Malignant cerebral hemispheric infarction with swelling and risk of herniation

INTRODUCTION — This topic will review the clinical features and management of life threatening malignant (ie, massive) hemispheric infarction.

The acute treatment of hemispheric infarction in the first few hours after stroke onset (prior to the development of severe brain swelling) is similar to other types of acute ischemic stroke, as discussed in detail elsewhere. (See "Initial assessment and management of acute stroke".)

MALIGNANT HEMISPHERIC INFARCTION

Description — Less than 10 percent of ischemic strokes are classified as malignant or massive, which is characterized by the development of space-occupying cerebral edema that is severe enough to produce brain tissue shifts and herniation [1,2]. The development of this malignant ischemic stroke syndrome is seen primarily following large hemispheric infarctions from cardioembolic or thrombotic etiologies. These hemispheric infarcts commonly result from occlusions of the internal carotid artery or the proximal segment (stem or M1 segment) of the middle cerebral artery (MCA). By one accepted definition, a large hemispheric infarction affects the total or subtotal territory of the MCA and at least partially affects the basal ganglia, with or without involvement of adjacent territories [3].

Patients with this syndrome can have a mortality rate as high as 78 percent due to local mass effect and herniation of the temporal lobe onto the brainstem, as well as major systemic complications [4]. This very high mortality rate may also be explained at least in part by withdrawal of life support measures because of the expected poor quality of life for these patients were they to survive.

Presentation and progression — On presentation, patients with a large hemispheric infarction at risk for swelling and herniation typically have severe neurologic deficits with forced gaze deviation, visual field deficits, hemiplegia, and aphasia or neglect, depending on the hemisphere involved. This combination of neurologic findings yields a National Institutes of Health Stroke Scale (NIHSS) score >15 for a right hemisphere infarction and >20 for a left hemisphere infarction; see table (table 1) and calculator (calculator 1) for determination of the NIHSS score.

Neuroimaging with head computed tomography (CT) or magnetic resonance imaging (MRI) should be obtained immediately upon presentation and repeated urgently if there is neurologic worsening. Neurologic deterioration in these patients is often due to new or worsening cerebral edema, sometimes accompanied by hemorrhagic transformation of the ischemic infarction. Hemorrhagic transformation is less frequently the primary cause of neurologic decline.

●Cerebral edema – The development of space-occupying cerebral edema due to large infarction leads to neurologic deterioration with signs that typically include impairments of consciousness, pupillary changes, and worsening of motor responses [5]. These neurologic signs can indicate the need to intervene urgently with measures to treat brain swelling, including hemicraniectomy for patients who want aggressive treatment (see 'Care considerations' below). Cerebral edema with mass effect and neurologic deterioration may develop with a rapid and fulminant course over 24 to 36 hours from stroke onset but can also follow a more gradual course over several days to a week [4-8].

●Hemorrhagic transformation – Hemorrhagic transformation of a cerebral infarction occurs when blood extravasates from vessels that are damaged by ischemia [9]. The risk of hemorrhagic transformation is thought to increase with increasing size of the infarction, such that patients with malignant hemispheric infarction are at particularly high risk. The clinical impacts of hemorrhagic transformation occur across a spectrum of severity [10]; hemorrhagic infarction with heterogenous or confluent hemorrhage within an infarct region generally does not cause mass effect and is asymptomatic, while confluent parenchymal hematoma with mass effect occurring outside of the infarct region causes symptoms ranging from minimal worsening to precipitous neurologic decline.

Clinical predictors of progression — A systematic review of 38 studies and over 3200 patients found that malignant edema after ischemic stroke was associated with younger age, higher NIHSS scores on admission, and parenchymal hypoattenuation of more than 50 percent of the MCA territory on initial head CT [11]. Additionally, in this study, revascularization within 24 hours after stroke onset was associated with a reduced risk of malignant edema.

The early development of decreased consciousness in patients with malignant hemispheric infarction who are awake at initial presentation is predictive of poor outcome. Supporting evidence comes from an analysis of prospective data from a randomized trial that included 564 placebo-treated patients with major anterior circulation ischemic stroke who did not have impairment in level of consciousness at baseline; the trial enrolled patients up to 12 hours after symptom onset [12]. Decreased level of consciousness at three hours after enrollment and maximum score on a level of consciousness scale (indicating no reaction to pain) in the first 24 hours were both significantly associated with increased mortality.

