Cardiovascular disease is the world's leading cause of morbidity and death. The incidence of cardiac arrest is rising despite use of proven and effective medical and surgical therapies. In the US alone, over 700,000 cardiac deaths will occur this year and half will be related to cardiac sudden death. Approximately 86% of these victims will experience cardiopulmonary resuscitation (CPR). Return of spontaneous circulation (ROSC) will occur in 17-49%, but most victims will expire at the scene.' Despite these dismal statistics, more victims are admitted to the hospital for advanced treatment because of improvements in the quality of the "chain of survival", bystander CPR, establishment of effective police/firefighter responder programs and public availability of automated external defibrillators.2 Of those patients who are admitted to the hospital, half will die within hours of cardiac causes or other comorbidities. If the victim survives, morbidity and mortality are more likely to be the result of neurologic rather than cardiac injury.1
In those patients successfully resuscitated and admitted to the hospital, almost 80% will be in a coma for varying lengths of time, making cardiac arrest the third leading cause of coma. Approximately 40% will enter a persistent vegetative state and 80% die within one year.3 The overall survival to hospital discharge is 20%.2 If a patient does survive to discharge without significant neurologic injury, meaning that s/he can participate in rehab or go home, that patient can expect a fair to good quality of life. Studies have shown no significant difference in anxiety, depression, vitality, and general wellbeing between patients and controls. There are, however, significant differences in energy levels, emotional reactions and sleep patterns. Additionally, 11-28% of patients discharged will be cognitively impaired with problems in memory, attention and executive function.4 These consequences of cardiac arrest are the result of neuronal injury.
Neuronal injury is the result of ischemia and hypoxia which leads to necrotic and apoptotic cell death. Animal models and limited human studies of neonatal hypoxia and cerebral hypoperfusion during cardiopulmonary bypass have shown that brain injury occurs during the event and evolves even after ROSC. The primary sources of cerebral injury are perfusion failure, reoxygenation injury and blood stasis leading to the formation of microthrombi. At the time of arrest, there is perfusion failure. A stop of cerebral circulation depletes the neuronal oxygen stores within 20 seconds and consciousness is lost.1 Within 5 minutes of complete cerebral ischemia and hypoxia, brain glucose and ATP stores are depleted. Dysfunction of neuronal membrane pumps leads to membrane depolarization, influx of calcium, and activation of lytic enzymes. The loss of cerebral autoregulation leads to prolonged global and multifocal hypoperfusion and to transient global hyperemia due to vasoparalysis.5
With the ROSC there are reoxygenation-induced chemical reactions that cause excitotoxicity, mitochondrial damage, and DNA fragmentation that results in primary necrosis or programmed cell death (apoptosis).1 Cerebral microcirculation is compromised because of increased blood viscosity, endothelial cell swelling, and coagulation activation with subsequent mircrocirculatory fibrin deposition that contributes to prolonged ischemia.6
The extent of neuronal injury depends on the duration of ischemia/hypoxia and on premorbid function. Neuronal injury manifests clinically as amnesia, focal or multifocal motor and sensory deficits, spinal cord compromise, seizures, myoclonus, persistent vegetative state or brain death. Patients with a short duration of ischemia/hypoxia, usually less than five minutes, are unlikely to have permanent neurologic deficits and will usually return to their previous functional level.7 However, patients who have longer events may have seizures, watershed infarcts, or spinal cord infarction. A patient who suffers prolonged, severe, global hypoxic/ischemic insult that results in widespread death of neurons and necrosis of the cerebral cortex will have myoclonus, be in a persistent vegetative state or possibly be brain dead.7
Given the range of neurological outcomes, prognosis has a significant effect on the choice of medical treatment and on the family. If a patient wakes up within 24 hours after the event, their neurological prognosis is good. For those who remain comatose, early prognosis can be difficult. Many studies have attempted to validate a reliable method for the correct prediction of outcome. Earlier studies of prognosis in coma after cardiac arrest focused on the initial examination and whether the patient would wake up. One even formulated a mathematical equation ( motor response + 3Xpupillary light response + spontaneous eye movements + blood glucose on admission) to help with early prognosis.8 The Glasgow Coma Scale (GCS), which is used to evaluate trauma patients, has been easily adapted to aid in early prognosis in nontraumatic coma patients. Unfortunately, an indeterminate GCS (5-9) means an indeterminate prognosis. The GCS does not have a high specificity for determining if the patient will wake up until 3-5 days after admission.7 Functional recovery was not taken into account until 1985 when Levy and Caronna, et al. followed 210 patients with nontraumatic coma, 150 following cardiac arrest. They performed clinical examinations at days 1,3,7 and 14 and assessed functional recovery at 1,3,6 and 12 months. They observed that the neurologic signs most reliably linked to outcome within the first 24 hours were pupillary light reflex, spontaneous eye movements, and corneal reflex. The presence or absence of a motor response did not make a difference in prognosis until 72 hours after the event.10 A meta-analysis showed that the absence of the corneal reflex, pupillary response, withdrawal response to pain at 24 hours and no motor response at 72 hours was consistent with a poor prognosis 97% (95% confidence interval (CI) 87-100%) of the time.11 The variables not associated with outcome were verbal response, which is one of the cornerstones of the GCS, age, sex, site of onset (meaning whether the event occurred in the home or in the hospital ICU), cause of coma and paroxysmal activity.10 No clinical findings strongly predicted good clinical outcome.11 Even though simple physical examination techniques strongly predicted death or poor outcome in coma patients after cardiac arrest, the clinical exam was useful in prognosis 24 hours after presentation and not before-hand. The reason the clinical exam was not useful acutely is that the exam itself may be difficult to interpret because of sedating medications, paralytics or a patient's comorbidities.11 In this situation, it is best to try to minimize these confounding factors, by reducing sedation, the effects of paralytics and the use of pain medications; by correcting underlying metabolic abnormalities, and by stabilizing the patient from a cardiopulmonary standpoint. If it is not possible to control confounding factors, objective tests may be helpful, but should not be performed before the clinical examination at 24 hours unless there is suspicion of status epilepticus or a structural lesion.
