Hypothalamic tumors impact gray and white matter volumes in fronto-limbic brain areas

Accepted Manuscript Hypothalamic Tumors Impact Grey and White Matter Volumes in Fronto-Limbic Brain Areas Jale Özyurt, Hermann L. Müller, Monika Warmuth-Metz, Christiane M. Thiel PII:

S0010-9452(17)30031-X

DOI:

10.1016/j.cortex.2017.01.017

Reference:

CORTEX 1930

To appear in:

Cortex

Received Date: 19 May 2016 Revised Date:

23 November 2016

Accepted Date: 20 January 2017

Please cite this article as: Özyurt J, Müller HL, Warmuth-Metz M, Thiel CM, Hypothalamic Tumors Impact Grey and White Matter Volumes in Fronto-Limbic Brain Areas, CORTEX (2017), doi: 10.1016/ j.cortex.2017.01.017. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Hypothalamic Tumors Impact Grey and White Matter Volumes in Fronto-Limbic Brain Areas Jale Özyurt1, Hermann L. Müller2, Monika Warmuth-Metz3, Christiane M. Thiel1,4

RI PT

Affiliations: 1Biological Psychology Lab, Department of Psychology, Faculty of Medicine and Health Sciences, Carl von Ossietzky Universität, Oldenburg, Germany; 2Department of Pediatrics and Pediatric Hematology and Oncology, Klinikum Oldenburg, Medical Campus University Oldenburg Oldenburg, Germany; 3Department of Neuroradiology, University Hospital Würzburg, Würzburg, Germany; 4Research Center Neurosensory Science and Cluster of Excellence “Hearing4all”, Carl von Ossietzky Universität, Oldenburg, Germany

M AN U

SC

Address correspondence to: Jale Özyurt, 1Biological Psychology Lab, Department of Psychology, Faculty of Medicine and Health Sciences, Carl von Ossietzky Universität, Oldenburg, Germany; email: [email protected]; fax: +49-441-798-3848; phone: +49-441-798-3891 Financial Disclosure: The authors have no financial relationships relevant to this article to disclose Conflict of Interest: The authors have no conflicts of interest to disclose

Keywords: Brain tumors, cognitive, hypothalamus, craniopharyngioma, social-emotional

AC C

EP

TE D

Clinical Trial Registration: KRANIOPHARYNGEOM 2000. Clinical Trial Registration Number: NCT00258453

ACCEPTED MANUSCRIPT

Abstract Patients with hypothalamic involvement of a sellar/parasellar tumor often suffer from cognitive and social-emotional deficits that a lesion in the hypothalamus cannot fully explain. It is conceivable that

RI PT

these deficits are partly due to distal changes in hypothalamic networks, evolving secondary to a focal lesion. Focusing on childhood-onset craniopharyngioma patients, we aimed at investigating the impact of hypothalamic lesions on grey and white matter areas densely connected to the hypothalamus, and to relate structural changes to neuropsychological deficits frequently observed in patients.

We performed a voxel-based morphometric analysis based on data of 11 childhood-onset

SC

craniopharyngioma patients with hypothalamic tumor involvement, and 18 healthy controls (median age: 17.2 and 17.4 yrs.). Whole-brain analyses were used to test for volumetric differences between the groups (T-tests) and subsequent regression analyses were used to correlate neuropsychological

M AN U

performance with grey and white matter volumes within the patient group.

Patients compared to controls had significantly reduced grey matter volumes in areas of the anterior and posterior limbic subsystems which are densely connected with the hypothalamus. In addition, a reduction in white matter volumes was observed in tracts connecting the hypothalamus to other limbic areas. Worse long-term memory retrieval was correlated with smaller grey matter volumes in the posterior cingulate cortex.

TE D

Our data provide the first evidence that hypothalamic tumor involvement impacts grey and white matter volumes in limbic areas, outside the area of tumor growth. Notably, the functional range of the two limbic subsystems affected, strikingly parallels the two major domains of psychological complaints in patients i.e. deficits in episodic memory and in socio-emotional functioning. We suggest that focal hypothalamic lesions may trigger distal changes in connected brain areas, which then

EP

contribute to the impairments in cognitive, social and emotional performance often observable in

AC C

patients, and not explicable by a hypothalamic lesion alone.

1

ACCEPTED MANUSCRIPT

1.

Introduction

The hypothalamus is a small multinuclear complex at the brain base (approx. 0.7 cm³ in each side), located below and anteriorly to the thalamus. It regulates and coordinates a wide range of functions, which are of vital importance for survival, such as food and water intake, control of the cardiovascular

RI PT

system, temperature regulation, and control of pituitary function (Lemaire et al., 2011; Lemaire et al., 2013). Lesions to this region may result in profound endocrine and autonomic dysfunctions with potentially life-threatening outcomes. Accordingly, the often more subtle deficits in cognitive and socio-emotional outcomes after hypothalamic lesions have for long received less attention. In addition, isolated lesions of the hypothalamus are rare in humans and knowledge on its role in neurobehavioral

SC

functioning is mainly based on animal research and a few single case studies in humans (Vann, 2010; Vann & Aggleton, 2004).

Craniopharyngiomas, however, as the most common tumors of the hypothalamic-pituitary region in

M AN U

children (Warmuth-Metz, Gnekow, Muller, & Solymosi, 2004), bear a high potential for investigating the effects of hypothalamic lesions on neurobehavioral functioning. These rare embryonal malformations have a low-grade histological malignancy and can be located anywhere along an axis extending from the sella turcica through the optic nerves, the pituitary stalk and the hypothalamus (Flitsch, Muller, & Burkhardt, 2011). Accordingly, the most frequent symptoms, which can seriously limit psychosocial functioning and quality of life, comprise visual field defects, loss of neurovegetative homeostasis, endocrine, and neurobehavioral disturbances. Long-term outcomes have

TE D

been shown to be particular worse in case of hypothalamic involvement. Rates reported for hypothalamic involvement range from 60-92% for preoperative tumor involvement, and 59-71% for postoperative lesions (de Vile et al., 1996; Fjalldal et al., 2013; Hoffmann et al., 2014; Laffond et al., 2012; Puget et al., 2007). In most cases, craniopharyngiomas do not involve any other brain areas

EP

relevant for cognition or socio-emotional performance. In patients with childhood-onset craniopharyngioma, the most consistent findings in the cognitive domain are impairments in learning and episodic memory, largely sparing other memory components,

AC C

(Ozyurt, Muller, & Thiel, 2015). This is in accordance with single case studies and animal research, showing that isolated lesions of the mammillary bodies in the posterior part of the hypothalamus cause episodic memory deficits similar to those of patients with hippocampal lesions, but to a lesser degree (Vann & Aggleton, 2004). Patients with hypothalamic lesions may however also suffer from deficits in executive functioning, relying on the integrity of the prefrontal cortex along with its subcortical pathways, and deficits that indicate fronto-limbic dysfunctioning, such as emotional lability, and rage attacks (Garnett, Puget, Grill, & Sainte-Rose, 2007; Muller, 2008; Pierre-Kahn et al., 2005). Moreover, some patients complain about difficulties in social interactions and relations, and in social-cognitive skills such as understanding others’ emotional states, thoughts and feelings (Hermann L. Müller, personal communication). 2

ACCEPTED MANUSCRIPT

While lesions to the mammillary bodies of the hypothalamus were shown to be sufficient to cause clear deficits in episodic memory (Vann, 2010), impairments in executive or social-emotional skills do not appear to be fully explainable by a hypothalamic lesion. Such impairments may arise from several disease- and treatment-related factors associated with brain injury outside the region of tumor growth, including, e.g., surgical approaches, obstructive hydrocephalus, peri- and postoperative complications,

RI PT

and radiation effects. Alternatively, they may be due to distal effects, which relate to the long known finding that brain pathology is likely not confined to a discrete region but is propagated along its axonal pathways to affect connected areas. Pathological processes in disease propagation include diaschisis (interruption of function in distal regions due to functional connectivity changes) and transneuronal degeneration (structural degeneration of distal brain regions/pathways). In both cases,

SC

patterns of disease spread are shaped and constrained by the connectivity of the lesioned area (Fornito, Zalesky, & Breakspear, 2015).

