Hypoxic-Ischemic Cerebral Injury

The cerebral effects of cardiac arrest are not strictly a neurological intensive care unit (neuro-ICU) problem; neurologists and intensivists are called on to give authoritative opinions regarding the treatment and outcome in these cases. In many ways, the clinical and scientific problems presented by this condition reflect the central challenge to neurointensivists, namely, the sparing and salvaging of damaged neurons. Coma, brain death, and the vegetative state are often derivatives of cardiac arrest. Furthermore, in comparison to the classic problems of critical care neurology such as raised intracranial pressure (ICP) and neuromuscular respiratory failure (which occupy much of the rest of this book), the frequency of cardiac arrest is considerably higher. The dimensions of the problem can be appreciated in that according to recent Centers for Disease Control data: Over 700,000 cardiac deaths occur in the United States every year, with half estimated to be related to cardiac sudden death.

RESUSCITATION

Some of the factors that adversely influence survival and recovery have been identified: advanced age, history of myocardial infarction, and the presence of congestive heart failure (1,2). The prognosis of the cardiac arrest victim depends mostly on how quickly appropriate resuscitation is initiated. Several studies indicate that mortality increases approximately 3% during each minute before cardiopulmonary resuscitation is initiated and continues to increase 4% during each additional minute until the first defibrillatory shock is administered (3). Therefore, increased survival rates after out-of-hospital cardiac arrest should be attainable by reducing the time to cardiopulmonary resuscitation (CPR) and defibrillation (4). The goal of neurologists, nonetheless, has been to attempt to reduce secondary cerebral damage. This is driven by experimental evidence that neurons may not be irrevocably damaged for minutes or hours after exposure to hypoxia-ischemia.

In a patient with cardiac arrest caused by ventricular fibrillation, closed chest massage is at best only a “holding” procedure until defibrillation can be administered. When patients who develop cardiac arrest are defibrillated immediately, as in a supervised cardiac rehabilitation program, 100% survival has been reported. In the future, widespread use of automatic defibrillators in heart attack victims may eliminate delays in defibrillation (5, 6 and 7). This device does not require that the user be skilled in the recognition of dysrhythmias and, therefore, permits an individual trained only in basic first aid to deliver the initial countershocks almost immediately after cardiac arrest. These devices are becoming more widely available, and are now commonly found in large workplaces and public areas such as airports and office buildings.

In addition to rapid defibrillation, effective perfusion of vital organs during CPR is critical for recovery (8). Conventional CPR provides only marginal cerebral perfusion, and cerebral blood flow (CBF) during resuscitation attains only a fraction (2% to 11%) of prearrest values. Both pharmacologic and mechanical attempts have been made to increase cerebral perfusion. Epinephrine in doses
greater than 0.2 mg/kg has increased CBF in animal models of cardiac arrest (9), and therefore has been emphasized in guidelines in patient resuscitation. “New” CPR (now over a decade old), which uses simultaneous chest and abdominal compression and negative intrathoracic pressure, provides greater CBF than present methods of manual CPR, but requires mechanical devices.

Once spontaneous effective cardiac function has been restored, attention must be directed toward assessing the ischemic damage to other organs of the body, which vary in their tolerance to ischemia. Renal tubular cells and myocardial cells can survive periods of circulatory arrest of up to 30 minutes; liver cells can survive up to 1 or 2 hours; and lung tissue can survive more than 2 hours. However, the brain’s ability to tolerate no more than a few minutes of circulatory arrest is the major factor limiting the success of CPR (10,11).

PHYSIOLOGIC ASPECTS OF HYPOXIC-ISCHEMIC BRAIN INJURY

It should be pointed out that ischemia, although difficult to separate from hypoxia in many circumstances, is probably the primary cause of neuronal damage after cardiac arrest. Hypoxia alone arises as a result of strangulation, suffocation, or anesthesia accident and gives rise to somewhat different patterns of neuropathologic damage. In both instances, the rapidity of decline in blood flow or blood oxygen content is a major factor in the degree of damage; this is perhaps truer for hypoxia insofar as we have observed patients to be awake and with minimal cognitive disruption with PaO₂ of 35 mm Hg. In the case of hypotension, however, certain approximate limits seem to be common to all individuals and below which consciousness is lost.

