Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF

Journal of Psychiatric Research xxx (2014) 1e7

Contents lists available at ScienceDirect

Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/psychires

Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF Xueqin Song a, *, 1, Meina Quan b, 1, Luxian Lv c, d, Xue Li a, Lijuan Pang a, David Kennedy b, Steven Hodge b, Amy Harrington b, Douglas Ziedonis b, Xiaoduo Fan b, ** a

The First Affiliated Hospital, Zhengzhou University, Zhengzhou, China UMass Memorial Medical Center, University of Massachusetts Medical School, Worcester, MA, USA Henan Province Biological Psychiatry Key Laboratory, Xinxiang Medical University, Xinxiang, China d Henan Province Mental Hospital, The Second Affiliated Hospital, Xinxiang Medical University, Xinxiang, China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 July 2014 Received in revised form 16 August 2014 Accepted 11 September 2014

This study was to examine cortical thickness in drug naïve, first episode schizophrenia patients, and to explore its relationship with serum levels of brain-derived neurotrophic factor (BDNF). Forty-five drug naive schizophrenia patients and 28 healthy controls were enrolled in the study. Freesurfer was used to parcellate cortical regions, and vertex-wise group analysis was used for whole brain cortical thickness. The clusters for the brain regions that demonstrated group differences were extracted, and the mean values of thickness were calculated. Serum levels of BDNF were measured using enzyme-linked immunosorbent assay (ELISA). After controlling for age and gender, significantly thinner cortical thickness was found in left insula and superior temporal gyrus in the patient group compared with the healthy control group (HC group) (p's < 0.001). Lower serum levels of BDNF were also found in the patient group compared with the HC group (p ¼ 0.001). Correlation analysis showed a significant positive relationship between thickness of left insula and serum levels of BDNF within the HC group (r ¼ 0.396, p ¼ 0.037) but there was no such relationship within the patient group (r ¼ 0.035, p ¼ 0.819). Cortical thinning is present in drug naïve, first episode schizophrenia patients, indicating neurodevelopmental abnormalities at the onset of schizophrenia. Left insula might be an imaging biomarker in detecting the impaired protective role of neurotrophic factor for the brain development in schizophrenia.

Keywords: Cortical thickness First episode schizophrenia BDNF

© 2014 Published by Elsevier Ltd.

1. Introduction Schizophrenia is a disorder characterized by mental dysfunction across a broad range of symptom domains, with a worldwide lifetime prevalence of 0.5e1%. Structural abnormalities have been found in various brain regions in patients with schizophrenia (Shenton et al., 2001). Brain volume changes were measured using either manual tracing (Levitt et al., 2010) or voxel-based morphometric approaches (Haijma et al., 2013) in previous studies. The

* Corresponding author. The First Affiliated Hospital, Zhengzhou University, Zhengzhou, 450052, China. Tel.: þ86 371 6796 6822. ** Corresponding author. UMass Memorial Medical Center, University of Massachusetts Medical School, One Biotech, 365 Plantation Street, Worcester, MA, 01605, USA. Tel.: þ1 508 856 3881. E-mail addresses: [email protected] (X. Song), [email protected] (X. Fan). 1 Xueqin Song, M.D., PH.D. and Meina Quan, M.D., PH.D. share co-first authorship.

regions with decreased volumes include frontal and temporal lobe, cingulate cortex, insular, the thalamus and hippocampus/amygdala (Fornito et al., 2009; Shepherd et al., 2012). With the development of neuroimaging techniques, recent studies examining cortical thickness adds complementary information to the volume data. In chronic schizophrenia, reduced cortical thickness has been found primarily in fronto-temporal regions (Voineskos et al., 2013; Zugman et al., 2013). Increasing attention has been focused on identifying brain structural and functional changes in the early stage of illness (Qiu et al., 2013). In first episode schizophrenia, studies have found reduced cortical thickness in areas such as frontal and temporal regions, as well as medial parietal, lateral occipital, and anterior cingulate cortex (Crespo-Facorro et al., 2011; Qiu et al., 2013; Sprooten et al., 2013). Brain-derived neurotrophic factor (BDNF), a trophic protein, plays an important role in neurodevelopment. It modulates activity-dependent synaptic plasticity among mature neurons, particularly in the neocortex and hippocampus (Gorski et al., 2003;

http://dx.doi.org/10.1016/j.jpsychires.2014.09.009 0022-3956/© 2014 Published by Elsevier Ltd.