In a prospective study of 140 patients with occlusion of the MCA diagnosed by magnetic resonance angiography (MRA) within six hours of stroke onset, malignant MCA infarction developed in 27 (19 percent) [13]. The severity of neurologic deficit on admission as measured by the NIHSS score was an independent predictor for the development of malignant MCA infarction (per point, odds ratio 1.18, 95% CI 1.01-1.38). However, an NIHSS score >18 predicted the development of malignant MCA infarction with only low to moderate sensitivity and specificity (70 and 63 percent).

The association of malignant MCA infarction with younger age has been identified in other studies as well [14-16], and may be explained by a protective effect afforded by cerebral atrophy that accommodates brain swelling in older patients [4,14].

Radiologic predictors of progression — Large infarct volume, typically defined by early infarction involving >50 percent of the MCA territory on initial head CT, is the most important predictor of malignant edema, herniation, and death [3]. Extracranial and intracranial carotid occlusion as well as poor collateral circulation (eg, due to an incomplete ipsilateral circle of Willis) are also plausible factors contributing to malignant infarction [17,18].

Because CT findings consistent with infarction may lag behind physiologic tissue infarction, other radiographic methods that are more sensitive to evolving infarction and edema may be more sensitive in predicting patients at risk of herniation in the first few hours after stroke onset. These methods include perfusion CT, diffusion and perfusion MRI, and several techniques that measure cerebral blood flow, including positron emission tomography (PET), xenon CT, and single photon emission computed tomography (SPECT). However, PET, xenon CT, and SPECT are not widely used for this indication in clinical practice.

●CT – Head CT is the most widely available neuroimaging study in acute stroke. The detection of a large area of early ischemic change (ie, involving >50 percent of the MCA territory) on CT is useful for predicting the subsequent development of malignant brain swelling (image 1) [3,14,19,20]. Additional CT findings that may be predictive of malignant edema include midline shift of the septum pellucidum >5 mm, infarction of additional vascular territories (eg, anterior cerebral artery territory plus MCA, or posterior cerebral artery territory plus MCA in patients with fetal origin of the posterior cerebral artery) [21], and low a ASPECTS (Alberta stroke program early CT score). ASPECTS is a method of assessing early ischemic changes in the MCA territory on head CT scan, as described separately (see "Neuroimaging of acute stroke", section on 'ASPECTS method'). The score divides the MCA territory into 10 regions of interest that are evaluated on two axial CT cuts (figure 1). When rating the ASPECTS, one point is subtracted for each identified area of early ischemic change from a total of 10 defined regions. Therefore, a normal CT scan has an ASPECTS value of 10 points, while diffuse ischemic change throughout the MCA territory gives a value of 0. In one retrospective report of 121 patients with large hemispheric infarction, an ASPECTS of ≤7 at baseline was independently associated with the development of malignant brain edema [22].

Several reports have found that perfusion CT can be used to predict malignant MCA infarction [2,23,24]. As an example, one study used perfusion CT to calculate infarct volume and routine CT to calculate intracranial cerebrospinal fluid volume, a measure of intracranial volume reserve [2]. The ratio of infarct volume to cerebrospinal fluid volume was a better predictor of a malignant course than other measures such as ischemic lesion volume or clinical characteristics.

●MRI – Similar to head CT, a large volume of infarction on MRI (image 2), as measured by restricted diffusion on diffusion-weighted imaging (DWI) and correspondingly low apparent diffusion coefficient, may predict a malignant course in MCA territory stroke that is characterized by cerebral edema. Risk of massive edema is particularly high in patients with DWI lesions >145 cm³ in volume [13,25]. The predictive utility of both CT and MRI are higher the later the scan is acquired (given the known growth of infarct over time). However, within critical acute time periods where these scans would provide the highest benefit for early risk detection, MRI may provide higher sensitivities/specificities compared with CT scans for developing malignant hemispheric infarct.

●Cerebral blood flow studies – Although rarely applied to patients with malignant MCA territory infarction in clinical practice, small studies suggest that quantifying cerebral blood flow with CT angiography (CTA), xenon CT, SPECT, or PET may be predictive of cerebral swelling [18,26-30].