The brain does not store oxygen; therefore, it functions only for seconds and survives only for minutes after its oxygen supply is reduced below critical levels. The extent of tissue injury is a product of both the degree of hypoxia and hypoperfusion, and the duration of exposure to inadequate oxygen delivery. The term “hypoxia” is necessarily ambiguous, because untreated anoxia is a complex state of decreased oxygen availability, systemic acidosis, hypercapnia, and eventual circulatory collapse.

As discussed later, it should be emphasized that the neuropathologic pattern of low blood flow is either heterogeneous infarction in the border zones between major cerebral vessels (“watershed infarction”) or global damage throughout the cerebrum, mainly in layers 3 and 6 of the cortex but also in the lenticular nuclei; by contrast, severe hypoxia causes preferential damage in special vulnerable areas, mainly the medial temporal lobes. To the extent that widespread cortical infarction typical of reduced blood flow also occurs at times from hypoxia, and the extraordinary medial temporal lobe Korsakoff syndrome that is typical of hypoxia can also result from cardiac arrest, the distinctions are not pure, but a differentiation between the various pathologic patterns serves a useful purpose. For brevity, “anoxia” is sometimes used in the following discussion as shorthand for the combined ischemia-hypoxia that is necessarily part of a cardiac arrest. Some of these issues are detailed in the following.

NEUROLOGICAL SYNDROMES AFTER CARDIAC ARREST

A spectrum of clinical disorders can result, depending on the severity of cerebral ischemia or anoxia (12,13) (Table 17.1).

Transient or Mild Neurological Deficits

Global ischemia can result in both metabolic dysfunction and structural injury to the central nervous system (CNS). Brief episodes of cerebral anoxia generally are well tolerated, and patients usually awaken promptly. Patients with slightly longer episodes of systemic circulatory arrest suffer mild degrees of cerebral anoxia and have a reversible metabolic encephalopathy. Drowsiness in these patients, if present, lasts only a few hours, usually less than 12.
Residual signs of confusion or amnesia may persist for hours to days. In general, recovery is rapid and complete, and these patients are able to resume their previous occupations.

TABLE 17.1. Neurological syndromes after cardiac arrest

Transient central nervous system (CNS) deficits after brief coma (<12 h)
	Pathology		No damage or scattered ischemic neurons
		Clinical	Transient confusion often followed by antegrade amnesia
		Outcome	Rapid, complete recovery; delayed deterioration (rare)
Persistent focal CNS deficits after coma (>12 h)
	Cerebral syndrome
		Pathology	Focal or multifocal infarcts of cortex, especially in boundary zones
		Clinical	Amnesia
			Dementia
			Bibrachial or quadriparesis
			Cortical blindness, visual agnosia
			Also may occur: seizures, myoclonus (acute state); ataxia, intention myoclonus, parkinsonism (chronic stage)
		Outcome	Slow, often incomplete, recovery
	Spinal cord syndrome (may occur in isolation or accompany cerebral syndrome)
		Pathology	Focal or multifocal infarcts of spinal cord, especially in the lower thoracic boundary zone
		Clinical	Flaccid paralysis of lower limbs
			Urinary retention
			Loss of pain and temperature sense
			Preserved touch and position sense
		Outcome	No or incomplete recovery
Global CNS damage (no recovery of consciousness)
	Destruction of hemispheres alone
		Pathology	Laminar necrosis of cortex
		Clinical	Vegetative state (awake but unaware)
		Outcome	Prolonged survival in vegetative state
	Brain death
		Pathology	Necrosis of cortex + brainstem ± spinal cord
		Clinical	No evidence of cortical activity, no brainstem reflexes, reflexes of purely spinal origin may persist
		Outcome	Systemic death within days

