Research Article Brain Connectomics Modification to

Research ArticleBrain Connectomicsrsquo Modification to Clarify Motor andNonmotor Features of Myotonic Dystrophy Type 1

Laura Serra1 Matteo Mancini12 Gabriella Silvestri3

Antonio Petrucci4 Marcella Masciullo5 Barbara Spanograve1 Mario Torso1

Chiara Mastropasqua1 Manlio Giacanelli4 Carlo Caltagirone67

Mara Cercignani8 Giovanni Meola9 and Marco Bozzali1

1Neuroimaging Laboratory IRCCS Santa Lucia Foundation Via Ardeatina 306 00179 Rome Italy2Department of Engineering ldquoRoma Trerdquo University Via Vito Volterra 62 00154 Rome Italy3Department of Geriatrics Orthopedics and Neuroscience Institute of Neurology Catholic University of Sacred HeartLargo Agostino Gemelli 8 00168 Rome Italy4UOC Neurologia e Neurofisiopatologia AO San Camillo Forlanini Via Portuense 332 00149 Rome Italy5SPInal REhabilitation Lab (SPIRE) IRCCS Santa Lucia Foundation Via Ardeatina 306 00179 Rome Italy6Department of Clinical and Behavioural Neurology IRCCS Santa Lucia Foundation Via Ardeatina 306 00179 Rome Italy7Department of Neuroscience University of Rome ldquoTor Vergatardquo Via Montpellier No 1 00133 Rome Italy8Clinical Imaging Sciences Centre Brighton amp Sussex Medical School University of Sussex Falmer BrightonEast Sussex BN1 9RR UK9Department of Neurology IRCCS Policlinico San Donato University of Milan Via Morandi 30 San Donato Milanese20097 Milan Italy

Correspondence should be addressed to Marco Bozzali mbozzalihsantaluciait

Received 8 January 2016 Accepted 17 April 2016

Academic Editor Zygmunt Galdzicki

Copyright copy 2016 Laura Serra et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

The adult form of myotonic dystrophy type 1 (DM1) presents with paradoxical inconsistencies between severity of brain damagerelative preservation of cognition and failure in everyday life This study based on the assessment of brain connectivity andmechanisms of plasticity aimed at reconciling these conflicting issues Resting-state functionalMRI and graph theoretical methodsof analysis were used to assess brain topological features in a large cohort of patients with DM1 Patients compared to controlsrevealed reduced connectivity in a large frontoparietal network that correlated with their isolated impairment in visuospatialreasoning Despite a global preservation of the topological properties peculiar patterns of frontal disconnection and increasedparietal-cerebellar connectivity were also identified in patientsrsquo brainsThe balance between loss of connectivity and compensatorymechanisms in different brain networks might explain the paradoxical mismatch between structural brain damage and minimalcognitive deficits observed in these patients This study provides a comprehensive assessment of brain abnormalities that fit wellwith both motor and nonmotor clinical features experienced by patients in their everyday life The current findings suggest thatmeasures of functional connectivity may offer the possibility of characterizing individual patients with the potential to become aclinical tool

1 Introduction

Myotonic dystrophy type 1 (DM1) is the most commonmuscular dystrophy observed in adults [1] It is caused by aCTG triplet repeat expansion within the myotonic dystro-phy protein kinase (DMPK) gene located on chromosome

19q133 whose inheritance is autosomal dominant [2] DM1is a multisystemic disorder dominated by muscular impair-ment but involving also other organs including the brain[1]

It is becoming increasingly clearer that most of theimpairment observed in patients with DM1 is driven by

Hindawi Publishing CorporationNeural PlasticityVolume 2016 Article ID 2696085 10 pageshttpdxdoiorg10115520162696085

2 Neural Plasticity

higher-level dysfunctions [3ndash7] DM1 brains have beendemonstrated to be structurally damaged in both tissues thegrey (GM) and white matter (WM) [7ndash11] with a specificanatomical distribution of abnormalities This structuraldamage has been consistently reported across independentstudies [7ndash11] and it was more recently associated with CTGtriplet expansions in theDMPK gene andmeasures of clinicalseverity [7] Abnormal patterns of brain connectivity havealso been reported in DM1 patients and have been demon-strated to account for patientsrsquo personality traits [6] Whilemental retardation is frequently observed in the congenitalform of DM1 [12] the adult forms are typically characterizedby isolated cognitive deficits [7 13]This relative preservationof global cognition contrasts with the pathological andneuroimaging evidence of diffuse WM abnormalities [7ndash10] and with the paradoxical failure of these patients ineveryday life In addition upon investigating functional con-nectivity within the so-called default mode network whosedisruption is typically associated with cognitive impairmentin degenerative dementia [14 15] DM1 patients reveal anincrease of connectivity in some critical nodes [6] Thesedata taken altogether suggest DM1 as neurodevelopmentaldisorder that associates with peculiar rearrangements inneuronal networksrsquo segregation and integration propertiesThese complex alterations which are hardly detectable bystructural brain assessments are likely to account for thedistinctive motor and nonmotor features observed in DM1Resting-state functional MRI (RS-fMRI) [16] is one of themost widely used methods to investigate brain connectivityin neurological and psychiatric diseases [17 18] with theadvantage of not requiring participants to perform any activetask RS-fMRI data can be analysed using different method-ological approaches One promising technique is based onthe whole brain analysis driven by graph theory [19] amathematical approach that describes complex systems asnetworks [20] In simple words the brain is conceptualized asa number of regions (nodes) that are functionally connectedto each other by edges and whose importance and efficiencywithin the whole network are determined by their functionalspecialization (ie segregation) and integration In this viewsome nodes are more critical (ie centrality) for informationprocessing (efficiency in information transferring) and arecalled ldquohubsrdquo Abnormal connectivity between ldquohubsrdquo isbelieved to cause more deficits than that between peripheralnodes [19 20] With these concepts in mind the currentwork aimed at assessing topological properties ofDM1 brainstheir potential relationship with genetics and their abilityin accounting for patientsrsquo motor and nonmotor clinicalfeatures To the best of our knowledge this is the first attemptto investigate brain connectomics in DM1 patients

2 Materials and Methods

21 Participants Thirty-one patients with a molecular diag-nosis of DM1 were recruited from the Neuromuscular andNeurological Rare Diseases Center at San Camillo ForlaniniHospital (Rome Italy) and the Institute of Neurology atthe Catholic University of Rome (Rome Italy) Data from

Table 1 Principal demographic characteristics of studied subjects

DM1patients119873 = 31

HS119873 = 26

119901 value

Mean (SD) age [years] 399 (114) 457 (132) nsa

Gender (FM) 160150 19070 nsb

Mean (SD) years offormal education 124 (22) 140 (33) nsa

aOne-way ANOVA bChi-squareDM1 = myotonic dystrophy type 1 HS = healthy subjects

Table 2 Principal genetic and clinical characteristics of DM1patients


Age at onsetChildhood-onset (age range 6ndash16 years) 12 (387)Adulthood-onset (age range 18ndash60 years) 19 (612)Size of CTG tripletsrsquo expansion on DMPK gene

Mean (SD) [range]6371

(4564)[54ndash2000]

IDMC nomenclatureE1 (CTG range 50ndash150) (119873 and ) 1 (30)E2 (CTG range 151ndash500) (119873 and ) 15 (484)