CARE CONSIDERATIONS

Level of care — Most patients with acute hemispheric infarction should be monitored and managed in an intensive care unit or dedicated stroke unit in a center with available expertise in neurology, neurosurgery, neuroradiology, and critical care. If such services are not available locally, these patients should be transferred to a higher-level facility. Exceptions include those with advanced directives to withhold resuscitation and life support measures.

Within hours or days of stroke onset, patients with large hemispheric infarction may deteriorate from mass effect caused by cerebral edema. Thus, it is important to quickly establish goals of care with patients, families, caregivers, or other relevant medical decision makers to clarify clinical care options and potential outcome expectations. In the acute phase, patients may require intubation and mechanical ventilation, blood pressure control, and pharmacologic interventions for cerebral edema with mass effect (see 'Medical management' below). Surgical decompressive hemicraniectomy may improve outcomes in selected patients, as discussed below.

Surgical decompression or medical care only? — Decompressive hemicraniectomy is a surgical technique used to remove the skull overlying the infarcted tissue in order to reverse mass effect and relieve brain tissue shifts (see 'Decompressive hemicraniectomy' below). There is evidence that hemicraniectomy for malignant hemispheric infarction substantially reduces mortality. However, many surviving patients are left with major disability. Thus, the dilemma for patients and families is that surgery may leave patients alive with severe disability [31], while medical management alone most often results in death. (See 'Efficacy' below.)

Given the dire prognosis for survival associated with medical treatment, we suggest decompressive hemicraniectomy within 48 hours of stroke onset for patients age ≤60 years with an indisputable diagnosis of large hemispheric infarction (with >50 percent infarction of middle cerebral artery [MCA] territory by head computed tomography [CT] or magnetic resonance imaging [MRI]) who are at high risk of developing malignant edema and who value survival despite the substantial likelihood of survival with severe disability. This surgery may also be offered on a case-by-case basis for otherwise healthy patients older than 60 years of age and beyond 48 hours after stroke onset.

The choice between surgical decompression and medical management ultimately depends upon the values and preferences of patients, families, caregivers, or health care proxy [32]. Hemicraniectomy is clearly an aggressive therapy and may not be appropriate for many patients, particularly for older patients or those who do not wish to survive with severe disability. Hemicraniectomy should proceed only after a thorough discussion with the patient (if cognizant), family, or other relevant medical decision-makers; they should understand that hemicraniectomy increases the odds of survival, but most patients who survive are left with severe to very severe disability. Hemicraniectomy may increase the chance of surviving with only moderate disability in younger patients, but not in those older than 60 years. (See 'Efficacy' below.)

Alternative treatment options for malignant hemispheric infarction include medical management of increased intracranial pressure and cerebral edema, and supportive medical care or withdrawal of medical care for those who do not desire aggressive therapy. Standard medical therapy for large hemispheric infarctions, regardless of surgical decision, involves close monitoring for neurologic deterioration secondary to developing or evolving mass effect; management of cerebral edema with measures such as elevation of the head of the bed, osmotic therapy (mannitol or hypertonic saline), and brief periods of hyperventilation as needed; blood pressure, temperature, and glycemic management; and prevention of secondary complications such as aspiration and deep venous thrombosis. (See 'Medical management' below and 'Palliative care' below.)

Other aspects that may influence the choice between aggressive medical treatment and surgery include the following:

●Quality of life – In addition to mortality and function, patient-centered outcomes are crucial considerations. However, there are few data about quality of life after hemicraniectomy; evidence from the three major clinical trials is limited by small patient numbers and the early stopping of some of these trials [33-37].

●Perspectives may differ before and after the procedure – The perspective of patients, families, and caregivers who have survived the hemicraniectomy procedure may differ from the perspective of those contemplating the procedure beforehand. A priori, many may feel that survival with severe disability (ie, unable to walk without assistance and requiring assistance to attend to bodily needs) after stroke is an unacceptable outcome tantamount to or worse than death [38]. By contrast, limited data from observational studies and trials suggest that most patients who survive after hemicraniectomy have favorable responses to the procedure and, in retrospect, would consent to having the procedure again [39]. As an example, among 64 survivors of hemicraniectomy after infarction in six separate observational reports, 44 patients and families (69 percent) reported that they would favor undergoing decompressive surgery if in the same situation [40]. However, the opinions of survivors of hemicraniectomy may reflect bias since the group is self-selected for the procedure. Survivors may have adapted to a degree of disability that they previously would have considered unacceptable, and their views may be altered by reduced cognitive capacity [31].

