**3. Functional Magnetic Resonance Imaging and Positron Emission Tomography**

At present a diagnosis of VS or MCS is made using prognostic markers from the patient's clinical history supported by detailed neurological and behavioral assessment by a multidisciplinary team over several weeks. However, the behavioral assessment of these patients predominately relies upon the subjective interpretation of observed spontaneous and volitional behavior. A diagnosis of VS is supported if the patient demonstrates no evidence of awareness of self or environment, no evidence of sustained, reproducible, purposeful or voluntary behavioral response to visual, auditory, tactile or noxious stimuli and critically no evidence of language comprehension or expression (MSTF, 1994). In contrast the patient in MCS demonstrates partial preservation of awareness of self and environment, responding intermittently, but reproducibly, to verbal command and therefore demonstrating some degree of basic language comprehension (Giacino et al., 2002).

PET and recently fMRI, by measurement of cerebral metabolism and brain activations in response to sensory stimuli, can provide important MR indices on the presence and location of any residual brain function.

PET is the most sensitive method to image trace amounts of molecules in vivo. Therefore this technique is used to measure in man or in the living animal biochemical and physiological processes in any organ with three dimensional resolution. The last 25 years have seen a rapid and still ongoing development in the production of positron emitters, radiochemical labeling techniques, tomograph technology and image reconstruction algorithms. Because of the possibility to see and measure quantitatively physiological disorders in an early stage, before permanent morphological damage has occurred, which will only then be visible in x-ray or magnetic resonance computer tomography, PET is finally finding its way from a sophisticated research tool into routine clinical diagnosis.

Resting cerebral metabolism derived from quantitative glucose uptake provides an indirect assessment of neuronal activity against which brain states may be compared quantitatively (Levy et al., 1987). All previous quantitative [18F] fluorodeoxyglucose–positron emission tomograph (FDG‐PET) investigations of VS have correlated the condition with a global reduction of brain metabolic activity: Laureys et al. (1999) have assessed regional cerebral glucose metabolism (rCMRGlu) and effective cortical connectivity in four patients in VS by means of statistical parametric mapping and FDG‐PET. Results showed a common pattern of impaired rCMRGlu in the prefrontal, premotor, and parietotemporal association areas and posterior cingulate cortex/precuneus in VS. In a next step, they demonstrated that in VS patients various prefrontal and premotor areas have in common that they were less tightly connected with the posterior cingulate cortex than in normal controls. Schiff et al. (2005) have described the first evidence of reciprocal clinical–pathological correlation with regional differences of quantitative cerebral metabolism. They studied five patients in VS with different behavioral features employing FDG‐PET, MRI and magnetoencephalographic (MEG) responses to sensory stimulation. Each patient's brain expressed a unique metabolic pattern. The specific patterns of preserved metabolic activity identified in these patients reflect novel evidence of the modular nature of individual functional networks that underlie conscious brain function. In three of the five patients, co‐registered PET/MRI correlate islands of relatively preserved brain metabolism with isolated fragments of behavior. Two patients had suffered anoxic injuries and demonstrated marked decreases in overall cerebral

At present a diagnosis of VS or MCS is made using prognostic markers from the patient's clinical history supported by detailed neurological and behavioral assessment by a multidisciplinary team over several weeks. However, the behavioral assessment of these patients predominately relies upon the subjective interpretation of observed spontaneous and volitional behavior. A diagnosis of VS is supported if the patient demonstrates no evidence of awareness of self or environment, no evidence of sustained, reproducible, purposeful or voluntary behavioral response to visual, auditory, tactile or noxious stimuli and critically no evidence of language comprehension or expression (MSTF, 1994). In contrast the patient in MCS demonstrates partial preservation of awareness of self and environment, responding intermittently, but reproducibly, to verbal command and therefore demonstrating some degree of basic language comprehension (Giacino et al.,

PET and recently fMRI, by measurement of cerebral metabolism and brain activations in response to sensory stimuli, can provide important MR indices on the presence and location

