Brain Advance Access originally published online on May 23, 2008
Brain 2008 131(7):1783-1792; doi:10.1093/brain/awn087
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Intra-cortical connectivity in multiple sclerosis: a neurophysiological approach
1Istituto Scienze e Tecnologie della Cognizione—CNR, Unità MEG, Roma, 2Casa di Cura San Raffaele Cassino e IRCCS San Raffaele Pisana, 3Dipartimento di Neuroscienze ed Unità Operativa Complessa Radiologia Diagnostica ed Interventistica, AFaR, Ospedale Fatebenefratelli and 4Neurologia Clinica, Universitá Campus Biomedico, Roma, Italy
Correspondence to: Dr ssa Franca Tecchio, Unità MEG, Dipartimento di Neuroscienze, Ospedale Fatebenefratelli, Isola Tiberina, 00186 Rome, Italy E-mail: franca.tecchio{at}istc.cnr.it
 |
Summary |
|---|
 Top Summary IntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
Multiple sclerosis is an autoimmune disease predominantly affecting the white matter of the CNS, causing—among functional sequelae—cortico-cortical partial or total disconnection. Since functional connectivity linking cerebral regions is reliably reflected by synchronization of their neuronal firing, in this study an electrophysiological parameter measured by magnetoencephalography was used to quantify an intra-cortical connectivity (ICC) index focused on the primary somatosensory cortical areas (S1). Twenty-one patients affected by mild (Extended Disability Scale Score, median 1,5) relapsing-remitting (RR) multiple sclerosis in the remitting phase without clinically evident sensory impairment were evaluated. Three dimensional MRI was used to quantify the lesion load, discriminating black hole and non-black hole portions, normalized by individual brain volumes. When matched with a control population, multiple sclerosis patients showed a reduced ICC combined with the complete loss of the finger-dependent functional specialization in S1 cortex of the dominant hemisphere. No association was found between ICC impairment and disease duration, or prolongation of the central sensory conduction time, presence of spinal cord lesions and ongoing disease modifying therapy. The ICC index slightly correlated with the lesion load. A local index of ICC in a circumscribed brain primary area was altered in mildly disabled RR-multiple sclerosis patients, also in absence of any impairment of central sensory conduction. In conclusion, the diffuse damage influencing the multi-nodal network subtending complex cerebral functions also affects intrinsic cortical connectivity. The S1 ICC index is proposed as a highly sensitive and simple-to-test functional measure for the evaluation of intra-cortical synchronization mechanisms in RR-multiple sclerosis.
Key Words: magnetoencephalography; cerebral connectivity; primary sensorimotor cortex; lesion load black holes
Abbreviations: BHV, black hole volume; CCT, central conduction time; EDSS, Extended Disability Scale Score; EEG, electroencephalography; FSS, functional source separation; ICC, intra-cortical connectivity; LrF, lesion relative fraction; MEG, magnetoencephalography; PD, proton density; S1, primary somatosensory cortical areas; RR, relapsing-remitting; TLV, total lesion volume; WMHy, white matter hyperintensities
Received December 3, 2007. Revised April 11, 2008. Accepted April 15, 2008.
|
Introduction |
|---|
|
TopSummaryIntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
Multiple sclerosis is an autoimmune disorder characterized by multiple lesions of the central myelin and accumulating clinical signs due to demyelination and progressive axonal damage (Keegan and Noseworthy, 2002
). As a consequence of white matter injury, a partial or total disconnection from subcortical and spinal targets occurs, in parallel with deficits of cortico-cortical connectivity. Current efforts are devoted to delineating brain connectivity changes from an anatomical perspective through MRI techniques such as voxel-based morphometry and diffusion tensor imaging-based fibre tracking (Cercignani et al., 2002; Audoin et al., 2007; Cader et al., 2007; Reich et al., 2007). However, such techniques only provide information on anatomical substrates mainly related to contingents of large, unidirectional fibres. Blood oxygen level-dependent functional MRI, assessing alterations of neural activations in brain diseases including multiple sclerosis, has been enriched with functional and effective connectivity algorithms to more directly characterize neural connectivity impairments (Leocani and Comi, 1999; Reddy et al., 2000; Horwitz, 2003; Au Duong et al., 2005; Horwitz et al., 2005; Audoin et al., 2006; Cader et al., 2006). Such an approach does not allow investigation of rapid (millisecond range) phenomena linked with transient increments/decrements of neural assembly activities and functional cortico-cortical connectivity. In fact, studies in animals and humans demonstrate that separate cerebral regions synchronize their neuronal firings when functionally connecting each other (Gray et al., 1989; Engel et al., 2001; Fellous et al., 2001). Electro- and magneto-encephalography (EEG and MEG, Del Gratta et al., 2001) represent non-invasive neurophysiological techniques with appropriate temporal resolution to detect such phenomena in humans. Previous EEG coherence data in progressive multiple sclerosis patients demonstrated a significant decrease of alpha and theta band coherence between both anteroposterior and inter-hemispheric areas correlated with cognitive impairment and subcortical lesion load on MRI (Leocani et al., 2000). Likewise, a decreased alpha band inter-hemispheric coherence was found by Cover and colleagues (2006) in relapsing-remitting (RR) multiple sclerosis patients through a MEG paradigm.
