Management of Large Hemispheric Infarction

K.E. Wartenberg and S.A. Mayer

z Introduction: Natural History of Large Hemispheric Infarction

Large hemispheric infarctions due to middle cerebral artery (MCA) or internal carotid artery (ICA) occlusion are an important cause of morbidity and mortality in the neurological intensive care unit (ICU). Neurological deterioration occurs as a consequence of malignant cerebral edema in approximately 5±10% of hemispheric ischemic strokes [1±3], but in over two-thirds of patients when the complete MCA territory is infarcted [1, 3]. The reported mortality of these `malignant' hemispheric infarctions varies between 42 and 80% [1±5]. Patients with complete MCA infarction are generally 10 years younger (mean age 56 years) than the average stroke patient [1]. The initial presentation usually in- cludes contralateral conjugate gaze paresis, hemineglect, and reduced level of consciousness in addition to the expected sensorimotor and language deficits [1, 6, 7]. Most patients experience neurological decline within 48 hours [1, 3]. Of those who deteriorate, worsening occurs within 24 hours in 36% and within 48 hours in 68% [3]. The first sign of transtentorial herniation is usually drowsiness, followed by pupillary asymmetry, hyperventilation, and contralateral motor posturing [8, 9]. Autonomic abnormalities may include hyper- or hypoventilation, bradycardia, and sustained hypertension or blood pressure lability [1, 3, 5, 7]. Bilateral motor posturing and lower extremity rigidity then follows as the midbrain and diencephalon are subjected to physical distortion and compression [8]. Without life support, death typically occurs within five days [1, 3, 5] as a result of brain death, respirato- ry failure, cardiac arrhythmia, or pneumonia [1±3]. Infarction of the brain parenchyma and the vasculature results in a delayed break down of the blood brain barrier with extravasation of serum proteases and worsening of brain edema 24 to 72 hours after the initial infarct signs [9]. Hemi- spheric brain swelling leads to brain tissue shifting with subsequent brain stem distortion, bihemispheric dysfunction through mechanical displacement, vascular compression, uncal and transtentorial herniation (Fig. 1). Intracranial pressure (ICP) is usually not elevated early in the process of transtentorial hermiation from large hemispheric infarction, but increases later as severe cytotoxic edema ensues. Ongoing ischemia is usually not the cause of neurological deterioration beyond 24 hours of onset, but this can result from vascular compression of the anterior and posterior cerebral arteries against the falx or tentorium, and is a universal finding in patients who become brain dead [4]. 648 K.E. Wartenberg and S. A. Mayer

Fig. 1. Schematic diagram of the importance of tissue shifts and hypothetical significance of pressure differentials in clinical worsening from large hemispheric infarction with edema. P1 represents the pressure in the injured hemisphere and P2 the pressure in the uninjured hemisphere. As edema ensues, pressure differentials occur and accentuate, leading to tissue shifts and clinical worsening. From [4] with permission

z Etiology of Large Hemispheric Infarctions

Large hemispheric infarctions occur as the consequence of an occlusion of the dis- tal ICA or proximal MCA trunk without sufficient collateral flow (Figs. 2 and 3). Total ICA occlusions lead to infarction of the anterior cerebral artery (ACA) and MCA territories [7]. Most patients have risk factors for vascular disease such as hypertension, dia- betes, hypercholesteremia, tobacco abuse, history of transient ischemic attacks or ischemic strokes, congestive heart failure (CHF), and coronary artery disease. Atrial fibrillation is more frequent in patients with MCA and ICA territory strokes compared to the remaining stroke population [1±3, 6]. ICA dissection is a significant cause of large territory infarctions in younger patients (12%) [6]. In one series of 610 patients with large hemispheric strokes 42% were attributed to focal or general atherosclerosis and 33% to a cardioembolic source [6].

z Diagnosis of Early MCA Infarction

Computed tomography (CT) of the brain obtained within 6 hours of symptom onset has a sensitivity of 82% for ischemic hemispheric infarctions [10]. Early infarct signs on CT include: z Hyperdense MCA sign (high contrast in the MCA that is brighter than the adjacent brain tissue and other intracranial arteries in the absence of calcification) (Fig. 4) Management of Large Hemispheric Infarction 649

