Altered topology of the functional speech production network in non-fluent/agrammatic variant of PPA

Publication date: Available online 11 August 2018

Source: Cortex

Author(s): M.L. Mandelli, E. Vilaplana, A.E. Welch, C. Watson, G. Battistella, J.A. Brown, K.L. Possin, H.I. Hubbard, Z.A. Miller, M.L. Henry, G.A. Marx, M.A. Santos-Santos, L.P. Bajorek, J. Fortea, A. Boxer, G. Rabinovici, S. Lee, J. Deleon, H.J. Rosen, B.L. Miller

Abstract

Non-fluent/agrammatic primary progressive aphasia (nfvPPA) is caused by neurodegeneration within the left fronto-insular speech and language production network (SPN). Graph theory is a branch of vmathematics that studies the network architecture (topology) by quantifying features based on its elements (nodes and connections). This approach has been recently applied to neuroimaging data to explore the complex architecture of the brain connectome, though few studies exploited this technique in PPA. Here, we used graph theory on functional MRI resting state data of a group of 20 nfvPPA patients and 20 matched controls to investigate topological changes in response to focal neurodegeneration. We hypothesized that changes in the network architecture would be specific to the affected SPN in nfvPPA, while preserved in the spared default mode network (DMN). Topological configuration was quantified by hubs' location and global network's metrics. Our findings showed a less efficiently wired and less optimally clustered SPN, while no changes were detected in the DMN. The SPN in the nfvPPA group showed a loss of hubs in the left fronto-parietal-temporal area and new critical nodes in the anterior left inferior-frontal and right frontal regions. Behaviorally, speech production score and rule violation errors correlated with the strength of functional connectivity of the left (lost) and right (new) regions respectively. This study shows that focal neurodegeneration within the SPN in nfvPPA is associated with network-specific topological alterations, with loss and gain of crucial hubs and decreased global efficiency that were better accounted for by functional rather than structural changes. These findings support the hypothesis of selective network vulnerability in nfvPPA and may offer biomarkers for future behavioral intervention.

https://ift.tt/2vXgxuF