White matter microstructural characteristics in Bipolar I and Bipolar II Disorder: A diffusion tensor imaging study

Journal of Affective Disorders 189 (2016) 176–183

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Research report

White matter microstructural characteristics in Bipolar I and Bipolar II Disorder: A diffusion tensor imaging study Elisa Ambrosi a,b, Chiara Chiapponi a,c, Gabriele Sani a,b,d, Giovanni Manfredi a,b,d, Fabrizio Piras a, Carlo Caltagirone a,e, Gianfranco Spalletta a,f,n a

Neuropsychiatry Laboratory, Department of Clinical and Behavioral Neurology, IRCCS Santa Lucia Foundation, Rome, Italy NESMOS Department, Sapienza University of Rome, Italy c Department of Systems Medicine, “Tor Vergata” University, Rome, Italy d Centro Lucio Bini, Rome, Italy e Department of Neuroscience, “Tor Vergata” University, Rome, Italy f Beth K. and Stuart C. Yudofsky Division of Neuropsychiatry, Menninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, Houston, TX, USA b

art ic l e i nf o

a b s t r a c t

Article history: Received 17 June 2015 Received in revised form 8 September 2015 Accepted 18 September 2015 Available online 25 September 2015

Background: Diffusion tensor imaging (DTI) studies of bipolar disorder (BD) report contrasting results and are mainly focused on bipolar I (BD-I) samples. We aimed at investigating how and where DTI parameters differ between BD-I and bipolar II (BD-II) and between BD and healthy control subjects (HC). Methods: We conducted a tract-based spatial statistics analysis of DTI derived parameters, namely fractional anisotropy (FA), axial diffusivity (AD) and radial diffusivity (RD) in a matched sample of 50 BD (25 BD-I and 25 BD-II) during the chronic course of the illness and 50 HC. Results: Compared to BD-I and HC, BD-II showed lower FA but no significant AD or RD differences in the right inferior longitudinal fasciculus (ILF). Both patient groups showed lower AD and RD in the left internal capsule and lower AD across the left ILF, the cortico-spinal tract within the right hemisphere and bilaterally in the cerebellum with respect to HC. Limitations: Patients were medicated at the time of scanning; the BD-II group had higher Hamilton Rating Scale for Depression scores than the BD-I group. Conclusions: BD-II patients differ from BD-I in the ILF. Both BD subtypes showed widespread white matter (WM) changes in the internal capsule, cortico-spinal tract and cerebellum. The loss of WM integrity in BD-II might be due to demyelination whereas WM changes common to both subgroups could be attributable to axonal damage. & 2015 Elsevier B.V. All rights reserved.

Keywords: Bipolar Disorder, type I Bipolar Disorder, type II Diffusion tensor imaging White matter integrity

1. Introduction Structural brain imaging studies in individuals with bipolar disorder (BD) have described white matter (WM) changes in early (Vita et al., 2009) and late phases (Vederine et al., 2011) of the illness. However, there is no definitive evidence that patients with BD type I (BD-I) and II (BD-II) are differently characterized by specific structural abnormalities (Mahon et al., 2009). Diffusion Tensor Imaging (DTI), a magnetic resonance technique sensitive to the movement of water molecules, has been used to investigate WM microstructure in BD; however, DTI studies have reported conflicting results. Indeed, fractional anisotropy n Corresponding author at: Neuropsychiatry Laboratory, Department of Clinical and Behavioral Neurology, IRCCS Santa Lucia Foundation, Via Ardeatina 306, 00179 Rome, Italy. Fax: þ39 06 51501575. E-mail address: [email protected] (G. Spalletta).

http://dx.doi.org/10.1016/j.jad.2015.09.035 0165-0327/& 2015 Elsevier B.V. All rights reserved.

(FA), the most replicated WM integrity index reflecting axonal coherence, was found reduced (Bauer et al., 2015; Haller et al., 2011; Sussmann et al., 2009; Vederine et al., 2011), increased (Wessa et al., 2009) or unchanged (Beyer et al., 2005) in patients with BD compared to healthy controls (HC). Fewer studies have investigated other DTI parameters, such as axial diffusivity (AD) and radial diffusivity (RD). These parameters give an in vivo measure of water diffusion parallel and perpendicular to WM fibers and reflect axonal and myelin integrity respectively (Beaulieu, 2002). RD values have been found increased at the onset of BD (Lu et al., 2012) and when comparing BD patients to those with unipolar depression (Benedetti et al., 2011) or unchanged in elderly BD (Haller et al., 2011). However, most of the above mentioned studies focused on patients suffering from BD-I; only a few included mixed groups of BD-I and BD-II (Bruno et al., 2008; Mahon et al., 2009). Thus, these findings may reflect the specific neurobiological substrate of BD-I.

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

From a clinical point of view, BD-II is characterized by more frequent episodes of depression (Vieta et al., 1997), shorter euthymic periods (Judd et al., 2003) and cognitive deficits comparable with those found in BD-I (Bora et al., 2011). Indeed, episodes of depression and depressive symptoms have been found to lead to psychosocial disability as well as manic or hypomanic events (Judd et al., 2005). Whether or not these phenotypic characteristics of BD-II have different brain structural correlates than those found in BD-I has still to be clarified. The under-recruitment of BDII patients in neuroimaging studies may be due to greater misdiagnosis of BD-II compared to BD-I (Fiedorowicz et al., 2011). Recently, functional and structural studies comparing the two subgroups suggested that BD-II patients have increased activity in the ventral striatum as well as higher volume in the left putamen (Caseras et al., 2013), while BD-I show reduced cortical volume in the orbitofrontal region and reduced thickness of the temporal cortex (Maller et al., 2014). Moreover, DTI studies on BD-II patients found widespread FA reductions in all major WM tracts, including cortico-cortical associative and interhemispheric fibers (Ambrosi et al., 2013; Yip et al., 2013). To date, few studies have compared BD-I, BD-II and HC for WM abnormalities. Compared to HC, both BD subgroups were shown to have FA reductions in the WM of the corpus callosum, cingulum and right frontal regions (Ha et al., 2011; Liu et al., 2010). A direct comparison between BD-I and BD-II indicated that BD-I have lower FA values in the right temporal WM (Ha et al., 2011) while BD-II have lower FA values in the right precuneus, frontal and prefrontal regions (Liu et al., 2010) and widespread prefrontal and temporal RD increase (Maller et al., 2014). Hence, we performed a cross-sectional study on DTI parameters of BD-I, BD-II and HC. We aimed at investigating FA, AD and RD differences among the three groups and at identifying potential axonal or myelination damage. We expected to find WM impairments in both BD-I and BD-II compared to HC. Due to the lack of data on BD-II, we could not make an a priori hypothesis about specific brain regions in which the two subgroups may have microstructural differences. We performed a whole brain analysis to address this issue.