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009

2

X. Song et al. / Journal of Psychiatric Research xxx (2014) 1e7

Table 1 Demographic and clinical characteristics of the study sample. Characteristics

HC groupa (N ¼ 28) (Mean ± SD)

SZ groupb (N ¼ 45) (Mean ± SD)

Age (years) Education (years) Gender (Male/Female) Smoking status (Yes/No) ICVc (cm3) Age of onset (years) Duration untreated psychosis (months) PANSSd-positive PANSS-negative PANSS-general PANSS-total

22.89 ± 5.17 13.61 ± 3.60 17/11 6/22 1633.00 ± 148.32 N/A N/A N/A N/A N/A N/A

22.09 ± 4.95 12.53 ± 2.18 27/18 9/36 1651.83 ± 159.60 21.64 ± 5.17 7.92 ± 7.57 21.29 ± 2.99 18.09 ± 2.28 38.89 ± 3.15 78.22 ± 4.63

a b c d

t

P 0.663 1.423

0.503

0.510 0.163 1.000 1.000 0.616

HC: healthy control group. SZ: schizophrenia patient group. ICV: intracranial volume. PANSS: the Positive and Negative Syndrome Scale, including positive symptoms, negative symptoms, general psychopathology subscales, and total scores.

Lu, 2003). It influences neural survival, differentiation, synaptogenesis, and maintenance (Webster et al., 2006; Xu et al., 2000). Alterations of serum levels of BDNF have been shown in a number of studies in patients with schizophrenia (Green et al., 2011; Jindal et al., 2010). In addition, a number of studies have linked BDNF gene polymorphism with neocortex (Ho et al., 2007) and hippocampus (Eisenberg et al., 2013) in schizophrenia, which are the two major areas involved in multiple cognitive domains that are impaired in schizophrenia (Elvevag and Goldberg, 2000). In addition, postmortem studies have shown reductions of BDNF mRNA expression and BDNF protein in cortical regions and hippocampus in schizophrenia (Hashimoto et al., 2005; Weickert et al., 2005). Direct measurement of BDNF levels in live human brain is challenging. BDNF in the peripheral blood is likely derived from central nervous system, and thus provides a less invasive measure for the BDNF levels in the brain (Huang and Reichardt, 2001). Previous animal studies have reported positive correlations between BDNF levels in peripheral blood (Pan et al., 1998, Sartorius et al., 2009) and those in the brain. A few previous studies have found a positive relationship between BDNF and hippocampal volume in patients with schizophrenia (Erickson et al., 2010; Rizos et al., 2011). The present study was performed to examine cortical thickness in drug naïve, first episode schizophrenia patients, and to explore its relationship with serum levels of BDNF. 2. Methods 2.1. Subjects All subjects provided written informed consent to participate in the study, which was approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University. The study sample was composed of a consecutive series of admissions between November 2011 and December 2012. Patients within the age range 18e45 years old, diagnosed with schizophrenia according to the criteria of DSM-IV, were recruited. They were never treated with antipsychotic medications or other psychotropics. The diagnosis of schizophrenia was further determined by a research psychiatrist (X.S.) using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-IV). Exclusion criteria included history of alcohol or other substance use, history of brain injury, or any ongoing significant medical conditions. Healthy control subjects were recruited through advertisement. A complete medical history, physical examination, and routine laboratory tests were obtained from all subjects to rule out possible medical conditions. In addition, the

same research psychiatrist (X.S.) conducted a comprehensive clinical interview to rule out any psychiatric conditions. None of them had a history of any psychiatric conditions, alcohol or substance use. All subjects were right handed. This study was conducted in accordance with the Helsinki Declaration as revised in 1989 (Harrison, 2005). 2.2. Clinical symptom measurement Symptoms of schizophrenia were assessed for all patient subjects using the Positive and Negative Syndrome Scale (PANSS). The PANSS was administered by the same rater (J.G.) throughout the study. 2.3. Serum levels of BDNF measurement Venous blood (5 mL) was collected between 7:00 and 8:00 AM to avoid circadian fluctuation of the parameters to be measured. The blood was drawn to a glass tube and allowed to clot at room temperature. Serum was obtained through centrifugation at 3000 rpm for 10 min, and was then divided into 3 eppendorf tubes, and stored at 70  C for assay. Serum levels of BDNF were measured using enzyme linked immunosorbent assay (ELISA, R&D Systems, USA). The assays were performed according to manufacturer's instructions. The sensitivity of BDNF was 20 pg/mL, with inter-assay variation coefficients of 9.0%, and intra-assay variation coefficients of 5.0%, respectively. Standard curve concentrations were calculated in triplicate for each plate. Absorbencies were measured by a microtiter plate reader (absorbency at 492 nm). All assays were carried out at the same time, and were conducted blinded to the subjects' group status (patients versus controls). 2.4. Image acquisition and processing On the same day after clinical assessment and before any antipsychotic or other psychotropic medication, T1-weighted Spoiled Gradient Echo (SPGR) images were collected for all the subjects on a 3.0 T Scanner (Signa HDxt 3T GEHCGEHC) at the Magnetic Resonance Chamber of the First Affiliated Hospital of Zhengzhou University. All scans were clinically reviewed by a neuroradiologist. The acquisition protocol included the following pulse sequence and parameters: repetition time (TR) ¼ 12 ms, echo time (TE) ¼ 4.5 ms, inversion time (TI) ¼ 1100 ms, flip angle 7, field of view (FOV) ¼ 256  256  220 mm3, matrix size 256  256, slice thickness 1 mm without spacing and scan piles 192, scan time 17 min. Scan results were optimized by high contrast in the gray/ white and gray/CSF boundaries to obtain the best structure and