PET is the most sensitive method to image trace amounts of molecules in vivo. Therefore this technique is used to measure in man or in the living animal biochemical and physiological processes in any organ with three dimensional resolution. The last 25 years have seen a rapid and still ongoing development in the production of positron emitters, radiochemical labeling techniques, tomograph technology and image reconstruction algorithms. Because of the possibility to see and measure quantitatively physiological disorders in an early stage, before permanent morphological damage has occurred, which will only then be visible in x-ray or magnetic resonance computer tomography, PET is finally finding its way from a sophisticated research tool into routine clinical diagnosis. Resting cerebral metabolism derived from quantitative glucose uptake provides an indirect assessment of neuronal activity against which brain states may be compared quantitatively (Levy et al., 1987). All previous quantitative [18F] fluorodeoxyglucose–positron emission tomograph (FDG‐PET) investigations of VS have correlated the condition with a global reduction of brain metabolic activity: Laureys et al. (1999) have assessed regional cerebral glucose metabolism (rCMRGlu) and effective cortical connectivity in four patients in VS by means of statistical parametric mapping and FDG‐PET. Results showed a common pattern of impaired rCMRGlu in the prefrontal, premotor, and parietotemporal association areas and posterior cingulate cortex/precuneus in VS. In a next step, they demonstrated that in VS patients various prefrontal and premotor areas have in common that they were less tightly connected with the posterior cingulate cortex than in normal controls. Schiff et al. (2005) have described the first evidence of reciprocal clinical–pathological correlation with regional differences of quantitative cerebral metabolism. They studied five patients in VS with different behavioral features employing FDG‐PET, MRI and magnetoencephalographic (MEG) responses to sensory stimulation. Each patient's brain expressed a unique metabolic pattern. The specific patterns of preserved metabolic activity identified in these patients reflect novel evidence of the modular nature of individual functional networks that underlie conscious brain function. In three of the five patients, co‐registered PET/MRI correlate islands of relatively preserved brain metabolism with isolated fragments of behavior. Two patients had suffered anoxic injuries and demonstrated marked decreases in overall cerebral

**3. Functional Magnetic Resonance Imaging and Positron Emission** 

**Tomography** 

2002).

of any residual brain function.

metabolism. Two other patients with non‐anoxic, multifocal brain injuries demonstrated several isolated brain regions with relatively higher metabolic rates. A single patient who suffered severe injury to the tegmental mesencephalon and paramedian thalamus showed widely preserved cortical metabolism. The variations in cerebral metabolism in chronic VS patients indicate that some cerebral regions can retain partial function in catastrophically injured brains.

fMRI is based on the increase in blood flow to the local vasculature that accompanies neural activity in the brain. This result in a corresponding local reduction in deoxyhemoglobin because the increase in blood flow occurs without an increase of similar magnitude in oxygen extraction (Roy & Sherrington, 1890; Fox & Raichle, 1985). Since deoxyhemoglobin is paramagnetic, it alters the T2 weighted magnetic resonance image signal (Ogawa et al, 1990). Thus, deoxyhemoglobin is sometimes referred to as an endogenous contrast enhancing agent, and serves as the source of the signal for fMRI. Using an appropriate imaging sequence, human cortical functions can be observed without the use of exogenous contrast enhancing agents on a clinical strength (1.5 T) scanner (Bandettini et al., 1992, 1993; Schneider et al, 1993).

Functional activity of the brain determined from the magnetic resonance signal has confirmed known anatomically distinct processing areas in the visual cortex (Schneider, et al, 1993), the motor cortex, and Broca's area of speech and language-related activities (Hinke et al., 1993; Kim et al., 1995). Further, a rapidly emerging body of literature documents corresponding findings between fMRI and conventional electrophysiological techniques to localize specific functions of the human brain (Atlas et al., 1996; Detre, et al, 1995; George, et al, 1995). Consequently, the number of medical and research centers with fMRI capabilities and investigational programs continues to escalate.

Several fMRI studies in the VS have confirmed the findings of previous PET studies. Di et al. (2007) used fMRI to evaluate differences between seven VS and four MCS patients in brain activation occurring in response to the presentation of the patient's own name, spoken by familiar voice (SON-FV). They prospectively studied residual cerebral activation to SON-FV in seven patients with VS and four with MCS. Two patients with VS failed to show any significant cerebral activation. Three patients with VS showed SON-FV induced activation within the primary auditory cortex. Only two of the VS patients, and all four MCS patients, showed activation not only in the primary auditory cortex but also in hierarchically higherorder associative temporal areas.

Three months after fMRI examination, these two VS patients had progressed to the MCS. This study showed that fMRI measurement might be a useful tool for pre-clinically distinguishing MCS-like cognitive processing in some patients behavioural classified as vegetative. Schiff et al. (2005) have tested the hypothesis that MCS patients retain active cerebral networks that underlie cognitive function. fMRI was employed to investigate cortical responses in two male adults with severe brain injuries resulting to MCS and in seven healthy volunteers. Three passive stimulation tasks were performed: tactile stimulation, auditory narratives of familiar events presented by a familiar person, and the same auditory passages without language-related content. Results have showed a residual brain activity of cortical systems involved in a potential cognitive and sensory function despite their inability to follow simple instructions or communicate reliably.

