Subthalamic Beta Activity in Parkinson's Disease May Be Linked to Dorsal Striatum Gray Matter Volume and Prefrontal Cortical Thickness: A Pilot Study

Abstract

Background: Excessive oscillations at beta frequencies (13-35 Hz) in the subthalamic nucleus (STN) represent a pathophysiological hallmark of Parkinson's disease (PD), which correlates well with parkinsonian symptoms and is reduced in response to standard disease treatments. However, the association of disease-specific regional gray matter (GM) atrophy or cortical thickness (CT) with the presence of STN beta oscillatory activity has been poorly investigated but is of relevance given the potential of these variables for extracting information about PD pathophysiology. This exploratory study investigated the involvement of regional GM volume and CT in the basal ganglia-cortical network and its potential association with the presence of STN beta oscillatory activity in PD.

Methods: We acquired preoperative GM densities on T1-weighted magnetic resonance imaging scans and we carried out regional estimation of GM volume and CT. LFP activities from the STN were recorded post-operatively in 7 cognitively preserved PD patients off dopaminergic medication undergoing deep-brain stimulation surgery. Oscillatory beta power was determined by power spectral density of 4-min resting state STN LFP activity. Spearman partial correlations and regression analysis were used to screen the presence of STN beta power for their relationship with GM volume and CT measurements.

Results: After controlling for the effects of age, educational level, and disease duration, and after correcting for multiple testing, enhanced STN beta power showed significant and negative correlations between, first, volume of the right putamen and left caudate nucleus, and second, smaller CT in frontal regions involving the left rostral middle frontal gyrus (MFG) and left medial orbitofrontal gyrus. A lower volume in the right putamen and a lower CT in the left MFG demonstrated the strongest associations with increased STN beta power.

Conclusions: These tentative results seem to suggest that STN LFP beta frequencies may be mainly linked to different but ongoing parallel neurodegenerative processes, on the one hand, to GM volume reduction in dorsal striatum, and on the other hand, to CT reduction of prefrontal-"associative" regions. These findings could further delineate the brain structural interactions underpinning the exaggerated STN beta activity commonly observed in PD patients.

Keywords: Parkinson's disease; beta oscillations; cortical thickness; gray matter volume; local field potentials; subthalamic nucleus.

Figure 1

(A) DBS electrode location. Images were reconstructed and 3-D visualized using the Lead-DBS toolbox to determine the coordinates of each electrode contact. Right: DBS electrode localizations of 7 PD patients; and left: simulated location of the position of each bipolar register (black spheres, calculated by averaging the location of each pair of adjacent contacts) in the STN used for LFP analysis. Figures are visualized in the context of a 3D representation of the Morel stereotactic atlas in MNI space (Posterior view). The bilateral subthalamic nucleus (STN; orange), external globus pallidus (GPe; blue), internal globus pallidus (GPi; green), and red nucleus (RN; red) are shown. (B) Resting state STN spectral power in PD patients. Averaged (orange solid line) log-transformed power spectral density for STN-LFPs in 7 patients with Parkinson's disease off medication are shown. Orange shaded areas depict S.E.M of the mean for each frequency bin across patients. The x-axis represents frequency (Hz), and the y-axis represents the normalized power spectral density (log PSD, μV²/Hz). The gray shaded area indicates the beta frequency spectrum (13–35 Hz). (C) Regional analysis of gray matter (GM) volume showing areas negatively correlated, after adjustment for multiple testing, with local field potential (LFP) power in the beta band in the subthalamic nucleus (STN). The yellow line shows the right putamen, which was the only structure retained in the regression model predicting LFP beta activity in the STN (p < 0.01). (E) Regional analysis of cortical thickness (CT) volume showing areas negatively correlated, after adjustemt for multiple testing, with LFP power in the beta band in the subthalamic nucleus (STN). The yellow line shows the left rostral middle frontal gyrus (MFG), which was the only structure retained in the regression model predicting LFP beta activity in the STN (p < 0.01). (C,E) The red color denotes brain areas with significant partial correlations after FDR adjustment between regional GM volumes or CT measures and STN LFP beta activity. The yellow color indicates brain areas with a trend toward statistical significance after FDR adjustment between regional GM volumes and STN LFP beta activity. The red and yellows images are overlaid on a standard MNI T1 template. (D–F) Scatterplot graphs illustrating the partial correlations between STN LFP beta power and (D) right putamen nucleus atrophy and (F) left rostral MFG thinning, controlling for age, educational level, and disease duration. For illustration purposes, the y-axis represents the log-transformed power spectral density (log PSD, μV²/Hz), and the x-axis of each scatterplot graph represents the averaged residuals of age, educational level, and disease duration of (D) regional GM volume in the right putamen and (F) CT in the left rostral MFG.