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009

X. Song et al. / Journal of Psychiatric Research xxx (2014) 1e7

3

Fig. 1. Comparison of cortical thickness between the schizophrenia patient group (SZ) and the healthy control group (HC). Clusters mapped on inflated surface display significant (p < 0.001) as well as trend level (0.001 < p < 0.01) group differences in cortical thickness in lateral, medial, anterior and posterior view respectively after controlling for age and gender. The threshold was set between 2 (p < 0.01) and 5 (p < 0.00001) for figure showing purpose. Red to yellow clusters show decreased thickness, while dark blue to light blue clusters show increased thickness in the patient group. The anatomical region representing the location of the peak vertex in the cluster is given a number. 1: left insula, 2: left superior temporal gyrus (STG), 3: right supramarginal, 4: right inferior parietal, 5: right superior frontal, 6: right lateral occipital, 7: right rostral middle frontal (dorsolateral prefrontal) area. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

surface images. One schizophrenia patient (SZ) and two healthy control subjects (HC) were excluded from the original sample size (46 SZ and 30 HC), due to bad imaging quality, which gave us the current sample size of 45 SZ and 28 HC.

MRI data were coded and cataloged, before they were transferred to the Psychotic Disorders Program, Department of Psychiatry at University of Massachusetts Medical School (UMMS) for blinded analysis. The FreeSurfer software package (version 5.1.0,

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009

4

X. Song et al. / Journal of Psychiatric Research xxx (2014) 1e7

Table 2 Cortical regions with significant between-group difference in thickness based on the vertex-wise group comparison. Cortical area Left hemisphere Insula Superior temporal Right hemisphere Supramarginal Inferior parietal Superior frontal Lateral occipital Rostral middle frontal (dorsolateral prefrontal)

Cluster size (mm2)

Number of vertices

490.23 905.83

1176 2162

3.2614 3.0952

594.06 485.04 177.33 192.72 21.54

1498 1312 368 259 33

2.9991 2.7865 2.6824 2.3215 2.0512

http://surfer.nmr.mgh.harvard.edu/) was used for processing T1weighted SPGR images and parcellating cerebral cortex into anatomical structures (Segonne et al., 2007). Cortical thickness was computed by identifying the shortest distance between a given point on the estimated pial surface and the grayewhite matter boundary and vice versa; and the two values were averaged (Fischl and Dale, 2000). Measurements of cortical thickness have been validated against manual measurements in schizophrenia (Kuperberg et al., 2003). In addition, FreeSurfer morphometric procedures have demonstrated good testeretest reliability across scanner manufacturers, field strengths and various sequence parameters (Wonderlick et al., 2009). Thickness measurements were mapped on the “inflated” surface of each participant's reconstructed brain. Maps were smoothed and averaged across participants using a non-rigid high-dimensional spherical averaging method to align cortical folding patterns. 2.5. Statistical analysis The data were analyzed using SPSS 20.0 (SPSS Inc., Chicago, IL). Demographic and clinical characteristics were reported using descriptive statistics (Table 1). ShapiroeWilk test was used to check the normality of data. Group comparisons were performed using Student's t-test for continuous variables, and Chi-square or Fisher's exact test for categorical variables. For cortical thickness, we used a general linear model to estimate group difference in cortical thickness at each vertex of the surface, controlling the effect of age and gender. Right and left hemispheres were tested separately. A Bonferroni corrected p value of less than 0.05/48 ¼ 0.001 (2-tailed, 48 represents number of cortical regions of interest (ROI) in each hemisphere) was used to identify clusters showing significant vertex-wise between group differences. MRI-derived cortical thickness measures have been demonstrated to be highly reliable. Within-scanner variability of global cortical thickness measurements was <0.03 mm, and the point-wise standard deviation of measurement error was approximately 0.12 mm (Han et al., 2006). In addition, these cortical thickness analyses have been performed on postmortem brains to validate these methods (Rosas et al., 2002), and have been validated against manual measurements in schizophrenia (Kuperberg et al., 2003). Serum levels of BDNF were compared for group difference using independent sample t test. A p value of less than 0.05 (2-tailed) was used for statistical significance. For correlation analysis, ShapiroeWilk test was done first to check the distribution. Based on distribution, Pearson or Spearman correlation was performed between averaged thickness of the clusters and the thickness as well as volume of the entire corresponding ROIs extracted from FreeSurfer. Pearson or Spearman correlation was also used to examine the relationship between averaged cortical thickness of the clusters that show group

Threshold (maximum)

Uncorrected p value <0.001 <0.001 0.0010 0.0016 0.0021 0.0048 0.0089

difference and serum levels of BDNF within each group, the relationships between averaged thickness of the clusters and the PANSS scores within the patient group, and the relationship between serum levels of BDNF and PANSS scores. Multiple regression analysis was then performed to identify demographic/clinical measures that are related to cortical thickness. A p value of less than 0.05 (2-tailed) was used for statistical significance.

