Introduction

       Multiple sclerosis (MS) is considered as a chronic inflammatory, autoimmune, demyelinating neurologic disease of the central nervous system that is characterized by disability and cognitive deficits clinically. Cognitive impairment is now considered one of the important clinical markers of MS. Conventional MRI offers the most sensitive technique for detecting multiple sclerosis lesions and has proven to be an important clinical tool for diagnosing MS and monitoring therapeutic trials. But the functional deficits of MS does not simply depend on the extent of tissue destruction, but rather represents a cortical reorganization of complex balance between tissue damage, and tissue repair^[1].

       During the past decade, numerous studies have reported that functional magnetic resonance imaging provides information about the extent and nature of brain plasticity, which follows structural injury in MS patients and might have the potential to limit the clinical manifestations of the disease^{[1, 2]}. Several fMRI studies have compared the activation patterns during motor, visual, and cognitive tasks of MS patients with those of healthy controls^[1,2,3,4,5]. Resting-state fMRI (rsfMRI) techniques has emerged as a novel informative method for investigating the functional networks in the human brain, to evaluate the brain dysfunction 'at rest' in MS patients.

1. Task functional MR imaging

1.1　Task functional MR imaging: Basic principle

       fMRI is a relatively new procedure that uses MR imaging to measure the tiny metabolic changes that take place in an active part of the brain by different types of physical sensation (sight, sound, touch, taste, smell) or activity such as problem solving and/or movement. fMRI provides a sensitive, noninvasive tool for mapping patterns of activation in the working human brain, which is an increasingly common tool for "brain mapping" in cognitive science.

       MRI is able to detect a small difference (a signal in the order of 3%) between the two types of hemoglobin (deoxygenated and oxygenated hemoglobin), this is called a blood-oxygen level dependent, or "BOLD" signal, first discovered in 1990 by Seiji Ogawa. As neurons do not have internal reserves for glucose and oxygen, more neuronal activity requires more glucose and oxygen to be delivered rapidly through the blood stream. Hemoglobin is diamagnetic when oxygenated (oxyhemoglobin) but paramagnetic when deoxygenated (deoxyhemoglobin). The MR signal of blood is therefore slightly different depending on the level of oxygenation in blood stream of capillaries. These changes can be either positive or negative depending on the relative changes in both cerebral blood flow (CBF) and oxygen consumption. Increases in CBF that outstrip changes in oxygen consumption will lead to increased BOLD signal^[6]. The signal difference is very small, but given many repetitions of a thought, action or experience, statistical methods can be used to determine the areas of the brain which reliably show more of this difference as a result, and therefore which areas of the brain are active during that thought, action or experience. BOLD effects are measured using rapid volumetric acquisition of images with contrast weighed by T1 or T2^*. Such images can be acquired with moderately good spatial and temporal resolution; images are usually taken every 1-4 seconds, and the voxels in the resulting image typically represent cubes of tissue about the millimeters scale, and as event-related fMRI, the full time course of a BOLD response to a briefly presented stimulus lasts about 15 seconds for the robust positive response^[6]. Like any technique, fMRI has its advantages and disadvantages, and in order to be useful, the experiments that employ it must be carefully designed and conducted to maximize its strengths and minimize its weaknesses^[6].

1.2　Advancement of task functional MR imaging (fMRI) in MS

       In MS patients, the severity of clinical signs is not closely related to indices of structural brain damage provided by conventional MRI. fMRI provides information regarding the extent and nature of brain plasticity, which follows neuronal/structural damage and might have the potential to limit the extent of clinical presentation of the disease. An altered recruitment of regions normally devoted to the performance of a given task and/or the recruitment of additional areas, which are not typically activated by healthy people for performing that given task, have been described in patients with MS, regardless of their clinical phenotype when investigating the visual, cognitive and motor systems^[4]. These functional cortical changes vary across patients at different stages of the disease, after an acute relapse, and in clinically stable patients^{[4, 7]}. The correlation found between measures of functional changes and MRI measures of structural damage suggest that brain plasticity might help to limit the clinical consequences of the extent and severity of brain tissue damage.