M AN U

Within a network supporting behavioral control, the hypothalamus has strong connections with major cortical and subcortical limbic areas, such as the orbitofrontal cortex, cingulum, insula, hippocampus, ventral striatum, amygdala, and thalamic nuclei. The dense hypothalamic projections to thalamic nuclei are of particular interest as they allow for substantial indirect influences of the hypothalamus on the ventral striatum and orbital/medial prefrontal cortex (Lemaire et al., 2011; Risold, Thompson, & Swanson, 1997; Toni, Malaguti, Benfenati, & Martini, 2004). The anatomical network of the hypothalamus is paralleled by resting state connectivity, showing significant functional connections

TE D

with the same brain areas Kullmann et al. (2014). Notably, the notion of a single limbic system has been recently questioned in favor of at least two functionally distinct limbic sub-systems: a posterior subsystem that constitutes a neural system supporting episodic memory and an anterior subsystem supporting emotional, motivational and social functioning (Catani, Dell'acqua, & Thiebaut de 2013;

Rolls,

2015).

Neurobehavioral

deficits

most

frequently

reported

for

EP

Schotten,

craniopharyngioma patients (episodic memory and socio-emotional deficits) strikingly correspond to each of these two subsystems’ functional range.

AC C

Cortical diaschis after thalamic stroke is a well known phenomenon and cortical areas affected were shown to depend on areas of thalamo-cortical fibre loss (Baron et al., 1992; De Witte et al., 2011). However, only few studies exist on distal effects of hypothalamic lesions. In a study with mice, damage to the mammillary bodies with a chemical agent resulted in distal hypoactivity of frontal cortex and hippocampus, assumed to arise from a reduction of excitatory mammillary body influence on thalamic target cells (Beracochea, Micheau, & Jaffard, 1995). In another study with rats, disconnection of efferents from the medial mammillary to anterior thalamic nuclei resulted in hypoactivity in the hippocampus, the retrosplenial and prefrontal cortex, which are all strongly involved in episodic memory performance (Vann, 2009). Distal effects of hypothalamic injury in humans were only recently shown in an fMRI study the current subsample is based on (Ozyurt, Thiel, et al., 2014). In patients compared to controls, task-induced deactivation during memory recognition 3

ACCEPTED MANUSCRIPT

was significantly reduced in orbital and adjacent medial prefrontal cortex. This failure of deactivation was assumed to be functionally related to the altered functional coupling which was observed between patients’ rostral medial prefrontal cortex and the thalamus. The current study was motivated by the assumption that some behavioral abnormalities in postsurgical

RI PT

patients with hypothalamic tumor involvement are not explicable by a hypothalamic lesion alone but partly due to distal lesion effects, such as transneuronal degeneration. We used voxel-based morphometry (VBM) to characterize tumor- and treatment-related structural brain pathology, with a special focus on the effects of secondary disease propagation following a circumscribed primary lesion to the hypothalamus. In addition, we aimed to relate findings on brain pathology to cognitive and

SC

social-emotional deficits associated with hypothalamic lesions. Based on hypothalamic connectivity and on patients’ known deficits in episodic memory and social-emotional competencies, we hypothesized that hypothalamic lesions are associated with structural changes in limbic subsystems

AC C

EP

TE D

M AN U

and associated thalamic nuclei.

4

ACCEPTED MANUSCRIPT

2. Materials and methods 2.1 Subjects Based on a larger sample recruited for a neuropsychological and an fMRI study in the context of the multinational trials on children and adolescents with craniopharyngioma (KRANIOPHARYNGEOM

RI PT

2000/2007; Clinical Trial Registration Number: NCT00258453) (Muller, 2010), we analyzed a subsample of 11 patients and 18 healthy controls (median age: 17.2 and 17.4 yrs.) for whom a high resolution anatomical image was available. The recruitment procedure for patients is detailed elsewhere (Ozyurt, Thiel, et al., 2014). Patients all had a history of childhood-onset craniopharyngioma, and hypothalamic involvement due to the tumor and/or surgical lesion, no major

SC

visual impairments that may interfere with neuropsychological task performance, and no shunts or other devices that prohibited MRI. Participants of the healthy control group had no known neurological or psychiatric disorders, normal or corrected-to-normal vision and, (for both groups)

M AN U

normal hearing abilities. They were selected to match the patient group with respect to age, sex and intelligence (Table 1).

The study was conducted in accordance with the Declaration of Helsinki and all procedures were carried out with the adequate understanding and written consent of the participants or the participants’ parents. Ethics approval was obtained from the ethics committee of the local university.

Information

related

to

TE D

< Insert Table 1 here >

patient

history

was

obtained

from

clinical

records

of

the

KRANIOPHARYNGEOM 2000/2007 studies. The body mass index (BMI) was calculated for each

EP

subject at diagnosis and at the time of the study. Scores are depicted as a standard deviation score (BMI SDS) (Rolland-Cachera et al., 1991). Based on Roth et al., (2007) a BMI >4 SDS was defined as severe obesity, a BMI between 2 and 4 SDS as a moderate obesity and a BMI <2 SDS as normal

AC C

weight. Data on BMI SDS was only available for the patient group. The anatomical MRI images were assessed by a neuroradiologist (one of the authors, Warmuth-Metz) who was blinded to the clinical data. For the purpose of the larger neuropsychological study (Ozyurt, Thiel, et al., 2014), the current status of hypothalamic involvement had been assessed by assigning patients to one of three grades based on a novel grading system for preoperative hypothalamic involvement/postoperative hypothalamic lesions (Grade 0: no involvement/lesion, Grade 1: involvement/lesion of the anterior hypothalamus, Grade 2: involvement/lesion of the anterior and posterior hypothalamic area (the latter involving the mammillary bodies and the posterior hypothalamic area beyond mammillary bodies) (Muller et al., 2011). For the purpose of the current VBM study, the anatomical MRI images were reassessed by the same neuroradiologist to identify, by visual inspection, lesions of grey and white 5

ACCEPTED MANUSCRIPT

matter structures outside the hypothalamus, related to e.g. surgical corridors or peri- and postoperative complications. 2.2. Statistical analyses of behavioral data. Note that all participants of our VBM study were also tested with formal neuropsychological tests as part of the larger neuropsychological study. For detailed

RI PT

information related to the neuropsychological methods we refer to a previous publication (Ozyurt, Thiel, et al., 2014). Statistical analyses of patients and controls included in the current analyses were performed with SPSS 19 for Windows. For between-group analyses we used the nonparametric Mann Whitney U-test (two-sided). Results were evaluated at p≤.05 (Bonferroni-corrected for multiple comparisons).

SC

2.3 MRI acquisition

A 1.5-T SONATA MRI system (Siemens, Erlangen, Germany) with an 8 channel headcoil was used to obtain anatomical images. The high-resolution structural volumes were collected for each participant

M AN U

as part of a larger neuroimaging and neuropsychology protocol. A high resolution T1-weighted scan (176 contiguous slices, each slice 224x256 voxels, voxel size = 1x1x1 mm³) was conducted with a magnetisation prepared rapid acquisition gradient echo sequence (MPRAGE, TR = 1.97 s, TE = 3.93 ms and α = 15°) for each participant. 2.4 Data analysis

Voxel-based morphometry (VBM) analysis of the data was performed with SPM8 (Wellcome Trust

TE D

Center for Neuroimaging, London; http://www.fil.ion.ucl.ac.uk/spm/software/spm8/). The technique enables a voxel-wise statistical comparison of tissue characteristics throughout the whole brain to identify subtle brain structural changes which are usually not detectable during visual insprection of the MR images (Ashburner & Friston, 2001). It is not recommended for severe brain lesions as the

EP

routines cannot reliably identify and delineate damaged regions (Rorden & Karnath, 2004). VBM analyses were performed following the steps depicted in Ashburner’s VBM Tutorial; http://www.fil.ion.ucl.ac.uk/~john/misc/VBMclass10.pdf. Prior to the preprocessing steps, the T1

AC C

images were assessed for gross anatomical abnormalities and approximately aligned to the template data released with SPM8. Subsequently, the images were automatically segmented into grey matter, white matter, and cerebrospinal fluid using the New Segment option implemented in SPM8. For an accurate inter-subject alignment of grey and white matter images a DARTEL (diffeomorphic anatomical registration through exponentiated lie algebra) approach was applied. The resulting DARTEL template, which was registered to MNI space, was then used to bring the individual spatially normalized scans into MNI space. As we were interested in volume changes, this step involved modulation, i.e. scaling by the amount of contraction to ensure that the total amount of grey and white matter volumes in the modulated images is the same as in the original images. Finally, the grey and white matter images were smoothed with an isotropic Gaussian kernel at 8-mm full-width half maximum. 6

ACCEPTED MANUSCRIPT

Preprocessed grey and white matter images were each entered into a general linear model for a twosample T-test comparing patients to controls in a whole-brain analyses. Here, the total intracranial volume was computed and used for global normalization. As we had a strong hypotheses on thalamic involvement, we additionally performed a small volume correction with a mask of the thalamus as included in the AAL ROI-Library (Tzourio-Mazoyer et al., 2002). To account for complications that

RI PT

may affect grey and white matter volumes in patients, we reanalyzed the data several times, each time excluding one of the patients with a subfrontal approach, a transcallosal approach, a basal ganglia infarct or a hydrocephalus.

