Structural 'connectomic' alterations in the limbic system of multiple sclerosis patients with major...

Multiple Sclerosis Journal

1 –10

DOI: 10.1177/ 1352458514558474

© The Author(s), 2014. Reprints and permissions: http://www.sagepub.co.uk/ journalsPermissions.nav

MULTIPLESCLEROSIS MSJJOURNAL

http://msj.sagepub.com 1

IntroductionMajor Depression (MD) is a common psychiatric dis-order in patients with multiple sclerosis (MS), the most frequent neurological disease in young people.1 Approximately 50–60% of MS patients may display recurrent episodes of MD that have a negative impact on their quality of life and may lead to devastating outcomes, as death by suicide.1 In clinical settings, MD in MS tends to be overlooked and attrib-uted to the overall neurological disability, rather than to specific brain changes; however, evidence is accumulating that MD in MS has a distinct neurobio-logical basis and does not only reflect a maladaptive psychological response to a chronic disease such as MS. Three different lines of research support this view.

First, clinical studies showed that MD in MS is not necessarily associated with disease duration and severity, but might also manifest itself at early stages of the disease.2 Second, MS-related immunological markers (e.g. T-cell activation) may not only be cor-related with motor or cognitive symptoms, but also with mood disturbances as MD.3 Third, neuroimaging studies revealed specific brain alterations in MS patients with MD, relative to non-depressed MS patients. Specifically, increased lesion load and reduced fractional anisotropy (FA), two measures of white-matter damage, were reported in the frontal and temporal lobes of MS patients with MD, relative to non-depressed MS patients.4,5 Furthermore, our past functional magnetic resonance imaging

Structural ‘connectomic’ alterations in the limbic system of multiple sclerosis patients with major depression

Salvatore Nigro, Luca Passamonti, Roberta Riccelli, Nicola Toschi, Federico Rocca, Paola Valentino, Rita Nisticò, Francesco Fera and Aldo Quattrone

AbstractBackground: Major depression (MD) is a common psychiatric disorder in multiple sclerosis (MS). Despite the negative impact of MD on the quality of life of MS patients, little is known about its underly-ing brain mechanisms.Objective: We studied the whole-brain connectivity patterns that were associated with MD in MS. Alter-ations were mainly expected within limbic circuits.Methods: Diffusion tensor imaging data were collected in 20 MS patients with MD, 22 non-depressed MS patients and 16 healthy controls. We used deterministic tractography and graph analysis to study the white-matter connectivity patterns that characterized MS patients with MD.Results: We found that MD in MS was associated with increased local path length in the right hippocampus and right amygdala. Further analyses revealed that these effects were driven by an increased shortest distance between both the right hippocampus and right amygdala and a series of regions including the dorsolateral and ventrolateral prefrontal cortex, orbitofrontal cortex, sensory-motor cortices and supplementary motor area.Conclusion: Our data provide strong support for neurobiological accounts positing that MD in MS is mediated by abnormal ‘communications’ within limbic circuits. We also found evidence that MD in MS may be linked with connectivity alterations at the limbic-motor interface, a group of regions that trans-lates emotions into survival-oriented behaviors.

Keywords: Diffusion Tensor Imaging (DTI), connectome, graph analysis, limbic circuits, major depression, multiple sclerosis, white-matter connectivity patterns

Date received: 8 August 2014; accepted: 14 October 2014

Correspondence to: Luca Passamonti Institute of Bioimaging and Molecular Physiology, National Research Council 88100, Catanzaro, Italy [email protected]

Salvatore Nigro Luca Passamonti Federico Rocca Rita Nisticò Institute of Bioimaging and Molecular Physiology, National Research Council, Catanzaro, Italy

Roberta Riccelli Paola Valentino Francesco Fera Department of Medical and Surgical Sciences, ‘Magna Graecia’ University, Catanzaro, Italy

Aldo Quattrone Institute of Bioimaging and Molecular Physiology, National Research Council, Catanzaro, Italy/Department of Medical and Surgical Sciences, ‘Magna Graecia’ University, Catanzaro, Italy

