Elsevier

NeuroImage

Volume 94, 1 July 2014, Pages 385-395
NeuroImage

Review
Working memory performance of early MS patients correlates inversely with modularity increases in resting state functional connectivity networks

https://doi.org/10.1016/j.neuroimage.2013.12.008Get rights and content

Highlights

  • Network modularity (Q) increased in early stage Multiple Sclerosis patients

  • Impaired dual task performance in patients may predict early cognitive decline.

  • Modularity is a promising biomarker for detection of early stage brain reorganization.

  • Modularity is a promising biomarker for possibly detection of disease progression.

Abstract

Multiple sclerosis (MS) is an autoimmune inflammatory demyelinating and neurodegenerative disorder of the central nervous system characterized by multifocal white matter brain lesions leading to alterations in connectivity at the subcortical and cortical level. Graph theory, in combination with neuroimaging techniques, has been recently developed into a powerful tool to assess the large-scale structure of brain functional connectivity. Considering the structural damage present in the brain of MS patients, we hypothesized that the topological properties of resting-state functional networks of early MS patients would be re-arranged in order to limit the impact of disease expression. A standardized dual task (Paced Auditory Serial Addition Task simultaneously performed with a paper and pencil task) was administered to study the interactions between behavioral performance and functional network re-organization. We studied a group of 16 early MS patients (35.3 ± 8.3 years, 11 females) and 20 healthy controls (29.9 ± 7.0 years, 10 females) and found that brain resting-state networks of the MS patients displayed increased network modularity, i.e. diminished functional integration between separate functional modules. Modularity correlated negatively with dual task performance in the MS patients. Our results shed light on how localized anatomical connectivity damage can globally impact brain functional connectivity and how these alterations can impair behavioral performance. Finally, given the early stage of the MS patients included in this study, network modularity could be considered a promising biomarker for detection of earliest-stage brain network reorganization, and possibly of disease progression.

Introduction

Multiple sclerosis (MS) is an inflammatory demyelinating and neurodegenerative disorder of the central nervous system (CNS). Its pathology is characterized by regional demyelination and axonal damage appearing as focal lesions. This leads to disruptions in connectivity patterns of subcortical areas, dysfunction of spinal cord pathways, as well as changes in cortico-cortical connections, which can translate to motor, sensory and/or cognitive dysfunction. However, these behavioral abnormalities are not always perceptible with conventional clinical examination. There is evidence that at early stages of the disease, functional adaptive changes occur as a means to limit clinical manifestations of tissue damage (Audoin et al., 2005, Mainero et al., 2004, Pantano et al., 2002a, Pantano et al., 2002b). Due to the scattered nature of the structural abnormalities, the relationship between lesions and clinical disability is not straightforward.

Aiming to reach a better understanding of the disease to facilitate diagnosis and treatment, preferably at very early stages, considerable research efforts are currently being devoted to neuroimaging methods, at both structural and functional levels (Bonzano et al., 2009, Rossi et al., 2012). Functional neuroimaging studies in MS have shown the potential of providing information about functional adaptive changes and their relation with cognitive performance (Audoin et al., 2005, Bonavita et al., 2011, Bonzano et al., 2009).

This field of research has recently benefited from the observation that the human brain at the large-scale level behaves as a complex network of functionally interacting elements (Bullmore and Sporns, 2009, Rubinov and Sporns, 2010, Sporns et al., 2004). This network can be modeled using tools such as graph theory, which—in combination with neuroimaging techniques—allow going beyond the mapping of local activity changes and studying the re-organization of structural and functional connectivity (Zhou et al., 2006)—for a review see Guye et al. (2010). Graph theoretical approaches have contributed to elucidate patterns of functional connectivity in different brain states (task-related or resting-state) and have shown to be a promising instrument to characterize pathological conditions (Guye et al., 2010). Since connectivity changes associated with brain pathology seem to be highly correlated with cognitive impairment (Audoin et al., 2005, Bonavita et al., 2011, Guye et al., 2008, He et al., 2007, Tecchio et al., 2008), it has been speculated that a quantification of the abnormalities in topology (Rossi et al., 2012) and efficiency (Cover et al., 2006) of brain networks could be a good indicator of disease progression.

