Despite recent advances in anesthesia techniques and monitoring measures, intraoperative and postoperative neurologic events remain the most devastating complications and continue to concern anesthesia providers. Even without any significant intraoperative events, there is a considerable risk for cerebral ischemia in specific surgical populations, such as patients undergoing cardiac surgeries and vascular surgeries.
The neurologic sequelae range from frank stroke to cognitive dysfunction. The incidence of perioperative stroke is reported to be from 1.6% to 5.2% in coronary artery bypass grafting (CABG) and from 0.25% to 7% in carotid endarterectomy (CEA), whereas the incidence of cognitive dysfunction ranges from 24% to 57% at 6 months after cardiac surgery.
There is a substantial amount of interest in research to identify neuroprotective strategies; however, most of the clinical trials have resulted in disappointment, and there are no formal guidelines based on the strongest clinical evidence. This is thought to be because of the complexity of the mechanisms in cerebral ischemia.
Most anesthesia providers strongly agree that maintaining adequate cerebral oxygenation and perfusion pressure is the most effective and important strategy in neuroprotection. Historical clinical evidence also advocates avoiding deleterious factors in the event of ongoing cerebral ischemia or in higher risk populations.
Neuroprotective strategies are classified into two concepts: passive, which refers to the avoidance of deleterious factors, and active, which refers to the application of beneficial interventions. Hans and Bonhomme proposed categorizing the neuroprotection measures into the following areas: physiology, anesthetics, nonanesthetic pharmacologic agents, and preconditioning. Along with these strategies, the role of monitoring in specific surgical populations will also be discussed.
Physiology: avoidance of hyperthermia, hyperglycemia, cerebral hypoxia, and hypoperfusion
Anesthetics: the use of certain anesthetics that are potentially neuroprotective because of reduction of energy requirements
Pharmacology: the use of potentially neuroprotective agents that can block the pathways of neuronal cell death. This may include N -methyl-d-aspartate (NMDA) receptor antagonists, excitatory amino acid (EAA) receptor antagonists, and erythropoietin (EPO)
Preconditioning: the use of physiologic or pharmacologic alterations that could mimic preconditioning for high-risk populations
Monitoring: the use of epiaortic echocardiographic scanning to manage severe atherosclerotic disease and near-infrared reflectance spectroscopy (NIRS) for assessment of bifrontal regional cortical oxygen saturation (rSO 2 ) in cardiac surgery
A number of studies have evaluated neuroprotective strategies and outcomes in the areas of physiology, anesthetics, pharmacology, and monitoring ( Table 63-1 ).
|Study (Year)||No. of Subjects||Patient Population||Study Design||Intervention||Control||Outcomes|
|Bernard (2002)||273||Comatose survivors of out-of-hospital cardiac arrest||Prospective randomized||Mild hypothermia||Normothermia||Favorable neurologic outcome|
|Kammersgaard (2002)||390||Acute stroke||Observational||Hypothermia (≤37° C)||Hyperthermia (>37° C)||Low admission temperature is an independent predictor of good short-term outcome|
|Grigore (2002)||165||CABG with CPB||Prospective not randomized||Slower rate of rewarming||Conventional rewarming||Better cognitive performance at 6 wk|
|Gentile (2006)||960||Acute ischemic stroke||Retrospective||Normalization of BG (<130 mg/dL) during first 48 hr||Hyperglycemia (BG ≥130 mg/dL)||Associated with a 4.6-fold decrease in mortality risk|
|Vicek (2003)||372||Acute ischemic stroke||Retrospective||Lowering DBP more than 25% from admission value||Maintained DBP||Associated with a 3.8-fold increased adjusted odds for poor neurologic outcome on day 5|
|Ahmed (2003)||201||Acute ischemic stroke||Retrospective||Lowering DBP with nimodipine||Maintained DBP||Worsened the neurologic outcome in nontotal anterior circulation infarct|
|Gold (1995)||251||CABG with CPB||Prospective randomized||High MABP (80-100 mm Hg) during CPB||MABP 50-60 mm Hg during CPB||Fewer myocardial and neurologic complications|
|Michenfelder (1987)||2223||Carotid endarterectomy||Retrospective chart review||Isoflurane||Enflurane, halothane||Lower critical CBF (10 mL/100 g/min) versus 15 in enflurane and 20 in halothane; lower incidence of EEG ischemic change (18% versus 26% in enflurane and 25% in halothane)|
|Messick (1987)||6||Carotid endarterectomy||Prospective single-arm||Isoflurane||Halothane||Lower critical CBF (less than 10 mL/100 g/min) versus 18-20 in halothane|