Nicola Toschi Department of Biomedicine and Prevention, University of Rome ‘Tor Vergata’, Rome, Italy/AA Martinos Center for Biomedical Imaging, Harvard Medical School, Boston, MA, USA

558474MSJ0010.1177/1352458514558474Multiple Sclerosis JournalS Nigro, L Passamonti et al.research-article2014

Original Research Paper

by guest on December 23, 2014msj.sagepub.comDownloaded from

mailto:[email protected]

http://msj.sagepub.com/


2 http://msj.sagepub.com

(fMRI) study showed that even non-depressed MS patients, relative to healthy controls, may display abnormal activations in limbic circuits, when pro-cessing emotional stimuli.6

Nonetheless, the brain mechanisms underlying MD in MS remain scarcely characterized. In particular, it is still unknown whether connectivity alterations mediate the manifestation of MD in MS, a demyeli-nating disorder that typically leads to ‘communica- tion’ deficits within large-scale brain networks.7,8 Addressing this issue is important, because it may enhance our understanding of the pathogenesis of a common psychiatric disorder in MS and may have implications for developing novel diagnostic markers in MS patients with MD.

In the current study, we explored the whole-brain structural connectivity patterns in MS patients with MD, relative to non-depressed MS patients and healthy controls. To this end, we analyzed diffusion tensor imaging (DTI) data using the graph theory, a sophisticated mathematical approach that allows the quantitative representation of large-scale networks.7 Two previous DTI studies using graph analysis revealed altered connectivity patterns within frontal, temporal and occipital networks in MS patients, rela-tive to healthy controls7,8; however, it remains to be determined whether graph analysis is able to iden-tify specific connectivity patterns that distinguish between MS patients with MD and non-depressed MS patients. Overall, graph analysis offers a series of key metrics that can be readily employed to charac-terize large-scale connectivity patterns. For example, the efficiency of information routing within a neural network can be estimated by calculating the length of the shortest pathway linking two brain ‘nodes’. This ‘path length’ represents an index of the potential of information transfer across brain nodes that constitute a certain circuit. An increased path length can be thus considered as an index of pathological changes that alter the information flow in the brain. Graph analysis also enables the quantification of the network clus-tering, which represents the level of segregation of a circuit and its capacity for processing the informa-tion locally. Estimating path lengths and network clustering allows thus to study how MS-related dam-age (e.g. axonal fibers’ loss) alters the capacity to transfer and integrate inputs within large-scale neural networks.

Our main hypothesis was that MS patients with MD, relative to non-depressed MS patients and healthy controls, would have displayed increased path length

and reduced clustering across those brain nodes, which play a key role in the pathophysiology of MD.9 Specifically, we predicted increased path length and reduced clustering between the hippocampus and/or amygdala and various parts of the prefrontal cortex (PFC), a key region implicated in the ‘top-down’ con-trol of emotional behaviors.9 Both the hippocampus and amygdala are directly and reciprocally con-nected with the ventrolateral PFC (VLPFC) and orbitofrontal cortex (OFC), and indirectly linked with the dorsolateral PFC (DLPFC).9 Overall, these path-ways are thought to mediate the abnormal hormonal, viscero-motor, and cognitive functions that charac-terize MD (e.g. hypercortisolemia and memory deficits).10 Hence, MS-related pathological changes that perturb the information routing within these key circuits are likely to mediate the development of MD in MS.5,11

Participants and methods

ParticipantsWe recruited 42 patients with clinically definite relapsing–remitting MS (RRMS). All RRMS patients met the following criteria:

•• No history of traumatic brain injury, past or current history of substance abuse, or other coexisting medical conditions;

•• No clinical relapses for at least 6 months prior to study entry;

•• No assumption of steroids or disease-modifying therapy in the 3 months before recruitment;

•• Expanded Disability Status Scale (EDSS) rang-ing from 1–4.5. This EDSS range was chosen because it reflected a disease severity status that enabled MS patients to take part in further fMRI experiments, which will be described in other manuscripts.