Fig. 2. a ICA occlusion after the bifurcation demonstrated by cerebral angiography with a common carotid artery injection. b MR Angiogram of the Circle of Willis shows no flow signal in the left ICA and cross- filling of the left MCA via the anterior communicating artery z Hyperdense ICA sign distinguishable from the opposite ICA and the surround- ing bone z Obscuration of the lentiform nucleus defined by decreased density compared to the contralateral nucleus z Effacement of the sylvian fissure with loss of grey-white matter distinction compared to the contralateral side z Involvement of other vascular territories seen as hypodensity in the ACA, anterior choroidal and posterior cerebral arteries (PCA) z Complete sylvian fissure obscuration and extensive effacement of the hemisphere as well as compression of the lateral ventricle demonstrating mass effect z Midline shift at the level of the pineal gland and the septum pellucidum (antero- septal shift) [2, 10] z Predictors of Fatal Deterioration

Several studies have identified risk factors for secondary fatal neurological deterioration after MCA infarction. A multivariate analysis of 201 patients with large hemispheric strokes [2] identified the following predictors of fatal brain swelling: z History of hypertension z History of CHF z An elevated white blood count (WBC) 650 K.E. Wartenberg and S. A. Mayer

Fig. 3. MCA main stem occlusion by cerebral angiogram with left common carotid injection

z CT involvement of > 50% of MCA territory (Fig. 5), and z CT involvement of additional territories [1, 2, 10, 11].

In a series of 37 patients with MCA stroke and proximal vessel occlusion a National Institute of Health Stroke Scale (NIHSS) on admission of 19 or greater was found to be highly predictive of severe neurological deterioration (sensitivity 96%, specificity 72%) [12]. Several studies have found that radiographic evidence of a large initial infarction volume can reliably identify those at greatest risk for neurological deterioration. In a case control study of 31 patients studied with contrast CT, attenuated corticomedullary contrast enhancement involving the entire MCA territory within 18 hours of onset was found to be the most reliable neuroradiological predictor of neurological deterioration, with a sensitivity of 87% and specificity of 97% [13]. In another study horizontal pineal displacement greater than 4 mm on CT performed within 48 hours of stroke onset was highly predictive of mortality with a specificity of 89% and a sensitivity of 46% in 127 patients [14]. In an analysis of magnetic resonance imaging (MRI) predictors, a reduction in the apparent diffusion coefficient (ADC) of greater than 82 ml was the most accurate predictor of deterioration, with Management of Large Hemispheric Infarction 651

Fig. 4. a Hyperdense MCA sign (right) on CT surrounded by hypoattenuation in the right frontal and temporal regions with loss of sulci and grey-white matter differention. b The Fluid Attenuated Inversion Re- covery (FLAIR) sequence reveals high signal in the right MCA consistent with a thrombus a sensitivity of 87% and specificity of 91% [12]. This finding is supported by a retrospective analysis that identified a diffusion weighted imaging (DWI) lesion volume exceeding 145 ml within 14 hours of symptom onset as the best predictor of a malignant clinical course, with a sensitivity and specificity of 100% in a multivariate model [15]. Krieger et al. identified nausea and vomiting within 24 hours of stroke onset, systolic blood pressure (SBP) >180 mmHg after 12 hours, and involvement of >50% of the MCA territory on CT as independent predictors of fatal brain swelling in a multivariate analysis of 135 patients [16]. Carotid ªTº occlusion was significantly associated with a fatal outcome in 74 MCA infarction patients with acute carotid artery distribution stroke [17]. Severe cerebral blood flow (CBF) reductions in the MCA territory, detected by Xenon-CT (mean CBF 8.6 ml/100 gm/minute) [18] or single photon emission CT (SPECT) [19] can also identify patients at risk for fatal brain edema. Neurochemical monitoring with cerebral microdialysis is another interesting tool to monitor the course of MCA infarction. An increase in extracellular glutamate, glycerin and lactate concentration and an augmentation of the lactate/pyruvate ra- tio in peri-infarct areas was thought to reflect developing brain edema with subsequent secondary neuronal ischemia as those changes of neurochemicals preceded an increase in ICP [20±22]. Bosche et al. found significantly lower non-transmitter amino acid concentrations in the areas adjacent to the infarct in patients who de- veloped malignant brain edema [23]. In summary, the CT criteria involvement of >50% of the MCA territory and other vascular territories, and the presence of a midline shift at the level of the pineal gland of septum pellucidum represent the most reliable predictors of fatal neurological deterioration. 652 K.E. Wartenberg and S. A. Mayer