       Changes in functional organization of the cerebral cortex have been reported in fMRI studies that have compared the activation patterns during motor, visual, and cognitive tasks in MS patients with those of healthy controls. An abnormal pattern of brain activation has also been related to fatigue in patients with MS^[4]. fMRI studies in MS patients have so far provide conflicting results that are difficult to compare because of differences in the criteria used for patient selection, the activation paradigm, the experimental design, and the MR acquisition parameters. Functional brain reorganization mainly consists of an increase in the extent of activation of the brain areas used by healthy subjects, as well as the recruitment of additional brain areas. There is increasing evidence that the severity of the clinical manifestations of MS patients does not simply depend on the extent of tissue destruction, but rather represents a complex balance among tissue damage, tissue repair and cortical reorganization^[4]. However, the precise role of brain functional changes in MS patients has yet to be fully understood^[8,9,10].

       Several studies have attempted to develop sophisticated statistical approaches to establish strength of activation and synchrony between specific brain areas by analysis of functional and effective connectivity^[11]. The optimization of analysis methods, as well as the comparison of models of activation between patients with MS and controls, might help to explain abnormalities of function of specific brain networks and their relation to clinical symptoms. The combination of measures of functional connectivity with measures of structural damage within specific white-matter fiber bundles is likely to improve our understanding of the relation between structural and functional abnormalities, as suggested by two studies in patients with RRMS and benign MS (Fig 1) ^{[12, 13]}. The role of longitudinal and multi-site fMRI studies in MS has yet to be fully explored. Previous findings support the use of fMRI in large-scale longitudinal trials to monitor the effect of motor and cognitive rehabilitation or pharmacological therapies on the enhancement of any beneficial effect of cortical adaptive plasticity^{[4, 14]}. Other aspects that should be considered are the development of fMRI paradigms unbiased by differences in task performance between patients with MS and controls, which could make the assessment of more disabled patients feasible, and the development of acquisition protocols specifically tailored to the imaging of function and structure of relatively small regions, now that high-field scanners are increasingly available making it possible to provide improved spatial resolution.

点击查看大图

Fig 1  Areas of increased activation in patients with benign MS compared with healthy controls during the analysis of the Stroop interference condition. (A, B) Patients with benign MS had increased activity in several areas located in the frontal and parietal lobes, bilaterally, including the anterior cingulate cortex, superior frontal sulcus, inferior frontal gyrus, precuneus, secondary sensorimotor cortex, visual cortex and cerebellum. (C) The analysis of functional connectivity, by use of dynamic causal modelling, showed different connectivity strengths between patients with benign MS and controls: within-group connections that were significant with a one-sample t-test are shown as black arrows in healthy controls and as dashed arrows in patients with MS. The arrows and p values resulting from the between-group t-test comparisons are shown in red in cases of increased strength of connection in patients versus controls, and in blue in cases of reduced strength of connection in patients versus controls (two p values are shown for all bi-directional associations). (Bakshi R, 2008, Reproduced with permission from John Wiley and Sons (Lancet Neurol.) Publishing)

2. Resting-state fMRI

2.1　Resting-state fMRI: Basic principle

       It has recently been shown that task-related increases of neuronal metabolism are small, less than 5% compared to the metabolism of the brain at rest ^{[15, 16]}. Spontaneous low-frequency fluctuations of the cerebral blood oxygenation level-dependent signal can be measured during rest (no task) with fMRI^{[17, 18]}, these fluctuations have shown strong temporal coherence between brain regions (synchronization) that represent functional systems, or so-called resting state networks, engaged in e.g. sensorimotor, attention and visual processing^[17]. New data-driven analysis techniques, such as independent component analysis^[19], have been developed recently and are well-suited for the analysis of fMRI data in the absence of an active task paradigm. A recent supplement to group-independent component analysis, called dual regression^{[20, 21]}, allows for a principled comparison between groups, using second level spatial correlations based on individual time-courses.