●Side of infarction and expected deficits – There has been controversy about offering hemicraniectomy for large dominant hemisphere infarcts, since survival may be accompanied by poor quality of life due to severe aphasia [41]. However, nondominant hemisphere strokes can lead to severe and disabling deficits related to behavioral change, depression, abulia, inattention, or neglect. These problems may interfere with rehabilitation efforts as much as or more than aphasia. Assessments of patients at risk of stroke have indicated that hemiplegia is sometimes viewed as worse than aphasia or death [38]. Thus, the side of infarction should not influence the decision regarding whether to proceed with hemicraniectomy. However, communication of these expected deficits with the patient, family, or other medical decision-makers are needed to help guide shared decision-making for treatments to follow.

In a 2004 systematic review, there was a bias toward surgery of the nondominant hemisphere [40]. Among the 27 patients who had decompression of the dominant hemisphere, functional outcome was no worse than among the 111 patients who had nondominant infarcts. These data must be interpreted cautiously, but it is possible that language deficits may be of less consequence in patients severely disabled by hemiplegia. Additionally, some patients may recover significantly from aphasia after decompressive hemicraniectomy, especially those who are younger. Support for this notion comes from a study that followed 14 patients with left MCA territory infarction who had surgical decompression; a significant improvement in different aspects of aphasia was observed in 13 patients (93 percent) [42].

Palliative care — Hemispheric infarction is a devastating event that should prompt discussion about the goals of care and life-sustaining treatment. Faced with a poor prognosis for a good functional recovery and a high probability of death with best medical treatment or severe disability following surgical treatment, some patients and families or caregivers may choose supportive medical care or withdrawal of medical care. However, prognosis for individual patients with acute ischemic stroke cannot be estimated precisely. Therefore, it is reasonable to implement full aggressive medical care for most patients while postponing consideration of limitations to care until at least the second full day of hospitalization after stroke onset. This approach avoids the self-fulfilling prophecy of poor outcomes caused by clinical nihilism. This recommendation does not apply to patients with preexisting advanced directive orders limiting care, nor to patients who present with minimal brainstem function in whom the possibility of clinical recovery is thought to be negligible. (See "Neuropalliative care of stroke".)

MEDICAL MANAGEMENT

Immediate interventions — As noted above, patients with large or "malignant" hemispheric infarction should be managed in an intensive care unit or stroke unit (with the exception of patients with advanced directives to withhold resuscitation and life support measures and patients who present with catastrophic hemispheric infarction and minimal brainstem function).

Basic supportive measures include the following [3,5]:

●Intubation and mechanical ventilation are indicated for patients who want aggressive care and have diminished level of consciousness, respiratory insufficiency, or neurologic deterioration. (See "The evaluation, diagnosis, and treatment of the adult patient with acute hypercapnic respiratory failure".)

●Analgesia and sedation should be used cautiously as needed to treat pain or agitation and to enable procedures. However, these measures should be minimized and stopped as soon as possible because they depress the neurologic status, making it difficult to accurately assess for neurologic deterioration secondary to stroke progression. Analgesia and sedation may also result in hypoventilation in nonventilated patients, leading to additive intracranial pressure burden from vasodilation.

●Specific blood pressure targets are undefined, but general guidelines for patients with ischemic stroke apply; hypertension with systolic blood pressure >220 mmHg or diastolic >105 mmHg may increase the risk of hemorrhagic transformation and should be treated. Conversely, hypotension should be avoided. (See "Initial assessment and management of acute stroke", section on 'Blood pressure goals in ischemic stroke'.)

●Low serum glucose (<60 mg/dL [3.3 mmol/L]) should be corrected rapidly. It is reasonable to treat hyperglycemia if the glucose level is >180 mg/dL (>10 mmol/L) with a goal of keeping serum glucose levels within a range of 140 to 180 mg/dL (7.8 to 10 mmol/L). (See "Initial assessment and management of acute stroke", section on 'Hypoglycemia' and "Initial assessment and management of acute stroke", section on 'Hyperglycemia'.)

●Initially, isotonic saline should be used for fluid maintenance; hypotonic/hypo-osmolar fluids should be strictly avoided. (See "Evaluation and management of elevated intracranial pressure in adults", section on 'Fluid management'.)

