Integrity of hypothalamic fibers and cognitive fatigue in multiple sclerosis

Multiple Sclerosis and Related Disorders (2015) 4, 39–46

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/msard

Integrity of hypothalamic fibers and cognitive fatigue in multiple sclerosis Katrin Hankena, Paul Elingb, Andreas Kastrupa, Jan Kleinc, Helmut Hildebrandta,d,n a

Klinikum Bremen-Ost, Department of Neurology, Züricher Str. 40, 28325 Bremen, Germany Radboud University Nijmegen, Donders Institute for Brain, Cognition and Behaviour, Nijmegen, the Netherlands c Fraunhofer-MeVis Institute, University of Bremen, Germany d University of Oldenburg, Institute of Psychology, 26111 Oldenburg, Germany b

Received 4 August 2014; received in revised form 31 October 2014; accepted 13 November 2014

KEYWORDS

Abstract

Multiple sclerosis; Fatigue; Posterior hypothalamus; Histamine; Brainstem; Diffusion tensor imaging

Cognitive fatigue is a common and disabling symptom of multiple sclerosis (MS), but little is known about its pathophysiology. The present study investigated whether the posterior hypothalamus, which is considered as the waking center, is associated with MS-related cognitive fatigue. We analyzed the integrity of posterior hypothalamic fibers in 49 patients with relapsing-remitting MS and 14 healthy controls. Diffusion tensor imaging (DTI) parameters were calculated for fibers between the posterior hypothalamus and, respectively, the mesencephalon, pons and prefrontal cortex. In addition, DTI parameters were computed for fibers between the anterior hypothalamus and these regions and for the corpus callosum. Cognitive fatigue was assessed using the Fatigue Scale for Motor and Cognitive Functions. Analyses of variance with repeated measures were performed to investigate the impact of cognitive fatigue on diffusion parameters. Cognitively fatigued patients (75.5%) showed a significantly lower mean axial and radial diffusivity for fibers between the posterior hypothalamus and the mesencephalon than cognitively non-fatigued patients (GroupnTarget areanDiffusion orientation: F=4.047; p=0.023). For fibers of the corpus callosum, MS patients presented significantly higher axial and radial diffusivity than healthy controls (GroupnDiffusion orientation: F =9.904; po0.001). Depressive mood, used as covariate, revealed significant interaction effects for anterior hypothalamic fibers (Target areanDiffusion orientationnDepression: F =5.882; p =0.021; HemispherenDiffusion orientationn Depression: F =8.744; p =0.008). Changes in integrity of fibers between the posterior hypothalamus and the mesencephalon appear to be associated with MS-related cognitive fatigue. These changes might cause an

n

Corresponding author. Tel./fax: +49 4214081599. E-mail address: [email protected] (H. Hildebrandt).

http://dx.doi.org/10.1016/j.msard.2014.11.006 2211-0348/& 2014 Elsevier B.V. All rights reserved.

40

K. Hanken et al. altered modulation of hypothalamic centers responsible for wakefulness. Furthermore, integrity of anterior hypothalamic fibers might be related to depression in MS. & 2014 Elsevier B.V. All rights reserved.

1.

Introduction

In multiple sclerosis (MS) patients, fatigue is rated as one of the most common and most disabling symptoms. Its prevalence ranges from 65% to 97% and it tends to seriously impair approximately one third of all MS patients (Bakshi et al., 2000; Fisk et al., 1994; Ford et al., 1998; Krupp et al., 1988; Murray, 1985; van der Werf et al., 1998). MS-related fatigue seems to be multifactorial in nature, including both cognitive and physical components (Krupp et al., 1988). Even though it has been subject of numerous studies (Bakshi, 2003; Kos et al., 2008; Schwid et al., 2002), the exact pathology of MSrelated fatigue remains poorly understood. The hypothalamus is recognized as a key center for the regulation of wakefulness and sleep (Lin, 2000). Reciprocal interactions between anterior and posterior hypothalamic areas constitute one of the important hypothalamic mechanisms underlying alternation of sleep and wakefulness. The anterior hypothalamus is regarded to be the sleep center with GABAergic neurons, located in the ventrolateral preoptic area of the anterior hypothalamus, initiating and maintaining sleep, whereas the posterior hypothalamus is considered to be the wake center of the brain. Histaminergic neurons located in the tuberomammillary nucleus (TMN) of the posterior hypothalamus constitute one of the ascending activating systems and their exclusive role for wakefulness and arousal has been demonstrated in animal and clinical studies (Gillson et al., 2002; Lin and Jouvet, 1988; Lin et al., 1989; Lin, 2000; Monti et al., 1986; Parmentier et al., 2002; Welch et al., 2002). Mechanisms of histaminergic arousal involve ascending and descending projections of histaminergic neurons and their interaction with diverse neuronal populations such as wake-promoting regions of the brainstem, including the dorsal raphe and central superior nuclei in the mesencephalon and the locus coeruleus located in the pons (Haas and Panula, 2003; Thakkar, 2011). In general, the hypothalamus contains projections to various brain regions and hypothalamic impairment has been related to cognitive disturbances (Copenhaver et al., 2006). Especially the histaminergic system with its projections to almost all regions of the central nervous system (CNS) has been found to play a major role in the regulation of cognition (Alvarez, 2009). Despite its relevance for the regulation of wakefulness, arousal and cognition, there are hardly any studies investigating the role of hypothalamic changes in relation to fatigue in MS patients. Huitinga et al. (2004) found hypothalamic lesions in 15 out of 16 MS patients. Moreover, this group identified demyelinating lesions in and adjacent to the hypothalamus in 95% of MS patients (Huitinga et al., 2001). Recently, Zellini et al. (2009) used magnetic resonance imaging (MRI) T1 relaxation time as a sensitive metric for the detection of pathological changes in the hypothalamus of 44 relapsing-remitting MS patients. They found a significant correlation between T1 relaxation time in the hypothalamus and the Fatigue Severity Scale (FSS) score, indicating an