E3 (CTG range 501ndash1000) (119873 and ) 12(3874)

E4 (CTG range gt 1000) (119873 and ) 3 (97)MIRS stageStage 1 (119873 and ) 3 (97)Stage 2 (119873 and ) 13 (419)Stage 3 (119873 and ) 11 (355)Stage 4 (119873 and ) 4 (128)DM1 = myotonic dystrophy type 1 DMPK = myotonic dystrophy proteinkinase IDMC = international myotonic dystrophy consortium and MIRS =muscular impairment rating scale

part of this patientsrsquo cohort were previously presented in anindependent study [6] Twenty-six healthy participants wererecruited from Santa Lucia Foundation in Rome (Italy) Asdetailed below CTG expansion size within the DMPK genewas assessed for DM1 participants and used to classify themaccording to the International Myotonic Dystrophy Consor-tium nomenclature [22] Demographic characteristics of theparticipants are summarized in Table 1 Principal geneticand clinical characteristics of DM1 patients are summarizedin Table 2 All participants were right-handed as assessedby the Edinburgh Handedness Inventory [23] All subjectsunderwent clinical assessment to exclude the presence ofmajor systemic and neurological illnesses in controls andpathologies different from known comorbidities in DM1patients The study was approved by the Ethical Committeeof Santa Lucia Foundation and written informed consent

Neural Plasticity 3

was obtained from all participants before study initiation Allprocedures performed in this study were in accordance withthe 1964 Helsinki declaration and its later amendments orcomparable ethical standards

22 Genetic Assessment Normal and protomutated alleleswere analysed using ldquotouchdownrdquo PCR on DNA obtainedfrom peripheral blood leukocytes (PBL) Briefly 50 pg ofPBL-DNA was amplified in a 20120583L volume with fluorescentlabelled primer 101 and primer 102 Reactions were cycledthrough eight rounds at 94∘C (3010158401015840) 68∘C (3010158401015840) (minus1∘C percycle) and 72∘C (3010158401015840) followed by 30 rounds at 94∘C(3010158401015840) 60∘C (3010158401015840) and 72∘C (3010158401015840) PCR products werethen analysed using an ABI PRISM 310 Genetic AnalyzerDetermination of expanded alleles was performed on 10 pgof PBL-DNA which underwent XL-PCR1 and 1 agarosegel electrophoresis PCR products were analysed by Southernblotting with subsequent hybridization to a 32P radiolabeled(CTG) 7-oligonucleotide probe anddetected using autoradio-graphy

23 Neuropsychological Assessment All participants under-went an extensive neuropsychological battery covering allcognitive domains which included (a) Mini Mental StateExamination (MMSE) [24] (b) Reyrsquos 15-Word List (Immedi-ate and 15-min Delayed recall) [25] to assess episodic verbalmemory (c) Digit Span and Corsi Block Tapping task [26]forward and backward as measures of short-term memory(d) naming objects and verbs subtests of the BADA (ldquoBatteriaper lrsquoAnalisi dei Deficit Afasicirdquo Italian for ldquoBattery for theanalysis of aphasic deficitsrdquo) [27] to assess language abilities(e) Ravenrsquos Coloured Progressive Matrices [25] to assessreasoning (f) copy of simple drawings with and withoutlandmarks [25] to evaluate constructional abilities and (g)Phonological Word Fluency [25] and Stroop test [28] toexplore executive functions Italian normative data were usedfor all tests both for score adjustment (gender age and educa-tion) and for defining normality cut-off scores (95 toleranceinterval) A series of thirteen one-way ANOVAs were usedto test differences in cognitive performance between patientsand controls (119901 values lt 0004 after Bonferronirsquos correction)Additionally in the patient group MMSE scores were cor-related with CTG triplet expansion age of clinical onsetand Muscular Impairment Rating Scale (MIRS) [29] scores(119901 values lt 002 after Bonferronirsquos correction) Statisticalanalyses were performed using SPSS-20 (SPSS Inc ChicagoIllinois)

24 Image Acquisition and Preprocessing of Resting-StatefMRI All participants underwent MRI at 3 T including(1) 3D Modified Driven Equilibrium Fourier Transform(MDEFT) scan (TR = 1338ms TE = 24ms matrix = 256 times224 119899 slices = 176 and thickness = 1mm) (2) T2lowast weightedecho planar imaging (EPI) sensitized to blood oxygenationlevel dependent (BOLD) contrast (TR = 2080ms TE = 30ms32 axial slices parallel to AC-PC line matrix = 64 times 64 pixelsize = 3 times 3mm2 slice thickness = 25mm and flip angle 70∘)for resting-state fMRI (RS-fMRI) BOLD EPI images were

collected during rest for 7min and 20 s resulting in a totalof 220 volumes During this acquisition participants wereinstructed to keep their eyes closed not to think of anythingin particular and not to fall asleep EPI images were pre-processed for resting-state fMRI using Statistical ParametricMapping 8 (SPM8 httpwwwfilionuclacukspm) and in-house MATLAB scripts

The first 4 volumes of each fMRI time series werediscarded to allow for T1 equilibration effects then imagesunderwent headmotion correction (using the standard SPM8realignment algorithm) compensation for slice-dependenttime shifts and coregistration to the corresponding MDEFTEach MDEFT-volume was segmented using the standardSPM8 algorithm and the resulting grey matter images wereused to compute each participantrsquos total grey matter volumeSegmentation derived normalization parameters were usedto warp the motion and slice-time corrected EPI imagesinto Montreal Neurological Institute (MNI) coordinates In-house software was used to remove the global temporal driftusing a 3rd-order polynomial fit Data were then filteredby regressing out movement vectors average white matterand cerebrospinal fluid signal EPI images were then filteredusing a phase-insensitive band-pass filter (passband 001ndash008Hz) to reduce effects of low frequency drift and highfrequency physiological noise and then smoothed with an8mm3 FWHM 3D Gaussian Kernel

25 Construction of Connectivity Matrices In order to definebrain nodes for each participant the whole brain wasparcellated into 116 regions of interests (ROIs) applying anautomated anatomical labelling (AAL) atlas adapted to theEPI space using linear registration Each ROI corresponds toa node of the network and its mean time course was calcu-lated as the average of the fMRI time series from all voxelswithin the region Correlationmatrices (ie adjacencymatri-ces) were then obtained calculating the correlation betweenall pairs of ROI mean signals The resulting adjacencymatrices were analysed using 2 complementary approaches(1) network-based statistics [30] which compares betweengroups the strength of connectivity for each pair of nodesin the network thus assessing the presence of subnetworkaltered connectivity in patients and (2) graph theory whichcharacterizes the network using special indices describing itsshape and properties including efficiency and resilience toattack (Brain Connectivity Toolbox) [19]

26 Networks-Based Analysis This analysis was preformedusing the ldquonetworks-based statisticsrdquo (NBS) tool developedby Zalesky and coauthors [30] and allowed testing for anysignificant difference of inter-Nodal functional connectiv-ity (in all possible pairwise associations) between DM1patients and controls Between-group comparison was basedon a two-sample 119905-test using 50000 permutations andsetting the significant 119901 value at 0005 (NBS-connectomecorrected for multiple comparisons) Nodes with highernumber of connections were considered more crucial forinformation processing and defined as ldquohubsrdquo According toprevious literature [31] we considered ldquohubsrdquo those nodes