3. Results Table 1 shows demographic and clinical characteristics of the study sample. There were no significant differences between the patient group and the HC group in age, gender and education (p's > 0.05). Fig. 1 and Table 2 show the clusters that demonstrated significant or trend level group difference in cortical thickness. These clusters include two in the left hemisphere e insula and superior temporal areas (p's < 0.001); and five in the right hemisphere e supramarginal, inferior parietal, superior frontal, lateral occipital, and rostral middle frontal areas (0.001 < p's < 0.01). Among them, most clusters, especially left insula and superior temporal areas had decreased cortical thickness, while only lateral occipital area had increased cortical thickness in the patient group compared with the HC group. To support the significant findings of left insula and superior temporal gyrus (STG) clusters that show group difference, further Pearson correlation analysis found that averaged thickness of the clusters was positively correlated with the entire corresponding ROI thickness (left insula: r ¼ 0.809, p < 0.001; left STG: r ¼ 0.655, p < 0.001) and volume (left insula: r ¼ 0.342, p ¼ 0.003; left STG: r ¼ 0.234, p ¼ 0.046) extracted from FreeSurfer (Fig. 2). Serum levels of BDNF were significantly lower in the patient group compared with the HC group (741.50 ± 177.06 pg/ml, 833.41 ± 111.55 pg/ml respectively, t(71) ¼ 2.456, p ¼ 0.016). (Fig. 3). After excluding 3 outliers in the patient group using the threshold of mean ± 2 sd, the result remained significant (t(68) ¼ 4.522, p < 0.001). Pearson correlation analysis found that, of the 2 cortical areas showing significant group difference, left insula cluster cortical thickness had significant positive relationship with serum levels of BDNF in the HC group (r ¼ 0.396, p ¼ 0.037) but not in the patient group (r ¼ 0.035, p ¼ 0.819, Fig. 4). No significant relationship between cortical thickness of left STG cluster and serum levels of BDNF was found in either group. Multiple regression analysis was performed between demographic/clinical measures and left insula cluster cortical thickness, and no other factors were significantly associated with insula cortical thickness except serum levels of BDNF (data not shown). No significant correlation was found between cortical thickness of any cortical areas and PANSS total or subscale scores within the patient group (p's > 0.05, data not shown), nor was significant

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009

X. Song et al. / Journal of Psychiatric Research xxx (2014) 1e7

B r=0.809 ***

3.2 3.0 2.8 2.6 2.4 2.2 2.6

2.8

3.0

3.2

3.4

3.6

Average cluster thickness (mm)

Average cluster thickness (mm)

A

3.2

r=0.342 **

3.0 2.8 2.6 2.4 2.2 5000

6000

7000

8000

9000

10000

ROI volume (mm3)

ROI thickness (mm)

r=0.655 ***

2.8 2.6 2.4 2.2 2.0 2.0

2.2

2.4

2.6

2.8

3.0

3.2

Aver age cluster thickness (mm)

D

C Average cluster thickness (mm)

5

r=0.234 *

2.8 2.6 2.4 2.2

2.0 8000 10000 12000 14000 16000 18000 20000

ROI thickness (mm)

ROI volume (mm3)

Fig. 2. Relationship between cluster thickness and entire region of interest (ROI) thickness as well as volume. A) relationship between cluster thickness and ROI thickness of left insula; B) relationship between cluster thickness and ROI volume of left insula; C) relationship between cluster thickness and ROI thickness of left STG; D) relationship between cluster thickness and ROI volume of left STG. *0.01 < p < 0.05, **0.001 < p < 0.01, ***p < 0.001.

4. Discussion This is the first study investigating the association between serum levels of BDNF and cortical thickness in drug naïve, first episode schizophrenia. We were able to replicate previous findings of decreased serum levels of BDNF in first episode schizophrenia compared with healthy controls (Green et al., 2011; Jindal et al., 2010). We further examined the cortical thickness using an automatic vertex-wise group comparison from FreeSurfer software, and found decreased thickness in left insula and STG in the patient group. Furthermore, the cluster thickness of left insula positively correlated with serum levels of BDNF in the control group.