177

2. Methods 2.1. Subjects Sixty patients with BD-I (n¼ 32) and BD-II (n ¼28) were initially recruited at the IRCCS Santa Lucia Foundation of Rome. The diagnosis of BD was made according to the Diagnostic and Statistical Manual of Mental Disorders IV-Edition, text revised (DSMIV-TR) (American Psychiatric Association, 2000). The clinician who had been treating the patients and knew their clinical history, but who was blind to the aims of the study, made the preliminary diagnosis. Then a senior research psychiatrist confirmed all preliminary diagnoses using the Structured Clinical Interview for DSM-IV-TR-Patient Edition (SCID-I) (First et al., 2002a). From this original group, 7 patients with BD-I and 3 patients with BD-II were excluded for the presence of moderate or severe brain vascular lesions (see exclusion criteria). The final sample for this study consisted of 50 patients (BD-I ¼25; BD-II ¼ 25). Inclusion criteria were: (a) age 4 16 years and o75 years; (b) no additional axis I diagnosis; (c) at least five years of education. Exclusion criteria were: (a) traumatic brain injury with loss of consciousness; (b) lifetime history of major medical (e.g. chronic kidney failure, chronic heart failure, decompensated diabetes, etc.) or neurological disorders; (c) history of substance abuse or dependence; (d) dementia or cognitive deterioration according to DSM-IV-TR criteria and a Mini-Mental State Examination (MMSE) (Folstein et al., 1975) score lower than 25, according to the normative data of the Italian population (Measso et al., 1993); (e) contraindication for MRI; (f) any potential brain abnormality or microvascular lesion as apparent on conventional FLAIR-scans through white matter hyperintensities (WMH); in particular, the presence, severity and location of WMH were computed using the semi-automated method recently published by our group (Iorio et al., 2013). Patients with BD-II were recruited only if they had a stable diagnosis for at least six years, to avoid as much as possible diagnostic type changes. Mood symptoms were rated using the Young Mania Rating Scale (YMRS) (Young et al., 1978) and the 17item Hamilton Rating Scale for Depression (HAM-D) (Hamilton, 1960). All but two patients were receiving drug treatment during the evaluation. We also recruited 50 HC subjects in the same geographical area,

Table 1 Sociodemographic and clinical characteristics of BD-I and BD-II patients and HC subjects. Characteristics

BD-I (n¼ 25)

BD-II (n¼ 25)

HC (n¼ 50)

t, F or χ2

df

p

Age (years), mean (SD) Males, n(%) Educational level (years), mean (SD) Duration of illness (years), mean (SD) Number of past manic/hypomanic episodes, mean (SD) Number of past depressive, mean (SD) episodes HAM-D score, mean (SD) YMRS score, mean (SD) Current medication , n(%) Antidepressants, n(%) Antipsychotics, n(%) Antiepileptics, n(%) Lithium, n(%) Benzodiazepines, n(%) Other treatments

48.6 13 14.2 20.3 5.09

48.4 13 14.3 23.9 7.05

48.3 (12.0) 26 (52) 14.2 (3.2) – –

0.006 0.00 0.986  1.13  0.97

2 2 2 45 38

0.99 4 0.999 0.98 0.26 0.34

–

 0.50

39

0.61

– – – – – – – – –

 2.39  0.52

48 45

0.02 0.60

2 2 2 2 2 –

0.34 0.32 0.64 0.13 0.07 –

(11.4) (52) (4.2) (10.7) (5.4)

7.5 (9.8) 9.2 (8.3) 3.2 (2.2) 23 (92) 10 (40) 16 (64) 11 (44) 16 (64) 6 (24) –

(12.7) (52) (3.6) (11.7) (7.41)

9.7 (17.9) 14.5 (7.4) 3.5 (2.8) 22 (88) 6 (24) 12 (48) 14 (56) 10 (40) 13 (52) –

2.1 2.22 0.87 4.01 5.24 –

BD-I¼ patients with type I bipolar disorder, BD-II ¼ patients with type II bipolar disorder, HC ¼healthy controls; SD ¼ standard deviation; df¼ degrees of freedom. Note that for some patients we were unable to collect data on duration of illness (n¼ 3), number of past manic/hypomanic (n¼ 10) and depressive episodes (n ¼9), and YMRS score (n¼ 3).

