Clinical manifestations

Presentation and course

Persistent vegetative state/unresponsive wakefulness syndrome (VS/UWS). The American Neurological Association Committee on Ethical Affairs and the Quality Standards Subcommittee of the American Academy of Neurology have individually, and in association with other specialty societies, defined the VS/UWS as follows (175):

	1. No evidence of awareness of self or environment and an inability to interact with others;
	2. No evidence of sustained, reproducible, purposeful, or voluntary behavioral responses to visual, auditory, tactile, or noxious stimuli;
	3. No evidence of language comprehension or expression;
	4. Intermittent wakefulness manifested by the presence of sleep-wake cycles;
	5. Sufficiently preserved hypothalamic brainstem autonomic functions to permit survival with medical and nursing care;
	6. Bowel and bladder incontinence;
	7. Variably preserved cranial-nerve reflexes (pupillary, oculocephalic, corneal, vestibulo-ocular, and gag) and spinal reflexes.

Thus, VS/UWS is typically a chronic condition that preserves the ability to maintain blood pressure, respiration, and cardiac function but not cognitive function (07; 175; 140). Patients in VS/UWS can survive for many years if adequately fed and nursed. A neurologic examination with neuroimaging is necessary to rule out treatable infarction, hemorrhage, or infection; if the cause was traumatic, a general physical examination would be beneficial to determine other sites of injury. Eye opening, language function, and purposeful movements in response to noxious stimulation should be determined to assess the level of consciousness (138). Absolutely critical is the need to differentiate VS/UWS the from locked-in syndrome. Searching for voluntary eye movements, if necessary, after passive eye opening by the examiner and with the help of a mirror, is essential to identify locked-in patients--these patients are immobile except for rudimentary eye and lid movements, but they are conscious (100).

The neuro-ophthalmologic examination further evaluates pupillary size and response to light, spontaneous eye movements, and oculocephalic (doll's eyes) and oculovestibular (ice water caloric) responses. In some patients with VS/UWS, spontaneous nystagmus during wakefulness or REM sleep may be present (81).

Patients in VS/UWS should be monitored for important predictors of recovery, including recovery of spontaneous motility, eye tracking (which would indicate transition to MCS), oculocephalic reflex, and disappearance of oral automatisms (51). The Glasgow Outcome Scale was not thought to be the best indicator for recovery as recovery of voluntary movements may not be relevant (134), but an extended and revised version, which includes the VS/UWS and minimally conscious state categories, has been found to be highly correlated with the CRS-R. It is brief, easy to use, and requires no special training (70).

Minimally conscious state (MCS). As determined by the Aspen Neurobehavioral Conference Workgroup, the MCS denotes patients with severely altered consciousness with intermittent but reproducible, minimal but definite, behavioral evidence of self, or environmental awareness (72).

The behavioral activities associated with the MCS include (72):

	1. Ability to follows simple commands 2. Ability to gesture or verbalize “yes” and “no” responses 3. Intelligible verbalizations 4. Purposeful behavior, including movements or affective behaviors that occur in contingent relation to relevant environmental stimuli that are not due to reflexive activity, for example:
		(A) Appropriate smiling or crying in response to emotional topics or stimuli (B) Vocalizations or gestures that occur in direct response to the linguistic content of questions (C) Reaching for objects, demonstrating a clear relationship between object localization and direction of reach (D) Touching or holding of objects that accommodates the size and shape of the object (E) Sustained visual pursuit or fixation in direct response to moving or salient stimuli.

Furthermore, it has been suggested that also auditory localization should be considered as a sign of minimally conscious state based on multimodal findings (30). The characteristic feature of the MCS is response inconsistency, which may require multiple examinations for diagnosis as the aforementioned behaviors can occur infrequently, or be ambiguous, or masked by other clinical features (77). The European Academy of Neurology Guideline on the Diagnosis of Coma and Disorders of Consciousness recommends never to rely on an isolated evaluation when assessing consciousness in unresponsive patients with brain injury (100). Instead, five assessments over several days (eg, within 10 days) appear appropriate to evaluate the level of consciousness in patients with prolonged disorders of consciousness (36% misdiagnosis with a single assessment vs. 5% with five assessments) (184). Once reproducible functional communication and object use are accomplished, the patient is considered to be in eMCS.

Of great concern is whether patients with VS/UWS and MCS are able to, or do, experience pain. Given their disturbances in consciousness, such patients cannot communicate feeling or suffering from pain. Medical professionals are uncertain as to the presence of pain perception in these patients (45).

Various methods have been devised to determine the presence of pain in patients with disorders of consciousness; the best, currently, is the Nociception Coma Scale-Revised (NCS-R), a validated scoring system assessing motor, verbal, and facial expression responses at rest and to non-noxious and noxious stimuli (35; 37; 137). Facial grimacing is thought to be linked to nonconscious brain processing and so can be seen in patients in VS/UWS (164). The weakness in the NCS-R is its dependency on behavioral or motor responses, which may underestimate subtle signs of pain in patients with severe motor limitations (137). Measures of autonomic functions during noxious stimulation may be helpful in this regard (see below).

The brain’s “pain matrix” is divided into two subsystems: a lateral sensory-discriminative subsystem consisting of the primary and secondary somatosensory cortices, thalamus, posterior insula, and cerebellum; and a medial affective-cognitive subsystem consisting of the anterior cingulate cortex and the anterior insula (90). Activation of the lateral sensory-discriminative subsystem was seen on PET scanning after median nerve stimulation in all 15 patients in VS/UWS (111). Incomplete activation of both pain subsystems was seen with PET scanning in post-anoxic patients in VS/UWS, thereby questioning the presence of conscious pain perception in these patients (93).

Activation of the anterior cingulate cortex with noxious stimulation was seen on PET scanning in some patients with disorders of consciousness and correlated with NCS-R total scores (36). Activation of one or more components of this pain matrix was seen on fMRI in 24 of 44 (54.5%) patients with VS/UWS when hearing the recorded pain cries of others (200). This is similar to the responses of normal subjects and may represent “affective consciousness,” a more fundamental layer of awareness than cognition (200). Celesia and Sannita raise the possibility that these activations may only be autonomic reflexes, not indicative of consciousness, absent other evidence of integrative brain processes (34).

Two other pain issues in patients with disorders of consciousness are important:

(1) Treatment of pain. From a beneficence perspective, if a patient with a disorder of consciousness is in pain, treatment should be instituted. However, over-treatment could inhibit cognition and potentially affect brain plasticity. Both over- or undertreatment could, theoretically, inhibit recovery albeit by different mechanisms; yet, these mechanisms and their long-term effects are unknown (164). The NCS-R has been proposed as a tool for evaluating clinical pain management (37), but assessing pain severity and response to analgesic therapy is clearly problematic.

(2) Withdrawal of artificial nutrition and hydration and, hence, life-sustaining therapy as in the Nancy Cruzan case (see Clinical vignette). Artificial nutrition and hydration have been revisited in patients with VS/UWS because of the possibility that it is painful if not managed properly (146).