4 Neural Plasticity

whose number of inter-Nodal connections exceeded onestandard deviation the mean number of connections forevery node within the significant network (estimated byNBS)

Then using the NBS toolbox for each DM1 patientwe extracted the ldquoextent of connectivityrdquo (a measure ofconnectivity strength within disrupted networks) to be usedfor correlations with clinicalgenetic and neuropsychologicalvariables For the choice of the most appropriate statisti-cal approach (parametric versus nonparametric correlationtest) the normality of data distribution was assessed usingthe Shapiro-Wilk test (119882) MIRS score and CTG tripletexpansion were chosen as clinicalgenetic variables in thecorrelation analyses with extent of connectivity (Bonferronirsquoscorrection 120572 = 0052 119901 = 0025) All neuropsychologicalscores were used as cognitive variables for the same corre-lation analyses (Bonferronirsquos correction 120572 = 00513 119901 =0004)

27 Graph Theory Analysis A connectivity matrix is a repre-sentation of a graph where the nodes correspond to the rowsand columns of the matrix and the edges are characterizedby the values at the intersection of rows and columns Weused Brain Connectivity Toolbox [19] and MATLAB customscripts to explore global and local topological propertiesof each participantrsquos brain Undirected binary connectivitymatrices were built as thresholding adjacency matrices withdifferent correlation values for each subject The reason isthat the choice of a single common threshold across differentsubjects leads to comparing networks with different densitieswhere density is defined as the ratio between the total numberof edges and the maximum possible number of edges inthe network In order to prevent spurious differences inthe network topologies due to different density values [32]multiple correlation thresholds were used for each subjectobtaining several connectivity matrices for each subject in aspecific density range In thisway subjects could be comparedby means of connectivity matrices with the same densityvalues The density range was determined checking at eachdensity value both the absence of isolated nodes and thepresence of small-world properties (ie most nodes can bereached from any other by a small number of steps) Inour case the lower and the upper limits of this range arerespectively 019 and 047 We chose to use a range step of001 in line with other clinical studies [33 34]

Details of graph theory and its application to brainnetworks can be found in previous papers [19 20 35] togetherwith a full description of all the indices that can be derivedfrom it For the purposes of the current investigation wefocused on the topological measures with a more directclinical interpretation [33ndash38]

As global metrics mean Clustering Coefficient (ie thefraction of one nodersquos neighbours that are also neighboursof each other) Characteristic Path Length (ie the averagelength of sequences of nodes that form routes) Assortativity(an index of the likelihood for high-degree nodes to be linkedtogether) and Modularity (a measure of the presence of acommunity structure in a network) were calculated [19] In

order to further avoid spurious results Normalized Cluster-ing Coefficient and path length were also employedThe nor-malization was performed evaluating the ratio between theoriginal metrics and those derived from random networksThen the ratio between the normalized mean ClusteringCoefficient and the normalized Characteristic Path Lengthalso known as Small-Worldness was calculated This indexis associated with both processing high efficiency and lowwiring cost [35] All global metrics describe brain abilities toprocess information in dedicated regions (segregation) andto coordinate this distributed processing to perform complexactivities (integration)

With respect to local metrics we considered Betweennesscentrality Nodal degree and Nodal efficiency Betweennesscentrality is defined as the fraction of all the shortest pathspassing through a given node Nodal degree expresses thenumber of connections for each node Nodal efficiency isinversely related to each nodersquos paths length and identifiesthe less efficient nodes along certain routes Higher valuesof these topological metrics identify key-nodes for brainintegration and resilience As explained above we lookedat a range of densities for each connectivity matrix andtherefore obtained a range for each topological parametereach corresponding to a different density value First weexamined for each density value the differences betweenthe two groups Then in order to obtain scalar valuesfor each metric the area under the curve (AUC) of thedensity distribution was calculated For each consideredglobal and local measure of connectivity a two-sample 119905-testwas used to assess group differences between DM1 patientsand controls For further analysis as an alternative to theAUC approach we averaged local and global measures acrossdensities to compute the ldquoindividual mean valuesrdquo (Imv-)These global measures of connectivity (Imv-Clustering Coef-ficient (Imv-C) the Imv-Normalized Clustering Coefficient(Imv-NC) the Imv-Characteristic Path Length (Imv-PL)the Imv-Normalized Characteristic Path Length (Imv-NPL)the Imv-Small-Worldness (Imv-SWN) the Imv-Modularity(Imv-M) and the Imv-Assortativity (Imv-A) the Imv-Nodaldegree (Imv-ND) the Imv-Betweenness centrality (Imv-BC) and the Imv-Nodal efficiency (Imv-NE)) were usedfor correlations with patientsrsquo MIRSCTG triplets expansion(Bonferronirsquos correction 120572 = 0052 119901 = 0025) and finallywith neuropsychological variables (Bonferronirsquos correction120572 = 00513 119901 = 0004) Moreover to define an appropriatestatistical analysis the Shapiro-Wilk test was used again

3 Results

31 Demographic Clinical and Neuropsychological Character-istics Patients and controls were not significantly different inage gender or years of formal education (F

155= 303 chi-

square = 275 and F155

= 352 resp all 119901 = ns) (Table 1)Table 2 shows the genetic and clinical characteristics of allrecruited patients As reported in Table 2 most DM1 patients(19 out of 31 612) had an adulthood-onset of disease 12out of 31 patients (387) had a childhood-onset Followingthe guidelines of the Myotonic Dystrophy Consortium [22]

Neural Plasticity 5

Table 3 Performance obtained by DM1 and HS groups on neuropsychological testing

Domain DM1 patients HS119901 valuea

Testsubtest 119873 = 31 119873 = 26

Cognitive efficiencyMMSE (normal cut-off ge 238) 279 (18) 297 (06) 0000Verbal episodic long-term memory15-word list

(i) Immediate recall (cut-off ge 285) 432 (60) 449 (95) ns(ii) Delayed recall (cut-off ge 46) 92 (32) 98 (28) ns

Verbal short-term memoryDigit span forward (cut-off ge 37) 53 (22) 61 (13) nsDigit span backward 49 (51) 51 (10) nsVisuospatial short-term memoryCorsi span forward (cut-off ge 35) 53 (34) 55 (09) nsCorsi span backward 44 (15) 55 (09) nsLanguageNaming of objects (cut-off ge 22) 273 (64) 289 (13) nsNaming of verbs (cut-off ge 22) 254 (61) 261 (20) nsReasoningRavenrsquos Coloured Progressive Matrices (cut-off ge 189) 246 (86) 325 (26) 0001Constructional praxisCopy of drawings (cut-off ge 71) 99 (58) 113 (12) nsCopy of drawings with landmarks (cut-off ge 618) 574 (197) 690 (18) nsExecutive functionsPhonological word fluency (cut-off ge 173) 301 (177) 360 (84) nsStroop-interference (cut-off le 369) 302 (151) mdash mdashDM1 = myotonic dystrophy type 1 HS = healthy subjectsFor each group of studied subjects the table shows the performance scores means (SDs) obtained on neuropsychological testing For each administered testappropriate adjustments for gender age and education were applied according to the Italian normative data Available cut-off scores of normality (ge95 ofthe lower tolerance limit of the normal population distribution) are also reported for each test aOne-way ANOVA

one out of 31 patients (30) was classified as E1 type 15out of 31 (484) were classified as E2 type 12 out of 31(387) were classified as E3 type and finally 3 out of 31(97) were classified as E4 type According to MIRS diseaseclassification 3 out of 31 (86) were at the first stage ofdisease 13 out of 31 (419) were at the second disease stage11 out of 31 (354) were at the third stage and 4 out of31 (129) were at the fourth stage CTG triplet expansionwas negatively associated with years of formal education (119903 =minus062 119901 = 0002) and age at disease onset (119903 = minus072 119901 lt0001) No significant correlationwas found between patientsrsquoMIRS and MMSE scores

