Frontal Anatomical Correlates of Cognitive and Speech Motor Deficits in Amyotrophic Lateral Sclerosis

Abstract

The goal of this study was to identify neurostructural frontal lobe correlates of cognitive and speaking rate changes in amyotrophic lateral sclerosis (ALS). 17 patients diagnosed with ALS and 12 matched controls underwent clinical, bulbar, and neuropsychological assessment and structural neuroimaging. Neuropsychological testing was performed via a novel computerized frontal battery (ALS-CFB), based on a validated theoretical model of frontal lobe functions, and focused on testing energization, executive function, emotion processing, theory of mind, and behavioral inhibition via antisaccades. The measure of speaking rate represented bulbar motor changes. Neuroanatomical assessment was performed using volumetric analyses focused on frontal lobe regions, postcentral gyrus, and occipital lobes as controls. Partial least square regressions (PLS) were used to predict behavioral (cognitive and speech rate) outcomes using volumetric measures. The data supported the overall hypothesis that distinct behavioral changes in cognition and speaking rate in ALS were related to specific regional neurostructural brain changes. These changes did not support a notion of a general dysexecutive syndrome in ALS. The observed specificity of behavior-brain changes can begin to provide a framework for subtyping of ALS. The data also support a more integrative framework for clinical assessment of frontal lobe functioning in ALS, which requires both behavioral testing and neuroimaging.

Figure 1

A 3D volume surface-rendered cerebral cortex showing the nine cortical regions used in the analyses with SABRE: axial view (left) and left hemisphere sagittal view (right). Regions of interest correspond to lateral superior, middle, and inferior frontal; medial superior, medial middle, and medial inferior frontal; precentral, postcentral, and occipital.

Figure 2

Results of the partial least square (PLS) regression analysis for selected behavioral measures. The SABRE brain volumetric predictors that supported our hypotheses based on Stuss' model [14] and speech production models are marked with an asterisk.