Previous reports of cortical thickness in patients with schizophrenia were inconsistent. Cortical thinning has been found in areas such as insula, temporal regions including superior, middle and inferior temporal gyri, fusiform, parahippocampal gyrus and occipital temporal gyrus (Crespo-Facorro et al., 2011; Qiu et al., 2013), cingulate cortex especially the anterior part (Fornito et al., 2008), insula (Roiz-Santianez et al., 2010), frontal regions including frontopolar cortex, inferior frontal gyrus, superior frontal gyrus and medial orbitofrontal cortex (Crespo-Facorro et al., 2011; Qiu et al., 2013; Sprooten et al., 2013), parietal regions (CrespoFacorro et al., 2011; Sprooten et al., 2013), and occipital regions (Crespo-Facorro et al., 2011; Sprooten et al., 2013). Other studies didn't find significant change in cortical thickness (Haller et al.,

Average cluster thickness (mm)

correlation found between serum levels of BDNF and PANSS total or subscale scores (p's > 0.05, data not shown).

3.2 3.0 2.8 2.6 2.4 2.2 300

600

900

1200

1500

Serum levels of BDNF (pg/ml) HC (r=0.396, p=0.037) Fig. 3. Comparison of serum levels of BDNF between SZ and HC. Significant level was set at p < 0.05.

SZ (r=0.035, p=0.819)

Fig. 4. Relationship between serum levels of BDNF and left insula cluster thickness. Significant level was set at p < 0.05.

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009

6

X. Song et al. / Journal of Psychiatric Research xxx (2014) 1e7

2009; Wiegand et al., 2004). Possible explanations of the differences include different imaging methods, ROIs examined, sample sizes, study populations, and medication status. The regions showing cortical thinning in our study are consistent with most of the earlier studies in patients with chronic schizophrenia, suggesting that morphological findings in chronic schizophrenia are unlikely to be fully attributable to medication related effects, and supporting the possibility of pathophysiological progression of the illness. Studies have shown decreased surface area, volume and cortical thickness in insula in first episode schizophrenia (Crespo-Facorro et al., 2000; Haijma et al., 2013; Rais et al., 2012; Roiz-Santianez et al., 2010; Shepherd et al., 2012). There is evidence suggesting that cortical thickness is positively correlated with cortical volume in both first episode patients and healthy controls (Wiegand et al., 2004). The two regions showing thickness reduction in our study are consistent with the regions that show volume reduction in first episode schizophrenia, especially left STG (McCarley et al., 2002). A recent imaging study in drug naïve first episode schizophrenia proposed that brain volume loss at illness onset is related to reduced surface of the left temporal and insula cortex (Rais et al., 2012), suggesting that cortical gray-matter volume is the product of cortical thickness and cortical surface area. Our findings of correlations between cluster thickness and ROI thickness and volume support this hypothesis. Further, our study suggests asymmetrical brain structural abnormalities in first episode schizophrenia. This finding is supported by another study showing increased cortical thickness asymmetry in regions such as insula, indicating that cortical thickness asymmetry analysis is more accurate than direct cortical thickness measurement in distinguishing first episode schizophrenia from healthy controls (Haller et al., 2009). Taken together, our findings support that schizophrenia is a primary neurodevelopment disorder with cerebral cortex abnormalities (Weinberger, 2002). Previous studies have consistently reported decreased serum or plasma levels of BDNF in schizophrenia even though the degree of decrease varies across different studies (Green et al., 2011). Most of studies have been conducted in chronic, medicated schizophrenia patient; a few recent studies reported decreased serum levels of BDNF in first episode schizophrenia (Jindal et al., 2010; Lee et al., 2011; Rizos et al., 2011), which are consistent with the findings from our study. Our study also found that serum levels of BDNF were positively associated with the cortical thickness of left insula in healthy controls but not in first episode schizophrenia patients. This is the first study reporting a positive correlation between serum levels of BDNF and thickness of left insula. Several previous studies found similar positive correlation between serum levels of BDNF and hippocampal volume in healthy controls (Erickson et al., 2010; Rizos et al., 2011); such a relationship was found in the left, but not in the right hippocampus in both studies, confirming the reported brain structural asymmetry (Erickson et al., 2009). Our significant finding was also in the left, but not in the right hemisphere, indicating asymmetry of insula. In Rizos' study, the correlation between serum levels of BDNF and hippocampus volume was stronger in healthy controls than in first episode schizophrenia (Rizos et al., 2011). Our study found similar correlation pattern. In a recent longitudinal study in 14 first episode schizophrenia patients, Rizos et al. found a positive association between changes of BDNF levels and changes of left hippocampal volume after 8 months of antipsychotic treatment (Rizos et al., 2014) suggesting that neuroprotective effect through modulation of BDNF levels could in turn protect against brain tissue loss (Siekmeier and vanMaanen, 2013; Wakade et al., 2002). Based on these findings, an absence or reduction of the positive relationship between brain morphology of

