**9. Sensory stimulation procedure**

194 Neuroimaging – Cognitive and Clinical Neuroscience

procedures for patients with DOC (Neumann & Kotchoubey, 2004). Cognition abilities with theory of mind tasks, decision-making tasks, social performance tests and expanded cognitive assessment, to further characterize post-traumatic or hypoxic-ischemic brain damaged vegetative patients after recovery remain under evaluation at this time. The cognitive recovery in patients with DOC is a continual process rather than a step-by-step phenomenon and confirms that a good recovery assessment should include objective measures of behavioural, cognitive and functional domains, and neurophysiological data to support diagnosis. Survivors from a coma frequently suffer from long-lasting disability, which is mainly related to cognitive deficits. Such deficits include slowed information processing, deficits of learning and memory, of attention, of working memory, and of executive functions, associated with behavioral and personality modifications (Azouvi et al., 2009). An accurate cognitive assessment during the very first phase of the convalescence, when it is possible, is the first step for the management and the implementation of an

Appropriate management requires an experienced inter-disciplinary as opposed to multidisciplinary team working style, whose skill repertoire equips them to recognize oftensubtle improvements in cognitive function and act to maximize individual patient's quality of life. The current paucity of service provision for this vulnerable group of patients is highlighted. In fact, predicting the chances of recovery of consciousness and communication in patients who survive their coma, but transit in a VS or MCS remains a major challenge for their medical caregivers. Very few studies have examined the slow neuronal changes

Determining the accurate prognosis of VS and MCS is a critical step in counseling families and determining appropriate treatment. Previous studies of prognosis in VS were limited by several factors: 1) because there were no accepted diagnostic criteria for MCS prior to 2002, some patients in MCS in those studies may have been diagnosed with VS; 2) it is more accurate to determine prognosis by the etiology of brain damage than merely by categorization in a clinical syndrome; and 3) retrospective experiential analysis of outcomes, such us that by the Multi-Society Task Force, committed the fallacy of the self-fulfilling prophecy because they included patients in their survival data who died primarily because their life-sustaining therapy was discontinued (Bernat et al., 2010). Nevertheless, the prognostic guidelines published in 1994 by Multi-Society Task Force on PVS have been generally accepted, showing a very low probability of recovering awareness once VS has been present for a year following TBI or for 3 months following hypoxic-ischemic neuronal injury (Bernat et al., 2009).Two recently published studies of prognosis in VS add useful data. Luautè and colleagues (Luautè et al., 2010), confirmed the prognostic guidelines of the Multi- Society Task Force in all the patients in VS. They studied and showed that age greater than 39 years and absence of the middle-latency auditory evoked potentials were independent early predictors of poor outcome irrespective of pathogenesis. Estraneo and colleagues (Estraneo et al., 2010), found that 88% of patients in VS in their serious conformed to the Multi- Society Task Force prognostic guidelines but 12% made late recoveries of awareness but only to the point of severe disability with MCS, most of whom had TBI. Because of varying pathophysiologies, prognostic indicators for MCS as a group have been difficult to establish whereas prognostic indicators in individual pathophysiologic subsets of

underlying functional recovery of consciousness from severe chronic brain damage.

individual and effective treatment.

**8. Prognosis and rehabilitation** 

The use of unimodal and multimodal sensory stimulation for the treatment of comatose patient, both in the acute and prolonged states, has been advocated (Johnson et al., 1988). The rationale behind the use of these techniques is that all aspects of the patient must be treated; it is insufficient to attend to the maintenance of bodily well being alone. Sensory stimulation should at the least not have any ill-effects on the patient and could enhance the processes of recovery. S.L. Wilson et al. (1991) have observed patients diagnosed as being in prolonged coma, routinely treated according to a sensory stimulation protocol. They reported an evaluation of the efficacy of this procedure using the comparison of behavioral measures taken immediately prior and post-stimulation. Sensory stimulation treatment appears to be widely used with patients who are in VS arising from traumatic causes, but the term has to be regarded as generic rather than specific since sensory stimulation procedures appear to differ widely in content (Wilson et al., 1993). A number of studies have been published evaluating the effects of these treatment; some have methodological flaws, but the major difficulty in evaluating any treatment with this group of patients is getting sufficient subjects, so most of the published studies use relatively small numbers. Ideally, a large-scale matched control study would be looked for, which examined rate of recovery and long-term outcome. If sensory stimulation is rejected on the basis of lack of empirical evidence, then logically many other treatments used with medical settings should also be rejected. In real life, however, where definitive empirical evidence is not yet available, then clinicians can reasonably make decisions on treatment by combining clinical experience with inferences from scientific knowledge concerning related populations. For example, stimulation treatments which involve the use of some constant background stimulation within the patient's environment, such as TV or radio, have been justifiably criticized. As Wood points out, it is likely to be damaged within the brain that mediate selective attention are highly likely to be damaged within these patients; therefore it is unlikely they are going to be able to differentiate between stimuli in a situation where they are being bombarded with sensory input. In addition, habituation may exacerbate the problem.