|Kanbak (2004)||20||CABG with CPB||Prospective randomized||Isoflurane||Propofol||Alleviated increase of S-100 beta protein|
|Hoffman (1998)||12||Middle cerebral artery occlusion||Prospective randomized||Desflurane||Etomidate||Increased brain tissue PO 2 and attenuated acidotic change|
|Mitchell (1999)||65||Left heart valve operation||Prospective randomized||Intravenous lidocaine||Placebo||Fewer incidences of decreased neuropsychological performance|
|Wang (2002)||118||CABG with CPB||Prospective randomized||Intravenous lidocaine||Normal saline||Decreased the occurrence of early postoperative cognitive dysfunction|
|Arrowsmith (1998)||171||CABG with CPB||Prospective randomized||Remacemide||Placebo||Overall postoperative change (reflecting learning ability in addition to reduced deficits) was favorable in treated group|
|Mathew (2004)||914||CABG with CPB||Prospective randomized||Pexelizumab||Placebo||Decreased visuospatial function impairment but not overall cognitive dysfunction|
|Ehrenreich (2002)||40||Acute ischemic stroke||Prospective randomized||Recombinant human erythropoietin||Saline||Improvement in clinical outcome at 1 mo|
|Bhudia (2006)||350||Cardiac surgery with CPB||Prospective randomized||Magnesium sulfate||No intervention||Improved short-term neurologic function|
|Pandharipande (2007)||106||Mechanically ventilated in ICU||Prospective |
|Dexmedetomidine||Lorazepam||More days alive without delirium or coma; more time at the targeted level of sedation|
|Royse (2000)||46||CABG with CPB||Prospective not randomized||Epiaortic echocardiography and exclusive Y graft||Digital palpation and aorta-coronary operations||Less incidence of late neuropsychological dysfunction|
|Murkin (2007)||200||CABG with CPB||Prospective randomized||Cerebral regional oxygen saturation monitoring and treatment protocol||No intervention||Avoids profound cerebral desaturation and is associated with fewer incidences of major organ dysfunction|
To ensure adequate cerebral oxygenation and cerebral perfusion, measures that reduce the cerebral metabolic rate (CMR) are known to be beneficial. Hypothermia has been proposed to offer neuroprotective effects for several decades, but in a recent larger clinical trial in patients with acute traumatic injury, hypothermia failed to improve neurologic outcomes. The effect of mild hypothermia (32° C to 35° C) on CMR is negligible, and only deep hypothermia (18° C to 22° C), which is used in specific types of cardiac surgeries, is neuroprotective. However, two prospective randomized trials in comatose survivors of out-of-hospital cardiac arrest demonstrated better neurologic outcomes in the patients treated with mild hypothermia. It was also reported that intraischemic or delayed hyperthermia worsens outcome. Grigore and colleagues reported that a slower rewarming rate with lower peak cerebral temperatures results in significantly better cognitive performance after cardiac surgery with hypothermic cardiopulmonary bypass (CPB).
Tight glucose control is associated with reduced mortality and morbidity rates in critically ill patients and postcardiac surgery patients. Persistent hyperglycemia after a stroke has been shown to increase the size of ischemic brain injury and worsen clinical outcomes. A retrospective study demonstrated decreased mortality rates when blood glucose levels were normalized after acute ischemic stroke. One should keep in mind that tight glucose control (80 to 110 mg/dL) is associated with a higher incidence of hypoglycemia. In a retrospective study of 172 patients with subarachnoid hemorrhage (SAH), lower nadir glucose levels were associated with progressively worse outcomes. The investigators also reported that patients with symptomatic vasospasm had lower nadir glucose levels than those without vasospasm. Nonetheless, on the basis of these clinical data, hyperglycemia should be avoided perioperatively.
The use of corticosteroids is not advocated in ischemic or traumatic brain injury because no strong evidence supports the benefit from treatment with corticosteroids. In addition, hyperglycemia induced by administration of corticosteroids is potentially harmful.
Maintaining baseline blood pressure is an essential measure that ensures vital organ perfusion, including that to the brain. Cerebral perfusion pressure is calculated by subtracting intracranial pressure from mean arterial blood pressure (MABP). Two clinical studies demonstrated that lowering diastolic blood pressure (DBP) in the early phase of ischemic stroke worsened the neurologic outcome.
