THE PERSISTENT VEGETATIVE STATE (PROLONGED POSTCOMA UNRESPONSIVENESS) AND POSTHYPOXIC BRAIN INJURY

Published on 10/04/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 1317 times

CHAPTER 9 THE PERSISTENT VEGETATIVE STATE (PROLONGED POSTCOMA UNRESPONSIVENESS) AND POSTHYPOXIC BRAIN INJURY

Prolonged survival in a vegetative state is one of the most harrowing events encountered in modern medicine. Although the condition provides intellectual challenges to the expertise of medical ethicists, legal practitioners, and legislators, the patient’s family members and others intimately involved in the daily care of the patient are usually caught in a web of despair, unresolved grief, anger, a pervading sense of futility, and fading hope. Adversarial relationships between medical staff and family may arise, especially if iatrogenic issues such as anesthetic catastrophes are present. Family members may clutch at slim hopes offered by media reports of recovery after prolonged coma or vegetative states. Physicians may be torn by the conflict between limited resources and the pressure to persevere with intensive support, potentially restricting access to patients with substantially better chances of recovery. Decisions to continue or withdraw support are often charged with ethical and medicolegal implications, inasmuch as either course of action can be construed to be contrary to the “best interests” of the patient.

The various terms vegetative state (persistent and permanent) and postcoma unresponsiveness are used to describe the clinical emergence of the patient from deep coma to a state in which the sleep-wake cycle is reestablished but the patient shows no clinical evidence of cognitive interaction with the world around him or her. When the state is deemed to be unremitting (more than 3 months after anoxic brain damage ^* or more than 12 months after traumatic brain injury) the term permanent vegetative state is recommended. Expert groups in the United States¹ and Europe have suggested using only the terms vegetative state and permanent vegetative state because the term persistent has become shrouded with ambiguity. The terms vegetative state and persistent vegetative state do not imply irreversibility; however, the term permanent vegetative state carries prognostic implications.

For the situation in which some minimal level of purposeful response to environmental stimuli is observed, the term minimally conscious state has been advocated.⁴ The terminological uses vary, and some authors⁵ have advocated avoiding the term vegetative in view of its potentially pejorative connotations, but the original description⁶ remains widely accepted.

CRITERIA

Vegetative State

The salient features of the vegetative state ^† are as follows:

1. Long periods of wakefulness (spontaneous eye opening) without clinical evidence of purposeful responsiveness. Reflex responses and withdrawal to noxious stimulation may occur.

2. Roving eye movements but no sustained tracking of visual stimuli.

3. No response to verbal command.

4. No articulation or mouthing of words.

5. No bladder or bowel control.

Movements such as yawning, grunting, grimacing, startle myoclonus, chewing, swallowing, bruxism, posturing, tearing, and even brief visual fixation may occur, but evidence of a purposeful act clearly linked to a stimulus is lacking. Such patients usually have varying degrees of spastic quadriparesis or decerebrate rigidity, often with preserved brainstem reflexes, including autonomic reflexes that control respiration and, occasionally, swallowing. There is usually a release (disinhibition) of primitive reflexes such as grasp reflexes and pouting reflexes. Joint contractures tend to develop with time.

Rudimentary evidence of cognitive function such as response to command, consistent visual fixation or orientation to and tracking of a visual stimulus, smiling at jokes, or vocalizing recognizable words are inconsistent with the diagnosis.

The Minimally Conscious/Responsive State

Patients may improve from the vegetative state to demonstrate clearly discernible evidence of self or environmental awareness. The term minimally conscious state⁴ or minimally responsive state⁷ has been proposed to describe the condition of such patients. Minimal criteria such as the ability to follow a simple command, the presence of intelligible verbalization, or sustained visual following have been proposed to identify this state. Such responses may be intermittent and more easily evoked by family than by medical staff. Repeated observations may be necessary to distinguish this state from the vegetative state. Such purposeful responses clearly exclude a diagnosis of vegetative state, but difficulties defining the level of cognitive function sufficient to exceed a diagnosis of the minimally conscious state have limited widespread application of the term. The reestablishment of interactive communication, including orientation to place, the ability to give accurate autobiographical information, and the appropriate use of objects, are generally considered to indicate that the patient is functioning at a level higher than the minimally conscious state.