Cognitive-motor dissociation. Patients who do not follow commands at the bedside but are able to follow commands by modifying their brain activity during fMRI- and EEG-based active consciousness paradigms are thought to be in a state of cognitive-motor dissociation (43; 44). This condition is also known as nonbehavioral MCS, MCS*, functional locked-in syndrome, or covert consciousness (160). Cognitive-motor dissociation can, thus, only be diagnosed in unresponsive patients who are investigated using advanced technology that is only available in a few highly specialized centers worldwide so far. However, this is an urgent area for further research: according to a meta-analysis, roughly 15% of chronic patients with a clinical diagnosis of VS/UWS are actually in cognitive-motor dissociation because they can follow commands by modifying their brain activity when assessed with fMRI or EEG consciousness paradigms (103). Cognitive-motor dissociation also exists in the acute stage. In a landmark study, 16 of 104 unresponsive patients (15%) in the intensive care unit had brain activation detected by EEG at a median of 4 days after brain injury (43). Interestingly, in cognitive motor dissociation patients with subcortical lesions, structural and functional data suggest impaired motor command integration at the anterior forebrain mesocircuit level with preserved thalamocortical network function, whereas cognitive motor dissociation patients with bilateral cortical lesions appear to have preserved corticocortical network (71). Prospective multicentric studies with increased sample sizes and using neuroimaging and physiology-based paradigms are needed to identify and further validate cerebral markers indicative of cognitive-motor dissociation (161). Importantly, recovery trajectories of clinically unresponsive patients diagnosed with cognitive-motor dissociation early after brain injury are distinctly different and better compared to unresponsive patients without cognitive-motor dissociation (57).

Prognosis and complications

The standard prognostic estimates, originally thought to be at high probability levels, are that the VS/UWS (aka, unresponsive wakefulness syndrome) is “permanent” after 3 months in nontraumatic cases and after 12 months in traumatic cases (175); but these estimates have been questioned, one reason being that there had been no categorization of the MCS at the time of this report in 1994. Therefore, some patients thought to be VS/UWS when these estimates were created may well have been MCS. Additionally, it is more accurate to determine prognosis by etiology than by syndrome, and mortality may have been overestimated in patients whose demise was due to discontinuation of life-sustaining therapy due to withdrawal of artificial nutrition and hydration (14). The earlier patients in VS/UWS transition to MCS, especially within the first 8 weeks, the more likely they are to continue recovering and achieve higher levels of functioning (94). Ultimately, nearly half the patients in VS/UWS and MCS followed for at least a year recovered to daytime independence at home, and 22% returned to work or school (94).

Several studies have shown that the outcome in traumatic VS/UWS and MCS tends to be better than in nontraumatic states and that some patients regained consciousness beyond these standard estimates (94; 74); Estraneo and colleagues found that six of 50 patients regained consciousness (59). Baricich and colleagues reported 3 of 10 patients who were in VS/UWS at 36 months and transitioned to the MCS over the next year (10). Late recovery was associated with younger age and a traumatic etiology (59). Estraneo and colleagues also found, although the numbers are small, that hemorrhagic cases of VS/UWS fare worse than anoxic cases (59). Regarding MCS, a third of patients in this state improved to emergence from MCS more than 1-year post-onset, whereas no patient in VS/UWS state improved after 1 year (114). Nonetheless, in these studies the improved patients remained severely disabled. In a study of traumatic brain injury cases followed for 5 years, those who began to improve during inpatient rehabilitation (approximately 90 days post-injury) continued to demonstrate meaningful recovery for at least 2 years and possibly for 5 years. Those who did not show early improvements still underwent limited recovery up to 2 years but not thereafter (187; 188).

A meta-analysis found that recovery of consciousness occurred in 78% of patients in the posttraumatic VS/UWS at 1 year, whereas such recovery occurred in 17% of nontraumatic VS/UWS patients at 6 months (74). Further, of the remaining patients in the VS/UWS (83%), only 7.5% recovered consciousness at 2 years.

In a study of patients emerging from the VS/UWS after 36 months, univariate analysis revealed male gender, young age, shorter duration of consciousness disorder, and the absence of seizure activity were the most important characteristics for transitioning from VS/UWS to MCS. Other variables, especially the etiology of coma and the CRS-R score on acute admission, were not found to be significant in predicting late recovery (107). Late recovery was seen in 32% of 27 patients in VS/UWS and seven patients in MCS, although they remained severely disabled (196).

In the acute setting, recovery of prolonged disorders of consciousness has also been noticed in several worldwide cohorts of critically ill patients with coronavirus disease 2019 (COVID-19) (54).

Clinical vignette

Case 1. A 22-year-old male student suffered a traumatic brain injury when hit by a car. He went into coma immediately with a GCS of 4. He had extensive cerebral edema, subdural and extradural hematomas, and bilateral frontotemporal contusions requiring emergency surgery. After 21 days in coma, there was spontaneous eye opening. He was transferred to a brain injury rehabilitation institute. A repeat CT scan 5 months later showed posttraumatic hydrocephalus status after shunt surgery. There were somatosensory-evoked responses bilaterally. EEG and brain perfusion SPECT scan showed diffuse slow-wave activity and decreased cerebral blood flow. Nine months after injury, a cranioplasty was performed. Two months after this procedure, the patient was smiling appropriately to his mother, thereby demonstrating transition from VS/UWS to MCS. At 19 months, he had consistent behavioral responses: appropriate smiling and pursuit eye movements, and he obeyed simple commands with his left arm. At 22 months, there were active movements in all limbs, and he could verbalize on request. At 24 months, he emerged from MCS and was discharged home with a rehabilitation program but was impulsive and lacked control. At 31 months, he could walk a few meters with a walker. At 4 years, he was able to walk with a cane. At 6 years, he was oriented to certain information, could perform language, spatial, and executive tasks. He obtained a part-time, noncompetitive job and resumed studying (154).

Case 2: Karen Quinlan. In 1975, Ms. Quinlan, age 21 years, imbibed a combination of alcohol and methaqualone following several days of not eating. She suffered a cardiopulmonary arrest and when found was unresponsive, apneic, pulseless, cyanotic and had dilated pupils. She received cardiopulmonary resuscitation and spontaneous respirations, and normal vital signs returned within the first hour. Aspiration and tracheostomy occurred on the second day; other brainstem reflexes returned in the first week. During the first 6 months she had unequivocal sleep-wake cycles but never demonstrated signs of awareness of the environment nor cognitive functions. Her movements were stereotypic with given stimuli but never purposeful; all consisting with VS/UWS. Cerebral angiograms were unremarkable. EEGs showed beta activity when awake and intermittent theta activity when asleep. Five years post-arrest a computed tomographic scan showed diffuse cerebral and cerebellar atrophy. She died 10 years later from overwhelming infection. Neuropathological evaluation showed significant diffuse atrophy (brain weight 835 g; normal weight 1300 g); the most severe damage being in both thalamic nuclei (96). Ms. Quinlan’s case received extensive media attention and raised important issues of right-to-die, medical ethics, legal guardianship, and euthanasia.

