Identifying 22q11.2 Deletion Syndrome and Psychosis Using Resting-State Connectivity Patterns

Abstract

The clinical picture associated with 22q11.2 deletion syndrome (22q11DS) includes mild mental retardation and an increased risk of schizophrenia. While the clinical phenotype has been related to structural brain network alterations, there is only scarce information about functional connectivity in 22q11DS. However, such studies could lead to a better comprehension of the disease and reveal potential biomarkers for psychosis. A connectivity decoding approach was used to discriminate between 42 patients with 22q11DS and 41 controls using resting-state connectivity. The same method was then applied within the 22q11DS group to identify brain connectivity patterns specifically related to the presence of psychotic symptoms. An accuracy of 84 % was achieved in differentiating patients with 22q11DS from controls. The discriminative connections were widespread, but predominantly located in the bilateral frontal and right temporal lobes, and were significantly correlated to IQ. An 88 % accuracy was obtained for identification of existing psychotic symptoms within the patients group. The regions containing most discriminative connections included the anterior cingulate cortex (ACC), the left superior temporal and the right inferior frontal gyri. Functional connectivity alterations in 22q11DS affect mostly frontal and right temporal lobes and are related to the syndrome's mild mental retardation. These results also provide evidence that resting-state connectivity can potentially become a biomarker for psychosis and that ACC plays an important role in the development of psychotic symptoms.

Large-scale functional network reorganization in 22q11.2 deletion syndrome revealed by modularity analysis

“…For this study, resting-state and structural data was acquired from 69 patients. We used the same exclusion criteria detailed in our previous paper (Scariati et al, 2014) resulting in the exclusion of a total of 22 scans: 18 scans were excluded for excessive motion (>3 mm translation or 3 rotation), 2 scans due to a part of the brain not being in the field of view (FOV) and 2 scans due to the participants falling asleep during the resting-state acquisition. In addition to the previous criteria, compared to our former study we also computed the Framewise Displacement (FD), which measures the amount of relative motion along the functional scanning session (Power, Barnes, Snyder, Schlaggar, & Petersen, 2012;Van Dijk, Sabuncu, & Buckner, 2012) and provides more strict control for motion confounds.…”

“…Table 1 details the demographic characteristics for these groups. Part of this sample was included in three previous papers from our group: 17 healthy control participants (9 females) and 17 patients with 22q11DS (8 females) were part of ; 27 control participants (18 females) and 26 patients (15 females) were included in (Padula et al, 2015); and all the participants except five control participants (2 females) and five patients (4 females) were included in our previous paper (Scariati et al, 2014 background. In order to avoid excessive head motion, the head was stabilized with material adapted to the participants' morphology.…”

“…edu/~richiard/software.html) to obtain functional connectivity matrices as described in previous work (Richiardi, Eryilmaz, Schwartz, Vuilleumier, & Van De Ville, 2011;Scariati et al, 2014). More specifically, the functional images were realigned to the mean functional scan.…”

“…The other was performed on an independent sample and showed a correlation between DMN dysconnectivity and social skills (Schreiner et al, 2013). A third paper, also published by our group, tested the wholebrain functional connectomes and described the presence of widespread functional connectivity alterations in 22q11DS, particularly in the frontal lobe (Scariati et al, 2014). However, even though altered functional connectivity has been described in 22q11DS, it is still unclear how this dysconnectivity affects functional network organization.…”

“…However, evidence suggests the presence of an altered neurodevelopment in 22q11DS, [reviewed in (Gothelf et al, 2008)] particularly in frontal lobe gray matter (Schaer et al, 2009;. Given that specific resting-state connectivity patterns have been associated with the presence of psychotic symptoms in this population (Scariati et al, 2014), identifying early course brain connectivity alterations is crucial as they may carry predictive value for the development of schizophrenia and may therefore potentially act as a biomarker for an increased risk of psychosis.…”