association between pathological changes in the hypothalamus and MS-related fatigue. Finally, inflammation-induced suppression of histaminergic neurons was found to be associated with impaired arousal (Gaykema et al., 2008). These findings suggest that the hypothalamus, especially the histaminergic system located in the TMN of the posterior hypothalamus, is associated with the underlying mechanism responsible for MS-related cognitive fatigue. Therefore, the aim of this study was to investigate the relationship between MS-related cognitive fatigue and the integrity of fibers originating or terminating in the posterior hypothalamus using diffusion tensor imaging (DTI). Axial diffusivity (AD) and radial diffusivity (RD) strength per voxel may enable us to distinguish between axonal loss and demyelination, the two major aspects of MS pathology (Budde et al., 2007; Fink et al., 2010). Animal studies have demonstrated an increase in RD shortly after inflammation and a decrease in AD that correlated with axonal loss (Budde et al., 2007). Therefore, we used these parameters as measures of fiber integrity. We expected cognitively fatigued MS patients to show an increased RD and a decreased AD for posterior hypothalamic fibers when compared with cognitively non-fatigued MS patients and healthy controls. To demonstrate the unique role of the posterior hypothalamus in the underlying mechanisms for MS-related cognitive fatigue, we also investigated fibers of the anterior hypothalamus and the anterior and posterior parts of the corpus callosum. If only the posterior hypothalamus is associated with cognitive fatigue, integrity of anterior hypothalamic fibers and the corpus callosum should not differ between cognitively fatigued and non-fatigued MS patients. To corroborate our findings, we replicated these analyses using an earlier collected data set of our group (see Fink et al. (2010)).

2. 2.1.

Methods Study population

A total of 49 MS patients with a relapsing-remitting disease course according to the McDonald criteria (Polman et al., 2011) and 14 age- and gender-matched healthy controls participated in this study. Patients were recruited from MS support groups in Bremen and surroundings or have been patients of the Department of Neurology of the Klinikum Bremen-Ost, Germany. Patients received either immunomodulatory therapy (60%) or no disease modifying drugs (40%). Individuals with an MS relapse or using corticosteroids during the last four weeks, under legal care and/or with a diagnosis of any neuropsychiatric illness according to the fourth edition of the diagnostic and statistical manual of mental disorders (American Psychiatric Association, 2000) were excluded from the study. The study was approved by

Integrity of hypothalamic fibers and cognitive fatigue in multiple sclerosis the ethical board of the Physicians Society Bremen and written informed consent was obtained from participants.

2.2.

Clinical and psychological assessment

Fatigue was assessed using the Fatigue Scale for Motor and Cognitive Functions (FSMC; Penner et al., 2009). This selfreported questionnaire, composed of 20 items, evaluates two main components of fatigue, namely motor and cognitive fatigue. The cut-off score between normal and pathological fatigue is 43 for the total scale and 22 for the cognitive and motor scale, with 22 being the cut-off score for mild, 28 for moderate and 34 for severe cognitive fatigue. Emotional distress was investigated using the Beck's Depression Inventory Scale (BDI; Beck et al., 1988), a selfrated multiple choice questionnaire. It can be divided into a psychological (items A-N) and a somatic (items O-U) component. Somatic items also concern sleepiness and sleep problems. Hence, we focused on the psychological items to correct for depressive mood in the statistical evaluation (see below). The BDI component P, which reflects sleep problems, was used as a measure for possible sleep disturbances. All patients were assessed by experienced neurologists and MS disability has been evaluated using the Expanded Disability Status Scale (EDSS; Kurtzke, 1983).

2.3.

Image acquisition and image processing

MR images of the brain were obtained using a 3T scanner (Siemens Verio, Erlangen, Germany). Patients were placed in a supine position and a circular-polarized array head coil was used with a 2-dimensional DTI echo planar imaging and 30 diffusion directions (b-factor 0 and 1000 mm 2 s). Sequence parameters were as follows: repetition time (TR) 10700 msec, echo time (TE) 84 msec, field of view (FOV) 207 mm, voxel size 1.80  1.80  1.98 mm3, 72 slices, number of excitations (NEX) 2, scanning time  10 min. Autoshimming and phase correction were activated. In addition, 176 sagittally oriented 3-dimensional MPRAGE T1-weighted images were obtained. Images were acquired with a 256  256 matrix over a 256 mm of FOV and reconstructed with a rectangular FOV of 85%. A Gaussian smoothing was performed on diffusion-weighted images using an infinite impulse algorithm with s=1. Subsequently, DTI data were resampled to an isotropic image resolution of 1.5 mm3 using a cubic B-spline filter in order to improve fiber tracking results. For coregistration of the T1 and the diffusion MRI data, we used a linear registration algorithm which performs an affine transformation. As similarity measure, we used normalized mutual information with an epsilon of 10 7.

