Brain circuitries of obsessive compulsive disorder: a systematic review and meta-analysis of diffusion tensor imaging studies.

Neuroscience and Biobehavioral Reviews 37 (2013) 2856–2877

Contents lists available at ScienceDirect

Neuroscience and Biobehavioral Reviews journal homepage: www.elsevier.com/locate/neubiorev

Review

Brain circuitries of obsessive compulsive disorder: A systematic review and meta-analysis of diffusion tensor imaging studies Federica Piras a , Fabrizio Piras a , Carlo Caltagirone a,b , Gianfranco Spalletta a,∗ a b

IRCCS Santa Lucia Foundation, Department of Clinical and Behavioral Neurology, Neuropsychiatry Laboratory, Via Ardeatina 306, 00179 Rome, Italy Department of Neuroscience, Tor Vergata University of Rome, Italy

a r t i c l e

i n f o

Article history: Received 10 August 2013 Received in revised form 27 September 2013 Accepted 19 October 2013 Keywords: Obsessive compulsive disorder Connectivity White matter micro-structure Diffusion tensor imaging Fractional anisotropy Mean diffusivity Cortico-striato-thalamo-cortical circuitry Intra-hemispheric bundles Posterior parietal/occipital cortices Corpus callosum

a b s t r a c t The potential role of white matter (WM) abnormalities in the pathophysiology of obsessive compulsive disorder (OCD) is substantially unexplored. Apart from alterations in the WM tracts within corticostriato-thalamo-cortical circuitry, recent theorizations predict the existence of more widespread WM abnormalities. In this paper we systematically reviewed the current diffusion tensor imaging literature in OCD and purposely evaluated the prevalence and functional significance of specific WM tissue changes in the disorder. The relationship between clinical variables (medication status, symptom severity) and WM microstructural changes was also assessed. The reviewed studies are consistent with the existence of microstructural alterations in the fronto-basal pathways targeting the orbitofrontal cortex and the anterior cingulate cortex. Moreover, altered anatomical connectivity between lateral frontal and parietal regions and microstructural abnormalities in intra-hemispheric bundles linking distinctive areas of the prefrontal cortex to posterior parietal and occipital association cortices, are consistently reported. Finally, microstructural abnormalities in the corpus callosum, characterized by decreased connectivity in the rostrum and hyperconnectivity in the genu, are substantiated by a large body of evidence. © 2013 Elsevier Ltd. All rights reserved.

Contents 1. 2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods and materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Review questions and objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Literature search, study selection and data extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. ALE and CMA meta-analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Neuroimaging and statistical methods employed in the reviewed studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Study moderators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. White matter tracts implicated in the classical cortical–striatal–thalamic-cortical model of OCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1. Evidence supporting the fronto-striatal model of OCD. Data synthesis and integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2. How might microstructural abnormalities within the classic frontostriatal circuit relate to the expression of OCD symptoms? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Additional white matter regions and fiber bundles putatively involved in OCD pathophysiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. Evidence supporting the extended model of OCD circuitry. Data synthesis and integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Clinical correlates of microstructural alterations in regions outside the OCD classic circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quantitative evaluation of FA alterations in OCD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Coordinate based ALE meta-analysis of whole brain voxel based studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. ROI based meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author. Tel.: +39 06 51501575; fax: +39 06 51501575. E-mail address: [email protected] (G. Spalletta). 0149-7634/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neubiorev.2013.10.008

2857 2858 2858 2858 2859 2859 2859 2862 2862 2862 2862 2865 2865 2872 2872 2872 2873

5.

F. Piras et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 2856–2877

2857

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2873 2874 2875 2875 2875

1. Introduction With its prevalence rate ranging between one and four percent, obsessive–compulsive disorder (OCD) is the fourth-most-common mental disorder worldwide (Fullana et al., 2009; Leonard et al., 2005). The main clinical manifestations seen in OCD patients are recurrent, intrusive and distressing thoughts (obsessions) and/or repetitive behaviors or mental acts (compulsions), which are executed to avoid anxiety or neutralize obsessions (American Psychiatric Association, 2000). OCD symptoms interfere significantly with subjects’ normal routine and patients are often chronically hampered by functional impairments including poor or failed occupational and educational performance (Koran et al., 2010). Over the last 2 decades, neuroimaging studies have indicated several neurobiological changes underlying the psychological and behavioral deficits of OCD. Evidence from functional and structural magnetic resonance imaging (MRI) and positron emission tomography (PET) has supported the notion that abnormalities in key gray matter (GM) regions, such as the orbitofrontal cortex (OFC), anterior cingulate cortex (ACC), thalamus and striatum play an important role in its pathophysiology (Graybiel and Rauch, 2000; Menzies et al., 2008a; Saxena and Rauch, 2000). These findings suggest that a dysfunctional cortico-striato-thalamo-cortical circuitry contributes to the pathophysiology of OCD. However, the overall picture is still rather heterogeneous and recent studies employing whole-brain analyses also indicate more distributed neuroimaging alterations in patients with OCD, implicating other brain regions such as parietal cortex, dorsolateral prefrontal cortex (DLPFC) and posterior temporo-parieto-occipital associative areas (Menzies et al., 2008a). On the other hand, the white matter (WM) tracts connecting the cortical and subcortical regions are relatively unexplored (Yoo et al., 2007). Consistently with current frontostriatal models of OCD pathophysiology, recent studies (Harrison et al., 2009; Stern et al., 2012; Zhang et al., 2011) showed altered functional connectivity among GM matter nodes of the cortico-striato-thalamo-cortical circuitry. Accordingly, there is growing evidence that OCD symptoms may be at least partly underpinned by reduced WM integrity (Douzenis et al., 2009; Fontenelle et al., 2009). Moreover, recent genetic studies in OCD demonstrated a biased transmission of polymorphisms in genes involved in myelination (Stewart et al., 2007; Zai et al., 2004), suggesting the existence of structural abnormalities of myelin in the disorder. The only published study reviewing the role of WM abnormalities in the pathophysiology of OCD (Fontenelle et al., 2009) suggests the existence of abnormalities in specific WM tracts (e.g. internal capsule (IC), cingulate bundle (CB), and corpus callosum (CC)) and in different brain regions (medial frontal and parietal WM), in the OCD population. These WM abnormalities may be familial (Menzies et al., 2008b) and responsive to serotonin reuptake inhibitor treatment (SSRI) (Fan et al., 2012; Yoo et al., 2007) and vary according to the severity of different symptom dimensions (Ha et al., 2009; Koch et al., 2012). Recently, diffusion tensor imaging (DTI) (Basser et al., 1994; Pierpaoli et al., 1996) has been used to detect possible microstructural WM abnormalities in OCD patients. Indeed, DTI is sensitive to the diffusion patterns of water molecules and by

measuring the direction and magnitude of restricted tissue water motility (Frodl et al., 2012), the orientation of WM tracts in the brain can be determined (Nobuhara et al., 2006). The commonly used parameters for measuring WM integrity are Fractional Anisotropy (FA), a measure of directionality of water diffusion, and Mean Diffusivity (MD), a measure of the magnitude of diffusion. Decreased FA indicates a loss of water directionality, likely due to a damage in structural organization of the tissue (Schulte et al., 2005), while increased MD is thought to be linked to an enlargement in the extracellular space due to altered cytoarchitecture, suggesting immaturity or degeneration of the tissue (Sykovà, 2004). For improving the specificity, the directional diffusivities derived from DTI measurements are separated into components parallel, Axial Diffusivity (AD), and perpendicular, Radial Diffusivity (RD), to the WM tract. Decreased AD is associated with axonal injury and dysfunction, whereas increased RD is associated with myelin injury in mouse models of WM injury (Song et al., 2003). The use of multiple DTI measures, such as a combination of FA, MD, RD and/or AD has already been proven to be helpful in understanding the different mechanisms underlying microstructural changes (Di Paola et al., 2010); however, AD and RD have been hardly reported in OCD and 5 out of the 8 studies reviewed by Fontenelle (Fontenelle et al., 2009) examined only FA without looking at other measures. Given that DTI provides a particular unique piece of microstructural information about WM organization and connectivity, which volumetric measurements cannot convey (since for example, a disorganized WM pathway is not expected to change substantially in volume), we considered essential to review the current DTI literature in OCD. Moreover, given the rapid evolution of methods for the acquisition and analysis of diffusivity measures, an update of the evidence on WM pathways integrity in OCD is of relevance. Indeed, the number of neuroimaging studies in OCD has grown exponentially in recent years, while findings from different investigations may sometimes be difficult to integrate into a coherent picture. Moreover, each method has advantages and drawbacks since region of interest (ROI) studies are affected by a limited and potentially biased inclusion of brain regions (Radua and Mataix-Cols, 2012), while the use of voxels, as in whole brain and in tract based voxel wise analyses, improves the correct localization of potential abnormalities, but is biased toward group differences that are localized in space (Davatzikos, 2004). Our main aim was to assess the evidence on WM microstructure abnormalities in the disorder, trying to characterize the specific WM tissue changes (i.e. axonal reduced organization or damage vs. abnormalities of myelin integrity) associated with OCD. As different microstructural indices are sensitive to diverse tissue properties, we tried, whenever possible, to interpret data considering the interrelations between complementary measures of WM microstructure. We also purposely analyzed the relationship between clinical variables such as medication status and symptom severity, and WM microstructural changes, in order to determine whether the observed abnormalities should be considered trait or state markers. Finally, we intentionally evaluated the evidence of WM microstructural changes outside the cortico-striato-thalamo-cortical loop, as to investigate whether OCD pathology involves WM tracts beyond this classically implicated circuit (Menzies et al., 2008a; Piras et al., 2013). In order to synthesize the findings,

2858


different meta-analyses were conducted on the database: the activation likelihood estimation (ALE) (Laird et al., 2005a) a quantitative voxel based meta-analysis method which can be used to estimate consistent activation across different imaging studies, and a ROI-based meta-analysis. Evidence of microstructural alterations in WM tracts within the cortico-striato-thalamo-cortical circuit is discussed first, while eventual empirical support for the existence of more widespread WM abnormalities in OCD, if confirmed, will be reported separately.

medication status) and diffusivity indices of WM integrity in order to ascertain whether the reported changes are pathogenic for OCD rather than an epiphenomenon of contingent comorbid illnesses or a by-product of medication usage. 2.2. Literature search, study selection and data extraction We accessed Pubmed, PsycNET (including the PsycINFO, PsycBOOKS, PsycCRITIQUES, PsycARTICLES and PsycEXTRA databases) and Scopus, searching for neuroimaging articles employing the DTI method in OCD patients. Identical searches were conducted in each database from September 2012 to March 2013, without limits on year of publication, entering in turn the Key words obsessive compulsive disorder and any of the following terms: diffusion magnetic resonance imaging, diffusion tensor imaging, DTI or tract-based spatial statistic. The reference list of identified articles and review papers were also hand searched to obtain additional articles. Studies were considered for inclusion if: (1) they were published in English in a peer-reviewed journal, (2) they included subjects with a primary diagnosis of OCD according to ICD or DSM criteria, (3) they used the DTI method (Basser et al., 1994; Pierpaoli et al., 1996) and (4) they compared an OCD group with an healthy subjects (HS) group. From the database search, a total of 266 studies were identified, including 20 review papers (Aoki et al., 2012; Assaf, 2008; Ayling et al., 2012; Diwadkar and Keshavan, 2002; Etkin and Wager, 2007; Fontenelle et al., 2009; Fornito and Bullmore, 2012; Huey et al., 2008; Huyser et al., 2009; Kwon et al., 2009; Leckman and Bloch,

2. Methods and materials 2.1. Review questions and objectives Although we did not follow a published pre-specified protocol during our systematic review, the papers inclusion/exclusion criteria, search strategy and primary assessed variables were defined a priori (the Prisma 2009 Checklist is Fig. S1 and the Prisma Flow Chart is Fig. 1). Our main aim was to assess the evidence on WM microstructure abnormalities in OCD, and we refined this principal issue by addressing three related more specific objectives: (1) to evaluate whether potential microstructural WM changes spread beyond the fronto-striatal pathways classically implicated in OCD, (2) to examine the prevalence and functional significance of specific WM tissue changes (as measured by different diffusivity indices) in the disorder, (3) to assess the possible correlation between clinical variables (OCD symptom profile and severity, comorbidity and

266 References from PubMed, PsycNET and Scopus databases 20 Previously Published Reviews

Unable to Retrieve 1 Not available in English 2 No DTI Studies Included 3 No OCD studies 5

85 Records after Duplicate Removed 22 Studies Identified

85 Records Screened

68 Papers Excluded

17 Studies Included in Systematic Review

No Patients with OCD 33

No Control Group 1

Other than the DTI method 31

Less than 10 Patients included 3

Fig. 1. Processes of literature search, study screening and study selection.


2008; Lim and Helpern, 2002; Maia et al., 2008; Menzies et al., 2008a; Milad and Rauch, 2012; Nucifora, 2010; Paul, 2011; PenaGarijo et al., 2010; Vloet et al., 2006; White et al., 2008). Within the latter, 1 was not traceable (White et al., 2008), 2 were not available in English (Pena-Garijo et al., 2010; Vloet et al., 2006), 3 did not include DTI studies (Huyser et al., 2009; Leckman and Bloch, 2008; Maia et al., 2008), and 5 did not involve OCD studies (Aoki et al., 2012; Diwadkar and Keshavan, 2002; Etkin and Wager, 2007; Fornito and Bullmore, 2012; Paul, 2011). The remaining 9 comprised 22 DTI articles, 21 already included in the original search. After duplicated were removed, 85 citations and abstracts were screened and 68 records excluded (no subjects with a primary diagnosis of OCD: 33 articles; no control group: 1; other than the DTI method: 31 papers; less than 10 individuals included: 3 studies). 17 full-text articles were assessed for eligibility and all fulfilled the inclusion/exclusion criteria (Bora et al., 2011; Cannistraro et al., 2007; Fan et al., 2012; Fontenelle et al., 2011; Garibotto et al., 2010; Ha et al., 2009; Jayarajan et al., 2012; Li et al., 2011; Lochner et al., 2012; Menzies et al., 2008b; Nakamae et al., 2008, 2011; Oh et al., 2011; Saito et al., 2008; Szeszko et al., 2005; Yoo et al., 2007; Zarei et al., 2011). These investigations comprised 350 OCD patients (309 adults, and 41 adolescents) and 358 HS (317 adults, and 41 adolescents). Eligibility assessment and study selection was performed by first author (FP) and independently verified by another (GS). We developed a data extraction sheet, pilot-tested it on five randomly selected included studies, and refined it accordingly. First author (FP) extracted the following data from included studies and the second author (GS) checked the extracted data: Sociodemographic characteristics: sample numerosity, age, proportion of males in the study populations, years of formal education; Clinical characteristics: age at onset, illness duration, Y-BOCS total score, presence of Axis I comorbid conditions, medication status; Methods for quantitative analysis and diffusion indices extracted; Image acquisition parameters: RM magnetic field strength, diffusion directions, b-factor; Image analysis: smoothing kernel size, intensity absolute threshold, statistical test, p-value threshold, correction for multiple comparisons, variables included in correlation analyses. For each study, the used standardized atlas (Montreal Neurological Institute or Talairach spaces) was identified, and MNI coordinates transformed into Talairach coordinates using the icbm2tal algorithm (Lancaster et al., 2007). The corresponding coordinates were extracted for each significant focus of difference between patients and controls; when coordinates were not available in the original publications, the corresponding authors were contacted by email requesting any details not included. Difference in diffusion metrics was the primary measure of white matter integrity, while we tried to discern whether WM abnormalities reflect the underlying “trait” of OCD, as assessed by comparing patients and HS, or a “state” marker, as assessed through a cross-sectional approach by correlating DTI indices with symptom ratings and clinical status. 2.3. ALE and CMA meta-analyses To summarize the results of the database searches, we employed two different meta-analytical methods: a coordinate-based metaanalysis using the activation likelihood estimation (ALE) approach (Laird et al., 2009) for data from whole brain voxel based DTI investigations, and Comprehensive Meta-Analysis (CMA) version 2 (Borenstein et al., 2005) for a ROI-based meta-analysis. ALE analyses were performed when at least two studies providing coordinates suitable for meta-analysis were available, and completed using the GingerALE software (BrainMap, University of Texas) (Laird et al., 2005a,b). Separate ALE maps were created for coordinates associated with increased or decreased FA in OCD patients compared with HS. Analyses were performed after recommendations