It is important to note that the lack of behavioral manifestations of self or external awareness cannot by themselves prove the absence of a residual or rudimentary subjective awareness. It may not be possible clinically to distinguish a minimally responsive patient with residual cognitive function from one who lacks adequate afferent processing or efferent control to reliably demonstrate understanding or performance of a requested task. Indeed, Owen and colleagues ⁸ reported three patients with clinical diagnoses of persistent vegetative state who demonstrated “covert cognitive processing” in positron emission tomographic scanning with a facial recognition task. It is also conceivable that patients may satisfy the criteria for the minimally conscious state on a “good” day and, perhaps as a consequence of infection, electrolyte disturbance, or fatigue, show no sign of awareness on a “bad” day.⁹ Because of such patients, the boundary between the vegetative state and the minimally conscious state may not be as discontinuous as the definitions suggest.

Locked-In Syndrome

The locked-in syndrome refers to a state in which cognitive processing is clearly evident but severe paralysis prevents articulation or gestural expression of awareness. Communication may be achieved by encoding eyelid or ocular movement. In severe cases, the diagnosis may depend on the demonstration of a normal reactive alpha rhythm on the electroencephalogram (EEG).¹⁰ Such patients usually have isolated lesions of the ventral pons with preserved cerebral hemispheres. Locked-in syndrome may also occur as a consequence of severe peripheral nerve or muscle disease.

Akinetic Mutism

The term akinetic mutism has been used to describe a range of cases in which varying degrees of consciousness, paralysis, and mutism may be present. Various pathological findings have been described,¹¹ including infarction of the anterior cingulate gyrus.¹² Cairns and associates¹³ described an adolescent with a craniopharyngioma who developed repeated episodes of “silent immobility” with open and apparently attentive eyes and occasional whispered monosyllables whose state was reversed on several occasions by aspiration of the tumor. A similar case with improvement after shunting has been described more recently.¹⁴ Akinetic mutism has been divided into frontal and mesencephalic types, the latter distinguished by the presence of vertical gaze palsy or ophthalmoplegia. Cyclosporine neurotoxicity¹⁵ and baclofen toxicity¹⁶ may each cause a reversible state of akinetic mutism. Severe frontal abulia (for example, caused by bilateral anterior cerebral artery infarction or a ruptured anterior communicating aneurysm) may be impossible to distinguish on clinical criteria from either akinetic mutism or the minimally conscious state.

EPIDEMIOLOGY

The incidence and prevalence of vegetative state in the community vary widely in accordance with the sources of the data. A comprehensive review ³ of the epidemiology of vegetative state revealed prevalence rates in the United States ranging from 24 to 168 per million of the population in studies dating from 1990 to 1994. The Multi-Society Task Force on Persistent Vegetative State (1994) accepted an estimate from Ashwal and colleagues¹⁷ of 56 to 140 per million.

The incidence of vegetative survivors after all acute causes at 1, 3, and 6 months has been estimated for the United Kingdom, the United States, and France.³ The incidence falls with time as patients either die or attain higher levels of function and is also heavily influenced by the level of head injury in the community. At 1 month after injury, the incidence ranges from 14 per million in the United Kingdom to 67 per million in France. By 6 months after the cerebral insult, the incidence of survival in a vegetative state is 5 per million in the United Kingdom, 17 per million in the United States, and 25 per million in France. The relative contribution of traumatic to nontraumatic causes of vegetative state also varies considerably, depending on the source of the data. Studies have shown the contribution of head injury ranging from 24%¹⁸ to 72%.¹ Some studies include patients who decline into a vegetative state as a consequence of progressive neurodegenerative disorders, as opposed to progressing to a vegetative state from coma caused by an acute cerebral insult. This and other case ascertainment issues continue to complicate the epidemiological data.