178

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

carefully matched with BD patients for age, gender and educational level (see Table 1). All HC were screened for a current or lifetime history of DSM-IV-TR Axis I and II disorders using the SCID-I/NP (First et al., 2002b) and SCID-II (First et al., 1997); they were also assessed to confirm that no first-degree relative had a history of psychosis. Exclusion criteria were the same as those for patients group. Sociodemographic and clinical characteristics of the HC, BD-I and BD-II samples are shown in Parkash, O., Yean, C.Y., Shueb, R.H Table 1. The study was approved and undertaken in accordance with the guidelines of the Santa Lucia Foundation Ethics Committee. All participants gave their written informed consent to participate in the research after they had received a complete explanation of the study procedures. 2.2. Image acquisition and processing All 100 participants included in this study underwent the same imaging protocol, which comprehended DTI, T2-weighted and FLAIR sequences using a 3T Allegra MR imager (Siemens, Erlangen, Germany) with a standard quadrature head coil. Diffusionweighted volumes were acquired using spin-echo EPI (TE/TR ¼ 89/ 8500 ms, bandwidth ¼2126 Hz/vx; matrix size 128  128; 80 axial slices, voxel size 1.8  1.8  1.8 mm3) with 30 isotropically distributed orientations for the diffusion sensitizing gradients at a bvalue of 1000 s/mm2 and two no diffusion weighted images (b0). Scanning was repeated three times to increase the signal-to-noise ratio (Cherubini et al., 2009). T2 and FLAIR sequences were acquired to screen for brain pathology. Two expert neuroradiologists examined all images of all subjects to exclude those with potential brain abnormalities and vascular lesions. DTI images were processed using FSL 4.1 software (www.fmrib. ox.ac.uk/fsl/). First, they were corrected for the distortion induced by eddy currents and head motions by applying a 3D full affine alignment of each image to the mean b0 image. After distortion corrections, DTI data were averaged and concatenated into 31 (1 b0þ 30 b1000) volumes. A diffusion tensor model was fitted at each voxel and it generated FA, AD and RD maps. We used tractbased spatial statistics (TBSS) (Smith et al., 2006) version 1.2, part of FSL, for the post processing of FA, AD and RD maps in the so called WM skeleton, i.e. an alignment invariant tract representation. Briefly, TBSS first projects all subjects' microstructural data onto the skeleton in the standard MNI space by means of the nonlinear registration tool FNIRT (Andersson et al., 2007a, 2007b), which uses a b-spline representation of the registration warp field (Rueckert et al., 1999). This process of projecting individual maps onto a mean skeleton helps confine the effect of cross-spatial subject variability that remains after classical non-linear registration. The resulting data are then fed into voxel-wise cross-subject statistics, i.e. the “randomise” command in the FSL package. 2.3. Statistical analyses Comparisons between the three diagnostic groups (i.e. HC, BD-I and BD-II) for sociodemographic and clinical characteristics were performed using t-test, F-test or chi-square test. We ran voxel-wise analyses of variance (ANOVAs) to explore the possible effect of diagnostic group (independent variable) in brain DTI indices (dependent variables) over the WM skeleton (pANOVA o 0.05, cluster threshold 50 voxels). Whenever a significant group effect emerged, pairwise post-hoc t-tests were performed to clarify the direction of the effect. We considered the post-hoc results that survived family wise error (FWE) correction (i) using the Threshold-Free Cluster Enhancement option (Smith and Nichols, 2009) in randomise to take into account the number

Table 2 White matter clusters in which an effect of diagnosis emerged. White matter area

Fractional anisotropy Temporal lobe R R Subcallosal cortex R area Cingulate gyrus R area Parietal lobe L Frontal lobe L Thalamic radiation R area Occipital lobe R Temporal lobe L Occipital lobe L Temporal lobe R Parietal lobe L R Temporal lobe R Thalamic radiation L area Frontal lobe R Parietal lobe R L Frontal lobe R R Parietal lobe R Axial diffusivity Corticospinal tract Precentral area

Occipital lobe

R R L R R R L R L R

Radial diffusivity Cerebellum

L

Insula area

MNI coordinates of statistical peak (x, y, z)

Cluster size (number of voxels)

0.005 0.017 0.021

52,  41,  22 53,  52,13 1,16,  1

1,122 352 329

0.007

9,39,7

268

0.02 0.006 0.025

 19,  72,42  6,13,  22 21, 38,7

261 211 187

0.023 0.016 0.013 0.018 0.021 0.028 0.035 0.029

26,  90,  1  35,  27,  29  12,  89,8 51,  60,2  7,  55,11 12,  59,57 26,  3,  11  17, 31,8

171 160 136 126 87 82 81 80

0.03 0.022 0.028 0.021 0.036 0.014

4,3,  15 9,  51,51  10,  63,29 27,38,34 24,21,  14 17,  57,61

76 69 64 61 61 52

o 0.001

5,  31,  43 9,  15,69  11,  24,69 14,  39,69 12,  40,69 14,  23,68  14,  10,67 38,14,  8  34,12,  13 38,  79,  17

94,323

Laterality p of statistical peak

o 0.001 0.01 o 0.001

R

 27,  55,  30  14,  53,  18 10,  50,  18 12,  46,  16 12,  49,  15 26,2,  12 24,4,  11 34,  82,  14  10,  69,50  7,  21,67  6,  15,36 9,  24,55 8,  59,46 34,  81,  18  35,16,  4  27,  59,15

844 553 267 252 210 162 146 136 131

41,734

Fronto-temporal lobe Occipital lobe Parietal lobe

R L

Cingulum area Parietal lobe

L R

Occipital lobe Insula area Posterior thalamic radiation Frontal lobe Occipital lobe Parietal lobe Postcentral area Frontal lobe Cingulate gyrus area Temporal lobe Frontal lobe

R L L

0.013 0.009 0.008 0.012 0.01 0.007 0.013 o 0.001 0.035

R R L L L R

0.026 0.027 0.021 0.017 0.021 0.013

8,32,54 14,  84,33  7,  61,19  27,  33,64  40,33,2 41,11,  6

126 119 109 96 91 85

R L

0.016 0.027 0.033 0.025 0.032 0.028 0.02

6,  28,42  50,  9,  23  14,58,3 8,8,64  10,54,28 32,38,11 23,  50,61

81 77 73 70 67 63 59

Parietal lobe

R

182 170 150

R L R

L ¼ left, R¼ right. p values are not corrected for multiple comparisons. The lower threshold for cluster size is 50 voxels.

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

of voxel-wise comparisons within a single t-test and (ii) taking into account the number of t-tests performed. Thus, since we performed 6 post-hoc tests for each DTI parameter, the final p value considered was ppost-hoc_MultipleComparisons ¼ 0.05/6¼ 0.008. Reported results were obtained after 5000 permutations in the randomise process.