●Elevation of the head of bed to 30 degrees is suggested for patients at risk for cerebral edema and elevated intracranial pressure. (See "Initial assessment and management of acute stroke", section on 'Head and body position'.)

●Therapeutic anticoagulation, if previously or currently clinically indicated, should be held following large hemispheric infarctions given the hemorrhagic transformation risk and the potential decompressive hemicraniectomy surgical indication/need.

●It is reasonable to avoid antiplatelets for the first 48 hours, until the absence of major hemorrhagic conversion is confirmed and once it is clear whether the patient will undergo decompressive hemicraniectomy.

●Venous thromboembolism prophylaxis is indicated for all patients with stroke who have restricted mobility using intermittent pneumatic compression plus prophylactic low-dose anticoagulation with subcutaneous low molecular weight heparin or unfractionated heparin. (See "Prevention and treatment of venous thromboembolism in patients with acute stroke", section on 'Approach in acute ischemic stroke'.)

●Dysphagia is common after stroke and is a major risk factor for developing aspiration pneumonia. Prevention of aspiration includes initial nulla per os (NPO) status (including medications) until swallowing function can be evaluated. (See "Complications of stroke: An overview", section on 'Dysphagia'.)

●Prophylactic osmotic therapy is not recommended prior to the development of brain edema or tissue shifts. However, osmotic therapy with mannitol and/or hypertonic saline is reasonable for patients who develop neurologic deterioration from cerebral edema. (See 'Osmotic and salvage therapies' below.)

●Fever should be treated if present, with the goal of achieving a normal temperature [3]. Fever is associated with unfavorable outcomes in patients with acute stroke, but pharmacologic treatment to lower temperature has not been proven to be beneficial. Therapeutic hypothermia may lower intracranial pressure but has not been demonstrated to be effective for treating malignant hemispheric infarction [5]. A randomized trial of moderate hypothermia after hemicraniectomy for large hemispheric infarction was stopped early, after only 50 patients were randomized, due to higher rate of serious adverse events among the patients treated with hypothermia without any evidence of improved functional outcomes [43]. (See "Initial assessment and management of acute stroke", section on 'Fever'.)

Frequent monitoring — Patients with large hemispheric infarction are at high risk for neurologic deterioration associated with the development of cerebral edema and herniation. Frequent monitoring (eg, hourly) for level of consciousness, ipsilateral pupillary dilation, and onset or progression of other neurologic deficits (including early signs such as new development of ipsilateral Babinski sign or cerebral ptosis [ie, inability to open the eyes despite preserved consciousness]) is recommended to identify patients who might benefit from timely urgent interventions (ie, osmotic and salvage therapies, decompressive hemicraniectomy) that might prevent an irreversible decline and fatal outcome [5]. Hourly neurologic checks are indicated in these patients for at least the first 48 hours following stroke onset.

Routine intracranial pressure (ICP) monitoring is not recommended; a number of studies suggest ICP monitoring has limited or no value for predicting herniation or improving clinical outcome in this setting [5,7,44,45].

Osmotic and salvage therapies — Acute cerebral edema with life-threatening mass effect can be treated, at least temporarily, with hypertonic saline, mannitol, or hyperventilation. These interventions are most effective when used as a bridging therapy to decompressive hemicraniectomy [3], and are not intended for prolonged use. (See "Evaluation and management of elevated intracranial pressure in adults", section on 'Osmotic therapy and diuresis'.)

●Intravenous mannitol quickly and effectively lowers intracranial pressure. Mannitol regimens vary between institutions and regions. It is most common to start with a bolus of 1 to 1.5 g/kg of mannitol 20 percent. Subsequent doses may be scheduled (eg, 0.5 g/kg every 6 hours) or given as needed for signs of clinical decline or radiologic progression. Patients receiving repeated doses of mannitol should have their serum osmolality and osmolar gap monitored to avoid nephrotoxicity; the risk is high when serum osmolality increases above 320 milliosmoles/kg and when the osmolar gap is greater than 20. (See "Evaluation and management of elevated intracranial pressure in adults", section on 'Osmotic therapy and diuresis' and "Complications of mannitol therapy".)

●Hypertonic saline is an effective hyperosmolar agent for lowering intracranial pressure. Bolus doses of various concentrations (from 3 to 23.4 percent) and volumes (most commonly 50 mL of 14 percent or 30 mL of 23.4 percent sodium chloride) can be administered. (See "Management of acute moderate and severe traumatic brain injury", section on 'Osmotic therapy'.)