2859

from BrainMap (http://www.brainmap.org) using the ALE algorithm described by Eickhoff (Eickhoff et al., 2009) which estimates the spatial uncertainty of each focus and takes into account the possible differences among studies. A full-width half-maximum of 10 mm was used and the threshold for statistical significance was set at p = 0.01 (false discovery rate corrected) with a minimum cluster size of 200 mm3 . We used MRIcron software (http://www.sph.sc.edu/comd/rorden/mricron/) to visualize ALE maps overlaid onto a Talairach template downloaded from the BrainMap web site. Regarding the ROI-based meta-analysis, we considered for examination only those brain regions where FA was measured by at least three studies, which was true for only 3 brain areas. Since studies were quite heterogeneous in terms of sample size or neuroimaging and statistical methods employed, the analyses were performed with a random-effects model, which considers both between-study and within-study variability. An effect size was considered significant when the 95% confidence interval (95% CI) excluded 0 and when the p value was < 0.05. For assessing study heterogeneity, the Q statistic was calculated and considered significant for p < 0.05. When a significant level of heterogeneity was reached, the I2 index, an estimate of the total variation across included studies that is due to heterogeneity rather than chance, was determined. I2 values of 25, 50, and 75 were indicative of a mild, moderate, and high heterogeneity between trials, respectively. 2.4. Neuroimaging and statistical methods employed in the reviewed studies Multiple methods were used for the quantitative analysis of data from diffusion tensor imaging in the 17 identified papers. VBM-style analysis of diffusion measures was the most used (7 investigations, Fan et al., 2012; Garibotto et al., 2010; Ha et al., 2009; Li et al., 2011; Nakamae et al., 2008; Szeszko et al., 2005; Yoo et al., 2007) followed by tract-based spatial statistic (TBSS) (Smith et al., 2006) (3 studies, Bora et al., 2011; Jayarajan et al., 2012; Nakamae et al., 2011) and ROI analysis (1 study, Saito et al., 2008). One study employed seeding tractography (Oh et al., 2011). Finally, 5 studies (Cannistraro et al., 2007; Fontenelle et al., 2011; Menzies et al., 2008b; Lochner et al., 2012; Zarei et al., 2011) combined different techniques. Table 1 synthesizes the methods and parameters for quantitative analysis and imaging acquisition employed in the reviewed studies, while Table 2 reports the sociodemographic and clinical characteristics of the studied populations. 2.5. Study moderators The majority of the identified studies (15 papers) analyzed the potential correlation between DTI measures and OCD symptom severity (Fan et al., 2012; Fontenelle et al., 2011; Garibotto et al., 2010; Ha et al., 2009; Jayarajan et al., 2012; Li et al., 2011; Lochner et al., 2012; Menzies et al., 2008b; Nakamae et al., 2008, 2011; Oh et al., 2011; Saito et al., 2008; Szeszko et al., 2005; Yoo et al., 2007; Zarei et al., 2011), while 9 investigations evaluated the confounding effect of comorbid depression (Ha et al., 2009; Lochner et al., 2012; Saito et al., 2008; Szeszko et al., 2005; Yoo et al., 2007) or anxiety symptoms (Bora et al., 2011; Fan et al., 2012; Li et al., 2011; Lochner et al., 2012; Oh et al., 2011; Yoo et al., 2007). A subgroup analysis to determine whether WM alterations were associated with different symptom dimensions was carried out in one study (Ha et al., 2009). Illness duration was correlated to diffusivity measures in two studies (Ha et al., 2009; Li et al., 2011), while the influence of medication status was investigated in three other papers (Bora et al., 2011; Fan et al., 2012; Jayarajan et al., 2012). Finally, one

2860

Table 1 Methods and parameters for quantitative analysis and imaging acquisition in DTI studies on subjects diagnosed with obsessive compulsive disorder. Imaging and analyses parameters Scanner

DTI processing

Voxel size (mm)

Directions

b-Factor (s/mm2 )

Smoothing (mm)

Theshold masking (%)

DTI index

Correction

Szezsko 2005

1.5 T

0.86 × 0.86 × 5

25

1000

3

40

FA

p < 0.005 uncorr.

Cannistraro 2007

1.5 T

1 × 1 × 1.33

6

600

1.5

N/A

FA

p < 0.005 uncorr.

Yoo 2007 Menzies 2008

1.5 T 1.5 T

1.71 × 1.71 × 4 N/A

6 25

600 1000

10 8

N/A 50

FA FA

Nakamae 2008

1.5 T

Thickness 3

15

1000

8

50

FA, ADC

Saito 2008 Ha 2009 Garibotto 2010

1.5 T 1.5 T 1.5 T

Thickness 4 2×2×2 1.88 × 1.88 × 2.3

6 12 35

1000 1000 1000

N/A 6 N/A

N/A N/A N/A

3T

1.875 × 1.875 × 2

28

N/A

N/A

20

Fontenelle 2011

1.5 T

Thickness 5

6

800

3

20

FA, MD FA FA, PDD FA, AD, RD FA, MD

p < 0.001 uncorr. p < 0.017 cluster level p < 0.05 FDR corr. for ROI p < 0.05 for the AC p < 0.01 for other brain regions p < 0.01 Bonferroni corr. p < 0.0001 uncorr. p < 0.005 uncorr.

Bora 2011

Whole brain voxel wise (on a WM mask); voxel wise analysis within the AC Voxel wise restricted to the CB and ALIC; ROI analysis for laterality effect Whole brain voxel wise Whole brain voxel wise (on a WM mask); ROI analysis Whole brain voxel wise using a WM mask for FA and a GM/WM mask for ADC CC ROIs Whole brain voxel wise Whole brain voxel wise; single subject fiber tracking Voxel wise using TBSS

Li 2011

3T

Thickness 3

15

1000

8

50

Oh 2011

1.5 T

2×2×2

12

1000

N/A

Nakamae 2011

1.5 T

Thickness 3

15

1000

Zarei 2011

1.5 T

2.5 × 2.5 × 2.5

60

Fan 2012

1.5 T

Jayarajan 2012 Lochner 2012

Thickness 5

3T

Whole brain voxel wise; tract based MD analysis Whole brain voxel wise (on a WM mask) Seeding tractography of fronto-callosal fibers Voxel wise using TBSS; post hoc tractography analysis Voxel wise using TBSS; ROI subgroup analysis Whole brain voxel wise (on a WM mask) Voxel wise using TBSS

3T

TBSS on selected ROIs

p < 0.05 cluster level corr. p < 0.05 corr.

15

FA, AD, RD FA

p < 0.05 FDR corr. voxel level p < 0.05 FDR corr.

N/A

20

FA

p < 0.05 cluster level corr.

1000

N/A

N/A

FA

p < 0.01 FDR corr.

25

1000

6

50

p < 0.001 uncorr.

Thickness 2.5

32

800

N/A

20

2×2×2

30

1000

N/A

20

FA, AD, RD FA, AD, RD FA, MD

p < 0.05 corr. p < 0.0125 Bonferroni corr.

AC, anterior cingulum; AD, axial diffusivity; ADC, apparent diffusion coefficient; ALIC, anterior limb of the internal capsule; CB, cingulum bundle; CC, corpus callosum; corr., corrected; DTI, diffusion tensor imaging; FA, fractional anisotropy; FDR, false discovery rate; GM, gray matter; MD, mean diffusivity; N/A, not available; PDD, principal diffusion direction; RD, radial diffusivity; ROI, region of interest; T, tesla; TBSS, tract based spatial statistic; uncorr., uncorrected; WM, white matter.


Study

Table 2 Sociodemographic and clinical characteristics of subjects diagnosed with obsessive compulsive disorder and healthy controls included in DTI studies. Sociodemographic characteristics

Clinical characteristics: OCD participants only

Age, years: mean (sd)

Males, n (%)

OCD

HS

OCD

HS

OCD

HS

OCD

Education, years mean (sd)

Szezsko 2005 Cannistraro 2007 Yoo 2007 Menzies 2008 Menzies 2008 (cont.)

15 8 13 30 30

38.5 (10.9) 26.4 (5.9) 27.8 (7.3) 32.2 (11.1)

38.5 (11.8) 23.3 (1.5) 26.9 (7.0) 33.7 (11.2) 37.2 (13.3)

10 (66) 3 (38) 8 (61) 9 (30)

10 (66) 4 (40) 8 (61) 10 (33) 9 (30)

Nakamae 2008 Saito 2008 Ha 2009 Garibotto 2010 Bora 2011 Fontenelle 2011c Li 2011 Oh 2011

15 16 25 15 21 9 23 20

15 10 13 30 30 I-degree relatives 15 16 25 16 29 9 23 19

29.7 (6.9) 28.7 (9.8) 23.5 (5.3) 31.8 (7.9) 34.4 (10.6) 26.2 (10.4) 27.6 (9.9) 25.9 (6.7)

29.1 (6.0) 29.9 (9.0) 23.7 (3.9) 29.6 (6.3) 31.4 (8.0) 28.0 (10.2) 26.7 (9.1) 24.6 (3.7)

9 (60) 7 (43) 25 (100) 15 (100) 11 (52) 7 (78) 16 (69) 13 (65)

Nakamae 2011

30

30

31.6 (9.3)

30.8 (8.4)

Zarei 2011a,b Fan 2012

26 27

26 23

16.6 (1.5) 25.5 (7.0)

Jayarajan 2012a,b Lochner 2012c

15 15

15 17

14.13 (1.7) Adults, matched for age

Age at onset Years: mean (sd)

Illness duration Years: mean (sd)

Y-BOCS score: mean (sd)

Medicated patients, n (%)

Comorbidity n (%)

14.8 (1.8) 15.9 (1.8) 15.6 (2.8) 15.9 (0.9) 15.3 (1.8) 15.1 (2.9) Matched for NART Matched for NART

16.9 (7.7) N/A N/A N/A

N/A N/A 7.2 (9.0) N/A

25.9 (4.4) 22.9 (7.0) 30.2 (4.6) 22.1 (5.4)

12 (80) Drug free Treat. naïve 23 (76)

5 (33) Pure OCD 3 (23) Pure OCD

9 (60) 7 (43) 25 (100) 15 (100) 14 (48) 7 (78) 16 (69) 13 (68)

15.2 (2.2) 15.7 (1.8) N/A N/A 13.1 (2.5) 14.0 (1.2) N/A N/A 14.6 (2.2) 15.0 (2.3) N/A N/A 13.3 (2.8) 13.4 (2.7) Matched for IQ

19.6 (8.0) N/A N/A 20.7 (10.8) N/A 11.5 (5.0) N/A 18.6 (7.2)

N/A N/A 6.3 (4.8) 10.1 (8.0) N/A N/A 6.3 (5.4) 7.7 (5.1)

29.0 (5.3) 26.0 (5.3) 20.2 (5.4) 28.2 (5.0) 19.2 (5.4) 28.5 (4.8) 23.2 (5.1) 20.5 (6.6)

2 (13) Pure OCD Pure OCD Pure OCD Pure OCD 7 (77) Pure OCD 2 (10)

14 (46)

15 (50)

N/A

25.0 (9.6)

6.7 (7.1)

23.8 (5.7)

16.5 (1.4) 28.8 (7.6)

14 (54) 17 (63)

14 (54) 15 (65)

Matched for IQ 14.0 (2.9) 14.6 (3.7)

11.2 (2.8) 20.4 (7.9)

N/A 4.8 (4.0)

19.5 (7.6) 22.0 (4.9)

14.31 (2.1)

8 (53) 10 (66)

8 (53) 9 (53)

8.13 (1.8) Matched for IQ

8.87 (2.3) 11.2 (4.9)

12.7 (1.8) N/A

1.40 (1.04) 22.9 (3.6)

15 (100) 13 (81) 10 (40) 13 (86) 10 (47) 7 (77) 13 (56) 15 (75) Treat. Naïve 5 (25) drug free 14 (46) Treat. Naïve 16 (53) drug free 16 (61) 13 (48) treat. naïve 14 (51) drug free 13 (86) 9 (60)

HS

Pure OCD Pure OCD Pure OCD 6 (40) Pure OCD


Sample size Author

IQ, intelligence quotient, N/A, not available; NART: National Adult Reading Test. a Studies including children or adolescent patients. b Studies including child onset patients assessed in adolescence. c Studies including child onset patients assessed in adult age

2861

2862


study (Garibotto et al., 2010) explored the association between fiber integrity and decision making and visuo-spatial impairments.