To relate the structural changes observed in patients to neuropsychological performance, results of

SC

neuropsychological tests that showed significant differences for patients compared to controls (within the current subsample) were included in subsequent regression analyses. Further, to relate the structural changes with functional changes described previously, additional regression analyses were

M AN U

performed with beta values obtained from an orbital/medial prefrontal cortex cluster showing abnormal activation in the patient group (Ozyurt, Lorenzen, et al., 2014). Finally, to assess whether reductions in patients’ grey and white matter volumes simply reflect detrimental effects obesity may have on the brain (Figley, Asem, Levenbaum, & Courtney, 2016), we performed regression analyses with patients’ BMI standard deviation scores (SDS). All regression analyses were confined to patients and the search for significant clusters was restricted to grey and white matter areas significantly differing between patients and controls in the two-sample T-test (including the small volume

TE D

correction). For all analyses, the initial voxel threshold was set to p<.001 uncorrected. Multiple testing was accounted for on cluster level, based on a corrected pFWE <.05 and activations were reported in MNI coordinates. Brain regions were identified using two atlases (Duvernoy, 2003; Oishi, Faria, van

AC C

EP

Zijl, & Mori, 2011).

7

RI PT

ACCEPTED MANUSCRIPT

3. Results

3.1 Demographic and clinical variables for the patient and control groups

The patient and control groups were balanced with respect to age, sex, and intelligence as tested with

SC

the German adaptation of the Culture Fair Intelligence Test - Scale 2 by R.B. Cattell (Weiß, 1987). (p>.05, see Table 1). Additionally, no significant between-group differences were found with respect to state and trait anxiety, and the depression score.

M AN U

3.2 Patient history and current clinical status

Preoperatively, only one of the patients suffered from severe obesity and three were moderately obese. Postoperatively, a considerable weight gain was detectable for most of the patients, thus showing symptoms closely associated with hypothalamic dysfunction (de Vile et al., 1996; Muller et al., 2011). At the time our study was conducted, the majority in our sample was severely obese (N=7). At the time of primary surgery/diagnosis, the median age of the patients was 10.5 years (7.6-17.6 yrs.). The

TE D

median interval between primary surgery and data collection within the current study was 8.0 years (4.3-18.9 yrs.). Preoperatively, hypothalamic tumor involvement could be confirmed in all but one patient (data not available), either with only anterior (grade 1; N=3) or with both anterior and posterior hypothalamic involvement (grade 2; N=7). Postoperatively, a hypothalamic lesion was observed in 10 patients (grade 1: N=3; grade 2: N=7) (Table 2). Figure 1A in the appendix aims to illustrate the extent

EP

and location of lesions in the patient group, presenting cases for hypothalamic involvement grade 1 and 2, together with a structural image of a healthy control participant.

AC C

Most of the surgical approaches were right hemispheric, with a pterional approach (N=8). In four of these patients, a surgical lesion was detectable along a surgical corridor fronto-laterally, in the insular region. In two cases, a transcallosal approach was chosen, with one of the patients showing a surgical lesion in the right frontal lobe near midline, and the other patient showing a lesion in corpus callosum only. A subfrontal surgical approach with an associated frontobasal midline defect was only seen in one patient. None of the patients had additional surgeries, e.g. for cyst decompression or ventricular drainage. Due to tumor progression or relapse, two of the patients had a second surgery along the former surgical corridor. Substance defects outside the hypothalamus or surgical corridors were identified in two patients. One of the patients had a right basal ganglion infarct and another patient had an enlarged ventricle due to obstructive hydrocephalus. No other than surgical lesions were detected in the frontal lobe (Table 3). 8

ACCEPTED MANUSCRIPT

RI PT

< Insert Table 2 and 3 here >

3.3 Neuropsychological Results

Significant differences between the patient and control group were found for long-term memory in the German version of the Auditory Verbal Learning Test (Helmstaedter, Lendt, & Lux, 2001). Patients compared to controls had worse performance when required to successfully recall verbal information

SC

after delay (p<.001, corrected). Lower performance scores in patients compared to the control group were also obtained in the set shifting task (Intra-Extradimensional Set Shift) of the Cambridge Neuropsychological Test Automated Battery (CANTAB; Cambridge Cognition, Cambridge, United

M AN U

Kingdom), which measures frontal lobe function (p<.05). 3.4. VBM Results

A two sample T-test with the grey matter images resulted in two large clusters with significantly reduced volumes for patients compared to controls (Figure 1A-B). One cluster (2797 voxel; pFWE <.001) comprised the ventral striatum (caudate nucleus and nucleus accumbens), orbital and adjacent medial prefrontal cortex, and sub- and perigenual anterior cingulate cortex. All regions were affected

TE D

bilaterally. The second cluster (1198 voxel; pFWE=.007) comprised the anterior and mid cingulum and the posterior cingulate cortex/retrosplenium. A small volume correction with a bilateral thalamic ROI resulted in a right-hemispheric thalamic cluster, comprising medial and posterior thalamic areas

EP

(511 voxel; pFWE=.003; Figure 1C).

< Insert Figure 1 and 2 here >

AC C

A two sample T-test with the white matter images resulted in two clusters, each including similar regions in the right and left hemisphere (4070 and 3587 voxel; both pFWE <.001) and with significantly less white matter volume for patients compared to controls (Figure 2). The peak activation of this cluster was located in the region of the right optic tract, near the mammillary bodies. Both clusters comprise the fornix, stria terminalis, ansa lenticularis, lenticular fasciculus, inferior fronto-occipital bundle, uncus, sagittal striatum, including inferior longitudinal fasciculus and inferior fronto-occipital

fasciculus,

the

hypothalamus/mammillary

bodies,

putamen,

amygdala,

superior/inferior and middle temporal white matter areas. Subsequent regression analyses within these areas revealed that patients’ BMI SDS scores were not significantly related to reductions in grey and white matter areas. Reanalyzing the grey and white matter volumes to account for complications and

9

ACCEPTED MANUSCRIPT

surgical pathways that may affect grey and white matter volumes in patients did not result in substantial changes. Significant differences between patients and controls were obtained in two neuropsychological domains: long-term memory recall and set shifting. To relate the structural changes observed in

RI PT

patients to neuropsychological performance, patients’ scores obtained with the respective tests were used for regression analyses. Results revealed a significant positive correlation between long-term memory performance and grey matter volumes in the left posterior cingulate cortex (Figure 3A). In other words, worse long-term memory performance was correlated with less grey matter in this brain region (399 voxel; pFWE =.013). No significant clusters were found within grey and white matter

SC

volumes when conducting a regression analysis with performance data from the set-shifting task.

An additional regression analysis, aimed to relate the structural changes in patients investigated here with functional changes described previously (Ozyurt, Lorenzen, et al., 2014). We found a significant

M AN U

negative correlation between BOLD activity (beta values) obtained from the orbital/medial prefrontal cortex cluster showing abnormal activation in patients in our previous fMRI study, with white matter volume changes in the left ansa lenticularis (578 voxel; pFWE =.007; Figure 3B). Thus, the stronger the abnormality in the functional signal, the higher the volumetric reduction in a small part of the left ansa lenticularis, which contributes to the connection of the medial thalamus and the hypothalamus

TE D

with the medial temporal and orbitofrontal lobes.