Our group has previously identified a MEG synchronization measure within the primary sensory hand cortical area (S1) showing finger-dependent functional specialization in the dominant hemisphere of right handed healthy subjects (Tecchio et al., 2007a). In fact, the thumb cortical representation, activated in response to a sensory stimulus, is more phase locked than the little finger representation. These phase-locking phenomena reflect activity in S1 cortex sensitive to the strength of cortico-cortical connections and, therefore, provide an intra-cortical connectivity index (ICC).
The aim of the present study was to test whether the ICC index measured in S1 is involved in a cohort of patients affected by RR-multiple sclerosis. In order to assess the test sensitivity, we purposefully selected patients with absent or minimal disability in remitting phase and without clinically evident sensory deficits at the time of evaluation. Particular care was also devoted to a precise quantification of the lesion-load. Three-dimensional high resolution anatomical images were coregistered with lesion sensitive images estimating white matter hyperintensities (WMHy). We focused on the relationship between the ICC and putative MRI markers of axonal loss (black holes), because they appear to correlate more closely to the development of multiple sclerosis disability than traditional MRI features (Truyen et al., 1996; Bitsch et al., 2001).
|
Patients and Methods |
|---|
|
TopSummaryIntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
Patients
Patients were recruited on the basis of the following inclusion criteria: diagnosis of multiple sclerosis according to McDonald et al.'s Criteria (2001
); RR clinical course (Lublin and Reingold, 1996); low degree of disability as estimated by Extended Disability Scale Score (EDSS, Kurtzke, 1983) < 3,5; absence of clinical relapse or radiological evidence of disease activity for at least 3 months preceding the study; absence of sensory symptoms at the moment of recruitment; no treatment with corticosteroids or psychotropic drugs within the previous 3 months. The study was approved by Ethic Committee of ‘S. Giovanni Calibita’ Hospital and all patients gave written consent.
Clinical, neurophysiological and neuroradiological examination
In addition to EDSS, upper limb sensory evoked potentials were collected and the central conduction time (CCT) calculated following standardized methods and criteria (Mauguière et al., 1999). All patients underwent cervical and brain MRI examination at least 6 months before and at the time of recruitment.
MRI exam
MR image acquisition
Imaging was performed at the Radiology Unit, equipped with an Achieva 1.5 T scanner (Philips Medical Systems, Best, The Netherlands), provided of 33 mT/m gradient amplitude, online 2D/3D geometric distortion correction and a standard quadrature head coil. The acquisition protocol consisted of one 3D high resolution anatomical sequence and a group of four 2D sequences for lesion characterization. The former was empirically optimized to increase grey/white matter image contrast (T1-weighted Turbo Field Echo TR/TE/FA = 9.5 ms/4 ms/8°; 2562 matrix, 160 coronal contiguous slice, in-plane resolution 0.98 x0.98 mm). The latter included a dual spin echo (TR/TE/FA of, respectively, 4000 ms/20–120 ms/90°, yielding to a T2 and a proton density—PD—weighted images); two T1-Spin Echo sequences (TR/TE/FA of, respectively, 550 ms/15 ms/60°, one before intravenous injection of the contrast agent gadolinium and the second 5 min later) and one FLAIR (TR/TE/FA = 10 000 ms/100 ms/90°). All 2D sequences were acquired as axial oblique contiguous 3 mm slices (2562 image matrix, 240 mm field of view), encompassing the bi-commisural plane. After entering the PACS system of the Department, the DICOM images were post-processed on a dual processor Linux workstation. All subjects underwent an identical imaging protocol. The MRI scanner did not undergo major hardware upgrades during data collection.
MR image analysis (lesion characterization, optimized segmentation and volumes calculation for normalization)
The high resolution 3D anatomical images were coregistered to the PD/T2 dataset in order to exactly match the regions of the brain in which an abnormal signal intensity could be detected by a trained neuroradiologist.
A region of interest's approach was used to manually trace the regions that showed abnormal signal intensity, either as hyper- or hypo-intensity, according to the visual inspection. MRICro software (Brett et al., 2001) allowed the delineation of such regions classified as being white matter PD/T2 hyperintensities (WMHy). Within the WMHy regions, the black hole areas, identified as pre-contrast T1 persistently hypo-intense and coincident T2 hyper-intense lesions, were submitted to a separate measurement (Fig. 1). According to post-contrast behaviour, black holes are divided into ‘acute’ black holes if showing gadolinium enhancement (Matthews, 1999; van Waesberghe et al., 1999) and ‘permanent’ black holes if not, while being detectable on a previous MRI session performed at least 6 months earlier (Bagnato et al., 2003).
