Neurobiological Bases of Coma, Vegetative, and Minimal Conscious States

FIGURE 1-1 Subcortical and cortical connections implicated in arousal (left). These mechanisms are impaired in unconscious patients (right) in relation to ascending reticular activating system (ARAS) or thalamic lesions or deafferentation of these structures and inhibition by globus pallidus interna.

ALTERATIONS OF CONSCIOUSNESS RELATED TO ALTERED CONNECTIVITY BETWEEN CORTICAL AREAS

The impairment of consciousness most often alters tasks requiring complex information processing for high-order cognitive functions, such as writing logic, working memory, and judgment, relative to simple sensory tasks. Conscious effortful tasks and responses to complex stimuli require more cortical processing than do elementary sensory-motor tasks, suggesting the importance of connectivity between areas through the thalamus (via the medial pulvinar essentially) [13]. Over the last few decades, the concept of “global neuronal workspace” has been progressively developed: this concept is based on complex architecture of distributed neurons with long-distance corticocortical connectivity [18]. This model implies that workspace neurons are in dense cortical layers 2 and 3 but are also differentially distributed across the cortical areas, ensuring interconnections and coordination between specialized brain areas. Unconscious processing by these “workspace neurons” distributed throughout the brain and their long-distance connectivity make the incoming information available to a variety of brain processes and would be necessary to access consciousness (see Chapter 2). Neurophysiological, anatomical, and brain-imaging data strongly support a major contribution of the prefrontal cortex (especially anterior cingulate regions) and the parietal cortex to the workspace [18]. Accordingly, the effort engaged in a complex task (as compared with an automatized task) is associated with specific brain activation, as in the dorsolateral prefrontal cortex and anterior cingulate, but declines with routinization of the process [18]. Lesion studies have reinforced this concept and, as an example, linked the prefrontal cortex with decision making and reward-based learning in dynamic environment. Recently, a variation on the “global work space” hypothesis was applied to seizures in terms of alteration of consciousness [2], and it stresses the importance of long-distance corticocortical synchronization between distant brain areas in a perioperative context. Intracerebral EEG patient data provided important information suggesting an excess of synchronization (“oversynchronization”) within corticothalamic networks associated with conscious impairment after temporal and parietal seizures.

Another example of diffuse alteration of connectivity between cortical areas would be anesthesia administration, which commonly involves functional impairment of the thalamus and frontoparietal areas. Most general anesthetics are GABAergic agents that decrease spontaneous neuronal firing in the cerebral cortex and slow cortical activity independently of subcortical structures. Conversely, ketamine induces unconsciousness in concomitance with active enhancement of EEG activity. These differences in the process of reaching unconsciousness suggest that different anesthetics interfere with, rather than inhibit, cortical activity. Moreover, depending on the dose, the sedative–hypnotic effects of different anesthetics may not affect the connections between the thalamus and primary sensory areas but may alter the functional connections between higher-level cortical areas. Positron emission tomography revealed that anesthesia-induced unconsciousness was associated with global reduction of regional cerebral blood flow, particularly with deactivation of the mesial parietal cortex, posterior cingulate cortex, and precuneus (PCC/C). The exploration of emergence from deep anesthesia to recovery of consciousness showed the temporal activation sequence in the subcortical and limbic structures before the frontal area and then the parietal area [28]. On the basis of such data, some authors have proposed the disruption of “top–down” functional connectivity from the frontal area to more posterior sensory areas as a mechanistic approach for the effects of general anesthetics [10]. It might be worth to note that, contrary to anesthesia, natural sleep induces first a thalamic deactivation, followed by cortical deactivation only minutes later [32].

Posterior parietal areas were also deactivated in VS and were the first areas to reactivate with recovery [29]. As a central hub of human brain anatomical connectivity, the integrity of the frontoparietal network may be necessary for consciousness to occur [19]. Interestingly, theses cortical areas are part of the “default mode network” (DMN) of brain function, which encompasses areas that were shown to be more active at rest compared with when the subjects were involved in cognitive tasks demanding attention. The potential major result of using this approach to study the default network in patients is that patient collaboration is not required. Functional magnetic resonance imaging (fMRI) resting state investigations have demonstrated that decreases in functional connectivity in patients, particularly in the posterior cingulate cortex and precuneus (PCC/C), are a function of the level of consciousness [53], which may differentiate MCS from unconscious patients. From a functional viewpoint, most recent investigations have suggested that alterations to the effective connectivity between DMN nodes, particularly precuneus and its inhibitory or excitatory influences, control the process of attention and efficient information integration [14, 30]. The functionality of DMN in MCS suggests the preservation of self-related processes and, among them, perception of pain (Chapter 10). Conversely, in unresponsive/VS patients the internal self-related mentation is disrupted. However, the anticorrelated pattern or switch between external awareness system (related to “global workspace”) and internal awareness system (DMN) seems to be of critical importance for conscious cognition [19].

Among the EEG signals, event-related potentials (ERPs) have been extensively applied to assess information-processing ability, with the interest to be interpretable in absence of explicit behavior from patient [22]. In particular, long-latency components of ERPs (recorded while healthy subjects perform cognitive tasks) have been described as a reliable marker of neuronal conscious perception involving fronto–temporo–parietal cortices and backward connection between these areas. The top–down process or late cortical interconnections might be reflected by the occurrence of a long-latency period as a reliable marker of conscious perception. Disruption of these task-related electrophysiological elements was observed in VS patient in comparison with MCS patients and healthy controls [9]. Some of these components may demonstrate a long-term prognostic value [49]. Those long-latency ERPs result from the synchronized activity of multiple brain regions and reflect the integrity of multiple brain connections. In patients, the persistence of these signals may indicate that these regions and connections are not too badly injured or that there may be a chance of recovery.

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