2.4.

DTI-based tractography and quantification

Image analysis was performed using the NeuroQLab3.531 software package (Fraunhofer MEVIS, Bremen, Germany; Weiler et al., 2009). The fiber tracking algorithm we employed (see Schlüter et al., 2005) is based on the deflection-based approach by Weinstein et al. (1999). The anatomical accuracy of the DTI fiber tracking method of NeuroQLab appeared to be quite high (Feigl et al., 2013). Fiber tracking of the hypothalamus was performed separately for each brain hemisphere and was started from two different seed regions of interest (ROI). ROIs

41

were defined manually on coregistered images. Two coronally oriented slices were used to track the fibers of the posterior (see Supplementary data 1A and 1B) and anterior (see Supplementary data 2A and 2B) hypothalamus. The posterior seed ROI was defined by the mammillary body in its most ventral extension, the wall of the third ventricle in the medial direction, the thalamus in the dorsal direction and the motor pathway of the internal capsule in the lateral extension. Major hypothalamic nuclei targeted by this seed ROI are the posterior hypothalamic nucleus, the lateral hypothalamic area and the nuclei of the mammillary bodies (see Supplementary data 1A). For the anterior seed ROI, the most rostral slice was chosen, showing the anterior commissure and anterior parts of the pituitary and posterior parts of the optic chiasm. This seed ROI was located just below the anterior commissure, encompassing a section through the preoptic nucleus (lateral and medial), the supraoptic nucleus and the lamina terminalis. The nucleus gyri diagonalis has been used as lateral border (see Supplementary data 2A). For quantifying fibers between the hypothalamus and the mesencephalon, the pons and the prefrontal cortex, two crop ROIs defining a start and an end of the section of measurements were determined respectively. The first crop ROI defining the hypothalamic starting point was always located directly around the respective seed ROI of the anterior or posterior hypothalamus. For the quantification of fibers between the posterior hypothalamus and the mesencephalon, we used an axial slice directly ventral to the inferior colliculi for defining the crop ROI. To analyze fibers between the posterior hypothalamic seed ROI and the pons, the crop ROI was defined in an axial oriented slice at the level of the posterior recessus of the fourth ventricle. These two fiber bundles include fibers of the lower part of the medial forebrain bundle including serotonergic and noradrenergic neurons of wake-promoting brainstem areas. For quantifying fibers between the posterior hypothalamus and the prefrontal cortex, including fibers of the upper part of the medial forebrain bundle, we identified the tip of the most anterior extension of the cingulate cortex from a sagittal viewpoint, and then on a coronal-oriented slice the second crop ROI was defined. The same crop ROIs were defined for quantifying fibers of the anterior hypothalamic seed ROI. The seed ROI for the corpus callosum has been defined in a sagittally oriented slice, around the red-coded region located above the lateral ventricles. Two parts of the corpus callosum were tracked separately, namely the frontal part (from the frontal tip to vertex) and the posterior part (from vertex to back tip; see Supplementary data 3A and 3B). For the quantification of each part, two crop ROIs were defined in sagittally oriented slices at the points where fibers run caudally to merge with pyramidal tracts. AD and RD scores were used as measures of fiber integrity. The AD and RD along each fiber bundle were determined and mean values were calculated for each participant (for description of the quantification method see Klein et al., 2007).

2.5.

Brain atrophy

Volumes for gray matter (GM), white matter (WM) and intracranial cavity were computed using the software module “Brain Volumetry”. The calculation of the brain parenchymal

42

K. Hanken et al.

fraction (BPF) was performed using the methodology described by Lukas et al. (2004), relying on skull-stripping based on watershed segmentation and histogram analysis. The software module “Ventricle Volumetry” was used to calculate volumes of lateral, third and fourth ventricles.

2.6.

Statistical analysis

Based on the FSMC cut-off score for moderate cognitive fatigue, patients were divided into two groups: cognitively fatigued (cognitive fatigue score Z 28; n =37), and cognitively non-fatigued MS patients (cognitive fatigue score o28; n =12). These two groups were compared to the healthy controls (two-tailed Student's t-tests). Five separate analyses of variance with repeated measures were performed for (I) fibers between the posterior hypothalamus and the brainstem areas (mesencephalon/pons), (II) fibers between the posterior hypothalamus and the prefrontal cortex, (III) fibers between the anterior hypothalamus and the brainstem regions, (IV) fibers between the anterior hypothalamus and the prefrontal cortex and (V) fibers of the corpus callosum. In all analyses, axial and radial diffusivity scores were used as dependent variables and Group (cognitively fatigued and cognitively non-fatigued MS patients, healthy controls) was used as between-subject factor. For the analyses I and III (fibers between hypothalamus and brainstem areas), we used Hemisphere (right/left), Diffusion orientation (axial/radial) and Target area (mesencephalon/pons) as within-subject factors. For the analyses II and IV (fibers between hypothalamus and the prefrontal cortex), Hemisphere (left/right) and Diffusion orientation (axial/radial) were used as within-subject factors. For the analysis V (fibers of the corpus callosum), Diffusion Table 1

3.