Neuroimaging and Outcome Assessment in Vegetative and Minimally Conscious State 197

Patients with severe brain damage who are unable to communicate present several ethical concerns. Foremost is the concern that diagnostic and prognostic accuracy is certain, as treatment decisions typically include the possibility of withdrawal of life-support. Although imaging techniques have the potential to improve both diagnostic and prognostic accuracy, careful and repeated neurological assessment by a trained examiner remains best practice. Accurate clinical assessments of patients in these conditions must be obtained before they undergo neuroimaging. Moreover, in reports of neuroimaging studies, all relevant clinical

Ethical concerns are commonly raised about the participation of patients with severe brain damage in neuroimaging studies. By definition, unconscious or minimally conscious patients cannot give informed consent to participate in clinical research and written approval must typically be obtained from family or legal representatives depending on governmental and hospital guidelines. Nonetheless, researchers studying these patients have been refused grants, ethics committee approval, and research publication; these decisions tend to be made on the basis that studies of patients who cannot provide consent are unethical. We prefer an ethical framework that balances access to research with medical advances alongside protection for defenseless patients. Severe brain damage represents an immense social and economic problem that warrants further research. Unconscious, minimally conscious, and locked-in patients deserve special procedural protections. However, it is important to stress that they are also at risk of being denied therapy that may

Patients who are in coma, VS, MCS, or locked-in syndrome present unique problems for diagnosis, prognosis, treatment, and everyday management. At the patient's bedside, assessment of cognitive function is difficult because voluntary movements may be very small, inconsistent, and easily exhausted. Functional neuroimaging will never replace the clinical assessment of patients with altered states of consciousness. Nevertheless, using population norms it can provide an objective measure of the regional distribution of cerebral activity at rest and under various conditions of stimulation. The quantification of brain activity differentiates patients who sometimes only differ by a brief and small movement of a finger. In our opinion, PET, MRS and fMRI will increase substantially our understanding

Adams, J.H.; Jennett, B.; McLellan, D.R.; Murray, L.S. & Graham D.I. The neuropathology of

Ammermann, H.; Kassube, K.J.; Lotze, M.; Gut, E.; Kaps, M. et al. MRI brain lesion patterns

Arfanakis K.; Haughton V.M.; Carew J.D.; Rogers, B.P.; Dempsey, R.J. et al. Diffusion tensor

Atlas, S.W.; Howard, I.I.R.S.; Maldijian, J.; Alsop, D.; Detre, J.A. et al. Functional magnetic

vol. 260, N° 1-2, (September 2007), pp. 65-70, ISSN 0022-510X

the vegetative state after head injury. *Journal of Clinical Pathology, Vol. 52, No. 11,* 

in patients in anoxia-induced vegetative state. *Journal of the Neurological Sciences*,

MR imaging in diffuse axonal injury. *American Journal of Neuroradiology*, Vol.23,

resonance imaging of regional brain activity in patients with intracerebral gliomas:

details must be available for comparisons between studies.

be life-saving if clinical research cannot be done on these patient groups.

*(November 1999), pp. 804-806,* ISSN 1472-4146

No.5, (May 2002), pp. 794-802, ISSN 1936-959X

of patients with severe brain damage.

**12. References** 

**11. Conclusions** 

#### **10. Neuroimaging of self-consciousness and recovery**

A recent meta-analysis by Northoff et al. (2006) of 27 PET and fMRI studies comparing hemodynamic brain responses obtained during active paradigms comparing processing of stimuli related to the self with those of non-self-referential stimuli identified activation in cortical midline structures in all studies and occurring across all functional domains (e.g. verbal, spatial, emotional, and facial). Cluster and factor analyses indicated functional specialization into ventral, dorsal, and posterior cortical midline areas. The latter encompasses the posterior cingulate cortex and adjacent precuneus and is considered to be involved in self-integration – that is linkage of self-referential stimuli to the personal context (Northoff et al., 2004). Neuroimaging studies during tasks involving self-processing (i.e. selfreflection, self-perspective and free thoughts) have also reported the activation of the medial prefrontal areas. Gusnard, Akbudak, Shulman , and Raichle (2001) for example showed medial frontal activation when subjects had to make two judgments in response to pleasant *vs* unpleasant pictures (i.e. self-referential) as compared to indoors *vs* outdoors pictures (i.e. not self-referential). The same area was also shown to be engaged when subjects had to make self-referential judgments about trait adjectives (i.e. self-referential processing) as compared to when they had to make case judgments (Kelley et al., 2002) and when subjects responded to statements requiring knowledge of, and reflection on, their own abilities, traits and attitudes , i.e. self-reflective thought (Johnson et al., 2002). Taking a self– perspective (i.e. being the agent of an history) also activated medial prefrontal/anterior cingulate cortices (Vogeley et al., 2001). Finally, activation of the mesiofrontal areas was describes in studies dealing with the conscious resting state, i.e. free thought (Mazoyer et al., 2001), a brain state which "instantiates functions that are integral to the self". The recovery of consciousness of one VS patient has previously been linked to an increase in the functional connectivity within fronto-parietal network, (Laureys et al, 1999) encompassing the areas known to be most active in resting–state conditions (Gusnard et al., 2001). A growing body of evidence from Positron emission tomography (PET) and fMRI studies of healthy volunteers in a variety of altered states of consciousness has emphasized the role of this "default-mode" network in the genesis of awareness. In keeping with this, functional impairments to this network have been observed during sleepwalking, absence of seizures, deep sleep and anesthesia (Bassetti et al., 2000). fMRI has also proved its utility in identifying a number of cognitive functions which may be preserved in DOC patients, the results of which have, in some cases, proved prognostic of positive outcomes (Owen et al., 2008). In one such fMRI study investigating language processing. Coleman et. al found evidence of speech processing in three out of seven behaviorally non-communicative VS patients (Coleman et al., 2007). Six months after the scan, each of these patients had made a marked behavioral recovery relative to those patients who did not demonstrate comparable activations. Similar findings have also been reported for the neural responses observed when patients hear their own name (Di et al., 2007). Multimodal imaging approach can provide a powerful tool for assessing the mechanisms involved in the recovery of consciousness in DOC patients. Further longitudinal studies with large cohorts will prove useful in assessing its full value in predicting outcome. Such insights may then provide guidance for decisions relating to rehabilitation programs by those orientating these towards the effective stimulation of those functions that appear preserved, in order to maintain their integrity.