|
To correctly normalize the lesion volume in each patient, procedures aimed at identifying the intra-cranial structures were carried out. The lesional region of interest's where then subtracted from the 3D anatomical images, yielding to 3D lesion-free images that entered an optimized segmentation process. Statistical parametric mapping software (SPM2, www.fil.ion.ucl.ac.uk/spm, Friston et al., 1995
) running under MATLAB 7.2 (the Mathworks, Inc., MA, USA) was used to manually reorient each subject's dataset, pointing the origin to the anterior commissure. To improve the accuracy of segmentation, a preliminary iterative normalization/segmentation procedure was performed on the 3D lesion-free images, as implemented in optimized voxel-based morphometry method (Good et al., 2001). A study-specific template was created for several reasons: (i) minimizing potential normalization problems related to differences of registration to the standard space among patients’ brains; (ii) to take into account the demographics of our patients, who had a frequent involvement of periventricular and corona radiata white matter and (iii) to reduce the magnetic field inhomogeneities and non-uniformities related to the static magnetic field. The normalization matrices derived from previously obtained lesion-free volumes were applied to their own lesional region of interest images, resampled to 1 x 1 x 1 mm for anatomical accuracy. The overall lesion-specific volumes were calculated firstly as number of voxels within each image, then transformed into millilitres.
MRI variables
Three MRI variables were considered that characterized the lesion load:
- the total lesion volume (TLV), i.e. the volume of the whole T2 WMHy;
- the black hole volume (BHV);
- the volume of the non-black hole portion of the total lesion (nBHV = TLV–BHV).
These three quantities were divided by the standardized normalization factors, i.e. total white matter and the brain parenchyma with and without the cerebrospinal fluid (for nomenclature see Table 1).
|
Functional ICC exam
Recordings
MEG data were separately acquired during the electrical stimulation of the thumb and little finger of the right and left hand (ring electrodes, 0.2 ms electric pulses, inter-stimulus interval of 641 ms, stimulus intensities set at about twice the subjective somatosensory threshold) and compared with the rest state. Recordings were carried out via a 28-channel system operating in a magnetically shielded room (Vacuumschmelze Gmbh), the active channels being evenly distributed on a spherical surface (13.5 cm of curvature radius) and covering a total area of about 180 cm2. Cerebral activity from rolandic areas have been recorded via a single position on the hemisphere contralateral to the stimulated hand, by centring the recording apparatus on the C3/C4 positions of the 10–20 international EEG system. MEG data were sampled at 1000 Hz (pre-sampling analogical filter 0.48–256 Hz), and collected for off-line processing.
ICC index
By submitting MEG signals to the procedure called functional source separation (FSS, Tecchio et al., 2007b), the source activities of cortical neural networks devoted to the contralateral thumb and little finger sensory perception were segregated in each hemisphere. FSS comes from the blind source separation algorithms, which separate sources, exploiting their own signal waveform, while remaining blind to their biophysical relationships with the recorded signal (Cichocki and Amari, 2002). The aim of FSS is to enhance the separation of relevant signals by exploiting some a priori knowledge without renouncing the advantages of using only information contained in original signal waveforms. In fact, starting from an independent component analysis model, FSS identifies a source on the basis of its functional behaviour. In addition to the identified functional source (FS) time signal (FS(t)), the field distribution it generates can be obtained by retro-projecting the source activity in the space scanned by the MEG sensors, being thus used as the input for inverse-problem solution algorithms. In the present study, FS(t) related to the thumb (FST(t)) and little finger (FSL(t)) cortical representations were identified by requiring maximal responsiveness at around 24 ms to the corresponding contralateral finger stimulation (Barbati et al., 2006). It is well known that this latency marks the primary sensory area (S1) response to specific projections coming from thalamus after hand fingers stimulation. Moreover, MEG is sensitive to the field generated by pyramidal neurons tangentially oriented to the scalp, i.e. topographically positioned within the walls of cortical sulci. The identified sources (FST and FSL) are thus localized in the post-central wall of the Rolandic sulcus in the area 3b, as demonstrated by Barbati and colleagues (2006). Once identified the time signals FST(t) and FSL(t), they were considered as two mutually interconnected portions of the hand representation in sensory area, as previously observed that FST responded also to the little finger stimulation—with around half the strength than when stimulating the thumb—and vice versa for FSL (Barbati et al., 2006). The sensory cortical connectivity was estimated as the phase locking between these source activities. Connectivity of the thumb sensory cortical representation was obtained as the phase locking after the stimulation of the thumb (ICCT). The same was done for the little finger representation ICCL (Tecchio et al., 2007a, see Appendix I). In right handed healthy subjects (mean age ± SD: 31 ± 8 years, 7 females and 7 males, Edinburgh Inventory Test Score > 85) the ICC in the dominant hemisphere demonstrated somatotopic properties. In fact, ICCT was higher than ICCL. Moreover, both ICCT and ICCL were correlated with the contralateral finger sensorimotor dexterity (Tecchio et al., 2007a), as quantitatively estimated by the ‘fingertip writing’ test (Lezak, 1976). The cited study documented that all these properties were evident selectively in the gamma frequency range [(33, 45)Hz].