PATHOLOGY

Hypoxic/ischemic cerebral injury, cerebrovascular disease, and traumatic brain injury are the major causes of the vegetative state. Although patients with advanced neurodegenerative disorders (e.g., Alzheimer’s disease) may enter a state clinically indistinguishable from the vegetative state, this chapter is restricted to discussion of patients in a state of postcoma unresponsiveness. Patients may also recover to a state of postcoma unresponsiveness after severe bacterial meningitis, viral encephalomyelitis, or acute disseminated encephalomyelitis.

In the setting of cerebral anoxia or diffuse ischemia, there is usually extensive neocortical laminar necrosis and bilateral hippocampal, amygdalar, and thalamic damage.^19,²⁰ The Purkinje cell layer of the cerebellum is similarly vulnerable to diffuse ischemic injury. A second pattern of brain damage occurs after a short episode of profound hypotension, in which the ischemic damage is confined to the arterial boundary zones. This produces wedge-shaped zones of ischemia with additional bilateral thalamic damage and varying involvement of the hippocampus.

After severe traumatic brain injury, the brunt of the pathology is subcortical and conforms to the diffuse shearing injuries now known as diffuse axonal injury (DAI), as described by Strich in 1956 ²¹ and 1961.²² DAI is graded into types I to III; grade I is defined as diffuse subcortical shearing injury without callosal or brainstem involvement, grade II indicates additional focal lesions in the corpus callosum, and grade III indicates additional lesions in both the corpus callosum and the dorsolateral region of the rostral brainstem. Patients with DAI grade I are unlikely to remain in a prolonged vegetative state unless there is coexisting evidence of a hypoxic/ischemic insult. Coexisting evidence of hypoxic/ischemic damage in posttraumatic cases is not an uncommon finding.²⁰ It is thought to reflect respiratory or circulatory failure at the time of head injury, although disturbances of cerebral autoregulation may also play a role. On the other hand, significant brainstem damage is a rare finding in prolonged survival in a vegetative state, which explains the recovery of sleep-wake function and autonomic cardiorespiratory control in this state.

It appears that thalamic damage may play a critical role in the genesis of prolonged postcoma unresponsiveness, as cases with relatively little cortical damage are well described. Adams and associates ²⁰ published detailed studies of the pathological findings in both nontraumatic and traumatic cases and compared the latter group with 35 traumatic cases in which patients recovered to a state better than vegetative state before death. In summary, patients who remained in a vegetative state generally had more extensive DAI, thalamic damage, or both. Patients with lesser degrees of subcortical damage nearly always had extensive thalamic damage. Higher functioning patients rarely had both bilateral thalamic damage and extensive DAI, and if thalamic damage was present, it was considered to be of a lesser degree. Thus, extensive bilateral damage to the thalamus appears to be critical for the development of the vegetative state, especially if damage elsewhere is minimal.

DIAGNOSIS

Misdiagnoses of the vegetative state and the minimally conscious state appear to be common.²³ Repeated clinical assessments are necessary to clearly establish whether reliable evidence of cognitive function is present in a severely brain-damaged patient. Patients are often in an intensive care environment when the diagnosis is first considered, and varying levels of potentially sedating medications may complicate the assessment. Similarly, metabolic manifestations of other major organ involvement may confound the clinical picture. Neurologists and intensive care physicians may disagree as to whether visual fixation or tracking is present or whether a particular movement is purposeful or reflexive. Nursing staff are often particularly helpful in determining whether an orienting or purposeful response can be consistently linked to a particular stimulus, especially if there is some fluctuation caused by extraneous factors such as seizures or bolus doses of anticonvulsants.

