Critical Care of Patients With Acute Ischemic and Hemorrhagic Stroke: Update on Recent Evidence and International Guidelines FREE TO VIEW

The management of ischemic and hemorrhagic stroke is a formidable challenge often guided by less than satisfactory clinical trial evidence. This review focuses on some of the acute clinical problems faced by those providing care to stroke patients in the critical care setting. For each area reviewed, we discuss recently completed and ongoing clinical trials and summarize current guideline recommendations.

A comprehensive review of Medline and PubMed databases was performed using the following keywords: cerebral infarction, hemicraniectomy, acute stroke, BP, mannitol, hypertonic saline, hypothermia, and intracerebral hemorrhage (ICH). International guidelines were reviewed, and content was compared across guidelines. These included the most recent guidelines from the American Stroke Association (ASA) from 2007 and 2010,^1,2 the National Institute for Health and Clinical Excellence (NICE) from 2008,³ the Canadian Stroke Strategy (CSS) from 2008,⁴ and the European Stroke Organization (ESO) from 2009.⁵ The Cochrane Library (http://www.thecochranelibrary.com) was searched for relevant content and ongoing trials, and the Internet Stroke Center at Washington University (http://www.strokecenter.org) was referenced to identify ongoing trials.

Both high-admission and low-admission BP have been associated with poorer outcome after acute stroke. Theoretically, elevations in pressure can worsen outcome by leading to increased cerebral edema and risk of hemorrhagic conversion, and decreases can worsen ischemia in penumbral areas. The largest trial of BP medications following acute stroke, the Controlling Hypertension and Hypotension Immediately Post-Stroke (CHHIPS) trial, was published recently.⁶ Early BP reduction with either labetalol or lisinopril for patients with a systolic BP > 160 mm Hg was not associated with any adverse outcomes at 2 weeks and was associated with a borderline significant decrease in mortality at 3 months (P = .05). Recruitment was lower than planned, so the study had inadequate power to detect an early benefit from treatment. The long-term decrease in mortality seen in CHHIPS is consistent with the effect of an oral angiotensin receptor blocker, candesartan, in the Acute Candesartan Cilexetil Therapy in Stroke Survivors Study Group (ACCESS) trial.⁷ In that trial, the benefit was due primarily to a decrease in myocardial ischemic events.

Small case series using phenylephrine⁸ and norepinephrine⁹ and one small randomized trial of nine subjects treated with phenylephrine followed by oral midodrine, fludrocortisones, and sodium chloride¹⁰ suggest that pressor therapy is safe, feasible, and sometimes associated with symptomatic improvement. A recent Cochrane review of this topic concluded that it is unclear if BP should be altered in acute stroke.¹¹

The Scandinavian Candesartan Acute Stroke Trial (SCAST) completed enrollment at 2,025 subjects, and results are pending. The Efficacy of Nitric Oxide in Stroke (ENOS) trial is assessing the effect of a nitric oxide donor, transdermal glyceryl trinitrate, in acute stroke, and > 1,900 subjects have been enrolled to date. The Continue or Stop Post-Stroke Antihypertensives Collaborative Study (COSSACS) has completed enrollment, and results are pending. All these trials include ischemic and hemorrhagic stroke patients.

The guidelines consistently recommend using caution when lowering BP following ischemic stroke. Immediately withholding antihypertensive medication is generally recommended unless systolic BP exceeds 220 mm Hg or diastolic BP exceeds 120 mm Hg. When treated, a goal reduction of approximately 15% in the first 24 h is suggested by the ASA. Specific exceptions to this include patients eligible for, or treated with, thrombolytics and those with other indications for tight BP control, such as hypertensive encephalopathy or other acute end-organ damage. Labetalol and nicardipine are specifically recommended by the ASA, whereas labetalol and urapidil (the latter not available in the United States) are recommended by the ESO. However, the guidelines indicate little evidence supporting the use of specific agents. Nitroprusside is now reserved for intractable severe hypertension (ASA). Resumption of outpatient BP medications after 24 h is generally considered safe except in specific situations when this seems contraindicated. Volume expanders are recommended for hypotensive stroke patients. Pressor therapy is not recommended routinely outside of trials but is presented as an option in the ASA guidelines.