All MS patients were interviewed with the Mood Disorder Section of the Structured Clinical Interview for the Diagnostic and Statistical Manual (DSM) of Mental Disorders, to assess for the presence of MD.12 Given the high frequency of fatigue symptoms in MS patients, the fatigue question was not considered when assessing the presence of MD in MS, and this enhanced the diagnostic validity of the DSM inter-view to assess for MD (i.e. five instead of four symp-toms were needed to reach a diagnosis of MD).4

Individual differences in fatigue levels were, nonethe-less, assessed using the Fatigue Severity Scale (FSS).



S Nigro, L Passamonti et al.


We found that 20 MS patients reached a DSM-IV diagnosis of MD, while 22 MS patients showed no evidence of clinically-relevant depression. Depression severity was also evaluated via the Beck Depression Inventory (BDI-II). All MS patients were either anti-depressant medication naïve or had undergone a washout period of at least five-half lives of a previ-ously prescribed antidepressant medication. We also recruited 16 healthy controls with no history of neuro-logical or psychiatric disease, and with a normal brain, as assessed by a structural MRI scanning. All participants underwent a number of neuropsychologi-cal tests (see Supplementary materials). Written informed consent was obtained from all participants. The study was approved by the Ethical Committee of the ‘Magna Graecia’ University of Catanzaro, Italy, according to the Helsinki Declaration.

Image acquisition, data pre-processing, and network definitionThis information is reported in Supplementary materials.

Graph analysesThe Brain Connectivity Toolbox (http://www.brain-connectivity-toolbox.net/) was used to compute the global and local network metrics.13 These analyses are described in detail in Supplementary materials.

Global network analysisThese analyses were aimed at characterizing the global (i.e. whole-brain) organization of the white-matter connectivity patterns in MS patients with MD, relative to non-depressed MS patients, and healthy controls (main effect of group). To this end, we calcu-lated the following graph measures:

•• the global clustering•• global path length•• small-worldness.

Local network analysesThese analyses were aimed at characterizing the local organization (i.e. in distinct interconnected nodes) of the white-matter connectivity patterns in MS patients with MD, relative to non-depressed MS patients, and healthy controls (main effect of group). To this end, we calculated the following graph measures:

•• the local clustering•• local path length •• the shortest distance between couples of brain

nodes.

Statistical analysesAnalyses of variance (ANOVA) were used to study the main effect of group for the following demo-graphic, clinical and neuroimaging variables:

•• Age;•• Neuropsychological scores;•• Anxiety levels;•• Total intracranial volume;•• Total grey-matter volume;•• Total white-matter volume;•• Global network metrics; •• Local network metrics.

We also employed 2-tailed t-tests to assess for statisti-cally-significant differences in disease duration and severity, and the total lesion load between MS patients with MD and non-depressed MS patients. Group dif-ferences in sex distribution were analyzed using a χ2 test. Global and local network analyses were cor-rected for multiple comparisons, using a stringent Bonferroni’s procedure (i.e. P < 0.05/90 brain regions). As there was a significant difference in sex distribution across groups, we re-ran the ANOVAs, assessing the main effect of group for the network measures by including sex as a covariate of no inter-est. Finally, we correlated global and local metrics with depression severity in the group of MS patients considered as a whole.

Results

ParticipantsTable 1 summarizes the demographic, clinical and neuropsychological data of the sample of MS patients and healthy controls included in this study. The MS and healthy control groups were compara-ble in terms of age and cognitive functions. A sig-nificant difference in sex distribution was found across groups, although this did not influence our main findings (Supplementary Tables S4 to S6). As reported in previous studies,14,15 MS patients with MD showed significantly higher levels of fatigue, relative to non-depressed MS patients; but again, this difference did not influence the main results (Supplementary Table S7). Relative to healthy con-trols, both groups of MS patients displayed reduced total white- and grey-matter volume, although the differences in total grey-matter volume across groups did not reach statistical significance (Table 1). MS patients with MD and non-depressed MS patients were comparable for disease duration and severity, total lesion load and distribution of white-matter lesions (see Supplementary materials).


http://www.brain-connectivity-toolbox.net/

http://www.brain-connectivity-toolbox.net/




These results are in line with previous findings and highlight the fact that MD in MS is not necessarily associated with disease severity and duration, or total lesion load.2 As expected,16 anxiety levels were significantly higher in MS patients with MD, relative to non-depressed MS patients and healthy controls.