3. Results

Table 3 Post-hoc pairwise comparisons of DTI indices between diagnostic groups. White matter area

Laterality p of statistical peak

Fractional anisotropy BD-II oBD-I ILF R

3.1. Sociodemographic and clinical characteristics The three diagnostic groups did not significantly differ for age, gender or educational level (see Table 1). BD-II showed increased HAM-D scores in comparison to BD-I. No other clinical variables differed significantly between BD-I and BD-II. 3.2. Brain microstructural characteristics A global effect of diagnosis emerged in clusters bilaterally distributed over the WM skeleton for each DTI parameter examined (pANOVA o 0.05) (see Table 2). Such main effect did not show significant variations when controlling for HAM-D score. AD and RD revealed a diagnosis effect in wide clusters ( 94,323 and 41,734 voxels) covering most parts of the WM skeleton. FA revealed a diagnosis effect in more localized clusters belonging to the bilateral temporal, parietal, frontal and occipital lobes. To localize the direction of these FA, AD and RD effects, Table 3 shows the results of the post-hoc t-tests. FA was reduced in BD-II patients: (i) with respect to both BD-I and HC in a WM area including the right inferior longitudinal fasciculus (ILF) (see Fig. 1), and (ii) with respect to HC only in the left fornix and left parietal lobe. It is interesting to note that at the uncorrected level BD-II patients had: (i) higher RD than both BD-I and HC in the right ILF, and (ii) higher RD than only HC in the left fornix and left parietal lobe (data available upon request). However, statistical significance of these RD relations did not survive correction for multiple comparisons. The two subgroups of patients did not significantly differ (ppost-hoc_MultipleComparisons 40.05) for either AD or RD. Compared to HC, both BD-I and BD-II patients had: (i) reduced AD in wide clusters bilaterally distributed over the cerebellum, right corticospinal tract (from the brainstem to the parietal lobe), left ILF and left internal capsule, and (ii) reduced RD in a cluster belonging to the left internal capsule. BD-I patients showed a further RD reduction with respect to HC in a cluster covering the left ILF, in areas bilaterally distributed in the fronto-parietal lobe and bilaterally in the cerebellum. The combined interpretation of the directional, interplaying diffusion-related parameters (i.e. AD and RD) allowed us to highlight regions of interest (left internal capsule, right cortico-spinal tract, bilateral cerebellum and left ILF) with peculiar microdiffusivity properties: (i) the left internal capsule was the only area in which both BD-I and BD-II subgroups showed a concurrent significant reduction of AD and RD with respect to HC (see Fig. 2), (ii) in the bilateral cerebellum, the right cortico-spinal tract and the left ILF both patients subgroups showed a significant reduction only in AD. In these latter regions (cerebellum, cortico-spinal tract and ILF) only BD-I had a significant RD reduction with respect to HC.

4. Discussion The aim of our study was to detect potential brain microstructural differences among BD-I, BD-II and HC by comparing FA,

179

BD-II oHC ILF Fornix Parietal lobe

Axial diffusivity BD-Io HC Corticospinal tract Cerebellum Cerebellum Internal capsule Precentral area ILF BD-II oHC Corticospinal tract Cerebellum Cerebellum Precentral area Internal capsule ILF

R L L

R

L

R

L

Radial diffusivity BD-Io HC ILF L Middle frontal gyrus area Precentral area

Internal capsule Cerebellum Cerebellum Postcentral area Precentral area

BD-II oHC Internal capsule

R

L

MNI coordinates of statistical peak (x, y, z)

Cluster size (number of voxels)

0.001 50,  35, 10 52,  37,  8 43,  39,  8 50,  40,  8 46,  37,  9 50,  38,  9 43,  34,  10

836

0.001 39,  31,  18 0.001  2,  11,15 0.002  18,  62,39

373 253 164

o 0.001 5,  31,  43 10,  26,68 9,  39,65 13,  62,  28  8,  63,  29  16,  9,  1  7,  22,67  40,  33,  1

76,347

o 0.001 5,  31,  43 13,  23,66 21, 35,66 13,  62,  28  8,  63,  29  8,  22,68  16,  9,  1  40,  33,  1

81,267

o 0.001

21,990

 40,  33,  1  31,3,52  16,12,47  24, 14,46  30,  15,46  24, 16,46  29,4,44  16,  9,  1  8,  63,  29 o 0.001 13,  62,  28 36,  26,41 41,  6,47 43,  6,46 46,  3,42 41,  5,42 39,  6,41 0.006

 16,  9,  1  17,  7,6  17,  11, 2

12,122

2,104

HC¼healthy controls, BD-I¼ patients with type I bipolar disorder, BD-II ¼ patients with type II bipolar disorder. L ¼ left, R¼ right. ILF¼ Inferior longitudinal fasciculus. The lower threshold for cluster size is 50 voxels. Results are corrected for multiple comparisons with p o0.008.

AD and RD. As expected, both BD-I and BD-II showed WM microstructural changes with respect to HC. Our main result is that BD-II had lower FA in the right ILF compared to both BD-I and HC. By contrast, the FA of BD-I was not different from that of HC in that region. Our second result refers to the directional parameters AD and RD. With respect to HC, both subgroups of BD patients showed

180

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

Fig. 1. Reduction of FA in BD-II. Sagittal, coronal and axial view of the cluster including the right inferior longitudinal fasciculus in which BD-II have reduced FA with respect to both BD-I and HC. R¼ right and L ¼ left. MNI coordinates are marked. Brain background is the MNI template. WM skeleton is represented in green. Results are superimposed in red and printed in bold to improve readability. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