In conclusion, results of these studies we analyzed confirm the idea that PET and fMRI activation profiles may constitute useful adjunctive diagnostic methods when behavioral

Neuroimaging and Outcome Assessment in Vegetative and Minimally Conscious State 187

the histograms derived from the subcortical white matter and the thalami and an increase in

In addition, DTI may be a valuable biomarker for the severity of tissue injury and a predictor for outcome. It reveals changes in the WM that are correlated with both acute GCS and Rankin scores at discharge (Huisman et al., 2004). Significant early reduction of anisotropy was observed in WM structures, in particular in the internal capsule and the corpus callosum, which are the sites most commonly involved by DAI (Arfanakis et al., 2002). Moreover, several regions recovered normal values of anisotropy 1 month after the injury (Arfanakis et al., 2002). Xu et al. (2007) found significant differences in the corpus callosum, internal and external capsule, superior and inferior longitudinal fascicles, and the fornix in TBI patients. They showed that FA and ADC measurements offered superior sensitivity compared to conventional MRI diagnosis of DAI. Salmond et al. (2006) reported increased diffusivity in TBI patients at least 6 months after their injury in the cerebellum, frontal, insula, cingulate, parietal, temporal, and occipital lobes. The anisotropy seems to be reduced both in the major WM tracts such as the corpus callosum and the internal and external capsule, and the associative fibers underlying the cortex. DTI has a number of advantages as an imaging biomarker of brain injury: first, it can be used to evaluate brain trauma in an unconscious or sedated patient; second, it could permit the evaluation of responses to treatment even when the clinical scores are inadequate for assessing the patient; third, quantitative DTI measurements are unlikely to be tainted by adverse central nervous system (CNS) effects of hypnotic drugs, unlike clinical scores; and fourth, DTI may be an important alternative marker, as low initial Glasgow Come Scale scores are of limited value in predicting the prognosis (Huisman et al., 2004). Finally, Perlbarg et al. (2009) showed significant FA differences between favorable and unfavorable 1-year outcome groups around four FA tracks: in inferior longitudinal fasciculus, posterior limb of the

internal capsule, cerebral peduncle, and posterior corpus callosum.

aminobutyric acid, Cre, Cho, myo-inositol, and scyllo-inositol (Figure 1).

Proton MRS (1H-MRS) is a non-invasive imaging technique that enables in vivo quantification of certain neurochemical compounds. Using the same equipment utilized for the conventional MRI, single-voxel 1H-MRS and multi-voxel Imaging (1H-MRSI) or Chemical Shift Imaging (CSI) provide metabolic information on brain damage that may not be visible with the conventional structural imaging methods. Then 1H-MRS, added to traditional MRI, offers the possibility to study the brain activity combining information on

Classically, the exploration of DOC is performed on 1,5 or 3 Tesla MR scanners and at intermediate or long echo time (TE) (135-288 ms). Long TE 1H-MRS detects the signal arising from four metabolites: N-acetyl-aspartate containing compounds (NAA), choline-containing compounds (Cho), creatine + phosphocreatine (Cre) and lactate (Lac). Short TE 1H-MRS identifies peaks from mobile lipids, Lac, alanine, NAA, Glutamate/Glutamine (Glx), γ-

NAA, which resonates at 2.02 parts per million (ppm), represents the largest proton metabolic concentration in the human brain after water. Indeed the concentration of NAA reaches on the order of 10 μmol/g. NAA is widely interpreted as a neuronal marker and implicated in several neuronal processes, mitochondrial functioning and osmoregulation.

**5. Magnetic Resonance Spectroscopy** 

structure and function.

the peak width of the thalamic histogram.

findings are very limited or ambiguous, helping in differential diagnosis, prognostic assessment and identification of pathophysiological mechanism.