Statistical analysis
ICCT and ICCL value distributions did not differ from a Gaussian (Kolmogorov-Smirnov P > 0.400).
The ICC of the network devoted to S1 representation of the thumb (ICCT) and of the little finger (ICCL) were studied by a repeated measure analysis of variance with Hemisphere (left, right) and Finger Representation (thumb, little finger) as within-subject factors and Group (controls, multiple sclerosis patients) as the between-subject factor. Factors of interest were Group, indicating different ICC properties in healthy subjects and patients, as well as the interaction factor Finger Representation x Group indicating alterations of the physiological somatotopy of the index in patients.
For those electrophysiological measures, which were different in patients from in controls (significance threshold P < 0.05), possible bivariate correlations with the clinical and neuroradiological measures were carried out. Since neuroradiological variable distributions differed from a Gaussian (consistently, Kolmogorov Smirnov test, P < 0.054), non-parametric tests were applied.
Patients’ and controls’ group age distributions did not differ from Gaussian (Kolmogorov-Smirnov P > 0.400 for both). The two group ages did not differ (independent sample t-test P > 0.400). No dependence on age was observed in the control group for any of the ICC index (Pearson P > 0.200 consistently). No effect of gender was present on any ICC index, as indicated by the between-subject factor Gender [F(1,10) = 0.006, P = 0.941] in the analysis of variance design, together with the absence of significance of any interaction effect with intra-subject factors Hemisphere (left, right) and Finger Representation (thumb, little finger).
|
Results |
|---|
|
TopSummaryIntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
Patients
Twenty-one right-handed (Edinburgh Inventory > 85, Oldfield, 1971
) RR-multiple sclerosis patients with a low degree of disability (median EDSS 1, 5; Table 2) were enrolled. They presented a mean disease duration of 10 ± 7 years and a median relapse rate of 0.33 (Table 2). Ten patients were on therapy with interferon-β-1a.
|
Although 13 out of the 21 patients claimed upper or lower limb sensory symptoms in the past (11 upper limb—4 of whom bilateral—2 lower limb), none of them presented any of this at the time of MEG evaluation in agreement with the inclusion criteria. Nine patients of the 11 with prior upper limb sensory symptoms showed delayed latency of scalp potentials with a normal cervical potential and prolonged CCT (Table 2).
Fifteen patients had a positive history of one or more cervical plaques, seven of whom still showed the lesions at the recruitment time (Table 2).
In agreement with the inclusion criteria of being in remitting phase, no patient had acute black holes (stable black hole behaviour in the two MRI exams performed at least 6 months apart one each other). No grey matter lesions were detected in 19 of our patients and very few millimetric lesions at cortical or juxtacortical level of frontal lobes and at the vertex were present in two patients. Normalized lesion fractions with respect to the total white matter volume are provided in Table 2.
Functional ICC
Since ICCT and ICCL behaved differently in the right and left hemispheres [Hemisphere x Finger Representation interaction factor F(1,29) = 6.086, P = 0.020, Fig. 3], separate investigations were carried out. In the right hemisphere, ICCT and ICCL considered together were lower in patients than in controls [Group factor F(1,31) = 4.189, P = 0.049, Fig. 2]. Neither ICCT or ICCL differed in the two groups when compared separately (P > 0.200 consistently). In the left hemisphere, the clear distinction between thumb and little finger representations, observed in healthy subjects, was lost in patients [Finger Representation x Group interaction factor F(1,29) = 11.915, P = 0.002, Fig. 3]. In fact, while in controls ICCT was greater than ICCL (paired t-test P = 0.001), ICCT and ICCL values did not differ in multiple sclerosis patients (P = 0.354; Fig. 3). Moreover, in the left hemisphere mean ICCT was smaller in patients than in controls (Fig. 3, Independent sample t-test, P = 0.047). No difference between the two groups was found for ICCL (P = 0.228). As a consequence, left hemisphere ICCT was studied in relationship with the clinical and neuroradiological findings. To assess whether the ICC reduction was due to reduced responsiveness from primary area, the strength of the M20 equivalent current dipole (Barbati et al. 2006) activated in response to the thumb and little finger was compared in patients and controls. The analysis of variance model with Finger as within-subject factor and Group as between-subject factor, indicated no significance of differences in patients and controls [F(1,21) = 1.364, P = 0.256].
|
|
Relationship among left hemispheric ICCT, clinical, neurophysiological and radiological state
The ICCT was not associated to right median nerve central sensory conduction time prolongation: in fact, similar ICCT values were equally found in patients with normal and increased CCT [paired t-test t(17) = 0.206, P = 0.839]. Neither ICCT was modified by the occurrence of a cervical lesion in the clinical history [independent sample t-test t(18) = –1.708, P = 0.105]. No association was present between ICCT and either EDSS (Spearman's rho P = 0.410, or disease duration (P = 0.958, Fig. 4) or history of sensory symptoms [independent t-test t(19) = –0.430, P = 0.671]. Moreover, in those patients with very long disease duration (
10 years, nine patients) with the present low EDSS values, suggesting a benign form, no difference of the ICCT value was found with respect to the other patients [independent sample t-test t(19) = 1.430, P = 0.170]. No association linked ICCT and relapses rate (Pearson’ r P = 0.966). Interferon therapy had no relationship with ICCT values (independent samples t-test P = 0.709).