Case 3: Nancy Cruzan. In 1983 Ms. Cruzan, age 25 years, was involved in a motor vehicle accident. She was resuscitated by paramedics at the scene but spent the next 7 years in VS/UWS. A landmark legal battle ensued when her parents asked that her feeding tube be removed; the feeding tube being a medical treatment that could be refused under the Due Process Clause of the United States Constitution. The United States Supreme Court decided that “clear and convincing evidence was needed” that a patient, when competent, would not have wanted such therapy before it could be terminated. The family gathered sufficient evidence in that regard, and a local judge ruled in their favor. The feeding tube was removed, artificial nutrition and hydration was stopped, and she died 12 days later. Ms. Cruzan’s case led to the creation of advance health directives.

Case 4: Terry Schiavo. Ms. Schiavo was 23 years old in 1990 when she suffered a cardiac arrest with significant anoxic brain damage. She spent the next 15 years in VS/UWS. Her case received massive media attention when her husband petitioned the Florida courts to have her feeding tube removed and her parents objected. Multiple neurologic examinations repeatedly confirmed her VS/UWS, but politicians at local, state, and federal levels tried to intervene. Ultimately, a court allowed the feeding tube to be removed in 2005. Ms. Schiavo’s autopsy revealed severe and widespread brain damage with atrophy, findings that precluded the possibility of consciousness or awareness.

Case 5: Terry Wallis. Mr. Wallis was 20 years old in 1984 when he was involved in a motor vehicle accident. He was comatose for approximately 3 weeks and then in VS/UWS. He was hospitalized for 4 months and then transferred to a nursing home. At some point thereafter he progressed to the MCS, but because MCS was not defined until 2002 he was always considered vegetative. In 2003, 19 years after his accident, he began to speak and made further progress since then, consistent with eMCS, although he has difficulty comprehending his age and the amount of time that has passed (64). Two MR DTI studies 18 months apart showed possible axonal regrowth with sprouting and pruning; the results were interpreted as demonstrating new neuronal connections (182; 190; 64).

Biological basis

Etiology and pathogenesis

Multiple etiologies may lead to disorders of consciousness; these include acute and traumatic, nontraumatic, degenerative or metabolic, and congenital abnormalities. The end result is severe neuronal loss of the cerebral hemispheres with sparing of at least the lower brainstem. Postmortem examinations of VS/UWS have demonstrated the loss of forebrain and neocortical structures with an intact brainstem (138). In some traumatic cases, hypoxic brain damage can be secondary and can contribute to the condition. Most nontraumatic cases involve hypoxia after cardiac arrest, drowning, or strangulation (91). Metabolic disorders, such as hypoglycemia, also can result in VS/UWS and MSC.

Crucial to the development of disorders of consciousness is a “disconnection” between the thalamus and striatum and widespread cortical structures, particularly with “downregulation” of the anterior forebrain, according to the mesocircuit model (157). Anatomical studies with diffusion tensor imaging tractography, of patients with disorders of consciousness, including VS/UWS, MCS-, MCS+, and eMCS, demonstrated corticocortical and thalamocortical structural damage, especially affecting the posterior region (posterior cingulate cortex, precuneus, or temporoparietal junction) connections to the thalamus, and the degree of structural connectivity correlated with the severity of the disorder of consciousness as measured by the standardized behavioral assessment of awareness, the CRS-R (61; 158).

The role of the thalamus itself in disorders of consciousness is unsettled. A study of arousal levels in acute stroke patients suggests that lesions confined to the thalamus do not impair arousal; rather, pathways responsible for arousal lie outside the thalamus and extend into the brainstem (89). Another study of direct neurophysiological (microelectrode) recordings from the thalamus found half the number of active neurons in the thalami of two patients in VS/UWS when compared to one patient in the MCS. Additionally, coupling of thalamic neuron discharges with EEG phases (thalami-cortical cross-frequency coupling) was only seen in the MCS patient (119).

Epidemiology

Limited information on the epidemiology of the disorders of consciousness is available, but a meta-analysis quotes the annual incidence of VS/UWS in the United States as approximately 4200 with a prevalence estimated between 5000 and 42,000. The prevalence of MCS was estimated at between 112,000 to 280,000 (74). In a group of nursing home patients, VS/UWS was diagnosed in 3% of the individuals. The mean age of the patients was approximately 65 years (range 19 years to 96 years), and the duration of VS/UWS ranged from 1 year to 16.8 years. Twenty-five percent of the patients were VS/UWS for longer than 5 years (179). However, it is important to note that coma, which is the initial condition leading to VS/UWS and MCS, indeed is very frequent; every year, approximately 2 in 1000 people will enter a coma (98).

Prevention

Although organs vary in their tolerance to loss of blood flow and metabolism, neurons of the central nervous system can survive for only minutes following reduction in blood supply below critical levels. Thus, efforts that eliminate or reduce the duration of ischemia during a cerebral insult may prevent the onset of the disorder of consciousness. Similarly, measures to prevent traumatic brain injury will be needed to decrease the prevalence of traumatic disorders of consciousness.

Differential diagnosis

Associated or underlying disorders

Misdiagnosis of VS/UWS is not uncommon (approximately 40%) (74), and unfortunately may severely affect patient care. Some individuals initially described as having VS/UWS do not meet the strict diagnostic criteria of this syndrome. Patients have been found to be interactive with their environment (179), to possess alternative diagnoses such parkinsonism (121), be in coma rather than in VS/UWS (104), or to be in MCS (72).

One factor that appears to result in misdiagnosis is the confusion in terminology used to describe vegetative patients. The terms "apallic syndrome," "akinetic mutism," "coma vigil," "alpha coma," and "neocortical death" should not be used to depict individuals in VS/UWS. In addition, the neurologic conditions of coma, "locked-in" syndrome, and brain death clearly distinct from VS/UWS.

Coma is a state of unresponsiveness in which the patient remains with eyes closed and is unarousable (138). The patient is not conscious of self or the environment. Blood pressure and respiratory function may be intact, but the patient lacks sleep-wake cycles. The presence of flaccidity and areflexia indicate severe brainstem depression, and this is frequently found in terminal coma or in severe sedative intoxication.

Unlike individuals in VS/UWS, locked-in syndrome patients are conscious, but lack the ability to communicate with their environment other than using rudimentary (usually vertical) eye movements (138). Loss of motor control of their facial musculature and extremities prevents these individuals from speaking and interacting with their surroundings, other than by blinking or by using (vertical) eye movements or sophisticated brain-computer interfaces. At times, even ocular movements may be lost. Usually, these patients are not aphasic and are able to comprehend language. In contrast, patients in VS/UWS are not conscious and cannot communicate with their surroundings.

Brain death is declared when there is irreversible loss of brain function, thus, brain death is death. People in VS/UWS are evidently alive (197).

Diagnostic workup

Behavioral measures. There is no “gold standard” to diagnose disorders of consciousness nor differentiate between them. Furthermore, in clinical practice, all disorders of consciousness are currently diagnosed on behavioral bases; the newer neuroimaging and electrophysiological techniques are still considered research modalities, although a few highly specialized centers use them to diagnose patients with cognitive-motor dissociation (103).