NEUROPHYSIOLOGICAL ASSESSMENT

Clinical neurophysiological evaluation of the comatose patient has long been recognized to play an important role in determining both diagnosis and prognosis. In the modern intensive care environment, patients who remain in a coma after anoxic cerebral injury typically undergo electroencephalographic testing and somatosensory evoked potential (SSEP) testing, in view of the powerful prognostic information that these tests can provide. Usually these tests have been undertaken well before the patient has “emerged” to a vegetative state, and the results, if gravely abnormal, assist in the decision to withdraw cardiopulmonary support. Conversely, unexpectedly favorable electrophysiological findings lead the clinician to reassess the prognosis.

Electroencephalography

Certain electroencephalographic patterns are known to be associated with extremely grave outcomes. Care must be taken to ensure that sedative medications, hypothermia, and severe metabolic disturbances are not confounding the interpretation of the record. In the setting of anoxic/ischemic cerebral injury, several electroencephalographic patterns have been associated with either a fatal outcome or, at best, survival with severe neurological sequelae.²⁴ Of these patterns, the isoelectric or “flat” EEG after the first 24 hours is invariably associated with a poor outcome. The burst-suppression pattern (Fig. 9-1), especially if accompanied by generalized (typically facial and axial) myoclonic status²⁵; the continuous bilateral periodic EEG or generalized epileptiform EEG (Fig. 9-2); and alpha (Fig. 9-3) and theta coma (Fig. 9-4) patterns usually but not invariably indicate a poor prognosis. Serial EEGs are sometimes necessary to identify a deteriorating trend.

Figure 9-1 Burst suppression electroencephalographic pattern two days after prolonged cardiac arrest.

(Reproduced with permission from the American Journal of EEG Technology. Drury I.: The EEG in hypoxic-ischemic encephalopathy. Am J EEG Technol 1988; 28:129-137.)

Figure 9-2 Continuous periodic anterior predominant sharp waves on electroencephalography after prolonged cardiac arrest.

(Reproduced with permission from the American Journal of EEG Technology.)

Figure 9-3 Diffuse unreactive alpha frequency activity on electroencephalography after cardiac arrest.

(Reproduced with permission from the American Journal of EEG Technology.)

Figure 9-4 Diffuse unreactive theta slowing on EEG in an elderly patient after prolonged cardiac arrest.

The isoelectric EEG is used in many countries as a confirmatory test for brain death, and the technique has been standardized.²⁶ The recording must be undertaken for a minimum of 30 minutes at a sensitivity of 2 μV/mm using a minimum of eight channels with double-distance (>10 cm) electrodes in the 10-to-20 montage with filters set at less than 1 Hz and at 70 Hz. At these sensitivities, artifact is common, and the technician must take great care to identify all sources of artifact. ECG artifact is usually present in all channels, and this can usually be readily distinguished from activity of cerebral origin.

The burst-suppression EEG pattern is usually readily identified; however, this pattern is also seen with the use of anesthetic agents—in particular, propofol and midazolam—and careful documentation of the use of such agents is necessary to avoid incorrect interpretation of the EEG. Even short-acting barbiturates can saturate adipose tissue after prolonged use, resulting in very prolonged excretion times.

The continuous bilateral periodic EEG and generalized epileptiform records can be confused with nonconvulsive electrographic status epilepticus. The distinction may well be semantic. In general, patients who develop clinically subtle or nonconvulsive status epilepticus after anoxic cerebral injury do not do well, and aggressive trials of anticonvulsants are typically futile. If doubts remain on clinical grounds (e.g., a known epileptic patient who suffers a near-drowning event), a suitable trial of an anticonvulsant therapy is prudent, and SSEPs can be used as a more reliable guide to prognosis. Periodic complexes are also characteristic of advanced renal and hepatic failure, but these are readily ruled out by appropriate biochemical investigations.