Fig. 5. a CT showing hypoattenuation in the total right MCA and ACA territory with effacement of the adjacent sulci and loss of the grey-white matter distinction. b Left:Diffusion-weighted MRI sequence (DWI) demonstrating restricted diffusion in the right MCA territory compatible with early acute ischemia. Right: Effacement of the sulci, loss of grey-white matter differentiation and local mass effect on the right lateral ventricle on the FLAIR sequence corresponding to panel on the left. c DWI (left) and T2-weighted (right) MRI sequence revealing significant cerebral edema from a left MCA territory infarction seen as hyperintense signal with effacement of the lateral ventricle and >2 cm midline shift at the level of the septum pellucidum Management of Large Hemispheric Infarction 653 z Acute Management

The mainstay of supportive care for large hemispheric infarction is endotracheal intubation and mechanical ventilation for airway protection and depressed level of consciousness. The costs of critical care management of stroke patients treated with mechanical ventilation has been calculated to be (1996 USD) $ 89400 for every patient discharged alive, and $ 174200 for each quality-adjusted life year saved. The functional status of most survivors is poor, with over 50% left severely disabled and completely dependent [24]. For this reason, aggressive efforts to attain reperfu- sion within an early time window or to minimize brain injury are justified. Intrave- nous recombinant tissue plasminogen activator (rt-PA) administered within 3 hours of onset of the first stroke symptom is the only FDA approved acute stroke treatment [25]. The PROACT trials demonstrated safety and efficacy of intra-arterially administered recombinant pro-urokinase within 6 hours of stroke onset. The reca- nalization rate was 66% with a 10% risk of intracerebral hemorrhage, and the like- lihood of survival with a good outcome was significantly increased [26, 27]. Several other methods to potentially salvage hypoperfused but not yet infarcted brain tissue are currently under investigation, such as bridging of intravenous and intra-arterial t-PA, mechanical thrombus removal, intra-arterial abciximab, stenting, and angioplasty. Induced hypertension with norepinephrine or phenylephrine augments cerebral perfusion pressure (CPP) and mean flow velocity and may improve perfusion in areas at risk for further infarction (e.g. the ischemic penumbra) [28]. One preliminary prospective clinical report indicates that a trial of induced hypertension can result in clinical improvement of neurological signs in 54% of patients, usually within 20 minutes [29]. z Intensive Care Management

Patients with large hemispheric strokes are best managed in an ICU. The mainstays of treatment for massive cerebal edema include osmotherapy and hyperventilation [7, 30]. The goal of osmotherapy is to reduce brain volume by creating an osmotic gradient between the intracellular and extracellular compartment [7]. Mannitol infusion leads to a reduction in brain water content first by creating an osmotic gradient between the interstitial and intracellular compartments and the intravascular space across the semipermeable blood brain barrier. Mannitol also reduces blood viscosity and improves microvascular CBF, which may result in reflex vasoconstriction and reduced cerebral blood volume, and has free radical scavenger properties. It has a fast onset of action, which is helpful in cases of impending transtentorial herniation. The recommended dose is 0.5±1.5 g/kg i.v. every 1 to 6 hours. Complications may include transient intravascular fluid overload, secondary volume contraction or hypokalemia resulting from the diuretic effect of repeated doses over longer intervals [7, 30]. In a series of seven patients with large hemispheric infarction the administration of a bolus of 1.5 g/kg mannitol did not have any effect on midline shift measured by MRI or neurological status [31]. It has also been suggested that mannitol might have a greater effect on normal rather than infarcted tissue due to a disruption of the blood brain barrier [4]. At least theoreti- cally, this might result in worsening of midline shift as mannitol extravasates in the infarcted tissue and dehydrates the normal contralateral hemisphere. In clinical 654 K.E. Wartenberg and S. A. Mayer