Behaviorally, standardized assessments of patients in disorders of consciousness have been in use for the past 3 decades to provide evidenced-based approaches to accurate diagnosis and as means to evaluate clinical course and therapeutic efficacy (166). The most widely accepted and validated behavioral measure of patient disorders of consciousness status is the JFK Coma Recovery Scale-Revised (CRS-R), which is a 24-point, semi-quantitative score of 6 clinical parameters: auditory, visual, motor, oromotor or verbal, and communicative abilities as well as arousal level (73). Standardized assessments are superior to clinical expert consensus, revealing conscious awareness in 30% to 40% of patients with an (incorrect) diagnosis of VS/UWS (163; 166; 74).

It has been shown that a score of greater than 10 on the CRS-R demonstrates consciousness (MCS or better) but misdiagnosed 22% of patients who were thought to be unconscious (VS/UWS) and actually were in the MCS as indicated by other behavioral measures. A cutoff CRS-R score of 8 yielded a diagnostic accuracy of 93%. Thus, a score of 10 or more will ensure the presence of MCS or better, whereas a score of 8 offers the best chance of avoiding false-positive and false-negative errors (18). However, an analysis of CRS-R subset scores in a large inpatient population (n 1190) revealed that 4.7% of subset scores yielded impossible combinations and 12.2% combinations were improbable (38). Additionally, in 282 patients diagnosed in MCS based on the CRS-R, certain responses were frequently seen (visual fixation and pursuit, and reproducible movements to command), whereas other responses were rarely or never seen alone (reaching for objects, object manipulation or recognition, verbalization) (183). The authors concluded that limiting the CRS-R examination to five items (fixation, visual pursuit, reproducible movements to command, automatic motor responses, and localization of noxious stimuli) detected 99% of patients in MCS. Assessing the presumed high degree of interdependence between level of consciousness and pain perception is complicated by the significant overlap between motor and verbal/oromotor CRS-R subscales and similar motor response scores to noxious stimuli in the NCS-R (39).

It has been determined that the CRS-R may miss subtle motor behaviors that can be captured with another measure, the Motor Behavioral Tool-Revised (MBT-r). In a series of patients with severe brain injury studied within 28 days of injury, 75% of those meeting the CRS-R criteria for coma or VS/UWS demonstrated some residual cognition on the MBT-r, 66.7% of whom had a favorable outcome. It is suggested that the MBT-r be used to complement the CRS-R (135). Furthermore, because the CRS-R is relatively time-consuming to administer (30-45 minutes), alternative scales that are easier to administer are being developed, eg, the Simplified Evaluation of CONsciousness Disorders (SECONDs) (155) and the CRSR-FAST (Coma Recovery Scale-Revised for Accelerated Standardized Testing) (20).

The recognition that some patients with disorders of consciousness may have some degree of consciousness, which is not behaviorally demonstrable (ie, cognitive-motor dissociation) because of motor, executive function, or sensory impairments, has propelled multiple empirical attempts to develop neural correlates (proxies, surrogates) of consciousness with neuroimaging or electrophysiologic methodologies (43). One promising measure is the Perturbational Complexity Index (PCI), which quantifies EEG responses to transcranial magnetic stimulation. The PCI assumes that consciousness requires activation of modules and networks connecting thalamocortical structures to produce complex cortical integration patterns. Casarotto and colleagues have compared healthy subjects and conscious brain-injured patients with patients who were in VS/UWS and MCS (31). The investigators were able to discriminate with 100% sensitivity and specificity between conscious and unconscious states (during NREM sleep and general anesthesia) in the control population and a 94.7% sensitivity in detecting consciousness in patients with MCS. They also determined that nine of 43 (21%) patients in VS/UWS had PCI levels in the consciousness range (as noted in, and comparable to, the control population) and may have a capacity for consciousness that is not expressed behaviorally. This raises the possibility that there are minimal anatomical and functional thalamocortical connectivity requirements for consciousness; the higher the PCI level, particularly above the cutoff in this study, the greater the potential presence of consciousness, irrespective of behavioral evidence. The authors suggest that these findings may have practical implications for prognosis and patient management (31); higher PCI scores may mandate more aggressive interventional techniques to increase behavioral output. The PCI was found to be correlated to structural integrity as measured with fractional anisotropy on diffusion-weighted MRI (17) and with metabolic activity as measured with FDG-PET (16).

Radiographic studies. CT or MRI can assist in determining the etiology of the disorder of consciousness and the extent of damage. These studies can determine the presence of an infarct, intracerebral hemorrhage, mass lesion involving the cortex or brainstem, or diffuse or multifocal axonal injuries as in trauma or cerebral anoxia. MRI brain lesions have certain patterns detected in anoxia-induced VS/UWS; predominant lesion patterns include white matter signals within the frontal and occipital lobes in the periventricular regions (05). With chronicity, voxel-based morphometry MRI scans revealed widespread structural brain injury. There was more injury and subsequent atrophy in the brainstem, thalamus, hypothalamus, basal forebrain, cerebellum, and posterior corpus callosum in patients with traumatic versus non-traumatic causes, eg, anoxia (84).

Lutkenhjoff and colleagues studied a large population of patients with disorders of consciousness with structural T1-weighted MRI and found significant widespread atrophy in subcortical brain regions (115). They were able to correlate basal forebrain atrophy to patient response to sensory stimulation and thalamic and basal ganglia atrophy to the CRS-R, consistent with the mesocircuit theory that associates thalamocortical projections to the prefrontal cortex with sustained organized behavior (157). Greater atrophy of the lateral and medial sections of the left thalamus were seen in patients with nontraumatic etiologies as compared to those with traumatic brain injury. Arousal, as measured by the CRS-R, was inversely correlated with the degree of bilateral atrophy of the basal ganglia. Interestingly, no significant differences in subcortical structural pathology were found between VS/UWS and MCS- and MCS+ patients, perhaps due to extensive within-category variance. But patients in VS/UWS had significant atrophy of the left putamen and globus pallidus, as well as small sections of the right hippocampus when compared to patients in MCS. The authors thought that their finding of greater left-lateralized atrophy might affect linguistic processing in patients with disorders of consciousness and suggest the development of nonlanguage-based procedures for assessing consciousness (115).

The role of the prefrontal cortex itself in consciousness is suspect. Considerable evidence from surgical procedures, such as bilateral prefrontal lobectomy and lobotomy and large bilateral prefrontal resections and leucotomy, which severed the prefrontal cortices from their thalamic connections did not impair consciousness (97). It may be that the connections between the prefrontal cortex or parietal lobe structures are more necessary than the prefrontal cortex itself.

PET scanning can demonstrate intact auditory pathways (110) and somatosensory pathways (111) that remain physiologically severed from higher cortical function. PET and fMRI scans can predict recovery; patients with atypical activation patterns involving a spread to higher-level associative cortices had a better chance of recovery from vegetative state (48). fMRI can be used to evaluate higher cortical functions in unresponsive patients, and it can reveal robust responses to visual, auditory, and tactile stimulation (53). Speech stimulation showed responses in the Broca and Wernicke areas. Gray matter to white matter density ratios on a CT scan also can be predictive after a cardiac arrest; a gray matter to white matter ratio of less than 1.22 predicted VS/UWS or death with a sensitivity of 63% and a specificity of 100% (40). Increased cerebral blood flow patterns in MCS patients as measured with arterial spin labeling (192) or functional transcranial Doppler (128) may help differentiate them from VS/UWS patients.