Alpha and theta coma patterns are rare electroencephalographic findings. The salient features of alpha coma pattern are widespread monorhythmic alpha frequency activity, devoid of the typical waxing and waning amplitude of normal alpha waves, and unreactive to eye opening. The alpha frequency activity in alpha coma pattern is often anterior predominant, as opposed to true alpha activity with its occipital predominance. Theta coma pattern tends to be seen more frequently in elderly patients and usually shows some degree of low-amplitude burst suppression. Although alpha and theta coma patterns are considered grave, rare patients have made a satisfactory recovery. The evolution from alpha or theta coma pattern to a burst-suppression pattern in the first week indicates a hopeless prognosis. In contrast, patients who develop continuous mixed frequencies or some degree of reactivity might be expected to show further gains.

Alternatively, more benign electroencephalographic findings are sometimes very helpful in the decision to continue intensive support. In general, near normal EEGs or EEGs with mixed frequencies and some degree of reactivity to auditory, visual, or noxious stimulation indicate a more favorable prognosis, although varying proportions of such patients are left with neurological disability. Similarly, the rare comatose patients who manifest sleeplike changes (vertex waves, K complexes, and spindles on EEG) usually have a good prognosis.

Apart from the technical difficulties often experienced in obtaining good-quality EEGs in patients with severe traumatic brain injury, prognostication for this group based on electroencephalographic findings is recognized to be less reliable. The presence of reactivity to auditory, visual, or noxious stimulation, which may be either “attenuation” or “paradoxical” (decrease or increase in background amplitude after stimulation), implies a better outcome.²⁷ Such a finding suggests that the patient is very unlikely to remain in merely a permanent vegetative state after emerging from coma. Electroencephalographic reactivity has been demonstrated in the cat to depend on an intact reticular activating system and preserved thalamocortical pathways.²⁸ Head-injured patients with less favorable electroencephalographic findings may still do well, however, and in this group the additional use of other neurophysiological parameters—in particular, SSEPs—may help refine prognostication.

SOMATOSENSORY EVOKED POTENTIALS

The bilateral absence of thalamocortical waveforms (N19/P22 or N1) in comatose patients after either anoxic cerebral injury or severe head injury ²⁹ indicates an extremely bleak prognostic outcome. If the study is performed with appropriate care, this is the most robust electrophysiological finding in the intensive care environment. In the setting of coma after anoxia, patients with this finding do not improve beyond the vegetative state. Repeated studies³⁰ have shown this finding to have 100% specificity when the SSEPs are performed after 24 hours of coma. In patients with traumatic brain injury, care should be taken to rule out significant cervical trauma or brainstem pathology that may interrupt large fiber sensory pathways at that level and hence may complicate SSEP interpretation. Access and technical factors such as prominent scalp edema or hematoma can also make results of SSEP studies more difficult to interpret in head injury. In addition, the use of barbiturate-induced coma in severely head-injured patients may attenuate low-amplitude waveforms (B. Day, personal observation, 2003); therefore, SSEP studies for prognostication in these patients should be delayed until the barbiturate-induced coma therapy is withdrawn. Well-formed, normal-latency Erb’s point potentials and cervical potentials are important quality assurance requirements before prognostication can be made without thalamocortical waveforms (Fig. 9-5). Although higher stimulus intensities can be used in comatose patients, the stimulus frequency should be less than 5 Hz. If these technical issues do not complicate interpretation, bilateral absence of cortical SSEP responses is also 100% specific for an outcome no better than vegetative state in patients with traumatic brain injury.

Figure 9-5 Bilateral (superimposed) absence of cortical waveforms with normal-latency, well-defined Erb’s point and cervical waveforms in a patient after prolonged anoxic cerebral injury. C₃, C₄′ are each 2 cm posterior to C₃, C₄.