Results

3.1.

Patient characteristics

Demographic, clinical and psychological characteristics of the three groups are presented in Table 1. Cognitively fatigued patients showed significantly higher motor fatigue scores (po0.001), BDI scores (po0.001) and a significantly shorter symptom duration (p=0.028) than cognitively non-fatigued patients. There were no significant differences on measures of brain atrophy between these two groups. The number of patients receiving immunomodulatory treatment did not differ between cognitively fatigued and cognitively non-fatigued MS patients. When comparing MS

Demographic, clinical and psychological data of study participants. CF MS patients

Number Male/female Age (years)a Education (years)a Disease duration (month)a Symptom duration (month)a Immunomod. treatment (%) EDSS scorea FSMCtotalscorea FSMCcognitivescorea FSMCmotorscorea BDItotalscorea BDIsomaticscorea BDIpsychologicalscorea BPF (%) Lateral ventricles (ml) Third ventricle (ml) Fourth ventricle (ml) a

orientation (axial/radial) and Fiber section (anterior/posterior) were used as within subject factors. In all analyses, the score on the psychological component of the BDI was used as covariate to control for depressive symptoms. To validate our findings, we replicated the analysis I and V using a data set from an earlier study of our group examining the interrelation between DTI measures and brain atrophy in 53 relapsing-remitting MS patients and 17 healthy controls (for detailed information on the study group and image acquisition, see Fink et al., 2010). In that study, fatigue severity was assessed with the Fatigue Severity Scale (FSS; Krupp et al., 1989). This self-reported instrument assesses severity and frequency of fatigue with higher scores representing stronger fatigue. A cutoff score of 5 was chosen to divide patients into those with (FSSZ5) and without (FSSo5) a moderate level of fatigue. For all statistical analyses, a p-value of less than 0.05 was considered statistically significant.

37 10/27 47.2 (710.7) 11.2 (71.6) 108.9 (7101.6) 145.0 (780.5) 61 3.2 (71.5) 77.1 (79.0) 38.2 (75.8) 38.9 (74.9) 13.4 (78.5) 5.6 (73.2) 7.8 (76.3) 79.1 (73.1) 30.8 (716.4) 1.5 (70.7) 1.8 (70.8)

CNF MS patients

12 2/10 43.8 (710.8) 11.6 (71.6) 170.5 (7113.6) 276.1 (7133.1) 56 4.1 (72.7) 46.4 (711.3) 20.6 (74.3) 25.8 (78.4) 5.7 (73.3) 4.6 (75.1) 2.6 (72.7) 78.4 (75.2) 35.3 (723.9) 1.9 (71.2) 2.1 (71.1)

p-values

Healthy controls

14 4/10 43 (79.6) 12 (71.6)

37.3 (711.9) 18 (75.9) 19.3 (76.3) 4 (74.4) 1.8 (71.9) 2.2 (72.8) 83.9 (71.6) 15.0 (75.3) 0.7 (70.3) 1.7 (70.6)

CF vs. CNF

CF vs. HC

CNF vs. HC

n.a n.s n.s n.s n.s 0.028 n.s n.s o0.001 o0.001 o0.001 o0.001 n.s o0.001 n.s n.s n.s n.s

n.a n.s n.s n.s n.a n.a n.a n.a o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 o0.001 n.s

n.a n.s n.s n.s n.a n.a n.a n.a n.s n.s 0.032 n.s n.s n.s 0.004 0.014 0.005 n.s

Mean (7SD); BDI: Beck's Depression Inventory; BPF: brain parenchymal fraction; CF: cognitively fatigued; CNF: cognitively nonfatigued; EDSS: Expanded Disability Status Scale; FSMC: Fatigue Scale for Motor and Cognitive Functions; HC: healthy controls; Immunomod.: immunomodulatory; MS: multiple sclerosis; n.a: not applicable; n.s: not significant.

Integrity of hypothalamic fibers and cognitive fatigue in multiple sclerosis patients with (BDI P40), to those without a sleep disorder (BDI P=0), no significant difference in their level of cognitive fatigue was found (35.3 vs. 32.3; p=0.331). Compared to healthy controls, MS patients presented a significantly higher degree of brain atrophy. Moreover, cognitively fatigued MS patients showed significantly higher fatigue and depression scores than healthy controls.

3.2.