All patients with DM1 reported a normal MMSE score(range 26ndash30) though significantly lower thanHS Exploringspecific cognitive domains (Table 3) DM1 patients performedsignificantly worse than controls on a visuospatial task only(Ravenrsquos Coloured Progressive Matrices F

155= 161 119901 lt

0001)

32 Resting-State fMRI

321 Network-Based Analysis NBS analysis showed a sig-nificant reduction of connectivity in patients with DM1compared to HS in a large brain network formed by 51

different nodes and 83 edges (Figure 1 grey blue andgreennodes together) Conversely no significant reduction inconnectivity was found in HS compared to DM1 patients Toisolate a more restricted pattern of disconnection (ie core ofthe dysfunctional network) we applied an extra Bonferronirsquoscorrection for multiple comparisons [120572 = 005number ofparticipants (119873 = 57)119901 = 00008]This119901 value correspondsin Studentrsquos 119879 distribution table to 119905-values ge 324 with55 degrees of freedom (ie dof = 119873 minus 2 where 119873 isthe number of participants) The core of the dysfunctionalnetwork included therefore pairs of nodes whose statisticalthreshold corresponded to 119905-values ge 324 resulting in 37nodes and 47 edges (Figure 1 blue and green nodes andTable 4) Within this ldquocorerdquo dysfunctional network the meannumber of connections was 38 with a standard deviationof 25 According to Materials and Methods we defined asldquohubsrdquo all nodes with at least 7 inter-Nodal connections

Upon comparing DM1 patients to controls dysfunctionalldquohubsrdquo were located in the bilateral anterior cingulum (show-ing connectionswith 19 other brain regions) the orbitofrontalcortex (showing connections with 14 other brain regions)and the right parahippocampal gyrus (showing connectionswith 7 other regions) Those nodes surviving Bonferronirsquoscorrection but whose number of inter-Nodal connectivity

6 Neural Plasticity

R R

Entire dysfunctional networkHubsPeripheral nodes

Figure 1 Reduced connectivity ofDM1 brains obtained by network-based analysis Widespread pattern of functional brain disconnec-tion in patients with DM1 compared to healthy subjects (grey blueand green nodes) When Bonferronirsquos correction was applied toidentify the most critical nodes of this dysfunctional network twopopulations were identified the ldquohubsrdquo (blue) and the peripheralnodes (green) ldquoHubsrdquo characterized by the largest number ofconnections were located in the anterior cingulum and in theorbitofrontal cortex bilaterally and in the right parahippocampalgyrus Peripheral nodes (green) characterized by a smaller numberof connections were mainly located in the prefrontal temporal andparietal cortices and in the cerebellum In the picture nodersquos size isproportional to the number of its connections The brain networkwas visualized using the BrainNet Viewer (httpswwwnitrcorgprojectsbnv) [21] See text for further details R = Right

was lower than 7 were defined as peripheral nodes (Figure 1green nodes) In the comparison between DM1 patientsand controls the peripheral nodes showing a significantreduction of connectivity included the prefrontal temporalparietal and cerebellar regions

In both groups patients and controls the normality ofdistribution of the extent of networkrsquos connectivity measurewas not satisfied (in DM1119882 = 075 119901 lt 0001 in HS119882 =063 119901 lt 0001) With respect to DM1 patients for whomcorrelations were investigated between connectivity and cog-nitive performance this extent of networkrsquos connectivity waspositively associated with patientsrsquo scores at Ravenrsquos ColouredProgressive Matrices (119877 = 061 119901 = 0001)

322 Graph Theory Analysis

Global Measures of Connectivity DM1 patients and controlsdid not differ in any considered AUC global measure In bothgroups patients and controls the normality of distributionof the global connectivity measures was not satisfied (datanot shown) In the DM1 group we found significant positivecorrelations between MIRS scores and Imv-NC (119877 = 048119901 = 0014) and Imv-SWN (119877 = 052 119901 = 0008) anda negative correlation between MIRS scores and the Imv-A(119877 = minus046 119901 = 0019)

Local Measures of Connectivity Upon considering the Nodaldegree (Figure 2(a) Supplementary Table in SupplementaryMaterial available online at httpdxdoiorg10115520162696085) DM1 patients showed a significant reduction in

Table 4 Between-groups difference of functional connectivity intopairwise brain regions

Pairwise brain regions 119905-valueslowast

R superior frontal gyrus (orbital part)harr R middlefrontal gyrus 367

L olfactory cortexharr L orbitofrontal gyrus 385L orbitofrontal gyrusharr L rectus gyrus 470L orbitofrontal gyrusharr R rectus gyrus 482L superior frontal gyrus (orbital part)harr L cingulum(anterior part) 417

L olfactory cortexharr L cingulum (anterior part) 430R olfactory cortexharr L cingulum (anterior part) 373L rectus gyrusharr L cingulum (anterior part) 443R rectus gyrusharr L cingulum (anterior part) 437L superior frontal gyrus (orbital part)harr R cingulum(anterior part) 384

R superior frontal gyrus (orbital part)harr R cingulum(anterior part) 363

L olfactory cortexharr R cingulum (anterior part) 381R rectus gyrusharr R cingulum (anterior part) 392R orbitofrontal gyrusharr L parahippocampal gyrus 371L orbitofrontal gyrusharr R parahippocampal gyrus 368R orbitofrontal gyrusharr R parahippocampal gyrus 422R olfactory cortexharr L amygdala 470L orbitofrontal gyrusharr R amygdala 385R orbitofrontal gyrusharr R amygdala 386R amygdalaharr L occipital gyrus 353L superior frontal gyrusharr L caudate 368R superior frontal gyrusharr L caudate 349R rectus gyrusharr L caudate 372L superior frontal gyrus (medial part)harr L pallidum 364R orbitofrontal gyrusharr L pallidum 348L superior frontal gyrus (medial part)harr R pallidum 361R superior frontal gyrus (medial part)harr R pallidum 362R parahippocampal gyrusharr R middle temporal gyrus 353R amygdalaharr R middle temporal gyrus 429R superior frontal gyrus (medial part)harr L temporalpole 383

R orbitofrontal gyrusharr L temporal pole 444L orbitofrontal gyrusharr L inferior temporal gyrus 360L inferior parietal gyrusharr L inferior temporal gyrus 351R orbitofrontal gyrusharr R inferior temporal gyrus 359R orbitofrontal gyrusharr L cerebellum (Crus1) 384R orbitofrontal gyrusharr L cerebellum (Crus2) 465R inferior frontal gyrusharr L cerebellum (Crus2) 366L cingulum (posterior part)harr L cerebellum (Lobule 9) 440R cingulum (posterior part)harr L cerebellum (Lobule 9) 434R precuneus R harr L cerebellum (Lobule 9) 355L cingulum (anterior part)harr L cerebellum (Lobule 9) 412R cingulum (anterior part)harr R cerebellum (Lobule 9) 366L cingulum (posterior part)harr R cerebellum (Lobule 9) 432R cingulum (posterior part)harr R cerebellum (Lobule 9) 353L cingulum (anterior part)harr vermis 352R cingulum (anterior part)harr vermis 347L cingulum (posterior part)harr vermis 381R cingulum (posterior part)harr vermis 358lowastTwo-sample 119905-test with 55 degrees of freedom 119905-values are reported R =right L = left andharr = bidirectional connections

Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)

R RBetweenness centrality

DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)

Figure 2 Abnormal topological properties of DM1 brains obtained by the graph theory analysis Patients with DM1 compared to healthysubjects showed both decreased (blue nodes) and increased (red nodes) functional connectivity in centrality measures Upon considering theNodal degree (a) and Betweenness centrality (b) DM1 patients showed two distinct patterns one located more anteriorly characterized bydecreased connectivity (blue nodes) and onemore posterior characterized by increased connectivity (red nodes)The formermainly involvedthe superior frontal and orbitofrontal gyri bilaterally The latter involved the supplementary motor area and the cerebellum Nodal efficiency(c) was reduced in the right superior frontal gyrus in the right orbitofrontal gyrus and in the left angular gyrus The brain network wasvisualized using the BrainNet Viewer (httpswwwnitrcorgprojectsbnv) [21] See text for further details R = Right

the superior frontal and in the orbitofrontal gyrus bilaterallyConversely this samemeasure was increased inDM1 patientscompared to controls in the supplementary motor areabilaterally and in the cerebellum

Upon considering Betweenness centrality (Figure 2(b)Supplementary Table) DM1 patients revealed reduced con-nectivity compared to controls in the right superior frontalgyrus in the right inferior parietal gyrus and in the rightputamen In contrast DM1 patients showed a significantincrease of connectivity in the right paracentral lobule in theright CRUS-I and in the right Lobule 10 of the cerebellum

Moreover upon considering Nodal efficiency (Fig-ure 2(c) Supplementary Table) DM1 patients revealedreduced connectivity compared to controls in the right

superior frontal gyrus in the right orbitofrontal cortex andin the left angular gyrus

Finally we found a significant negative correlationbetween the Nodal degree (Imv-ND) of right CRUS-1 and thepatientsrsquo CTG triplet expansions (119877 = minus054 119901 = 0003)

4 Discussion

Here we provide the first evidence for widespread abnormal-ities in functional brain topological properties of DM1 brainsThis fits with previous evidence of widespread structuralalterations in DM1 patients which are dominated by whitematter involvement [7ndash11]

8 Neural Plasticity

The majority of our patients had an adult or childhooddisease onset and their muscular impairmentsrsquo severity wasmild to moderate Consistent with previous studies [6 710] there was no significant association between patientsrsquoCTG triplet expansion and MIRS scores while a negativeassociation was found between CTG triplet expansion andpatientsrsquo age at disease onset This indicates that a moresevere genetic load associates with an earlier clinical onsetas supported by a previous study by our own group [7]On the other hand it is known that in other neurologicalconditions brain function does not linearly respond to theaccumulation of structural abnormalities with a significantmodulation determined by environmental factors In thisview topological brain characteristics (ie measures of net-worksrsquo integrity and efficiency) might reflect better thanstructural information the global modifications occurring toDM1 brains

We first performedNBS analysis which revealed reducedconnectivity in DM1 patients within a widespread brainnetwork involving several nodes and edges with the anteriorcingulate and the orbitofrontal cortices as core regionsTheseregions are wired with several other nodes of the associationcortex thus playing the role of ldquohubsrdquo It is interesting toobserve that although our DM1 patients performed withinthe range of normality in most cognitive tests they reportedpoor scores at Ravenrsquos Coloured Progressive Matrices Thistest assesses visuospatial reasoning which is also regarded asa component of general intelligence [39] andmainly engagesthe orbitofrontal and parietal cortices [40] Consistently ourDM1 patientsrsquo performance at Ravenrsquos Coloured ProgressiveMatrices test correlated with their extent of brain connec-tivity Taken altogether these results indicate that abnormalbrain connectivity explains at least partially the patientsrsquodeficits in reasoning

Upon considering graph theoretical measures DM1patients did not reveal significant differences in global topo-logical indices with respect to controls Changes to globalmeasures typically reflect the presence of considerable braindisorganization such as that observed in patients with severedementia [41] The lack of significant reductions in globalsegregationintegration measures in DM1 patients [41] istherefore consistent with their relative preservation of generalcognitive efficiency Nevertheless some global measures ofconnectivity (ie Assortativity Normalized Clustering Coef-ficient and Small-Worldness) were associated with patientsrsquoMIRS scores The MIRS score returns a global measure ofdisability in DM1 patients It is based on the assessmentof progression of muscular impairment and in principle itshould reflect also some contribution of CNS involvementIndeed it was previously shown that MIRS scores correlatewith the extension of white matter damage in DM1 patients[7 10] Among global measures of connectivity Assortativity[19] which was negatively associated with our DM1 patientsrsquoMIRS scores has been previously regarded either as a mea-sure of network vulnerability to pathological insults [42] or asan index of brain tissue resilience Additionally NormalizedClustering Coefficient and Small-Worldness which werepositively associated with our patientsrsquo MIRS scores are con-sidered as measures of network segregation and integration

respectively Taken altogether higher network segregationin the absence of a sufficient integration might reflect thepresence of dysfunctional communication between nodes[20] Considering this strict association between measuresof global brain connectivity and patientsrsquo MIRS scoreswe believe they might be useful biomarkers for patientsrsquomonitoring in clinical routine and trials

Upon considering the local measures of brain connectiv-ity we found two distinct patterns in DM1 patients namelya more anterior decrease and a more posterior increase ofconnectivity With respect to the anterior pattern consistentwith the findings obtained by NBS analysis DM1 patientsshowed decreased connectivity in prefrontal (orbitofrontaland superior frontal gyrus) and parietal (inferior parietaland angular gyrus) regions As mentioned above these brainregions are implicated in higher-level functions and mayaccount for patientsrsquo cognitive and behavioural symptoms[6 7] Consistently a selective pattern of topological brainalterations including Nodal degree Betweenness centralityand Nodal efficiency was found in a selection of the nodesbelonging to this network This indicates that some of thesenodes which are the most ldquocentralrdquo for an efficient transfer ofinformation are affected The reduction of Nodal efficiencyand Betweenness centrality that we found in patientsrsquo angulargyrus and inferior parietal gyrus respectively is only inapparent contrast with the increase of functional connectivitythat we previously observed in a critical node of patientsrsquodefault mode network [6] These brain regions are locatedin neighbour but different brain areas and these differentfindings seem to be complementary to each other rather thanin contrast Functional connectivity estimated by indepen-dent component analysis [6] returns ameasure of segregationsimilarly to that expressed by Nodal degree and ClusteringCoefficient as obtained by the graph theory ConverselyNodal efficiency and Betweenness centrality (found to bereduced here in DM1 patients) are regarded as measuresof integration [43] In this view as previously suggested byvan den Heuvel and Sporns [43] segregation and integrationmeasures produce complementary information and ourprevious and current findings are likely to detect differentaspects of the pathophysiological processes occurring toDM1brains