insula and serum levels of BDNF that present in the healthy cohort may suggest a breakdown in the protective role or neurotrophic effect of BDNF in these brain regions in psychosis, which is consistent with the neurotrophic theory of schizophrenia (Weinberger, 1987). We didn't find significant correlation between cortical thickness and clinical symptoms measured by PANSS, or between serum levels of BDNF and clinical symptoms. While some studies have shown significant correlations between cortical thickness and clinical symptoms (Buckley et al., 2007), most other studies have shown that cortical thinning does not seem to influence clinical and functional outcomes (Cobia et al., 2012; Crespo-Facorro et al., 2011; Roiz-Santianez et al., 2010). The fact that cortical thinning is state independent further supports the neurodevelopmental hypothesis of schizophrenia, and suggests that cortical thickness may serve as an endophenotype biomarker of the illness (Crespo-Facorro et al., 2011; Roiz-Santianez et al., 2010). Little is known about the relationship between serum levels of BDNF and clinical symptoms. One study didn't find correlation between them at onset, but found significant positive correlation of BDNF increase with PANSS scores after treatment (Lee et al., 2011). Further studies are needed to address the question of the relationship between BDNF and symptoms. Our study has several limitations. First, we used an automatic method to identify cortical regions, which is limited in accuracy as compared to manual tracing. Second, multiple comparisons were performed in the vertex-wise group comparison analysis, and we only did a Bonferroni correction for the number of cortical regions, instead of number of vertices, which could raise the possibility of false positive results. Third, BDNF levels were assessed in serum, not directly in the brain. Zuccato et al. mentioned the reliability issue of measuring serum BDNF as a biomarker for Hungtinton's disease, as well as for patients of other illness conditions (Zuccato et al., 2011). Fourth, the study sample size, in particular in number of subjects in the HC group, is relatively small. Fifth, our findings should be considered preliminary as no statistical correction was used for multiple correlations. Future studies are needed to examine the longitudinal change of cortical thickness, and its relationship with serum levels of BDNF and antipsychotic therapy. In conclusion, our data suggest that cortical thinning and decreased serum levels of BDNF exist in the early stage of schizophrenia and before antipsychotic treatment. The protective and neurotrophic role of BDNF in insula is impaired in first episode schizophrenia. These findings support the theory of neurodevelopmental contributions to the development of schizophrenia. Role of the funding source Funding for this study was provided by the National Natural Science Foundation of China (No.30971058 to X-QS; No.81071090 to L-XL), the Natural Science Foundation of Henan (No.102300413208, 112300413226 to L-XL). Contributors Dr. Song and Dr. Quan were responsible for the analysis and interpretation of the data for this paper. All authors contributed to the writing of the paper. Conflict of interest Dr. Fan has received research support or honoraria from Eli Lilly, AstraZeneca, Bristol-Myer-Squibb, Janssen, and Pfizer. Other authors report no competing interests.