The high specificity of the bilateral absence of cortical waveforms is, however, accompanied by a low sensitivity (28% to 73%).³⁰ Many patients with preserved but abnormal cortical waveforms nonetheless have very poor neurological outcomes, and the finding of normal cortical responses is not necessarily predictive of a favorable outcome. Unilateral delayed or low-amplitude cortical responses tend to be associated with a more severe residual neurological deficit in the stimulated limb, whereas a well-formed normal-latency cortical waveform is suggestive of a higher level of function in the corresponding limb, but the positive predictive value of such findings varies sufficiently between studies that their clinical utility is limited.

For patients in whom the cortical waveforms (N19/P22 or N1) are present, some attempt to improve the sensitivity of SSEP studies has been undertaken by examining later waveforms, particularly the N70 (also called N3)^31,³² (Fig. 9-6). No patient with bilaterally absent N19/P22 or N70 potentials “awakened” (“awakening” being defined as following commands or having comprehensible speech). No patient with absent N19/P22 potentials had preserved N70 waveforms. Patients in whom the N70 waveforms were present but prolonged beyond 176 milliseconds did not “awaken.” When the absence of the N70 or an N70 latency of more than 176 milliseconds was used instead of bilaterally absent N19/P22 potentials for predicting nonawakening, the sensitivity of SSEP testing increased from 55% to 67%, and the specificity remained at 100%. Identification of the N70 waveform and determination of its peak latency can be technically more difficult, however, and it is more sensitive to the effects of anesthetic medications than is the N19/P22 waveform.

Figure 9-6 Normal-latency, well-defined N70 waveform.

BIOCHEMICAL MARKERS

Cerebrospinal Fluid Creatine Kinase BB activity

Elevated creatine kinase BB (205 U/L) isoenzyme activity measured in the cerebrospinal fluid 48 to 72 hours after cardiopulmonary arrest has been reported to have 100% specificity and 48% sensitivity for never awakening ³³ (as defined previously). Using a lower cutoff increases the sensitivity but lowers the specificity. Using this parameter in conjunction with SSEP studies further increases sensitivity to 78% without sacrificing the 100% specificity required for making decisions about withdrawal of life support.³¹

Serum Astroglial S-100 Protein

The calcium-binding modulator protein S-100 is an established biochemical marker of central nervous system injury.³⁴ Increased serum levels (0.2 μg/L) in comatose patients 24 to 48 hours after out-of-hospital cardiopulmonary arrest were invariably associated with a fatal outcome within 14 days.³⁵

Serum Neuron-Specific Enolase

Serum concentrations of neuron-specific enolase exceeding 33 ng/mL were also predictive of a poor prognostic outcome in comatose patients, with a high degree of specificity (100%), when measured within 72 hours after out-of-hospital cardiopulmonary arrest.³⁶

The role of biochemical markers, especially when used in conjunction with neurophysiological tests, is likely to become increasingly important in establishing the early prediction of outcome from coma caused by anoxic ischemic injury or severe head injury.³⁰

IMAGING STUDIES

Although computed tomographic and magnetic resonance imaging (MRI) studies are often routinely performed in patients with prolonged postcoma unresponsiveness after hypoxic brain injury, these studies only rarely provide useful prognostic information. Diffusion-weighted MRI in the acute stage when the patient is still comatose may show diffuse cortical enhancement that is consistent with laminar infarction (Fig. 9-7A-D). Late imaging usually shows diffuse cortical atrophy (see Fig. 9-7E and F), but correlation between the degree of atrophy and outcome is imprecise. In one notable case, a 60-year-old academician was able to resume his career after emerging from a vegetative state of 8 weeks’ duration, despite generalized cerebral atrophy evident on the computed tomographic head scan.³⁷

Figure 9-7 Magnetic resonance imaging (MRI) findings in a 50-year-old patient who suffered a severe anoxic injury, emerging from coma to a vegetative state after 2 weeks to remain in a permanent vegetative state for 6 months before death. A, Sagittal T1-weighted MRI performed on day 6 after anoxic injury. B, Axial T1-weighted MRI on day 6 after anoxic injury. C and D, Diffusion-weighted images performed on day 6 after the anoxic event, showing diffuse cortical enhancement. E and F, Sagittal and axial T1-weighted images from the same patient 3 months after the anoxic event showing diffuse cortical atrophy and ex vacuo dilatation of the ventricles.