3. Results 3.1. White matter tracts implicated in the classical cortical–striatal–thalamic-cortical model of OCD 3.1.1. Evidence supporting the fronto-striatal model of OCD. Data synthesis and integration Consistently with the assumption that the cortical–striatal– thalamic-cortical loop is a major site of synaptic dysfunction in OCD, the evidence we review here reports abnormal microstructure in OCD patients in the major WM tracts forming the physical foundation for connectivity of cortical and subcortical brain regions implicated in the pathophysiology of the disorder. Specifically, 10 of the considered studies show abnormalities in the cingulum (Cannistraro et al., 2007; Fontenelle et al., 2011; Garibotto et al., 2010; Ha et al., 2009; Jayarajan et al., 2012; Li et al., 2011; Lochner et al., 2012; Nakamae et al., 2011; Szeszko et al., 2005; Zarei et al., 2011), while 6 papers report altered WM integrity in the anterior and posterior limb of the IC (Cannistraro et al., 2007; Fontenelle et al., 2011; Jayarajan et al., 2012; Lochner et al., 2012; Nakamae et al., 2011; Yoo et al., 2007). Such evidence suggests altered microstructure in the affective loop subserved by the cinglum bundle and linking the limbic system to the ACC and the medial PF cortex, and abnormal connectivity in corticothalamic projections, forming the major part of the IC. On the other hand, the direction of the findings is contradictory among studies since the reviewed evidence reports either increased FA (putatively related to increased myelination and neuronal remodeling and probably leading to functional hyperconnectivity) or decreased FA (a marker of decreased or disrupted myelination, or reduced coherence of fibers, functionally determining hypoactivity of the circuit) in both WM tracts. Table S1 summarizes the brain regional diffusivity changes reported in individual studies, while a summary of findings regarding the frontal-subcortical circuit in OCD is depicted in Table 3. A first consideration is that medication status may have determined the observed inconsistency in findings, given that previous pre- and post-treatment neuroimaging studies showed that the hyperactivity in frontal-sub-cortical circuits in untreated OCD patients was normalized in response to treatment (Benkelfat et al., 1990; Saxena et al., 2002; Swedo et al., 1992). There are also reports that the SSRI Fluoxetine can increase astrocytic glycogenolysis, thereby ameliorating neuronal energy supply and improving diffusivity coefficients on DTI (Sijens et al., 2008). Indeed, if we consider the only 3 studies conducted on drug-naïve patients, 2 of them (Cannistraro et al., 2007; Yoo et al., 2007) consistently report increased FA in the IC and the CB, while the 3rd (Nakamae et al., 2011) showed just a trend for lower FA in the right cingulum and in the ALIC. Moreover, 12 weeks of treatment with citalopram induced a normalization of FA in drug naïve OCD patients compared with HS (Yoo et al., 2007), suggesting that WM alterations are associated with the pathophysiology of OCD and that abnormalities may be partly reversible with treatment. All the above considered, we propose that the contradictory picture emerged from the reviewed studies might have been determined by the inclusion of mixed samples of treated and untreated patients. Also the small samples and the varied methods employed may have contributed to inconsistencies in findings. However, within moderator variables (i.e. study characteristics that may modify the effect estimate) medication status and demographic characteristics, more than variations in the analysis technique, seem to exert a crucial effect on the variable of interest. As a case in

point, contradictory findings in the cingulum were obtained using the same technique, while homogeneous findings are reported in the same region, despite heterogeneity in the employed method. On the other hand, while the FA metric gives a representation of WM integrity, additional measures of diffusivity along the axon (AD), perpendicular to the main axis (RD) or the invariant magnitude of water diffusion (MD) or even the estimation of the main fiber direction within each voxel (PDD) yield more specific information about the biological processes that underlie the observed changes in FA (Song et al., 2003). In this regard, changes in AD, suggestive of axonal damage, possibly deriving from inadequate development, were described in the cingulate WM (Li et al., 2011). Similarly, increased water diffusion only in the direction of the fibers (increased AD, no difference in FA) was reported in the right and left cingulum (Jayarajan et al., 2012). Although the interpretation of overlapping diffusion indices is not univocal (Zhang et al., 2009) and largely based on animal models studies (e.g.: DeBoy et al., 2007), increased AD without FA modifications suggests a generalized increase in extracellular space due, for example, to the axonal atrophy expected in wallerian degeneration. In this pathophysiological model of WM damage, degradation of WM microstructure occurs secondary to GM pathology and indeed, several VBM studies consistently show that GM volume reduction in the cingulate gyrus is coupled with reduced WM density in both the anterior (Riffkin et al., 2005; Togao et al., 2010) and posterior cingulum (Riffkin et al., 2005). Likewise, incipient axonal loss (i.e. increased MD or AD) with no changes in fiber directionality (i.e. no PDD or FA alterations) is described in the left and right CB (Fontenelle et al., 2011; Garibotto et al., 2010) and eventually coupled with a decline in myelin integrity (i.e. increased AD and RD, no difference in FA (Jayarajan et al., 2012)). Turning to the IC, the FA abnormalities reported in the reviewed studies point toward hyper (i.e. increased FA) (Cannistraro et al., 2007; Lochner et al., 2012; Yoo et al., 2007; Zarei et al., 2011), as well as hypo-connectivity (i.e. decreased FA) (Cannistraro et al., 2007; Fontenelle et al., 2011) among the frontal cortex and the thalamus. Considering the profiles of FA, MD, parallel and perpendicular diffusivity, abnormal structural connectivity (increased AD, Jayarajan et al., 2012) coupled with tract poor compaction (increased MD and RD) is suggested in the P-LIC (Fontenelle et al., 2011; Jayarajan et al., 2012). On the other hand, increased connectivity in regions closer to the thalamus (increased AD, Jayarajan et al., 2012; increased FA, no changes in MD, Lochner et al., 2012) and disrupted or decreased myelination in regions closer to the cortex were reported in the A-LIC (increased RD, Jayarajan et al., 2012; decreased FA, Lochner et al., 2012). Such findings point toward a decrease in the neuronal density of the cingulate WM and CB, with no changes in the structural organization of the tissue, while pathologically increased connectivity in the IC would be likely due to increased directional coherence. 3.1.2. How might microstructural abnormalities within the classic frontostriatal circuit relate to the expression of OCD symptoms? The question of whether clinical variables correlate with WM microstructural alterations is particularly relevant to the evolving model of OCD pathophysiology. If indeed, local changes in WM connectivity show a relationship with symptom ratings, then the observed microstructural alterations should be considered state markers reflecting a more severe clinical condition, and not underlying traits of OCD. On the other hand, WM tissue changes in certain regions might be an epiphenomenon of contingent comorbid illnesses, a by-product of medication usage, or progressive alterations evolving along dynamic trajectories during the illness course. Finally, partially distinct neural circuits may mediate different symptom dimensions, thus accounting for the phenotypic heterogeneity of OCD.

Table 3 Fractional anisotropy changes in the classical cortical–striatal–thalamic-cortical circuit in subjects diagnosed with obsessive compulsive disorder. Controls mean ± sd

% change OCD

Additional measures

Cingulum FADOC < FAHS Nakamae 2011

NA

NA

↓

Ha 2009

NA

NA

↓

Szezsko 2005

NA

NA

↓

FADOC > FAHS Li 2011

0.41± 0.02

0.37 ± 0.02

+8

Zarei 2011

NA

NA

↑

FADOC = FAHS Jayarajan 2012

NA

NA

None

Increased AD, no changes in RD

Lochner 2012

NA

NA

None

Decreased MD in L AC and R C body

Nakamae 2008

NA

NA

None

Yoo 2007

NA

NA

None

CB FADOC < FAHS Garibotto 2010

NA

NA

↓

Cannistraro 2007

NA

NA

↓

Clinical correlation

Interpretation

No significant correlation between Y-BOCS score and the FA Mean FA value of the AC was significantly correlated with the Y-BOCS total score and obsessive subscale but not with the IQ, duration of illness, and BDI score.

Lower FA in the CC, the cingulum, and the ALIC represent trait markers rather than state markers A significantly lower FA was observed in the L AC in patients with a predominant aggressive/checking symptom dimension. The finding further support that limbic circuits may be more involved in obsessions An abnormality that involves the white matter in the AC could contribute to the pathogenesis of OCD through aberrant connectivity with other cortical and subcortical brain regions. A defect in connectivity that involves the PC may also play a role in the neurobiology of OCD

Results remained significant when patients with comorbid depression were excluded. No significant correlation between Y-BOCS score and the average measure of FA across the 3 regions within the AC and negative correlation between Y-BOCS scores and the average measure of FA across the 4 regions outside the AC (including the PC) Increased AD and normal RD in the R sup frontal WM

No significant correlations between FA, AD, RD and Y-Bocs score, HARS, and HDRS scores or illness duration

No significant correlation between FA in the L cingulum and total CY-BOCS severity scores

No significant correlation between the severity of OCD on the CY-BOCS scale and diffusivity parameters nor between dosage of medication and duration of treatment with diffusivity values Significant positive correlation between the total Y-BOCS score and MD in the L AC and a significant negative correlation between MADRS and HAM-A scores and MD in the R C body

Increased FA might reflect increased WM connectivity, while increased AD could result from reduced axonal density or caliber, increasing the EAS. Microstructural abnormalities in frontal-subcortical circuits are important in OCD pathophysiology and not in determining clinical status The findings support a developmental hypothesis of OCD that is characterized by hyperconnectivity of different brain regions resulting from premature myelination. Increased FA may also reflect increased crossing fibers or less organized connectivity in OCD The potential factors that might have contributed to the differential findings include younger population, the possible influence of brain development on WM architecture, and medication effects Decreased MD may reflect the increased activity and connectivity in the cingulum


OCD mean ± sd

The studied subjects were asians and this may be an important reason for the inconsistency with other studies The inconsistency with other studies might be due to the younger age, lower amount of co-morbid illness, and the drug-naive state of the studied subjects

No changes in PDD were associated with decreased FA in B CB Significant difference in the asymmetry index (left > right) in the OCD group

No significant correlation between Y-BOCS scores or degree of impairment in decision-making and visuospatial ability and FA values in the CB Findings of decreased FA within R CB could represent decreased or disrupted myelination, or reduced coherence of fibers within that tract 2863

OCD mean ± sd

Controls mean ± sd

% change OCD

Additional measures


Interpretation

FADOC > FAHS Cannistraro 2007

NA

NA

↑

Significant difference in the asymmetry index (left > right) in the OCD group

Exaggerated coherence with respect to WM orientation, as indicated by elevated FA in L CB, may mediate or parallel the exaggerated function and functional connectivity among these nodes in OCD

FADOC = FAHS Fontenelle 2011

NA

NA

None

Higher MD in the L CB

The discrepant finding regarding CB anisotropy may be due to the utilization of nonisotonic voxels in some studies and the assessment of small sexually and clinically heterogeneous samples

FADOC < FAHS Lochner 2012

0.39 ± 0.03

0.43 ± 0.04

−9

No significant differences in MD

No significant correlations between the total Y-BOCS score and FA.

Fontenelle 2011

NA

NA

↓

Higher MD in the B P-LIC

Overall Y-BOCS was positively correlated to MD values in the B A-LIC

Nakamae 2011

NA

NA

↓

Trend for lower FA in the L ALIC

No significant correlation between Y-BOCS score and FA

FADOC > FAHS Lochner 2012

0.68 ± 0.04

0.63 ± 0.04

+7


No significant correlations between the total Y-BOCS score and FA. However, the MADRS and HAM-A scores correlated significantly with FA in B ALIC

Zarei 2011

NA

NA

↑

Cannistraro 2007

NA

NA

↑

Yoo 2007

NA

NA

↑


NA

NA

None

2864

Table 3 (Continued)

IC

No significant laterality differences in FA in the ALIC

Drug-naïve OCD patients showed increased FA in the B P-LIC. None of the clinical measures, such as the Y-BOCS, HDRS, BDI, and BAI, was significantly correlated to FA. The increased FA was normalized after 12 weeks of citalopram

Increased AD in B A-LIC and L P-LIC and increased RD in the L P-LIC

No significant correlation between the severity of OCD on the CY-BOCS scale and diffusivity parameters nor between dosage of medication and duration of treatment with diffusivity values

Increased FA in B ALIC in patients with OCD may reflect increased connectivity in regions closer to the thalamus. The correlation between increased FA in the ALIC and depression and anxiety scores, is consistent with increased activity/connectivity in this area The implication of multiple WM tracts in pediatric OCD supports a developmental hypothesis of OCD characterized by hyperconnectivity of different brain regions and premature myelination Exaggerated coherence with respect to WM orientation, as indicated by elevated FA in left ALIC, may mediate or parallel the exaggerated function and functional connectivity among these nodes in OCD The higher FA in OCD patients represents a rapid or increased information transmission caused by filtering failures at the level of the thalamus. Consistently with the observation that hyperactivity in frontal sub-cortical circuits in untreated OCD patients is normalized in response to treatment, 12 weeks of treatment with citalopram induced a normalization of FA in B P-LIC Areas showing increase in both AD and RD might be indicative of pathologically hyperconnected WM tracts with poor compaction or demyelination. The presence of microstructural abnormalities indicative of reduced coherence in the A-LIC is broadly in agreement with the widely accepted neurobiological model of OCD.

AC, anterior cingulate; AD, axial diffusivity; ALIC, anterior limb of the internal capsule; B, bilateral; BAI, Beck Anxiety Inventory; BDI, Beck Depression Inventory; C, cingulum; CB, cingulum bundle; CC, corpus callosum; CY-BOCS, Children Yale Brown Obsessive Compulsive Scale; EAS, Extra Axonal Space; FA, fractional anisotropy; HAM-A, Hamilton Anxiety Scale; HARS, Hamilton Anxiety Rating Scale; HDRS, Hamilton Depression Rating Scale; IC, internal capsule; IQ, intelligence quotient; L, left; MADRS, Montgomery-Asberg Depression Rating Scale; MD, mean diffusivity; NA, non available; OCD, obsessive compulsive disorder; PC, posterior cingulum, PDD, principal diffusion direction; PF, prefrontal; P-LIC, posterior limb of the internal capsule; R, right; RD, radial diffusivity; sd, standard deviation; sup, superior; WM, white matter; Y-BOCS, Yale Brown Obsessive Compulsive Scale.


Symptom severity was positively correlated with WM FA in various regions included the ALIC and P-LIC

Decreased FA in the R ALIC closer to PF projections may indicate disrupted or decreased myelination in those regions or decreased fiber density or coherence Findings with regard to the IC dovetail with current models suggesting the occurrence of a hypoactivity of frontal-striatal circuits, particularly the so-called “indirect” pathways, among patients with OCD Findings of lower FA in the CC, the cingulum, and the ALIC represent trait markers rather than state markers


Pertaining to the fronto-striato-thalamo-cortical loop classically implicated in OCD, while microstructural abnormalities in the cingulum are not correlated with symptom severity (Jayarajan et al., 2012; Li et al., 2011; Nakamae et al., 2011; Szeszko et al., 2005; Zarei et al., 2011) (but see Ha et al., 2009; Lochner et al., 2012), changes in fiber directionality within the IC seem to be interrelated with clinical variables (Fontenelle et al., 2011; Lochner et al., 2012; Zarei et al., 2011) (see Table 2). Such observation might suggest that altered microstructure in the cingulate WM plays a role in the neurobiology of OCD and may represent a trait, rather than a state marker. On the other hand, abnormal connectivity among the frontal cortex and the thalamus, as expressed by changes in diffusivity measures within the IC, may mediate symptoms expression. However, different patterns of regional changes in WM integrity may characterize patients with different symptom dimensions, as the cingulum, a WM structure important in cortico-limbic circuitry, might be primarily involved in obsessions and mediate the expression of aggressive/checking symptoms (Ha et al., 2009). Indeed, the reported correlation between altered microstructure and symptom severity was specifically observed in the ALIC (Fontenelle et al., 2011; Zarei et al., 2011) a major efferent tract which includes the anterior thalamic radiation carrying fibers between the thalamus, the frontal cortex and the AC. It is therefore possible that OCD symptoms emerge from disturbed connectivity among the thalamus and the frontal cortex, either consequent to decreased fiber density or coherence in the ALIC, (possibly leading to hypoactivity in fronto-striatal circuits, Fontenelle et al., 2011) or resulting from exaggerated coherence in the orientation of the WM tracts connecting the thalamus to the frontal cortex (potentially determining increased activity/connectivity in the circuit, Lochner et al., 2012; Zarei et al., 2011). The hyperconnectivity hypothesis might be reinforced by the observation of a positive correlation between increased FA and anxiety symptom severity in bilateral anterior regions of the IC close to the thalamus (Lochner et al., 2012). The symmetrical finding of a positive relationship between OCD severity scores and tissue density in the same region (Duran et al., 2009) would suggest that the neurobiological properties underlying individual variations in WM volume and FA are probably somewhat overlapping in the disorder. Specifically, it is probable that increased FA in the ALIC of OCD patients is primarily related to increased packing density and fiber diameter or directional coherence (Song et al., 2003) as increased myelination will probably increase WM volume, but may decrease FA (Fjell et al., 2008). However, since the contributors to DTI anisotropy in living white matter have not yet been resolved, while the molecular basis and pathophysiological consequences of abnormal FA in OCD remain unclear (Cannistraro et al., 2007), such interpretation is necessarily, tentative. It is also important to notice that the three studies showing a correlation between symptom severity and ALIC microstructure were performed on child onset OCD patients, assessed in adult age or during adolescence. This observation would suggest that both genetic and environmental risk factors may contribute to the hypothesized abnormal connectivity among the frontal cortex and the thalamus implied by changes in diffusivity measures. Concurrently, two studies showed that alterations in frontal–subcortical tracts (the P-LIC, the retrolenticular part of the IC and the WM around the striatum) (Fan et al., 2012; Yoo et al., 2007) normalized after SSRI treatment. Since the reported changes were observed in regions that contain thalamic radiations (Mori et al., 2008), projecting to the PF cortex via the superior thalamic radiation and to occipital, temporal and parietal cortices through the posterior thalamic radiation, it is conceivable that the clinical efficacy of SSRI may be related to a post-treatment normalization of functional connectivity in dorsalfrontal-striatal networks and/or in thalamocortical pathways to posterior brain regions.