practice, however, this does not occur, and there are animal studies demonstrating that mannitol dehydrates infarcted tissue at least as effectively as normal brain tissue [32]. Hypertonic saline solutions improve microvascular perfusion and CPP by caus- ing volume expansion, increases in cardiac output and systemic blood pressure, and decreased production of cerebrospinal fluid (CSF). Hypertonic saline also can modify inflammatory responses, interact with the neuroendocrine system and expand intracranial elastance. We use continuous infusion of 3% saline solution or 23.4% saline boluses for ICP control in our neurocritical ICU. There are only a few prospective clinical trials that have investigated the effect of hypertonic saline in fatal large hemispheric infarctions with conflicting results, and none of these studies have evaluated functional outcome [30]. In one study, hypertonic saline hydroxyethyl starch solution successfully decreased elevated ICP by 34% within 15 minutes in patients with space-occupying infarctions [33]. In general, the administration of hypertonic saline over a longer period seems to be safe in the absence of renal failure and CHF, though careful monitoring of central venous pressure (CVP) and fluid balance is prudent. There is no evidence supporting the use of glucocorticoids for the treatment of brain edema from acute stroke [30]. Barbiturates decrease the cerebral metabolic rate, resulting in diminished CBF and blood volume with reduction of cerebral edema formation. The neuroprotective properties (free radical scavenging) of barbiturates make them attractive for the treatment of brain swelling related to ischemia on a theoretical basis [30], but practical experience suggests that barbiturates may in fact be harmful when used to control ICP in patients with MCA infarction. Of 60 patients with large hemispheric infarctions who were subjected to barbiturate coma with thiopental (3±5 mg/kg bolus followed by continuous infusion) to achieve burst suppression pattern on a continuous electroencephalogram (EEG) for 48 hours, only 5 patients (8%) survived. Thiopental infusion resulted in a decrease of ICP, which was not sustained and was complicated by arterial hypotension requir- ing vasopressors, pneumonia, sepsis and hepatic dysfunction [34]. Hypotension in this setting may have been provoked by previous dehydration from osmotherapy [7]. Hyperventilation has long been considered a mainstay of management for increased ICP and cerebral edema. Decreased PCO2 (30±35 mmHg) is achieved by in- creasing the ventilation rate, leading to vasoconstriction, diminished cerebral blood volume, and lower ICP. This effect is rapid ± a reduction in ICP usually occurs within 30 minutes ± but may be limited by excessive vasoconstriction resulting in further cerebral ischemia if the patient is excessively hyperventilated. Normocapnia should be reinstituted slowly because of a possible rebound effect [7, 30]. Hyper- ventilation is a useful tool for the acute stabilization of impending herniation, but is often not effective over longer periods of time.

z Decompressive Surgery

Hemicraniectomy for large hemispheric infarction was first reported in 1935. Hori- zontal and vertical tissue shifts, and ventricular and vascular compression by massive brain edema are relieved by removal of the bone flapover the frontal, temporal and parietal lobe at the infarct site. This allows the edematous brain to expand ex- Management of Large Hemispheric Infarction 655 tracranially, improves CPP and retrograde flow in the MCA, preserves CBF, and may prevent further ongoing ischemia. The diameter of the craniectomy should be at least 12 cm (14 to 15 cm anterior-posterior, and 10 to 12 cm from the temporal base to the vertex is recommended) [7]. A small diameter hemicraniectomy can result in compression and kinking of bridging veins, or mushroom-like herniation of the brain with shearing distortion and additional ischemic lesions (Fig. 6) [35].

Fig. 6. CT of a 35 year old woman with left MCA infarction (day 0) who deteriorated clinically and devel- oped additional left ACA territory infarction and significant midline shift to the right on day 2. After decompressive hemicraniectomy the CT showed slight improvement of the midline shift on day 3. On day 5 she proceeded to infarct the left PCA territory because the diameter of the hemicraniectomy was too small 656 K.E. Wartenberg and S. A. Mayer

After resection of the temporal bone to the skull base the dura is opened, adjusted, and a biconvex dural patch is placed into the incision (duroplasty). Resection of the infarction is not advisable as the margins between infarct and penumbra are poorly defined. The bone flap can be conserved in the abdominal subcutaneous tissue or in a cooled sterile isotonic solution. Reimplantation of the bone flap is possible 6 to 12 weeks after removal, once the swelling has resolved. Potential complications include intracranial, wound and bone flapinfection, subdural and epidural hematoma, subdural CSF hygroma, paradoxical herniation after the swelling period and hydrocephalus [7, 30, 35, 36]. The first prospective but uncontrolled clinical trial of hemicraniectomy included 32 patients with a midline shift of more than 10 mm at the septum pellucidum level. Most of the patients who underwent decompressive surgery at a mean time of 39 hours after stroke onset suffered a non-dominant MCA territory infarction and did not have any medical complications. The mortality rate in the surgical group was 34%. The control group encompassed patients with more severe medical co- morbidities and inability to consent and had a mortality rate of 76% [36]. The next prospective, uncontrolled trial included 31 patients undergoing hemicraniectomy considerably earlier in the course (within the first 24 hours of symptom onset), with a mortality rate of 16% [37]. In another prospective, non-randomized series of 34 patients, mortality at 3 months was 67% in the conservatively managed group and 16% in the groupreceiving surgery immediately after the onset of neurological deterioration [38]. Ultra-early hemicraniectomy performed within 6 hours of stroke onset in 12 patients resulted in a mortality of 8% compared to 36% in 30 patients undergoing surgery after 6 hours of symptom onset and 80% in 10 patients managed with maximal medical therapy [39]. Hemicraniectomy is clearly a life-saving procedure. However, analysis of functional outcome of patients after hemicraniectomy has revealed conflicting results. Multiple case series, clinical trials and a metaanalysis suggest that timing of hemicraniectomy and age of the patients are crucial factors in determining outcome: Early surgery has a greater impact on reduction of mortality, and young patients (<50 years) tend to have a better outcome [37±41]. Although many clinicians are reluctant to offer hemicraniectomy to patients with dominant hemisphere infarcts, a meta-analysis found no difference in functional outcome comparing left versus right-sided [40]. A retrospective analysis of 188 patients identified age 50 years and involvement of more than one vascular territory as predictors of death among patients undergoing hemicraniectomy [42]. In another analysis of long-term out- comes, none of 36 patients who had a hemicraniectomy attained an independent outcome at a 6-month follow-up[43]. The HeaDDFIRST (Hemicraniectomy and Durotomy for Deterioration From In- farction Related Swelling Trial) was the first multicenter, prospective, randomized trial that investigated mortality and functional outcome in patients undergoing hemicraniectomy versus comprehensive standardized medical therapy. Enrollment in this study was limited to 26 randomized patients. Mortality was reduced from 46 to 27%, but this reduction was not statistically significant [44]. Two multicenter, prospective randomized trials comparing hemicraniectomy and medical management are still ongoing: z DESTINY (Decompressive Surgery for the Treatment of Malignant Infarction of the Middle Cerebral Artery) and z HAMLET (Hemicraniectomy After MCA infarction with Life-threatening Edema Trial). Management of Large Hemispheric Infarction 657 z Hypothermia

Mild systemic hypothermia (33±368C) can be achieved with surface and intravascular cooling devices. Cooling decreases cerebral metabolism, preserves the blood brain barrier, and reduces inflammatory responses as well as excitotoxic neuro- transmitter release [7, 30]. In 25 patients with large MCA territory infarcts hypothermia (338C) was started after an interval of 14 hours after stroke onset and maintained for 48±72 hours, which significantly reduced the ICP. The patients were subjected to passive rewarming (17 to 24 hours) which resulted in a continuous rise in ICP with subsequent herniation and death in 9 patients thought to be attribut- able to a hypermetabolic response (44% mortality). Complications of hypothermia included shivering, sepsis, pneumonia, thrombocytopenia, coagulopathy, and eleva- tion of serum amylase and lipase levels. Cardiac arrhythmias, including prolonga- tion of the PR and QT interval, ventricular ectopy and fibrillation, are adverse effects that limit cooling to levels below 328C [7, 45]. More gradual controlled rewarming appears to be safer than passive or active rewarming. In one series of MCA infarction patients treated with hypothermia, a mean temperature increase of 0.1 to 0.28C over 2 to 4 hours correlated with a more gradual increase in ICP and improved cerebral cellular and metabolic compensation mechanisms [46]. In a large prospective uncontrolled trial of 50 patients undergoing mild-to-moderate hypothermia (32±338C) for 72 hours after large hemispheric infarction ICP was significantly reduced [47]. In this study, passive rewarming within 16 hours was associated with a pronounced rise in ICP compared to controlled rewarming over a longer period of time (>16 hours) (Fig. 7). The most common side effects of hypothermia in this study were arrhythmia (sinus bradycardia, prolonged PR and QT intervals), arterial hypotension, pneumonia, decreased serum potassium, decreased platelet count, and coagulopathy. Mor- tality was 38% overall. All patients received midazolam and propofol for sedation, morphine and fentanyl for analgesia, and neuromuscular blockade with vecuronium and atracurium during hypothermia and rewarming [47]. In summary, the use of hypothermia for control of cerebral edema in large hemispheric infarction is feasible, but the risks are prominent. It seems most likely that cooling shows greater promise as an acute intervention designed to reduce infarct volume, than a primary form of treatment for malignant brain edema.

Fig. 7. ICP levels after rewarming of different durations (>16 and <16 hours). From [47] with permission 658 K.E. Wartenberg and S. A. Mayer

z Conclusion

Space-occupying edema after large hemispheric infarctions is difficult to control with conservative intensive medical management including osmotherapy, hyperventilation and barbiturate coma. It is of major importance to predict which patients may experience severe neurological deterioration from tissue shifts and edema expansion. Decompressive surgery is extremely effective in decreasing mortality, especially when performed early in the course, but the quality of life of those who survive may be poor, particularly if the patient is older than 50 years. Hypothermia is a promising additional tool for short-term ICP control but our experience suggests that it is not as robust as hemicraniectomy as a life-saving procedure for patients with malignant edema. Additional research is needed to developICU treatment strategies that minimize brain edema and the need for hemicraniectomy among victims of large hemispheric infarction.

References

1. Hacke W, Schwab S, Horn M, Spranger M, De Georgia M, von Kummer R (1996) `Malig- nant' middle cerebral artery territory infarction: Clinical course and prognostic signs. Arch Neurol 53:309±315 2. Kasner SE, Demchuk AM, Berrouschot J, et al (2001) Predictors of fatal brain edema in massive hemispheric ischemic stroke. Stroke 32:2117±2123 3. Qureshi AI, Suarez JI, Abutaher MY, et al (2003) Timing of neurologic deterioration in massive middle cerebral artery infarction: A multicenter review. Crit Care Med 31:272±277 4. Frank JI (1995) Large hemispheric infarction, deterioration, and intracranial pressure. Neu- rology 45:1286±1290 5. Berrouschot J, Sterker M, Bettin S, Kæster J, Schneider D (1998) Mortality of space-occupying (`malignant') middle cerebral artery infarction under conservative intensive care. Inten- sive Care Med 24:620±623 6. Heinsius T, Bogousslavsky J, Van Melle G (1998) Large infarcts in the middle cerebral artery territory: Etiology and outcome patterns. Neurology 50:341±350 7. Steiner T, Ringleb P, Hacke W (2001) Treatment options for large hemispheric stroke. Neu- rology 57 (Suppl 2):S61±S68 8. Ropper AH, Shafran B (1984) Brain edema after stroke. Clinical outcome and intracranial pressure. Arch Neurol 41:26±29 9. Ayata C, Ropper AH (2002) Ischaemic brain oedema. J Clin Neurosci 9:113±124 10. Von Kummer R, Nolte PN, Schnittger H, Thron A, Ringelstein EB (1996) Detectability of cerebral hemisphere ischaemic infarcts by CT within 6 h of stroke. Neuroradiology 38:31±33 11. Moulin T, Cattin F, Crepin-Leblond T, et al (1996) Early CT signs in acute middle cerebral artery infarction: Predictive value for subsequent infarct locations and outcome. Neurology 47:366±375 12. Thomalla GT, Kucinski T, Schoder V, et al (2003) Prediction of malignant middle cerebral artery infarction by early perfusion- and diffusion-weighted magnetic resonance imaging. Stroke 34:1892±1900 13. Haring HP, Dilitz E, Pallua A, et al (1999) Attenuated corticomedullary contrast: An early cerebral computed tomography sign indicating malignant middle cerebral artery infarction. A case control study. Stroke 30:1076±1082 14. Pullicino PA, Alexandrov AV, Shelton JA, Alexandrova NA, Smurawska LT, Norris JW (1997) Mass effect and death from severe acute stroke. Neurology 49:1090±1095 15. Oppenheim C, Samson Y, Mana R (2000) Prediction of malignant middle cerebral artery infarction by diffusion-weighted imaging. Stroke 31:2175±2181 Management of Large Hemispheric Infarction 659

16. Krieger DW, Demchuk AM, Kasner SE, Jauss M, Hantson L (1999) Early clinical and radi- ological predictors of fatal brain swelling in ischemic stroke. Stroke 30:287±292 17. Kucinski T, Koch C, Gryzska U, Freitag HJ, Kræmer H, Zeumer H (1998) The predictive value of early CT and angiography for fatal hemispheric swelling in acute stroke. Am J Neuroradiol 19:839±846 18. Firlik AD, Yonas H, Kaufmann AM, et al (1998) Relationshipbetween cerebral blood flow and the development of swelling and life-threatening herniation in acute stroke. J Neurosurg 89:243±249 19. Berrouschot J, Barthel H, von Kummer R, Knapp WH, Hesse S, Schneider D (1998) 99m technetium-ethyl-cysteinate-dimer single-photon emission CT can predict fatal ischemic brain edema. Stroke 29:2556±2562 20. Schneweis S, Grond M, Staub F, et al (2001) Predictive value of neurochemical monitoring in large middle cerebral artery infarction. Stroke 32:1863±1867 21. Berger C, Annecke A, Aschoff A, Spranger M, Schwab S (1999) Neurochemical monitoring of fatal middle cerebral artery infarction. Stroke 30:460±463 22. Heiss WD, Dohmen C, Sobesky J, et al (2003) Identification of malignant brain edema after hemispheric stroke by PET-imaging and microdialysis. Acta Neurochir 86:237±240 23. Bosche B, Dohmen C, Graf R, et al (2003) Extracellular concentrations of non-transmitter amino acids in peri-infarct tissue of patients predict malignant middle cerebral artery infarction. Stroke 34:2908±2915 24. Mayer SA, Copeland D, Bernardini GL, et al (2000) Cost and outcome of mechanical ventilation for life-threatening stroke. Stroke 31:2346±2353 25. The National Institute of Neurological Disorders, and Stroke rt-PA Study Group(1995) Tis- sue plasminogen activator for acute ischemic stroke. N Engl J Med 333:1581±1587 26. Del Zoppo GJ, Higashida RT, Furlan AJ, et al (1998) PROACT: A phase II randomized trial of recombinant pro-urokinase by direct arterial delivery in acute middle cerebral artery stroke. Stroke 29:4±11 27. Furlan AJ, Higashida RT, Wechsler LR (1999) Intra-arterial prourokinase for acute ischemic stroke: The PROACT study: A randomized controlled trial. JAMA 282: 2003±2011 28. Schwarz S, Georgiadis D, Aschoff A, Schwab S (2002) Effects of induced hypertension on intracranial pressure and flow velocities of the middle cerebral arteries in patients with large hemispheric stroke. Stroke 33:998±1004 29. Rordorf G, Koroshetz WJ, Ezzeddine MA, Segal AZ, Buonnanno FS (2001) A pilot study of drug-induced hypertension for treatment of acute stroke. Neurology 56:1210±1213 30. Hofmeijer J, van der Worp B, Kappelle J (2003) Treatment of space-occupying cerebral infarction. Crit Care Med 31:617±625 31. Manno EM, Adams RE, Derdeyn CP, Powers WJ, Diringer MN (1999) The effects of mannitol on cerebral edema after large hemispheric cerebral infarct. Neurology 52: 583±587 32. Paczynski RP, He YY, Diringer MN, Hsu CY (1997) Mulitple-dose mannitol reduces brain water content in a rat model of cortical infarction. Stroke 28:1437±1443 33. Schwarz S, Schwab S, Bertram M, Aschoff A, Hacke W (1998) Effects of hypertonic saline hydroxyethyl starch solution and mannitol in patients with increased intracranial pressure after stroke. Stroke 29:1550±1555 34. Schwab S, Spranger M, Schwarz S, Hacke W (1997) Barbiturate coma in severe hemispheric stroke: useful or obsolete? Neurology 48:1608±1613 35. Wagner S, Schnippering H, Aschoff A, Koziol JA, Schwab S, Steiner T (2001) Suboptimum hemicraniectomy as a cause of additional cerebral lesions in patients with malignant infarction of the middle cerebral artery. J Neurosurg 94:693±696 36. Rieke K, Schwab S, Krieger D, et al (1995) Decompressive surgery in space-occupying hemispheric infarction: Results of a open, prospective trial. Crit Care Med 23:1576±1587 37. Schwab S, Steiner T, Aschoff A, et al (1998) Early hemicraniectomy in patients with complete middle cerebral artery infarction. Stroke 29:1888±1893 38. Mori K, Aoki A, Yamamoto N, Horinaka N, Maeda M (2001) Aggressive decompressive surgery in patients with massive hemispheric embolic cerebral infarction associated with severe brain swelling. Acta Neurochir (Wien) 143:483±492 39. Cho DY, Chen TC, Lee HC (2003) Ultra-early decompressive craniectomy for malignant middle cerebral artery infarction. Surg Neurol 60:227±233 660 K.E. Wartenberg and S. A. Mayer:Management of Large Hemispheric Infarction

40. Gupta R, Connolly ES, Mayer SA, Elkind MSV (2004) Hemicraniectomy for massive middle cerebral artery territory infarction: A systematic review. Stroke 35:539±543 41. Walz B, Zimmermann C, Bættger S, Haberl RL (2002) Prognosis of patients after hemicraniectomy in malignant middle cerebral artery infarction. J Neurol 249:1183±1190 42. Uhl E, Kreth FW, Elias B, et al (2004) Outcome and prognostic factors of hemicraniectomy for space occupying cerebral infarction. J Neurol Neurosurg Psychiatry 75:270±274 43. Foerch C, Lang JM, Krause J, et al (2004) Functional impairment, disability, and quality of life outcome after decompressive hemicraniectomy in malignant middle cerebral artery infarction. J Neurosurg 101:248±254 44. Frank JI, Chyatte D, Thisted R, et al (2003) Hemicraniectomy and durotomy upon deterioration from infarction related swelling trial (HeADDFIRST): First public presentation of the primary study findings. Neurology 60 (Suppl 1):S52.004 45. Schwab S, Schwarz S, Spranger M, Keller E, Bertram M, Hacke W (1998) Moderate hypothermia in the treatment of patients with severe middle cerebral artery infarction. Stroke 29:2461±2466 46. Steiner T, Friede T, Aschoff A, Schellinger PD, Schwab S, Hacke W (2001) Effect and feasibility of controlled rewarming after moderate hypothermia in stroke patients with malignant infarction of the middle cerebral artery. Stroke 32: 2833±2835 47. Schwab S, Georgiadis D, Berrouschot J, Schellinger PD, Graffagnino C, Mayer SA (2001) Feasibility and safety of moderate hypothermia after massive hemispheric infarction. Stroke 32:2033±2035