Brain edema following cerebral infarction is an important and potentially treatable cause of neurologic deterioration and mortality seen primarily in large hemispheric and cerebellar infarctions. Hypertonic solutions are used widely because of their diuretic action and ability to decrease intracranial pressure (ICP) transiently following cerebral infarction but have yet to be shown to improve outcome.^12-14 General measures followed include treatment of fever, pain, and agitation, and keeping the head elevated to 30°. Medication regimens used most commonly are Mannitol and hypertonic saline in concentrations of 3%, 10%, and 23.4%.

Ventriculostomy and suboccipital craniotomy for patients with large cerebellar infarction and brainstem compression are of unquestioned benefit and have not been the subject of any trials. Surgery for supratentorial infarction, however, has recently been the object of close scrutiny and study.¹⁵ Three similar European trials of hemicraniectomy were completed and they support the belief that this intervention can both save lives and improve functional outcome.¹⁶ These trials included patients < 60 years of age with severe neurologic deficits as defined by the National Institutes of Health Stroke Scale, and radiographic changes of a large hemispheric infarction. Anecdotal reports and many editorials have highlighted the difficulties in determining what a patient will consider to be an “acceptable” outcome and the importance social support and access to rehabilitation services play in maximizing recovery.¹⁷ Therapeutic hypothermia has been the subject of a few small trials but remains an unproven measure in this situation (see “Temperature Control/Therapeutic Hypothermia” section). The Decompressive Surgery for the Treatment of Malignant Infarction of the Middle Cerebral Artery 2 (DESTINY 2) trial recently began in Germany and includes patients > 60 years of age.¹⁸ The protocol is otherwise very similar to that used in the recently completed European trials.

Close monitoring of patients at risk of severe cerebral edema at a facility with expertise is recommended in the ASA and NICE guidelines. Ventricular drainage and decompression for cerebellar infarctions are uniformly recommended by all organizations. Osmotherapy is reviewed and recommended with little evidence by the more detailed ASA and ESO guidelines. Decompression for large cerebral infarctions is recommended by all organizations. The ESO and NICE specifically recommend that this be done early (ie, < 48 h) and be limited to those < 60 years of age. NICE defines the radiographic criteria for decompression in detail (> 50% MCA territory on CT scan, or infarct volume > 143 cm³ on diffusion weighted imaging). The use of corticosteroids is discouraged by all the guidelines.

Elevated temperature is common after ischemic stroke and has been associated with poorer outcome. The prevention of hyperthermia following stroke has been the subject of recent trials. Scheduled administration of acetaminophen and ibuprofen in normothermic and hyperthermic patients following acute ischemic and hemorrhagic stroke had no effect on outcome in small trials and one recent large trial.^19,20 Despite the many potential neuroprotective effects of hypothermia seen in animal stroke models and the benefit of hypothermia seen in humans following cardiac arrest, there is still no evidence demonstrating improved outcomes in stroke patients.²¹

Surface and endovascular cooling systems with temperature control systems have been used in stroke studies.^21-24 Goal temperatures usually range from 33° to 35°C. IV infusion of ice-cold saline (1 L over 1 h and repeated again over 2 h if needed) has been shown to induce hypothermia more rapidly.²⁵ Pharmacologic agents, including meperidine, buspirone, and magnesium, and concurrent skin warming inhibit shivering, allow patients to tolerate treatment with less sedation, and have been used in more recent reports.²⁴ More controlled rewarming techniques seem to prevent rebound brain edema and have been incorporated into the more recent trials.²⁴ Therapeutic hypothermia has been compared with hemicraniectomy alone,²⁶ in combination with hemicraniectomy,²⁷ and following intravenous thrombolysis up to 6 h.²³ Trials using more mild hypothermia (eg, 35°C) and slower rewarming have found lower complication rates.²⁷ A combination of mild hypothermia and caffeinol started very soon after IV tissue plasminogen activator was found to be safe in one recent small study.²⁸ The Cerebral Hypothermia in Ischemic Lesion (CHIL) trial is a randomized trial initiated recently that will evaluate the effect on penumbral salvage of endovascular cooling (Australia) and local head cooling (China) (http://www.anzctr.org.au/trial_view.aspx?ID=308341).

Treating fever and identifying and treating the underlying cause is recommended by all guidelines. The ESO recommends treatment with fanning along with acetaminophen for a temperature greater than 37.5°C. The guidelines found insufficient evidence to recommend therapeutic hypothermia.