#### **4. Diffusion Tensor Imaging**

Diffusion tensor imaging (DTI) is an emerging technique that complements traditional MRI and may be able to provide erstwhile unavailable information about the pathological substrates of DOC. DTI is a modified MRI technique that is sensitive to microscopic, threedimensional water motion within tissue. In cerebrospinal fluid, water motion is isotropic, i.e., roughly equivalent in all directions. In white matter, however, water diffuses in a highly directional or anisotropic manner. Due to the structure and insulation characteristics of myelinated fibers, water in these white matter bundles is largely restricted to diffusion along the axis of the bundle. DTI can thus be used to calculate two basic properties: the overall amount of diffusion and the anisotropy (Douaud et al., 2007; Benson et al., 2007; Kraus et al., 2007; Ringman et al., 2007; O'Sullivan et al., 2004). It is only very recently that DTI has been used to evaluate white matter integrity in patients with DOC. For example, Voss et al. (2006) described two patients with traumatic brain injury: one who had remained MCS for 6 years and one who had recovered expressive language after 19 years diagnosed as MCS. In both cases, widespread changes in white matter integrity were observed. Interestingly, however, the increased anisotropy and directionality in the bilateral medial parieto-occipital regions that was observed in the second patient reduced to normal values in a follow-up scan performed 18 months later. This coincided with increased metabolic activity, leading the authors to interpret these observations as evidence of axonal regrowth in this region. Although this is certainly a landmark finding in two high spectrum MCS patients, it remains to be seen whether DTI has any diagnostic or prognostic utility in a broader group of patients with disorders of consciousness. To this end, Tollard et al. (2009) and Perlbarg et al. (2009) have recently demonstrated that DTI measures in sub-acute severe traumatic brain injury may be a relevant biomarker for predicting the recovery of consciousness at 1 year. However, VS and MCS patients were classified in the same outcome category and potential differences between these two groups were not investigated. Although, in this context DTI has been generally used to address specific clinical problems, the study of white matter integrity in behaviorally defined states has a more basic relevance to understanding the relationship between brain and behavior in both health and disease. For example, in healthy volunteers, DTI techniques have been used recently to examine how structural changes underpin the behavioral changes that are related to learning a complex skill (Scholz et al., 2009). In a very recent study (Espejo et al., 2011), the integrity of white and grey matter regions was assessed in a group of 25 VS and MCS patients in vivo. In accordance with previous post-mortem work (Jennett et al., 2001; Adams et al., 1999) significant changes were observed in the integrity of the tissue in subcortical, thalamic and brainstem regions in the patients when compared to healthy volunteers. The precise location of this damage was not different between the MCS and VS sub-groups, which, again, accords well with previous post-mortem studies. However, an analysis of the MD values within two of these regions of interest (subcortical white matter and thalami), revealed significant differences between the patients meeting the clinical (behavioral) criteria defining VS and those who met the criteria defining MCS. Specifically, the VS patient group exhibited a decrease in the peak height of

findings are very limited or ambiguous, helping in differential diagnosis, prognostic

Diffusion tensor imaging (DTI) is an emerging technique that complements traditional MRI and may be able to provide erstwhile unavailable information about the pathological substrates of DOC. DTI is a modified MRI technique that is sensitive to microscopic, threedimensional water motion within tissue. In cerebrospinal fluid, water motion is isotropic, i.e., roughly equivalent in all directions. In white matter, however, water diffuses in a highly directional or anisotropic manner. Due to the structure and insulation characteristics of myelinated fibers, water in these white matter bundles is largely restricted to diffusion along the axis of the bundle. DTI can thus be used to calculate two basic properties: the overall amount of diffusion and the anisotropy (Douaud et al., 2007; Benson et al., 2007; Kraus et al., 2007; Ringman et al., 2007; O'Sullivan et al., 2004). It is only very recently that DTI has been used to evaluate white matter integrity in patients with DOC. For example, Voss et al. (2006) described two patients with traumatic brain injury: one who had remained MCS for 6 years and one who had recovered expressive language after 19 years diagnosed as MCS. In both cases, widespread changes in white matter integrity were observed. Interestingly, however, the increased anisotropy and directionality in the bilateral medial parieto-occipital regions that was observed in the second patient reduced to normal values in a follow-up scan performed 18 months later. This coincided with increased metabolic activity, leading the authors to interpret these observations as evidence of axonal regrowth in this region. Although this is certainly a landmark finding in two high spectrum MCS patients, it remains to be seen whether DTI has any diagnostic or prognostic utility in a broader group of patients with disorders of consciousness. To this end, Tollard et al. (2009) and Perlbarg et al. (2009) have recently demonstrated that DTI measures in sub-acute severe traumatic brain injury may be a relevant biomarker for predicting the recovery of consciousness at 1 year. However, VS and MCS patients were classified in the same outcome category and potential differences between these two groups were not investigated. Although, in this context DTI has been generally used to address specific clinical problems, the study of white matter integrity in behaviorally defined states has a more basic relevance to understanding the relationship between brain and behavior in both health and disease. For example, in healthy volunteers, DTI techniques have been used recently to examine how structural changes underpin the behavioral changes that are related to learning a complex skill (Scholz et al., 2009). In a very recent study (Espejo et al., 2011), the integrity of white and grey matter regions was assessed in a group of 25 VS and MCS patients in vivo. In accordance with previous post-mortem work (Jennett et al., 2001; Adams et al., 1999) significant changes were observed in the integrity of the tissue in subcortical, thalamic and brainstem regions in the patients when compared to healthy volunteers. The precise location of this damage was not different between the MCS and VS sub-groups, which, again, accords well with previous post-mortem studies. However, an analysis of the MD values within two of these regions of interest (subcortical white matter and thalami), revealed significant differences between the patients meeting the clinical (behavioral) criteria defining VS and those who met the criteria defining MCS. Specifically, the VS patient group exhibited a decrease in the peak height of

assessment and identification of pathophysiological mechanism.