Amnestic Syndrome

An amnestic syndrome may follow a brief period of postanoxic confusion or may occur as an isolated phenomenon. Finklestein and Caronna (14) followed 16 patients after cardiac arrest, none of whom remained in coma after resuscitation, but developed an amnestic syndrome as their only neurological sequel. All had severe antegrade amnesia and variable retrograde memory loss with preservation of immediate and remote memory, resembling Korsakoff psychosis, and a bland, unconcerned affect, with confabulation. Recovery was complete within 7 to 10 days in 12 of the 16 patients; amnesia persisted for a month or longer in the other four; all later recovered. The time required for recovery and the occasional instances of incomplete recovery distinguished this syndrome from transient global amnesia and the postictal state, from which recovery is rapid and complete. Electroencephalograms (EEGs) and computed tomography (CT) scans failed to identify a cerebral lesion, although changes in the hippocampal region are almost always appreciated nowadays with magnetic resonance imaging (MRI). In view of the vulnerability of the hippocampal regions to anoxia, transient post-cardiac arrest amnesia is thought to represent partially reversible bilateral damage to the hippocampi. Subtle but more permanent cognitive impairment may also follow cardiac arrest. Individuals with anoxic-ischemic coma of more than 6 hours’ duration but with unremarkable cranial MRI and CT have demonstrated persistent poor learning and recall of paired associations when compared with controls matched for age and intelligence quotient (IQ). Impaired neuro-transmitter
synthesis may be responsible for these mild amnestic syndromes of anoxic amnesia, possibly through impairment of cholinergic memory circuits (15).

After apparently recovering from the immediate effects of an anoxic insult to the brain, rare patients develop a progressive cerebral disorder and relapse into unconsciousness (16). In fatal cases, pathologic lesions have been primarily restricted to the deep white matter of the parietal and occipital lobes. The clinical syndrome of this delayed anoxic leukoencephalopathy is distinctive: Days to weeks after a period of improvement or recovery from global anoxia, patients suffer progressive neurological deterioration and often die or remain comatose. Occasional cases are encountered that have a slow recovery (17). Delayed neurological deterioration has been reported after all types of anoxic insults but follows no more than one or two of each thousand arrests and is not predictable by the type of insult, duration of anoxia, period of coma, or any identifiable clinical feature (18).

Syndromes of Persistent Central Nervous System Damage

Focal Cerebral Syndromes

Several types of strokelike focal structural brain lesions occur after severe or prolonged hypotension (19). Patients in this group usually remain in coma for 12 hours or more and on awakening have lasting focal or multifocal motor, sensory, and intellectual deficits. Among the focal signs clinically manifest in this group of patients are three main syndromes: partial or complete cortical blindness, bibrachial paresis, and quadriparesis.

Cortical blindness, often transient (20) but occasionally permanent, probably results from disproportionate ischemia of both occipital poles as a result of their location in an arterial border zone, as alluded to in the preceding (21). The visual syndromes are commonly seen in children following hypoxic events. Magnetic resonance imaging has demonstrated injury in the striate and parastriate cortices and the regions of the optic radiations (22). The extent of visual recovery is most related to the age at insult and extent of injury demonstrated in the optic radiations.

Bilateral infarction of the cerebral motor cortex in the region that innervates the shoulder in the border zone between the anterior and middle cerebral arteries appears to be responsible for the syndrome of bibrachial paresis sparing the face and legs. The deficit of more proximal than distal pyramidal weakness has been described as “man in a barrel,” and in one report was seen in 11 of 34 comatose patients following hypoxic injury (23). Many of those patients were severely affected and did not survive. The “man in a barrel” syndrome can be seen with less severe injury, and has been reported in a hypoperfused cardiac patient who never lost consciousness (24). Recovery often is incomplete and delayed over a period of weeks to months. Some of these patients eventually are able to lead a relatively independent existence at home (25), whereas others remain in nursing homes, severely disabled and dependent.