Electrophysiological studies. In patients with disorders of consciousness, EEG is of great value in assessing cortical function or dysfunction by recording spontaneous electrical activity, identifying the presence of occult seizure activity, evaluating possible cognitive functions and responses indicative of consciousness, and perhaps providing a means of communication (77; 55). EEG has several advantages over neuroimaging technologies: temporal resolution of activity in the millisecond range as opposed to the lower temporal resolution of fMRI and PET in the seconds range; portability; availability; limited expense (82), and almost no contraindications; it is a marker of corticothalamic integrity (68) and a direct measure of the behavior of large-scale neuronal networks (63). On the other hand, EEG’s spatial resolution is more limited, particularly with signal depth from the scalp and cortex.

In general, an increased amount of slow wave (delta) activity was associated with severity of disorders of consciousness, eg, more delta activity in VS/UWS than MCS; conversely alpha and theta activity was significantly less present in VS/UWS than patients with MCS (62; 150). Alpha activity was mainly focused in the parieto-occipital region in patients in the minimally conscious state whereas when present in patients in VS/UWS, it was more uniformly distributed, in support of posterior brain activity distinguishing MCS from patients in VS/UWS (62; 150). Persistent alpha and beta rhythms were associated with greater levels of consciousness and better prognoses (63). The presence of normal or near-normal EEG background activity and sleep spindles increases the possibility of covert cognition in patients with disorders of consciousness (68). EEG background activity correlated with overall mean brain metabolic activity as measured by 18FDG-PET (68).

There were greater fluctuations in spectral activity (which measures the strength of neuronal oscillations) in patients with MCS, in accordance with the behavioral fluctuations that define this condition. These findings suggest that: “a stable state of increased delta and reduced alpha-theta power is a solid sign of unconsciousness” (168). Other measures “of signal complexity (spectral entropy, permutation entropy, algorithmic complexity) discriminated VS/UWS from MCS patients [with] both the average and the intertribal stability of EEG complexity increas[ing] monotonically with the patients’ state of consciousness” (168). Unfortunately, although these findings clearly differentiate VS/UWS from MCS at the group level, they cannot be used for individual patients; for example, 24% of patients with MCS were classified as VS/UWS, and 33% of patients in VS/UWS were classified as minimally conscious electrophysiologically (168). This may indicate that, per fMRI, VS/UWS is not a homogeneous category, and some patients in VS/UWS may have electrophysiological evidence of consciousness. In fact, more patients in VS/UWS classified as MCS based on their EEG activity later showed signs of consciousness than those classified as VS/UWS on EEG (p=0.025). The authors suggest that rather than 30-minute recordings, 24-hour recordings might be more helpful (168). Bai and colleagues have written a more complete review of resting-state EEG analyses in disorders of consciousness (09).

Single pulse transcranial magnetic stimulation when combined with high-density EEG was employed to study patients with disorders of consciousness. Patients in VS/UWS showed simple local responses, whereas in MCS patients, complex activations were seen ipsi- and contralateral to the site of stimulation. These differences were seen early in the course of the disorder of consciousness and may offer a modality to detect and track recovery of consciousness in nonbehavioral patients (149). A similar study found different synchronicities for slow and fast frequency bands after transcranial magnetic stimulation delivered over the primary motor area in normal subjects versus patients in VS/UWS and MCS (69). Patients in vegetative state demonstrated cortical circuit silence (OFF-period) on EEG during transcranial magnetic stimulation, a finding never seen in healthy awake individuals (other than during normal sleep) (148).

EEG studies of fronto-parietal gamma synchrony following sensory stimulation in patients with disorders of consciousness revealed short-range parietal and long-range fronto-parietal gamma frequency coherences in normal patients and those in the MCS but not in patients in VS/UWS. This suggests a lack of connectivity in fronto-parietal network circuitry and, by inference, a lack of information integration in patients in VS/UWS (33). However, there is other evidence that gamma synchrony can occur in the absence of consciousness (97). Using resting-state EEG recordings, “network-based statistical analysis reveals a subnetwork of decreased functional connectivity in UWS compared to MCS patients, mainly involving the interhemispheric fronto-parietal connectivity patterns” (28).

Waveland subband entropy was found to be a useful tool for predicting poor outcome in patients in VS/UWS. The median wavelet subband entropy of the poor outcome group was significantly lower, thus, making it a predictor of poor outcome (58).

Brain-evoked potentials to external stimuli (auditory, visual, somatosensory) can be used to determine voluntary brain processing and have been proposed as signatures of consciousness (97). Patients in MCS, but not patients in VS/UWS had P3 potentials when hearing their name (162). A subsequent study revealed P3a (bottom-up attention orienting potentials) in patients in MCS but not in patients in VS/UWS. No patient in either group showed P3b (top-down) event-related potentials (79). Some patients in these impaired consciousness states show better event-related potential responses to self-relevant stimuli, personalized objects (172), or their own names (95) than non-self-relevant stimuli. Other studies have questioned whether evoked potentials are sufficient markers of consciousness as many clearly comatose and sedated patients have them (180; 97).

Polysomnographic recordings may be helpful in evaluating patients with disorders of consciousness. Landsness and colleagues found no slow-wave sleep or rapid eye movement sleep stages in patients in VS/UWS whereas there were near-to-normal sleep patterns in the patients with MCS (109). However, these findings were not confirmed in subsequent studies: circadian EEG changes were seen in all of 55 patients in VS/UWS (19 had periodic amplitude reductions during sleep when compared to waking periods) and in all 36 patients with MCS (151; 106); the better the clinical condition, the more sleep spindles were present and the longer the episodes of distinct non-REM sleep. In all but 1 of 15 patients in the VS/UWS, sleep patterns were present but disturbed (133). Pediatric patients in VS/UWS had power increases in the delta frequency band and decreased alpha, sigma, and beta activities during sleep (122). These authors support the concept that the slower the EEG activity during sleep and the narrower the power modulation between sleep and the waking state, the poorer the clinical outcome (122). In 44 patients in VS/UWS, MCS, and eMCS, polysomnography revealed vertex waves in 89%, sleep spindles in 77%, and slow-wave sleep activity in 54% (68). In a larger study of 142 patients (85 VS/UWS, 57 MCS) 18-hour recordings showed sleep activity in 85% (150). Wislowska and colleagues found entropy and oscillatory differences during polysomnography at consciousness group levels, but this was not helpful in distinguishing VS/UWS from patients with MCS on an individual level (191).

A preliminary study of the use of electromyography found electrical muscle activity in all MCS, eMCS, and locked-in syndrome patients but not in any of 18 VS/UWS patients (113).

In view of the lack of concordance of single or several electrophysiologic markers in disorders of consciousness, Sergent and colleagues conducted a preliminary study of eight different parameters in a single 1.5-hour session and found the combined protocol was able to distinguish patients in VS/UWS state from those in MCS at the single-subject level (167).

Multimodal studies. Multimodal studies, combining fMRI and EEG, are increasingly done in both chronic and acute settings, including the intensive care unit (03; 04). Machine learning algorithms are being tested to predict recovery of consciousness in patients with acute brain injury with improved positive predictive value and sensitivity (03; 04).