An important property of functional connectivity networks is their non-trivial modular structure. Network modularity is a topological property that provides information about the degree of decomposability of a network, by identifying modules of densely interconnected nodes, which are sparsely connected with nodes in other modules (Newman, 2006, Newman and Girvan, 2004). Several studies have reported an overlap between modular organization in the anatomical brain network and its functional specializations (Chen et al., 2008, Ferrarini et al., 2009, Meunier et al., 2009b, Wu et al., 2011b). Furthermore, changes in the modular structure of functional networks have been observed in the healthy brain related to normal aging (Chen et al., 2011, Meunier et al., 2009a, Wu et al., 2011a), learning processes (Bassett et al., 2011), sleep (Tagliazucchi et al., 2012), and in diverse pathological conditions such as chronic back pain (Balenzuela et al., 2010), Alzheimer's disease (de Haan et al., 2012, He et al., 2008, Sanz-Arigita et al., 2010), schizophrenia (Alexander-Bloch et al., 2010, Bassett et al., 2008) and epilepsy (Vaessen et al., 2012).

In this study we investigated the network modularity and the ensuing modular structure of functional connectivity networks derived from resting-state fMRI data of early MS patients and matched healthy controls (HC). Here and throughout the manuscript the term “early” refers to patients with minimal or bare of any clinical disability in established clinical measures (such as the EDSS) rather than disease duration. We hypothesized that, as a consequence of the structural abnormalities present in the brain of MS patients, communication between brain modules will be impaired and therefore connectivity will be re-arranged to favor network segregation, resulting in increased modularity when compared to healthy controls. Additional to these analyses, we performed the Paced Auditory Serial Addition Task (PASAT) in a dual task fashion to assess deficiencies in cognitive function, namely working memory, attention and speed of information processing. Even though the group of early MS patients taking part in the study was cognitively normal, our aim when introducing the high cognitive load in the dual task was to limit the brain network resources available during memory performance to compensate for subliminal dysfunction. The correlations between lesion load volume, modularity and behavioral data were then computed in order to test interactions between structure and function.

Section snippets

Subjects

A total of 16 patients in the initial phase of MS [CIS or RRMS according to the revised McDonald criteria (Polman et al., 2011), EDSS ≤ 2.5, T2-lesion volume < 15 mL] (11 females, mean age = 35.3, SD = 8.3, see Table 1) and 20 healthy volunteers (10 females, mean age = 29.9, SD = 7.0) participated in this study. All participants were assessed in sessions 1 and 2, however only 15 healthy controls (9 females, mean age = 31.1, SD = 7.1 years) and 15 patients (10 females, mean age = 35.0, SD = 8.5 years) joined session

Results

The results of the neuropsychological tests are summarized in Table 2.

Discussion

In the present study we show increased modularity values over a wide range of link densities (δ) in large-scale brain networks of the MS group. A decreased accuracy in the dual task performance and a negative correlation between modularity and dual task accuracy were found in the MS group.

In the present study we report increased network modularity values for the MS group. It is known that localized brain lesions can impair global metrics of network topology such as modularity (Gratton et al.,

Funding

This work was supported by the Hessisches Ministerium für Wissenschaft und Kultur—LOEWE Neuronale Koordination Forschungsschwerpunkt Frankfurt (NeFF).

Acknowledgments

This work was supported by the Hessisches Ministerium für Wissenschaft und Kultur—LOEWE Neuronale Koordination Forschungsschwerpunkt Frankfurt (NeFF).

We thank Christoph Mayer for patient recruitment, Marion Behrens for methodological advice and valuable discussions, Ulrike Nöth, Ralf Deichmann, Steffen Volz, Astrid Morzelewski and Julia Heiken for extensive MRI support, Julia Dieter for helping with patient recruitment and providing PASAT training to the participants and all subjects for their

References (74)

  • D. Meunier et al.