The EEG is not a good tool for prognosis because it is easily contaminated by the use of sedating medications, the ICU environment and the metabolic state of the patient.12 It is useful in ruling out the possibility of status epilepticus. Additionally, imaging is only useful in ruling out the possibility of stroke or other structural abnormalities. The most reliable objective tool for prognosis is the somatosensory evoked potential (SEP). This test is not influenced by medications, environment or the metabolic state of the patient and it is easy to perform in the ICU. If both cortical responses are absent, that the chance of the patient awakening is less than 1%.12,13 However, caution needs to be exercised because this test does not take into account functional recovery. Biochemical markers, such as S-100 protein and neuron specific enolase (NSE) that assess neuronal tissue injury are under investigation.14 The physical exam, performed at 24 hours and at regular intervals thereafter, is the best tool for predicting meaningful recovery in patients in coma after cardiac arrest.
A meaningful recovery in survivors of coma after cardiac arrest depends not only on cardiopulmonary resuscitation, but also on cerebral resuscitation. Studies have suggested that just as patients have "stunned" myocardium after cardiac arrest, there may also be "stunned", viable neurons after ischemic/hypoxic cerebral injury.15 Many different neuroprotective strategies have been tested and are still being developed to help mitigate neuronal injury that occurs during the event and continues to evolve after the ROSC. Pharmacological agents that seemed promising in animal studies, but failed in controlled trials include corticosteroids, barbiturates, inhalational anesthetics and calcium channel blockers. Corticosteroids were used to reduce cerebral swelling; however, in clinical trials they were shown to be ineffective and in fact may worsen cerebral damage because of elevation of serum glucose. Steroids are NOT recommended after arrest even if the head CT shows swelling. Barbiturates reduce cerebral metabolic requirements and have neuroprotective effects when given immediately after cardiac arrest, but are no longer used because of their tendency to produce hypotension and arrhythmias. Calcium channel blockers were believed to improve cerebral blood flow and reduce calcium entry into cells; however, in clinical trials they failed to demonstrate any improvement. 15 Other pharmacologie agents under investigation include inhibitors of neuronal apoptosis, like antiapoptotic proteins Bcl-2 and Bcl-X1, excitatory amino acid receptor blockers, such as blockade of glutamate with NMDA and AMPA receptor antagonists, and free radical scavengers, such as superoxide dismutase and the use of normoxic ventilation after cardiac arrest.16 The canine model showed that resuscitation with 21% versus 100% inspired O2 resulted in lower levels of oxidized brain lipids and improved neurological outcome after 24 hours of reperfusion.17 Also under investigation are agents that improve microcirculation by inhibiting vasocontrictive mediators, leukocytes and coagulation. Improvement in neurologic outcome was achieved in animals with heparin, dextran and rTPA.6
Hypothermia is perhaps the most promising neuroprotective strategy. Hypothermia is "a state of body temperature which is below normal in a homeothermic organism". Small clinical trials were performed in the 1960s, but because of management problems and adverse reactions such as arrhythmia, coagulopathy, and sepsis it was not widely used. At that time it was believed that hypothermia must be moderate,
Therapeutic hypothermia acts on many different targets of this damaging ischemic/hypoxic cascade. The protective effects include the slowing of the destructive enzymatic processes, protection of lipid membrane fluidity, reduction in oxygen requirements without impairing microvascular blood flow, inhibition of lipid peroxidation, attenuation of brain edema and reduction of intracellular acidosis. Animal studies have shown that mild hypothermia reduces cell death and has beneficial effects on white matter injury and astroglial cell proliferation.18