Management

Clinical management of patients with disorders of consciousness is problematic as the diagnosis is based on behavior, but the underlying neural damage may be relatively focal or widely generalized, and no one treatment is likely to benefit all such patients (185; 56). With that in mind, there are two main goals: (1) prevention of secondary medical complications and (2) restoration of cognitive-behavioral functions (77; 56). The medical complications of most concern are contractures, spasticity, urinary tract and other infections, agitation, sleep disturbances, hyperkinesia, deep vein thrombosis, and pneumonia. Multiple prophylactic treatments are necessary for management of these issues (77; 185).

There are no approved therapies for restoration of cognitive and behavioral functions, although individual case reports and small case series suggest some possibilities. Foremost are aggressive physical therapy treatments. Fins and others have argued that, given the commonality of late progress in disorders of consciousness, even after 2 or more years, current relatively short-term physical therapy protocols, some dictated by medical insurance companies and programs (eg, the Milliman Care Guidelines, McKesson’s InterQual criteria), are inadequate and too often force the end of active physical therapies in favor of custodial care (186; 77; 64). The early transfer of patients out of hospital has the benefit of reducing hospital complications but runs the risk of foreclosing on aggressive therapies before maximal improvements are reached.

Pharmacological agents that have been studied include amantadine, a 4-week course of which seemed to accelerate the pace of functional recovery in patients with VS/UWS and MCS, but once stopped no differences were seen between treatment and placebo groups (78). Nonetheless, guidelines suggest that amantadine at 100 to 200 mg BID when given to traumatic VS/UWS or MCS patients aged 16 to 65 for 4 weeks between 4 and 16 weeks postinjury probably hastens functional recovery (75; 74). Zolpidem was not found beneficial in patients with disorders of consciousness (185; 21; 129). Intrathecal baclofen has proven beneficial (173; 136).

Anodal transcranial direct current stimulation to the left DLPFC for 20 minutes transiently improved consciousness in 13 of 30 patients (43%) with MCS but not in 25 patients in VS/UWS (176). Repetitive transcranial magnetic stimulation (rTMS) (10 Hz X 10 seconds repeated 10 times--1000 pulses QD X 20 days) improved CRS-R scores in five of five patients with MCS and four of 11 patients in VS/UWS (193). However, other studies with other rates of stimulation of other brain areas, eg, motor cortex in patients in VS/UWS, have been negative (41). The varied rTMS protocols limit interpretation and recommendations.

Deep-brain stimulation was found beneficial in a patient who was in the MCS for 6 years (159). A 5-month study of bilateral deep-brain electrical stimulation of the central thalamus improved cognition, limb control, and oral feeding function during deep-brain stimulation as compared to stimulation-off periods (159). Another study of bilateral chronic stimulation (18 to 48 months) of the anterior intralaminar thalamic nuclei and adjacent paralaminar regions in two patients in VS/UWS and one patient in MCS revealed improved EEG and CRS-R scores and reductions in limb spasticity and pathological movements, but no patient returned to full consciousness (118). A review of thalamic deep brain stimulation methodology discussed the many challenges to this treatment, including informed consent (76). Cervical spinal cord stimulation in 12 VS/UWS patients resulted in significant improvement in average CRS-R score (6.25 to 10.8) after an average follow-up period of 11.1 months, with three patients recovering to eMCS or better and two to MCS (195).

Management of the patient in VS/UWS can raise several ethical concerns. The decision to assume an aggressive course or to gradually terminate care must be reached after careful consideration of the patient's prior documented wishes. If the patient has not provided an advance directive, then the course for further management should be decided between the patient's physician and health care surrogate (guardian, spouse, children, family members). Decisions for care should focus on the utility of future treatment modalities, quality of life, and possible resource constraints. It is argued that a legacy of the right-to-die movement “influences practice patterns, viewing patients with severe brain injury through an end-of-life prism, leaving them marginalized and sequestered from the evolving fruits of neuroscience” (77).

Certain studies are trying to detect signs of learning in patients in VS/UWS; this involves participants using an electronically regulated optic microswitch mounted on eyeglasses that allowed experimenters to monitor levels of response to different phases of stimuli. The results showed that there was an increased response level during intervention periods with stimuli dependent on responses, thus, allowing physicians to better classify patients’ ability despite their being in VS/UWS (108).

Music therapy can be used to rate behavioral responses across multiple domains, can provide information for interdisciplinary assessment, and has been found to have therapeutic value in patients with disorders of consciousness and other severe neurologic disorders (116; 117).

Working on the assumption that patients with these disorders of consciousness can willfully modulate their brain activities despite no or minimal behavioral evidence of such, nascent attempts to communicate with them via brain-computer interfaces have begun (55; 56). Three separate technologies can be employed: fMRI; functional near-infrared spectroscopy (fNIRS); and EEG. Each has advantages and disadvantages, and it is anticipated that not all patients will be able to use such systems despite showing evidence of consciousness with other methods (124).

For comprehensive reviews on emerging treatment option, please see the articles written by Thibaut and colleagues and Kondziella and colleagues (178; 99).

Special considerations

Pregnancy

In rare instances, pregnancy can be complicated by prolonged coma or VS/UWS. In a report, a pregnant woman suffered massive head injuries in an automobile accident at 6 weeks' gestation, but she successfully carried the fetus an additional 7 months. The patient remained comatose during the gestation and was aggressively supported (153). Other reports describe spontaneous vaginal deliveries in patients aggressively supported during VS/UWS (08). Some studies have shown that the mean latency between maternal brain injury and delivery was shorter in patients diagnosed with brain death than with those individuals in VS/UWS (27). The gestational ages at delivery were shorter and the birth weights were smaller in brain-dead mothers than in VS/UWS mothers (27). It is unclear whether intensive-care support of a brain-dead mother or VS/UWS mother can produce a healthy newborn.

Neural correlates of consciousness. Much of the interpretations of neuroimaging and electrophysiological findings rely on models of cerebral structure, functions, and activities that relate to neural networks and their intrinsic connections. The concept of neural networks has been around since the 19th century, but with the advent of functional scanning, PET and fMRI, and electrophysiological studies, at least 10 resting-state functional networks have been identified so far (143). Perhaps the most important network, from the disorders of consciousness perspective, is the default mode network, which includes the posterior cingulate cortex or precuneus, anterior medial prefrontal cortex and left and right temporo-parietal junctions, and connections to the thalamus. These subregions are functionally and anatomically connected and active during the awake resting state; they represent baseline stimulus-independent neural activity (181; 61; 88), which Havlik calls “endogenous.” Qin and colleagues have also found the salience network, especially the connections between the supragenual anterior cingulate cortex and the left anterior insula, important in assessing disorders of consciousness (141). Additionally, other authors subdivide the brain’s consciousness activities into two networks: a medial “internal or intrinsic” network similar to the default mode network, responsible for awareness of self, thoughts, and mental imagery, and a lateral “external or extrinsic” frontoparietal control network, which includes the left dorslolateral prefrontal cortex (DLPFC) responsible for awareness of the environment during attention-demanding, goal-oriented cognitive activities (181; 32). For extensive reviews of the role and evaluation of functional networks and functional connectivity in disorders of consciousness please see works by Bodien and colleagues (19) and Amico and colleagues (02).