With respect to the posterior pattern DM1 patientsshowed increased connectivity in the supplementary motorarea (SMA) and in the right cerebellum (CRUS-1 and lobule10) Similar abnormalities have also been observed in patientswith autism spectrum disorders (ASD) [44ndash46] In this casethe atypical pattern of increased connectivity between thecerebellum and the sensorimotor areas was interpreted as apossible substrate for the repetitive stereotyped behavioursobserved in autism [44] Interestingly autism-like traits havealso been previously reported in patients with congenitalDM1 [47] Moreover in the current work we also founda significant association between patientsrsquo genetic load andtheir Nodal degree connectivity in CRUS-I The SMA theCRUS-1 and cerebellar lobule 10 are all regions implicated inldquointernally generatedrdquo planning of movements in the plan-ning of sequences of movements and in motor coordination[48 49]We believe that the severity of abnormal connectivity

Neural Plasticity 9

within this set of regions might be relevant for the clinicalinterpretation of motor impairment in patients with DM1which can differ across individuals Motor impairment inDM1 is indeed predominantly due to muscular weaknessHowever a recent fMRI study reported an associationbetween myotonia and SMA activation in patients with DM1[50] thus suggesting CNS contribution to patientsrsquo motordeficits [50] In this view we speculate that our findingsof abnormally high connectivity within the motor networkof DM1 patients might represent a compensatory albeitnonefficient or maladaptive mechanism of brain plasticityto contrast muscular impairment This might be of clinicalinterest for tailored neurorehabilitation programs in DM1

In conclusion the current study describes for the firsttime the topological properties of brain functional net-works in patients with DM1 Though preliminary our datacontribute to clarifying some relevant pathophysiologicalaspects of DM1 with a focus on both motor and nonmotorsymptoms Not only is this of speculative interest but itintroduces a new potential tool for patientsrsquo assessment andmonitoring Finally this work opens an important ques-tion on the possibility of implementing novel therapeuticapproaches in clinical settings by using for instance thetranscranial magnetic stimulation (TMS) or the transcranialdirect current stimulation (TDCS) These techniques thatexploit the unique plasticity of the human brain have alreadyproven their usefulness inmodulating brain connectivity andin improving the clinical outcome in various neurologicalpatients [51ndash53]

Competing Interests

None of the Authors has any conflict of interests to disclose

Acknowledgments

TheNeuroimaging Laboratory is in part supported by grantsobtained by the Italian Ministry of Health

References

[1] G Meola and R Cardani ldquoMyotonic dystrophies an update onclinical aspects genetic pathology and molecular pathomech-anismsrdquo Biochimica et Biophysica Acta (BBA)mdashMolecular Basisof Disease vol 1852 no 4 pp 594ndash606 2015

[2] J D Brook M E McCurrach H G Harley et al ldquoMolecularbasis of myotonic dystrophy expansion of a trinucleotide(CTG) repeat at the 31015840 end of a transcript encoding a proteinkinase family memberrdquo Cell vol 68 pp 799ndash808 1992

[3] G Meola V Sansone D Perani et al ldquoExecutive dysfunctionand avoidant personality trait in myotonic dystrophy type 1(DM-1) and in proximal myotonic myopathy (PROMMDM-2)rdquo Neuromuscular Disorders vol 13 no 10 pp 813ndash821 2003

[4] A Modoni G Silvestri M G Vita D Quaranta P A Tonaliand C Marra ldquoCognitive impairment in myotonic dystrophytype 1 (DM1) a longitudinal follow-up studyrdquo Journal of Neu-rology vol 255 no 11 pp 1737ndash1742 2008

[5] M Douniol A Jacquette D Cohen et al ldquoPsychiatric andcognitive phenotype of childhood myotonic dystrophy type 1rdquo

DevelopmentalMedicine and Child Neurology vol 54 no 10 pp905ndash911 2012

[6] L Serra G Silvestri A Petrucci et al ldquoAbnormal functionalbrain connectivity and personality traits in myotonic dystrophytype 1rdquoThe JAMA Neurology vol 71 no 5 pp 603ndash611 2014

[7] L Serra A Petrucci B Spano et al ldquoHow genetics affectsthe brain to produce higher-level dysfunctions in myotonicdystrophy type 1rdquo Functional Neurology vol 30 no 1 pp 21ndash31 2015

[8] A Di Costanzo F Di Salle L Santoro V Bonavita and GTedeschi ldquoBrain MRI features of congenital- and adult-formmyotonic dystrophy type 1 case-control studyrdquo NeuromuscularDisorders vol 12 no 5 pp 476ndash483 2002

[9] A Di Costanzo F Di Salle L Santoro A Tessitore V Bonavitaand G Tedeschi ldquoPattern and significance of white matterabnormalities in myotonic dystrophy type 1 an MRI studyrdquoJournal of Neurology vol 249 no 9 pp 1175ndash1182 2002

[10] M Minnerop B Weber J C Schoene-Bake et al ldquoThe brain inmyotonic dystrophy 1 and 2 evidence for a predominant whitematter diseaserdquo Brain vol 134 part 12 pp 3530ndash3546 2011

[11] J RWozniak BAMueller E EWardKO Lim and JWDayldquoWhite matter abnormalities and neurocognitive correlates inchildren and adolescents with myotonic dystrophy type 1 adiffusion tensor imaging studyrdquo Neuromuscular Disorders vol21 no 2 pp 89ndash96 2011

[12] B Echenne and G Bassez ldquoCongenital and infantile myotonicdystrophyrdquo Handbook of Clinical Neurology vol 113 pp 1387ndash1393 2013

[13] G Meola and V Sansone ldquoCerebral involvement in myotonicdystrophiesrdquoMuscle andNerve vol 36 no 3 pp 294ndash306 2007

[14] T Gili M Cercignani L Serra et al ldquoRegional brain atrophyand functional disconnection across Alzheimerrsquos disease evolu-tionrdquo Journal of Neurology Neurosurgery and Psychiatry vol 82no 1 pp 58ndash66 2011

[15] WKoch S Teipel SMueller et al ldquoDiagnostic power of defaultmode network resting state fMRI in the detection ofAlzheimerrsquosdiseaserdquoNeurobiology of Aging vol 33 no 3 pp 466ndash478 2012

[16] M P van den Heuvel and H E Hulshoff Pol ldquoExploringthe brain network a review on resting-state fMRI functionalconnectivityrdquo European Neuropsychopharmacology vol 20 no8 pp 519ndash534 2010

[17] M E Raichle and A Z Snyder ldquoA default mode of brainfunction a brief history of an evolving ideardquo NeuroImage vol37 no 4 pp 1083ndash1090 2007

[18] R L Buckner J R Andrews-Hanna and D L Schacter ldquoThebrainrsquos default network anatomy function and relevance todiseaserdquo Annals of the New York Academy of Sciences vol 1124pp 1ndash38 2008

[19] M Rubinov and O Sporns ldquoComplex network measures ofbrain connectivity uses and interpretationsrdquo NeuroImage vol52 no 3 pp 1059ndash1069 2010

[20] E Bullmore and O Sporns ldquoComplex brain networks graphtheoretical analysis of structural and functional systemsrdquoNature Reviews Neuroscience vol 10 pp 186ndash198 2009

[21] M Xia J Wang and Y He ldquoBrainNet Viewer a networkvisualization tool for human brain connectomicsrdquo PLoS ONEvol 8 no 7 Article ID e68910 2013