Global network resultsThere was no main effect of group in the global clus-tering (Figure 1, left, and Supplementary Table S1 for statistics). In contrast, there was a significant main effect of group in the global path length that was driven by an increased path length in MS patients

(both depressed and non-depressed) relative to healthy controls (Figure 1, right panel, and Supplementary Table S1 for statistics).

Small world measuresNo main effect of group was found for small world measures (P’s > 0.12, Bonferroni corrected). This result is consistent with previous findings showing that MS patients and healthy controls display a similar small-world organization of the white-matter net-works.7 In particular, both MS patients and healthy controls exhibited a much higher γ coefficient and a similar λ coefficient (two measures to assess ‘small worldness’) relative to random networks (healthy

Table 1. Demographic, clinical, neuropsychological, and neuroimaging data of the sample.

HC (n = 16) MS-NODEP (n = 22)

MS-DEP (n = 20)

F/t/χ2-value p values

Demographic and clinical data

Gender (% female) 43 60 85 Χ2(2) = 7.33 P = 0.03

Age in years (range) 35.3 ± 7.2 (24–51)

31.5 ± 7.9 (20–48)

37.1 ± 8.9 (21–51)

F(2,55) = 2.6 P = 0.08

Duration of disease in years (range)

– 6.4 ±4.2 (0.8–14)

8.4 ± 6.5 (0.1–20)

t(40) = 1.2 P = 0.23

EDSS (range) – 2.3 ± 1.0 (1–4.5)

2.8 ± 0.9 (1.5–4.5)

t(40) = 1.9 P = 0.06

FSS (range) – 2.7 ± 1.4 (1–5.6)

5.3 ± 1.4 (1.4–7.6)

t(40) = 6.0 P < 0.0001

Neuroimaging data

Total gray matter volume (ml) 623.1 ± 49.1 604.3 ± 57.0 581.7 ± 61.6 F(2,55) = 2.4 P = 0.10

Total white matter volume (ml) 623.05 ± 49.1 460.0 ± 52.6 473.2 ± 56.0 F(2,55) = 5.5 P = 0.007

Intracranial volume (ml) 1352.7 ± 101.8 1296.4 ± 110.8 1275.0 ± 117.3 F(2,55) = 2.3 P = 0.11

Lesion load (ml) – 13.1 ± 11.5 12.3 ± 10.5 t(40) = 0.2 P = 0.84

Neuropsychological data

WLG (language) 19.8 ± 4.1 16.0 ± 3.4 14.6 ± 4.6 F(2,55) = 7.8 P = 0.001

LTS (verbal memory) 33.5 ± 12.3 38.8 ± 14.3 33.4 ± 13.6 F(2,55) = 0.4 P = 0.67

CLTR (verbal memory) 25.1 ± 10.8 28.4 ± 14.3 26.0 ± 13.5 F(2,55) = 0.1 P = 0.90

SRT D (verbal memory) 6.5 ± 3.0 6.7 ± 2.4 6.3 ± 2.5 F(2,55) = 0.0 P = 0.99

SPART I (spatial memory) 19.6 ± 4.7 18.4 ± 4.4 18.4 ± 4.1 F(2,55) = 0.5 P = 0.61

SPART D (spatial memory) 6.8 ± 2.1 6.2 ± 2.4 5.4 ± 2.2 F(2,55) = 1.7 P = 0.19

SDMT (information processing speed)

45.0 ± 16.0 41.3 ± 13.3 37.0 ± 11.4 F(2,55) = 1.6 P = 0.22

STROOP (executive control) 19.5 ± 5.8 20.0 ± 5.3 18.2 ± 7.5 F(2,55) = 0.4 P = 0.67

Global cognitive impairment (Z scores)