Posterior hypothalamic fibers

For fibers between the posterior hypothalamus and brainstem areas, a significant interaction effect between Group, Target area and Diffusion orientation was found (F= 4.047; p= 0.023; see Fig. 1). With respect to the fibers between the posterior hypothalamus and the mesencephalon, cognitively fatigued patients showed significantly lower axial (p=0.025) and radial (p=0.033) diffusivity values, averaged over both hemispheres, than cognitively non-fatigued patients (two-tailed Student's t-test). For fibers between the posterior hypothalamus and the frontal cortex, no significant interaction effect with the factor Group was found.

3.3.

Anterior hypothalamic fibers

Concerning the anterior hypothalamic fibers, no significant interaction with Group was observed. Instead, significant interaction effects of the psychological BDI component were found for fibers between the anterior hypothalamus and the two brainstem regions (Target areanDiffusion orientationn Depression: F =5.882; p= 0.021) and for fibers between the anterior hypothalamus and the frontal cortex (HemispherenDiffusion orientationnDepression: F= 8.744; p= 0.008).

Fig. 1 Mean axial and radial diffusivity (averages of both brain hemispheres) for fibers between the posterior hypothalamus and the mesencephalon according to different groups. Covariates appearing in the model are evaluated at a BDIpsychological of 5.657. Error bars reflect standard errors of respective means. This figure shows that cognitively fatigued MS patients present a lower mean of axial and radial diffusivity for fibers between the posterior hypothalamus and the mesencephalon compared to healthy controls and cognitively non-fatigued MS patients. CF: cognitively fatigued multiple sclerosis patients; CNF: cognitively non-fatigued multiple sclerosis patients; HC: healthy controls; npo0.05.

3.4.

43

Fibers of the corpus callosum

The analysis of fibers of the corpus callosum revealed a significant interaction effect between Group and Diffusion orientation (F=9.904; po0.001; see Fig. 2) with healthy controls showing significantly lower means of axial and radial diffusivity than cognitively fatigued and cognitively nonfatigued MS patients (po0.001). Cognitively fatigued and cognitively non-fatigued MS patients did not differ significantly concerning diffusion parameters of these fibers (two-tailed Student's t-test).

3.5.

Comparison to earlier data set

The finding that MS patients without cognitive fatigue and healthy controls show higher axial and radial diffusion values for fibers between the posterior hypothalamus and the mesencephalon than cognitively fatigued MS patients is somewhat unexpected. To check these results, we replicated analysis I on a different set of data, collected in 53 relapsing-remitting MS patients (28 fatigued and 25 non-fatigued MS patients based on the FSS) and 17 healthy controls (see Table 2 and Fink et al., 2010). Even though the fatigue scale used in this study does not allow to differentiate between cognitive and motor fatigue, the analysis showed the same results with respect to overall fatigue. A significant interaction effect between Group (fatigued and non-fatigued MS patients, healthy controls), Target area and Diffusion orientation was found for fibers between the posterior hypothalamus and brainstem areas (F=3.688; p=0.034). Fatigued patients had lower means for axial and radial diffusivity for these fibers than non-fatigued MS patients. Diffusivity values of fatigued MS patients and healthy controls did not differ significantly. We also replicated analysis V on this data set to ensure that fibers of the corpus callosum do not show similar trends in diffusivity. A significant interaction effect between Group and Diffusion orientation was found (F= 6.16; p= 0.004) with

Fig. 2 Mean axial and radial diffusivity for fibers of the corpus callosum according to different groups. Covariates appearing in the model are evaluated at a BDIpsychological of 5.708. Error bars reflect standard errors of respective means. This figure demonstrates that MS patients show increased axial and radial diffusivity for fibers of the corpus callossum compared to healthy controls. CC: corpus callosum; CF: cognitively fatigued multiple sclerosis patients; CNF: cognitively non-fatigued multiple sclerosis patients; HC: healthy controls;npo0.001.

44

K. Hanken et al.

Table 2

Demographic, clinical and psychological data of the earlier study group. F MS patients

Number Male/female Age (years)a Education (years)a Disease duration (month)a Symptom duration (month)a Immunomod. treatment (%) EDSS scorea FSS scorea BDItotalscorea BDIsomaticscorea BDIpsychologicalscorea BPF (%) Lateral ventricles (ml) Third ventricle (ml) Fourth ventricle (ml)

28 4/24 41.9 (76.8) 11.3 (71.5) 88.8 (793.3) 148.2 (7115.5) 82 2.9 (72.0) 54.6 (75.1) 13.3 (77.3) 6.1 (73.0) 7.1 (75.1) 79.3 (73.6) 26.6 (715.5) 2.6 (71.7) 2.1 (70.9)

NF MS patients

25 6/19 44.5 (711.2) 11.0 (71.6) 82.8 (770.1) 155.0 (7107.2) 72 2.8 (71.7) 29.8 (79.3) 6.9 (74.4) 3.2 (71.9) 3.7 (73.4) 78.7 (73.2) 28.2 (716.0) 2.6 (71.1) 2.0 (70.8)

p-values

Healthy controls

17 6/11 37.4 (79.9) 11.3 (71.4)

24.3 (77.8) 5.7 (75.1) 2.6 (72.7) 3.1 (72.8) 81.4 (73.0) 17.3 (76.9) 2.0 (71.0) 1.9 (70.8)