AC C

EP

< Insert Figure 3 here >

10

RI PT

ACCEPTED MANUSCRIPT

4. Discussion

SC

Comparing craniopharyngioma patients with hypothalamic involvement to healthy controls, the current study provides the first evidence for grey and white matter volume reductions outside the area of tumor growth. Volumetric reductions were found in the anterior and posterior limbic subsystems

M AN U

and some of their major pathways (see Catani et al., 2013; Rolls, 2015). Most importantly, the functional range of each of the subsystems strikingly corresponded to the two major domains of behavioral impairments in patients (i.e. episodic memory and socio-emotional skills). Metabolic abnormalities in most of these limbic areas, amongst some others, were also shown in a recent postoperative PET-study, which was conducted with 50 childhood-onset craniopharyngioma patients before proton therapy. Compared to the patients in our study, which all had hypothalamic involvement and were highly selected due to their participation in an fMRI study, patients in the PET-study

TE D

suffered from a much higher rate of complications and additional surgeries (Hua et al., 2015). Surgical pathways and complications may considerably affect the integrity of white and grey matter volumes and hence it is worth to note that results of our study even hold when these variables were accounted for as far as possible. In addition, the lack of any correlation between BMI standard deviation scores

EP

and patients’ grey or white matter volumes indicate that volumetric differences between patients and controls are not simply due to detrimental effects of obesity. These results imply additional causes for volumetric reductions in patients and based on the literature we propose that they are, to a large extent,

AC C

a consequence of pathological processes secondary to the hypothalamic lesion, such as anterograde or retrograde transneuronal degeneration of connected areas and pathways. In brain networks, such pathological processes can spread between connected elements to affect surprisingly large areas of the system (Fornito et al., 2015). It has been shown, for example, that lesions to a small region such as the anterior thalamic nucleus of rats may result in severe spatial learning deficits together with decreases of immediate-early gene expression in a wide range of limbic and limbic related cortical structures, indicating decreased neuronal activity (Aggleton & Nelson, 2015). In the following discussion, we will focus on regions and pathways which exhibited volume reductions in our patient group and outline their functional relevance for cognitive and socio-emotional behavior which are frequently impaired in patients with hypothalamic lesions. 11

ACCEPTED MANUSCRIPT

4.1. Grey and white matter reductions in limbic circuits involved in episodic memory performance The mammillary bodies in the posterior part of the hypothalamus and their pathways are part of the posterior limbic system (also known as the extended hippocampal system), vital for encoding and

RI PT

retrieval of episodic memory (Aggleton & Brown, 1999). Their crucial role for episodic memory is now well established by results obtained in animal studies (Vann, 2010; Vann & Aggleton, 2004) and single case studies in humans (Dusoir, Kapur, Byrnes, McKinstry, & Hoare, 1990; Hildebrandt, Muller, Bussmann-Mork, Goebel, & Eilers, 2001; Kapur et al., 1998; Squire, Amaral, Zola-Morgan, Kritchevsky, & Press, 1989; Tanaka, Miyazawa, Akaoka, & Yamada, 1997; Teuber, Milner, &

SC

Vaughan, 1968; Vann et al., 2008). Significant episodic memory deficits compared to controls were also found in our recently published neuropsychological study on craniopharyngioma patients with hypothalamic involvement (Ozyurt, Thiel, et al., 2014) and the subsample of patients the current

M AN U

analysis is based on. Notably, the memory deficits in our patient group are in accordance with recent evidence from patients with childhood-onset craniopharyngioma, particularly from those with hypothalamic involvement (for a review see Ozyurt et al., 2015).

The structural connectivity pattern of the posterior limbic system for episodic memory is outlined in Figure 4A. The mammillary bodies receive strong projections from the hippocampus (by the postcommissural fornix) and mainly project to the anterior thalamic nuclei (by the mammillothalamic tract). Importantly, the fornix projections, which arise from the subiculum of the hippocampus

TE D

terminate in both, the mammillary bodies and the anterior thalamic nuclei so that the latter receive both, direct and indirect hippocampal input. Connections between the anterior thalamic nuclei and the hippocampus are bidirectional (Lemaire et al., 2011; Vann & Aggleton, 2004). The mammillary bodies are linked via the mammillothalamic tract and thalamo-cingulate pathway, to the retrosplenial

EP

cortex, which is densely and reciprocally connected with the anterior thalamic nuclei, the prefrontal cortex, and the hippocampus (Maguire, 2001; Yu et al., 2011). The prefrontal cortex was assumed to exert control of memory mainly via projections of the retrosplenial cortex to the hippocampus

AC C

(Aggleton, 2014).

< Insert Figure 4 here >

We here report white matter reductions within the patients’ posterior limbic system, which were observed in a considerable proportion of the fornix, the right mammillary body, and the region of the mammillo-thalamic tract bordering the lateral hypothalamic area (most obvious in the right hemisphere). Grey matter reductions in patients compared to controls were found in the retrosplenial cortex, which is connected to the hypothalamus by an anterior thalamic pathway. The anterior 12

ACCEPTED MANUSCRIPT

thalamus is not affected by volume reductions, but it is known that the retrosplenial cortex is extremely sensitive to indirect deafferentiation by lesions of the mammillothalamic tract (Vann & Albasser, 2009). In a regression analysis which aimed to relate the structural changes observed in patients to their neuropsychological performance, we found a positive correlation between long-term memory performance and grey matter volume in the posterior cingulate cortex (Brodmann area 31). In

RI PT

other words, better memory performance within the patient group was correlated with higher posterior cingulate grey matter volume, an area which is strongly connected to the retrosplenial cortex and the mediodorsal thalamus and also implicated in memory (Beckmann, Johansen-Berg, & Rushworth, 2009; Maguire, 2001). Note that albeit research in animals has shown that isolated lesions of the mammillary bodies are sufficient to cause clear deficits in episodic memory (Vann, 2010), the

SC

volumetric reductions in grey and white matter shown in large parts of the posterior limbic system in our study may well accentuate individual patients’ memory impairments.

M AN U

4.2. Grey and white matter reductions in limbic circuits involved social and emotional performance

Hypothalamic regions anterior to the mammillary bodies form part of the anterior limbic system, vital for motivational, emotional and social behavior (Morgane, Galler, & Mokler, 2005; Rolls, 2015). Deficits in these functional domains have been frequently reported for patients with childhood-onset craniopharyngioma, based on investigations with standardized questionnaires (for a review see Zada, Kintz, Pulido, & Amezcua, 2013). The role of the hypothalamus for the regulation of behaviors

TE D

essential for survival, such as aggression, sexual and maternal behavior, eating, drinking, and appetitive behavior is well evidenced from studies with animals and humans. In cats, electrical stimulation of different hypothalamic areas results in different coordinated affective behavior patterns (e.g. defence, attack, flight or rage) with concomittant autonomic signs, such as increased heart rate

EP

and alertness (Brown, Hunsperger, & Rosvold, 1969; Dalgleish, 2004). In the light of our knowledge on the role of the hypothalamus for emotion, however, it is conceivable that impairments in socialemotional competencies are not explainable by hypothalamic lesions alone but based on damage to a

AC C

broader limbic network.

The structural connectivity pattern of the anterior limbic network for motivational, emotional and social behavior is outlined in Figure 4B. The hypothalamus, its lateral and medial zones in particular, are densely connected with other brain regions of the anterior limbic system. Connections are both, direct and indirect via the thalamus. The orbital/medial prefrontal cortex and the subgenual/perigenual cingulate cortex are directly and reciprocally connected with the hypothalamus by the medial forebrain bundle, but connections also exist with ventral parts of the supracallosal anterior cingulate cortex. Indirect hypothalamic connections with these areas synapse in the thalamus and are mediated by the inferior thalamic peduncle and ansa lenticularis (which also contributes to the medial forebrain bundle). The ventral striatum receives hypothalamic information by way of the thalamus and projects 13

ACCEPTED MANUSCRIPT

directly and indirectly (after synapsing in the substantia innominata) to the lateral hypothalamic zone. The amygdala projects to the lateral and medial parts of the hypothalamus by the ansa lenticularis, and to its rostral medial part by the stria terminalis. The lateral part of the hypothalamus projects in turn, via the ansa peduncularis, to the amygdala. All regions of the limbic network connecting to the hypothalamus are in turn richly interconnected (Beckmann et al., 2009; Lemaire et al., 2011; Ongur,

RI PT

An, & Price, 1998; Rempel-Clower & Barbas, 1998; Risold et al., 1997; Rolls, 2015; Toni et al., 2004).

We here report white matter reductions in subcortical nuclei, such as the hypothalamus, amygdala and putamen, and in major connecting pathways, such as the medial forebrain bundle, ansa lenticularis

SC

(which contributes to the medial forebrain bundle and the ansa peduncularis), lenticular fasciculus (which contributes with the ansa lenticularis to the thalamic fasciculus), and a considerable portion of the stria terminalis. Grey matter reductions in patients compared to controls were shown in three

M AN U

regions of the anterior limbic system: the orbitofrontal cortex, ventral striatum (nucleus accumbens and ventral parts of the caudate nucleus), and anterior cingulate cortex (including peri- and subgenual cingulate areas).

Damage to the human orbitofrontal cortex frequently results in affective changes such as euphoria, lack of affect, aggressiveness and impulsiveness. Likewise, lesions of the anterior part of the cingulate cortex have been associated with emotional instability, apathy, and autonomic dysregulation (Rolls, 2015). These signs and symptoms are similar to the neurobehavioral abnormalities described for

TE D

patients with cranipharyngioma (Zada et al., 2013). Apart from these broader emotional and motivational functions, the orbitofrontal cortex, anterior cingulate cortex and ventral striatum are also implicated in processing the reinforcement value of stimuli (reward and punishment), representing expected outcomes, detecting errors in reward prediction, adjusting value information based on

EP

changed contingencies and utilizing this information to guide behavior (Rolls, 2015; Schoenbaum, Roesch, Stalnaker, & Takahashi, 2009). It is well established that the orbitofrontal cortex is crucial for representing and learning the motivational and emotional values of social and non-social stimuli (e.g.