AD reductions in the left ILF, right cortico-spinal tract, bilateral cerebellum and left internal capsule and RD reductions in the left internal capsule. Only BD-I had a further RD reduction with respect to HC in the left ILF, right cortico-spinal tract and bilateral cerebellum. Our results are partially in line with those of previous studies that investigated BD-I, BD-II and HC. With respect to BD-I, BD-II showed RD increase in the left ILF as well as in the superior longitudinal fasciculus, internal capsule and in the corona radiata without any significant FA difference between subgroups (Maller et al., 2014). Conversely, other studies found that both BD subgroups presented FA reductions with respect to HC in the WM of the corpus callosum, cingulum and right inferior frontal cortex (Ha et al., 2011; Liu et al., 2010). Looking at the direct comparison between BD-I and BD-II, our results are in contrast with those of Ha and colleagues which indicated that BD-II had higher FA than BD-I in the right temporal WM (Ha et al., 2011) and with those presented by Liu and coworkers which found that BD-II patients had lower FA than BD-I in the right precuneus, frontal and prefrontal WM (Liu et al., 2010). In our study, the FA reduction in the right ILF was present only in patients with BD-II and the concomitant RD increase (at the uncorrected statistical level) suggests a damage in the continuity of the myelin sheaths restricted to this WM tract (Song et al., 2002). The ILF is known as a long association fiber bundle that connects the occipital cortex to the ventral temporal cortex and the posterior para-hippocampal gyrus. Furthermore, the ILF anteriorly joins the uncinate fasciculus to convey information to the orbito-frontal cortex. It plays a role in the visual ventral stream involved in object recognition, discrimination and memory (Schmahmann et al., 2007). In the ILF, FA was found to be bilaterally reduced in mood stabilizer/antipsychotic-naive BD-II and BD not-otherwise-specified patients (Yip et al., 2013) as well as in those with BD-I at the first episode (Lu et al., 2011), bilaterally increased in depressed BD-I (Poletti et al., 2015) or unchanged in euthymic BD-I (Torgerson et al., 2013). Although FA reduction in the left ILF was noted in the early phase of major depressive disorders (Bessette et al., 2014) and in depressed patients suffering from neurological disorders (Huang et al., 2014), decreased FA in the right ILF has been identified as one of the most consistent changes involved in the pathophysiology of depression (Liao et al., 2013). Thus, dysregulated myelination of the ILF may play a role in depression mechanisms of patients suffering from BD. This is consistent with the recent finding that demyelination processes are present at the onset of depression in animal models (Hemanth Kumar et al., 2014). Since depressive symptoms have been found to be predominant during the course of BD-II (Judd et al., 2003), in line with the higher HAM-D score present in our BD-II sample, changes in FA of the right ILF may be an indicator of the

predominant depressive polarity of this subtype. We found simultaneous reductions of AD and RD in the internal capsule of both BD subtypes. This may be due to low water diffusivity and may be associated with the axonal swelling formation after axonal injury (Liu et al., 2013). The swelling is related to the shift of water from extracellular to intracellular space. These results are in line with the findings showing that BD patients present a higher concentration of neurofilament light chain, a neurofilament subunit that reflects the degradation of the axonal membrane (Jakobsson et al., 2014) in the cerebrospinal fluid. The internal capsule is a major WM tract divided into an anterior limb, which mostly contains the thalamocortical radiations and projection fibers of the corona radiata, and a posterior limb, which conveys descending fibers from the premotor and motor cortices. Abnormalities of the internal capsule in the direction of higher RD and unchanged AD were found at the onset of BD (Lu et al., 2012, 2011), and reduced FA was encountered during the chronic course (Bauer et al., 2015; Sussmann et al., 2009). Although in contrast with our results, reduced FA in the internal capsule has been indicated as a trait and a vulnerability marker of BD as it was also present in unaffected BD relatives (Sprooten et al., 2011). The selective reduction of AD in the right cortico-spinal tract and cerebellum found in our BD patients can be interpreted in terms of the diameter of axons (Beaulieu, 2002). A selective reduction of AD may be related to axonal injury (Boretius et al., 2012) and can occur in the presence of high cellularity associated with inflammation (Wang et al., 2014). Although our cross-sectional study cannot establish the causality, one possible interpretation is that the axonal microstructural damage might be related with inflammation (Munkholm et al., 2013; Wang et al., 2014). Elevated levels of different cytokines have already been detected in BD, supporting the hypothesis that chronic low-grade inflammation has a key role in the disease (Munkholm et al. 2013). Our results on the cortico-spinal tract and cerebellum are not in line with previous findings. In fact, Mahon et al. (2009) found lower RD and higher FA whereas others detected higher RD and lower FA (Bauer et al., 2015) in the cortico-spinal tract of BD. Despite structural (Redlich et al., 2014) and functional (Liang et al., 2013) studies reported cerebellar abnormalities in BD, diffusionrelated data on the cerebellum appear lacking and conflicting. Increased RD in the left cerebellum was found in BD (Mahon et al., 2009), while unaltered cerebellar fibers were detected in euthymic patients (Houenou et al., 2007). However, the above mentioned studies enrolled heterogeneous samples in which BD-II patients were only a small minority and they included patients with high lifetime numbers of mood episodes (Bauer et al., 2015) or a history of drug abuse (Mahon et al., 2009). Discrepancies between our results and those of other studies

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

181

Fig. 2. Reduction of AD and RD in bipolar patients. Top panel: 2D and 3D representations of the cluster within the left internal capsule where both BD-I and BD-II have reduced AD and RD with respect to HC. R ¼ right and L¼ left. MNI coordinates are marked. Brain background is the MNI template. WM skeleton is represented in green. Results are superimposed in red and are printed in bold to improve readability. Bottom panel: schematic view of the effect of AD and RD reduction on the diffusion tensor, represented as an ellipsoid. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

could be due to some methodological differences (i.e. voxel-wise approach vs. tractography). Although the advantages of tractography include overcoming alignment issues by operating in the individual space of subjects and higher sensitivity in detecting subtle WM microstructural changes, TBSS has been designed to bring together the strengths of both tractography and classic VBMstyle approaches. In fact, it is fully automated, it solves the alignment and smoothing problems and investigates the whole WM tissue without the need to pre-specify the tracts of interest (Smith et al., 2006). The present findings have to be interpreted in light of some limitations. First, we included a sample with different mood states at the time of evaluation. Although the greater depression severity of our BD-II subgroup did not impact our results, future studies on larger samples should deeply address the effect of mood variables

and phases on microstructural indexes. Second, excluding participants with WMH might decrease the heterogeneity of the population. However, the presence of WMH could influence the results of DTI studies (Iverson et al., 2011; Lange et al., 2014) and could be related to the comorbid cardiovascular and metabolic disease in BD sample (Cardoso de Almeida and Phillips, 2013). Third, psychotropic medications and other clinical characteristics may influence WM microstructure. Nevertheless, due to the severity of the disorder, clinicians have to keep patients under pharmacological treatment to prevent relapses. Thus, for ethical and clinical reasons it is unrealistic to enrol drug-free patients. Moreover, our BD subgroups did not differ for duration of illness, number of past manic/hypomanic or depressive episodes and current pharmacotherapy. This minimizes the risk of confounding factors.