2865

3.2. Additional white matter regions and fiber bundles putatively involved in OCD pathophysiology 3.2.1. Evidence supporting the extended model of OCD circuitry. Data synthesis and integration The observation that OCD mechanisms involve a more widespread network of cerebral dysfunctions than previously thought (Menzies et al., 2008a; Piras et al., 2013) suggests that functional and anatomical connectivity, in particular of the long intrahemispheric fibers connecting the orbitofrontal cortex to parietal and occipital cortices or of the interhemispheric connections spanning left and right PF areas and superior temporal/posterior parietal cortices, might be altered in the disorder. Likewise, WM abnormalities may be present also in regions outside the classically implicated circuit, and indeed a defect in connections between the parietal lobe and areas of the orbitofronto-striatal circuit or altered anatomical connectivity between lateral frontal and parietal regions, or among the OFC and posterior parietal and occipital association cortices may provide a pathological basis for both the neuropsychological deficits and the behavioral symptoms observed in OCD patients (Menzies et al., 2008b; Okasha et al., 2000). Finally, changes in the inter-hemispheric connectivity subserved by the CC, spanning left and right PF areas and superior temporal/posterior parietal cortices, might explain some of the neuropsychological disorders evident in OCD patients (Di Paola et al., 2012). The evidence we review here confirms the involvement of the previously mentioned circuits in OCD pathogenesis since 6 studies found WM abnormalities in frontal and PF regions (Fan et al., 2012; Ha et al., 2009; Menzies et al., 2008b; Nakamae et al., 2008) and in parietal, temporal and occipital areas (Fan et al., 2012; Menzies et al., 2008b; Szeszko et al., 2005; Yoo et al., 2007). Table S1 summarizes the brain regional diffusivity changes reported in individual studies while Table 4 includes a summary of the findings on WM microstructural alterations in regions outside the orbitofrontostriatal circuit. Concurrently, 4 additional investigations report changes in fiber directionality in long-range intra-hemispheric tracts (Fontenelle et al., 2011; Garibotto et al., 2010; Jayarajan et al., 2012; Zarei et al., 2011), while 10 studies (Bora et al., 2011; Fontenelle et al., 2011; Garibotto et al., 2010; Jayarajan et al., 2012; Li et al., 2011; Nakamae et al., 2011; Oh et al., 2011; Saito et al., 2008; Yoo et al., 2007; Zarei et al., 2011) showed microstructural WM abnormalities in the CC. A summary of the findings regarding intra-hemispheric WM bundles is detailed in Table 5 while Table 6 includes a synopsis of the evidence about inter-hemispheric WM bundles. Consistently with the evidence demonstrating structural and functional abnormalities in parietal and medial frontal regions (Menzies et al., 2008a) the results of this review indicate reduced coherency in WM fibers in both areas, thus suggesting a defect in connections between the parietal lobe and frontal regions, critical for cognitive control processes. Intriguingly, changes in diffusion indices in the medial frontal and parietal areas showed to be significantly correlated in OCD patients, implying a pathological relationship between WM abnormalities in these two regions in OCD (Menzies et al., 2008b). Moreover, the fact that the same abnormalities were found in healthy subjects with a genetic risk for OCD (Menzies et al., 2008b) suggests they are endophenotypes, probably mediating cognitive deficits in response inhibition evident in both OCD patients and unaffected first-degree relatives (Menzies et al., 2008b). We also found consistent evidence of altered anatomical connectivity in long-range intra-hemispheric fibers linking parietotemporal–occipital regions to distinctive areas of the PF cortex and vice versa. Structural anatomical abnormalities in these bundles positively correlated with both symptom severity and

2866

Table 4 Fractional anisotropy changes in regions outside the orbitofronto-striatal circuit in subjects diagnosed with obsessive compulsive disorder. Controls mean ± sd

% change OCD

Additional measures


Interpretation

Frontal WM FADOC < FAHS Fan 2012

NA

NA

↓

Increased RD and MD, but normal AD in the L medial sup frontal WM and around the L striatum

None of the diffusion multi-parameters correlated with clinical measures. After 12-weeks SSRI treatment, no significant changes in DTI parameters in the L medial superior frontal WM, but significant changes in RD in the L striatum. The latter significantly correlated with decreases in Y-BOCS compulsive score

Higher RD in combination with no change in AD among OCD patients suggests a disruptions of myelin integrity

FADOC > FAHS Ha 2009

NA

NA

↑

Significantly higher FA was observed in the B middle PF WM and L sup PF WM in patients with a predominant contamination/cleaning symptom dimension

Menzies 2008

0.38 ± 0.02

0.36 ± 0.02

+6

Mean regional FA in the medial frontal and parietal WM was significantly correlated between regions for OCD but not relatives or controls. No significant correlation between abnormal FA and symptom severity scores

Nakamae 2008

R 0.48 ± 0.03 L 0.48 ± 0.03

0.42 ± 0.01 0.43 ± 0.02

+12 +11

Higher ADC in the L medial frontal gyrus

Significant correlation between FA values in bilateral semioval center extending to subinsular regions in patients. No significant correlations among FA, ADC and Y-BOCS scores

Frontostriatal circuits may be more crucial in compulsions. The PF involvement in contamination/cleaning dimension provide additional evidence that OCD may be a heterogeneous condition, while two kinds of important neural circuits in the pathophysiology of the disorder are differentially involved in different symptom dimensions The finding of increased FA in R medial frontal WM in patients and relatives could be interpreted in terms of increases in myelination or increased numbers of fibers in this region, potentially caused by differences during brain structure development in individuals at increased risk of OCD Higher FA could reflect an increase of connectivity between some brain areas leading to OCD symptoms. Abnormalities in the subinsular WM involved in disgust processing, may play a role in OCD symptoms. Higher ADC in the medial frontal cortex may reflect less GM and dysfunction of this region may cause dysregulation of disgust

FADOC = FAHS Fan 2012

NA

NA

None

Increased AD, RD and MD, but normal FA in the right frontal lobe and left insula

After 12-weeks SSRI treatment, no significant changes in DTI-derived parameters were found in the right frontal lobe and left insula

Higher AD along with higher RD and MD, and normal FA suggests a pattern of both myelination and axonal deficits

Parietal WM FADOC < FAHS Fan 2012

NA

NA

↓

Increased RD and MD accompanied the decreased FA in the tempo-parietal lobe

None of the diffusion multi-parameters correlated significantly with Y-BOCS total score, HAM-A, HAMD, age at onset or illness duration. No significant changes in DTI-derived parameters were found in the T/P lobe

Garibotto 2010

NA

NA

↓

Menzies 2008

0.38 ± 0.01

0.40 ± 0.02

−5

Decreased FA, increased MD and higher RD, in combination with no change in AD, suggest a disruption of myelin integrity. The temporo-parietal junction is a complex sensory cortical area, which is involved in perception and awareness and probably participates in the frontal-subcortical circuit All but one cluster reported by Szeszko and coll. (2005) were also significant in the present sample, as the right parietal FA reduction previously reported by Menzies and coll. (2008) Decreased FA in R parietal WM is compatible with data suggesting dysfunction within large-scale neural systems in OCD and supportive of dysconnectivity within these systems

Mean regional FA in the medial frontal and parietal WM was significantly correlated between regions for OCD but not relatives or controls. No significant correlation between abnormal FA and symptom severity scores


OCD mean ± sd

Szezsko 2005

NA

NA

↓

Greater Y-BOCS scores correlated significantly with lower FA in the parietal lobe WM bilaterally. Lower parietal lobe FA correlated significantly with the obsession subtotal. No significant correlation with the HAMA or HAMD scores

FADOC = FAHS Yoo 2007

NA

NA

None

The OCD group showed a significant pre- to post-treatment decrease in FA in the right posterior thalamic radiation on the R parietal lobe. The change in FA from the baseline to the follow-up was not correlated to the changes in the scores of the clinical measures

R 0.36 ± 0.02

0.38 ± 0.02

−5

Significantly decreased FA in OCD patients compared to their first degree relatives. No significant correlation in patients between FA in these temporal regions and Y-BOCS score

The observed abnormality may represent a marker of disease state

L 0.40 ± 0.02

0.42 ± 0.02

−4.7

FADOC > FAHS Yoo 2007

NA

NA

↑

None of the clinical measures, such as the Y-BOCS, HDRS, BDI, and BAI, was significantly correlated to FA. No effect of pharmacotherapy on FA values in the R sup temporal region

Hyperconnectivity (increased FA = higher density, more directional coherence, and greater degree of myelination of fibers) between the sup temporal region and other cortical (OFC) and subcortical (putamen, caudate) areas may cause OCD symptoms.

FADOC = FAHS Li 2011

NA

NA

None

Positive correlation between FA in the L middle temporal gyrus WM and Y-BOCS total score and obsessive subscale score

The WM of temporal lobe might have a role in the pathology of OCD

Occipital WM FADOC < FAHS Fan 2012

NA

NA

↓

Decreased FA, increased MD and higher RD, in combination with no change in AD, suggest a disruption of myelin integrity

Szezsko 2005

NA

NA

↓

None of the diffusion multi-parameters correlated significantly with Y-BOCS total score, HAMA, HAMD, age at onset or illness duration. No significant changes in FA after SSRI treatment No correlation with Y-BOCS score, HAMA and HAMD

Increased RD and MD


Temporal WM FADOC < FAHS Menzies 2008

Parietal lobe WM microstructure plays a role in mediating obsessions and compulsive behavior, possibly through disruption of cortical-cortical and/or cortical-subcortical connectivity with other brain regions implicated in the pathophysiology of OCD

The lingual gyrus has a role in processing emotionally charged visual stimuli. Altered WM connectivity in this area might have relevance for anxiety disorders such as OCD, where an abnormality in arousal and sensory processing is considered important to phenomenology

AC, anterior cingulate; AD, axial diffusivity; ALIC, anterior limb of the internal capsule; B, bilateral; BAI, Beck Anxiety Inventory; BDI, Beck Depression Inventory; C, cingulum; CB, cingulum bundle; CC, corpus callosum; CY-BOCS, Children Yale Brown Obsessive Compulsive Scale; EAS, Extra Axonal Space; FA, fractional anisotropy; HAM-A, Hamilton Anxiety Scale; HARS, Hamilton Anxiety Rating Scale; HDRS, Hamilton Depression Rating Scale; HS, healthy subjects; IC, internal capsule; IQ, intelligence quotient; L, left; MADRS, Montgomery-Asberg Depression Rating Scale; MD, mean diffusivity; NA, non available; OCD, obsessive compulsive disorder; PC, posterior cingulum, PDD, principal diffusion direction; PF, prefrontal; P-LIC, posterior limb of the internal capsule; R, right; RD, radial diffusivity; sd, standard deviation; sup, superior; WM, white matter; Y-BOCS, Yale Brown Obsessive Compulsive Scale.

2867

2868

Table 5 Fractional anisotropy changes in intra-hemispheric WM bundles in subjects diagnosed with obsessive compulsive disorder. Controls mean ± sd

% change OCD

Additional measures


Interpretation

IFOF FADOC < FAHS Garibotto 2010

NA

NA

↓

Significant changes in the distribution of PDD in B IFOF

No correlation with Y-BOCS scores. Negative correlation between FA, Iowa Gambling Task scores (disadvantageous choices) and the time required for the Trail Making Test-A task in B IFOF

Changes in fiber coherency (FA reduction) and fiber directionality (PDD alterations) might be explained by an abnormal development of WM tracts, as this is associated with regionally decreased fiber organization and myelinization. Microstructural abnormalities in the IFOF are particularly relevant for the study of OCD, since this bundle represents the main long intrahemispheric connection to and from the orbitofrontal cortex, which has long been implicated in OCD pathogenesis.

FADOC > FAHS Zarei 2011

NA

NA

↑


NA

NA

None

Increased AD in L IFOF and increased RD in B IFOF

No significant correlation between the severity of OCD on the CY-BOCS scale and diffusivity parameters nor between dosage of medication and duration of treatment and diffusivity values

Areas showing increase in both AD and RD might be indicative of pathologically hyperconnected WM tracts with poor compaction or demyelination

SLF FADOC < FAHS Fontenelle 2011

NA

NA

↓

Increased MD in B SLF

Overall Y-BOCS was positively correlated to MD values in the B SLF

Garibotto 2010

NA

NA

↓

Significant changes in the distribution of PDD

Significant correlation between higher Y-BOCS scores and lower FA value in the major long fiber tracts (SLF included). Negative correlation between FA, Iowa Gambling Task scores (disadvantageous choices) and the time required for the Trail Making Test-A task in B SLF

Since the SLF is an heterogeneous set of bi-directional fibers that connects the postrolandic regions (i.e., parieto-temporal association areas) with distinctive areas of the PFl cortex and vice versa, the finding is consistent with the increasingly recognized role played by posterior brain regions in OCD Signal changes in OCD may be due primarily to disorganization of WM fibers, reflected by local changes in fiber directionality, as specifically indicated by the PDD alterations. This alteration might result from changes in neural migration as postulated for developmental diseases


NA

NA

↑


NA

NA

None

The implication of multiple WM tracts in pediatric OCD supports a developmental hypothesis of OCD characterized by hyperconnectivity of different brain regions and premature myelination

Increased AD and RD in B SLF

Symptom severity was positively correlated with WM FA in various regions including the B SLF



Areas showing increase in both AD and RD might be indicative of pathologically hyperconnected WM tracts with poor compaction or demyelination. Disturbances in connectivity in the SLF can result in impaired attention and spatial working memory


OCD mean ± sd

ILF FADOC > FAHS Zarei 2011

NA

NA

↑


NA

NA

None

Increased AD and RD in the L ILF


Garibotto 2010

NA

NA

None

Significant changes in PDD in B ILF

No significant correlation between FA, Y-BOCS scores and degree of neuropsychological impairment

UF FADOC > FAHS Zarei 2011

NA

NA

↑


NA

NA

None

CST FADOC > FAHS Zarei 2011

NA

NA

↑


Symptom severity was positively correlated with WM FA in various regions, particularly the L UF



The presence of increased RD raises the possibility of defective oligodendrocyte development, defective maturation of proteins, and lipids forming myelin or demyelination in the WM. Microstructural abnormalities in both the UF and ILF could serve as neurobiological correlates for the emotional processing deficits seen on neuropsychological studies of OCD patients.