For patients in a posttraumatic vegetative state, more extensive MRI changes of DAI involving the corpus callosum and dorsolateral brainstem are correlated with continuing postcoma unresponsiveness.³⁸

Magnetic resonance spectroscopy and diffusion tensor imaging have yet to be subjected to prospective longitudinal randomized studies in patients with prolonged postcoma unresponsiveness after anoxic cerebral injury or head injury, although early studies ³⁹ suggest that such techniques hold promise.

Measurements of cerebral metabolism and brain activation with positron emission tomographic scanning and functional MRI studies after sensory stimulation are providing important insights into the pathophysiology of the vegetative state (see Laureys et al, 2004).⁴⁰ These techniques are methodologically complex and careful interpretation is required, but important differences between the vegetative state and the minimally conscious state are beginning to emerge. Overall, patients who remain in a vegetative state have cerebral metabolic rates of the order of 30% to 40% of normal⁴¹ (Fig. 9-8), marginally lower than those in minimally conscious patients. A notable finding in the vegetative state is the extent of the metabolic impairment in the polymodal association cortices. These regions are important for higher order processing of incoming afferent information necessary for orientation, attention, recognition, memory, and language⁴² (Fig. 9-9). Such studies will potentially provide important new insights into prognostic and ethical issues, such as the presence of “covert cognitive processing”; however, current methodological complexities and problems with analysis and interpretation limit their clinical application.

Figure 9-8 Cerebral metabolism in various states.

(From Laureys S, Owen AM, Schiff ND: Brain function in coma, vegetative state and related disorders. Lancet Neurol 2004; 3:537-546. Copyright 2004. Reprinted with permission from the American Academy of Opthalmology.)

Figure 9-9 Resting cerebral metabolism in a healthy individual (“Conscious control”) and patients in the vegetative state, with locked-in syndrome, and in the minimally conscious state. Sagittal images showing reduced activity in the medial posterior cortex (precuneus and adjacent cingulate cortex).

PROGNOSTIC, MANAGEMENT, AND ETHICAL CONSIDERATIONS

The phenomenon of prolonged survival in a vegetative state after coma is a direct consequence of advances in modern medical practice, particularly of the popular promotion of techniques such as cardiopulmonary resuscitation and highly expeditious emergency rescue and retrieval systems, coupled with access to sophisticated intensive care facilities. As a consequence, many busy intensive care units are faced on a daily basis with acutely comatose patients with potentially very poor but ultimately uncertain prognoses for independent recovery. Although ethical concerns dictate that economic considerations be set aside in the management of such patients, many intensive care services in large tertiary hospitals have problems with both significant access and cost containment. Ongoing futile management represents a serious misallocation of limited health care resources, as well as an unnecessary addition to the family’s distress. There is little doubt that prolonged survival in a vegetative state represents a disastrous outcome, particularly if it results from overly zealous application of intensive care support of comatose patients in the presence of extremely grave prognostic indicators. There is therefore an urgent need to refine neurophysiological, biochemical, and cerebral metabolic activity imaging studies further as early prognostic indicators to assist clinicians and families make the painful decisions regarding ongoing support.