In 2002, two prospective randomized studies showed that hypothermia may be an effective neuroprotective strategy. The Australian trial included 77 comatose survivors from cardiac arrest with a primary rhythm of ventricular fibrillation (VF) or pulseless ventricular tachycardia (VT). Patients were randomly assigned to hypothermia (33C, 12 hours; achieved with ice packs) or normothermia. The primary outcome measure was survival to hospital discharge with sufficiently good neurologic function to be discharged home or to rehab. Twenty-one of the 43 patients (49%) treated with hypothermia survived and had favorable neurological recovery, compared to nine of the 34 patients (26%) treated with normothermia (P=0.046). After adjustment for baseline differences, the odds ratio for good outcome with hypothermia compared with normothermia was 5.25 (95% CI 1.47-18.76).19 The European multicenter trial included 275 patients who were seen in the emergency department. They were known to have either VF or VT as the cause of their arrest. They were randomly assigned to receive therapeutic hypothermia (32-34 C, 24 h; achieved with cold air) or to standard treatment with normothermia. The primary outcome measure was favorable neurologic out come at 6 months. Seventy-five of the 136 patients (55%) in the hypothermia group had favorable neurological recovery after six months, compared with 54 out of 137 patients (39%) in the normothermia group. This translates to a number needed to treat of 6. In addition, there was a significant reduction in mortality at six months in the hypothermia group, compared to normothermia.20 In both studies, age, gender and time from collapse to spontaneous circulation did not have a significant effect. There was no difference between the two groups in adverse events.
Many strategies have been developed to initiate and control cooling; e.g., ice bags, blankets containing circulating coolant, cold artery infusion, a cooling helmet and endovascular convection cooling using a balloon tipped catheter. Despite these innovations there are still many questions that need to be answered about the use of hypothermia, such as when is the optimal time to initiate cooling, what is the optimal time to keep a patient cooled, is there permanent benefit and what is the best strategy for rewarming the patient?
With improvements in CPR, neurologic morbidity and mortality will continue to be a consequence of cardiac arrest. Accurate prognosis is possible, but the care of the patient should not stop there. Continued research into neuroprotective agents and strategies is necessary to improve the neurologic short and long term outcomes of victims of cardiac arrest.
1. Holzer M, Behringer W. Curr Opin Anest 2005; 18: 163-8.
2. Rea T, Peredes, V. Curr Opin Crit Care 2004; 10: 218-23.
3. Madl C, Holzer, M. Curr Opin Crit Care 2004; 10: 213-7.
4. Saner H, Rodriguez EB, et al. Resuscitation 2002; 53: 7-13.
5. Sudgreen C., Larsen S. et al. Stroke 2001; 32: 128-32.
6. Bottiger, BW, Martin E. Curr Opin Crit Care 2001; 7: 176-83.
7. Ropper Allan H, Gress Daryl R, et al. eds. Neurological and Neurosurgical Intensive Care 4th ed. Lippincott-Williams and Wilkins, Phila, Pa. 2004.
8. Longstreth WT, Diehr P, et al. NEJM 1983; 308: 1378-82.
9. Sterz F, Holzer M, et al. Curr Opin Crit Care 2003; 9:205-10.
10. Levy DE., Caronna JJ. et al. JAMA 1985; 253: 1420-6.
11. Booth CM, Boone RH, et al. JAMA 2004; 291: 870-9.
12. Bassetti C, Bomio F, et al. J. Neurol. Neurosurg, Psychiatry 1996; 61: 610-5.
13. Zandbergen EG, de Haan RJ, et al. Lancet 1998; 352: 1808-12.
14. Martens P, Raabe A, Johnsson P. Stroke 1998; 29: 2363-6.
15. Grubb NR. Heart 2001; 85: 6-8.
16. Fiskum G. J Neurosurg Anesthesiol 2004; 16:108-10.
17. Liu Y, Rosenthal RE, et al. Stroke 1998; 29: 1679-86.
18. Holzer, M, Bernard SA, et al. Crit Care Med 2005; 33: 414-8.
19. Bernard, SA, Gray TW, et al. NEJM 2002; 346: 557-63.
20. Hypothermia After Cardiac Arrest Study Group. NEJM 2002; 346: 549-56.
MICHELLE L. MELLION, MD
Michelle L. Mellion, MD, is a clinicial neurophysiology fellow, Rhode Island Hospital, Brown Medical School.
Michelle L. Mellion, MD
Phone: (401) 444-4882; e-mail: Mmellion@comcast.net
Copyright Rhode Island Medical Society Nov 2005
Provided by ProQuest Information and Learning Company. All rights Reserved