Recommendations about the use of fMRI, PET, and EEG in the diagnosis of disorders of consciousness have been put forward by the European and American Academies of Neurology (74; 100).

Neuroimaging studies. Two neuroimaging technologies, PET and fMRI, are most frequently used to assess three different cerebral functional states in disorders of consciousness: resting wakefulness, passive response to stimuli, and active involvement in a cognitive task. Most PET scanning studies measure resting state brain glucose metabolism with 18F-Fluorodeoxyglucose (FDG). fMRI scanning measures the blood oxygenation level dependent (BOLD) signal. Changes in neuronal activity result in localized changes in FDG uptake (PET) or oxygen utilization (fMRI). In normal awake consciousness there is a great deal of energy-utilizing neural activity, spontaneous low-frequency fluctuations in the range of 0.01 to 0.1 Hz (181), distributed amongst various neural networks. With decreases of consciousness due to sleep, anesthesia, or brain disorders these neural activities decrease as demonstrated in decreases in FDG (PET) or BOLD (fMRI) signaling. Alternatively, localized or distributed increases in neuronal activity during cognitive tasks are associated with increases in the FDG (PET) or BOLD (fMRI) signals.

PET. FDG-PET studies show markedly reduced resting state brain glucose metabolism in chronic VS/UWS (30% to 40% of normal) and greater glucose metabolism in MCS (50% to 70% of normal), with maintenance of consciousness at levels of less than 45% rare (170; 171; 82); the metabolic index of 3.19 (42%) was considered an optimal cutoff (171).

These metabolic changes seen on FDG-PET affect multiple networks but especially the default mode network, which is thought essential for consciousness (06). Annen and colleagues also found correlations between default mode network metabolism and its white matter integrity on MRI-DWI (06). PET studies may have prognostic value; preservation of frontoparietal glucose metabolism was associated with imminent recovery of consciousness, whereas complete loss of metabolic activity in frontoparietal areas was not associated with long-term recovery (169). When FDG-PET was used to study patients in the MCS, those in MCS+ showed higher cerebral metabolism in left-sided cortical areas encompassing the language network and premotor, presupplementary motor, and sensorimotor cortices with greater functional connections to Broca area than seen in MCS patients (25). When comparing patients in VS/UWS, MCS, and eMCS, brain metabolism increased in frontal, parietal, and occipital midline regions, the thalami, basal ganglia, and the lateral front-parietal parieto-occipital, and temporal regions--more so on the left at each level of consciousness (50).

fMRI. In the resting state, Qin and colleagues, using fMRI, found reduced connectivity within the salience network, particularly between the supragenual anterior cingulate cortex and left anterior insula, in patients in VS/UWS but not in patients with MCS or with other brain lesions and full consciousness (141). Thus, the authors thought salience network connectivity could distinguish unconscious from conscious patients. Default mode network connectivity was reduced in the patients in VS/UWS, more so in the patients who remained in that state as opposed to those who emerged from it after 3 or more months and, thus, may have predictive value regarding recovery of consciousness, although, as a stand-alone finding, this has been questioned (50). Salience network connectivity did not show a predictive effect. A review and meta-analysis confirmed the reduced resting state activity in default mode network structures, more so in patients in VS/UWS compared to those in the MCS; the reduced activity was greatest in the thalamus and posterior cingulate cortex or precuneus (85).

Several studies show that in acutely comatose patients studied with resting state fMRI, having an intact default mode network is a good prognostic sign; those patients with a preserved default mode network recovered consciousness at follow-up, whereas those whose default mode network was disrupted acutely did not regain consciousness subsequently (130; 165; 102). Silva and colleagues found functional connectivity disruptions from the posterior cingulate cortex, especially with the mPFC, and the strength of this connectivity in the acute stage had predictive value in regard to recovery (165). Dynamic changes in the default mode network coactivation patterns suggesting impaired spatial connectivity were seen in vegetative state patients (49). Resting state activity was also able to differentiate patients in VS/UWS from those with locked-in syndrome and controls; the default mode network was observed in every control and patient with locked-in syndrome but not in the patients in VS/UWS (147). Widespread (global) connectivity when measured during rest was correlated with the CRS-R total score and arousal subscale score (01).

Passive paradigms assess covert narrative capacity, not requiring the patient to perform a task but recording brain activation during passive stimulation and comparing these activities with those seen in normal persons under identical circumstances, eg, watching a movie during fMRI scanning (86). An 8-minute Alfred Hitchcock suspense movie was employed to assess frontoparietal executive functions; highly synchronized activity fluctuations that tracked common cognitive experiences were seen in normal subjects while watching the movie. In two patients in VS/UWS exposed to the same movie, one showed auditory cortex synchronization without visual or executive function synchrony, whereas in the second patient, who had been behaviorally unresponsive for 16 years, not only was there auditory synchronization, but also visual, and most significantly, frontal and parietal synchronization in regions known to support executive processing, similar to that seen in normal subjects (125). The authors hypothesize that “the degree to which each individual’s frontoparietal brain activity could be predicted from the rest of the group’s represented a reliable neural index of how similar his/her cognitive experience was to the others’ … these results demonstrate that similar conscious experiences in different individuals are supported by a common neural code … this approach interprets brain activity and concomitant mental states without recourse to behavior … as in the case of behaviorally unresponsive patients” (125).

Active paradigms are “command-following paradigms,” for example, respond to auditory stimuli (127), or “pretend you are playing tennis” and “pretend you are navigating the rooms of your house” and do so for 30 second intervals (132; 123; 131). The best and most tested--motor imagery (tennis/house) “employ a highly artificial experimental context and as such can offer only a highly circumscribed picture of a patient’s mental life … these tasks can only probe cognition within the purview of the particular task the patient performs [which] are very narrow when compared to the open-ended nature of conscious experience in day-to-day life…. For this reason, the mental imagery paradigm cannot address questions about the extent to which behaviorally nonresponsive patients …. may consciously process the dynamically evolving sensory environment around them.” Can these patients sustain attention “while filtering out irrelevant distractors … [can] any such patient continuously integrate the inputs from various sensory modalities into a unified whole … so as to understand complex, real-life events unfolding over time”? (126).

Unfortunately, the false-negative rate is high. In the seminal fMRI study of active paradigms in VS/UWS and MCS, only 1 of 31 MCS patients exhibited positive fMRI findings. It is obviously possible that the fMRI studies occurred when only one person (3%) was responsive to the external environment, given the intermittent nature of consciousness in MCS. However, because four of the 23 patients (17%) in VS/UWS had positive responses, this possibility presupposes that even with intermittent consciousness in the MCS, the patients in VS/UWS were greater than five times more likely to be conscious at the time of testing, despite the fact that they never demonstrated behavioral evidence of consciousness at any time under prior observation. This is a long shot. Even if we accept that two of the patients in VS/UWS were recategorized as being in the MCS because post-fMRI they were found to have some behavioral evidence of awareness, this still means that only 9% of patients in either group (two of 21 VS/UWS, three of 33 MCS) demonstrated fMRI evidence of consciousness.