Another retrospective study in patients who sustained sudden cardiac arrest demonstrated that good neurologic recovery was independently and directly related to MABP during the first 2 hours after return of spontaneous circulation. Gold and colleagues found fewer myocardial and neurologic complications after CABG surgery when targeted MABP during CPB was between 80 and 100 mm Hg rather than 50 and 60 mm Hg. In this study, the incidence of cognitive dysfunction at 6 months after surgery was low, and no relation was found between arterial pressure and cognitive outcome. However, maintaining a “higher” MABP target is considered to be acceptable, safe, and useful for patients at high risk of neurologic complications.
Accumulating experimental evidence confirms the neuroprotective effect of inhalational anesthetics in both focal and global ischemia. The mechanism involves inhibition of excitatory neurotransmission and potentiation of inhibitory receptors, resulting in suppression of energy requirements. Preconditioning from inhalational agents is proposed as an additional mechanism of neuroprotection. The tolerance against ischemia is increased in the future event by activation of adenosine triphosphate (ATP)–dependent potassium channels and adenosine A1 receptors. In contrast to the multitude of experimental studies, clinical evidence on the neuroprotective effect of inhalational anesthetics has been scant. Hoffman and colleagues reported that desflurane, in comparison with etomidate, increased brain tissue oxygen pressure and reduced acidosis in patients subjected to temporary middle cerebral artery occlusion. Another prospective study in patients undergoing CEA determined that critical regional cerebral blood flow, which is the flow rate when electroencephalographic signs of ischemia are evident, during isoflurane anesthesia was much lower than during halothane or enflurane anesthesia. In a retrospective study, the incidence of ischemic change was lower with isoflurane anesthesia when compared with halothane or enflurane anesthesia, and no difference was found in neurologic outcomes, despite the fact that the isoflurane group had a higher risk of an adverse outcome. Sufficiently powered, prospective, randomized controlled studies evaluating neurologic outcomes using more appropriate endpoints such as long-term neurocognitive function are still needed. However, the use of volatile anesthetics can be considered as a part of an anesthetic plan when the risk of neuronal injury is anticipated.
Xenon is the most potent inert gas to be used as an anesthetic agent and has been shown to have potential neuroprotective properties because it can inhibit the NMDA receptor. A large number of recent studies report that xenon affords neuroprotection in a variety of animal models, including focal cerebral ischemia, neonatal asphyxia, neurocognitive deficit after CPB, and traumatic brain injury. Xenon has also been shown to be neuroprotective in preconditioning paradigms. Despite promising data from animal studies, very few clinical trials have addressed xenon neuroprotection. The use of xenon is limited because of the significantly high cost of production and complexity of administration that requires closed circuits. Small clinical trials examining the efficacy of xenon in decreasing postoperative cognitive dysfunction after noncardiac surgery failed to show an advantage compared with propofol or desflurane.
Lidocaine was shown to have a neuroprotective effect in an in vivo study due to the reduction of energy consumption by delay of ischemia-induced membrane depolarization and also by alleviation of apoptosis. In a small clinical trial, lidocaine infusion at an antiarrhythmic dose demonstrated improved long-term neuropsychologic performance in 65 patients with left heart valve procedures. More recently, Wang and colleagues reported that intraoperative administration of lidocaine decreased the occurrence of early postoperative cognitive dysfunction in patients who had undergone CABG surgery. Neither of these studies had enough power to conclude that lidocaine infusion should be used routinely as a neuroprotective agent. Larger clinical trials determining the optimal dosing regimen and long-term results are still needed.
The neuroprotective ability of barbiturates has long been postulated, and their mechanism of action is thought to originate from reduction of the CMR and blockade of glutamate receptors. However, mixed results have been published, and their clinical perioperative efficacy remains controversial. One of the problems with using barbiturates is their prolonged duration of action, thus causing delayed emergence. Because the volatile anesthetics have been shown to have similar effects as barbiturates but with shorter emergence time, the popularity of barbiturates has declined.
Propofol and ketamine have also been postulated to be neuroprotective agents; however, both drugs failed to improve long-term neurocognitive performance. A recent randomized two-arm prospective study compared propofol with isoflurane using plasma S100beta as a brain injury marker; neurologic outcomes at 6 months postoperatively in the patients with traumatic brain injury demonstrated that propofol attenuated the increase of S100beta and improved neurologic outcomes; however, the latter did not reach statistical significance because of small sample size.