AC C

of individual faces or objects). It is however, still under debate whether it is of vital importance for the rapid learning and readjustment of values in view of changed reinforcement contingencies or whether it only contributes to readjustments in other brain regions (Rolls, 2015; Schoenbaum et al., 2009). The anterior cingulate cortex is suggested to receive value information about expected outcomes and reward magnitude from caudal orbitofrontal cortex and amygdala and to link these outcome representations with action-related information represented in the midcingulate cortex. It thus serves learning about actions needed to obtain goals, by additionally considering the costs of the actions (Rolls, 2015). In reward learning and decision making, reward-related responses in the ventral striatum may play an important role in computing predictions errors, i.e. deviations from reward expectations (Hare, O'Doherty, Camerer, Schultz, & Rangel, 2008; Peters & Buchel, 2010). 14

ACCEPTED MANUSCRIPT

The ability to recognize and properly evaluate others’ emotional states and feelings (e.g. in facial expressions, speech or gestures) is an essential basis for inferences about others’ thoughts. In addition, the ability to flexibly update valuations in view of changing contingencies is of equal importance for stable social interactions. As shown in a study on patients with brain lesions, orbitofrontal and/or anterior ventral cingulate cortex lesions can impair the identification of affective voice or face

RI PT

expressions. Severe changes in social behavior and emotional state were in particular associated with bilateral lesions to the orbitofrontal cortex (Hornak et al., 2003). The sensitivity of the orbitofrontal and anterior cingulate cortices to changing contingencies in social cues was shown in an fMRI study on discrimination reversal, when a choice reversal was required based on a face that was no longer rewarding (smiling) but showing an angry expression (Kringelbach & Rolls, 2003). The anterior

SC

rostral medial prefrontal cortex, adjacent to the orbitofrontal cortex, which also revealed a reduced grey matter volume in our patients, was in several studies shown to be involved during social

M AN U

interactions and theory of mind tasks (Amodio & Frith, 2006).

Based on the functional anatomy of the anterior limbic system as outlined above, and on patterns of limbic brain pathology as shown in our study, it is conceivable that patients with hypothalamic lesions suffer from impairments in social-cognitive skills, such as understanding others’ emotional states, thoughts and feelings, which remains to be investigated.

4.3. Grey and white matter reductions in hypothalamic circuits involved in executive

TE D

performance

Hypothalamic lesions have been associated with deficits in frontal lobe based higher cognitive functions in several studies (Cavazzuti, Fischer, Welch, Belli, & Winston, 1983; Laffond et al., 2012). Significant deficits compared to controls were also found in a set-shifting task in our neuropsychological study on patients with hypothalamic involvement (Ozyurt, Thiel, et al., 2014) and

EP

the subsample of patients the current analysis is based on. Note that within the hypothalamus the dorsomedial nucleus mainly connects to the medial thalamus, which has strong reciprocal connections to the orbital/medial prefrontal cortex. These pathways are assumed to be involved in behavioral

AC C

control, including higher cognitive functioning (Lemaire et al., 2011). In the current study, the posterior/mediodorsal thalamus and the orbital/medial prefrontal cortex revealed reduced grey matter volumes in patients compared to controls. However, regression analyses with data from the setshifting task of our neuropsychological study did not yield significant results, which is probably due to reduced power with our small sample size. Abnormalities in similar regions of the posterior/mediodorsal thalamus and orbital/medial prefrontal cortex were also found in the fMRI study the current subsample is based on (Ozyurt, Lorenzen, et al., 2014). In this study, task-induced deactivation in patients was reduced in the medial prefrontal node of the default mode network during memory recognition. Based on the role of the thalamus in modulating (depending on external demands) activation changes in the default-mode network (Gusnard, Raichle, 15

ACCEPTED MANUSCRIPT

& Raichle, 2001; McKiernan, Kaufman, Kucera-Thompson, & Binder, 2003), we investigated its coupling with the area of deactivation failure. As hypothesized, we found an altered functional connectivity between the thalamus and the orbital/medial prefrontal cortex in patients, which was assumed to be related to the deactivation failure in the latter brain area (Ozyurt, Lorenzen, et al., 2014). To relate these functional changes with structural changes found in the current VBM study, we

RI PT

performed regression analyses with beta values obtained from the orbital/medial prefrontal cortex cluster (Ozyurt, Lorenzen, et al., 2014). Lower task-induced deactivation was correlated with higher white matter reductions in the ansa lenticularis, which contributes to the connection of the thalamus with the orbitofrontal lobes. Thus the altered functional connectivity found in the fMRI study and assumed to induce distal functional changes in orbital/medial prefrontal cortex may be due to a

SC

reduction of connecting fibers between the thalamus and orbitofrontal cortex.

M AN U

4.4. Limitations

The current study is limited by the small sample size, which is due to the low prevalence of the tumor and to inclusion criteria such as hypothalamic involvement, well-preserved visual abilities and the absence of any MRI exclusion criteria. Hence it would be important to replicate the current results with a larger sample size but also with additional measures such as diffusion weighted MRI to better assess changes in white matter volumes. The inclusion of a non-limbic tumor group would also add to

TE D

better clarify the relative contribution of mechanisms leading to structural changes outside the area of tumor growth (e.g., transneuronal degeneration, effects of the tumor and surgical procedures,

AC C

EP

complications).

16

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

5. Conclusion

To the best of our knowledge, this is the first study to identify subtle volumetric changes outside the area of tumor growth in craniopharyngioma patients with hypothalamic tumor involvement. Changes in grey and white matter volumes were largely confined to areas and pathways of two limbic

TE D

subsystems: a) the posterior limbic system involving the mammillary bodies of the hypothalamus and subserving episodic memory and b) the anterior limbic system involving the extramammillary part of the hypothalamus and subserving social-emotional functions. Noteworthy, the two major domains of behavioral impairments in craniopharyngioma patients, episodic memory and socio-emotional skills, strikingly correspond to the functional range of each of the two subsystems. Thus, our results indicate

EP

that behavioral abnormalities in patients, which are partly not explicable by a hypothalamic lesion alone, are at least partly due to distal lesion effects in limbic subsystems.

AC C

Patients with childhood-onset craniopharyngioma and hypothalamic involvement may greatly differ in their clinical presentation, not only with respect to the severity of their cognitive and social-emotional impairments, but also with respect to other symptoms, like hypothalamic obesity or daytime sleepiness (Ozyurt, Thiel, et al., 2014; Roth et al., 2015). Importantly, healthy states in all these domains were shown to rely on intact limbic circuits (Betley & Sternson, 2015; Killgore, Schwab, Kipman, DelDonno, & Weber, 2012; Rolls, 2015; Vann, 2010). Based on these studies and results on volumetric changes in our study, we propose that the severity of many symptoms may not only be related to the site and exact nature of hypothalamic involvement but also to structural and functional changes in associated networks (Fornito et al., 2015).

17

ACCEPTED MANUSCRIPT

Acknowledgments

Table and Figure Captions

M AN U

SC

RI PT

Funding: This word was supported by the German Childhood Cancer Foundation [DKS 2014.13] and the Research Pool of the Faculty of Medicine and Health Sciences, Carl von Ossietzky Universität, Oldenburg, Germany.

Figure 1. A-B. Sagittal and axial views showing significantly higher amounts of grey matter volumes in controls compared to childhood-onset craniopharyngioma patients (i.e. reduced grey matter

TE D

volumes in patients). Activations are displayed at pFWE <.05 on cluster level C. As shown on a sagittal view, a small volume correction revealed a cluster of reduced grey matter volume in patients in the right posterior and medial thalamus. The color bar depicts fMRI signal level (T-scores). Ac=nucleus accumbens, Cd=caudate nucleus, CG=cingulate gyrus, OFC =orbitofrontal/medial frontal cortex, Pu=Putamen.