●Hyperventilation causes a rapid lowering of intracranial pressure by inducing cerebral vasoconstriction, but the effect is short-lived. Brief hyperventilation may be used as a bridge to more definitive therapy for deteriorating patients with signs of brainstem herniation [3,46]. A partial pressure of carbon dioxide (PaCO₂) goal of 30 to 35 mmHg is suggested. More aggressive (ie, a PaCO₂ goal of 26 to 30 mmHg) or prolonged hyperventilation may result in brain ischemia and worse outcomes.

DECOMPRESSIVE HEMICRANIECTOMY

Description and goals of surgery — Decompressive hemicraniectomy (DHC) with durotomy is a surgical technique used to relieve the increased intracranial pressure and brain tissue shifts that occur in the setting of large cerebral hemisphere mass or space-occupying lesions (image 3). The main goal of hemicraniectomy for malignant hemispheric infarction is to prevent further brain injury by reversing mass effect, reducing brain tissue shifts, decreasing intracranial pressure, and improving cerebral perfusion pressure [47]. DHC should be considered as an adjunct to optimal medical care, and never a substitute for it.

In general, the technique involves removal of bone tissue (skull) and incision of the restrictive dura mater covering the brain, allowing swollen brain tissue to herniate upwards through the surgical defect rather than downwards to compress the brainstem (image 4 and image 5). Bone resection should be extensive (ideally 12 cm or greater) to allow adequate brain pressure alleviation and avoid brain tissue injury around the borders of the intact skull.

Efficacy — Evidence from several small randomized controlled trials demonstrates that DHC for massive middle cerebral artery (MCA) territory infarction increases survival, compared with standard medical therapy, but most patients who survive are left with moderately severe to severe disability (ie, a modified Rankin Scale (mRS) (table 2) score of 4 or 5). However, several of the trials were stopped early, which could lead to an overestimation of effect size [48]. In particular, the individual trial risk estimates for achieving a mRS score of ≤3 (moderate disability or less) are of low to moderate certainty. Additionally, these separate trials differed in the upper age limit of included patients and the time window within which surgical decompression was performed.

A 2021 meta-analysis of seven trials, with patient-level data for 488 patients (including unpublished data from one trial), found that a favorable outcome (defined as a mRS ≤3) at one year was more likely with DHC compared with medical treatment alone (37 versus 15 percent, absolute difference, 22 percent, adjusted odds ratio [OR] 2.95, 99% CI 1.55-5.60) [49]. In addition, DHC reduced the risk of death at one year (29 versus 71 percent, absolute difference 42 percent, adjusted OR 0.16, 95% CI 0.10-0.24). The favorable outcome benefits of decompressive surgery were consistent across subgroups defined by age (<60 versus >60 years), sex, presence of aphasia, baseline NIH Stroke Scale score, time to randomization (<24, 24 to 48 and >48 hours), and extent of infarction by vascular territory. However, among patients >60 years old, the proportion who achieved a favorable outcome varied markedly across studies, from 0 to 66 percent, and the authors concluded that the evidence of benefit in older patients remained uncertain. Similarly, too few patients were treated beyond 48 hours to draw reliable conclusions about late treatment. And parallel to the individual trial data, it was again seen that DHC led to increased numbers of patients with moderately severe or severe disability (mRS 4 or 5), although these disability outcomes were not individually analyzed.

Thus, as demonstrated by meta-analyses of randomized controlled trials, the primary benefit of DHC across age groups is increased survival. Though DHC increases the number of survivors with moderate or severe disability as illustrated in the figure (figure 2), it also increases the chances of more favorable outcomes, particularly in patients less than 60 years of age [49-51]. Although these studies assessed clinical efficacy using mRS as the neurological outcome assessment, they did not assess other metrics of outcome such as quality of life or cognitive outcomes.

While DHC confers survival benefit in patients older than 60, it is less clear whether DHC improve clinical outcomes in these patients given that they generally do not achieve good neurological outcomes (ie, moderate or slight disability) regardless of treatment.

The available data also suggest that DHC within 48 hours is more likely to be beneficial than later hemicraniectomy, as noted in earlier meta-analyses [48,52]. Less robust data suggest but do not establish that later surgery (ie, from 48 to 96 hours) may be beneficial [49,53].