Symptom severity was positively correlated with WM FA in various regions, particularly the CST



Increased RD in B UF, no changes in AD

Areas showing increase in both AD and RD might be indicative of pathologically hyperconnected WM tracts with poor compaction or demyelination. Microstructural abnormalities in both the UF and ILF could serve as neurobiological correlates for the emotional processing deficits seen on neuropsychological studies of OCD patients Signal changes in OCD may be due primarily to disorganization of WM fibers, reflected by local changes in fiber directionality, as specifically indicated by the PDD alterations

AD, axial diffusivity; B, bilateral; CST, cortico-spinal tract; CY-BOCS, Children Yale Brown Obsessive Compulsive Scale; FA, fractional anisotropy; HS, healthy subjects; IFOF, inferior fronto-occipital fasciculus; ILF, inferior longitudinal fasciculus; L, left; MD, mean diffusivity; NA, non available; OCD, obsessive compulsive disorder; PDD, principal diffusion direction; PF, prefrontal; RD, radial diffusivity; sd, standard deviation; SLF, superior longitudinal fasciculus; UF, uncinate fasciculus; WM, white matter; Y-BOCS, Yale Brown Obsessive Compulsive Scale.

2869

2870

Table 6 Fractional Anisotropy changes in inter-hemispheric WM bundles in subjects diagnosed with obsessive compulsive disorder. Controls mean ± sd

% change OCD

Rostrum FADOC < FAHS Oh 2011

0.38 ± 0.02

0.4 ± 0.02

−5

Saito 2008

0.55 ± 0.07

0.69 ± 0.06

−20

Genu FADOC < FAHS Oh 2011

0.34 ± 0.03

0.37 ± 0.03

−8

FADOC > FAHS Li 2011

0.49 ± 0.03

0.44 ± 0.03

+10

Zarei 2011

NA

NA

↑


NA

NA

Lochner 2012

NA

Fontenelle 2011

Additional measures


Interpretation


Positive correlation between FA, BAI score and Y-BOCS obsessive subtotal, once comorbidity, depression level and past medication were regressed out Negative correlation between FA and Y-BOCS total score

Reduced integrity in ventral CC tracts (important for task-switching) and imbalanced dorsal/ventral networks originate the anxiety symptoms Microstructural abnormalities in the rostrum and fiber inegrity abnormalities in the orbital prefrontal region projecting into the rostrum are implicated in OCD symptom severity

Negative correlation between FA and Y-BOCS compulsive subtotal, once comorbidity, depression level and past medication were regressed out

Reduced fiber integrity in dorsal CC tracts (important for working memory and executive functions) correlates with higher compulsions

AD was higher in OCD, while RD was not significantly different

No significant correlations between FA, AD and RD and compulsive subscale, HARS, and HDRS scores or illness duration no significant effect of medication on FA changes, nor correlation with global OCD severity

Deficits of specific WM circuits may be pathophysiologically significant for OCD and account for the observed increased AD and FA the findings support a developmental hypothesis of OCD that is characterized by hyperconnectivity of different brain regions

None

Increased AD and RD

No significant correlation between FA, Y-BOCS score and dosage of medication

NA

None

No significant differences in terms of FA and MD

No significant correlations between the total Y-BOCS score and FA or MD

NA

NA

None

No differences in FA or MD

Changes in FA in the genu of CC were correlated to total Y-BOCS. Overall Y-BOCS was positively correlated to MD values

Areas showing increase in both AD and RD might be indicative of pathologically hyperconnected WM tracts with poor compaction or demyelination. Possibly, changes in the CC are particularly apparent in pediatric patients with OCD. Nevertheless, no association was found between age at onset of OCD and DTI measures The significant correlation between both FA and MD and the severity of OC symptoms in the genu suggests changes in the WM connectivity spanning left and right prefrontal areas and the superior temporal cortex

Body FADOC < FAHS Bora 2011

NA

NA

↓

Increased RD with no changes in AD

Nakamae 2011

NA

NA

↓

Depression and anxiety symptoms in OCD were not associated with significant changes in WM connectivity in any of the voxels, including the body of the CC No significant correlation between FA and Y-BOCS score

Impaired WM integrity in the body of the CC in OCD is driven by a myelin abnormality

WM abnormalities in the anterior body of CC might cause dysfunction of DLPFC and set-shifting deficits. Lower FA in the CC body might be a trait marker


OCD mean ± sd

NA

NA

↑


NA

NA

None

Splenium FADOC < FAHS Garibotto 2010

NA

NA

↓


NA

NA

↑


NA

NA

None

Fontenelle 2011

NA

NA

None

None of the clinical measures, such as the Y-BOCS, HDRS, BDI, and BAI, was significantly correlated to FA. No normalization after citalopram

The higher FA in OCD patients represents a rapid or increased information transmission caused by filtering failures at the level of the thalamus

Increased AD


Increased AD indirectly implicates increased connectivity probably suggestive of excessively active brain circuits in OCD

Significant differences in the distribution of PDD in the CC genu and body

Negative correlation between FA, Iowa Gambling Task scores (disadvantageous choices) and the time required for the Trail Making Test-A task

FA reduction and altered PDD could be primarily interpreted as local changes in fiber directionality. This alteration might result from changes in neural migration as postulated for developmental diseases

Positive correlation between FA and CY-BOCS total score, no significant effect of medication on FA changes

OCD may be characterized by premature myelination. Increased FA may also reflect increased crossing fibers or less organized connectivity in participants with OCD

Increased AD


Increased MD, no differences in FA

Positive correlation between FA and MD and Y-BOCS score

Increased AD indirectly implicates increased connectivity probably suggestive of excessively active brain circuits in OCD Suggested changes in OCD in the WM connectivity spanning L and R PF areas and STC

AD, axial diffusivity; BAI, Beck Anxiety Inventory; BDI, Beck Depression Inventory; CC, corpus callosum; CY-BOCS, Children Yale Brown Obsessive Compulsive Scale; DLPFC, dorsolateral prefrontal cortex; DTI, diffusion tensor imaging; FA, fractional anisotropy; HARS, Hamilton Anxiety Rating Scale; HDRS, Hamilton Depression Rating Scale; HDRS, Hamilton Depression Rating Scale; HS, healthy subjects; MD, mean diffusivity; OC, obsessive compulsive; OCD, obsessive compulsive disorder; PDD, principal diffusion direction; PF, prefrontal; RD, radial diffusivity; sd, standard deviation; STC, superior temporal cortex; WM, white matter; Y-BOCS, Yale Brown Obsessive Compulsive Scale.


FADOC > FAHS Yoo 2007

2871

2872


neuropsychological performance (Garibotto et al., 2010) suggesting that symptom expression and the specific profile of cognitive deficits in OCD patients might be subtended by intra-hemispheric disconnection. Finally, the most reliable finding in the reviewed papers is the substantiation of microstructural abnormalities in the CC. Indeed, 10 out of the 17 considered investigations report changes in axonal integrity in various inter-hemispheric tracts, with strong evidences of decreased connectivity in the rostrum, containing fibers from the OCF, and hyperconnectivity in the genu, projecting fibers into the PF cortex and involved in inhibition of cortical activity. On the other hand, results regarding the body and the splenium of CC are less consistent and probably affected by patients’ medication status. 3.2.2. Clinical correlates of microstructural alterations in regions outside the OCD classic circuit As for the additional WM regions putatively involved in OCD, whereas one study (Menzies et al., 2008b) suggested that WM abnormalities in medial frontal and parietal areas are endophenotypes, representing genetic risk for OCD, another one (Szeszko et al., 2005) showed that symptom scores significantly correlate with microstructural alterations in the parietal WM, thus entailing that this region plays a role in mediating obsessions and compulsive behavior. On the other hand, two studies (Fan et al., 2012; Yoo et al., 2007) found no correlation between symptom severity and diffusivity indices in temporo-parieto-occipital regions demonstrating that abnormalities in WM microstructure of posterior brain regions may contribute to OCD pathogenesis. Similarly, while one study (Jayarajan et al., 2012) revealed no correlation between symptom severity scores and microstructural abnormalities in intra-hemispheric bundles, three others (Fontenelle et al., 2011; Garibotto et al., 2010; Zarei et al., 2011) demonstrated that aberrant connectivity and reduced coherence in multiple WM tracts are associated with disease expression and selective cognitive dysfunctions (Garibotto et al., 2010). The reported evidence suggests that microstructural abnormalities in posterior brain regions could contribute to the cognitive deficits evident in OCD, rather than being responsible for symptom expression. Concurrently, the substantiation of altered anatomical connectivity between frontal and parieto-occipital associative cortices may provide a pathological basis for the hypothesis that OCD cognitive deficits might be underpinned by posterior cortical

underactivation and anterior overactivation (Ciesielski et al., 2005). Indeed, several neuroimaging (Nordahl et al., 1989; van der Wee et al., 2003) and electrophysiological studies (Ciesielski et al., 1981, 2005) in OCD patients demonstrated increased cortical activation in the prefrontal–striatal, thalamic, and anterior cingulate networks and decreased activation over the occipital and parietal regions during different cognitive tasks. While the decreased activation of posterior brain regions might be linked to processing difficulties during complex visuo-spatial tasks, the pattern of increased anterior brain activation could be consequent to a compensatory mechanism of effortful inhibition and be acquainted with one of the most debilitating symptoms in cognitive functioning of patients with OCD, the persistent doubt (Ciesielski et al., 2005). Finally, significant positive correlations between DTI measures of WM integrity and symptom scores were found in the CC rostrum (Oh et al., 2011; Saito et al., 2008) and not in other subdivisions of the CC (Jayarajan et al., 2012; Li et al., 2011; Nakamae et al., 2011; Yoo et al., 2007; Zarei et al., 2011), thus implying that microstructural abnormalities in ventral callosal projections connecting left and right OFCs, determine clinical status, being particularly implicated in obsession symptoms (Oh et al., 2011). Tissue changes in the CC genu, body and splenium may, on the other hand, be trait markers possibly associated with the selective cognitive dysfunctions observed in OCD (Garibotto et al., 2010). 4. Quantitative evaluation of FA alterations in OCD patients 4.1. Coordinate based ALE meta-analysis of whole brain voxel based studies Between the 8 whole brain voxel-based studies, 7 reported data suitable for meta-analysis, while for 1 paper (Ha et al., 2009) coordinates were retrieved by asking to the corresponding author. The identified articles included 35 foci, among which 12 corresponded to increased FA (4 studies, 81 subjects) (Li et al., 2011; Menzies et al., 2008b; Nakamae et al., 2008; Yoo et al., 2007) and 23 to decreased FA (5 studies, 112 subjects) (Fan et al., 2012; Garibotto et al., 2010; Ha et al., 2009; Menzies et al., 2008b; Szeszko et al., 2005) in OCD patients compared with HS. The results from ALE meta-analysis are reported in Table 7A. FA was increased in the body of CC and anterior corona radiata, although only one focus contributed to each cluster (Li et al.,

Table 7 Results from meta-analyses. (A) FA changes in OCD patients as revealed by ALE meta-analysis on whole brain voxel-based studies. (B) Overall Standardized Mean Differences (SMD) between OCD and healthy subjects in the ROI based meta-analysis. (A)

Talairach coordinates

Brain region

x

Increased FA Ant corona radiata CC Body

22 3

22 13

−8 −24 30

11 −50 −48

Decreased FA Cingulum CC Splenium Superior Longitudinal Fasciculus

y

Volume mm3

ALE value

41 19

224 256

0.007 0.008

34 20 30

496 448 400

0.01 0.01 0.009

z

(B) Brain region

N. of investigations

N. of OCD

N. of HS

Test for overall effect SMD

CC genu CC splenium

4 3

32 52

36 52

0.838 0.049

L ALIC

3

42

41

−0.727

ALE, Activation likelihood estimation; CI, confidence interval; HS, healthy subjects; L, left.

95% CI −0.519 to 2.194 −0.795 to 0.894 −1.389 to −0.065

z

p 1.21 0.114

0.226 0.909

−2.152

0.031


2011; Menzies et al., 2008b, respectively). FA was decreased in OCD patients in the left cingulum, splenium of corpus callosum and right superior longitudinal fasciculus. Two foci contributed to the first cluster (Garibotto et al., 2010; Szeszko et al., 2005), two to the second (Fan et al., 2012; Garibotto et al., 2010) and two to the third (Garibotto et al., 2010; Szeszko et al., 2005). Though there are no community-accepted criteria for metaanalysis results, if 6 or more foci contribute to a cluster, it is considered very robust, and if 3–5 foci contribute to a cluster, it is acceptable (see the forum of GingerALE, http://www/brainmap.org/ forum/) (Li et al., 2010). Given the insignificant number of contributors foci to the reported clusters (either of increased or decreased FA) and the small size of our study, the described results should be taken with caution. 4.2. ROI based meta-analysis Among the 6 ROI based studies included in our systematic review, 7 investigations, from 6 articles, contributed to at least three assessments for a given cerebral region (Cannistraro et al., 2007; Fontenelle et al., 2011; Lochner et al., 2012; Menzies et al., 2008b; Oh et al., 2011; Saito et al., 2008). FA differences in the following 3 WM regions were assessed by at least three studies: genu of CC (Fontenelle et al., 2011; Oh et al., 2011; Saito et al., 2008), splenium of CC (Fontenelle et al., 2011; Menzies et al., 2008b; Saito et al., 2008) and left ALIC (Cannistraro et al., 2007; Fontenelle et al., 2011; Lochner et al., 2012). Overall analyses are reported in Table 7B. The only significant overall effect size was in the L ALIC (SDM = −0.727; CI = −1.389 to −0.065; p = 0.031) where the SDM indicates a decrease in FA, possibly suggestive of reduced coherence of fibers. However, the analysis included only a limited number of subjects, while the Q test was significant (p < 0.001) and the between study heterogeneity quite high (I2 = 89). On the other hand, the lack of significant difference between group means for the CC genu was unlikely due to heterogeneity between studies, as no significant level of heterogeneity was reached (Q test p = 0.10). High heterogeneity was conversely observed between studies investigating the splenium of the CC (Q test p = 0.014; I2 = 76), suggesting that the non-significant effect size was probability due to differences in sample characteristics or neuroimaging methods. 5. Conclusions In this paper systematically reviewing DTI data in OCD patients, we assert that discrete abnormalities of different WM tracts play an important role in the neurobiology of the disorder. We also emphasize that the reported WM alterations complement the broader gray matter abnormalities identified in OCD (Piras et al., 2013), suggesting that the disorder is associated with large-scale disruption in brain systems or networks, rather than being a consequence of disturbances in isolated brain regions (Menzies et al., 2008a). Consistently with predictions of current fronto-striatal models of the disorder, we first report evidence of microstructural alterations in the fronto-basal pathways targeting the OFC and ACC. Specifically, the tissue changes identified through the combination of multiple diffusion measures are essentially represented by a decrease in the neuronal density of the cingulate WM and CB, with no changes in the structural organization of the tissue, and by abnormal structural connectivity in the IC, likely due to increased directional coherence. Although the significance of diffusivity changes in regions with complex fiber-crossing such as the cingulate WM, is intricate, the reported microstructural alterations in this region and CB are supported by morphometric findings. Actually, significant WM volume