Acute hyperglycemia is common after ischemic stroke and has been associated with poorer outcomes. Acute tight control after stroke has been advocated and supported by earlier trials in critically ill patients but has been questioned more recently.²⁹ The Normoglycemia in Intensive Care Evaluation-Survival Using Glucose Algorithm Regulation (NICE-SUGAR) study questioned the benefit of intensive insulin therapy and extremely tight glucose control (goal 81-108 mg/dL) in critically ill patients and supports a more conservative goal of 120 to 180 mg/dL.³⁰ Concerns have been raised that insulin infusions containing potassium may lower BP. Tight glycemic control may also lead to low cerebral extracellular glucose as measured by microdialysis and may be detrimental to patients with severe brain injury, although this was seen in a study of mostly hemorrhagic stroke and traumatic brain injury.³¹

The UK Glucose Insulin Stroke Trial (GIST-UK) trial published in 2007 found no benefit from a continuous glucose-potassium-insulin infusion with a goal range of between 72 and 126 mg/dL in patients with ischemic and hemorrhagic stroke.³² Of note, this trial was stopped prematurely because of low enrollment and it excluded patients with admission glucose measurements > 300 mg/dL. Further, the study infusion was given for only 24 h after admission and achieved only a modest lowering of blood glucose (< 20 mg/dL between groups). The Glucose Regulation in Acute Stroke Patients (GRASP) trial compared different intensities of acute glycemic control using an insulin infusion and found that patients with more severe strokes and higher admission glucose levels might benefit from tight control.³³ A recently published MRI-based trial showed that insulin infusion titrated to a glucose concentration of 72 to 126 mg/dL attenuated an increase in brain lactate but actually increased the growth of infarction in the subset of patients with persistent arterial occlusion.³⁴ One proposed explanation is that the hypoglycemia or hypotension associated with insulin infusion may have been detrimental in patients with borderline perfusion reserve when compared with a regimen of intravenous saline.

Treatment with insulin is recommended when blood glucose levels are “elevated” (CSS) and for values “possibly greater than 140 to 185 mg/dL” (ASA), 180 mg/dL (ESO), and 200 mg/dL (NICE). These guidelines do not distinguish between patients with and without diabetes mellitus.

The broad category of intracranial hemorrhage includes epidural, subdural, subarachnoid, and ICHs. ICH shares epidemiologic, pathophysiologic, and diagnostic features with ischemic stroke and is reviewed here.

Elevated BP after ICH is common and may be a marker of baseline hypertension; it may also reflect associated pain or agitation, or serve as a compensatory response to the injury or associated mass effect. Permissive hypertension raises concerns for extension of bleeding, whereas BP reduction raises concerns for neurologic deterioration from perilesional hypoperfusion. However, both have been called into question.^2,35 More recently, the Intensive BP Reduction in Acute Cerebral Hemorrhage (INTERACT) trial has shown that a systolic BP goal of 140 mm Hg compared with 180 mm Hg when begun within 6 h of symptoms can attenuate hematoma expansion. Outcome data are forthcoming.³⁶ Likewise, the Antihypertensive Treatment of Acute Cerebral Hemorrhage (ATACH-II) trial is enrolling hypertensive ICH patients who can be treated within 3 h of symptom onset with nicardipine with a goal of systolic BP of 140 mm Hg vs the standard goal (systolic BP < 180 mm Hg) to look at 3-month outcome.

ASA² recommendations for optimal management based on current data suggest a treatment threshold of systolic BP of 180 mm Hg if suspicion of elevated ICP, with a target BP goal of 160/90 mm Hg. Acute lowering of the systolic BP to 140 mm Hg is probably safe. NICE recommends a treatment threshold in ICH of 200 mm Hg. The ESO³⁷ recommends a treatment threshold of 180/105 mm Hg or reducing mean arterial pressure below 130 mm Hg, with a target goal of 160/100 mm Hg.