**4. Diffusion Tensor Imaging** 

the histograms derived from the subcortical white matter and the thalami and an increase in the peak width of the thalamic histogram.

In addition, DTI may be a valuable biomarker for the severity of tissue injury and a predictor for outcome. It reveals changes in the WM that are correlated with both acute GCS and Rankin scores at discharge (Huisman et al., 2004). Significant early reduction of anisotropy was observed in WM structures, in particular in the internal capsule and the corpus callosum, which are the sites most commonly involved by DAI (Arfanakis et al., 2002). Moreover, several regions recovered normal values of anisotropy 1 month after the injury (Arfanakis et al., 2002). Xu et al. (2007) found significant differences in the corpus callosum, internal and external capsule, superior and inferior longitudinal fascicles, and the fornix in TBI patients. They showed that FA and ADC measurements offered superior sensitivity compared to conventional MRI diagnosis of DAI. Salmond et al. (2006) reported increased diffusivity in TBI patients at least 6 months after their injury in the cerebellum, frontal, insula, cingulate, parietal, temporal, and occipital lobes. The anisotropy seems to be reduced both in the major WM tracts such as the corpus callosum and the internal and external capsule, and the associative fibers underlying the cortex. DTI has a number of advantages as an imaging biomarker of brain injury: first, it can be used to evaluate brain trauma in an unconscious or sedated patient; second, it could permit the evaluation of responses to treatment even when the clinical scores are inadequate for assessing the patient; third, quantitative DTI measurements are unlikely to be tainted by adverse central nervous system (CNS) effects of hypnotic drugs, unlike clinical scores; and fourth, DTI may be an important alternative marker, as low initial Glasgow Come Scale scores are of limited value in predicting the prognosis (Huisman et al., 2004). Finally, Perlbarg et al. (2009) showed significant FA differences between favorable and unfavorable 1-year outcome groups around four FA tracks: in inferior longitudinal fasciculus, posterior limb of the internal capsule, cerebral peduncle, and posterior corpus callosum.

#### **5. Magnetic Resonance Spectroscopy**

Proton MRS (1H-MRS) is a non-invasive imaging technique that enables in vivo quantification of certain neurochemical compounds. Using the same equipment utilized for the conventional MRI, single-voxel 1H-MRS and multi-voxel Imaging (1H-MRSI) or Chemical Shift Imaging (CSI) provide metabolic information on brain damage that may not be visible with the conventional structural imaging methods. Then 1H-MRS, added to traditional MRI, offers the possibility to study the brain activity combining information on structure and function.

Classically, the exploration of DOC is performed on 1,5 or 3 Tesla MR scanners and at intermediate or long echo time (TE) (135-288 ms). Long TE 1H-MRS detects the signal arising from four metabolites: N-acetyl-aspartate containing compounds (NAA), choline-containing compounds (Cho), creatine + phosphocreatine (Cre) and lactate (Lac). Short TE 1H-MRS identifies peaks from mobile lipids, Lac, alanine, NAA, Glutamate/Glutamine (Glx), γaminobutyric acid, Cre, Cho, myo-inositol, and scyllo-inositol (Figure 1).

NAA, which resonates at 2.02 parts per million (ppm), represents the largest proton metabolic concentration in the human brain after water. Indeed the concentration of NAA reaches on the order of 10 μmol/g. NAA is widely interpreted as a neuronal marker and implicated in several neuronal processes, mitochondrial functioning and osmoregulation.

Neuroimaging and Outcome Assessment in Vegetative and Minimally Conscious State 189

NAA/Cr ratios were able to differentiate patients in VS who recovered awareness from those who remained in persistent VS. However, this alteration was not found in the

**COO**

**OH**

**-**

**3 CH OH**

**N(CH3)3 +**

**COO-**

**Lactate (Lac)**

**OH OH OH**

**OH**

**OH**

**2 NH H**

**Glutamate**

**O**

**OH**

**Myo-inositol (mI)** 

**O**

**OH**

**OH**

Fig. 1. Chemical structure and spectrum of main cerebral metabolites detected by 1H-MRS. In some studies has been shown that the combination of imaging techniques may be useful to predict the long-term neurological outcome. A 1H-MRS study in the pons allowed separating of patients who recovered from patients with severe neurological impairment, death or in VS. In addition, 1H-MRS metabolic alterations were not correlated with anatomical MRI lesions, suggesting that these two techniques are strongly complementarity (Carpentier et al., 2006). Tollard et al. (2009) reported the first study on patients with TBI based on a combined quantitative analysis of 1H-MRS and DTI. This combined analysis was

thalamus of patients in VS resulting from mild TBI (Kirov et al., 2007).

