Obsessive-compulsive disorder (OCD) is a generally chronic disease that can significantly affect the functionality of the person, with repetitive obsessions and/or compulsions (Abramowitz et al., 2009). Studies about brain regions associated with OCD have been conducted with various neuroimaging methods, especially the fMRI, which provides the opportunity to examine the functional connectivity and task-related activity changes. Numerous functional neuroimaging studies have revealed that dysfunction in the frontal subcortical circuits plays a role in the pathogenesis of OCD (Saxena et al., 1998; Hou et al., 2012, 2014). The orbito-fronto-subcortical circuit proposed by Saxena et al. (1998) has been accepted as a basic model in the neurobiology of OCD. In this model, the existence of an orbitofrontal loop consisting of projections starting from the orbitofrontal cortex (OFC) to the caudate nucleus and ventral striatum, then to the dorsomedial thalamus via internal pallidus and back to the OFC from the thalamus was asserted. Subsequent research has claimed that the underlying pathology of the disease is not limited to orbitofronto-striatal regions and associated limbic structures, but also involves abnormalities in additional brain areas, particularly more lateral frontal and parietal regions which may be considered as dorsolateral prefronto-striatal circuit (Menzies et al., 2008). Additionally, many fMRI studies have reported activity changes in the OFC, anterior cingulate cortex (ACC), subcortical areas (Hou et al., 2012; Fan et al., 2017; Ping et al., 2013) and also temporal cortex (Yang et al., 2019), parietal cortex (Xu et al., 2019) and occipital cortex (Chen et al., 2016; Tian et al., 2016).

Many studies using resting-state data exhibited abnormal functional connectivity within and between the default mode network (DMN), salience network (SN), and frontoparietal network (FPN) (Stern et al., 2012; Chen et al., 2018; Fan et al., 2017). Gürsel et al. (2018) conducted a meta-analysis of resting-state networks in OCD, in which impaired connectivity between the DMN, SN and FPN was reported. DMN is involved in reflective and introspective self-referring processes, while FPN plays a crucial role in action planning and response inhibition. SN is associated with regulating external and internal salient information (Cui et al., 2020). Abnormal functional connectivity both within and between the FPN, SN and DMN, which is named the “triple network” model (Menon, 2011), provides important insight about the pathophysiology of OCD (Gürsel et al., 2018). In this model, the SN is involved in both filtering and detecting the salient information interfering with the FPN and DMN functions. In other words, SN plays an important role in switching the activities of the DMN (Chen et al., 2018) and FPN (Gürsel et al., 2020).

Selective serotonin reuptake inhibitors (SSRIs) and cognitive-behavioral therapy (CBT) are the first-line treatment options for OCD. Few functional neuroimaging studies on the potential neurobiological mechanisms and predictors of response to CBT in OCD have provided results associated with either fronto-striatal circuitry or with ICNs (Yang et al., 2015; Moody et al., 2017; Li et al., 2018; Reggente et al., 2018). Yang et al. (2015) showed decreased regional homogeneity (ReHo) in the bilateral middle frontal cortex, right OFC, and increased ReHo in the left caudate along with the clinical improvement after CBT. Li et al. (2018) reported that CBT normalized the aberrant resting-state functional connectivity (rs-FC) between the right OFC and left dorsolateral prefrontal cortex. Reggente et al. (2018) showed that rs-FC within the DMN and visual network predicted OCD symptom severity after CBT.

Similarly, neuroimaging studies on the SSRI effects in OCD have reported either regulation of the increased connectivity in the orbito-fronto-striatal circuit by antidepressant medication through the reduction of the local ventral striatal connectivity (Beucke et al., 2013), or treatment-related changes in large-scale connectivity networks (Shin et al., 2014b). The authors reported reduced small-world efficiency and functional association between the DMN and FPN in OCD, while the elevation of the small-world efficiency, modular organization, and connectivity degree accompanied clinical improvement after 16 weeks of SSRI treatment.

Although there is a growing literature focusing on the “triple network” model in the pathogenesis of OCD, current literature indicates that other networks are also in play. For example, the dorsal attention network (DAN), used as seed only in a few resting-state fMRI (rs-fMRI) studies, draws attention by showing aberrant connectivity in OCD (Brennan et al., 2019).

Our study, adding the large-scale ICNs to the conventional fronto-striatal circuitry, investigated rs-FC changes in adult OCD patients and the effects of medication (escitalopram) on them with a seed-based approach, where the seed regions include basal ganglia as core structures of the fronto-striatal circuit as well as the main nodes of the FPN, DMN and SN of the triple network model. Additionally, main nodes of the DAN, recently shown to be associated with the obsession severity (Lee et al., 2020), were also included in the analyses. By computing the seed-to-voxel connectivities between these regions and all brain voxels, we assume that a comprehensive view of the pathophysiological changes within and between the related circuits or networks can be obtained. Strict inclusion criteria, recruitment of unmedicated patients with no comorbidity and detailed comparison of the pre- and post-treatment findings further helped in obtaining a more complete view of the neurobiology and pathogenesis of OCD.