[22] International Myotonic Dystrophy Consortium (IDMC) ldquoNewnomenclature andDNA testing guidelines for myotonic dystro-phy type 1 (DM1)rdquoNeurology vol 54 no 6 pp 1218ndash1221 2000

10 Neural Plasticity

[23] D Busch N Hagemann andN Bender ldquoThe dimensionality ofthe edinburgh handedness inventory an analysis with modelsof the item response theoryrdquo Laterality vol 15 no 6 pp 610ndash628 2010

[24] M F Folstein S E Folstein and P R McHugh ldquolsquoMini-mentalstatersquo A practical method for grading the cognitive state ofpatients for the clinicianrdquo Journal of Psychiatric Research vol12 no 3 pp 189ndash198 1975

[25] G A Carlesimo C Caltagirone and G Gainotti ldquoThe MentalDeterioration Battery normative data diagnostic reliability andqualitative analyses of cognitive impairmentThe Group for theStandardization of the Mental Deterioration Batteryrdquo EuropeanNeurology vol 36 no 6 pp 378ndash384 1996

[26] A Orsini D Grossi E Capitani M Laiacona C Papagnoand G Vallar ldquoVerbal and spatial immediate memory spannormative data from 1355 adults and 1112 childrenrdquo The ItalianJournal of Neurological Science vol 8 pp 539ndash548 1987

[27] G Miceli A Laudanna C Burani and R Capasso Batteria perlrsquoAnalisi dei Deficit Afasici Berdata Ed Milano Italy 1991

[28] P Caffarra G Vezzadini F Dieci A Zonato and A VennerildquoUna versione abbreviata del test di Stroop dati normativi nellapopolazione italianardquo Nuova Rivista di Neurologia vol 12 pp111ndash115 2012

[29] J Mathieu H Boivin D Meunier M Gaudreault and P BeginldquoAssessment of a disease-specific muscular impairment ratingscale in myotonic dystrophyrdquoNeurology vol 56 no 3 pp 336ndash340 2001

[30] A Zalesky A Fornito and E T Bullmore ldquoNetwork-basedstatistic identifying differences in brain networksrdquoNeuroImagevol 53 no 4 pp 1197ndash1207 2010

[31] T Itahashi T Yamada H Watanabe et al ldquoAltered networktopologies and hub organization in adults with autism aresting-state fMRI studyrdquo PLoS ONE vol 9 no 4 Article IDe94115 2014

[32] B CM vanWijk C J Stam and A Daffertshofer ldquoComparingbrain networks of different size and connectivity density usinggraph theoryrdquo PLoS ONE vol 5 no 10 Article ID e13701 2010

[33] Z Liu Y Zhang H Yan et al ldquoAltered topological pat-terns of brain networks in mild cognitive impairment andAlzheimerrsquos disease a resting-state fmri studyrdquo PsychiatryResearchmdashNeuroimaging vol 202 no 2 pp 118ndash125 2012

[34] M A Rocca P Valsasina A Meani A Falini G Comi and MFilippi ldquoImpaired functional integration in multiple sclerosis agraph theory studyrdquo Brain Structure and Function vol 221 no1 pp 115ndash131 2016

[35] D S Bassett and E Bullmore ldquoSmall-world brain networksrdquoThe Neuroscientist vol 12 no 6 pp 512ndash523 2006

[36] A J Lerner P K Ogrocki and P J Thomas ldquoNetwork graphanalysis of category fluency testingrdquo Cognitive and BehavioralNeurology vol 22 no 1 pp 45ndash52 2009

[37] Z Yao Y Zhang L Lin et al ldquoAbnormal cortical networksin mild cognitive impairment and Alzheimerrsquos diseaserdquo PLoSComputational Biology vol 6 no 11 Article ID e1001006 2010

[38] M Shim D-W Kim S-H Lee and C-H Im ldquoDisruptions insmall-world cortical functional connectivity network during anauditory oddball paradigm task in patients with schizophreniardquoSchizophrenia Research vol 156 no 2-3 pp 197ndash203 2014

[39] G Gainotti P DrsquoErme G Villa and C Caltagirone ldquoFocalbrain lesions and intelligence a study with a new version ofRavenrsquos Colored Matricesrdquo Journal of Clinical and ExperimentalNeuropsychology vol 8 no 1 pp 37ndash50 1986

[40] CWendelken ldquoMeta-analysis how does posterior parietal cor-tex contribute to reasoningrdquo Frontiers in Human Neurosciencevol 8 article 1042 2015

[41] Z Dai and Y He ldquoDisrupted structural and functional brainconnectomes in mild cognitive impairment and Alzheimerrsquosdiseaserdquo Neuroscience Bulletin vol 30 no 2 pp 217ndash232 2014

[42] M E J Newman ldquoAssortative mixing in networksrdquo PhysicalReview Letters vol 89 no 20 Article ID 208701 2002

[43] M P van den Heuvel and O Sporns ldquoAn anatomical substratefor integration among functional networks in human cortexrdquoThe Journal of Neuroscience vol 33 no 36 pp 14489ndash145002013

[44] A M DrsquoMello and C J Stoodley ldquoCerebro-cerebellar circuitsin autism spectrum disorderrdquo Frontiers in Neuroscience vol 9article 408 2015

[45] S K Noonan F Haist and R-A Muller ldquoAberrant functionalconnectivity in autism evidence from low-frequency BOLDsignal fluctuationsrdquo Brain Research vol 1262 pp 48ndash63 2009

[46] A J Khan A Nair C L Keown M C Datko A J Lincolnand R-A Muller ldquoCerebro-cerebellar resting-state functionalconnectivity in children and adolescents with autism spectrumdisorderrdquo Biological Psychiatry vol 78 no 9 pp 625ndash634 2015

[47] A-B Ekstrom L Hakenas-Plate L Samuelsson M Tuliniusand E Wentz ldquoAutism spectrum conditons in myotonic dys-trophy type 1 a study on 57 individuals with congenital andchildhood formsrdquo American Journal of Medical Genetics PartB Neuropsychiatric Genetics vol 147 no 6 pp 918ndash926 2008

[48] A N Carlsen J S Eagles and C D MacKinnon ldquoTranscranialdirect current stimulation over the supplementary motor areamodulates the preparatory activation level in the human motorsystemrdquo Behavioural Brain Research vol 279 pp 68ndash75 2015

[49] J C Houk and S P Wise ldquoDistributed modular architectureslinking basal ganglia cerebellum and cerebral cortex their rolein planning and controlling actionrdquo Cerebral Cortex vol 5 no2 pp 95ndash110 1995

[50] A Toth E Lovadi S Komoly et al ldquoCortical involvementduring myotonia in myotonic dystrophy an fMRI studyrdquo ActaNeurologica Scandinavica vol 132 no 1 pp 65ndash72 2015

[51] G Koch P Porcacchia V Ponzo et al ldquoEffects of two weeks ofcerebellar theta burst stimulation in cervical dystonia patientsrdquoBrain Stimulation vol 7 no 4 pp 564ndash572 2014

[52] M Kojovic P Kassavetis M Bologna et al ldquoTranscranial mag-netic stimulation follow-up study in early Parkinsonrsquos disease adecline in compensation with disease progressionrdquoMovementDisorders vol 30 no 8 pp 1098ndash1106 2015