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009

X. Song et al. / Journal of Psychiatric Research xxx (2014) 1e7

Acknowledgments None. References Buckley PF, Mahadik S, Pillai A, Terry Jr A. Neurotrophins and schizophrenia. Schizophr Res 2007;94:1e11. Cobia DJ, Smith MJ, Wang L, Csernansky JG. Longitudinal progression of frontal and temporal lobe changes in schizophrenia. Schizophr Res 2012;139:1e6. Crespo-Facorro B, Kim J, Andreasen NC, O'Leary DS, Bockholt HJ, Magnotta V. Insular cortex abnormalities in schizophrenia: a structural magnetic resonance imaging study of first-episode patients. Schizophr Res 2000;46:35e43. Crespo-Facorro B, Roiz-Santianez R, Perez-Iglesias R, Rodriguez-Sanchez JM, Mata I, Tordesillas-Gutierrez D, et al. Global and regional cortical thinning in firstepisode psychosis patients: relationships with clinical and cognitive features. Psychol Med 2011;41:1449e60. Eisenberg DP, Ianni AM, Wei SM, Kohn PD, Kolachana B, Apud J, et al. Brain-derived neurotrophic factor (BDNF) Val(66)Met polymorphism differentially predicts hippocampal function in medication-free patients with schizophrenia. Mol Psychiatry 2013;18:713e20. Elvevag B, Goldberg TE. Cognitive impairment in schizophrenia is the core of the disorder. Crit Rev Neurobiol 2000;14:1e21. Erickson KI, Prakash RS, Voss MW, Chaddock L, Heo S, McLaren M, et al. Brainderived neurotrophic factor is associated with age-related decline in hippocampal volume. J Neurosci 2010;30:5368e75. Erickson KI, Prakash RS, Voss MW, Chaddock L, Hu L, Morris KS, et al. Aerobic fitness is associated with hippocampal volume in elderly humans. Hippocampus 2009;19:1030e9. Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci U S A 2000;97:11050e5. Fornito A, Yucel M, Patti J, Wood SJ, Pantelis C. Mapping grey matter reductions in schizophrenia: an anatomical likelihood estimation analysis of voxel-based morphometry studies. Schizophr Res 2009;108:104e13. Fornito A, Yucel M, Wood SJ, Adamson C, Velakoulis D, Saling MM, et al. Surfacebased morphometry of the anterior cingulate cortex in first episode schizophrenia. Hum Brain Mapp 2008;29:478e89. Gorski JA, Zeiler SR, Tamowski S, Jones KR. Brain-derived neurotrophic factor is required for the maintenance of cortical dendrites. J Neurosci 2003;23: 6856e65. Green MJ, Matheson SL, Shepherd A, Weickert CS, Carr VJ. Brain-derived neurotrophic factor levels in schizophrenia: a systematic review with meta-analysis. Mol psychiatry 2011;16:960e72. Haijma SV, Van Haren N, Cahn W, Koolschijn PC, Hulshoff Pol HE, Kahn RS. Brain volumes in schizophrenia: a meta-analysis in over 18 000 subjects. Schizophr Bull 2013;39:1129e38. Haller S, Borgwardt SJ, Schindler C, Aston J, Radue EW, Riecher-Rossler A. Can cortical thickness asymmetry analysis contribute to detection of at-risk mental state and first-episode psychosis? A pilot study. Radiology 2009;250:212e21. Han X, Jovicich J, Salat D, van der Kouwe A, Quinn B, Czanner S, et al. Reliability of MRI-derived measurements of human cerebral cortical thickness: the effects of field strength, scanner upgrade and manufacturer. NeuroImage 2006;32: 180e94. Harrison JE. Orthodontic Clinical Trials III: reporting of ethical issues associated with clinical trials published in three orthodontic journals between 1989 and 1998. J Orthod 2005;32:115e21. Hashimoto T, Bergen SE, Nguyen QL, Xu B, Monteggia LM, Pierri JN, et al. Relationship of brain-derived neurotrophic factor and its receptor TrkB to altered inhibitory prefrontal circuitry in schizophrenia. J Neurosci : Off J Soc Neurosci 2005;25:372e83. Ho BC, Andreasen NC, Dawson JD, Wassink TH. Association between brain-derived neurotrophic factor val66met gene polymorphism and progressive brain volume changes in schizophrenia. Am J Psychiatry 2007;164:1890e9. Huang EJ, Reichardt LF. Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci 2001;24:677e736. Jindal RD, Pillai AK, Mahadik SP, Eklund K, Montrose DM, Keshavan MS. Decreased BDNF in patients with antipsychotic naive first episode schizophrenia. Schizophr Res 2010;119:47e51. Kuperberg GR, Broome MR, McGuire PK, David AS, Eddy M, Ozawa F, et al. Regionally localized thinning of the cerebral cortex in schizophrenia. Arch Gen Psychiatry 2003;60:878e88. Lee AH, Lange C, Ricken R, Hellweg R, Lang UE. Reduced brain-derived neurotrophic factor serum concentrations in acute schizophrenic patients increase during antipsychotic treatment. J Clin Psychopharmacol 2011;31:334e6. Levitt JJ, Bobrow L, Lucia D, Srinivasan P. A selective review of volumetric and morphometric imaging in schizophrenia. Curr Top Behav Neurosci 2010;4: 243e81.