Selection of patients — General eligibility criteria for DHC are listed in the table (table 3).

Timing of surgery — Some investigators have advocated early (within 24 hours of onset), prophylactic surgical decompression for large hemispheric infarction, before the appearance of signs of herniation. This approach has intuitive appeal since it is logical that early decompressive surgery (before the development of life-threatening herniation with secondary brain injury) should result in the best clinical outcomes [47]. There is also experimental evidence that early surgery is more likely to be beneficial [54,55]. Furthermore, there is clinical evidence that aggressive medical reversal of a single episode of transtentorial herniation in the absence of radiographic evidence of midbrain injury may permit good long-term outcome [56]. These findings suggest that herniation alone does not preclude benefit from surgical decompression and that medical management may be an appropriate initial therapeutic option to a bridge to DHC.

However, the data are limited and conflicting regarding the utility of selecting patients for surgical decompression based upon time since symptom onset to treatment. Timing of surgery has been analyzed for <24 hours, 24 to 48 hours, and 48 to 96 hours.

●In subgroup analyses of pooled data from three randomized controlled trials, there was no difference in outcome between patients assigned to early (<24 hours after stroke onset, n = 45) or later (24 to 48 hours, n = 38) decompressive surgery [52]. In the HAMLET trial, which evaluated DHC up to 96 hours after stroke onset, there was no benefit on any outcome measure for patients assigned to surgery after 48 hours (n = 25) [35]. However, small numbers of patients treated beyond 48 hours in the meta-analysis [52], HAMLET [35], and HeADDFIRST [57] preclude definitive conclusions about of the possible efficacy of surgical decompression after 48 hours.

●In a 2004 systematic review of 12 studies with over 100 patients who were treated with hemicraniectomy for malignant MCA territory infarction, the time to treatment, clinical signs of herniation, sex, and side of infarction did not predict outcome with decompressive surgery [40]. Other reports have not found time to treatment to be a predictive factor for outcome, although this may be confounded by variables such as age [58].

●A retrospective study from the United States analyzed 1301 patients from the Nationwide Inpatient Sample (2002-2011) who had decompressive craniectomy for stroke [59]. Of these, 726 had surgery within 48 hours. When time was evaluated as a continuous variable, later surgery was associated with increased risk of discharge to institutional care (OR 1.17, 95% CI 1.05-1.31) and poor outcome (OR 1.12, 95% CI 1.02-1.23).

Procedure — The surgical technique of DHC and durotomy varies in the literature. In general, a large question mark-shaped incision is made in the scalp, starting from the midline, that includes the frontal, parietal, and temporal lobes. A large incision is essential to the success of the procedure to avoid both incomplete release of pressure and further injury to the brain where it is forced against the edges of the craniotomy. A diameter of 12 cm or larger is recommended [3].

The protocol for one pilot trial defined required margins of the skull defect as follows [57]:

●Anterior: from the floor of the anterior cranial fossa at the mid-pupillary line

●Posterior: to 4 cm posterior to the external auditory canal

●Superior: to 1 cm lateral to the superior sagittal sinus

●Inferior: to the floor of the middle cranial fossa

Bone is removed and can be stored in a frozen tissue bank or sewn into the peritoneal cavity of the patient if this is to be reused for later cranioplasty (replacement of skull). The dura is opened with a cruciate incision to allow the brain to swell outwards. In some reports, infarcted tissue is resected ("strokectomy") [60,61]. Other surgeons advocate against tissue resection, as the removal of islands of normal functioning tissue not apparent at the time of surgery may worsen outcome. However, removal of large areas of infarcted tissue may be required in certain cases to ensure adequate decompression and reduce the risk of herniation.

Complications — Possible complications of craniectomy include hydrocephalus, external brain tamponade, infections, seizures, and paradoxical herniation [62]. The "sinking skin flap" syndrome (SSFS), also termed "syndrome of the trephined," is a delayed complication of craniectomy that can occur when atmospheric pressure exceeds intracranial pressure [62]. The major clinical features are the sunken appearance of the skin over the skull defect and severe orthostatic headache. Associated symptoms may include focal neurologic deficits, seizures, and altered mental status. Unchecked, SSFS may progress to paradoxical brain herniation, coma, and death.