Spinal Cord Syndromes

The spinal cord generally is more resistant to transient ischemia than more rostral parts of the CNS. Nevertheless, rare cases of isolated spinal cord infarction after cardiac arrest occur without evidence of cerebral injury. Necrosis of central structures of the spinal cord can occur in the periphery of the territory supplied by a main contributory artery. These border zones or “watersheds” in the upper thoracic and lumbar regions of the spinal cord are at risk from any profound drop in perfusion pressure. The syndrome of spinal stroke from hypotension is characterized by flaccid paralysis of the lower limbs, urinary retention, and a sensory level in the thoracic region, with pain and temperature more affected than light touch or position sense.

Syndromes of Global Cerebral Damage

A third group of resuscitated patients has more widespread destruction of the brain, progressing
to either a vegetative state or brain death. Some patients with severe irreversible brain damage who survive for more than 1 week regain eye opening, sleep-wake cycles, spontaneous roving eye movements, and other reflex activities at brainstem and spinal cord levels but remain in a functionally decorticate state of wakefulness without awareness. This state, distinct from the sleeplike condition of coma, is referred to as (perhaps unfortunately) the “vegetative state” (26). In rare instances, recovery of cognition has occurred after prolonged unconsciousness (see later).

Extrapyramidal tract dysfunction after anoxia can produce a clinical syndrome with elements of parkinsonism. It has been reported particularly after carbon monoxide poisoning but may follow an episode of anoxia or ischemia. In some cases, parkinsonian features are only a small part of widespread cerebral injury, whereas in others, the signs of parkinsonism (i.e., rigidity, akinesia, and tremor) are the only neurological disability. Some patients have responded to treatment with L-dopa, but this is exceptional in our experience (27).

Myoclonus

Myoclonic jerks and convulsions may acutely follow episodes of acute cerebral ischemia, especially when they are severe enough to cause coma. Myoclonus refers to irregular, asynchronous shocklike jerks of one or more limbs, and it is a relatively common manifestation of disturbance in function of diverse regions of the CNS. The acute myoclonus has not been well understood, although the brainstem is likely involved in the origin of the movement disorder (28). The myoclonic activity may resolve rapidly over the first 24 to 48 hours, and specific treatment is difficult. There is only an inconsistent relationship to paroxysmal EEG activity, and traditional anticonvulsants have little effect. Nonetheless, if the movements are upsetting to the family or interfere with ventilation they can sometimes be suppressed by large doses of a benzodiazepine or by morphine. When severe and protracted, the myoclonic movements are associated with a dismal prognosis. In a report of 107 consecutive patients comatose after cardiac arrest, myoclonic status was seen in over one third, and associated with burst suppression on EEG, cerebral edema, and infarctions on CT imaging (29). Mortality was 100% in that series.

An entirely different “action myoclonus” syndrome described by Lance and Adams has been recognized after recovery from coma secondary to cerebral ischemia (30). Here, the awake patient is incapacitated by jerking during attempted use of the limbs. Jerks in this condition may also be stimulus-activated and brought on by light, sound, or touch. Thus, these involuntary jerks can incapacitate the patient in walking, eating, and using the upper limbs for carrying out other activities of daily living. Some control of this intention myoclonus has been reported with a combination of 5-hydroxytryptophan (31), clonazepam (32), and valproic acid (33). We have found benefit from piracetam and more recently with levetiracetam. The anatomic origin of this syndrome is a matter of debate. Chronic posthypoxic myoclonus is most commonly considered a type of cortical reflex myoclonus, but reticular reflex myoclonus and an exaggerated startle response also may occur (28). In at least some cases the evidence favors a diffuse cerebellar cortical lesion, rather than or in addition to the more commonly proposed cerebral cortical lesion (34).

Cerebellar ataxia is another, albeit infrequent postanoxic syndrome that appears to be related to the selective vulnerability of Purkinje cells to ischemia. These patients are ataxic in all movements and gait but tend not to have nystagmus or severe dysarthria.