182

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

Further studies are needed to clarify the involvement and eventual role of the different WM tracts in BD and its subtypes. Although common structural changes (Goodkin et al., 2015) and genetic variants (Braskie et al., 2013) have been identified across psychiatric disorders, enrolling clinically pure populations may help to elucidate different underlying etiologies. Particularly, such studies should be performed investigating the directional diffusivity-derived parameters. A point worth noticing is that, as suggested by our results, WM abnormalities can be detected even where FA is unchanged, as a concomitant change of RD and AD in the same direction does not affect the FA value. Hence, it would be incorrect to take into account only FA to interpret WM integrity. In conclusion, we found that BD-II differs from BD-I in a localized region belonging to the right ILF while both BD-I and BD-II patients have widespread WM changes in the right cortico-spinal tract, left internal capsule and bilateral cerebellum in comparison to HC. Our data suggest that the loss of WM integrity in BD-II might be due to demyelination, whereas WM changes common to both BD subgroups could be attributable to axonal damage. Further longitudinal DTI studies focused on the different phases of the illness in balanced samples of BD-I and BD-II are essential to determine the association between the clinical course of the illness and its neurobiological underpinning.

Refrences Ambrosi, E., Rossi-Espagnet, M.C., Kotzalidis, G.D., Comparelli, A., Del Casale, A., Carducci, F., Romano, A., Manfredi, G., Tatarelli, R., Bozzao, A., Girardi, P., 2013. Structural brain alterations in bipolar disorder II: a combined voxel-based morphometry (VBM) and diffusion tensor imaging (DTI) study. J. Affect. Disord. 150, 610–615. American Psychiatric Association, 2000. Diagnostic and statistical manual of mental disorders, Diagnostic and Statistical Manual of Mental Disorders Foruth Edition-Text Revision (DSM-IV-TR), Washington. Andersson, J.L.R., Jenkinson, M., Smith, S., 2007a. Non-Linear Optimisation. Oxford, United Kingdom, FMRIB Technial Report TR07JA1. Andersson, J.L.R., Jenkinson, M., Smith, S., 2007b. Non-Linear Registration aka Spatial Normalisation. Oxford, United Kingdom, FMRIB Technial Report TR07JA2. Bauer, I.E., Ouyang, A., Mwangi, B., Sanches, M., Zunta-Soares, G.B., Keefe, R.S.E., Huang, H., Soares, J.C., 2015. Reduced white matter integrity and verbal fluency impairment in young adults with bipolar disorder: a diffusion tensor imaging study. J. Psychiatr. Res. 62, 115–122. Beaulieu, C., 2002. The basis of anisotropic water diffusion in the nervous system – a technical review. NMR Biomed. 15, 435–455. Benedetti, F., Absinta, M., Rocca, M.A., Radaelli, D., Poletti, S., Bernasconi, A., Dallaspezia, S., Pagani, E., Falini, A., Copetti, M., Colombo, C., Comi, G., Smeraldi, E., Filippi, M., 2011. Tract-specific white matter structural disruption in patients with bipolar disorder. Bipolar Disord. 13, 414–424. Bessette, K.L., Nave, A.M., Caprihan, A., Stevens, M.C., 2014. White matter abnormalities in adolescents with major depressive disorder. Brain Imaging Behav. 4, 531–541. Beyer, J.L., Taylor, W.D., MacFall, J.R., Kuchibhatla, M., Payne, M.E., Provenzale, J.M., Cassidy, F., Krishnan, K.R.R., 2005. Cortical white matter microstructural abnormalities in bipolar disorder. Neuropsychopharmacology 30, 2225–2229. Bora, E., Yücel, M., Pantelis, C., Berk, M., 2011. Meta-analytic review of neurocognition in bipolar II disorder. Acta Psychiatr. Scand. 123, 165–174. Boretius, S., Escher, A., Dallenga, T., Wrzos, C., Tammer, R., Brück, W., Nessler, S., Frahm, J., Stadelmann, C., 2012. Assessment of lesion pathology in a new animal model of MS by multiparametric MRI and DTI. Neuroimage 59, 2678–2688. Braskie, M.N., Kohannim, O., Jahanshad, N., Chiang, M.C., Barysheva, M., Toga, A.W., Ringman, J.M., Montgomery, G.W., McMahon, K.L., de Zubicaray, G.I., Martin, N. G., Wright, M.J., Thompson, P.M., 2013. Relation between variants in the neurotrophin receptor gene, NTRK3, and white matter integrity in healthy young adults. Neuroimage 82, 146–153. Bruno, S., Cercignani, M., Ma, R., Bruno, S., Cercignani, M., Ron, M.A., 2008. White matter abnormalities in bipolar disorder : a voxel-based diffusion tensor imaging study. Bipolar Disord. 10, 460–468. Caseras, X., Lawrence, N.S., Murphy, K., Wise, R.G., Phillips, M.L., 2013. Ventral striatum activity in response to reward: differences between bipolar I and II disorder. Am. J. Psychiatry 170, 533–541. Cardoso de Almeida, J.R., Phillips, M.L., 2013. Distinguishing between unipolar depression and bipolar depression: current and future clinical and neuroimaging perspectives. Biol. Psychiatry 73, 111–118. Cherubini, A., Péran, P., Caltagirone, C., Sabatini, U., Spalletta, G., 2009. Aging of subcortical nuclei: microstructural, mineralization and atrophy modifications