       Human brain is a network, which consists of spatially distributed, but functionally linked regions that continuously share information with each other^[22]. Resting-state networks are grouped by their anatomical and functional properties. Some studies^{[22, 23]} have revealed interesting new findings about the functional connections of specific brain regions and local networks, as well as important new insights in the overall organization of functional communication in the brain network (Fig 2). Resting-state networks are divided into groups based on their anatomical and functional properties and include basal ganglia (BG), auditory (AUD), sensorimotor (MOT), visual (VIS), default-mode (DMN), attentional (ATTN), and frontal (FRONT) networks^{[24, 25]}. The default network is a network of brain regions that are active when the individual is not focused on the outside world and the brain is at wakeful rest. Also called the default mode network (DMN), it is characterized by coherent neuronal oscillations at a rate lower than 0.1 Hz (one every ten seconds). DMN is an interconnected and anatomically defined brain system that preferentially activates when individuals focus on internal tasks such as daydreaming, envisioning the future, retrieving memories and gauging others' perspectives. It is negatively correlated with brain systems that focus on external visual signals. Its subsystems include part of the medial temporal lobe for memory, part of the medial prefrontal cortex for theory of mind, and the posterior cingulate cortex for integration^[26], along with the adjacent precuneus and the medial, lateral and inferior parietal cortex. In the infant brain, there is limited evidence of the default network, but default network connectivity is more consistent in children aged 9-12 years, suggesting that the default network undergoes developmental change^[27]. DMN has been hypothesized to be relevant to disorders including Alzheimer's disease, autism, and schizophrenia^[26]. Lower connectivity was found across the default network in people who have experienced long term trauma, such as childhood abuse and Post Traumatic Stress Disorder^[28].

       Functional connectivity is defined as the temporal dependency of neuronal activation patterns of anatomically separated brain regions and in the past years numerous neuroimaging studies have initiated exploration of functional connectivity by measuring the level of co-activation of resting-state fMRI time-series between brain regions^[22]. Resting state functional connectivity measures low frequency (<0.08 Hz) blood oxygen level dependent (BOLD) signal fluctuations between regions at rest. Resting state functional connectivity MRI (fcMRI) has become a particularly useful tool for studying regional relationships in typical and atypical populations.

       Resting state functional connectivity measures are of interest for several reasons. First, some consider them to reflect human anatomical connectivity^{[29, 30]}. Second, the reliance of functional connectivity MRI (fcMRI) on resting state data unburdens experimental design, subject compliance, and training demands making it attractive for studies of development and clinical groups^[31]. Because many investigators have already obtained large task related datasets of atypical populations, the ability to take advantage of existing task data and extract resting state data is of considerable interest^[32].

点击查看大图

Fig 2  Sagittal, coronal, and axial spatial maps of the six RSNs: RSNs 1 (default), RSNs 2 (dorsal attention), RSNs 3 (visual), RSN 4 (auditory), RSN 5 (somato-motor), and RSN 6 (self-referential) (Mantini D, 2007, with permission from Proc. Natl. Acad. Sci. (USA) Publishing)

2.2　Advancement of resting-state fMRI in MS

       Resting state functional magnetic resonance imaging is to analyze no-task functional magnetic resonance imaging data. Because task-fMRI in multiple sclerosis may only reveal the proverbial tip of the iceberg, it is critical to take into account brain activity that occurs in the absence of external stimulation in order to better understand how the brain functions and adapts in patients with multiple sclerosis.

       Patients with clinically isolated syndrome show increased synchronization in six of the eight resting state networks, including the default mode network and sensorimotor network, compared to controls or relapsing remitting patients (Fig 3). No significant decrease is found in patients with clinically isolated syndrome. No significant resting state synchronization differences are found between relapsing remitting patients and controls. Normalized grey matter volume is decreased and white matter diffusivity measures is abnormal in relapsing remitting MS patients compared to controls, whereas no atrophy or diffusivity changes are found for the clinically isolated syndrome group. Thus, early synchronization changes are found in patients with clinically isolated syndrome that are suggestive of cortical reorganization of resting state networks. These changes are lost in patients with relapsing remitting MS with increasing brain damage, indicating that cortical reorganization of resting state networks is an early and finite phenomenon in multiple sclerosis^[9].

       In patients with secondary progressive (SP) and primary progressive (PP) multiple sclerosis, between-group differences in DMN activity are found in the left medial prefrontal cortex (mPFC), left precentral gyrus (PcG) and anterior cingulate cortex (ACC). Compared to controls, patients with SPMS have reduced activity in the mPFC and PcG, while patients with PPMS have reduced activity in the PcG and the ACC. Compared to patients with PPMS, patients with SPMS have increased ACC activity. Reduction of RS activity in the ACC is more pronounced in cognitively impaired v.s. cognitively preserved patients with MS. In patients with MS, DMN abnormalities correlate with the PASAT and word list test scores and diffusion tensor MRI changes in the corpus callosum and the cingulum. These results suggest that a dysfunction of the anterior components of the default-mode network may be among the factors responsible for the accumulation of cognitive deficits in patients with progressive multiple sclerosis^[33].