7

Lu B. BDNF and activity-dependent synaptic modulation. Learn Mem 2003;10: 86e98. McCarley RW, Salisbury DF, Hirayasu Y, Yurgelun-Todd DA, Tohen M, Zarate C, et al. Association between smaller left posterior superior temporal gyrus volume on magnetic resonance imaging and smaller left temporal P300 amplitude in firstepisode schizophrenia. Arch Gen Psychiatry 2002;59:321e31. Pan W, Banks WA, Fasold MB, Bluth J, Kastin AJ. Transport of brain-derived neurotrophic factor across the blood-brain barrier. Neuropharmacology 1998;37: 1553e61. Qiu A, Gan SC, Wang Y, Sim K. Amygdala-hippocampal shape and cortical thickness abnormalities in first-episode schizophrenia and mania. Psychol Med 2013;43: 1353e63. Rais M, Cahn W, Schnack HG, Hulshoff Pol HE, Kahn RS, van Haren NE. Brain volume reductions in medication-naive patients with schizophrenia in relation to intelligence quotient. Psychol Med 2012;42:1847e56. Rizos E, Papathanasiou MA, Michalopoulou PG, Laskos E, Mazioti A, Kastania A, et al. A longitudinal study of alterations of hippocampal volumes and serum BDNF levels in association to atypical antipsychotics in a sample of first-episode patients with schizophrenia. PloS One 2014;9:e87997. Rizos EN, Papathanasiou M, Michalopoulou PG, Mazioti A, Douzenis A, Kastania A, et al. Association of serum BDNF levels with hippocampal volumes in first psychotic episode drug-naive schizophrenic patients. Schizophr Res 2011;129: 201e4. Roiz-Santianez R, Perez-Iglesias R, Quintero C, Tordesillas-Gutierrez D, Mata I, Ayesa R, et al. Insular cortex thinning in first episode schizophrenia patients. Psychiatry Res 2010;182:216e22. Rosas HD, Liu AK, Hersch S, Glessner M, Ferrante RJ, Salat DH, et al. Regional and progressive thinning of the cortical ribbon in Huntington's disease. Neurology 2002;58:695e701. Sartorius A, Hellweg R, Litzke J, Vogt M, Dormann C, Vollmayr B, et al. Correlations and discrepancies between serum and brain tissue levels of neurotrophins after electroconvulsive treatment in rats. Pharmacopsychiatry 2009;42:270e6. Segonne F, Pacheco J, Fischl B. Geometrically accurate topology-correction of cortical surfaces using nonseparating loops. IEEE Trans Med imaging 2007;26: 518e29. Shenton ME, Dickey CC, Frumin M, McCarley RW. A review of MRI findings in schizophrenia. Schizophr Res 2001;49:1e52. Shepherd AM, Laurens KR, Matheson SL, Carr VJ, Green MJ. Systematic meta-review and quality assessment of the structural brain alterations in schizophrenia. Neurosci Biobehav Rev 2012;36:1342e56. Siekmeier PJ, vanMaanen DP. Development of antipsychotic medications with novel mechanisms of action based on computational modeling of hippocampal neuropathology. PloS One 2013;8:e58607. Sprooten E, Papmeyer M, Smyth AM, Vincenz D, Honold S, Conlon GA, et al. Cortical thickness in first-episode schizophrenia patients and individuals at high familial risk: a cross-sectional comparison. Schizophr Res 2013;151:259e64. Voineskos AN, Foussias G, Lerch J, Felsky D, Remington G, Rajji TK, et al. Neuroimaging evidence for the deficit subtype of schizophrenia. JAMA Psychiatry 2013;70:472e80. Wakade CG, Mahadik SP, Waller JL, Chiu FC. Atypical neuroleptics stimulate neurogenesis in adult rat brain. J Neurosci Res 2002;69:72e9. Webster MJ, Herman MM, Kleinman JE, Shannon Weickert C. BDNF and trkB mRNA expression in the hippocampus and temporal cortex during the human lifespan. Gene Expr Patterns 2006;6:941e51. Weickert CS, Ligons DL, Romanczyk T, Ungaro G, Hyde TM, Herman MM, et al. Reductions in neurotrophin receptor mRNAs in the prefrontal cortex of patients with schizophrenia. Mol psychiatry 2005;10:637e50. Weinberger DR. Implications of normal brain development for the pathogenesis of schizophrenia. Arch Gen Psychiatry 1987;44:660e9. Weinberger DR. Neurotoxicity, neuroplasticity, and magnetic resonance imaging morphometry what is happening in the schizophrenic brain? JAMA Psychiatry 2002;59:553e8. Wiegand LC, Warfield SK, Levitt JJ, Hirayasu Y, Salisbury DF, Heckers S, et al. Prefrontal cortical thickness in first-episode psychosis: a magnetic resonance imaging study. Biol psychiatry 2004;55:131e40. Wonderlick JS, Ziegler DA, Hosseini-Varnamkhasti P, Locascio JJ, Bakkour A, van der Kouwe A, et al. Reliability of MRI-derived cortical and subcortical morphometric measures: effects of pulse sequence, voxel geometry, and parallel imaging. NeuroImage 2009;44:1324e33. Xu B, Zang K, Ruff NL, Zhang YA, McConnell SK, Stryker MP, et al. Cortical degeneration in the absence of neurotrophin signaling: dendritic retraction and neuronal loss after removal of the receptor TrkB. Neuron 2000;26:233e45. Zuccato C, Marullo M, Vitali B, Tarditi A, Mariotti C, Valenza M, et al. Brain-derived neurotrophic factor in patients with Huntington's disease. PloS One 2011;6: e22966. Zugman A, Gadelha A, Assuncao I, Sato J, Ota VK, Rocha DL, et al. Reduced dorsolateral prefrontal cortex in treatment resistant schizophrenia. Schizophr Res 2013;148:81e6.

Please cite this article in press as: Song X, et al., Decreased cortical thickness in drug naïve first episode schizophrenia: In relation to serum levels of BDNF, Journal of Psychiatric Research (2014), http://dx.doi.org/10.1016/j.jpsychires.2014.09.009