|
A slightly inverse correlation was found between the ICCT and the lesion relative fraction (LrF) (Table 3). Further investigating this relationship, two patients’ subgroups could be identified according to the extent of LrF (Fig. 4). The inverse correlation was confirmed even in patients with smaller lesion volumes (Table 3). Moreover, in this subgroup, the ICCT value tended to decrease with respect to the healthy subjects [independent t-test for equal variances not assumed t(14.068) = 1.871, P = 0.082].
|
To understand whether a different role is played by black hole and non-black hole lesion portions, their correlation with the ICCT was considered separately. We found that the association was present with non-black hole relative fractions and not with black hole relative fractions (Table 3). The black hole relative fraction was slightly smaller than the non-black hole relative fraction [paired t-test t(20) = –2.568, P = 0.018].
|
Discussion |
|---|
|
TopSummaryIntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
An altered pattern of ICC with a significant reduction of ICC index within primary sensory areas in multiple sclerosis patients with low or minimal disability was found. ICC did not associate to the clinical state, probably as a consequence of the homogeneity of EDSS toward lower scores. ICC was independent of any prolongation or delay of CCT or the radiological evidence of spinal lesions. No reduction of S1 cortical responsiveness was documented in patients. These findings support the idea of a mainly cortical origin of the alteration. Furthermore, the finger-dependent functional specialization of hand sensorimotor network in the left hemisphere was lost in patients. In fact, while in healthy subjects the ICC of primary neural networks devoted to the thumb was higher than those related to the little finger, this pattern vanished in multiple sclerosis patients.
It is worth remembering that by ICC the level of connection between specific neuronal pools in the primary hand sensory region (area 3b) is tested by measuring the level of synchronization of their oscillating activities in the gamma band. Within cortical areas, the intrinsic functional connectivity measured by ICC stands on the anatomical topology of lateral connections, modulated by extrinsic inputs, whose architecture embodies specific predictions that have been acquired and stored during evolution and through experience-dependent learning (Hebb, 1949; Riehle et al., 2000; von Stein et al., 2000; Petersen and Sakmann, 2001; Lutzenberger et al., 2002; Kaiser and Lutzenberger, 2005). Previous findings showed that the ICC index is possibly sensitive to the shaping of the lateral connections of the cortical ‘modules’ affected by dynamic learning processes. In fact in healthy subjects the ICC of both thumb and little finger primary sensory cortical networks correlated with contralateral finger dexterity, a parameter related to a sensorimotor perception skill (Tecchio et al., 2007b). The relationship between high gamma band phase locking and the dexterity of the represented finger came out in the dominant hemisphere and not in the non-dominant one. This fits the modelling proposed by the ‘temporal binding’ theory of perception (Singer, 1999). This theory defines neural synchrony as crucial to enhance the saliency of neural responses. Our findings in healthy subjects indicate that the synchrony mechanisms become ‘structural’ of the network devoted to the most dexterous district. On these bases, the dynamic gamma band phase locking measured by ICC is suggested as a code for finger dexterity, in addition to the magnification of somatotopic central maps. Sensory perception relies on an anatomical hierarchy of neural networks made of synaptic relays—nodes—recruited by a feed-forward information flow, modulated by inter-nodal feed-back (Singer, 1999; Engel et al., 2001; Varela et al., 2001). Present results of reduced S1 ICC in patients could be understood in light of previous findings showing that within a multi-modal neural network, the disruption of connection fibres yields to changes of activation properties in single nodes (for a review Steriade, 2004). Differently from previous electrophysiological methods characterizing long-range anteroposterior and inter-hemispheric connectivity in multiple sclerosis (Leocani et al., 2000; Cover et al., 2006; for a review see Hoffmann, 2007), the ICC index is an innovative approach providing information about local ICC within a single primary area. The privileged role of S1, main origin of projections to primary motor cortex (Rizzolatti and Luppino, 2001), makes this node a ‘second-last’ station of almost all complex network processing. This hierarchical key role of S1, together with possible node activity impairments induced by inter-nodes’ disconnections, suggests S1 ICC as a valuable marker of multiple sclerosis disconnection damage.
The slight inverse correlation between ICC and lesion load, in absence of any detectable S1 lesion, indicates a contribution of WMHy to ICC alteration. On the other hand, significant ICC reduction in patients, only slightly associated to the lesion load, can be viewed as a result of the widespread (cortical and subcortical) damage in multiple sclerosis, whose extent is more strictly associated to the clinical manifestations of the disease than the extent of focal pathology (for review see Barkhof, 2002; De Stefano et al., 2002; Filippi and Rocca, 2005). An alternative hypothesis behind ICC alteration in patients could be that it is an epiphenomenon of circumscribed S1 cortical damage below the level of lesion load analysis. The documented unaltered responsiveness of the thumb and little finger primary sensory areas indicates that possible S1 lesions do not alter the activation of the same neuronal pools whose interconnections are measured by ICC. These findings, thus, speak against the hypothesis of ICC reduction due to local lesions not detectable by our anatomical data assessment.