0.78 ± 0.28 0.77 ± 0.33 0.76 ± 0.26 F(2,55) = 0.0 P = 0.99

STAI – I (state anxiety) 34.9 ± 4.3 35.9 ± 5.1 51.4 ± 8.0 F(2,55) = 45.0 P < 0.0001STAI – II (trait anxiety) 34 (4.8) 34.7 (6.7) 50.9 (8.8) F(2,55) = 35.2 P < 0.0001

All data are expressed as mean ± standard deviation. Minimum and maximum values for demographic and clinical values (age, disease duration and EDSS) are also provided. CLTR: Consistent Long Term Storage; EDSS: Expanded Disability Status Scale; FSS: Fatigue Severity Scale; HC: healthy controls; LTS: Long Term Storage; MS: multiple sclerosis; MS-DEP: MS patients with major depression; MS-NODEP: non-depressed MS patients; SDMT: Symbol Digit Modality Test; SRT D: Selective Reminding Test Delayed; SPART D: Spatial Recall Test Delayed; SPART I: Spatial Recall Test Immediate; STAI: Spielberg Trait Anxiety Inventory; WLG: Word List Generation.





controls: γ = 3.308; λ = 1.197; MS patients with MD: γ = 3.488; λ = 1.20; MS patients with no depression: γ = 3.484;λ = 1.217 (Supplementary Table S1)).

Local network resultsANOVAs and t-tests were carried out for each brain node (90 in total). We found that the right hippocampus and right amygdala were the only two regions in the brain showing a highly significant main effect of group

for the local path length (Figure 2, and Table 2 for statis-tics). In contrast, no significant main effect of group was found for the local clustering coefficient (P’s > 0.11; Bonferroni corrected).

Shortest distance resultsRelative to non-depressed MS patients and healthy con-trols, MS patients with MD displayed increased shortest distance between the right hippocampus and right

Figure 1. Global network results (global clustering and global path length). No main effect of group was found in the global network clustering (left). In contrast, there was a significant main effect of group in the global path length (right) that was driven by an increased global path length in both groups of MS patients, relative to healthy controls. HC: Healthy controls; MS: multiple sclerosis; MS-DEP: MS patients with major depression; MS-NODEP: non-depressed MS patients. The data plot represents the mean values ± the standard deviation.

Figure 2. Local network results (local path length). The right hippocampus (a) and right amygdala (b) were the only two regions in the brain showing a significant main effect of the group in the local path length. In particular, MS patients with MD showed an increased local path length in the right hippocampus and right amygdala, relative to non-depressed MS patients and HCs. The data plot represents the mean values ± the standard deviation.HC: Healthy controls; MD: major depression; MS: multiple sclerosis; MS-DEP: MS patients with major depression; MS-NODEP: non-depressed MS patients.





amygdala, and a series of frontal and PFC areas includ-ing the DLPFC, VLPFC, OFC, sensory-motor cortices, and the supplementary motor area (SMA) (Figure 3 and Supplementary Tables S2-S3). Direct comparisons between MS patients with MD and non-depressed MS patients also showed significant effects in similar regions (i.e. DLPFC, VLPFC, OFC and SMA).

Correlation between connectivity metrics and depression severityIn the group of MS patients considered as a whole, BDI-II scores were not correlated with global network metrics (P’s > 0.69, Bonferroni corrected) or local clustering coefficient (P’s > 0.38, Bonferroni corrected); although a trend was found between BDI -II scores and the local path length in the right hippocampus (r = 0.49; P = 0.08, Bonferroni corrected) and inferior frontal gyru spars orbitalis (r = 0.51; P = 0.06, Bonferroni corrected).