    Age-related changes in modular organization of human brain functional networks

    Neuroimage

    (2009)
  • R. Oldfield

    The assessment and analysis of handedness: the Edinburgh inventory

    Neuropsychologia

    (1971)
  • P. Pantano et al.

    Contribution of corticospinal tract damage to cortical motor reorganization after a single clinical attack of multiple sclerosis

    Neuroimage

    (2002)
  • J. Reichardt et al.

    When are networks truly modular?

    Physica D

    (2006)
  • M. Rubinov et al.

    Complex network measures of brain connectivity: uses and interpretations

    Neuroimage

    (2010)
  • A. Solari et al.

    The multiple sclerosis functional composite: different practice effects in the three test components

    J. Neurol. Sci.

    (2005)
  • O. Sporns et al.

    Organization, development and function of complex brain networks

    Trends Cogn. Sci.

    (2004)
  • E. Tagliazucchi et al.

    Automatic sleep staging using fMRI functional connectivity data

    Neuroimage

    (2012)
  • E. Tagliazucchi et al.

    Large-scale brain functional modularity is reflected in slow electroencephalographic rhythms across the human non-rapid eye movement sleep cycle

    Neuroimage

    (2013)
  • N. Tzourio-Mazoyer et al.

    Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain

    Neuroimage

    (2002)
  • A. Zalesky et al.

    On the use of correlation as a measure of network connectivity

    Neuroimage

    (2012)
  • A.F. Alexander-Bloch et al.

    Disrupted modularity and local connectivity of brain functional networks in childhood-onset schizophrenia

    Front. Syst. Neurosci.

    (2010)
  • B. Audoin et al.

    Compensatory cortical activation observed by fMRI during a cognitive task at the earliest stage of MS

    Hum. Brain Mapp.

    (2003)
  • B. Audoin et al.

    Magnetic resonance study of the influence of tissue damage and cortical reorganization on PASAT performance at the earliest stage of multiple sclerosis

    Hum. Brain Mapp.

    (2005)
  • A.D. Baddeley et al.

    Testing central executive functioning with a pencil and paper test

  • P. Balenzuela et al.

    Modular organization of brain resting state networks in chronic back pain patients

    Front. Neuroinform.

    (2010)
  • D.S. Bassett et al.

    Hierarchical organization of human cortical networks in health and schizophrenia

    J. Neurosci.

    (2008)
  • D.S. Bassett et al.

    Dynamic reconfiguration of human brain networks during learning

    Proc. Natl. Acad. Sci. U. S. A.

    (2011)
  • J. Bellmann-Strobl et al.

    Poor PASAT performance correlates with MRI contrast enhancement in multiple sclerosis

    Neurology

    (2009)
  • V.D. Blondel et al.

    Fast unfolding of communities in large networks

    J. Stat. Mech.

    (2008)
  • S. Bonavita et al.

    Distributed changes in default-mode resting-state connectivity in multiple sclerosis

    Mult. Scler.

    (2011)
  • U. Brandes et al.

    Maximizing Modularity is Hard. J ArXiv Physics e-prints

    (2006)
  • J.M. Buldu et al.

    Reorganization of functional networks in mild cognitive impairment

    PLoS One

    (2011)
  • E. Bullmore et al.

    Complex brain networks: graph theoretical analysis of structural and functional systems

    Nat. Rev. Neurosci.

    (2009)
  • K.S. Cardinal et al.

    A longitudinal fMRI study of the paced auditory serial addition task

    Mult. Scler.

    (2008)
  • C.-C. Chang et al.

    LIBSVM: a library for support vector machines

    ACM Trans. Intell. Syst. Technol.

    (2011)
  • Z.J. Chen et al.

    Revealing modular architecture of human brain structural networks by using cortical thickness from MRI

    Cereb. Cortex

    (2008)
  • Cited by (0)

    1

    Authors contributed equally to this work.

    View full text