To understand the neural mechanism that underlies treatment resistant auditory verbal hallucinations (AVH), is still an important issue in psychiatric research. Alterations in functional connectivity during rest have been frequently reported in patients with schizophrenia. Though the default mode network (DN) appears to be abnormal in schizophrenia patients, little is known about its role in resistant AVH. We collected resting-state functional magnetic resonance imaging (R-fMRI) data with a 3 T scanner from 19 schizophrenia patients with chronic AVH resistant to pharmacological treatment, 14 schizophrenia patients without AVH and 20 healthy controls. Using seed-based correlation analysis, we created spherical seed regions of interest (ROI) to examine functional connectivity of the two DN hub regions (posterior cingulate cortex and anteromedial prefrontal cortex) and the two DN subsystems: dorsomedial prefrontal cortex subsystem and medial temporal lobe subsystem (p < 0.0045 corrected). Patients with hallucinations exhibited higher FC between dMPFC ROI and bilateral central opercular cortex, bilateral insular cortex and bilateral precentral gyrus compared to non hallucinating patients and healthy controls. Additionally, patients with hallucinations also exhibited lower FC between vMPFC ROI and bilateral paracingulate and dorsal anterior cingulate cortex. As the anterior cingulate cortex and the insula are two hubs of the salience network, our results suggest cross-network abnormalities between DN and salience system in patients with persistent hallucinations.
Auditory verbal hallucinations (AVH) are sensory experiences perceived in auditory modality in absence of external stimuli. They represent one of the most common and distressing symptoms of schizophrenia suffered by 60%-80% of patients (Andreasen and Flaum, 1991). Up to 25% of these patients are resistant to antipsychotic pharmacological treatment (Shergill et al., 1998).
Functional magnetic resonance imaging during resting state periods (R-fMRI) offer a fertile ground to examine multiple brain systems and alterations associated with psychiatric diseases. Measures of functional connectivity (FC) by means of R-fMRI refer to temporal correlation of low frequency fluctuations in the blood oxygen level dependent signal between two (or more) brain regions (Fox and Raichle, 2007). Previous R-fMRI studies have investigated disruptions of FC related to patients with schizophrenia with chronic AVH as one of the symptoms, although investigating different brain regions and using a variety of methods (Lawrie et al., 2002, Mechelli et al., 2007, Gavrilescu et al., 2010, Rotarska-Jagiela et al., 2010, Vercammen et al., 2010, Hoffman et al., 2011 and Wolf et al., 2011). Reduced frontotemporal FC was reported in patients with schizophrenia and AVH (Lawrie et al., 2002 and Rotarska-Jagiela et al., 2010). Further, reduced interhemispheric FC of the auditory cortex was found in patients with schizophrenia or schizoaffective disorder and AVH (Gavrilescu et al., 2010). Also reduced FC between left superior temporal and anterior cingulate cortex (Mechelli et al., 2007) and between temporoparietal cortex and anterior cingulate cortex (Vercammen et al., 2010) was found in patients with schizophrenia and AVH. In contrast, stronger FC between left inferior frontal gyrus and Wernicke area was reported in patients with schizophrenia and AVH (Hoffman et al., 2011). Many factors might have been involved in such variety of results. In particular, some studies have investigated FC in treatment resistant hallucinating patients (Vercammen et al., 2010 and Wolf et al., 2011) while others, in non refractory hallucinating patients (Lawrie et al., 2002, Mechelli et al., 2007, Gavrilescu et al., 2010, Rotarska-Jagiela et al., 2010 and Hoffman et al., 2011). Another aspect is the different methodological approaches to determining FC (Independent Component Analyses (Rotarska-Jagiela et al., 2010 and Wolf et al., 2011), seed region (Lawrie et al., 2002, Gavrilescu et al., 2010, Vercammen et al., 2010 and Hoffman et al., 2011) or dynamic causal modeling approach (Mechelli et al., 2007)).
The dysfunction of the Default Network (DN), one of the most well-known resting state networks, seems to play a prominent role in schizophrenia (Menon, 2011 and Woodward et al., 2011). This large-scale brain network appears to have increased activation during rest and decreased activation during stimulus-induced activity (Raichle et al., 2001), and comprises the medial prefrontal cortex (mPFC), posterior cingulate cortex (PCC) and medial temporal lobe (MTL) including the hippocampus and the lateral temporoparietal area (Buckner et al., 2008). This network is more activated in patients with schizophrenia and has been significantly correlated with greater psychopathology (Garrity et al., 2007, Whitfield-Gabrieli et al., 2009, Camchong et al., 2011, Bastos-Leite et al., 2014 and Zhang et al., 2014). However, little is known about its role in treatment resistance AVH in patients with schizophrenia. Bearing this in mind, Northoff and Qin (2011) hypothesised that an abnormal interaction between the DN and the auditory cortex may induce hallucinations. However, a recent study (Wolf et al., 2011), which investigated several resting state networks including the DN, failed to find connectivity differences in the DN between schizophrenia patients with persistent AVH and healthy controls.
The aim of the present study is to examine the role of DN activity in schizophrenia patients with persistent AVH as compared to schizophrenia patients without previous history of hallucinations and to healthy controls. Resting state connectivity was analysed through seed region approach including manually defined regions of interest within the DN based on the validated 11 seeds (Andrews-Hanna et al. 2010). These authors described a dorsal mPFC DN subsystem that is activated when attention is directed to the self and a MTL DN subsystem that is activated when awareness is directed to the future. In a previous work we found abnormalities specific to the dMPFC DN subsystem in the early stages of psychosis (Alonso-Solis et al., 2012). We hypothesise that patients with persistent AVH may show a specific neuro-functional correlate in both subsystems compared to patients without hallucinations and healthy controls, suggesting that persistent hallucinations alters broadly the DN.