EP

Figure 2. Areas of higher amounts of white matter volumes in controls compared to childhood-onset craniopharyngioma patients (i.e. reduced white matter volumes in patients) . Activations are displayed

AC C

at pFWE .05 on cluster level. The color bar depicts fMRI signal level (T-scores). al=ansa lenticularis, Amg=amygdala, fx/st=fornix (crus)/stria terminalis, Hy=hypothalamus, ifo=inferior fronto-occipital fasciculus, lenf=lenticular fasciculus, MB=mammillary bodies, mfb=medial forebrain bundle, Opt=optic tract, Pu=Putamen; ss=sagittal stratum, unc=uncinate fasciculus. Figure 3. A Regression analysis with craniopharyngioma patients’ data on long-term memory performance resulted in a significant grey matter cluster in left posterior cingulate cortex. As shown by the regression plot, better task performance was correlated with a higher grey matter volume in this cluster B. Regression analysis with beta values from the previous fMRI study the current subsample is based on (Ozyurt, Lorenzen, et al., 2014). Successful task induced deactivation in orbital/medial prefrontal cortex in the fMRI study correlated with increased white matter volumes in the ansa lenticularis (al) and superior longitudinal fasciculus (slf) in the current VBM study, indicating a higher 18

ACCEPTED MANUSCRIPT

integrity in pathway connecting the thalamus with the orbital/medial prefrontal cortex. The color bars depict fMRI signal level (T-scores). Figure 4. Main connections within limbic subsystems. A. posterior limbic subsystem supporting episodic memory and B. anterior limbic subsystem supporting social-emotional skills.

RI PT

Table 1: Demographic and clinical characteristics depicted for the control and craniopharyngioma patient groups. The Fisher’s exact test was used to test for possible between-group differences in gender. For between-group differences of all other variables the Mann-Whitney U-test was used. All tests were performed two sided.

SC

Table 2: Craniopharyngioma patient history and current clinical status I

M AN U

Table 3: Craniopharyngioma patient history and current clinical status II

Appendix

Figure A1: Coronal, sagittal and axial views on regions affected by craniopharyngiomas and their removal are depicted from left to right. The sellar/parasellar area is shown within a red circle. A.

TE D

Structural images of a 17 years old healthy female control subject. B. Structural images of a 17 years old female patient with hypothalamic involvement grade 1 (P02, for patient characteristics see Table 2). This patient presents with a lesion of the hypothalamus at the region of the floor of the third ventricle (coronal view). The mammillary bodies of the hypothalamus appear to be intact (axial view). C. Structural images of a 16.6 years old female patient with hypothalamic involvement grade 2 (P01).

EP

This patient has a lesion of the left hypothalamus at the region of the floor of the third ventricle (coronal view). The mammillary bodies of the hypothalamus appear to be reduced compared to previous images (axial view). D. Structural images of a 27 years old male patient with childhood-onset

AC C

craniopharyngioma and hypothalamic involvement grade 2 (P15). This patient has a large lesion of the hypothalamus at the region of the floor of the third ventricle (coronal view). The right mammillary body of the hypothalamus appears to be reduced in volume (axial view).

19

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Literature

AC C

EP

TE D

Aggleton, J. P. (2014). Looking beyond the hippocampus: old and new neurological targets for understanding memory disorders. Proc Biol Sci, 281(1786). doi: 10.1098/rspb.2014.0565 Aggleton, J. P., & Brown, M. W. (1999). Episodic memory, amnesia, and the hippocampal-anterior thalamic axis. Behav Brain Sci, 22(3), 425-444; discussion 444-489. Aggleton, J. P., & Nelson, A. J. (2015). Why do lesions in the rodent anterior thalamic nuclei cause such severe spatial deficits? Neurosci Biobehav Rev, 54, 131-144. doi: 10.1016/j.neubiorev.2014.08.013 Amodio, D. M., & Frith, C. D. (2006). Meeting of minds: the medial frontal cortex and social cognition. Nat Rev Neurosci, 7(4), 268-277. doi: 10.1038/nrn1884 Ashburner, J., & Friston, K. J. (2001). Why voxel-based morphometry should be used. Neuroimage, 14(6), 1238-1243. doi: 10.1006/nimg.2001.0961 Baron, J. C., Levasseur, M., Mazoyer, B., Legault-Demare, F., Mauguiere, F., Pappata, S., . . . et al. (1992). Thalamocortical diaschisis: positron emission tomography in humans. J Neurol Neurosurg Psychiatry, 55(10), 935-942. Beckmann, M., Johansen-Berg, H., & Rushworth, M. F. (2009). Connectivity-based parcellation of human cingulate cortex and its relation to functional specialization. J Neurosci, 29(4), 11751190. doi: 10.1523/JNEUROSCI.3328-08.2009 Beracochea, D. J., Micheau, J., & Jaffard, R. (1995). Alteration of cortical and hippocampal cholinergic activities following lesion of the mammillary bodies in mice. Brain Res, 670(1), 53-58. Betley, J. N., & Sternson, S. M. (2015). Applying the Brakes: When to Stop Eating. Neuron, 88(3), 440441. doi: 10.1016/j.neuron.2015.10.034 Brown, J. L., Hunsperger, R. W., & Rosvold, H. E. (1969). Defence, attack, and flight elicited by electrical stimulation of the hypothalamus of the cat. Exp Brain Res, 8(2), 113-129. Catani, M., Dell'acqua, F., & Thiebaut de Schotten, M. (2013). A revised limbic system model for memory, emotion and behaviour. Neurosci Biobehav Rev, 37(8), 1724-1737. doi: 10.1016/j.neubiorev.2013.07.001

20

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Cavazzuti, V., Fischer, E. G., Welch, K., Belli, J. A., & Winston, K. R. (1983). Neurological and psychophysiological sequelae following different treatments of craniopharyngioma in children. J Neurosurg, 59(3), 409-417. doi: 10.3171/jns.1983.59.3.0409 Dalgleish, T. (2004). The emotional brain. Nat Rev Neurosci, 5(7), 583-589. doi: 10.1038/nrn1432 de Vile, C. J., Grant, D. B., Hayward, R. D., Kendall, B. E., Neville, B. G., & Stanhope, R. (1996). Obesity in childhood craniopharyngioma: relation to post-operative hypothalamic damage shown by magnetic resonance imaging. J Clin Endocrinol Metab, 81(7), 2734-2737. doi: 10.1210/jcem.81.7.8675604 De Witte, L., Brouns, R., Kavadias, D., Engelborghs, S., De Deyn, P. P., & Marien, P. (2011). Cognitive, affective and behavioural disturbances following vascular thalamic lesions: a review. Cortex, 47(3), 273-319. doi: 10.1016/j.cortex.2010.09.002 Dusoir, H., Kapur, N., Byrnes, D. P., McKinstry, S., & Hoare, R. D. (1990). The role of diencephalic pathology in human memory disorder. Evidence from a penetrating paranasal brain injury. Brain, 113 ( Pt 6), 1695-1706. Duvernoy, H. M. (2003). The human Brain. Surface, blood supply, and three-dimensional sectional anatomy. (2nd Ed. ed.). Wien, New York: Springer. Figley, C. R., Asem, J. S., Levenbaum, E. L., & Courtney, S. M. (2016). Effects of Body Mass Index and Body Fat Percent on Default Mode, Executive Control, and Salience Network Structure and Function. Front Neurosci, 10, 234. doi: 10.3389/fnins.2016.00234 Fjalldal, S., Holmer, H., Rylander, L., Elfving, M., Ekman, B., Osterberg, K., & Erfurth, E. M. (2013). Hypothalamic involvement predicts cognitive performance and psychosocial health in longterm survivors of childhood craniopharyngioma. J Clin Endocrinol Metab, 98(8), 3253-3262. doi: 10.1210/jc.2013-2000 Flitsch, J., Muller, H. L., & Burkhardt, T. (2011). Surgical strategies in childhood craniopharyngioma. Front Endocrinol (Lausanne), 2, 96. doi: 10.3389/fendo.2011.00096 Fornito, A., Zalesky, A., & Breakspear, M. (2015). The connectomics of brain disorders. Nat Rev Neurosci, 16(3), 159-172. doi: 10.1038/nrn3901 Garnett, M. R., Puget, S., Grill, J., & Sainte-Rose, C. (2007). Craniopharyngioma. Orphanet J Rare Dis, 2, 18. doi: 10.1186/1750-1172-2-18 Gusnard, D. A., Raichle, M. E., & Raichle, M. E. (2001). Searching for a baseline: functional imaging and the resting human brain. Nat Rev Neurosci, 2(10), 685-694. doi: 10.1038/35094500 Hare, T. A., O'Doherty, J., Camerer, C. F., Schultz, W., & Rangel, A. (2008). Dissociating the role of the orbitofrontal cortex and the striatum in the computation of goal values and prediction errors. J Neurosci, 28(22), 5623-5630. doi: 10.1523/JNEUROSCI.1309-08.2008 Helmstaedter, C., Lendt, M., & Lux, S. (2001). Verbaler Lern- und Merkfähigkeitstest. Göttingen: Beltz Test. Hildebrandt, H., Muller, S., Bussmann-Mork, B., Goebel, S., & Eilers, N. (2001). Are some memory deficits unique to lesions of the mammillary bodies? J Clin Exp Neuropsychol, 23(4), 490-501. doi: 10.1076/jcen.23.4.490.1234 Hoffmann, A., Warmth-Metz, M., Gebhardt, U., Pietsch, T., Pohl, F., Kortmann, R. D., . . . Muller, H. L. (2014). Childhood craniopharyngioma - changes of treatment strategies in the trials KRANIOPHARYNGEOM 2000/2007. Klin Padiatr, 226(3), 161-168. doi: 10.1055/s-00341368785 Hornak, J., Bramham, J., Rolls, E. T., Morris, R. G., O'Doherty, J., Bullock, P. R., & Polkey, C. E. (2003). Changes in emotion after circumscribed surgical lesions of the orbitofrontal and cingulate cortices. Brain, 126(Pt 7), 1691-1712. doi: 10.1093/brain/awg168 Hua, C., Shulkin, B. L., Indelicato, D. J., Li, Y., Li, X., Boop, F. A., & Merchant, T. E. (2015). Postoperative cerebral glucose metabolism in pediatric patients receiving proton therapy for craniopharyngioma. J Neurosurg Pediatr, 1-7. doi: 10.3171/2015.4.PEDS159 Kapur, N., Crewes, H., Wise, R., Abbott, P., Carter, M., Millar, J., & Lang, D. (1998). Mammillary body damage results in memory impairment but not amnesia. Neurocase, 4(6), 509-517. 21

