Changes of resting state brain networks in amyotrophic lateral sclerosis

Abstract

The defining feature of amyotrophic lateral sclerosis is degeneration of upper and lower motor neurons but extramotor involvement, evidenced for example by executive dysfunction, has also been demonstrated. Here we employed a novel functional imaging approach, the analysis of resting state activity, followed by the definition of functionally connected brain networks by independent component analysis (ICA) to assess differences between ALS patients (n = 20) and healthy controls (n = 20). ICA analysis revealed 5 typical brain networks among which the so-called default mode network and the sensori-motor network showed distinct differences between patients and controls. The default mode network showed less activation in patients in several regions including the ventral anterior cingulate cortex, posterior cingulate cortex and the left and right inferior parietal cortex, regions that have been linked previously to executive functions. The sensori-motor network showed group differences in the premotor cortex. We propose that resting state analysis affords a new and simple means to assess disease-related neurofunctional alterations in widespread brain networks. A decisive advantage is that no task is demanded from the subjects and, thus, the problem of differential task difficulty and effort between groups is circumvented.

Introduction

Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease involving preliminary upper and lower motor neuron with a rapid progress (Kollewe et al., 2008). Even early descriptions of ALS pointed out that some patients develop dementia (Ziegler, 1930) and more recently neuropsychological (Frank et al., 1997, Irwin et al., 2007, Lakerveld et al., 2008, Murphy et al., 2007, Woolley and Katz, 2008), electrophysiological (Munte et al., 1999, Paulus et al., 2002, Vieregge et al., 1999), and neuroimaging (Abrahams et al., 1996, Abrahams et al., 2004, Kew et al., 1993a) results suggest that the disease process involves other parts of the nervous system.

The demonstration of functional involvement of brain networks outside of the motor system proper is of great importance for our understanding of ALS. The present contribution seeks to provide evidence for such extra-motor involvement using a novel functional magnetic resonance imaging (fMRI) analysis method, independent component analysis (ICA), applied to blood oxygen level-dependent (BOLD) time-series obtained during rest.

Recently, the analysis of “functional connectivity” of spatially remote brain regions has become a focus of neuroimaging research. Using diverse techniques, functional connectivity analyses seek to delineate inter-regional neural interactions during the involvement in particular cognitive or motor tasks or, as done in the present study, during rest. The idea is that during rest there exist spontaneous coherent fluctuations of the BOLD signal in different brain areas that are functionally connected.

ICA methods are particularly suited to recover the sources (or components) underlying the observed signal, i.e. the spatio-temporal patterns of the fMRI BOLD-signal, by assuming that the sources are statistically independent (Calhoun et al., 2001b, Calhoun and Adali, 2006, Esposito et al., 2005, Garrity et al., 2007, McKeown et al., 1998). A decisive advantage of the ICA method is that it can be applied easily to “resting state” scans. These only take minutes to acquire and do not suffer from performance confounds that may be present in patients with cognitive or motor impairments (Beckmann et al., 2005, Greicius et al., 2004, Sorg et al., 2007).

Importantly, different typical resting state networks can be recovered from the BOLD signal with high reliability across individuals and studies (Beckmann et al., 2005, Damoiseaux et al., 2006, De Luca et al., 2006, van den Heuvel et al., 2008). One of the consistently recovered networks is the default mode network (DMN) which is conceptualized as a stand alone cognitive network (Raichle et al., 2001, Raichle and Snyder, 2007). Another often reported network is the sensorimotor network (Beckmann et al., 2005, Damoiseaux et al., 2006, De Luca et al., 2006).

A number of studies have provided initial evidence that resting state activity might be altered in neuropsychiatric conditions. For example, the DMN has been reported be changed in autism (Kennedy et al., 2006, Kennedy and Courchesne, 2008), Alzheimer's disease (Greicius et al., 2004), minimal cognitive impairment (Sorg et al., 2007), depression (Greicius et al., 2007), schizophrenia (Liang et al., 2006, Williamson, 2007) and attention deficit hyperactivity disorder (Tian et al., 2006). These studies underscore the potential of resting state fMRI analysis to reveal impaired network activity in neuropsychiatric conditions.

In the present study we therefore analyzed resting state networks in ALS patients for the first time by applying ICA. Given the “motor”-nature of ALS we expected differences between in ALS patients and healthy controls in the sensorimotor network. Importantly, any differences in the resting-state sensorimotor network could not be attributed to differences in effort or task difficulty between patients and control subjects, an issue that has been raised repeatedly in conjunction with brain imaging studies involving active movements of the limb in ALS (Konrad et al., 2002, Konrad et al., 2006, Schoenfeld et al., 2005, Stanton et al., 2007). Moreover, in view of the repeated demonstrations of extramotor involvement in ALS (see above) we also suspected differences between ALS patients and healthy controls in the default mode network.