Brain Network Analysis Based On Magnetic Resonance Imaging In Multiple System Atrophy Patients With Mild Cognitive Impairment

Objective: Multiple system atrophy(MSA)is a sporadic,adult-onset,progressive neurodegenerative disease characterized by various combinations of autonomic failure,parkinsonism,cerebellar ataxia,and pyramidal signs.Cognitive dysfunction is traditionally believed to be rare in MSA.However,accumulating evidence has indicated that cognitive decline is an integral part of MSA,and the prevalence rates of cognitive impairment(CI)in MSA range from 22% to 37% in autopsy-confirmed MSA patients.In some cases,CI even precede motor symptoms.The spectrum of cognitive function in MSA spans from normal cognition through mild cognitive impairment(MCI)to MSA patients with dementia.Recognizing MCI in MSA is critical in clinical settings,as affected individuals are likely to progress toward dementia.However,the mechanism underlying MCI in MSA remains controversial.Multimodal magnetic resonance imaging(MRI)is comprised of three dimensional T1 weighted imaging and blood oxygen level dependent functional MRI,which can provide a comprehensive picture of the progression of neurodegenerative diseases from a morphological and functional perspective.A battery of multimodal MRI studies indicated that CI in MSA results from primary cortical damage or subcortical degeneration with secondary cortical deafferentation,which remains controversial.Current studies focus on the morphological changes in MSA-MCI,little is known about functional alterations in MSA-MCI.The brain is a unity of functional segregation and functional integration.Specific brain regions have different division of function,and separate brain regions perform cognitive tasks in the form of brain networks through mutual coordination and interaction.The amplitude of low-frequency fluctuation(ALFF)is a reliable and stable metric in rs-f MRI,examining regional brain activity at the voxel level.Previous studies have observed ALFF alterations in Parkinson’s disease(PD),MCI with depression,and early Alzheimer’s disease.Alterations of spontaneous activity can induce corresponding network disruptions.Limited studies utilized seed-based functional connectivity(FC)to investigate network alterations in MSA,and found disruption of CEN and DMN,as well as cerebello-cerebral in MSA.The selections of seed regions based on prior studies or network analysis results.Alterations of functional segregation and integration in MSAMCI is unknown.Among the intrinsic connected networks,the triple brain network is thought to be central to higher cognitive functions,including the central executive network(CEN),default mode network(DMN)and salience network(SN).The equilibrium of triple network is the foundation of normal cognition.The abnormal mode of triple networks has been demonstrated in healthy aging brain,amnestic MCI,Alzheimer’s disease,and psychiatric disorders such as depression and obsessivecompulsive disorder,but it is unclear in MSA.Independent component analysis(ICA)can identify internal brain networks and quantitatively analyze the interactions between networks and within networks,which is an effective method to explore the pattern of cognitive regulation in psychoneurological diseases.Objective: Based on BOLD and 3DT1 sequences,this study utilized functional segregation indicator and network analysis,to explore network disruptions in MSA-MCI.Combined with the triple network model,the intra-network and inter-network interaction of MSA-MCI internal connectivity network were analyzed to explore the pathophysiological mechanism of MSA-MCI and explore the early effective imaging markers of MSA-MCI.Participants and Methods: 44 MSA patients with mild cognitive impairment(MSAMCI),36 MSA patients with normal cognition(MSA-NC)and 40 healthy controls(HC)were included in PART I.The clinical diagnosis of MSA was based on the second consensus clinical criteria.We systematically collected the demographic and clinical information of each participant,including age,gender,years of education,disease duration,and drug dose.For patients,the Hoehn and Yahr(H & Y)scale was used to measure disease stage,and the severity of motor disability was evaluated with the Unified Multi-System Atrophy Scale-Part II(UMSARS-II).All participants underwent neuropsychological assessments including the Mini-Mental State Examination(MMSE),the Montreal Cognitive Assessment(Mo CA),and the Hamilton Depression Scale(HAMD-24).In PART II,70 MSA-MCI,66 MSA-NC and 63 HC were included.The clinical diagnosis of MSA was based on the second and third consensus clinical criteria.Besides data involved in PART I,we further evaluated cognitive function of MSA patients in five cognitive domains.Details are as follows,i)Executive functions: Symbol Digit transformation Test(SDMT),10 points Clock Drawing Test(CDT).ii)Visual space function: Clock Copying Test(CLOX-2),Copy task of the Rey-Osterrieth Complex Figure Test(ROCF);Clock copying test(Clox-2),copy task of the Rey-Osterrieth complex figure test(ROCF);iii)Attention and working memory: Trail Making-Part A(TPT-A),Stroop Color interference Test-C(CWT-C);iv)Memory: The Rey Auditory Verbal Learning Test(RAVLT),Brief Visuospatial memory test-revised(BVMT-R).v)Language: Boston naming test(BNT),category