The slight inverse ICC correlation with the lesion load was more evident with non-black hole than with black hole portions. A hypothesis could be drawn that the ICC is more affected by non-black hole-mediated aberrant inputs than by impoverished inputs carried on by fibres bordering functionally idle black hole areas. Minimal load of both non-black hole and black hole areas in our patients, makes mandatory confirmations on patients’ cohorts characterized by a wider range of lesion load.
In a previous study on RR-multiple sclerosis patients it was suggested that in the relapse phase of multiple sclerosis there is a cortical disinhibition possibly as compensation counteracting the damage in the course of the disease. In the remitting phase normal cortical excitability was observed (Caramia et al., 1991, 2004). The extent and excitability of the recruited neuronal pool do not affect the ICC index, which only accounts for the level of intra-node connectivity, and was altered even in the remitting phase. S1 ICC estimate, if supported by correlation with clinical data in larger population follow-up studies, might represent an innovative index for functional evaluation in multiple sclerosis.
In our cohort of mildly affected patients, there seems to be little, if any, correlation between clinical and neuroradiological pictures and MEG results. Moreover, patients with rather short disease duration presented with the same ICC alteration pattern as patients with long disease duration. These findings suggest that the S1 connectivity impairment could be a functional marker of the presence of the disease.
In conclusion, an altered pattern of ICC within primary sensory areas in multiple sclerosis patients with low disability was found. S1 ICC estimate is proposed as a non-invasive index for functional connectivity loss in RR-multiple sclerosis.
|
Appendix I |
|---|
|
TopSummaryIntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
ICC estimate
The network phase locking dynamic measure was obtained accordingly with the procedure previously applied by our group (Tecchio et al., 2007). After the extraction of sources related to the activity of thumb (FST(t)) and little finger (FSL(t)) neuronal network (Barbati et al., 2006
), the network phase locking index was obtained by the following steps:
(1) the FST(t) and FSL(t) signals were forward–backward band pass filtered in the gamma band [
= (33–45) Hz] by a Butterworth filter of the second order, obtaining 
signals, where X = T, L;
(2) the analytic signals of were calculated:
 | (1) |
Hilbert transform. The analytic signal was studied rather than the signal itself, to separately express amplitude and phase of FSX, maintaining the same time resolution of the original signal;
(3) the phase 
of the analytic signal is:
 | (2) |
(4) the dynamic synchronization between the FST and FSL sources was defined as:
 | (3) |
, tk 
[T1, T2], with T2–T1 the intervals length between two successive stimuli, NX is the number of considered stimuli for each finger X = {T, L}. 
is independent of the signal amplitudes; it is a dimensionless number varying between 0 and 1: if the two signals are strictly locked (
= constant) then equals 1, whereas if varies randomly, then tends toward zero.
(5) the phase locking index ICC was finally obtained by averaging in the time interval lasting 50 ms centred in the time point corresponding to maximal value automatically found between 20 and 75 ms after the stimulus onset (tMAX):
 | (4) |
To be noted, the time-dependent behaviour of the phase locking index was obtained in the two different conditions of thumb and little finger stimulations. In conclusion, the index 
estimates the phase synchronization between the two finger source rhythmic activities in the gamma frequency band while stimulating each of two contra-lateral fingers (X = thumb, little finger).
|
References |
|---|
|
TopSummaryIntroductionPatients and MethodsResultsDiscussionAppendix IReferences |
|---|
Audoin B, Au Duong MV, Malikova I, Confort-Gouny S, Ibarrola D, Cozzone PJ, et al. Functional magnetic resonance imaging and cognition at the very early stage of MS. J Neurol Sci (2006) 245:87–91.[CrossRef][Web of Science][Medline]
Audoin B, Guye M, Reuter F, Au Duong MV, Confort-Gouny S, Malikova I, et al. Structure of WM bundles constituting the working memory system in early multiple sclerosis: a quantitative DTI tractography study. Neuroimage (2007) 36:1324–30.[CrossRef][Web of Science][Medline]
Audoin B, Ibarrola D, Ranjeva JP, Confort-Gouny S, Malikova I, Ali-Chérif A, et al. Compensatory cortical activation observed by fMRI during a cognitive task at the earliest stage of MS. Hum Brain Mapp (2003) 20:51–8.[CrossRef][Web of Science][Medline]
Au Duong MV, Boulanouar K, Audoin B, Treseras S, Ibarrola D, Malikova I, et al. Modulation of effective connectivity inside the working memory network in patients at the earliest stage of multiple sclerosis. Neuroimage (2005) 24:533–8.[CrossRef][Web of Science][Medline]
Bagnato F, Jeffries N, Richert ND, Stone RD, Ohayon JM, McFarland HF, et al. Evolution of T1 black holes in patients with multiple sclerosis imaged monthly for 4 years. Brain (2003) 126:1782–9.