measured in vivo using MRI. Neuroimage 48, 29–36. Fiedorowicz, J.G., Endicott, J., Leon, A.C., Solomon, D.A., Keller, M.B., Coryell, W.H., 2011. Subthreshold hypomanic symptoms in progression from unipolar major depression to bipolar disorder. Am. J. Psychiatry 168, 40–48. First, M.B., Gibbon, M., Spitzer, R.L., Williams, J.B., Benjamin, L., 1997. Structured Clinical Interview for DSM-IV Axis II Personality Disorders, (SCID-II). American Psychiatric Press, Inc., Washington DC. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B., 2002a. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P). State Psychiatric Institute, New York, New York: Biometrics Research. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B., 2002b. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Non-patient Edition, (SCID-I/NP). State Psychiatric Institute, New York Biometrics Research, New York. Folstein, M.F., Folstein, S.E., McHugh, P.R., 1975. Mini-mental state. J. Psychiatr. Res. 12, 189–198. Goodkind, M., Eickhoff, S.B., Oathes, D.J., Jiang, Y., Chang, A., Jones-Hagata, L.B., Ortega, B.N., Zaiko, Y.V., Roach, E.L., Korgaonkar, M.S., Grieve, S.M., GalatzerLevy, I., Fox, P.T., Etkin, A., 2015. Identification of a common neurobiological substrate for mental illness. JAMA Psychiatry 72, 305–315. Ha, T.H., Her, J.Y., Kim, J.H., Chang, J.S., Cho, H.S., Ha, K., 2011. Similarities and differences of white matter connectivity and water diffusivity in bipolar I and II disorder. Neurosci. Lett. 505, 150–154. Haller, S., Xekardaki, A., Delaloye, C., Canuto, A., Lövblad, K.O., Gold, G., Giannakopoulos, P., 2011. Combined analysis of grey matter voxel-based morphometry and white matter tract-based spatial statistics in late-life bipolar disorder. J. Psychiatry Neurosci. 36, 391–401. Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56–62. Hemanth Kumar, B.S., Mishra, S.K., Trivedi, R., Singh, S., Rana, P., Khushu, S., 2014. Demyelinating evidences in CMS rat model of depression: a DTI study at 7T. Neuroscience 275, 12–21. Houenou, J., Wessa, M., Douaud, G., Leboyer, M., Chanraud, S., Perrin, M., Poupon, C., Martinot, J.L., Paillere-Martinot, M.L., 2007. Increased white matter connectivity in euthymic bipolar patients: diffusion tensor tractography between the subgenual cingulate and the amygdalo-hippocampal complex. Mol. Psychiatry 12, 1001–1010. Huang, P., Xu, X., Gu, Q., Xuan, M., Yu, X., Luo, W., Zhang, M., 2014. Disrupted white matter integrity in depressed versus non-depressed Parkinson's disease patients: a tract-based spatial statistics study. J. Neurol. Sci. 346, 145–148. Iorio, M., Spalletta, G., Chiapponi, C., Luccichenti, G., Cacciari, C., Orfei, M.D., Caltagirone, C., Piras, F., 2013. White matter hyperintensities segmentation: a new semi-automated method. Front. Aging Neurosci. 5, 76. Iverson, G.L., Hakulinen, U., Wäljas, M., Dastidar, P., Lange, R.T., Soimakallio, S., Öhman, J., 2011. To exclude or not to exclude: white matter hyperintensities in diffusion tensor imaging research. Brain Inj. 25, 1325–1332. Jakobsson, J., Bjerke, M., Ekman, C.J., Sellgren, C., Johansson, A.G., Zetterberg, H., Blennow, K., Landén, M., 2014. Elevated concentrations of neurofilament light chain in the cerebrospinal fluid of bipolar disorder patients. Neuropsychopharmacology 39, 2349–2356. Judd, L.L., Akiskal, H.S., Schettler, P.J., Coryell, W., Maser, J., Rice, J.A., Solomon, D.A., Keller, M.B., 2003. The comparative clinical phenotype and long term longitudinal episode course of bipolar I and II: a clinical spectrum or distinct disorders? J. Affect. Disord. 73, 19–32. Judd, L.L., Akiskal, H.S., Schettler, P.J., Endicott, J., Leon, A.C., Solomon, D.A., Coryell, W., Maser, J.D., Keller, M.B., 2005. Psychosocial disability in the course of bipolar I and II isorders:b a prospective, comparative, longitudinal study. Arch. Gen. Psychiatry 62, 1322–1330. Lange, R.T., Shewchuk, J.R., Heran, M.K., Rauscher, A., Jarrett, M., Brubacher, J.R., Iverson, G.L., 2014. To exclude or not to exclude: further examination of the influence of white matter hyperintensities in diffusion tensor imaging research. J. Neurotrauma 31, 198–205. Liao, Y., Huang, X., Wu, Q., Yang, C., Kuang, W., Du, M., Lui, S., Yue, Q., Chan, R.C., Kemp, G.J., Gong, Q., 2013. Is depression a disconnection syndrome? Metaanalysis of diffusion tensor imaging studies in patients with MDD. J. Psychiatry Neurosci. 38, 49–56. Liang, M.J., Zhou, Q., Yang, K.R., Yang, X.L., Fang, J., Chen, W.L., Huang, Z., 2013. Identify changes of brain regional homogeneity in bipolar disorder and unipolar depression using resting-state FMRI. PLoS One 8, 1–8. Liu, J.-X., Chen, Y.-S., Hsieh, J.-C., Su, T.-P., Yeh, T.-C., Chen, L.-F., 2010. Differences in white matter abnormalities between Bipolar I and II Disorders. J. Affect. Disord. 127, 309–315. Liu, M., Gross, D.W., Wheatley, B.M., Concha, L., Beaulieu, C., 2013. The acute phase of Wallerian degeneration: longitudinal diffusion tensor imaging of the fornix following temporal lobe surgery. Neuroimage 74, 128–139. Lu, L.H., Zhou, X.J., Fitzgerald, J., Keedy, S.K., Reilly, J.L., Passarotti, A.M., Sweeney, J. A., Pavuluri, M., 2012. Microstructural abnormalities of white matter differentiate pediatric and adult-onset bipolar disorder. Bipolar Disord. 14, 597–606. Lu, L.H., Zhou, X.J., Keedy, S.K., Reilly, J.L., Sweeney, J.A., 2011. White matter microstructure in untreated first episode bipolar disorder with psychosis: comparison with schizophrenia. Bipolar Disord. 13, 604–613. Mahon, K., Wu, J., Malhotra, A.K., Burdick, K.E., DeRosse, P., Ardekani, B.A., Szeszko, P.R., 2009. A voxel-based diffusion tensor imaging study of white matter in bipolar disorder. Neuropsychopharmacology 34, 1590–1600. Maller, J.J., Thaveenthiran, P., Thomson, R.H., McQueen, S., Fitzgerald, P.B., 2014.