A few clinical trials with encouraging results deserve mention. Remacemide, an NMDA receptor antagonist, has been shown to improve some measures of postoperative psychometric performance in cardiac surgery patients.
Mathew and colleagues reported that pexelizumab, a humanized monoclonal antibody against the C5 complement component, led to less visuospatial impairment up to 1 month after CABG surgery but failed to decrease the overall incidence of cognitive dysfunction.
EPO has been used for the treatment of anemia and is known to be safe. It blocks apoptosis, blocks inflammation, and induces vasculogenesis and neurotrophic factors. In a clinical trial, high-dose intravenous EPO was shown to improve clinical outcomes at 1 month in patients with acute ischemic stroke.
Magnesium, because of its direct vasodilatory effects and antagonism of the NMDA receptor, has been investigated as a potential treatment for patients with SAH. It also reduces the ischemia-related rise in intracellular Ca 2+ , thereby preventing cell death. Ma and colleagues reported a meta-analysis of six prospective studies evaluating the beneficial effect of serum magnesium concentration–targeted therapy on outcomes after SAH. In this analysis, magnesium sulfate was found to reduce the relative risk of poor outcomes as well as the risk of delayed cerebral ischemia. However, the use of magnesium was associated with a higher incidence of hypotension, arrhythmia, renal failure, respiratory arrest, myocardial infarction, and phlebitis. A single randomized placebo-controlled trial enrolling 350 patients undergoing CABG demonstrated that magnesium administration improved short-term postoperative neurologic function but not long-term outcomes.
Dexmedetomidine is a potent and highly selective alpha-2 adrenoceptor agonist with sedative, amnestic, and analgesic properties. It has been used in many clinical settings, including as an anesthetic adjunct and as a sedative for awake intubation or for critically ill patients. There is increasing evidence of its organ-protecting effects against ischemic and hypoxic injury, including cardioprotection, neuroprotection, and renoprotection. It has been shown to attenuate hypoxic–ischemic brain injury in developing brains and to improve functional neurologic outcomes after brain injury. No randomized controlled trials have directly evaluated the neuroprotective effects of dexmedetomidine. The Maximizing Efficacy of Targeted Sedation and Reducing Neurological Dysfunction (MENDS) study demonstrated that the use of dexmedetomidine infusion reduced the incidence of delirium or coma and postoperative mortality.
Numerous pharmacologic agents have been investigated for their potential ability to limit neuronal injury. Despite promising data from laboratory work, all had disappointing clinical results. This was mainly due to the complexity of the mechanisms of neuronal injury and the difficulty in controlling physiologic factors. The combination of multiple strategies, including the use of compounds targeting different pathways and the control of physiologic variables, may afford the most meaningful results in perioperative neuroprotection.
Preconditioning is a novel concept of neuroprotection in which a prior exposure to minor insults will induce an increased tolerance to more serious injury. The mechanism of preconditioning is activation of ATP-dependent potassium channels and adenosine A 1 receptors. Other than a history of transient ischemic attack before acute stroke promoting ischemic tolerance in the human brain, many factors and various drugs can mimic preconditioning, such as hyperoxia, hypothermia, electroconvulsive shock, volatile anesthetics, the potassium channel opener diazoxide, and erythromycin.
In some specific surgical procedures, the use of specific monitoring measures may have an impact on neurologic outcome. The change of surgical approach led by intraoperative epiaortic echocardiography has been shown to lower the incidence of late neurologic dysfunction in a large observational study and also in a smaller case-control study. This strategy is rated as class IIb (acceptable, safe, and useful) for patients undergoing CABG surgery at high risk of neurologic injury in an evidence-based rating by Hogue and colleagues.
The use of NIRS for assessment of rSO 2 has demonstrated a correlation between coronary artery bypass patients having low rSO 2 values and cognitive dysfunction, prolonged hospital length of stay, and cerebrovascular accident. A recent randomized controlled study by Murkin and colleagues demonstrated that the treatment of declining rSO 2 prevented prolonged desaturations and was associated with a shorter intensive care unit length of stay and a significantly reduced incidence of perioperative major organ morbidity and mortality. This result may have been a reflection of the good clinical practice of optimizing organ perfusion instead of the direct effect of rSO 2 monitoring. However, the monitoring would allow early detection and rapid improvement of end-organ compromise.