Barbati G, Sigismondi R, Zappasodi F, Porcaro C, Graziadio S, Valente G, et al. Functional source separation from magnetoencephalographic signals. Hum Brain Mapp (2006) 27:925–34.[CrossRef][Web of Science][Medline]
Barkhof F. The clinico-radiological paradox in multiple sclerosis revisited. Curr Opin Neurol (2002) 15:239–45.[CrossRef][Web of Science][Medline]
Bitsch A, Kuhlmann T, Stadelmann C, Lassmann H, Lucchinetti C, Bruck W. A longitudinal MRI study of histopathologically defined hypointense multiple sclerosis lesions. Ann Neurol (2001) 49:793–6.[CrossRef][Web of Science][Medline]
Brett M, Leff AP, Rorden C, Ashburner J. Spatial normalization of brain images with focal lesions using cost function masking. Neuroimage (2001) 14:486–500.[CrossRef][Web of Science][Medline]
Cader S, Cifelli A, Abu-Omar Y, Palace J, Matthews PM. Reduced brain functional reserve and altered functional connectivity in patients with multiple sclerosis. Brain (2006) 129:527–37.
Cader S, Johansen-Berg H, Wylezinska M, Palace J, Behrens TE, Smith S, et al. Discordant white matter N-acetylasparate and diffusion MRI measures suggest that chronic metabolic dysfunction contributes to axonal pathology in multiple sclerosis. Neuroimage (2007) 36:19–27.[CrossRef][Web of Science][Medline]
Caramia MD, Cicinelli P, Paradiso C, Mariorenzi R, Zarola F, Bernardi G, et al. Excitability changes of muscular responses to magnetic brain stimulation in patients with central motor disorders. Electroencephalogr Clin Neurophysiol (1991) 81:243–50.[CrossRef][Web of Science][Medline]
Caramia MD, Palmieri MG, Desiato MT, Boffa L, Galizia P, Rossini PM, et al. Brain excitability changes in the relapsing and remitting phases of multiple sclerosis: a study with transcranial magnetic stimulation. Clin Neurophysiol (2004) 115:956–65.[CrossRef][Web of Science][Medline]
Cercignani M, Bozzali M, Iannucci G, Comi G, Filippi M. Intra-voxel and inter-voxel coherence in patients with multiple sclerosis assessed using diffusion tensor MRI. J Neurol (2002) 249:875–83.[CrossRef][Web of Science][Medline]
Cichocki A, Amari SI. Adaptive blind signal and image processing (2002) New York: Wiley.
Cover KS, Vrenken H, Geurts JJ, van Oosten BW, Jelles B, Polman CH, et al. Multiple sclerosis patients show a highly significant decrease in alpha band interhemispheric synchronization measured using MEG. Neuroimage (2006) 29:783–8.[CrossRef][Web of Science][Medline]
Del Gratta C, Pizzella V, Tecchio F, Romani GL. Magnetoencephalography - a noninvasive brain imaging method with 1 ms time resolution. Rep Prog Phys (2001) 64:1759–814.[CrossRef]
De Stefano N, Narayanan S, Francis SJ, Smith S, Mortilla M, Tartaglia MC, et al. Diffuse axonal and tissue injury in patients with multiple sclerosis with low cerebral lesion load and no disability. Arch Neurol (2002) 59:1565–71.
Engel AK, Fries P, Singer W. Dynamic predictions: oscillations and synchrony in top-down processing. Nat Rev Neurosci (2001) 2:704–16.[CrossRef][Web of Science][Medline]
Fellous JM, Houweling AR, Modi RH, Rao RP, Tiesinga PH, Sejnowski TJ. Frequency dependence of spike timing reliability in cortical pyramidal cells and interneurons. J Neurophysiol (2001) 85:1782–7.
Filippi M, Rocca MA. MRI evidence for multiple sclerosis as a diffuse disease of the central nervous system. J Neurol (2005) 252(Suppl 5):v16–24.[CrossRef][Web of Science][Medline]
Friston KJ, Holmes AP, Poline JB, Grasby PJ, Williams SC, Frackowiak RS, et al. Analysis of fMRI time-series revisited. Neuroimage (1995) 2:45–53.[CrossRef][Web of Science][Medline]
Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS. A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage (2001) 14:21–36.[CrossRef][Web of Science][Medline]
Gray CM, Konig P, Engel AK, Singer W. Oscillatory responses in cat visual cortex exhibit inter-columnar synchronization which reflects global stimulus properties. Nature (1989) 338:334–7.[CrossRef][Medline]
Hebb DO. The organization of behavior (1949) New York: Wiley.
Horwitz B. The elusive concept of brain connectivity. Neuroimage (2003) 19:466–70.[CrossRef][Web of Science][Medline]
Horwitz B, Warner B, Fitzer J, Tagamets MA, Husain FT, Long TW. Investigating the neural basis for functional and effective connectivity. Application to fMRI. Philos Trans R Soc Lond B Biol Sci (2005) 360:1093–108.