2873

decreases in the anterior (Riffkin et al., 2005; Togao et al., 2010) and posterior cigulum (Riffkin et al., 2005) are described in samples of medicated and unmedicated OCD patients. Such evidence is coupled with volume reduction in the cingulate cortex (Carmona et al., 2007; Gilbert et al., 2008; Kopˇrivová et al., 2009; Matsumoto et al., 2010; Pujol et al., 2004; Valente et al., 2005; Yoo et al., 2008), suggesting that these inter-related brain alterations may occur in parallel as distinct effects of a common upstream pathological process. The evidence of decreased levels of N-acetylaspartate (NAA), a marker of neurons integrity and metabolic status (Bertolino et al., 2001), in the ACC and frontal WM of adults with OCD (Ebert et al., 1997; Jang et al., 2006; Yücel et al., 2007), and the fact that such reduction was reversed after treatment with the SSRI citalopram (Jang et al., 2006), indicate revocable abnormalities in axonal or dendritic arbors and synapses, rather than decreases in the number of neurons (Barker et al., 2001). On the other hand, the reported evidence of hyperconnectivity in the IC supports the notion that hyperactivity in the orbitofrontalstriatal loop serves a key role in the pathogenesis of OCD. Such assumption is sustained by the observation that a lesion placed in the ALIC may produce significant improvement in severe and refractory patients (Greenberg et al., 2003; Oliver et al., 2003), while deep brain stimulation of the tract can determine a significant reduction in Y-BOCS scores (Greenberg et al., 2010; Koning et al., 2011) and increased perfusion in the circuitry (Rauch et al., 2006). The observation that a neurosurgical lesion in the ALIC is beneficial in the treatment of highly refractory OCD has been also considered evidence for a glutamatergic dysfunction in the disorder. Indeed, although there are various neurotransmitters that modulate the activity of the orbitofrontal–striatal loop, efferent axons from the OFC are glutamatergic neurons and the neurochemical effect of capsulotomies is probably to interrupt increased glutamatergic transmission between the OFC and the caudate nucleus (Blier et al., 2006). Moreover, the significant reduction in OFC volumes reported in several morphometric studies (e.g.: Togao et al., 2010) might be a secondary degeneration to the hyperactivity of the circuit (Togao et al., 2010) since excess glutamate has long been known to lead to neuronal death, a phenomenon known as excitotoxicity (Olney, 1969). The fact that hyperconnectivity of different brain regions was observed also in children and adolescent OCD patients (Jayarajan et al., 2012; Zarei et al., 2011) supports a developmental hypothesis of the disorder which may be characterized by premature myelination (Zarei et al., 2011). All the above considered, we suggest that microstructural changes in the fronto-basal pathways involving the OFC and the ACC may contribute to the causation of the illness, while decreased neuronal density in the cingulate WM and CB, coupled with metabolic or neuropil abnormalities in the ACC (Yücel et al., 2007), might explain the consistent impairment in volitional suppression of actions leading to compulsions. On the other hand, the hyperactive loop between the OFC and the striatum might mediate circular and repetitive thoughts (Modell et al., 1989; Rauch and Jenike, 1993). Such abnormalities seem to be specific to OCD, as within DSM-IV anxiety disorders, microstructural alterations in the affective loop subserved by the cingulum were reported only in panic disorder. In the latter, the observed hyperconnectivity in anterior and posterior cingulate regions was interpreted as consequent to increased attention to anxiety-provoking stimuli (Han et al., 2008), a symptom shared by OCD patients. However, although results of both meta-analyses should be taken with caution, our hypothesis is not strengthen by statistical assessment, as decreased FA, putatively linked to hypoconnectivity, was observed in both the cingulum and the L ALIC in each meta-analytical method. A second important point is that altered anatomical connectivity between lateral frontal and parietal regions and in intra-hemispheric bundles linking distinctive areas of the PF cortex

2874


to posterior parietal and occipital association cortices, is consistently reported in the reviewed studies. Parietal lobe dysfunction might be relevant in OCD pathophysiology since it might interact with the frontal subcortical circuitry of OCD through direct anatomical connections between associative parietal areas and some of the key regions implicated in the disorder, including the lateral OFC (Zald and Kim, 1996), the striatum (Yeterian and Pandya, 1993), and the mediodorsal thalamic nucleus (Giguere and Goldman-Rakic, 1988). Moreover, the fact that parietal WM abnormalities occur also in healthy first-degree relatives of OCD patients (Menzies et al., 2008a) suggests they may be endophenotypes associated with increased genetic risk for the illness. Indeed, a recent study (Stewart et al., 2007) focusing on the frequency of the gene oligodendrocyte lineage transcription factor 2 (OLIG2) in families of patients with OCD, showed a biased transmission of polymorphisms in probands with OCD. Since the OLIG2 is an important regulator in the development of cells that produce myelin, such finding would suggest a genetic origin of WM abnormalities in OCD. The increasingly recognized role played by posterior brain regions in OCD pathophysiology is further supported by the evidence of WM abnormalities in temporo-parietal-occipital regions and in long-range and cortico-cortical bundles connecting the frontal, parietal, temporal and occipital lobes. Interestingly, structural anatomical abnormalities in these tracts positively correlate with both symptom severity and neuropsychological performance (Garibotto et al., 2010) suggesting that symptom expression, and the specific profile of cognitive deficits in OCD patients might be subtended by intra-hemispheric disconnection. For example, the SLF branches extensively in the frontal, parietal, and temporal lobes in humans (Makris et al., 2005) and disturbances in connectivity in this bundle can result in impaired attention and spatial working memory. The IFOF, connects the ventrolateral PF cortex and medial OFC to posterior parietal and occipital associative cortices and atrophy in its frontal branch is strongly correlated with the presence of signs of fronto-subcortical dysfunction, such as dysexecutive problems, apathy and personality change. Finally, the ILF, connecting the anterior part of the temporal lobe to the occipital lobe, plays a role in emotional processing, probably through a wide network also involving the UF. Thus, the involvement of both the UF and ILF could serve as neurobiological correlates for the emotional processing deficits seen in neuropsychological studies of OCD patients. Moreover, microstructural alterations in the UF were reported in several DSM-IV anxiety disorders (Ayling et al., 2012), while scores on anxiety-related personality traits are related, in healthy subjects, to WM integrity of the anterior thalamic radiations, and the large association fibers (IFOF and right SLF), wiring frontal, occipital, parietal, and temporal lobes (Westlye et al., 2011). Such observation would suggest that alterations in WM microstructure might be a marker of the biological psychopathologic susceptibility to anxiety disorders, including OCD. A further reliable finding in the reviewed papers is the substantiation of microstructural abnormalities in the CC with strong evidences of decreased connectivity in the rostrum, containing fibers from the OCF, and hyperconnectivity in the genu, projecting fibers into the PF cortex and involved in inhibition of cortical activity. Reduced fibers coherency (i.e. decreased FA) in the whole CC significantly correlates with selective cognitive dysfunctions (Garibotto et al., 2010), thus corroborating the hypothesis that changes in inter-hemispheric connectivity might explain some of the neuropsychological disorders evident in OCD patients (Di Paola et al., 2012). Speculatively, we can hypothesize that decreased integrity in the rostrum might underlie a connectivity defect in the system involved in mapping reinforcement contingencies subserved by bilateral OFC, while the abnormal functional hyperconnectivity observed in the genu could subtend the difficulties in

Fig. 2. Reduced/increased connectivity of white matter tracts in obsessive compulsive disorder. White matter tracts in which reduced (left side) or increased (right side) connectivity has been described on the basis of cases/controls differences in different diffusivity (FA, MD, AD or RD) indices. ALIC, anterior limb of internal capsule; CB, cingulum bundle; CC, corpus callosum; IFOF, inferior fronto-occipital fasciculus; ILF, inferior longitudinal fasciculus; P-LIC, posterior limb of internal capsule; SLF, superior longitudinal fasciculus; UF, uncinate fasciculus.

filtering and selecting relevant information often observed in OCD patients. Moreover, the involvement of parts of the CC other than the rostrum in both adult and juvenile OCD, strongly implicates that additional areas such as lateral PF and parietal cortices, may contribute to the causation of the illness (Fig. 2). In summary, although generalization of findings is limited by the small samples and the varied methods employed in DTI studies, results of the present review support the notion that brain alterations responsible for OCD are represented at a system or network level (Menzies et al., 2008a), while abnormalities across several different regions, and altered anatomical connectivity may account for both symptom expression and impairments in cognition. 5.1. Future directions Although DTI represents, at present, the best-equipped current technique for work on the WM issue in OCD, conclusions from the existing body of DTI research in the disorder must be drawn with extreme caution. Compared with volumetric measurements, this method is potentially more sensitive to detecting subtle and early changes in the microstructure and organization of WM fiber tracts and provides neuroscientist a unique tool for characterizing and defining the extent of pathologic and microstructural alterations that occurs in OCD. Nevertheless, the underlying neurobiological structural alterations resulting in the measured changes in microstructure have not yet been elucidated, thus limiting its potential application in the clinical setting. Moreover, psychiatric neuroimaging is plagued by heterogeneous and inconsistent findings, mainly due to technical differences in MRI data acquisition and analysis, clinical heterogeneity of psychiatric diagnostic criteria, divergences in inclusion criteria and clinical characteristics of samples, differential exposure to medication, and so on. Since there is no benchmark DTI method and questions still remain regarding how measures obtained using one technique correspond to those obtained using other methods (Kubicki et al., 2007), the application of standard DTI acquisitions should improve the clinical significance of diffusion weighted imaging in OCD patients. Furthermore, while some etiological theories of OCD have emphasized the role of structural WM abnormalities (e.g. Menzies et al., 2008b), others have focussed on the function of aberrant synaptic plasticity in cortico-striatal circuits in the causation of the disorder (e.g.: Allendes et al., 2008). Such hypothesis make multimodal studies necessary, since the combination of different


biomedical tools (e.g., oxygenation with fMRI, neurochemistry with nuclear magnetic resonance spectroscopy (MRS), electrodynamics via EEG, genetic profile via microarray technology, and diffusion characteristics using DTI) may simultaneously capture the structural, diffusion, and biochemical properties of OCD key brain regions in-depth. In conclusion, future studies incorporating DTI and at least another complementary method will shed light on the role of WM abnormalities in the pathophysiology of OCD and will establish and refine neurocircuitry models of the disorder. 5.2. Limitations Our systematic review has several limitations. First, is grounded on the available published results, which often do not report null findings or discard as false positives or artifacts abnormalities not thought to be related to the disorder (Borgwardt et al., 2012). Moreover, our conclusions were mainly drawn on cross-sectional DTI studies in adult OCD patients in which, due to lack of a longitudinal perspective, the causal relationship between the observed neurobiological abnormality and diagnosis is often unclear. However, as microstructural WM anomalies are an inherent trait of OCD (Menzies et al., 2008b) contributed to by particular genes involved in the disorder (Stewart et al., 2007), their potential pathogenic role seems plausible. Second, results from the reviewed studies are necessarily “binarized” (significant vs non-significant or increased vs decreased) with a loss of information on the effect sizes (Radua and MataixCols, 2012). However, we tried to overcome some of the weaknesses that typically limit label-based reviews, as we attempted to “weight” the considered studies by sample size, while conclusions were mostly drawn limiting our consideration to studies reporting findings surviving multiple comparison correction. We also tried to deal with the question of opposite findings in the same region as we always considered the potential effect of confounding variables in case of inconsistent evidence. Moreover, we specifically addressed the interrelation between clinical variables and DTI measures of WM integrity, a potential confounding factor that was explicitly modeled in a limited number of studies, and tried to provide a comprehensive account of the relationship between illness variables and WM microstructural abnormalities in the disorder. Nevertheless, future multimodal studies integrating different biomedical tools, will increase the translational impact on clinical practice of diffusion weighted imaging and contribute to refine the current neurobiological model of OCD circuitry. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.neubiorev.2013.10.008. References Allendes, F., Lozano, A., Hutchison, W., 2008. Attenuation of long-term depression in human striatum after anterior capsulotomy. Stereotact. Funct. Neurosurg. 86, 224–230. American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders, 4th ed., text revision. American Psychiatric Association, Washington, DC. Aoki, Y., Inokuchi, R., Gunshin, M., Yahagi, N., Suwa, H., 2012. Diffusion tensor imaging studies of mild traumatic brain injury: a meta-analysis. J. Neurol. Neurosurg. Psychiatry 83, 870–876. Assaf, Y., 2008. Can we use diffusion MRI as a bio-marker of neurodegenerative processes? BioEssays 30, 1235–1245. Ayling, E., Aghajani, M., Fouche, J.-P., Wee, N., 2012. diffusion tensor imaging in anxiety disorders. Curr. Psychiatry Rep. 14, 197–202.

