidence for the Role of Decompressive Craniectomy
Effect on Mortality and Functional Outcome
The employment of trephinations or craniectomies to relieve brain swelling is among the oldest practices in the history of neurosurgery. Following its increasingly common use in the management of severe head trauma, decompressive craniectomy (DC) was first used to treat ischemic stroke-related malignant edema in the 1970s and early 1980s.10 – 12 For the following two decades, several retrospective uncontrolled series were published addressing the question of whether DC provided benefit. Compared with the 80% mortality described for best medical therapy, these studies uniformly showed lower mortality rates, ranging from 11 to 34% with series including more than four patients. The fraction of patients with a good functional outcome, however, varied widely from 8 to 57% for series including more than four patients.
Gupta et al13 reviewed studies in the literature from 1970 to 2003 in which data for individual patients were available; they found 13 series with 138 patients. The pooled results demonstrated an overall mortality rate of 24%. They further separated the results into “good” (Barthel Index [BI] ≥60; modified Rankin Scale [mRS] ≤3; or Glasgow Outcome Scale [GOS] ≥4) and “poor” (BI <60; mRS >3; GOS <4) functional outcomes (see Table 16.2 for an explanation of the outcome scales). With this dichotomization, 42% of all patients experienced a good functional outcome, and 58% experienced a poor outcome, including those who died. Subgroup analyses showed that patient age >50 predicted poor outcome, with 80% poor outcomes in the older group compared with 32% poor outcomes in the age group ≤50. Time between stroke ictus and surgery, presence of uncal herniation signs preoperatively, and laterality did not affect functional prognosis, but the available data for the latter two comparisons were sparse.
One of the larger early series not included in the above review was that of Schwab et al,14 which included 63 DC patients. The authors reported an overall mortality of 27%, and relatively good functional outcomes, with a mean BI score of 65. They additionally found that surgery within the first 24 hours after stroke led to a reduced mortality of 16%, compared with 34% if surgery was performed after 24 hours. Within the early surgery group, only 13% had signs of uncal herniation, compared with 75% in the late surgery group.
To answer persistent questions about the role of DC and the prognostic factors affecting patient selection, three randomized controlled trials were initiated in Europe in the mid-2000s. The DECIMAL (Decompressive Craniectomy in Malignant Middle Cerebral Artery Infarcts) trial was initiated in France,15 the DESTINY (Decompressive Surgery for the Treatment of Malignant Infarction of the Middle Cerebral Artery) trial in Germany,16 and the HAMLET (Hemicraniectomy After Middle Cerebral Artery Infarction With Life-Threatening Edema Trial) in the Netherlands.17 These trials enrolled patients 18 to 55 or 60 years of age with a unilateral stroke occupying at least two thirds or 145 cc of the MCA distribution. Patients were randomized to receive either DC or best medical therapy, including all of the management options discussed in the previous section.
Both DECIMAL and DESTINY were aborted prematurely in early 2006 when a preliminary analysis demonstrated significant reduction in mortality in the surgical arm. Rather than subject further patients to the unnecessary risk of randomization, data from these two studies, along with the patients enrolled to that point in HAMLET, were pooled for analysis, a decision facilitated by the trials’ largely similar design. By that time, DECIMAL had enrolled 38 patients, DESTINY 32 patients, and HAMLET 23 patients. In the pooled data (N = 93), a total of 42 patients were treated medically and 51 surgically. Mortality was 71% in the medical arm compared with 22% in the surgical arm, a significant absolute risk reduction of 50%.18
In terms of the functional outcome of survivors, surgery resulted in a larger fraction of patients with only slight disability (mRS 2) than did medical management (14% versus 2%), but also produced a larger number of survivors with moderately severe disability (mRS 4; 31% versus 2%). Therefore surgery resulted in a significant absolute risk reduction of death or severe disability (mRS >4) of 51%, from 76% to 25%. But because of the large number of DC survivors with mRS 4, the trials did not individually show a significant reduction in the risk of moderately severe disability or worse (mRS >3). In the pooled data, however, there was a small but significant absolute risk reduction of mRS >3 related to surgery of 23%, from 79% to 57%.18
The HAMLET trial continued to enroll patients well into 2007, and results were published in early 2009.19 Of the 64 patients enrolled, 32 underwent DC, and 32 were treated medically. Again, a significant absolute risk reduction of 38% for mortality was afforded by DC, from 59% for patients treated medically to 22% for patients treated surgically. The functional outcome of survivors in this study was poorer compared with the other two randomized trials. The fraction of patients with moderately severe disability (mRS 4) was higher in the surgical arm (34% versus 16%), as was the fraction of patients with severe disability (mRS 5; 19% versus 0%).
The HAMLET publication also pooled its completed data with that of DESTINY and DECIMAL for a revised aggregate result (N = 134). This meta-analysis represents the best class I evidence to date available on this topic. The results confirmed the significant reduction in mortality risk of 50% in patients undergoing DC, from 71% to 21%. Risk for avoiding severe disability or death (mRS >4) was also reduced by 42%, from 75% to 33%. Risk for avoiding an outcome worse than moderately severe disability (mRS >3) was reduced just short of significance, by 16%, from 76% to 60%. All these results are summarized in Table 16.3 .
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