F vs. NF

F vs. HC

NF vs. HC

n.a n.s n.s n.s n.s n.s n.s n.s o0.001 o0.001 o0.001 0.006 n.s n.s n.s n.s

n.a n.s n.s n.s n.a n.a n.a n.a o0.001 o0.001 o0.001 0.002 0.056 0.024 n.s n.s

n.a n.s 0.04 n.s n.a n.a n.a n.a n.s n.s n.s n.s 0.008 0.005 n.s n.s

a Mean (7SD); BDI: Beck's Depression Inventory; BPF: brain parenchymal fraction; EDSS: Expanded Disability Status Scale; F: fatigued; FSS: Fatigue Severity Scale; HC: healthy controls; Immunomod.: immunomodulatory; MS: multiple sclerosis; n.a: not applicable; NF: non-fatigued; n.s: not significant.

MS patients showing higher axial and radial diffusivity than healthy controls. Non-fatigued and fatigued patients did not differ significantly in diffusivity for these fibers.

4.

Discussion

The primary aim of this study was to investigate the association between cognitive fatigue and the integrity of posterior hypothalamic fibers. Our results reveal that cognitively fatigued MS patients have a significantly lower axial and radial diffusivity for fibers between the posterior hypothalamus and the mesencephalon than cognitively non-fatigued patients. Replicating the analysis on a different data set of our group, using the FSS to assess fatigue and including a larger number of non-fatigued MS patients (total number of both investigations: 102 MS patients, 31 healthy controls), we found the same results. These findings suggest that disease-related changes in fibers between the posterior hypothalamus and the mesencephalon are associated with cognitive fatigue in MS patients. Fibers between the posterior hypothalamus and the mesencephalon include afferents of the vagal nerve, serotonergic and noradrenergic projections of wakepromoting brainstem regions as well as descending histaminergic neurons, innervating these wake-promoting areas. Altered integrity of these fibers might cause changes in the innervation of the posterior hypothalamus via afferents of the vagal nerve and via brainstem structures as well as changes in the innervation of brainstem structures via the posterior hypothalamus. Consequently, this might result in altered activity of neurotransmitter systems causing cognitive fatigue. To demonstrate the specificity of the relationship between cognitive fatigue and the integrity of fibers between the posterior hypothalamus and brainstem structures, we also investigated the integrity of fibers of the

anterior hypothalamus and the corpus callosum. As expected, the analysis of anterior hypothalamic fibers did not show significant effects in cognitive fatigue groups. Instead, significant interaction effects with the psychological BDI component were found, suggesting that depression may be associated with the integrity of fibers between the anterior hypothalamus and the mesencephalon, pons and prefrontal cortex. This finding is compatible with the depression model of Mayberg (1997) that considers depression to be a system level disorder affecting integrated pathways linking cortical, subcortical and limbic sites. Fiber analysis of the corpus callosum revealed significant group effects in both data sets. In both studies, healthy controls had the lowest means for axial and radial diffusivity. MS patients had higher diffusion values which might indicate pathological changes, such as demyelination and axon degeneration. AD and RD scores for the corpus callosum were both increased in MS patients. This difference with animal models (Budde et al., 2007) may be due to the fact that the latter concerns acute inflammation, whereas in our study chronic lesions play a major role. The results of the analyses of anterior hypothalamic fibers and the corpus callosum corroborate the specific finding that cognitively fatigued patients showed lowest diffusivity only for fibers between the posterior hypothalamus and the mesencephalon. In both of our studies, fatigued MS patients showed lower axial and radial diffusivity for fibers between the posterior hypothalamus and the mesencephalon than healthy controls, with non-fatigued MS patients having the highest diffusion scores. In contrast, MS patients always showed higher diffusivity scores than healthy controls for fibers of the corpus callosum. Assuming that higher diffusivity scores reflect tissue loss, fibers between the posterior hypothalamus and the mesencephalon of cognitively non-fatigued patients might be affected by increased demyelination and degeneration. This loss of fiber

Integrity of hypothalamic fibers and cognitive fatigue in multiple sclerosis integrity might contribute to the decreased perception of fatigue. Fibers between the posterior hypothalamus and the mesencephalon include afferent fibers of the vagal nerve. Afferents of the vagal nerve transmit information about peripheral inflammation to the CNS, resulting in suppressed histaminergic activity and in the initiation of symptoms of sickness behavior including fatigue (Gaykema et al., 2008). Marvel et al. (2004) have shown that mice with interrupted afferents of the vagal nerve do not experience sickness behavior, such as fatigue. Therefore, damage to these fibers might prevent non-fatigued patients to experience fatigue. In contrast, a normal or slightly above normal degree of fiber integrity between the brainstem and the hypothalamus, as reflected by the low AD and RD values of the cognitively fatigued MS patients, may cause fatigued patients to experience increased fatigue.

4.1.