Once the patient has emerged from coma into a vegetative state with autonomous respiratory function, the ethical issues regarding support become much more vexed. As indicated previously, the vegetative state, especially in its early phase, cannot be considered irreversible. The significant difference in the chances for recovery from the vegetative state that arise as a consequence of anoxic cerebral injury as distinct from that caused by head injury is reflected in the difference in elapsed time before the vegetative state can be considered permanent (Fig. 9-10). Even after these times have elapsed, there are occasional single case reports of late recovery. Review of many of these cases reveals a small error rate for declaring permanence.³ Since the Multi-Society Task Force review of late recovery cases in 1994, only two cases of late recovery have been reported,^43,⁴⁴ which perhaps reflects more widespread acceptance of the definitional criteria and time constraints on the determination of permanence.

Figure 9-10 Outcome for patients in a persistent vegetative state after traumatic or nontraumatic injury. Note that after nontraumatic causes of persistent vegetative state (PVS) in both children and adults, if consciousness has not been achieved by 3 months, no further recovery can be anticipated.

(From Multi-Society Task Force on Persistent Vegetative State: Medical aspects of the persistent vegetative state. Parts 1 and 2. N Engl J Med 1994; 330:1499-1508 and 1572-1579. Copyright 1994 Massachusetts Medical Society. All rights reserved.)

There are few useful data to guide clinicians in predicting recovery from a vegetative state once the diagnosis is established. It is generally considered that the prognosis for emergence is better after trauma than after hypoxia and that younger patients have a better chance of recovery. A short duration of coma before eye opening has been postulated to indicate an improved likelihood of further gains in function beyond a vegetative state.⁵ Recovery of visual pursuit or of a blink response to visual threat are often the first signs of recovery beyond vegetative state, and the early finding of such signs may suggest a better chance of further gains. Evidence on the quality of recovery from vegetative state is also very limited. A large study from France of 522 patients who were vegetative 1 month after head injury showed that although 61% had regained consciousness after 1 year of follow-up, only 14% had become independent. The statistics were markedly affected by age: only 5% of adults became independent, whereas 24% of those younger than 20 years reached this level.⁴⁵ In the nontraumatic vegetative state, only 1% of patients who remained vegetative at 3 months and none who remained vegetative at 6 months achieved an independent outcome by 1 year.¹

Conversely, very prolonged survival in the vegetative state is also unusual, although survival (up to 41 years)⁴⁶ has been described. In the Multi-Society Task Force review, mean survival in a vegetative state after acute cerebral injury ranged from 2 to 5 years¹ with a 70% rate of mortality by 3 years and an 84% rate of mortality by 5 years. However, long-term survival figures are likely to be heavily influenced by decisions to limit treatment.

The decision to withdraw or withhold support (usually only artificial hydration and nutritional support, but occasionally antibiotics for intercurrent infections) from a patient in the permanent vegetative state is an emotional, ethical, and medicolegal challenge. The relevant legal protocols vary between jurisdictions. The ethical considerations hinge on the concepts of patient autonomy, the presence of advance directives, the wishes of family or other proxies, and the ability of a surrogate to act on behalf of an incompetent patient. A full discussion of the ethical and legal issues involved is beyond the scope of this chapter but is well covered in several monographs.^3,⁴⁷

KEY POINTS

• Diagnosis of a vegetative state (postcoma unresponsiveness) depends on careful and repeated clinical assessment for the absence of evidence of a purposeful response to environmental stimuli.

• Diagnosis of a vegetative or persistent vegetative state does not necessarily indicate irreversibility. Diagnosis of permanent vegetative state is reasonably appropriate for patients remaining in a vegetative state for 3 months after anoxic cerebral injury and 12 months after severe head injury. According to these criteria, a diagnosis of permanent vegetative state carries an extremely high likelihood that this state will continue indefinitely.

• Certain neurophysiological and biochemical investigations undertaken early in the course of comatose patients after anoxic cerebral injury and severe head injury can provide highly predictive indicators that a patient will not recover to a state better than vegetative.

• Further studies of neurophysiological and biochemical markers of cerebral injury, both alone and in combination, are likely to refine further the ability to predict such a devastating outcome and thus help prevent ongoing futile interventions.