ACCEPTED MANUSCRIPT

TE D

M AN U

SC

RI PT

Killgore, W. D., Schwab, Z. J., Kipman, M., DelDonno, S. R., & Weber, M. (2012). Voxel-based morphometric gray matter correlates of daytime sleepiness. Neurosci Lett, 518(1), 10-13. doi: 10.1016/j.neulet.2012.04.029 Kringelbach, M. L., & Rolls, E. T. (2003). Neural correlates of rapid reversal learning in a simple model of human social interaction. Neuroimage, 20(2), 1371-1383. doi: 10.1016/S10538119(03)00393-8 Kullmann, S., Heni, M., Linder, K., Zipfel, S., Haring, H. U., Veit, R., . . . Preissl, H. (2014). Resting-state functional connectivity of the human hypothalamus. Hum Brain Mapp, 35(12), 6088-6096. doi: 10.1002/hbm.22607 Laffond, C., Dellatolas, G., Alapetite, C., Puget, S., Grill, J., Habrand, J. L., . . . Chevignard, M. (2012). Quality-of-life, mood and executive functioning after childhood craniopharyngioma treated with surgery and proton beam therapy. Brain Inj, 26(3), 270-281. doi: 10.3109/02699052.2011.648709 Lemaire, J. J., Frew, A. J., McArthur, D., Gorgulho, A. A., Alger, J. R., Salomon, N., . . . De Salles, A. A. (2011). White matter connectivity of human hypothalamus. [Research Support, Non-U.S. Gov't]. Brain Res, 1371, 43-64. doi: 10.1016/j.brainres.2010.11.072 Lemaire, J. J., Nezzar, H., Sakka, L., Boirie, Y., Fontaine, D., Coste, A., . . . De Salles, A. (2013). Maps of the adult human hypothalamus. Surg Neurol Int, 4(Suppl 3), S156-163. doi: 10.4103/21527806.110667 Maguire, E. A. (2001). The retrosplenial contribution to human navigation: a review of lesion and neuroimaging findings. Scand J Psychol, 42(3), 225-238. McKiernan, K. A., Kaufman, J. N., Kucera-Thompson, J., & Binder, J. R. (2003). A parametric manipulation of factors affecting task-induced deactivation in functional neuroimaging. J Cogn Neurosci, 15(3), 394-408. doi: 10.1162/089892903321593117 Morgane, P. J., Galler, J. R., & Mokler, D. J. (2005). A review of systems and networks of the limbic forebrain/limbic midbrain. Prog Neurobiol, 75(2), 143-160. doi: 10.1016/j.pneurobio.2005.01.001 Muller, H. L. (2008). Childhood craniopharyngioma. Recent advances in diagnosis, treatment and follow-up. [Research Support, Non-U.S. Gov't Review]. Horm Res, 69(4), 193-202. doi: 10.1159/000113019 Muller, H. L. (2010). Childhood craniopharyngioma--current concepts in diagnosis, therapy and follow-up. [Research Support, Non-U.S. Gov't

AC C

EP

Review]. Nat Rev Endocrinol, 6(11), 609-618. doi: 10.1038/nrendo.2010.168 Muller, H. L., Gebhardt, U., Teske, C., Faldum, A., Zwiener, I., Warmuth-Metz, M., . . . Study Committee of, K. (2011). Post-operative hypothalamic lesions and obesity in childhood craniopharyngioma: results of the multinational prospective trial KRANIOPHARYNGEOM 2000 after 3-year follow-up. Eur J Endocrinol, 165(1), 17-24. doi: 10.1530/EJE-11-0158 Oishi, K., Faria, A., van Zijl, P. C. M., & Mori, S. (2011). MRI Atlas of Human White Matter (2nd Ed. ed.). Amsterdam, Boson: Elsevier. Ongur, D., An, X., & Price, J. L. (1998). Prefrontal cortical projections to the hypothalamus in macaque monkeys. J Comp Neurol, 401(4), 480-505. Ozyurt, J., Lorenzen, A., Gebhardt, U., Warmuth-Metz, M., Muller, H. L., & Thiel, C. M. (2014). Remote effects of hypothalamic lesions in the prefrontal cortex of craniopharygioma patients. Neurobiol Learn Mem, 111, 71-80. doi: 10.1016/j.nlm.2014.03.007 Ozyurt, J., Muller, H. L., & Thiel, C. M. (2015). A systematic review of cognitive performance in patients with childhood craniopharyngioma. J Neurooncol, 125(1), 9-21. doi: 10.1007/s11060-015-1885-z Ozyurt, J., Thiel, C. M., Lorenzen, A., Gebhardt, U., Calaminus, G., Warmuth-Metz, M., & Muller, H. L. (2014). Neuropsychological outcome in patients with childhood craniopharyngioma and hypothalamic involvement. J Pediatr, 164(4), 876-881 e874. doi: 10.1016/j.jpeds.2013.12.010 22