**NH**

**+**

**N CH3**

**Creatine (Cre)** 

**Choline - containing compounds (Cho)** 

**N(CH3)3 +**

**2 NH**

**COO**

**NH**

**N-Acetylaspartate (NAA)** 

**OH**

**O P O**

**O**

**O -**

**- OOC -**

**CO <sup>3</sup> CH**

**OH**

NAA synthesis occurs in mitochondria and requires acetyl-CoA and L-aspartic acid as substrates. NAA has been proposed to serve as a mitochondrial shuttle of acetyl-CoA used for fatty acid synthesis. Its peak decreases when there is neuron suffering or loss. The Cho peak (3.2 ppm) represents a combination of several choline-containing compounds, including free Cho, phosphorylcholine and glycerophosphorylcholine, and to a small extent acetylcholine. Free Cho acts as a precursor to acetylcholine, while glycerophosphorylcholine is a product of breakdown of membrane phosphatidylcholine and acts as an osmoregulator. Its peak increases when there is greater membrane turnover, cell proliferation or inflammatory process. The peak of Cre at 3.03 ppm represents total creatine and phosphocreatine supplies phosphate for conversion of ADP to ATP in creatine kinase reaction. Indeed these metabolites buffer the energy use and energy storage of cells. The level of total Cre mainly remains constant in many neuronal diseases. Thus, total Cre is often used as an internal reference (i.e., a denominator in metabolite signal ratio). The Lac (1.3 ppm) is an end product of anaerobic glycolysis, thus increase in Lac concentrations often serves as an index of altered oxidative metabolism, i.e., in ischemia, hypoxia, and cancer. Increases of Lac in the brain are often accompanied by decreased intracellular pH and high-energy phosphates. The proposed role of Lac is a source of energy for neurons and the transport of Lac plays an essential role in the concept of metabolic coupling between neurons and glia. Glutamate (Glu) is the highest excitatory neurotransmitter in concentration in the CNS. Its peak increases when neuronal and astrocytic activation impairs mitochondrial function and energy utilization. Indeed this process impairs Glu transport and its following enhancement is associated to cellular toxicity.

1H-MRS has been used for at least 15 years in the exploration of patients with altered consciousness, both to investigate the mechanisms of vigilance and to predict the possibilities of regaining consciousness.

Predicting outcome of patients with DOC is an integral part of clinical care, facilitating medical decision making and therapeutic intervention. Current neurological and neurophysiological methods do not enable prediction of outcome of these patients in early stages. Although conventional neuroimaging can provide important information for acute clinical management, its prognostic value is limited, particularly at early stage of injury resolution, owing to its poor sensitivity.

Several studies present in literature have demonstrated the value of 1H-MRS as an accurate tool to predict patient's clinical outcome. Indeed many investigators have shown that correlation exists between metabolite changes and outcome of patients with DOC.

Previous studies using single-voxel technique have shown in brain-injured subjects a significant correlation between unfavorable outcome and reduction of marker NAA in occipitoparietal white and gray matter (WM and GM) (Brooks et al., 2000; Friedman et al., 1999; Ross et al., 1998; Yoon et al., 2005), frontal WM (Garnett et al., 2000), parietal WM (Shutter et al., 2004), brainstem (Carpentier et al., 2006), splenium of the corpus callosum (Sinson et al., 2001; Cecil et al., 1998), and thalamus (Uzan et al., 2003), increase in choline a marker for cell membrane disruption in frontal WM (Garnett et al., 2000) and occipitoparietal WM and GM (Brooks et al., 2000; Cecil et al., 1998; Ross et al., 1998; Yoon et al., 2005), and increase in Glx in occipital GM and parietal WM (Shutter et al., 2004).

In particular, NAA levels seem to discriminate patients who recovered from coma from those who died or remained in persistent VS (Ricci et al., 1997). Uzan et al. (2003) carried out a thalamic proton MRS in patients in VS resulting from severe TBI. They found that