Methodological limitation

A major limitation of our study is that we only obtained information about structural changes of hypothalamic fibers and their association with cognitive fatigue. To prove that these structural changes result in altered modulation of neurotransmitter systems causing cognitive fatigue in MS patients, functional brain imaging methods would be necessary. Another limitation of our study is the difference in baseline characteristics between groups. The groups that we compared did not only differ in fatigue scores, but also in level of depression and symptom duration. However, we controlled for depressive mood in the statistical analysis to ensure that the group differences that we found are due to differences in the fatigue level. Moreover, we did not perform gadolinium-enhanced MRI and we did not calculate overall lesion volume or lesion volume in the hypothalamus. Thus, we were not able to draw conclusions about acute disease activity. However, we excluded patients who had an MS relapse within the last four weeks and we used brain parenchymal fraction, white and gray matter volume as well as ventricle volumes as measures of global brain atrophy and checked for differences in atrophy between groups. Last but not least, one minor limitation of our study is that we used only subjective measures for the evaluation of cognitive fatigue. The use of objective measures, such as performance decline in sustained attention tasks, might add important information about the severity of cognitive fatigue in patients with MS.

5.

Conclusion

The finding that cognitively fatigued MS patients differ from cognitively non-fatigued patients and healthy controls with respect to axial and radial diffusivity for fibers between the posterior hypothalamus and brainstem areas suggests that changes in the integrity of these fibers are associated with cognitive fatigue in MS patients. Fibers between the posterior hypothalamus and the mesencephalon include afferents of the vagal nerve as well as serotonergic and noradrenergic fibers. Changes in the integrity of these fibers might result in altered neurotransmitter activity causing fatigue. Nevertheless, the results of our study only provide information on

45

structural changes of posterior hypothalamic fibers. To proof our hypothesis, studies investigating the direct relationship between neurotransmitter activity and fatigue in MS patients are necessary.

Financial support The study was funded by Bayer HealthCare Inc., MerckSerono Inc. and Novartis Inc.

Conflict of interest statement The authors confirm that there are no conflicts of interest and patent holdings related to this article.

Acknowledgments We thank Pia Lehmann, Anja Gossmann and Frauke Fink for their help in patient recruiting and investigation. We would also like to thank Imke Gillich and Ute Boekhoff for supporting the image analysis.

Appendix A.

Supporting information

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/ j.msard.2014.11.006.

References Alvarez EO. The role of histamine on cognition. Behav Brain Res 2009;199:183–9. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders.Fourth Edition-Text Revision Washington, DC: American Psychiatric Association; 2000. Bakshi R. Fatigue associated with multiple sclerosis: diagnosis, impact and management. Mult Scler 2003;9:219–27. Bakshi R, Shaikh ZA, Miletich RS, Czarnecki D, Dmochowski J, Henschel K, et al. Fatigue in multiple sclerosis and its relationship to depression and neurologic disability. Mult Scler 2000;6:181–5. Beck AT, Steer RA, Garbin MG. Psychometric properties of the Beck Depression Inventory: twenty five years of evaluation. Clin Psychol Rev 1988;8:77–100. Budde D, Kim JH, Liang H-F, Schmidt RE, Russell JH, Cross AH, et al. Toward accurate diagnosis of white matter pathology using diffusion tensor imaging. Magn Reson Med 2007;57:688–95. Copenhaver BR, Rabin LA, Saykin AJ, Roth RM, Wishart HA, Flashman LA, et al. The fornix and mammillary bodies in older adults with Alzheimer's disease, mild cognitive impairment, and cognitive complaints: a volumetric MRI study. Psychiatry Res: Neuroimaging 2006;147:93–103. Feigl GC, Hiergeist W, Fellner C, Schebesch K-M M, Doenitz C, Finkenzeller T, et al. Magnetic resonance imaging diffusion tensor tractography: evaluation of anatomic accuracy of different fiber tracking software packages. World Neurosurg 2013;81:144–50. Fink F, Klein J, Lanz M, Mitrovics T, Lentschig M, Hahn HK, et al. Comparison of diffusion tensor-based tractography and quantified brain atrophy for analyzing demyelination and axonal loss in MS. J Neuroimaging 2010;20:334–44.