Fluency Test(CFT).1、ALFF and seed-based FC analysis were combined in this part.Firstly,we compared ALFF among 3 groups,and the alterations were correlated with clinical data.The brain regions showing ALFF alterations between MSA-MCI and MSA-NC and correlated with Mo CA scores in MSA patients were selected as seeds.Secondary FC analysis was employed among groups,and the alterations were correlated with clinical data.We also conducted a whole-brain VBM analysis to detect gray matter(GM)volume reduction among groups.2、Based on the findings in PART I and prior literature,we employed ICA analysis in this part,to explore intra-network and inter-network FC among left CEN(LCEN),right CEN(RCEN),anterior anterior DMN,(a DMN),superior-posterior DMN,(sp DMN),inferior-posterior DMN(ip DMN),SN and cerebellar network(CN).We also selected core brain regions in each network as region of interest(ROI).Alterations of FC within and between networks,as well as between ROIs were analyzed and compared among groups,and the alterations were correlated with the global cognitive function,as well as specific cognitive domain.Results: 1.Compared with MSA-NC,MSA-MCI exhibited decreased ALFF in the right dorsal lateral prefrontal cortex(DLPFC)and increased ALFF in the right cerebellar lobule IX and lobule IV-V.In the secondary FC analyses,decreased FC in the left inferior parietal lobe(IPL)was observed when we set the RDLPFC as the seed region.Decreased FC in the bilateral cuneus,left precuneus,and left IPL and increased FC in the right middle temporal gyrus were shown when we set the right cerebellar lobule IX as the seed region.Further,FC of DLPFC-IPL and cerebello-cerebral circuit,as well as ALFF alterations,were significantly correlated with Mo CA scores in MSA patients.We also employed whole-brain voxel-based morphometry(VBM)analysis,but no gray matter atrophy was detected between the patient subgroups.2.In intra-network analysis,compared with MSA-NC,MSA-MCI exhibited altered FC in the RCEN,sp DMN,ip DMN,and CN,showing decreased FC in the right superior frontal gyrus,the left calcarine,the left precuneus,the right temporal gyrus and the right cerebellar IV/V,as well as increased FC in the vermis.In the inter-network analysis,compared with MSANC,MSA-MCI exhibited decreased FC between the LCEN and the a DMN,the CN1 and the RCEN,the SN and the RCEN,and increased FC between the RCEN and the ip DMN.In the inter-ROI analysis,compared with MSA-NC,we found decreased FC between the left middle frontal gyrus and the right insula,between the right angula gyrus and the left medial superior frontal gyrus,between the right angular gyrus and the right insula,between the right cerebellar Crus I and the left precuneus,and increased FC between the left cerebellar lobule VI and the left medial superior frontal gyrus in MSA-MCI.Further,in MSA patients,FC within CEN was positively correlated with Mo CA scores(r =0.2293,p = 0.007),and negatively correlated with TMTA scores in the right superior frontal gyrus(r =-0.3431,p < 0.001).FC within sp DMN in the left calcarine was positively correlated with Mo CA scores(r = 0.3158,p < 0.001),FC within ip DMN in left precuneus was positively correlated with Mo CA scores(r = 0.3704,p < 0.001)and CDT scores(r = 0.4270,p < 0.001)in MSA patients.FC in the right middle temporal gyrus was positively correlated with Mo CA scores in MSA patients(r =0.3032,p < 0.001).FC within CN was positively correlated with Mo CA scores in MSA patients(r = 0.2866,p < 0.001).The FC between the LCEN and the ip DMN was negatively correlated with Mo CA(r =-0.2555,p = 0.027)and CDT scores(r =-0.2956,p < 0.001)in MSA patients.The FC between the CN and the RCEN was positively correlated with CWT scores in MSA patients(r = 0.4411,p < 0.001).The FC between the SN and the RCEN was negatively correlated with TMTA scores in MSA patients(r =-0.2523,p = 0.003).The FC between the left middle frontal gyrus and the right insula was negatively correlated with TMTA scores in MSA patients(r =-0.2069,p = 0.0156).The FC between the right angular gyrus and the left medial superior frontal gyrus was positively correlated with Mo CA scores(r = 0.2237,p = 0.0089)and negatively correlated with TMTA scores(r =-0.2462,p = 0.0039)in MSA patients.The FC between the right cerebellar Crus I and the left precuneus was positively correlated with CWT(r = 0.3048,p < 0.001)and Mo CA scores(r = 0.2562,p =0.0026)in MSA patients.Conclusions: Our results indicated that early cognitive decline in MSA patients is more attributable to functional changes than to anatomical damage.MCI in MSA was related to hypo-activation of the right DLPFC and corresponding DLPFC–IPL network dysfunction as well as hyper-activation of the right cerebellum and corresponding cerebello–DMN network disruption.Furthermore,we found intra-network alterations within CEN,DMN,CN and inter-network disruptions among triple networks as well as cerebello-cerebral circuit in MSA-MCI,correlated with global and specific cognitive function in MSA patients.The CEN plays a critical role in the network regulation in MSA-MCI.In conclusion,our findings provide a further understanding that primary cortical damage and subcortical degeneration with secondary cortical deafferentation might simultaneously contribute to early cognitive decline in MSA.