In a comparison of active and passive paradigms using fMRI and EEG, patients in MCS were more likely than patients in VS/UWS to follow commands during active paradigms (32% vs. 14%) and to show preserved functional connectivity during passive paradigms (55% vs. 26%). Passive paradigms were more likely to show consciousness in both disorders of consciousness states. Resting state paradigms were unable to be evaluated. The authors conclude that active paradigms may underestimate the presence of consciousness as compared to passive paradigms. There are, however, inconsistencies at the single patient level, eg, patients from both categories may show command-following despite no responses during passive paradigms (103).

PET versus fMRI. In a head-to-head comparison of FDG-PET and active paradigm fMRI, PET scanning was found to be the more sensitive technology in identifying patients in minimally conscious state and predicting outcome in both VS/UWS and MCS, when the CRS-R scores were used as a behavioral standard. Thirteen of 41 patients (42%) diagnosed with VS/UWS were found to have positive results on one or the other of the scanning technologies, nine of whom progressed to the MCS over the following year (three died and one remained in the VS/UWS). One of the technical problems with fMRI was that spontaneous movements and large motion artifacts made results impossible to interpret in almost half of the patients studied. Of note is that the referring clinical diagnosis failed to identify responsiveness in 33% of patients when compared to the CRS-R. The authors are well aware that the CRS-R is not a gold standard; therefore, comparisons between CRS-R and the imaging techniques can be problematic (169).

Electrophysiological studies. Transcranial direct current stimulation (tDCS) of the DLPFC, administered in a single session, has been used to study resting-state fMRI activity. Approximately 40% of patients in MCS demonstrated increased connectivity of the DLPFC with the inferior frontal gyrus, relative gray matter preservation, and residual brain metabolic activity on FDG-PET, associated with a transitory recovery of consciousness (176; 177; 32). It is hypothesized that such studies may detect neuronal conditions necessary to improve behavior in the MCS (32). Noninvasive brain-computer interface programs utilizing vitro-tactile stimulators to measure P300 responsivity was able to determine command following in 2 of 12 vegetative state patients (83).

Autonomic responses. A “Central Autonomic Network (CAN) has been proposed as an integrative model where neural structures and heart function are involved and functionally linked [together in] affective, cognitive and autonomic regulation”; brainstem, limbic, and prefrontal cortex structures are involved via feed-forward and feedback loops (145). Evaluating heart rate variability (47; 145; 144) and total peripheral resistance (47) in sedated VS/UWS and MCS patients (as assessed with the CRS-R) in the resting state (145) and during noxious stimuli (47; 144), MCS patients demonstrated greater heart rate variability (47; 145; 144) and total peripheral resistance (47) than VS/UWS patients. These results correlate with the CRS-R scores (145; 144) and fMRI findings (145). The authors suggest that these measures are easy, fast, and inexpensive and can differentiate vegetative state and minimally conscious state patients.

Combination studies. Three studies of brain structure, one utilizing 7 tesla MRI (174) and function, one utilizing resting state fMRI (46), and high-density EEG and other measures (120) in normal subjects and patients with disturbances of consciousness found widespread coordinated network neuronal activity and normal white matter connectivity in conscious subjects in support of the Global Neuronal Workspace theory. These studies also show that patients in VS/UWS had global network disruption with significant reductions in widespread white matter connectivity and long-distance interareal coordination as compared to normal subjects (46). There was also enhancement of modular/clustered/focal networks that decreased with clinical improvement (174).

Those patients who meet the clinical/behavioral criteria for VS/UWS or MCS, but who, with complementary methodologies, eg, neuroimaging, electrophysiology, exhibit some command following, are thought to have cognitive-motor dissociation (156; 53; 87; 43; 135) and may carry a more favorable prognosis (53). Braiman and colleagues raise the possibility that cognitive-motor dissociation can be demonstrated passively utilizing high-density EEG recordings of responses to self-referential spoken language (the natural sleep envelope). The patients with normal EEG responses showed positive findings on fMRI whereas those with abnormal EEG responses had no evidence of covert consciousness on fMRI (22). Although it has been suggested that normal EEG responses to spoken language in behaviorally vegetative state patients may detect covert cognition, it is not volitional command following, the current standard for determining consciousness (52).

Ethical considerations. Disorders of consciousness, especially in view of neuroimaging and neurophysiological research findings, raise complicated and vexing ethical issues (199; 198):

	(1) Determination of the accuracy of the diagnosis of a specific disorder of consciousness is essential. It has been noted that many hospitals do not regularly check consciousness levels once the aforementioned time frames have elapsed, thereby potentially missing conscious behavior months or years after the event (60). Misdiagnosis of either disorder of consciousness state, but particularly the failure to find behavioral evidence of consciousness in patients in VS/UWS that would indicate that they are actually (at least) in MCS, is uncomfortably common (approximately 40%) (163). Misdiagnosis affects medical decision-making from several important perspectives: experiencing subjective states such as pain, suffering, or pleasure; communication; beneficence; prognosis; and informed consent (60).
	(2) A correct prognosis is essential for therapy. It is argued that current healthcare practices too tightly limit continuing rehabilitation therapies in patients with disorders of consciousness either because of misdiagnosis or failure to appreciate that the biological potential for continued recovery may occur at a slower rate than current standards recognize (77; 64). These authors also state that families of patients in the acute care setting are too often advised of end-of-life, palliative care, or organ donation alternatives before a patient’s prognosis or diagnosis is clear (77; 64). Of note is the role of VS/UWS (in the case of Karen Quinlan) in advancing right-to-die discussions; now similar patients are being cited to create a right-to-live dialogue (77; 64). Debates on the subjects of right-to-die and right-to-live continue (29). Fins and Wright argue that active rehabilitation of patients with severe brain injuries is a civil right in accordance with the Americans with Disabilities Act and the UN Convention of the Rights of Persons with Disabilities (66).
	(3) A particular disorder of consciousness diagnosis may not be static, and the original timeframes of the Multi-Society Task Force Study may be too short-lived. Detailed analyses of multiple patients in VS/UWS and MCS have demonstrated patterns of improvements that have occurred far later than the 3- to 12-month timeframes of VS/UWS and years later for MCS (eg, Terry Wallis) (64; 131; 74a; 74b).
	(4) Some authors have opined that there is a moral imperative to use fMRI in patients with disorders of consciousness in order to support accurate diagnoses, investigate communication possibilities, and assess decision-making capacities (13; 23). Others have interviewed relatives of patients with disorders of consciousness and found significant reservations about fMRI use in their affected patient-relatives (152).
	(5) It is important for physicians to manage the media as well. With media coverage of Terri Schiavo (a patient in VS/UWS) in America, there were multiple inaccuracies: 21% of the articles reported that Ms. Schiavo “might improve,” and 7% reported she “might recover.” Other inaccuracies included that Terri “responds” (10%), “reacts” (9%), “is incapacitated” (6%), “smiles” (5%), and “laughs” (5%). These claims in the media were not consistent with the diagnosis of a VS/UWS; thus, physicians need strategies to communicate about VS/UWS and MCS with professional communities, patient communities, families, media, and others (142; 101). Similar media misrepresentations of Terry Wallis also occurred (190).