Connectome dysfunction in patients at clinical high risk for psychosis and modulation by oxytocin

Abstract

Abnormalities in functional brain networks (functional connectome) are increasingly implicated in people at Clinical High Risk for Psychosis (CHR-P). Intranasal oxytocin, a potential novel treatment for the CHR-P state, modulates network topology in healthy individuals. However, its connectomic effects in people at CHR-P remain unknown. Forty-seven men (30 CHR-P and 17 healthy controls) received acute challenges of both intranasal oxytocin 40 IU and placebo in two parallel randomised, double-blind, placebo-controlled cross-over studies which had similar but not identical designs. Multi-echo resting-state fMRI data was acquired at approximately 1 h post-dosing. Using a graph theoretical approach, the effects of group (CHR-P vs healthy control), treatment (oxytocin vs placebo) and respective interactions were tested on graph metrics describing the topology of the functional connectome. Group effects were observed in 12 regions (all p_FDR < 0.05) most localised to the frontoparietal network. Treatment effects were found in 7 regions (all p_FDR < 0.05) predominantly within the ventral attention network. Our major finding was that many effects of oxytocin on network topology differ across CHR-P and healthy individuals, with significant interaction effects observed in numerous subcortical regions strongly implicated in psychosis onset, such as the thalamus, pallidum and nucleus accumbens, and cortical regions which localised primarily to the default mode network (12 regions, all p_FDR < 0.05). Collectively, our findings provide new insights on aberrant functional brain network organisation associated with psychosis risk and demonstrate, for the first time, that oxytocin modulates network topology in brain regions implicated in the pathophysiology of psychosis in a clinical status (CHR-P vs healthy control) specific manner.

Fig. 1

Schematic overview of the network analysis steps.

Fig. 2. Overview of group, treatment and interaction effects across all nodal metrics (betweenness-centrality, node degree and local efficiency).

The shade of colour represents the t-statistic for each region. Only regions surviving FDR-corrected significance threshold p < 0.05 shown. In the top panel depicting main effects of group, regions in purple and green depict lower and greater graph metrics (respectively) in CHR-P relative to healthy controls (HC). In the middle panel depicting treatment effects, regions in green and purple depict increases and decreases (respectively) in graph metrics under oxytocin (OT) relative to placebo (PL). Although significant, the brainstem (t = 2.34) is not depicted as it is not included in the ENIGMA visualisation template. In the lower panel depicting group x treatment interaction effects, regions in green depict where oxytocin increased (↑) graph metrics in the CHR-P group but decreased (↓) them in healthy controls; regions in purple depict where oxytocin decreased graph metrics in the CHR-P group but increased them in controls. All results are presented here irrespective of the metric; where there were effects in the same region in more than one nodal metric, the larger t-value is shown. All corresponding statistics are presented in Table 2 and individual figures for each metric are appended in Supplementary Figs. 2–4.

Fig. 3. Overlap between the group, treatment and interaction findings and the large-scale canonical resting-state networks (RSNs).

We calculated the percentage of overlap between our result maps—which included binary masks of all cortical regions showing differences in nodal metrics (for group, treatment and interaction effects, separately)—and the large-scale RSNs described in the atlas from Yeo et al. [64]. As subcortical structures are not covered by the Yeo atlas, these were omitted from our result maps to prevent artificial reduction of the overlap estimate. The Yeo atlas includes a coarse parcellation of 7 canonical RSNs: the default-mode (DMN), dorsal attention (DAN), frontoparietal (FPN), limbic (LIM), somatomotor (SOM), visual (VIS) and ventral attention (VAN) networks. We created a proxy DKA>Yeo atlas for each of the 7 Yeo RSNs by combining individual DKA regions, allocating each to a single RSN based on the RSN for which each region had the highest number of overlapping vertices based on the confusion matrix from a previous study [100]. Overlap was quantified using the Dice-kappa coefficient, which measures the percentage of voxels of each RSN overlapping with our group/treatment/interaction effect maps. In the upper section, we provide an overview of all regions where we found group, treatment or interaction effects, irrespective of the specific graph metric, rendered in a 3D glass brain (semi-transparent) surface model. In the lower section, we provide a heatmap summarising the percentage of overlap (Dice-kappa coefficient) between our results and each of the 7 networks, with each network rendered in a 3D glass brain (semi-transparent) surface model. Note that despite the visualisation, regions belonging to the different RSNs do not overlap, for example, the FPN contains the rostral and caudal middle frontal gyri, whereas the DMN contains the superior and inferior frontal gyri, which are difficult to differentiate in rendered models.