2875

Barker, V., Middleton, G., Davey, F., Davies, A., 2001. TNFalpha contributes to the death of NGF-dependent neurons during development. Nat. Neurosci. 4, 1194–1198. Basser, P.J., Mattiello, J., LeBihan, D., 1994. MR diffusion tensor spectroscopy and imaging. Biophysics 66, 259–267. Benkelfat, C., Nordahl, T.E., Semple, W.E., King, A., Murphy, D.L., Cohen, R.M., 1990. Local cerebral glucose metabolic rates in obsessive–compulsive disorder: patients treated with clomipramine. Arch. Gen. Psychiatry 47, 840–848. Bertolino, A., Callicott, J.H., Mattay, V.S., Weidenhammer, K.M., Rakow, R., Egan, M.F., Weinberger, D.R., 2001. The effect of treatment with antipsychotic drugs on brain N-acetylaspartate measures in patients with schizophrenia. Biol. Psychiatry 49, 39–46. Blier, P., Habib, R., Flament, M., 2006. Pharmacotherapies in the management of obsessive–compulsive disorder. Can. J. Psychiatry 51, 417–430. Bora, E., Harrison, B., Fornito, A., Cocchi, L., Pujol, J., Fontenelle, L., Velakoulis, D., Pantelis, C., Yücel, M., 2011. White matter microstructure in patients with obsessive–compulsive disorder. J. Psychiatry Neurosci. 36, 42–46. Borenstein, M., Hedges, L., Higgins, J., Rothstein, H., 2005. Comprehensive Metaanalysis Version 2. Biostat, Englewood, NJ. Borgwardt, S., Radua, J., Mechelli, A., Fusar-Poli, P., 2012. Why are psychiatric imaging methods clinically unreliable? Conclusions and practical guidelines for authors, editors and reviewers. Behav. Brain Sci. 8, 46. Cannistraro, P.A., Makris, N., Howard, J.D., Wedig, M.M., Hodge, S.M., Wilhelm, S., Kennedy, D.N., Rauch, S.L., 2007. A diffusion tensor imaging study of white matter in obsessive–compulsive disorder. Depress. Anxiety 24, 440–446. Carmona, S., Bassas, N., Rovira, M., Gispert, J.-D., Soliva, J.-C., Prado, M., Tomas, J., Bulbena, A., Vilarroya, O., 2007. Pediatric OCD structural brain deficits in conflict monitoring circuits: a voxel-based morphometry study. Neurosci. Lett. 421, 218–223. Ciesielski, K., Beech, H., Gordon, P., 1981. Some electrophysiological observations in obsessional states. Br. J. Psychiatry, 479–484. Ciesielski, K.T., Hämäläinen, M.S., Lesnik, P.G., Geller, D.A., Ahlfors, S.P., 2005. Increased MEG activation in OCD reflects a compensatory mechanism specific to the phase of a visual working memory task. Neuroimage 24, 1180–1191. Davatzikos, C., 2004. Why voxel-based morphometric analysis should be used with great caution when characterizing group differences. Neuroimage 23, 17–20. DeBoy, C.A., Zhang, J., Dike, S., Shats, I., Jones, M., Reich, D.S., Mori, S., Nguyen, T., Rothstein, B., Miller, R.H., Griffin, J.T., Kerr, D.A., Calabresi, P.A., 2007. High resolution diffusion tensor imaging of axonal damage in focal inflammatory and demyelinating lesions in rat spinal cord. Brain 130, 2199–2210. Di Paola, M., Di Iulio, F., Cherubini, A., Blundo, C., Casini, A., Sancesario, G., Passafiume, D., Caltagirone, C., Spalletta, G., 2010. When, where, and how the corpus callosum changes in MCI and AD: a multimodal MRI study. Neurology 74, 1136–1142. Di Paola, M., Luders, E., Rubino, I.A., Siracusano, A., Manfredi, G., Girardi, P., Martinotti, G., Thompson, P.M., Chou, Y.Y., Toga, A.W., Caltagirone, C., Spalletta, G., 2012. The structure of the corpus callosum in obsessive compulsive disorder. Eur. Psychiat. 28 (8), 499–506. Diwadkar, V.A., Keshavan, M.S., 2002. Newer techniques in magnetic resonance imaging and their potential for neuropsychiatric research. J. Psychosom. Res. 53, 677–685. Douzenis, A., Michalopoulou, P.G., Voumvourakis, C., Typaldou, M., Michopoulos, I., Lykouras, L., 2009. Obsessive–compulsive disorder associated with parietal white matter multiple sclerosis plaques. World J. Biol. Psychiatry 10, 956–960. Duran, F.L.d.S., Hoexter, M.Q., Valente Jr., A.A., Miguel, E.C., Busatto, G.F., 2009. Association between symptom severity and internal capsule volume in obsessive–compulsive disorder. Neurosci. Lett. 452, 68–71. Ebert, D., Speck, O., König, A., Berger, M., Hennig, J., Hohagen, F., 1997. 1 H-magnetic resonance spectroscopy in obsessive–compulsive disorder: evidence for neuronal loss in the cingulate gyrus and the right striatum. Psychiatry Res. 74, 173–176. Eickhoff, S.B., Laird, A.R., Grefkes, C., Wang, L.E., Zilles, K., Fox, P.T., 2009. Coordinatebased activation likelihood estimation meta-analysis of neuroimaging data: a random-effects approach based on empirical estimates of spatial uncertainty. Hum. Brain Mapp. 30, 2907–2926. Etkin, A., Wager, T., 2007. functional neuroimaging of anxiety: a meta-analysis of emotional processing in PTSD, social anxiety disorder, and specific phobia. Am. J. Psychiatry 164, 1476–1488. Fan, Q., Yan, X., Wang, J., Chen, Y., Wang, X., Li, C., Tan, L., You, C., Zhang, T., Zuo, S., Xu, D., Chen, K., Finlayson-Burden, J.M., Xiao, Z., 2012. Abnormalities of white matter microstructure in unmedicated obsessive–compulsive disorder and changes after medication. PLoS ONE 7, e35889. Fjell, A.M., Westlye, L.T., Greve, D.N., Fischl, B., Benner, T., van der Kouwe, A.J.W., Salat, D., Bjørnerud, A., Due-Tønnessen, P., Walhovd, K.B., 2008. The relationship between diffusion tensor imaging and volumetry as measures of white matter properties. Neuroimage 42, 1654–1668. Fontenelle, L.F., Bramati, I.E., Moll, J., Mendlowicz, M.V., de Oliveira-Souza, R., TovarMoll, F., 2011. White matter changes in OCD revealed by diffusion tensor imaging. CNS Spectr. 16, 101–109. Fontenelle, L.F., Harrison, B.J., Yucel, M., Pujol, J., Fujiwara, H., Pantelis, C., 2009. Is there evidence of brain white-matter abnormalities in obsessive–compulsive disorder? A narrative review. Top. Magn. Reson. Imaging 20, 291–298. Fornito, A., Bullmore, E.T., 2012. Connectomic intermediate phenotypes for psychiatric disorders. Front Psychiatry 3, 32. Frodl, T., Carballedo, A., Fagan, A., Lisiecka, D., Ferguson, Y., Meaney, J., 2012. Effects of early-life adversity on white matter diffusivity changes in patients at risk for major depression. J. Psychiatry Neurosci. 37, 37–45.

2876


Fullana, M.A., Mataix-Cols, D., Caspi, A., Harrington, H., Grisham, J.R., Moffitt, T.E., Poulton, R., 2009. Obsessions and compulsions in the community: prevalence, interference, help-seeking, developmental stability, and co-occurring psychiatric conditions. Am. J. Psychiatry 166, 329–336. Garibotto, V., Scifo, P., Gorini, A., Alonso, C.R., Brambati, S., Bellodi, L., Perani, D., 2010. Disorganization of anatomical connectivity in obsessive compulsive disorder: a multi-parameter diffusion tensor imaging study in a subpopulation of patients. Neurobiol. Dis. 37, 468–476. Giguere, M., Goldman-Rakic, P.S., 1988. Mediodorsal nucleus: areal, laminar, and tangential distribution of afferents and efferents in the frontal lobe of rhesus monkeys. J. Comp. Neurol. 277, 195–213. Gilbert, A.R., Mataix-Cols, D., Almeida, J.R.C., Lawrence, N., Nutche, J., Diwadkar, V., Keshavan, M.S., Phillips, M.L., 2008. Brain structure and symptom dimension relationships in obsessive–compulsive disorder: a voxel-based morphometry study. J. Affect. Disord. 109, 117–126. Graybiel, A.M., Rauch, S.L., 2000. Toward a neurobiology of obsessive–compulsive disorder. Neuron 28, 343–347. Greenberg, B., Price, L., Rauch, S., Friehs, G., Noren, G., Malone, D., Carpenter, L., Rezai, A., Rasmussen, S., 2003. Neurosurgery for intractable obsessive–compulsive disorder and depression: critical issues. Neurosurg. Clin. N. Am. 14, 199–212. Greenberg, B., Rauch, S., Haber, S., 2010. Invasive circuitry-based neurotherapeutics: stereotactic ablation and deep brain stimulation for OCD. Neuropsychopharmacology 35, 317–336. Ha, T.H., Kang, D.-H., Park, J.S., Jang, J.H., Jung, W.H., Choi, J.-S., Park, J.-Y., Jung, M.H., Choi, C.-H., Lee, J.-M., Ha, K., Kwon, J.S., 2009. White matter alterations in male patients with obsessive–compulsive disorder. Neuroreport 20, 735–739. Han, D.H., Renshaw, P.F., Dager, S.R., Chung, A., Hwang, J., Daniels, M.A., Lee, Y.S., Lyoo, I.K., 2008. Altered cingulate white matter connectivity in panic disorder patients. J. Psychiatry Res. 42, 399–407. Harrison, B., Soriano-Mas, C., Pujol, J., et al., 2009. Altered corticostriatal functional connectivity in obsessive–compulsive disorder. Arch. Gen. Psychiatry 66, 1189–1200. Huey, M.D.E., Zahn, M.D.P.D.R., Krueger, P.D.F., Moll, M.D.P.D.J., Kapogiannis, M.D.D., Wassermann, M.D.E., Grafman, P.D.J., 2008. A psychological and neuroanatomical model of obsessive–compulsive disorder. J. Neuropsychiatry Clin. Neurosci. 20, 390–408. Huyser, C., Veltman, D.J., de Haan, E., Boer, F., 2009. Paediatric obsessive–compulsive disorder, a neurodevelopmental disorder? Evidence from neuroimaging. Neurosci. Biobehav. Rev. 33, 818–830. Jang, J.H., Kwon, J.S., Jang, D.P., Moon, W.J., Lee, J.M., Ha, T.H., Chung, E.C., Kim, I.Y., Kim, S.I., 2006. A proton MRSI study of brain N-acetylaspartate level after 12 weeks of citalopram treatment in drug-naive patients with obsessive–compulsive disorder. Am. J. Psychiatry 163, 1202–1207. Jayarajan, R.N., Venkatasubramanian, G., Viswanath, B., Reddy, Y.C.J., Srinath, S., Vasudev, M.K., Chandrashekar, C.R., 2012. White matter abnormalities in children and adolescents with obsessive–compulsive disorder: a diffusion tensor imaging study. Depress. Anxiety 29, 780–788. Koch, K., Wagner, G., Schachtzabel, C., Christoph Schultz, C., Straube, T., Güllmar, D., Reichenbach, J., Peikert, G., Sauer, H., Schlösser, R.G., 2012. White matter structure and symptom dimensions in obsessive–compulsive disorder. J. Psychiatry Res. 46, 264–270. Koning, P., Figee, M., Munckhof, P., Schuurman, P.R., Denys, D., 2011. Current status of deep brain stimulation for obsessive–compulsive disorder: a clinical review of different targets. Curr. Psychiatry Rep. 13, 274–282. Kopˇrivová, J., Horáˇcek, J., Tintˇera, J., Praˇsko, J., Raszka, M., Ibrahim, I., Höschl, C., 2009. Medial frontal and dorsal cortical morphometric abnormalities are related to obsessive–compulsive disorder. Neurosci. Lett. 464, 62–66. Koran, L.M., Bromberg, D., Hornfeldt, C.S., Shepski, J.C., Wang, S., Hollander, E., 2010. Extended-release fluvoxamine and improvements in quality of life in patients with obsessive–compulsive disorder. Compr. Psychiatry 51, 373–379. Kubicki, M., McCarley, R., Westin, C., Park, H., Maier, S., Kikinis, R., Jolesz, F., Shenton, M., 2007. A review of diffusion tensor imaging studies in schizophrenia. J. Psychiatric Res. 41, 15–30. Kwon, J.S., Jang, J.H., Choi, J.-S., Kang, D.-H., 2009. Neuroimaging in obsessive–compulsive disorder. Expert Rev. Neurother. 9, 255–269. Laird, A.R., Eickhoff, S.B., Kurth, F., Fox, P.M., Uecker, A.M., Turner, J.A., Robinson, J.L., Lancaster, J.L., Fox, P.T., 2009. ALE meta-analysis workflows via the BrainMap database: progress towards a probabilistic functional brain atlas. Front. Neuroinform. 25 (1), 155–164. Laird, A.R., Fox, P.M., Price, C.J., Glahn, D.C., Uecker, A.M., Lancaster, J.L., Turkeltaub, P.E., Kochunov, P., Fox, P.T., 2005a. ALE meta-analysis: controlling the false discovery rate and performing statistical contrasts. Hum. Brain Mapp. 25, 155–164. Laird, A.R., Lancaster, J.L., Fox, P.T., 2005b. BrainMap: the social evolution of a human brain mapping database. Neuroinformatics 3, 65–78. Lancaster, J.L., Tordesillas-Gutiérrez, D., Martinez, M., Salinas, F., Evans, A., Zilles, K., Mazziotta, J.C., Fox, P.T., 2007. Bias between MNI and Talairach coordinates analyzed using the ICBM-152 brain template. Hum. Brain Mapp. 28, 1194–1205. Leckman, M.D.J., Bloch, M.D.M., 2008. A developmental and evolutionary perspective on obsessive–compulsive disorder: whence and whither compulsive hoarding? Am. J. Psychiatry 165, 1229–1233. Leonard, H.L., Ale, C.M., Freeman, J.B., Garcia, A.M., Ng, J.S., 2005. Obsessive–compulsive disorder. Child Adolesc. Psychiatr. Clin. N. Am. 14, 727–743. Li, F., Huang, X., Yang, Y., Li, B., Wu, Q., Zhang, T., Lui, S., Kemp, G.J., Gong, Q., 2011. microstructural brain abnormalities in patients with obsessive–compulsive disorder: diffusion-tensor MR imaging study at 3.0 T. Radiology 260, 216–223.