Kaiser J, Lutzenberger W. Human gamma-band activity: a window to cognitive processing. Neuroreport (2005) 16:207–11.[CrossRef][Web of Science][Medline]
Keegan BM, Noseworthy JH. Multiple sclerosis. Annu Rev Med (2002) 53:285–302.[CrossRef][Web of Science][Medline]
Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology (1983) 33:1444–52.
Leocani L, Comi G. EEG coherence in pathological conditions. J Clin Neurophysiol (1999) 16:548–55.[CrossRef][Web of Science][Medline]
Leocani L, Locatelli T, Martinelli V, Rovaris M, Falautano M, Filippi M, et al. Electroencephalographic coherence analysis in multiple sclerosis: correlation with clinical, neuropsychological, and MRI findings. J Neurol Neurosurg Psychiatry (2000) 69:192–8.
Lezak MD. Neuropsychological assessment (1976) New York: Oxford University Press.
Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis: results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on clinical trials of new agents in multiple sclerosis. Neurology (1996) 46:907–11.
Lutzenberger W, Ripper B, Busse L, Birbaumer N, Kaiser J. Dynamics of gamma-band activity during an audiospatial working memory task in humans. J Neurosci (2002) 22:5630–8.
Matthews PM. Axonal loss and demyelation in multiple sclerosis. J Neurol Neurosurg Psychiatry (1999) 67:708–9.
Mauguière F, Allison T, Babiloni C, Buchner H, Eisen AA, Goodin DS, et al. Somatosensory evoked potentials. The international federation of clinical neurophysiology. Electroencephalogr Clin Neurophysiol Suppl (1999) 52:79–90.[Medline]
McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al. Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Ann Neurol (2001) 50:121–7.[CrossRef][Web of Science][Medline]
Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia (1971) 9:97–113.[CrossRef][Web of Science][Medline]
Petersen CC, Sakmann B. Functionally independent columns of rat somatosensory barrel cortex revealed with voltage-sensitive dye imaging. J Neurosci (2001) 21:8435–46.
Reddy H, Narayanan S, Arnoutelis R, Jenkinson M, Antel J, Matthews PM, et al. Evidence for adaptive functional changes in the cerebral cortex with axonal injury from multiple sclerosis. Brain (2000) 123(Pt 11):2314–20.
Reich DS, Smith SA, Zackowski KM, Gordon-Lipkin EM, Jones CK, Farrell JA, et al. Multiparametric magnetic resonance imaging analysis of the corticospinal tract in multiple sclerosis. Neuroimage (2007) 38:271–9.[CrossRef][Web of Science][Medline]
Riehle A, Grammont F, Diesmann M, Grun S. Dynamical changes and temporal precision of synchronized spiking activity in monkey motor cortex during movement preparation. J Physiol Paris (2000) 94:569–82.[CrossRef][Web of Science][Medline]
Rizzolatti G, Luppino G. The cortical motor system. Neuron (2001) 31:889–901.[CrossRef][Web of Science][Medline]
Singer W. Neuronal synchrony: a versatile code for the definition of relations? Neuron (1999) 24:49–65. 111–25.[CrossRef][Web of Science][Medline]
Steriade M. Neocortical cell classes are flexible entities. Nat Rev Neurosci (2004) 5:121–34.[CrossRef][Web of Science][Medline]
Tecchio F, Graziadio S, Barbati G, Sigismondi R, Zappasodi F, Porcaro C, et al. Somatosensory dynamic gamma-band synchrony: a neural code of sensorimotor dexterity. Neuroimage (2007a) 35:185–93.[CrossRef][Web of Science][Medline]
Tecchio F, Porcaro C, Barbati G, Zappasodi F. Functional source separation and hand cortical representation for a brain-computer interface feature extraction. J Physiol (2007b) 580:703–21.
Truyen L, van Waesberghe JH, van Walderveen MA, van Oosten BW, Polman CH, Hommes OR, et al. Accumulation of hypointense lesions (‘black holes’) on T1 spin-echo MRI correlates with disease progression in multiple sclerosis. Neurology (1996) 47:1469–76.
van Waesberghe JH, Kamphorst W, De Groot CJ, van Walderveen MA, Castelijns JA, Ravid R, et al. Axonal loss in multiple sclerosis lesions: magnetic resonance imaging insights into substrates of disability. Ann Neurol (1999) 46:747–54.[CrossRef][Web of Science][Medline]
Varela F, Lachaux JP, Rodriguez E, Martinerie J. The brainweb: phase synchronization and large-scale integration. Nat Rev Neurosci (2001) 2:229–39.[CrossRef][Web of Science][Medline]
von Stein A, Chiang C, Konig P. Top-down processing mediated by interareal synchronization. Proc Natl Acad Sci USA (2000) 97:14748–53.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  Y. He, A. Dagher, Z. Chen, A. Charil, A. Zijdenbos, K. Worsley, and A. Evans Impaired small-world efficiency in structural cortical networks in multiple sclerosis associated with white matter lesion load Brain, May 12, 2009; (2009) awp089v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||