46 Fisk JD, Ritvo PG, Ross L, Haase DA, Marrie TJ, Schlech WF. Measuring the functional impact of fatigue: initial validation of the fatigue impact scale. Clin Infect Dis 1994;18:79–83. Ford H, Trigwell P, Johnson M. The nature of fatigue in multiple sclerosis. J Psychosom Res 1998;45:33–8. Gaykema RPA, Park S-A, McKibbin CR, Goehler LE. Lipopolysaccharide suppresses activation of the tuberomamillary histaminergic system concomitant with behavior: a novel target of immune sensory pathways. Neuroscience 2008;152:273–87. Gillson G, Richards TL, Smith RB, Wright JV. A double-blind pilot study of the effect of Prokarin on fatigue in multiple. Mult Scler 2002;8:30–5. Haas H, Panula P. The role of histamine and the tuberomamillary nucleus in the nervous system. Nat Rev Neurosci 2003;4:121–30. Huitinga I, De Groot CJ, van der Valk P, Kamphorst W, Tilders FJ, Swaab DF. Hypothalamic lesions in multiple sclerosis. J Neuropathol Exp Neurol 2001;60:1208–18. Huitinga I, Erkut ZA, van Beurden D, Swaab DF. Impaired hypothalamus- pituitary-adrenal axis activity and more severe multiple sclerosis with hypothalamic lesions. Ann Neurol 2004;55:37–45. Klein J, Hermann S, Konrad O, Hahn HK, Peitgen H-O. Automatic quantification of DTI parameters along fiber bundles. In: Proceeding of image processing for medicine BVM. 2007. p. 272–276. Kos D, Kerckhofs E, Nagels G, D'hooghe M, Ilsbroukx S. Origin of fatigue in multiple sclerosis: review of the literature. Neurorehabil Neural Repair 2008;22:91–100. Krupp LB, Alvarez LA, LaRocca NG, Scheinberg LC. Fatigue in multiple sclerosis. Arch Neurol 1988;45:435–7. Krupp LB, LaRocca NG, Muir-Nash J, Steinberg AD. The fatigue severity scale: application to patients with multiple sclerosis and systemic lupus erythematosus. Arch Neurol 1989;46:1121–3. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an variability with the Expanded Disability Status Scale (EDSS) and expanded disability status scale (EDSS). Neurology 1983;33:1444–52. Lin J-S. Brain structures and mechanisms involved in the control of cortical activation and wakefulness, with emphasis on the posterior hypothalamus and histaminergic neurons. Sleep Med Rev 2000;2000(4):471–503. Lin J-S, Jouvet M. Evidence for histaminergic arousal mechanisms in the hypothalamus of cat. Neuropharmacology 1988;27:111–22. Lin J-S, Sakai K, Vanni-Mercier G, Jouvet M. A critical role of the posterior hypothalamus in the mechanisms of wakefulness determined by microinjection of muscimol in freely moving cats. Brain Res 1989;479:225–40. Lukas C, Hahn HK, Bellenberg B, Rexilius J, Schmid G, Schimrigk SK, et al. Sensitivity and reproducibility of a new fast 3D segmentation technique for clinical MR-based brain volumetry in multiple sclerosis. Neuroradiology 2004;46:906–15. Marvel FA, Chen CC, Badr N, Gaykema RP, Goehler LE. Reversible inactivation of the dorsal vagal complex blocks lipopolysaccharide-

K. Hanken et al. induced social withdrawal and c-Fos expression in central autonomic nuclei. Brain Behav Immun 2004;18:123–34. Mayberg HS. limbic-cortical dysregulation: a proposed model of depression. J Neuropsychiatry Clin Neurosci 1997;9:471–81. Monti JM, Pellejero T, Janlos H. Effects of H1- and H2 histamine receptor agonists and antagonists on sleep and wakefullness in the rat. J Neural Transm 1986;66:1–11. Murray TJ. Amantadine therapy for fatigue in MS. Can J Neurol Sci 1985;12:251–4. Parmentier R, Ohtsu H, Djebbara-Hannas Z, Valatx J-L, Watanabe T, Lin J-S. Anatomical, physiological, and pharmacological characteristics of histidine decarboxylase knock-out mice: evidence for the role of brain histamine in behavioral and sleep–wake control. J Neurosci 2002;22:7695–711. Penner IK, Raselli C, Stöcklin M, Opwis K, Kappos L, Calabrese P. The Fatigue Scale for Motor and Cognitive Functions (FSMC): validation of a new instrument to assess multiple sclerosisrelated fatigue. Mult Scler 2009;15:1509–17. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, et al. Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann neurol 2011;69:292–302. Schlüter M, Konrad-Verse O, Hahn HK, Stieltjes B, Rexilius J, Peitgen H-O. White matter lesion phantom for diffusion tensor data and its application to the assessment of fiber tracking (Medical imaging 2005: physiology, function, and structure from medical images). Proc SPIE 2005;5746:835–44. Schwid SR, Covington M, Segal BM, Goodman AD. Fatigue in multiple sclerosis: current understanding and future directions. J Rehabil Res Dev 2002;39:211–24. Thakkar MM. Histamine in the regulation of wakefulness. Sleep Med Rev 2011;15:65–74. van der Werf SP, Jongen PJH, Lycklama a‘ Nijeholt GJ, Barkhof F, Hommes OR, Bleijenberg G. Fatigue in multiple sclerosis: interrelations between fatigue complaints. Cerebral MRI abnormalities and neurological disability. J Neurol Sci 1998;160:164–70. Weiler F, Rexilius J, Klein J, Hahn HK (2009): NeuroQLab –A Software Assistant for Neurosurgical Planning and Quantitative Image Analysis. GI-Edition - Lecture Notes in Informatics (LNI) - Proceedings 154: 1352-58. Weinstein DM, Kindlmann GL, Lundberg E. Tensorlines: advection diffusion based propagation through diffusion tensor fields. In: Proceeding of the IEEE Visual. San Francisco; 1999. p. 249–253. Welch MJ, Meltzer EO, Simons FE. H1-antihistamines and the central nervous system. Clin Allergy Immunol 2002;17:337–88. Zellini F, Niepel G, Tench CR, Constantinescu CS. Hypothalamic involvement assessed by T1 relaxation time in patients with relapsing-remitting multiple sclerosis. Mult Scler 2009;15:1442–9.