NEUROPATHOLOGIC FEATURES OF CEREBRAL ANOXIA

Infarction in Cerebral and Spinal Boundary Zones

A period of circulatory arrest, preceded or followed by appreciable periods of hypotension, often leads to ischemic alterations concentrated in the boundary zones between major
cerebral arteries in the cerebral cortex, basal ganglia, cerebellum, and spinal cord (35). In the cerebral cortex, ischemic necrosis after profound hypotension is most severe in the parieto-occipital regions, where the territories of the anterior, middle, and posterior cerebral arteries meet (Figs. 17.1 and 17.2). Necrosis is less prominent toward the frontal and temporal poles, and along the boundary zones of the anterior-middle and middle-posterior cerebral arteries. The cerebral cortex is commonly affected bilaterally but may occur unilaterally if there is carotid stenosis on one side. Cortical boundary zone infarcts often have petechial hemorrhages because of reperfusion when blood pressure is restored. Infarctions may coexist with focal injury to vulnerable neurons, as well as with diffuse cortical necrosis.

FIG. 17.1. Axial computed tomography scan 24 hours after cardiac arrest showing widespread infarction of convexities, including border zones between anterior cerebral and both middle and posterior cerebral arteries. There is also infarction within the anterior cerebral artery territory, demonstrating that cerebral ischemic damage does not always conform to watershed flow regions.

FIG. 17.2. Magnetic resonance imaging with FLAIR sequences demonstrates ischemic change in the border zone regions.

Within the spinal cord are border zones between the anterior spinal artery and the segmental arteries from the aorta. Severe hypotension or obstruction of the segmental arteries causes infarction of the spinal cord in the cervicothoracic and the thoracolumbar regions, as mentioned previously. The thoracic cord usually is involved at the boundary zone between the anterior spinal artery and the artery of Adamkiewicz (36).

The relationship between vascular territories subjected to anoxia by microemboli, thrombi, or hypotension; and the location, size, and number of cerebral infarcts is often perplexing. Nevertheless, the boundary zone hypothesis best explains the focal neurological deficits observed in patients after profound hypotension. The predilection of infarcts for the boundary zones may reflect the fact that the tissue effects of reduced oxygen and glucose delivery are likely to be maximal in those regions most distant from the origin of a major artery, but neither the topography nor the severity of cerebral infarcts can be predicted from clinical estimates of the duration of hypotension or circulatory arrest. In addition, some clinical and experimental studies have not consistently found infarctions in boundary zones after hypotension. In primates subjected to cardiac arrest, Miller and Myers (37) observed that the brainstem was regularly affected, but cerebral infarcts did not occur as long as hypotension before and after circulatory arrest was prevented. Clinical and pathologic verification of infarction, predominantly of the brainstem, has been provided by Boisen and Siemkowicz (38), who observed the “locked-in” syndrome in three survivors of cardiac arrest, but such cases are rarely encountered.

Diffuse Central Nervous System Injury

Prolonged cardiac arrest causes widespread death of neurons. The following groups of neurons are vulnerable to even moderate degrees of anoxia: pyramidal cells in Sommer’s sector of the hippocampus, Purkinje cells of the cerebellum, and pyramidal cells of the third and fifth layers of the cerebral cortex. In cases of the vegetative state, neuropathologic examination has revealed widespread necrosis of the cortex and thalamus, whereas the brainstem and spinal cord remained intact. Profound anoxia affects the cortex as well as the basal ganglia and brainstem nuclei and is not compatible with prolonged survival.

DIAGNOSTIC APPROACH TO ANOXIC COMA

Neurological Monitoring

Patients in coma initially require intensive care to support life and prevent complications. In most cases of coma, sequential observation of physical signs is paramount because they may indicate the degree of reversibility of CNS damage and predict likely outcome, or detect a loss of neurological function that precedes deterioration in cardiorespiratory function. The neurological examination of the comatose patient consists of an assessment of the level of consciousness as determined by eye opening, verbal responses, and reflex or purposive movements in response to noxious stimulation of the face, arms, and legs (39); neuroophthalmologic function as indicated by pupillary size and response to light, spontaneous eye movements, oculocephalic (doll’s eyes) and oculovestibular (ice water caloric) responses; and vegetative function as reflected mainly by the respiratory pattern (39). Neurological signs can be correlated with specific anatomic sites to establish the severity and extent of CNS dysfunction (Table 17.2).

Only gold members can continue reading. Log In or Register to continue