[53] S Cipollari D Veniero C Razzano C Caltagirone G Kochand P Marangolo ldquoCombining TMS-EEG with transcranialdirect current stimulation language treatment in aphasiardquoExpert Review of Neurotherapeutics vol 15 no 7 pp 833ndash8452015

Submit your manuscripts athttpwwwhindawicom

Neurology Research International

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Alzheimerrsquos DiseaseHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

ScientificaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


BioMed Research International


Research and TreatmentSchizophrenia

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Neural Plasticity


Parkinsonrsquos Disease


Research and TreatmentAutism

Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Neuroscience Journal

Epilepsy Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Psychiatry Journal


Computational and Mathematical Methods in Medicine

Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Brain ScienceInternational Journal of

StrokeResearch and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Neurodegenerative Diseases


Journal of

Cardiovascular Psychiatry and NeurologyHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

2 Neural Plasticity

higher-level dysfunctions [3ndash7] DM1 brains have beendemonstrated to be structurally damaged in both tissues thegrey (GM) and white matter (WM) [7ndash11] with a specificanatomical distribution of abnormalities This structuraldamage has been consistently reported across independentstudies [7ndash11] and it was more recently associated with CTGtriplet expansions in theDMPK gene andmeasures of clinicalseverity [7] Abnormal patterns of brain connectivity havealso been reported in DM1 patients and have been demon-strated to account for patientsrsquo personality traits [6] Whilemental retardation is frequently observed in the congenitalform of DM1 [12] the adult forms are typically characterizedby isolated cognitive deficits [7 13]This relative preservationof global cognition contrasts with the pathological andneuroimaging evidence of diffuse WM abnormalities [7ndash10] and with the paradoxical failure of these patients ineveryday life In addition upon investigating functional con-nectivity within the so-called default mode network whosedisruption is typically associated with cognitive impairmentin degenerative dementia [14 15] DM1 patients reveal anincrease of connectivity in some critical nodes [6] Thesedata taken altogether suggest DM1 as neurodevelopmentaldisorder that associates with peculiar rearrangements inneuronal networksrsquo segregation and integration propertiesThese complex alterations which are hardly detectable bystructural brain assessments are likely to account for thedistinctive motor and nonmotor features observed in DM1Resting-state functional MRI (RS-fMRI) [16] is one of themost widely used methods to investigate brain connectivityin neurological and psychiatric diseases [17 18] with theadvantage of not requiring participants to perform any activetask RS-fMRI data can be analysed using different method-ological approaches One promising technique is based onthe whole brain analysis driven by graph theory [19] amathematical approach that describes complex systems asnetworks [20] In simple words the brain is conceptualized asa number of regions (nodes) that are functionally connectedto each other by edges and whose importance and efficiencywithin the whole network are determined by their functionalspecialization (ie segregation) and integration In this viewsome nodes are more critical (ie centrality) for informationprocessing (efficiency in information transferring) and arecalled ldquohubsrdquo Abnormal connectivity between ldquohubsrdquo isbelieved to cause more deficits than that between peripheralnodes [19 20] With these concepts in mind the currentwork aimed at assessing topological properties ofDM1 brainstheir potential relationship with genetics and their abilityin accounting for patientsrsquo motor and nonmotor clinicalfeatures To the best of our knowledge this is the first attemptto investigate brain connectomics in DM1 patients

2 Materials and Methods

21 Participants Thirty-one patients with a molecular diag-nosis of DM1 were recruited from the Neuromuscular andNeurological Rare Diseases Center at San Camillo ForlaniniHospital (Rome Italy) and the Institute of Neurology atthe Catholic University of Rome (Rome Italy) Data from

Table 1 Principal demographic characteristics of studied subjects


HS119873 = 26

119901 value

Mean (SD) age [years] 399 (114) 457 (132) nsa

Gender (FM) 160150 19070 nsb

Mean (SD) years offormal education 124 (22) 140 (33) nsa

aOne-way ANOVA bChi-squareDM1 = myotonic dystrophy type 1 HS = healthy subjects

Table 2 Principal genetic and clinical characteristics of DM1patients


Age at onsetChildhood-onset (age range 6ndash16 years) 12 (387)Adulthood-onset (age range 18ndash60 years) 19 (612)Size of CTG tripletsrsquo expansion on DMPK gene

Mean (SD) [range]6371

(4564)[54ndash2000]

IDMC nomenclatureE1 (CTG range 50ndash150) (119873 and ) 1 (30)E2 (CTG range 151ndash500) (119873 and ) 15 (484)

E3 (CTG range 501ndash1000) (119873 and ) 12(3874)

E4 (CTG range gt 1000) (119873 and ) 3 (97)MIRS stageStage 1 (119873 and ) 3 (97)Stage 2 (119873 and ) 13 (419)Stage 3 (119873 and ) 11 (355)Stage 4 (119873 and ) 4 (128)DM1 = myotonic dystrophy type 1 DMPK = myotonic dystrophy proteinkinase IDMC = international myotonic dystrophy consortium and MIRS =muscular impairment rating scale

part of this patientsrsquo cohort were previously presented in anindependent study [6] Twenty-six healthy participants wererecruited from Santa Lucia Foundation in Rome (Italy) Asdetailed below CTG expansion size within the DMPK genewas assessed for DM1 participants and used to classify themaccording to the International Myotonic Dystrophy Consor-tium nomenclature [22] Demographic characteristics of theparticipants are summarized in Table 1 Principal geneticand clinical characteristics of DM1 patients are summarizedin Table 2 All participants were right-handed as assessedby the Edinburgh Handedness Inventory [23] All subjectsunderwent clinical assessment to exclude the presence ofmajor systemic and neurological illnesses in controls andpathologies different from known comorbidities in DM1patients The study was approved by the Ethical Committeeof Santa Lucia Foundation and written informed consent

Neural Plasticity 3









4 Neural Plasticity








3 Results


155= 303 chi-



Neural Plasticity 5



Testsubtest 119873 = 31 119873 = 26






155= 161 119901 lt

0001)





6 Neural Plasticity

R R
















Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)


DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)







4 Discussion


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


Neural Plasticity 3









4 Neural Plasticity








3 Results


155= 303 chi-



Neural Plasticity 5



Testsubtest 119873 = 31 119873 = 26






155= 161 119901 lt

0001)





6 Neural Plasticity

R R
















Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)


DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)







4 Discussion


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


4 Neural Plasticity








3 Results


155= 303 chi-



Neural Plasticity 5



Testsubtest 119873 = 31 119873 = 26






155= 161 119901 lt

0001)





6 Neural Plasticity

R R
















Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)


DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)







4 Discussion


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


Neural Plasticity 5



Testsubtest 119873 = 31 119873 = 26






155= 161 119901 lt

0001)





6 Neural Plasticity

R R
















Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)


DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)







4 Discussion


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


6 Neural Plasticity

R R
















Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)


DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)







4 Discussion


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


Neural Plasticity 7

R RNodal degree

DM1 lt HSDM1 gt HS

(a)


DM1 lt HSDM1 gt HS

(b)

Nodal efficiencyR R

DM1 lt HS(c)







4 Discussion


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


8 Neural Plasticity







Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of


Neural Plasticity 9



Competing Interests


Acknowledgments


References




































































Neural Plasticity










Psychiatry Journal









Journal of














































Neural Plasticity










Psychiatry Journal









Journal of














Neural Plasticity










Psychiatry Journal









Journal of


Documents

Research Article Brain Connectomics Modification to