In the DECIMAL trial, with follow-up data available from 27 patients, SSFS developed at three to five months after hemicraniectomy in three patients (11 percent) [63]. Radiologic SSFS developed in another four patients (15 percent) and was generally asymptomatic except for partial seizures affecting one patient.

Management of SSFS and paradoxical herniation, which is a neurocritical care emergency, requires measures that increase intracranial pressure, such as Trendelenburg position, intravenous hydration, clamping of cerebrospinal fluid drainage, and discontinuation of hyperosmolar measures [62]. In the absence of paradoxical herniation, SSFS may respond to intravenous fluid administration and supine position with head turned down to the side of the craniectomy [63]. Cranioplasty is the definitive treatment for this syndrome.

Independent of SSFS, cranioplasty is a necessary second-phase procedure for stroke survivors with hemicraniectomies. While the timing of cranioplasties varies, there is evidence that earlier cranioplasty (potentially within same hospitalization) is associated with enhanced neurologic recovery [64], potentially attributable to restoration of cerebral hemo- and hydrodynamics.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Stroke in adults".)

SUMMARY AND RECOMMENDATIONS

●Malignant hemispheric infarction – This type of stroke is characterized by the development of space-occupying cerebral edema that is severe enough to produce brain tissue shifts and herniation. Clinical features typically include forced gaze deviation, visual field deficit, hemiplegia, and aphasia or neglect, depending on the hemisphere involved. The most common etiology is cardioembolic or thrombotic occlusion of the internal carotid artery or the proximal segment (stem or M1 segment) of the middle cerebral artery (MCA). The mortality is as high as 78 percent due to herniation of the temporal lobe on to the brainstem. (See 'Description' above and 'Presentation and progression' above.)

●Predictors of malignant edema – Markers that may predict the development of malignant edema after a large hemispheric infarction include the early onset of decreased consciousness, younger age, higher National Institutes of Health Stroke Scale (NIHSS) scores on admission, and evidence of early infarction involving >50 percent of the MCA on initial head computed tomography (CT) or magnetic resonance imaging (MRI). (See 'Clinical predictors of progression' above and 'Radiologic predictors of progression' above.)

●Medical treatment options – These include management of increased intracranial pressure and cerebral edema, and supportive medical care or withdrawal of medical care for those who do not desire aggressive therapy. Standard medical therapy for large hemispheric infarctions involves close monitoring for neurologic complications, such as depressed mental status and herniation; management of intracerebral pressure, with measures such as elevation of the head of the bed, osmotic therapy, and brief periods of hyperventilation as needed; and prevention of secondary complications such as aspiration and deep venous thrombosis. (See 'Medical management' above and 'Palliative care' above.)

●Outcomes with hemicraniectomy – Decompressive hemicraniectomy (DHC) for malignant hemispheric infarction substantially reduces mortality. However, most surviving patients are left with major disability. Thus, the dilemma for patients, families, and medical decision-makers is that surgery may leave patients alive with severe disability, while medical management alone most often results in death. Among patients older than 60 years of age with malignant hemispheric infarction, DHC can increase survival, yet it is less clear whether DHC will confer clinical benefit in these patients, and most who survive are left with severe to very severe disability. (See 'Surgical decompression or medical care only?' above and 'Efficacy' above.)

Possible complications of DHC include hydrocephalus, infections, seizures, intracranial hemorrhage and fluid collections, sinking skin flap syndrome, and paradoxical herniation. (See 'Complications' above.)

●Our approach – For patients age 60 years or younger with infarction involving >50 percent of the MCA territory associated with a decreased level of consciousness, who are thus at high risk of developing malignant edema, we suggest decompressive hemicraniectomy (DHC) if surgery can be initiated within 48 hours of stroke onset (Grade 2B). Hemicraniectomy should occur only after a thorough discussion with the patient (if cognizant), family, or appropriate medical decision-makers (eg, health care proxy); they should understand that hemicraniectomy increases the odds of survival and may increase the chance of surviving with only moderate disability, but most patients who survive are left with severe to very severe disability. The alternative treatment of aggressive medical therapy for malignant edema may be considered if surgery is deferred. (See 'Surgical decompression or medical care only?' above and 'Decompressive hemicraniectomy' above and 'Medical management' above.)

ACKNOWLEDGMENT — The editorial staff at UpToDate acknowledges Mitchell SV Elkind, MD, MS, FAAN, who contributed to an earlier version of this topic review.