       In hippocampal formation of MS patients, right hippocampal volume is significantly smaller in MS patients as compared with controls. Left hippocampal volume is also smaller in MS patients compared with controls, but not significantly so. Resting-state functional connectivity between the hippocampus and its anatomic input or target areas, including the anterior cingulate gyrus, thalamus and prefrontal cortex, is significantly decreased in MS patients. Decreased hippocampal functional connectivity is more pronounced in a subgroup of MS patients with hippocampal atrophy, although subtle decreases of functional connectivity are also found in patients with normal hippocampal volume. These results suggest that substantial abnormalities of hippocampal functional connectivity are already present before spatial memory function is impaired, especially in those patients with more pronounced hippocampal atrophy. Longitudinal studies should now assess whether these functional connectivity and structural changes may precede memory impairment in MS patients^[8].

       Coefficients of effective connectivity of the sensorimotor network are similar in control subjects and pediatric MS patients. The preservation of brain adaptive properties might explain the favorable medium-term clinical outcome of pediatric MS patients. The progressive recruitment of cortical networks over time in patients with the adult RR forms of the disease might result in a loss of their plastic reservoir, thus possibly contributing to subsequent disease evolution^[34]. PPMS patients show increased activations and abnormal functional connectivity measures in several areas of the sensorimotor network. Such changes are correlated with the structural damage to the white matter fiber bundles connecting these regions^[35]. Compared with controls, MS patients have more significant activations of several areas of the cognitive network involved in Stroop performance, bilaterally. Compared with controls, BMS patients also have increased connectivity strengths between several cortical areas of the sensorimotor network and the right (R) inferior frontal gyrus and the R cerebellum, as well as decreased connectivity strengths with the anterior cingulate cortex. Coefficients of altered connectivity are moderately correlated with structural MRI metrics of tissue damage within intra- and inter-hemispheric cognitive-related WM fiber bundles, while no correlations are found with the remaining fiber bundles studied. The findings suggest that functional cortical changes in patients with benign multiple sclerosis (BMS) might represent an adaptive response driven by damage of specific WM structures^{[13, 36]}. Functional connectivity and structural damage to some of the major brain motor white matter bundles suggest an adaptive role of functional connectivity changes in limiting the clinical consequences of structural damage in patients with relapsing-remitting multiple sclerosis^[13].

点击查看大图

Fig 3  The upper rows show resting state networks identified with independent component analysis; the lower rows show increased synchronization in patients with clinically isolated syndrome for each network. (A) Executive functioning network; synchronization is increased compared to controls in the left medial prefrontal cortex. (B) Sensorimotor network: increased synchronization compared to controls in the right premotor cortex and inferior parietal gyrus. (C) Ventral and dorsal attention system: increased synchronization compared to controls in the bilateral precuneus. Synchronization is also increased compared to relapsing remitting patients in the precuneus. (D) Default mode network: increased synchronization compared to relapsing remitting patients in the posterior cingulate gyrus. (E) Right frontoparietal network: increased synchronization compared to relapsing remitting patients in the left inferior temporal gyrus and right superior temporal gyrus. (F) Left frontoparietal network: increased synchronization compared to relapsing remitting patients in the left superior parietal gyrus and the occipital lobe. (G) Visual processing and (H) auditory and language processing: no significant differences between groups are found in these networks. (Roosendaal SD, 2010, with permission from Oxford University Press (Brain) Publishing)

3. Conclusions

       fMRI studies have demonstrated evidence for cortical reorganization in patients with MS performing cognitive or movement associated tasks, which partly explains the discrepancy between conventional MRI findings, such as T2 lesion load and the extent and severity of clinical disability. fMRI may have a role in monitoring treatment response since there is a correlation between cortical reorganization and cognitive recovery. Resting-state fMRI (rsfMRI) has emerged as a novel informative method for investigating the development of large-scale functional networks in the human brain. The brain 'at rest' may provide additional information in the evaluation of patients with MS. Several studies have reported regarding resting state networks, including the default mode network and sensorimotor network, and showed increased synchronization in clinically isolated syndrome (CIS), secondary progressive (SP) and primary progressive (PP) MS. This article reviewed evidence that indicated that cortical reorganization of resting state networks was an early and finite phenomenon in multiple sclerosis with task and resting state fMRI. In the future, more studies should be performed to assess the functional and structural substrates of cognitive network changes in patients with multiple sclerosis. A combination of functional connectivity maps and diffusion tensor fiber tractography for mapping axonal connections between cortical areas may be a plausible and powerful tool in the evaluation of MS patients.