Li, H., Chan, R.C.K., McAlonan, G.M., Gong, Q.-y., 2010. Facial emotion processing in schizophrenia: a meta-analysis of functional neuroimaging data. Schizophrenia Bull. 36, 1029–1039. Lim, K.O., Helpern, J.A., 2002. Neuropsychiatric applications of DTI – a review. NMR Biomed. 15, 587–593. Lochner, C., Fouchè, J.-P., du Plessis, S., Spottiswoode, B., Seedat, S., Fineberg, N., Chamberlain, S.R., Stein, D.J., 2012. Evidence for fractional anisotropy and mean diffusivity white matter abnormalities in the internal capsule and cingulum in patients with obsessive–compulsive disorder. J. Psychiatry Neurosci. 37, 193–199. Maia, T.V., Cooney, R.E., Peterson, B.S., 2008. The neural bases of obsessive–compulsive disorder in children and adults. Dev. Psychopathol. 20, 1251–1283. Makris, N., Kennedy, D.N., McInerney, S., Sorensen, A.G., Wang, R., Caviness, V.S., Pandya, D.N., 2005. Segmentation of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cereb. Cortex 15, 854–869. Matsumoto, R., Ito, H., Takahashi, H., Ando, T., Fujimura, Y., Nakayama, K., Okubo, Y., Obata, T., Fukui, K., Suhara, T., 2010. Reduced gray matter volume of dorsal cingulate cortex in patients with obsessive–compulsive disorder: a voxel-based morphometric study. Psychiat. Clin. Neurosci. 64, 541–547. Menzies, L., Chamberlain, S.R., Laird, A.R., Thelen, S.M., Sahakian, B.J., Bullmore, E.T., 2008a. Integrating evidence from neuroimaging and neuropsychological studies of obsessive–compulsive disorder: the orbitofronto-striatal model revisited. Neurosci. Biobehav. Rev. 32, 525–549. Menzies, L., Williams, G., Chamberlain, S., Ooi, C., Fineberg, N., Suckling, J., Sahakian, B., Robbins, T., Bullmore, E., 2008b. White matter abnormalities in patients with obsessive–compulsive disorder and their first-degree relatives. Am. J. Psychiatry 165, 1308–1315. Milad, M., Rauch, S., 2012. Obsessive–compulsive disorder: beyond segregated cortico-striatal pathways. Trends Cogn. Sci. 16, 43–51. Modell, J.G., Mountz, J.M., Curtis, G.C., Greden, J.F., 1989. Neurophysiologic dysfunction in basal ganglia/limbic striatal and thalamocortical circuits as a pathogenetic mechanism of obsessive–compulsive disorder. J. Neuropsychiatry Clin. Neurosci. 1, 27–36. Mori, S., Oishi, K., Jiang, H., Jiang, L., Li, X., Akhter, K., Hua, K., Faria, A.V., Mahmood, A., Woods, R., Toga, A.W., Pike, G.B., Neto, P.R., Evans, A., Zhang, J., Huang, H., Miller, M.I., van Zijl, P., Mazziotta, J., 2008. Stereotaxic white matter atlas based on diffusion tensor imaging in an ICBM template. Neuroimage 40, 570–582. Nakamae, T., Narumoto, J., Sakai, Y., Nishida, S., Yamada, K., Nishimura, T., Fukui, K., 2011. Diffusion tensor imaging and tract-based spatial statistics in obsessive–compulsive disorder. J. Psychiatry Res. 45, 687–690. Nakamae, T., Narumoto, J., Shibata, K., Matsumoto, R., Kitabayashi, Y., Yoshida, T., Yamada, K., Nishimura, T., Fukui, K., 2008. Alteration of fractional anisotropy and apparent diffusion coefficient in obsessive–compulsive disorder: a diffusion tensor imaging study. Prog. Neuropsychopharmacol. Biol. Psychiatry 32, 1221–1226. Nobuhara, K., Okugawa, G., Sugimoto, T., Minami, T., Tamagaki, C., Takase, K., Saito, Y., Sawada, S., Kinoshita, T., 2006. Frontal white matter anisotropy and symptom severity of late-life depression: a magnetic resonance diffusion tensor imaging study. J. Neurol. Neurosurg. Psychiatry 77, 120–122. Nordahl, T., Benkelfat, C., Semple, W., Gross, M., King, A., Cohen, R., 1989. Cerebral glucose metabolic rates in obsessive compulsive disorder. Neuropsychopharmacology 2, 23–28. Nucifora, P., 2010. Evolving role of modern structural and functional MR imaging techniques for assessing neuropsychiatric disorders. PET Clin. 5, 169–183. Oh, J.S., Jang, J.H., Jung, W.H., Kang, D.-H., Choi, J.-S., Choi, C.-H., Kubicki, M., Shenton, M.E., Kwon, J.S., 2011. Reduced fronto-callosal fiber integrity in unmedicated OCD patients: a diffusion tractography study. Hum. Brain Mapp. 33, 2441–2452. Okasha, A., Rafaat, M., Mahallawy, N., Nahas, G.E., Dawla, A.S.E., Sayed, M., Kholi, S.E., 2000. Cognitive dysfunction in obsessive–compulsive disorder. Acta Psychiatry Scand. 101, 281–285. Oliver, B., Gascón, J., Aparicio, A., Ayats, E., Rodriguez, R., Maestro de León, J., García-Bach, M., Soler, P., 2003. Bilateral anterior capsulotomy for refractory obsessive–compulsive disorders. Stereotact. Funct. Neurosurg. 81, 90–95. Olney, J.W., 1969. Brain lesions, obesity, and other disturbances in mice treated with monosodium glutamate. Science (New York, NY) 164, 719–721. Paul, L., 2011. Developmental malformation of the corpus callosum: a review of typical callosal development and examples of developmental disorders with callosal involvement. J. Neurodev. Disord. 3, 3–27. Pena-Garijo, J., Ruipérez-Rodríguez, M.A., Barros-Loscertales, A., 2010. The neurobiology of obsessive–compulsive disorder: new findings from functional magnetic resonance imaging (II). Rev. Neurol. 50, 0541–0550. Pierpaoli, C., Jezzard, P., Basser, P.J., Barnett, A., Di Chiro, G., 1996. Diffusion tensor MR imaging of the human brain. Radiology 201, 637–648. Piras, F., Piras, F., Chiapponi, C., Girardi, P., Caltagirone, C., Spalletta, G., 2013. Widespread structural brain changes in OCD: a systematic review of voxel-based morphometry studies. Cortex, pii: S0010-9452(13)00046-4, Epub ahead of print. Pujol, J., Soriano-Mas, C., Alonso, P., et al., 2004. Mapping structural brain alterations in obsessive–compulsive disorder. Arch. Gen. Psychiatry 61, 720–730. Radua, J., Mataix-Cols, D., 2012. Meta-analytic methods for neuroimaging data explained. Biol. Mood Anxiety Disord. 2, 6. Rauch, S.L., Dougherty, D.D., Malone, D., Rezai, A., Friehs, G., Fischman, A.J., Alpert, N.M., Haber, S.N., Stypulkowski, P.H., Rise, M.T., Rasmussen, S.A., Greenberg, B.D., 2006. A functional neuroimaging investigation of deep brain stimulation in patients with obsessive–compulsive disorder. J. Neurosurg. 104, 558–565.

F. Piras et al. / Neuroscience and Biobehavioral Reviews 37 (2013) 2856–2877 Rauch, S.L., Jenike, M.A., 1993. Neurobiological models of obsessive–compulsive disorder. Psychosomatics 34, 20–32. Riffkin, J., Yücel, M., Maruff, P., Wood, S.J., Soulsby, B., Olver, J., Kyrios, M., Velakoulis, D., Pantelis, C., 2005. A manual and automated MRI study of anterior cingulate and orbito-frontal cortices, and caudate nucleus in obsessive–compulsive disorder: comparison with healthy controls and patients with schizophrenia. Psychiatry Res. 138, 99–113. Saito, Y., Nobuhara, K., Okugawa, G., Takase, K., Sugimoto, T., Horiuchi, M., Ueno, C., Maehara, M., Omura, N., Kurokawa, H., Ikeda, K., Tanigawa, N., Sawada, S., Kinoshita, T., 2008. Corpus callosum in patients with obsessive–compulsive disorder: diffusion-tensor imaging study 1. Radiology 246, 536–542. Saxena, S., Maidment, K.M., Vapnik, T., Golden, G., Rishwain, T., Rosen, R.M., Tarlow, G., Bystritsky, A., 2002. Obsessive–compulsive hoarding: symptom severity and response to multimodal treatment. J. Clin. Psychiatry 63, 21–27. Saxena, S., Rauch, S.L., 2000. Functional neuroimaging and the neuroanatomy of obsessive–compulsive disorder. Psychiatr. Clin. North Am. 23, 563–586. Schulte, T., Sullivan, E.V., Müller-Oehring, E.M., Adalsteinsson, E., Pfefferbaum, A., 2005. Corpus callosal microstructural integrity influences interhemispheric processing: a diffusion tensor imaging study. Cereb. Cortex 15, 1384–1392. Sijens, P.E., Mostert, J.P., Irwan, R., Potze, J.H., Oudkerk, M., De Keyser, J., 2008. Impact of fluoxetine on the human brain in multiple sclerosis as quantified by proton magnetic resonance spectroscopy and diffusion tensor imaging. Psychiat. Res. Neuroim. 164, 274–282. Smith, S.M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T.E., Mackay, C.E., Watkins, K.E., Ciccarelli, O., Cader, M.Z., Matthews, P.M., Behrens, T.E.J., 2006. Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data. Neuroimage 31, 1487–1505. Song, S.-K., Sun, S.-W., Ju, W.-K., Lin, S.-J., Cross, A.H., Neufeld, A.H., 2003. Diffusion tensor imaging detects and differentiates axon and myelin degeneration in mouse optic nerve after retinal ischemia. Neuroimage 20, 1714–1722. Stern, E.R., Fitzgerald, K.D., Welsh, R.C., Abelson, J.L., Taylor, S.F., 2012. Resting-state functional connectivity between fronto-parietal and default mode networks in obsessive–compulsive disorder. PLoS ONE 7, e36356. Stewart, S., Platko, J., Fagerness, J., Birns, J., Jenike, E., Smoller, J., Perlis, R., Leboyer, M., Delorme, R., Chabane, N., Rauch, S., Jenike, M., Pauls, D., 2007. A genetic familybased association study of olig2 in obsessive–compulsive disorder. Arch. Gen. Psychiatry 64, 209–214. Swedo, S., Leonard, H., Rapoport, J., 1992. Childhood-onset obsessive compulsive disorder. Psychiatr. Clin. North Am. 15, 767–775. Sykovà, E., 2004. Extrasynaptic volume transmission and diffusion parameters of the extracellular space. Neuroscience 129, 861–876. Szeszko, P., Ardekani, B., Ashtari, M., et al., 2005. White matter abnormalities in obsessive–compulsive disorder: a diffusion tensor imaging study. Arch. Gen. Psychiatry 62, 782–790. Togao, O., Yoshiura, T., Nakao, T., Nabeyama, M., Sanematsu, H., Nakagawa, A., Noguchi, T., Hiwatashi, A., Yamashita, K., Nagao, E., Kanba, S., Honda, H., 2010. Regional gray and white matter volume abnormalities in obsessive–compulsive disorder: a voxel-based morphometry study. Psychiatry Res. 184, 29–37.

2877

Valente, A.A., Miguel, E.C., Castro, C.C., Amaro Jr., E., Duran, F.L.S., Buchpiguel, C.A., Chitnis, X., McGuire, P.K., Busatto, G.F., 2005. Regional gray matter abnormalities in obsessive–compulsive disorder: a voxel-based morphometry study. Biol. Psychiat. 58, 479–487. van der Wee, N.J.A., Ramsey, N.F., Jansma, J.M., Denys, D.A., van Megen, H.J.G.M., Westenberg, H.M.G., Kahn, R.S., 2003. Spatial working memory deficits in obsessive compulsive disorder are associated with excessive engagement of the medial frontal cortex. Neuroimage 20, 2271–2280. Vloet, T.D., Neufang, S., Herpertz-Dahlmann, B., Konrad, K., 2006. Bildgebungsbefunde bei Kindern und Jugendlichen mit ADHS, Tic-Störungen und Zwangserkrankungen. Z. Kinder Jugendpsychiatr. Psychother. 34, 343–355. Westlye, L.T., Bjørnebekk, A., Grydeland, H., Fjell, A.M., Walhovd, K.B., 2011. Linking an anxiety-related personality trait to brain white matter microstructure: diffusion tensor imaging and harm avoidance. Arch. Gen. Psychiatry 68, 369–377. White, T., Nelson, M., Lim, K., 2008. Diffusion tensor imaging in psychiatric disorders. Top. Magn. Reson. Imaging 19, 97–109. Yeterian, E., Pandya, D., 1993. Striatal connections of the parietal association cortices in rhesus monkeys. J. Comp. Neurol. 332 (2), 175–197. Yoo, S.Y., Jang, J.H., Shin, Y.W., Kim, D.J., Park, H.J., Moon, W.J., Chung, E.C., Lee, J.M., Kim, I.Y., Kim, S.I., Kwon, J.S., 2007. White matter abnormalities in drug-naïve patients with obsessive–compulsive disorder: a diffusion Tensor Study before and after citalopram treatment. Acta Psychiatr. Scand. 116, 211–219. Yoo, S.Y., Roh, M.-S., Choi, J.-S., Kang, D.-H., Ha, T.H., Lee, J.-M., Kim, I.Y., Kim, S.I., Kwon, J.S., 2008. Voxel-based morphometry study of gray matter abnormalities in obsessive–compulsive disorder. J. Korean Med. Sci. 23, 24–30. Yücel, M., Harrison, B.J., Wood, S.J., Fornito, A., Wellard, R., Pujol, J., Clarke, K., Phillips, M., Kyrios, M., Velakoulis, D., Pantelis, C., 2007. Functional and biochemical alterations of the medial frontal cortex in obsessive–compulsive disorder. Arch. Gen. Psychiatry 64, 946–955. Zai, G., Bezchlibnyk, Y.B., Richter, M.A., Arnold, P., Burroughs, E., Barr, C.L., Kennedy, J.L., 2004. Myelin oligodendrocyte glycoprotein (MOG) gene is associated with obsessive–compulsive disorder. Am. J. Med. Genet. B: Neuropsychiatr. Genet. 129B, 64–68. Zald, D.H., Kim, S.W., 1996. Anatomy and function of the orbital frontal cortex, II. Function and relevance to obsessive–compulsive disorder. Zarei, M., Mataix-Cols, D., Heyman, I., Hough, M., Doherty, J., Burge, L., Winmill, L., Nijhawan, S., Matthews, P.M., James, A., 2011. Changes in gray matter volume and white matter microstructure in adolescents with obsessive–compulsive disorder. Biol. Psychiat. 70, 1083–1090. Zhang, J., Jones, M., DeBoy, C., Reich, D., Farrell, J., Hoffman, P., Griffin, J., Sheikh, K., Miller, M., Mori, S., Calabresi, P., 2009. Diffusion tensor magnetic resonance imaging of Wallerian degeneration in rat spinal cord after dorsal root axotomy. J. Neurosci. 29, 3160–3171. Zhang, T., Wang, J., Yang, Y., Wu, Q., Li, B., Chen, L., Yue, Q., Tang, H., Yan, C., Lui, S., Huang, X., Chan, R., Zang, Y., He, Y., Gong, Q., 2011. Abnormal small-world architecture of top-down control networks in obsessive–compulsive disorder. J. Psychiatry Neurosci. 36, 23–31.

Diffusion tensor imaging (DTI) studies in patients with obsessive-compulsive disorder (OCD): a review.

Obsessive-compulsive disorder is a heterogeneous disorder: evidence from diffusion tensor imaging and magnetization transfer imaging.

Microstructural brain abnormalities and symptom dimensions in child and adolescent patients with obsessive-compulsive disorder: a diffusion tensor imaging study.

Neurobiology of obsessive compulsive disorder: a review.

Treatment of comorbid bipolar disorder and obsessive-compulsive disorder: a systematic review.

Trauma-related obsessive-compulsive disorder: a review.

Brain MRI in obsessive-compulsive disorder.

Diffusion tensor imaging studies in late-life depression: systematic review and meta-analysis.

Obsessive-compulsive disorder onset after removal of a brain tumor.

Suicidality in obsessive compulsive disorder (OCD): a systematic review and meta-analysis.

Neuropsychological investigations in obsessive-compulsive disorder: A systematic review of methodological challenges.

Are 5-HT3 antagonists effective in obsessive-compulsive disorder? A systematic review of literature.

Atypical antipsychotic augmentation in SSRI treatment refractory obsessive-compulsive disorder: a systematic review and meta-analysis.

The relationship between obsessive-compulsive personality disorder traits, obsessive-compulsive disorder and excessive exercise in patients with anorexia nervosa: a systematic review.

Obsessive-compulsive disorder.

Obsessive-compulsive disorder.

Obsessive-compulsive disorder.

Brain-derived neurotrophic factor: a biomarker for obsessive-compulsive disorder?

Obsessive-Compulsive Personality Traits in Youth with Obsessive-Compulsive Disorder.

Obsessive-compulsive disorder.

Obsessive-compulsive disorder.

Obsessive-compulsive disorder.

The first 1000 days of the autistic brain: a systematic review of diffusion imaging studies.

Managing obsessive compulsive disorder.