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Peters, J., & Buchel, C. (2010). Neural representations of subjective reward value. Behav Brain Res, 213(2), 135-141. doi: 10.1016/j.bbr.2010.04.031 Pierre-Kahn, A., Recassens, C., Pinto, G., Thalassinos, C., Chokron, S., Soubervielle, J. C., . . . Sainte Rose, C. (2005). Social and psycho-intellectual outcome following radical removal of craniopharyngiomas in childhood. A prospective series. Childs Nerv Syst, 21(8-9), 817-824. doi: 10.1007/s00381-005-1205-6 Puget, S., Garnett, M., Wray, A., Grill, J., Habrand, J. L., Bodaert, N., . . . Sainte-Rose, C. (2007). Pediatric craniopharyngiomas: classification and treatment according to the degree of hypothalamic involvement. J Neurosurg, 106(1 Suppl), 3-12. doi: 10.3171/ped.2007.106.1.3 Rempel-Clower, N. L., & Barbas, H. (1998). Topographic organization of connections between the hypothalamus and prefrontal cortex in the rhesus monkey. J Comp Neurol, 398(3), 393-419. Risold, P. Y., Thompson, R. H., & Swanson, L. W. (1997). The structural organization of connections between hypothalamus and cerebral cortex. Brain Res Brain Res Rev, 24(2-3), 197-254. Rolland-Cachera, M. F., Cole, T. J., Sempe, M., Tichet, J., Rossignol, C., & Charraud, A. (1991). Body Mass Index variations: centiles from birth to 87 years. Eur J Clin Nutr, 45(1), 13-21. Rolls, E. T. (2015). Limbic systems for emotion and for memory, but no single limbic system. Cortex, 62, 119-157. doi: 10.1016/j.cortex.2013.12.005 Rorden, C., & Karnath, H. O. (2004). Using human brain lesions to infer function: a relic from a past era in the fMRI age? Nat Rev Neurosci, 5(10), 813-819. doi: 10.1038/nrn1521 Roth, C. L., Eslamy, H., Werny, D., Elfers, C., Shaffer, M. L., Pihoker, C., . . . Dobyns, W. B. (2015). Semiquantitative analysis of hypothalamic damage on MRI predicts risk for hypothalamic obesity. Obesity (Silver Spring), 23(6), 1226-1233. doi: 10.1002/oby.21067 Roth, C. L., Hunneman, D. H., Gebhardt, U., Stoffel-Wagner, B., Reinehr, T., & Muller, H. L. (2007). Reduced sympathetic metabolites in urine of obese patients with craniopharyngioma. Pediatr Res, 61(4), 496-501. doi: 10.1203/pdr.0b013e3180332cd6 Schoenbaum, G., Roesch, M. R., Stalnaker, T. A., & Takahashi, Y. K. (2009). A new perspective on the role of the orbitofrontal cortex in adaptive behaviour. Nat Rev Neurosci, 10(12), 885-892. doi: 10.1038/nrn2753 Squire, L. R., Amaral, D. G., Zola-Morgan, S., Kritchevsky, M., & Press, G. (1989). Description of brain injury in the amnesic patient N.A. based on magnetic resonance imaging. Exp Neurol, 105(1), 23-35. Tanaka, Y., Miyazawa, Y., Akaoka, F., & Yamada, T. (1997). Amnesia following damage to the mammillary bodies. Neurology, 48(1), 160-165. Teuber, H. L., Milner, B., & Vaughan, H. G. (1968). Persistent Anterograde Amnesia after Stab Wound of Basal Brain. Neuropsychologia, 6(3), 267-&. doi: Doi 10.1016/0028-3932(68)90025-0 Toni, R., Malaguti, A., Benfenati, F., & Martini, L. (2004). The human hypothalamus: a morphofunctional perspective. J Endocrinol Invest, 27(6 Suppl), 73-94. Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N., . . . Joliot, M. (2002). Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. Neuroimage, 15(1), 273-289. doi: 10.1006/nimg.2001.0978 Vann, S. D. (2009). Gudden's ventral tegmental nucleus is vital for memory: re-evaluating diencephalic inputs for amnesia. Brain, 132(Pt 9), 2372-2384. doi: 10.1093/brain/awp175 Vann, S. D. (2010). Re-evaluating the role of the mammillary bodies in memory. Neuropsychologia, 48(8), 2316-2327. doi: 10.1016/j.neuropsychologia.2009.10.019 Vann, S. D., & Aggleton, J. P. (2004). The mammillary bodies: two memory systems in one? [Review]. Nat Rev Neurosci, 5(1), 35-44. doi: 10.1038/nrn1299 Vann, S. D., & Albasser, M. M. (2009). Hippocampal, retrosplenial, and prefrontal hypoactivity in a model of diencephalic amnesia: Evidence towards an interdependent subcortical-cortical memory network. Hippocampus, 19(11), 1090-1102. doi: 10.1002/hipo.20574

23

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Vann, S. D., Denby, C., Love, S., Montaldi, D., Renowden, S., & Coakham, H. B. (2008). Memory loss resulting from fornix and septal damage: impaired supra-span recall but preserved recognition over a 24-hour delay. Neuropsychology, 22(5), 658-668. doi: 10.1037/a0012542 Warmuth-Metz, M., Gnekow, A. K., Muller, H., & Solymosi, L. (2004). Differential diagnosis of suprasellar tumors in children. Klin Padiatr, 216(6), 323-330. doi: 10.1055/s-2004-832358 Weiß, R. H. (1987). Grundintelligenztest Skala 2 CFT-20. Göttingen: Verlag für Psychologie, Hogrefe. Yu, C., Zhou, Y., Liu, Y., Jiang, T., Dong, H., Zhang, Y., & Walter, M. (2011). Functional segregation of the human cingulate cortex is confirmed by functional connectivity based neuroanatomical parcellation. Neuroimage, 54(4), 2571-2581. doi: 10.1016/j.neuroimage.2010.11.018 Zada, G., Kintz, N., Pulido, M., & Amezcua, L. (2013). Prevalence of neurobehavioral, social, and emotional dysfunction in patients treated for childhood craniopharyngioma: a systematic literature review. PLoS One, 8(11), e76562. doi: 10.1371/journal.pone.0076562

24

ACCEPTED MANUSCRIPT

Table 1: Demographic and clinical characteristics depicted for the control and patient groups. The Fisher’s exact test was used to test for possible between-group differences in gender. For betweengroup differences of all other variables the Mann-Whitney U-test was used. All tests were performed two sided.

Control

Craniopharyngioma

18 8/10

11 6/5

IQR 6.0 21

Median 17.4 101

5

9

5

33 34

15 8

AC C

EP

TE D

IQR=Interquartile range *CFT-20-R: Culture Fair Intelligence Test **State-Trait Anxiety Inventory

.71

IQR 8.8 32

.44 .86

11

.87

SC

Median 17.2 103

30 32

M AN U

Age IQ* Beck Depression Inventory Trait anxiety** State anxiety**

RI PT

p-value N Gender (male/female)

16 6

.33 .20

ACCEPTED MANUSCRIPT

RI PT

Table 2: Patient history and current clinical status I Age at diagnosis/ study (yrs.)

Surgery1

Recur / Surgery2

HI Grade preop

HI Grade postop

Tumor Volume cm³

BMI SDS preop

BMI SDS postop

F-01

12.3 / 16.6

2

1/5

2

2

33.83

8.27**

7.77**

F-02

08.9 / 17.0

1

-

2

1

1.15

-2.15

4.04**

M-03

07.6 / 16.8

1

-

1

0

18.20

0.80

1.8

F-04

10.0 / 17.1

1

-

2

F-07

08.0 / 14.6

2

1/5

2

M -09

11.4 / 17.4

2

-

2

M -10

15.8 / 26.1

1

2/3

1

M -11

14.0 / 23.0

1

-

F-12

10.5 / 18.2

2

-

M -13

17.6 / 25.6

2

1/5

M -15

08.3 / 27.2

3

1/4

M AN U

SC

Case

11.03

-1.15

0.93

1

5.40

2.50*

-0.97

2

2.93

0.93

2.63*

2

3.64

0.40

6.15**

2

2

19.64

1.60

5.48**

1

1

4.37

2.67*

5.07**

2

2

28.56

1.40

6.42**

n.a.

2

0.3

2.42*

5.97**

EP

TE D

2

AC C

Case: M=male, F= female; patient number Surgery1 (primary surgical approach): 1= transcranial, complete; 2= transcranial, incomplete; 3= not available, incomplete Recur/Surgery2 (surgical interventions after progression/relapse): 1=progression; 2=relapse / 3=transcranial, complete; 4=transcranial, incomplete; 5=no surgery after progression/relapse HI Grade: Preoperative (preop) and postoperative (postop) hypothalamic involvement (HI), resp. 0=no hypothalamic involvement/lesion; 1=involvement/lesion of the anterior hypothalamus; 2= involvement/lesion of the anterior and posterior hypothalamic area (the latter involving the mammillary bodies and the posterior hypothalamic area beyond) Tumor volume: Tumor size in patient M -15 was only measured two-dimensional BMI SDS: standard deviation score (SDS) for the body mass index preoperative (preop) and postoperative (postop). *Moderate obesity (SDS between 2 and 4); **severe obesity (SDS >4) n.a.= not available

ACCEPTED MANUSCRIPT

Table 3. Patient history and current clinical status II Surgical intervention (N)

Frontal Lobe (size of the surgical lesion)

Thalamus

Striatum

Signs of increased pressure

Complications

Irradiation

F-01

right pterional (1)

normal

normal

normal

-

-

external photon after primary surgery

F-02

right pterional (1)

normal

normal

normal

-

M-03

right pterional (1)

normal

normal

normal

F-04

right pterional (1)

right frontolateral (small; i.e. max. 6 mm)

normal

normal

-

-

F-07

right pterional (1)

normal

normal

normal

-

-

M -09

transcallosal (1)

right frontal, near midline (small; max. 6 mm)

normal

normal

-

-

M -10

right pterional (2)

right frontolateral (medium; i.e. max. 10mm)

normal

normal

-

-

M -11

transcallosal (1)

defect in corpus callosum (large; 16 mm)

normal

normal

lateral ventricle enlarged

hydrocephalus

M -13

subfrontal (1)

frontobasal, midline (large; >10 mm)

M -15

right pterional (2)

right frontolateral (2x small; i.e. max. 6mm)

normal

SC -

M AN U

TE D

EP

right pterional (1)

-

-

normal

-

-

normal

normal

-

-

right
infarction, right

-

right basal ganglion infarct

AC C

F-12

right frontolateral (medium; i.e. max. 10mm)

RI PT

Case

external photon after progression

external photon after primary surgery gamma knife after progression external photon after second surgery

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT