Hypoxic Brain Injury

Etiology

Anoxic and hypoxic brain injury can occur whenever oxygen delivery to the brain is compromised. Oxygen delivery is a function of the blood flow to the brain and the oxygen content of the blood.[1] Consequently, hypoxic brain injury can result from interruption of blood flow to the brain, such as cardiac arrest or strangulation, or from systemic derangements that affect the oxygen content of the blood. Severe anemia, systemic hypotension, and systemic hypoxia can result in hypoxic brain injury if left untreated. In the United States, cardiac arrest is the most common cause of hypoxic brain injury. Other causes include traumatic vascular injuries; near-drowning; smoke inhalation or carbon monoxide poisoning; shock, including hemorrhagic and septic shock; drug overdoses; and acute lung injury.

Epidemiology

Because the potential causes of hypoxic brain injury are diverse, the precise incidence is difficult to quantify. The best data comes from cardiac arrest and resuscitation literature. According to data from the American Heart Association, over half a million patients in the U. S. suffer a cardiac arrest each year. Unfortunately, the vast majority of patients who experience a cardiac arrest do not survive till hospital discharge; of those who do, the majority (50 to 83%) experienced clinically significant cognitive symptoms.[2] Among those who do not survive hospitalization, death often results when families decide to limit life-sustaining interventions due to concern for severe brain injury.[3]

Clinical outcomes among survivors are variable, but some degree of neurological disability is common; even with state-of-the-art post-resuscitation care, only about 10% of cardiac arrest survivors presenting with non-shockable rhythm achieved a good neurological outcome, defined as neurological function adequate to perform activities of daily living, at 90 days.[4] A recent European study indicated that only 5% of cardiac arrest survivors achieved full neurological recovery at 30 days.[5]

Pathophysiology

The brain depends on a constant energy supply provided by glucose and oxygen but is unable to store energy. With the cessation of blood flow, intracellular production of adenosine triphosphate is diminished. This results in the dysfunction of energy-dependent ion channels, which contributes to intracellular sodium accumulation and cytotoxic edema. Ongoing ischemia results in the release of glutamate, an excitatory neurotransmitter, which promotes calcium influx through N-methyl-D-aspartate (NMDA) receptors. Calcium influx exacerbates neuronal injury by activating lytic enzymes, precipitating free radical formation, and interfering with mitochondrial function. This process, known as excitotoxicity, can ultimately lead to cell death.[6][7]

Histopathology

The mechanisms that lead to delayed cell death following hypoxic-ischemic injury in the brain are complex. Ischemic cell death occurs via two different pathways: necrosis and apoptosis. During hypoxia-ischemia of the brain, acute energy failure leads to loss of ion homeostasis where intracellular sodium and calcium accumulate, creating osmotic swelling, which can lead to cell lysis.[8] This process releases glutamate and free radicals, which are cytotoxic and exacerbate the injury. A secondary phase of neuronal death can occur hours later.

Moderate global ischemia leads to infarcts in watershed areas (e.g., the area lying between regions fed by the anterior and middle cerebral arteries). These infarcts can damage highly vulnerable areas such as pyramidal neurons of the hippocampus (CA1 region) and pyramidal neurons of the cerebral cortex (layers 3, 5, and 6), which leads to laminar necrosis, the death of neurons in the basal ganglia (caudate nucleus and putamen), and the Purkinje cell layer of the cerebellum.[9]

The cells of these areas are high in metabolic demand and contain a high concentration of excitatory neurotransmitter receptors. Other histologic findings include a shrunken eosinophilic neuron (anoxic neuron) and a red neuron representing neuronal cells that die from hypoxia.

History and Physical

Hypoxic or anoxic brain injury often results in an impaired level of consciousness. Patients are often unable to follow verbal commands and may present with coma. Therefore, relevant history must typically be obtained from emergency personnel, family members, or other bystanders who may have witnessed the event. Information regarding the circumstances of brain injury is critical. For example, victims of strangulation may suffer from associated injuries to the cervical spine; the same may be true for patients suffering a near-drowning event after diving into shallow water. For cardiac arrest patients, in particular, information about the initial rhythm, whether or not the arrest was witnessed, and the duration of resuscitation efforts will be important to note. A focused medical history, including medication use, chronic medical conditions, and exposure to toxins or illicit substances, should be obtained if feasible.

Although obtaining a thorough neurological exam is important, the priority for patients presenting with hypoxic brain injury is adequate resuscitation. Specific therapeutic interventions should be targeted at the underlying cause of the injury, with the goal of returning normal cerebral blood flow and oxygen delivery. Systemic hypotension, hypoxia, and hypovolemia should be corrected before attempting a neurological evaluation. 

After appropriate resuscitation, factors that may confound an adequate assessment of neurological function must be excluded. The most prevalent confounder is drug exposure, including both illicit substances and medications that may have been administered during resuscitation efforts. Sedative agents and neuromuscular blockade are common, but toxic levels of certain antibiotics, anticholinergic medications, and antiepileptic drugs may also lead to a depressed level of consciousness. Significant metabolic derangements, such as severe acidosis, acute kidney failure, and acute liver injury, may also preclude an accurate assessment. 

Once confounders have been excluded, the first step of the neurological assessment is an evaluation of the level of consciousness, which involves determining if the patient is arousable. Observe for eye-opening in response to voice, and determine whether the patient follows verbal commands. If not, gentle tactile stimuli may be employed. If this does not result in eye-opening, noxious stimulation is required. This may entail supraorbital pressure, pressure on the temporal-mandibular joint, or other methods of eliciting central pain. Sternal rub and nail-bed pressure are not recommended because these techniques may precipitate spinal reflexes. While applying noxious stimulation, the examiner observes the patient for eye-opening and motor response. Motor response in the setting of noxious stimulation can be described as localizing, i.e., the patient reaches for the source of pain, withdrawal, flexor posturing, extensor posturing, or no response.

Cranial nerve examination should be performed, including an assessment of pupillary reactivity, corneal reflexes, and oculocephalics. Importantly, testing of oculocephalic reflexes is contraindicated if there is a suspected injury to the cervical spine. The position of the eyes at rest should be evaluated, as a dysconjugate gaze or gaze deviation may indicate a focal brain injury. A full examination of the cranial nerves may not be possible for patients with an endotracheal tube, but the gag and cough reflexes should be assessed. Repeating the neurological assessment at intervals can be valuable for determining prognosis; this is discussed in detail below.

Evaluation

In the acute period after the presentation to the hospital, laboratory and radiological evaluation of a patient with hypoxic brain injury are dictated by the underlying cause of the injury. Initial studies should include basic blood work, including blood glucose, electrolyte panel, a complete blood count, blood urea nitrogen, serum creatinine, and liver function studies. An arterial blood gas is often indicated to evaluate the acid-base status and rule out hypercarbia. A urine drug screen and/or blood alcohol level is useful, but it is important to note that many medications and drugs of abuse are not detected by routine urine tests; consequently, a negative UDS is not sufficient to exclude the possibility of drug intoxication or overdose.

A non-contrast head CT should be performed in all patients with depressed levels of consciousness to evaluate for structural lesions. Head CT is sufficient to detect acute hemorrhage, hydrocephalus, and evidence of traumatic injuries such as skull fractures. The primary indication for obtaining a head CT is to identify mass lesions that may require intervention, such as a subdural hematoma or acute hydrocephalus. Often in the setting of an acute hypoxic brain injury, the CT may be relatively unremarkable. However, the loss of gray-white differentiation may be appreciated. This can be quantified by measuring the Hounsfield units of the cerebral cortex and underlying white matter; the ratio of these values has been shown to correlate with prognosis.[10]

For patients who remain comatose after resuscitation, further evaluation may be required. CT angiography and/or CT perfusion may be performed if an acute stroke is suspected and may also be valuable to rule out vascular injuries in patients who have experienced cervical trauma. Electroencephalography is valuable in ruling out nonconvulsive status epilepticus. MRI is more sensitive for detecting hypoxic injury than CT and has been correlated with prognosis; this is discussed in more detail in subsequent sections. 

Treatment / Management

In the setting of brain anoxia due to cardiac arrest, resuscitation efforts, optimizing blood pressure, and maintaining systemic perfusion are critical components of treatment.[9] Clinical management is focused on supportive care, treatment of the underlying cause of the hypoxia, and prevention of ongoing brain injury. A variety of neuroprotective strategies have been evaluated in an effort to prevent cell death by interrupting or attenuating the cascade of events by which hypoxia precipitates neuronal apoptosis and necrosis. The most promising of these is therapeutic hypothermia or targeted temperature management, which has become the standard of care for comatose cardiac arrest survivors.[11] Initial trials evaluated the utility of induced hypothermia, to a target of 32-34 degrees Celsius, for patients with cardiac arrest due to ventricular fibrillation or pulseless ventricular tachycardia.[12][13] (A1)

These randomized controlled trials indicated that patients who received hypothermia had lower mortality rates and improved neurological outcomes. In a 2006 meta-analysis including roughly 400 patients, authors calculated the numbers needed to treat to be seven patients to save one life and five patients to improve neurologic outcomes. There was no evidence of treatment-limiting side effects.[14] In the latest version of the Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiac Care, the American Heart Association recommends that all comatose adult patients after cardiac arrest should receive targeted temperature management, with a goal temperature between 32 to 36 degrees Celsius, maintained for 24 hours after the arrest.[15](A1)

However, in a recent meta-analysis of patient data included in the Targeted Temperature Management (TTM) and TTM2 trials, hypothermia at 33°C was found to be of no benefit compared with normothermia with regard to survival or functional outcome at six months in adult patients surviving out-of-hospital cardiac arrest due to a cardiac etiology. This result contradicts the findings from the HYPERION trial that showed an apparent benefit of hypothermia on functional outcomes. The guidelines should incorporate these findings, and recommendations should be made regarding temperature management by targeting fever prevention.[16]

Although a complete discussion targeted temperature management is outside the scope of this document, there are a few important points to note. Mild hypothermia acts via a number of mechanisms to inhibit apoptosis, excitotoxicity, and mitochondrial dysfunction. Lowering the body temperature also acts to decrease the cerebral metabolic rate, leading to a reduction in oxygen consumption and cerebral blood flow, and stabilizes the blood-brain barrier, which decreases edema formation.[17] To optimize the clinical benefit, treatment must be initiated as soon as possible following the hypoxic brain injury, ideally within 6 hours. Furthermore, clinicians must be aware of the various physiological effects of mild to moderate hypothermia, including hemodynamic effects, effects on glucose and electrolyte homeostasis, renal function, and drug metabolism. Mild to moderate hypothermia often results in bradycardia and can sometimes result in hypotension. Although bradycardia is often well-tolerated, pressors may be required to maintain organ perfusion.

If clinically significant bradyarrhythmias occur, careful rewarming by 1 to 2 degrees Celsius - i.e., from 33 degrees to 34 degrees - is sufficient to resolve the issue. Hypothermic patients are often hyperglycemic due to decreased insulin secretion and relative insulin resistance—additionally, induction of hypothermia results in an intracellular influx of potassium, which can result in hypokalemia. In addition, hypothermia may produce a "cold diuresis," which can contribute to electrolyte abnormalities and result in volume depletion. During the rewarming phase, the sequestration of potassium is reversed, which may result in hyperkalemia, especially if hypokalemia was overtreated during induction and maintenance of hypothermia. For this reason, frequent monitoring of blood glucose and electrolyte levels is required during the induction and rewarming phases. Adequate sedation and analgesia are critical during targeted temperature management, often requiring continuous infusions of sedative agents such as propofol or midazolam, as well as opiate medications. However, these agents can confound neurological assessment. Furthermore, seizures are a common complication of anoxic brain injury. Electroencephalography can, therefore, be very valuable.[18] 

Finally, shivering is a common complication of induced hypothermia or targeted temperature management. Shivering is problematic because it increases the metabolic rate, slows the rate of cooling, and can increase intracranial pressure. Consequently, shivering can seriously undercut the beneficial effects of therapeutic hypothermia. Several methods for treating and preventing shivering have been described in the literature. Adequate analgesia and sedation are important; opiates, in particular, have been shown to decrease the shivering threshold. Counterwarming the hands, face, and upper chest can be very effective. Other pharmacological interventions, including antipyretics, such as acetaminophen; magnesium sulfate; and central alpha-2 blockers, such as dexmedetomidine, have been shown to be valuable.[19] It is recommended that patients undergoing therapeutic hypothermia be monitored for shivering on an ongoing basis and treated accordingly. 

Differential Diagnosis

• Epidural hemorrhage
• Ischemic stroke
• Seizure or post-ictal state
• Subarachnoid hemorrhage
• Subdural hemorrhage
• Traumatic brain injury

Pertinent Studies and Ongoing Trials

Patients with anoxic brain injury have a dismal prognosis despite our understanding of the complex pathways leading to anoxic brain injury. More clinical and basic science research are necessary to determine ways to slow down the anoxic and hypoxic brain injury. Overall, the best treatment solution is prevention.

Prognosis

Anoxic brain injury can be devastating after prolonged cerebral hypoxia; further, it is often challenging to predict the prognosis based on clinical findings.[20] Although the clinical exam provides valuable prognostic information, much of the literature on exam findings pre-dates the advent of therapeutic hypothermia and targeted temperature management. It is also important to recognize that induced hypothermia can significantly affect the pharmacokinetics of many commonly used sedative and analgesic medications, including propofol and fentanyl.[21]

Consequently, examiners must ensure that the patient is normothermic and that sedating medications have been held for an adequate period before performing a neurological assessment for prognostic purposes. It has long been appreciated that patients who demonstrate one or more absent brainstem reflexes in the hours immediately following an arrest can often be observed to improve with time. For that reason, it was often recommended to delay prognostication until 72 hours after the arrest; in particular, the absence of corneal and/or pupillary reflexes at 72 hours has been established as a reliable sign of poor neurological outcome. The motor exam at 72 hours has also been noted to correlate with prognosis in non-hypothermic patients. Specifically, an absent or extensor response to central pain at 72 hours has a high positive predictive value for poor outcomes. Some investigators have attempted to validate these findings in the era of therapeutic hypothermia. In one such study, bilateral absence of pupillary light reflexes was found to be a reliable indicator of poor neurological outcome in patients treated with hypothermia; however, the authors note that it is necessary to exclude surgical pupils and pre-existing pupillary abnormalities. Unfortunately, corneal reflexes and motor examination appear to be somewhat less reliable.[22]

In addition to clinical assessment, electrophysiological studies have been shown to be useful for prognostication. Somatosensory evoked potentials (SSEPs) can be useful because they can typically be evaluated even in the setting of hypothermia and/or moderate sedation. Specifically, bilaterally absent cortical N20 responses following median nerve stimulation SSEPs are highly specific for poor prognosis. It is important to note that the absence of cortical responses is unreliable unless peripheral and spinal responses are preserved. If this is not the case, the possibility of a confounding spine or peripheral nerve process cannot be excluded. Unfortunately, SSEPs are not very sensitive, as this finding is present in a relatively low percentage of patients who go on to have poor outcomes. In contrast, electroencephalography (EEG) can be valuable for identifying patients who may be more likely to recover. EEG monitoring is often used in patients with anoxic brain injury to identify status epilepticus. Postanoxic status epilepticus is associated with an increased likelihood of a poor neurological outcome, but increasing evidence suggests that this entity may respond to aggressive antiepileptic therapy. In addition, investigators have identified several EEG patterns with prognostic implications. Malignant patterns such as generalized suppression, alpha coma, and burst suppression not attributable to medication effect have been demonstrated to correlate with poor outcomes. Conversely, reactive and continuous EEG patterns have been shown to predict awakening from a coma.[23]

The value of magnetic resonance imaging (MRI) in survivors after cardiac arrest has been investigated thoroughly. The literature describes magnetic resonance imaging (MRI) findings in anoxic brain injury in four phases: an acute phase which lasts 24 hours after anoxia or hypoxia; an early subacute phase lasting from one to thirteen days, a late subacute phase lasting from fourteen to twenty days; and a chronic phase, starting at day twenty-one. MRI will show swelling, a hyper signal of basal ganglia, delayed white matter degeneration, cortical laminar necrosis, and atrophy occurring in succession.[24] From a practical standpoint, an MRI obtained 2 to 7 days after the arrest is likely to provide the highest yield. Diffusion restriction and/or hyperintensity on Fluid-Attenuated Inversion Recovery (FLAIR) sequences have been shown to correlate with poor prognosis; involvement of cortical gray matter is more specific than similar findings in the deep nuclei.[25]

Although there is considerable interest in identifying serum biomarkers for prognostication in the setting of brain injury, these have not been widely adopted in practice. Neuron-specific enolase (NSE) has been evaluated extensively for this purpose, and levels > 33 micrograms/liter have been correlated with poor prognosis.[26] However, studies have shown that hypothermia attenuates neuron-specific enolase levels, leading to higher false-positive rates in patients treated with hypothermia when established cutoff values are used. Neuron-specific enolase is also released in the setting of hemolysis, and it can be elevated in patients with other mechanisms of brain injury, including trauma. Consequently, NSE levels should not be used in isolation and should be interpreted cautiously in patients treated with therapeutic hypothermia.

Complications

Anoxic brain injury often results in a prolonged coma, but the pattern of recovery ranges from recovery to mild cognitive deficits to coma and possibly death. Trials show that 27% of patients with post-hypoxic coma regained consciousness within 28 days, Nearly 9% remained comatose or in a vegetative state, and unfortunately, 64% died. Common sequelae can occur after sustaining an anoxic brain injury and may vary from persistent vegetative states, seizures, myoclonus, movement disorder, cognitive dysfunction, and other neurological abnormalities.[27]