NAA synthesis occurs in mitochondria and requires acetyl-CoA and L-aspartic acid as substrates. NAA has been proposed to serve as a mitochondrial shuttle of acetyl-CoA used for fatty acid synthesis. Its peak decreases when there is neuron suffering or loss. The Cho peak (3.2 ppm) represents a combination of several choline-containing compounds, including free Cho, phosphorylcholine and glycerophosphorylcholine, and to a small extent acetylcholine. Free Cho acts as a precursor to acetylcholine, while glycerophosphorylcholine is a product of breakdown of membrane phosphatidylcholine and acts as an osmoregulator. Its peak increases when there is greater membrane turnover, cell proliferation or inflammatory process. The peak of Cre at 3.03 ppm represents total creatine and phosphocreatine supplies phosphate for conversion of ADP to ATP in creatine kinase reaction. Indeed these metabolites buffer the energy use and energy storage of cells. The level of total Cre mainly remains constant in many neuronal diseases. Thus, total Cre is often used as an internal reference (i.e., a denominator in metabolite signal ratio). The Lac (1.3 ppm) is an end product of anaerobic glycolysis, thus increase in Lac concentrations often serves as an index of altered oxidative metabolism, i.e., in ischemia, hypoxia, and cancer. Increases of Lac in the brain are often accompanied by decreased intracellular pH and high-energy phosphates. The proposed role of Lac is a source of energy for neurons and the transport of Lac plays an essential role in the concept of metabolic coupling between neurons and glia. Glutamate (Glu) is the highest excitatory neurotransmitter in concentration in the CNS. Its peak increases when neuronal and astrocytic activation impairs mitochondrial function and energy utilization. Indeed this process impairs Glu

transport and its following enhancement is associated to cellular toxicity.

possibilities of regaining consciousness.

resolution, owing to its poor sensitivity.

1H-MRS has been used for at least 15 years in the exploration of patients with altered consciousness, both to investigate the mechanisms of vigilance and to predict the

Predicting outcome of patients with DOC is an integral part of clinical care, facilitating medical decision making and therapeutic intervention. Current neurological and neurophysiological methods do not enable prediction of outcome of these patients in early stages. Although conventional neuroimaging can provide important information for acute clinical management, its prognostic value is limited, particularly at early stage of injury

Several studies present in literature have demonstrated the value of 1H-MRS as an accurate tool to predict patient's clinical outcome. Indeed many investigators have shown that

Previous studies using single-voxel technique have shown in brain-injured subjects a significant correlation between unfavorable outcome and reduction of marker NAA in occipitoparietal white and gray matter (WM and GM) (Brooks et al., 2000; Friedman et al., 1999; Ross et al., 1998; Yoon et al., 2005), frontal WM (Garnett et al., 2000), parietal WM (Shutter et al., 2004), brainstem (Carpentier et al., 2006), splenium of the corpus callosum (Sinson et al., 2001; Cecil et al., 1998), and thalamus (Uzan et al., 2003), increase in choline a marker for cell membrane disruption in frontal WM (Garnett et al., 2000) and occipitoparietal WM and GM (Brooks et al., 2000; Cecil et al., 1998; Ross et al., 1998; Yoon et

correlation exists between metabolite changes and outcome of patients with DOC.

al., 2005), and increase in Glx in occipital GM and parietal WM (Shutter et al., 2004).

In particular, NAA levels seem to discriminate patients who recovered from coma from those who died or remained in persistent VS (Ricci et al., 1997). Uzan et al. (2003) carried out a thalamic proton MRS in patients in VS resulting from severe TBI. They found that NAA/Cr ratios were able to differentiate patients in VS who recovered awareness from those who remained in persistent VS. However, this alteration was not found in the thalamus of patients in VS resulting from mild TBI (Kirov et al., 2007).

Fig. 1. Chemical structure and spectrum of main cerebral metabolites detected by 1H-MRS.

In some studies has been shown that the combination of imaging techniques may be useful to predict the long-term neurological outcome. A 1H-MRS study in the pons allowed separating of patients who recovered from patients with severe neurological impairment, death or in VS. In addition, 1H-MRS metabolic alterations were not correlated with anatomical MRI lesions, suggesting that these two techniques are strongly complementarity (Carpentier et al., 2006). Tollard et al. (2009) reported the first study on patients with TBI based on a combined quantitative analysis of 1H-MRS and DTI. This combined analysis was

Neuroimaging and Outcome Assessment in Vegetative and Minimally Conscious State 191

spindle and alpha-like rhythms, but they are more diffusely distributed than in the typical posterior regions and are not reactive to sound, pain, and light stimuli (Chokroverty, 1975; Huges, 1978). During sleep, fewer muscle twitches are observed, but a REM sleep remains (Oksenberg et al., 2001). In most patients, the transition from wakefulness to sleep is accompanied by some desynchronization of the background activity. Very-low-voltage EEG activity is all that can be detected in some patients. In others, persistent alpha activity is the most remarkable feature. In around 10% of patients with VS, the EEG is nearly normal late in the course of disease but without evidence of vision-induced alpha blocking (Danze et al., 1989). There have been occasional reports of isoelectric EEGs in patients in a VS, although it has not been confirmed (Higashi et al., 1977; Mizrahi et al., 1985). Typical epileptiform activity is unusual in patients in VS, as seizure activity is (The Multi-Society Task Force on PVS, 1994). Clinical recovery from the vegetative state may be paralleled by diminished delta and theta activity and reappearance of reactive alpha rhythm. Indeed, Babiloni (2009) has observed that occipital source power in the alpha band (8-13 Hz) of resting EEG, when calculated with low-resolution electromagnetic tomography (LORETA), is correlated with recovery outcome at 3-month follow-up in a group of VS patients; those who made a behavioural recovery had higher resting alpha band power than those who did not make a