E. Ambrosi et al. / Journal of Affective Disorders 189 (2016) 176–183

Volumetric, cortical thickness and white matter integrity alterations in bipolar disorder type I and II. J. Affect. Disord. 169, 118–127. Measso, G., Cavarzeran, F., Zappalà, G., Lebowitz, B.D., Crook, T.H., Pirozzolo, F.J., Amaducci, L.A., Massari, D., Grigoletto, F., 1993. The mini-mental state examination: normative study of an Italian random sample. Dev. Neuropsychol. 9, 77–85. Munkholm, K., Braüner, J.V., Kessing, L.V., Vinberg, M., 2013. Cytokines in bipolar disorder vs. healthy control subjects: a systematic review and meta-analysis. J. Psychiatr. Res. 47, 1119–1133. Poletti, S., Bollettini, I., Mazza, E., Locatelli, C., Radaelli, D., Vai, B., Smeraldi, E., Colombo, C., Benedetti, F., 2015. Cognitive performances associate with measures of white matter integrity in bipolar disorder. J. Affect. Disord. 174, 342–352. Redlich, R., Almeida, J.J., Grotegerd, D., Opel, N., Kugel, H., Heindel, W., Arolt, V., Phillips, M.L., Dannlowski, U., 2014. Brain morphometric biomarkers distinguishing unipolar and bipolar depression. A voxel-based morphometry-pattern classification approach. JAMA Psychiatry 71, 1222–1230. Rueckert, D., Sonoda, L.I., Hayes, C., Hill, D.L., Leach, M.O., Hawkes, D.J., 1999. Nonrigid registration using free-form deformations: application to breast MR images. IEEE Trans. Med. Imaging 18, 712–721. Schmahmann, J.D., Pandya, D.N., Wang, R., Dai, G., D’Arceuil, H.E., De Crespigny, A.J., Wedeen, V.J., 2007. Association fibre pathways of the brain: parallel observations from diffusion spectrum imaging and autoradiography. Brain 130, 630–653. 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. Smith, S.M., Nichols, T.E., 2009. Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference. Neuroimage 44, 83–98. Song, S.K., Sun, S.W., Ramsbottom, M.J., Chang, C., Russell, J., Cross, A.H., 2002. Dysmyelination revealed through MRI as increased radial (but unchanged axial)

183

diffusion of water. Neuroimage 17, 1429–1436. Sprooten, E., Sussmann, J.E., Clugston, A., Peel, A., McKirdy, J., Moorhead, T.W.J., Anderson, S., Shand, A.J., Giles, S., Bastin, M.E., Hall, J., Johnstone, E.C., Lawrie, S. M., McIntosh, A.M., 2011. White matter integrity in individuals at high genetic risk of bipolar disorder. Biol. Psychiatry 70, 350–356. Sussmann, J.E., Lymer, G.K.S., Mckirdy, J., Moorhead, T.W.J., Maniega, S.M., Job, D., Hall, J., Bastin, M.E., Johnstone, E.C., Lawrie, S.M., Mcintosh, A.M., 2009. White matter abnormalities in bipolar disorder and schizophrenia detected using diffusion tensor magnetic resonance imaging. Bipolar Disord. 11, 11–18. Torgerson, C.M., Irimia, A., Leow, A.D., Bartzokis, G., Moody, T.D., Jennings, R.G., Alger, J.R., Van Horn, J.D., Altshuler, L.L., 2013. DTI tractography and white matter fiber tract characteristics in euthymic bipolar I patients and healthy control subjects. Brain Imaging Behav. 7, 129–139. Vederine, F.-E., Wessa, M., Leboyer, M., Houenou, J., 2011. A meta-analysis of wholebrain diffusion tensor imaging studies in bipolar disorder. Prog. Neuropsychopharmacol. Biol. Psychiatry 35, 1820–1826. Vieta, E., Gastó, C., Otero, A., Nieto, E., Vallejo, J., 1997. Differential features between Bipolar I and bipolar II disorder. Compr. Psychiatry 38, 98–101. Vita, A., De Peri, L., Sacchetti, E., 2009. Gray matter, white matter, brain, and intracranial volumes in first-episode bipolar disorder: a meta-analysis of magnetic resonance imaging studies. Bipolar Disord. 11, 807–814. Wang, X., Cusick, M.F., Wang, Y., Sun, P., Libbey, J.E., Trinkaus, K., Fujinami, R.S., Song, S.K., 2014. Diffusion basis spectrum imaging detects and distinguishes coexisting subclinical inflammation, demyelination and axonal injury in experimental autoimmune encephalomyelitis mice. NMR Biomed. 27, 843–852. Wessa, M., Houenou, J., Leboyer, M., Chanraud, S., Poupon, C., 2009. Microstructural white matter changes in euthymic bipolar patients : a whole-brain diffusion tensor imaging study. Bipolar Disord. 11, 504–514. Yip, S.W., Chandler, R.A., Rogers, R.D., Mackay, C.E., Goodwin, G.M., 2013. White matter alterations in antipsychotic- and mood stabilizer-naïve individuals with bipolar II/NOS disorder. Neuroimage Clin. 3, 271–278. Young, R.C., Biggs, J.T., Ziegler, V.E., Meyer, D.A., 1978. A rating scale for mania: reliability, validity and sensitivity. Br. J. Psychiatry 133, 429–435.