The thalamus and brainstem act as key hubs in alterations of human brain network connectivity induced by mild propofol sedation

Abstract

Despite their routine use during surgical procedures, no consensus has yet been reached on the precise mechanisms by which hypnotic anesthetic agents produce their effects. Molecular, animal and human studies have suggested disruption of thalamocortical communication as a key component of anesthetic action at the brain systems level. Here, we used the anesthetic agent, propofol, to modulate consciousness and to evaluate differences in the interactions of remote neural networks during altered consciousness. We investigated the effects of propofol, at a dose that produced mild sedation without loss of consciousness, on spontaneous cerebral activity of 15 healthy volunteers using functional magnetic resonance imaging (fMRI), exploiting oscillations (<0.1 Hz) in blood oxygenation level-dependent signal across functionally connected brain regions. We considered the data as a graph, or complex network of nodes and links, and used eigenvector centrality (EC) to characterize brain network properties. The EC mapping of fMRI data in healthy humans under propofol mild sedation demonstrated a decrease of centrality of the thalamus versus an increase of centrality within the pons of the brainstem, highlighting the important role of these two structures in regulating consciousness. Specifically, the decrease of thalamus centrality results from its disconnection from a widespread set of cortical and subcortical regions, while the increase of brainstem centrality may be a consequence of its increased influence, in the mildly sedated state, over a few highly central cortical regions key to the default mode network such as the posterior and anterior cingulate cortices.

Figure 1.

The pipeline used to calculate an eigenvector centrality map. A correlation matrix calculated by considering ∼40,000 gray matter (including brainstem and cerebellum) voxels' time series was squared. The Erdös–Rényi entropy S = log(N)/log((∑d_i)/N), i = 1, … N (N being the number of nodes and d_i the degree of the ith node) of the network was set equal to 2 across subjects. As a result, a binary matrix whose entries were weighted by above threshold r² values (weighted adjacency matrix) was obtained. The eigenvector belonging to the normalized largest eigenvalue of the weighted adjacency matrix was calculated and its entries provided a centrality measure for each voxel.

Figure 2.

Eigenvector centrality changes induced by propofol mild sedation. A paired t test was calculated both for the decrease (A, AWAKE > SEDATED) and for the increase (B, AWAKE < SEDATED) of the eigenvector centrality. We report clusters that survived correction for multiple comparisons across space, via permutation testing, with FWE correction at P values <0.05.

Figure 3.

Thalamus seed-based R² connectivity analysis. A paired t test was calculated for the condition AWAKE > SEDATED. A significant reduction of R² was observed in the caudate (bilaterally), putamen (bilaterally), left premotor cortex, left primary somatosensory cortex, lingual gyrus (bilaterally), right anterior cingulate cortex, right hippocampus, left superior and inferior temporal gyrus, left lateral occipital cortex, cerebellar vermis, and crus I (bilaterally). We report clusters that survived correction for multiple comparisons across space, via permutation testing, with FWE correction at P values <0.05.

Figure 4.

Brainstem seed-based R² connectivity analysis. A paired t test was calculated for the condition AWAKE < SEDATED. As highlighted by arrows and circles, significant reductions of R² were observed in the left premotor cortex (yellow), right anterior cingulate cortex (red), right posterior cingulate cortex (green) and left angular gyrus (blue). We report clusters that survived correction for multiple comparisons across space, via permutation testing, with FWE correction at P values <0.05.