DiscussionThis study provides new evidence that specific pat-terns of white-matter connectivity differentiate MS patients with MD from non-depressed MS patients and healthy controls. Specifically, the right hippocam-pus and right amygdala were the only two regions in the brain showing increased local path length in MS patients with MD, relative to non-depressed MS patients and healthy controls. Additional analyses revealed that these findings were driven by increased shortest distance between the right hippocampus and right amygdala and a series of PFC regions involved in emotional behavior (e.g. the OFC, VLPFC), cogni-tive functions (DLPFC, VLPFC), and motor control (SMA). The fact that no effects were found in global network measures such as global clustering and global path length provided further support to the notion that

MD in MS is mediated by abnormalities in specific neural networks that have been previously implicated in affective disorders.9 Contrary to our expectations, we did not find reduced local clustering in MS patients with MD: this may have depended on the presence of adaptive mechanisms that maintained a relatively nor-mal level of information processing across neighbor brain nodes.17

Two possibilities exist to explain why MS patients with MD showed increased local path length and shortest distance within limbic circuits. First, an MS-related damage of the bundles of fibers that directly connect the hippocampus (or amygdala) and PFC may have reduced the capacity to exchange information between these nodes. As a reduced number of axonal fibers were available, the infor-mation transferring from the hippocampus (or amygdala) to PFC and vice versa was reduced (Figure 4, Hypothesis 1). Alternatively, the increased local path length and shortest distance within limbic circuits may have been increased because inputs had to ‘travel’ via neural pathways that include additional intermediate areas (e.g. basal ganglia) before reaching their final destination (e.g. the PFC) (Figure 4, Hypothesis 2). Whichever the explana-tion, our data provided clear evidence that MS-related pathological changes significantly inter-fered with the information flow within PFC-hippocampus and PFC-amygdala circuits, and this may be important for mediating the development of MD in MS. It is also likely that an aberrant routing of inputs within key limbic circuits is responsible for the altered hormonal responses that have been described in MS patients with MD,11 although additional research simultaneosly assessing hormo-nal and connectivity measures is necessary to verify this hypothesis.

Table 2. Main effect of group in the local path length of the right hippocampus and amygdala, expressed as mean (SD).

Regions HC MS-NODEP MS-DEP F values T value of post-hoc test

HC versus MS-NODEP

HC versus MS-DEP

MS-NODEP versus MS-DEP

Right Hippocampus 0.015 (0.002) 0.015 (0.002) 0.019 (0.002) F(2,55)=17.9c t(36)=0.1ns t(34)=4.7b t(40)=5.4c

Right Amygdala 0.019 (0.002) 0.020 (0.003) 0.024 (0.003) F(2,55)=12.0a t(36)=0.9ns t(34)=4.4b t(40)=3.8a

F-values and t-values were obtained from ANOVAs and 2-sample t-tests, respectively. Subscripted numbers in parentheses represent the degree of freedom. A Bonferroni’s procedure was applied to correct for multiple comparisons.aP < 0.05bP < 0.01cP < 0.001 ANOVA: analysis of variance; HC: healthy controls; MS-DEP: multiple sclerosis patients with major depression; MS-NODEP: non-depressed multiple sclerosis patients; ns: not significant.





Of interest, relative to non-depressed MS patients and healthy controls, MS patients with MD also displayed increased shortest distance between the right hip-pocampus (and right amygdala) and the DLPFC. The DLPFC plays a critical role in modulating emotional behaviors,18 although few direct anatomical connec-tions exist between this PFC area and both the hip-pocampus and amygdala.19 Nonetheless, the DLPFC may influence the core emotional systems via its strong links with the OFC and VLPFC, two PFC areas that showed highly significant effects in the connec-tivity patterns with the right hippocampus and right amygdala in MS patients with MD, relative to non-depressed MS patients and healthy controls. Previous fMRI studies also showed that the DLPFC is critically involved in emotional reappraisal and voluntary sup-pression of negative feelings,20 while recent research

emphasized that the cross-talk between limbic and cognitive PFC regions may improve the ability to control affective behaviors.21 Overall, the altered cross-talk between the DLPFC and right hippocam-pus (or amygdala) in MS patients with MD may be mediated by an abnormal flow of information within the PFC, and between the PFC and key limbic areas in the medial temporal lobe.