Hemostatic therapy is recommended to correct coagulopathy with protamine sulfate for heparin-associated ICH, and IV vitamin K along with clotting factor replacement (fresh-frozen plasma [FFP] or prothrombin complex concentrate [PCC]) for warfarin-associated ICH. International normalized ratio reversal may be quicker with PCC than with FFP,³⁸ with lower infusion volume, but improved outcomes with PCC have not been reported. Recombinant factor VIIa can quickly lower the international normalized ratio but does not replenish other vitamin-K dependent clotting factors. The use of hemostatic therapy in patients with spontaneous ICH is more controversial because the pivotal phase 3 trial³⁹ for recombinant factor VIIa showed attenuation of hematoma growth but more frequent thromboembolic events with active treatment but no functional outcome or survival benefit. ICH after thrombolytic therapy is managed with six to eight units of platelets and cryoprecipitate. Antithrombotics, including antiplatelets and anticoagulants, are discontinued after acute ICH. The Platelet Transfusion in Cerebral Hemorrhage (PATCH) trial will look at transfusion effect on hematoma expansion and 6-month outcome in ICH patients on concurrent antiplatelet agents.

Acknowledging the relative paucity of randomized data, recommendations from the ASA, ESO, and NICE suggest stopping anticoagulants and using PCC or FFP with IV vitamin K in cases of anticoagulation-associated ICH.

Management of ICP by ASA guidelines is underscored by the recommendation to keep cerebral perfusion pressure (CPP) > 60 to 80 mm Hg. An inherent limitation of this recommendation and in targeted treatments is the difficulty of gauging CPP (CPP = MAP − ICP). Many patients with ICH do not have ICP monitors. Even in those who do, accurate measurements are constrained by limitations of specific device type, location of monitor, and length of monitor usage. Supportive measures to reduce ICP, including positioning the head midline to avoid impaired jugular venous drainage, setting the head of the bed at 30°, and treating pain, are recommended.

The use of hyperosmolar agents, most commonly Mannitol and hypertonic saline, have not been shown to improve outcome after ICH in randomized trials.⁴⁰ It has been suggested that they be used as osmotic diuretics capable of reducing edema. The additional effect of improving blood viscosity to induce cerebral vasoconstriction and further decrease ICP is also touted.⁴¹ However, because these agents may cross the damaged blood-brain barrier into the hemorrhage site to worsen mass effect or may induce diuresis-associated azotemia, their benefit is questionable. The use of hyperventilation has fallen out of favor. Decreasing Paco₂ has been shown to decrease ICP by decreasing cerebral blood flow through reflex vasoconstriction. However, decreased cerebral blood flow can reduce CPP, whereas the goal of ICP reduction is to preserve CPP. Furthermore, ionic and pH corrective responses to induced hypocarbia occur within hours and may cause rebound increased ICP.

The ASA and ESO recommend ICP monitoring when aggressive measures are needed based on clinical features. ASA guidelines suggest a target CPP of 50 to 70 mm Hg. The ESO recommends medical treatment of ICP when clinical deterioration is thought to be due to edema.

The role of surgery in the treatment of ICH is evolving. Because of the risk of clinical deterioration associated with damage to structures adjacent to the hemorrhage site, surgical clot evacuation or decompressive craniectomy are appealing. In cerebellar hemorrhages where mass effect on the brainstem and herniation are catastrophic risks, surgical therapy has been proven superior to medical.⁴² The International Surgical Trial in Intracerebral Hemorrhage (STICH) trial⁴³ randomized patients with supratentorial ICH > 2 cm in diameter and Glasgow Coma Scale (GCS) > 5 presenting within 72 h to medical therapy or surgery. There was a trend toward a 6-month outcome benefit in the surgical group that did not reach significance. In the subgroup analyses, there was a suggestion of preferential surgical benefit for patients presenting with GCS > 9 and those with lobar hemorrhages, particularly within 1 cm of the cortical surface. Timing of surgery has been examined in early (3-24 h) surgery studies^44,45 and later surgery studies (eg, the STICH trial), which show that the benefit of early surgery is undercut by the risk of rebleeding, and that late surgery (eg, > 96 h) may expose patients to surgical risk without conferring benefit. The STICH II trial to better define a subset of patients most amenable to surgery is ongoing.

The CCS recommends that patients with acute cerebellar hemorrhage have neurosurgical consultation for consideration of hemorrhage evacuation. The ASA and ESO recommend surgery for patients with brainstem compression, clinical deterioration, or hydrocephalus. According to NICE guidelines, surgery is unequivocally recommended only for patients with hydrocephalus. The NICE recommends initial medical management for small, deep, lobar, and cerebellar hemorrhages. The ASA and ESO do not recommend routine surgical treatment of supratentorial ICH, with the exception of those hemorrhages within 1 cm of the cortical surface where treatment may be considered. The ESO suggests that stereotactic aspiration of deep ICH with mass effect be considered.

The armamentarium of treatment of ICH remains small. As part of supportive therapy, prophylaxis against common known complications is often used. Seizures occur after ICH in 8%, with the lobar location being specifically predictive.⁴⁶ Seizures after ICH are associated with worse outcome,⁴⁷ but prophylactic treatment has not been shown to convincingly improve outcome.⁴⁸ Antiepileptic medication use may place patients at the risk of prolonged hospital stays because of drug fever or impaired alertness.

The use of corticosteroids is not recommended for the treatment of ICH. An early trial showed no mortality benefit and was terminated early because of excess infectious and hyperglycemic complications.⁴⁹

DVT is a common complication after stroke, particularly in hemiplegic patients. Prophylaxis against DVT formation includes intermittent pneumatic compression devices. For added efficacy, beginning 3 days after cessation of bleeding, the use of low-dose, low-molecular-weight or unfractionated heparin is recommended in hemiplegic patients. Treatment of DVT in ICH with anticoagulants is contraindicated, so placement of an inferior vena caval filter to prevent pulmonary embolism is recommended. Because of the paucity of long-term data, no specific recommendations are given regarding the starting of anticoagulation subacutely.

Current ESO recommendations suggest 7 days of prophylactic antiepileptic use after lobar ICH. The ASA recommends against prophylactic treatment and suggests treating clinical seizures. For DVT prophylaxis in patients with ICH, pneumatic compression devices are highly recommended together with heparin after 24 h by the ESO, and beginning 1 to 4 days after cessation of bleeding by the ASA.

Because ICH carries a high risk of morbidity and mortality,² providers are frequently called on to give prognostic information to family members. Accurate predictions are limited by insufficient data on outcome and by the protean features in an individual patient with ICH that may affect outcome. However, some recent studies shed light on these variables. The ICH score calculated at initial presentation uses age, GCS score, intracerebral blood volume and location, and presence of intraventricular blood to predict 30-day mortality.⁵⁰ Perhaps of more interest to family members is the functional outcome, addressed more recently by the Functional Outcome After Intracerebal Hemorrhage score, which assessed the likelihood of 90-day functional independence based on age, premorbid cognitive impairment, GCS score, and blood volume and location.⁵¹ Nonetheless, the available outcome data may be contaminated by patients with preexisting unrelated do-not-resuscitate orders and by pessimism, which can lead to early withdrawal of care.⁵²

The ASA recommends consideration of aggressive care with postponement of new do-not-resuscitate orders for the first 24 h after diagnosis.

A persistently thorny issue in ICH care involves decisions about the reinstitution of antithrombotic agents. Because prospective trials are lacking, data are gleaned from expert opinion and small case series. Overall, because of the risk of recurrent hemorrhage in patients with prior ICH, decisions must balance the risks of reinstituting against withholding these agents.

The risk of recurrent hemorrhage may depend on the cause, with ICH due to amyloid angiopathy carrying a higher risk of recurrence than deep hemorrhage.⁵³ By contrast patients with nonvalvular atrial fibrillation or mechanical valves have a high risk of ischemic stroke.² In patients with suspected amyloid-related lobar ICH but an indication for anticoagulation such as atrial fibrillation, the risk of subsequent ischemic and hemorrhagic stroke is often managed with antiplatelets.

The ASA suggests avoiding anticoagulation in lobar ICH patients with nonvalvular atrial fibrillation but indicates that resumption may be considered in patients with nonlobar ICH. In patients with a very high risk of thromboembolism (eg, mechanical heart valve), ASA guidelines suggest resuming anticoagulation 7 to 10 days after ICH. The ESO suggests forestalling resumption in patients who need treatment for 10 to 14 days.

Ischemic and hemorrhagic strokes represent a group of diseases associated with a significant risk of morbidity and mortality. Adequate treatments are still lacking and recommendations from American, Canadian, UK, and European groups have points of divergence, reflecting persistent ambiguity in the optimal application of current treatments. It is hoped that ongoing trials will provide further direction and guidance to improve outcomes.

Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.