The EEG in MCS shows diffuse slowing brain activity, mainly of the theta band, and in most cases responsive to external stimuli. However, there are insufficient data as well as the typical pattern of MCS concerns. Evoked potentials have been studied in patients in a VS and showed normal brainstem auditory responses but abnormal somatosensory responses: prolonged conduction time or absence of scalp potentials. ERPs are more useful than EEG in the differential diagnosis between VS and MCS. ERPs studies focusing on the assessment of conscious awareness have frequently examined four specific components: the N100, the

In a recent work, the authors focused on the prediction of consciousness recovery in patients with post-traumatic VS. They used a classical two-stimulus oddball task to elicit the P300 using the patient's own name as deviant and a pure tone as standard stimulus (''subject's own name" paradigm). There is evidence that the amplitude of the P300 wave increases when more salient stimuli are used, such as the own first name instead of visual or auditory deviants. The authors found that P300 is a strong predictor of future recovery of consciousness in VS. This finding is in line with several studies that have confirmed the utility of P300 evoked by deviant tones to predict awakening and favourable outcome from coma and VS (Cavinato et al., 2009). In another study Cavinato et al., (2011) continue to using the "subject's own name" paradigm, but add a pure tone and an "other first name" paradigm. The authors instructed their patients to count the occurrence of deviant stimuli to better differentiate between patients in VS and MCS. The study indicates that in 6 out of 11 patients fulfilling the behavioral criteria for VS a reliable P300 component could be observed in all two conditions. These findings corroborate earlier reports showing that 38% of patients in VS generate a P300 wave. The patients in MCS exhibit significantly longer P300 latencies for the "subject's own name" and the "other first name" paradigms than patients in VS. The increase of P300 latencies for more complex and salient paradigms in MCS but not

The TMS, for high temporal resolution, was proposed as an additional functional imaging technique for the study of cognitive function. To date only some studies have assessed VS and MCS patients with TMS. Moosavi et al. (1999) applied TMS to the hand and leg motor

mismatch negativity (MMN), the P300, and the N400 (Connoly & D'Arcy, 2000).

in VS might help in the difficult differential diagnosis of MCS vs. VS.

significant recovery.

97% specific for predicting an unfavorable outcome after 1 year, compared with 85% for DTI and 75% for 1H-MRS. Similarly, sensitivity was better with the combined analysis (86%) than with either DTI (79%) or MRS (75%).

To study metabolite changes from a wider area of the brain, with the advantage of identifying more anatomical and functional details, a few investigators have used 1H-MRSI (Holshouser et al., 2006; Marino et al. 2007; Shutter et al., 2006; Signoretti et al., 2002, 2008). This technique has an advantage over single-voxel 1H-MRS because generates individual spectra from multiple voxels at the same time. Also 1H-MRSI studies have highlighted close correlation between metabolite alterations and potential recovery.

Neurometabolite concentrations obtained soon after injury may be useful for predicting individual outcome. The decrease of NAA and the increase of Lac, seen by Marino et al. (2007) early after brain injury, were correlated with GOS score. Then these 1H-MRS data may be, at this stage, a reliable index of injury severity and disease outcome.

However it is need to note that 1H-MRS studies in patients with DOC are heterogeneous in terms of patient nature, injury types, time from cerebral damage, voxel location, methods and timing outcome assessment. In addition, in many studies the metabolite concentrations were expressed in terms of semiquantitative ratios. The assumption that the concentration of Cr as reference metabolite remains constant may be incorrect, especially in acute conditions. It is therefore advisable to obtain concentration expressed in standard units by applying absolute quantification. Some studies have expressed metabolite concentrations in term of absolute quantification (Brook et al., 2000; Friedman et al., 1999; Marino et al. 2007; Ross et al., 2000; Shutter et al., 2004).

Data reported so far demonstrate that MRS measure have the potential to provide new and important biological brain markers able to predict clinical outcome, helping in the therapeutic interventions, clinical and rehabilitative management of these patients, as well as to assist with family education.