Intriguing results were also obtained in the sen-sory-motor cortices and SMA, two critical regions for motor control.22 Specifically, we found an increased shortest distance between the right hip-pocampus and right amygdala, and sensory-motor cortices and SMA in the MS patients with MD, relative to non-depressed MS patients and healthy controls. Very recently, a DTI study in humans

Figure 3. Local network results (shortest distance). A series of frontal and prefrontal cortex (PFC) regions showed an increased shortest distance with the right hippocampus (a) and right amygdala (b), as a function of the main effect of group (0.03 ≤ P ≤ 0.0006, Bonferroni corrected, for the right hippocampus and 0.04 ≤ P ≤ 0.001, Bonferroni corrected, for the right amygdala). In particular, MS patients with MD showed, relative to non-depressed MS patients and healthy controls, an increased shortest distance between the right hippocampus and right amygdala and the dorsolateral PFC (i.e. middle frontal gyrus, superior frontal gyrus), ventrolateral PFC (i.e. inferior frontal gyrus), and a part of the orbitofrontal cortex (e.g. rectus gyrus). The data plot represents the mean values ± the standard deviation.HC: Healthy controls; MD: major depression; MS: multiple sclerosis; MS-DEP: MS patients with major depression; MS-NODEP: non-depressed MS patients.





reported direct anatomical connections between the amygdala and sensory-motor cortices and SMA.23 This finding not only confirmed previous data in the monkey brain,24 but also emphasized the significant influence of emotional networks on motor systems and vice versa. There is also evi-dence in rats that increasing levels of electrical stimulation of the amygdala resulted in progres-sively more active motor behaviors.25 Furthermore, an fMRI study in humans showed that processing signals of threat prompted the individuals to pre-pare a series of motor responses that were medi-ated by the interplay between the amygdala and motor cortices.26

Finally, we note that some results were lateralized to the right hemisphere (i.e. the local path length in the right hippocampus and right amygdala). Similarly, two earlier studies found alterations in the right but not left hippocampus or medial

temporal lobe areas in MS with MD,27,28 although a very recent review of the literature reported mixed effects in the laterality of previous findings (i.e., either on the left or right hemisphere or bilaterally).29

LimitationsThe first limitation of the study may be represented by the fact that female participants were not equally distributed across groups. Although this could reflect the fact that female patients with MS may show a higher prevalence of MD than males,30 an increased prevalence of MD in female patients with MS has not been consistently confirmed.31,32 Nonetheless, we did not find any evidence for an influence of gender on the main results (Supplementary Table S4-S6); hence, it was unlikely that differences in gender distribution across groups

Figure 4. Summary of study findings. Two possibilities exist to explain the increased shortest distance between the right hippocampus and right amygdala, and the PFC in MS patients with MD. The first possibility (Hypothesis 1) is that MS-related damage of the bundles of fibers that directly connect the hippocampus (or amygdala) and the PFC reduces the flow of information that these nodes can exchange, because there are less axonal fibers available to transport the information (this is represented by a thinner red line, connecting the limbic regions and the PFC). In other words, the shortest distance between the hippocampus (or amygdala) and the PFC is increased because less routes (i.e. axonal fibers) are available for input travelling. Alternatively (Hypothesis 2), inputs need to pass throughout indirect neural pathways to reach their final destination (e.g. the PFC). This results in an increased shortest distance between two brain nodes.HC: Healthy controls; MD: major depression; MS: multiple sclerosis; MS-DEP: MS patients with major depression; MS-NODEP: non-depressed MS patients; PFC: prefrontal cortex.





played a significant role in mediating the abnormal patterns of brain connectivity seen in MS patients with MD.

Second, we employed deterministic tractography to define the whole-brain white-matter networks. Although this method was repeatedly used in previous DTI studies assessing pathophysiological mechanisms underlying MS,7,8 this procedure can not resolve fiber crossings at a voxel level. To overcome this issue, oth-ers have proposed to employ probabilistic tractogra-phy,33 although this technique also presents its own disadvantages.34 Overall, it was beyond the scope of the present study to determine the best methodological approach to define white-matter networks, although we acknowledge that further studies using deterministic and probabilistic tractography are warranted to repli-cate current results.

Lastly, it remains to be determined whether the abnor-mal connectivity patterns that we identified in RRMS patients were generalized to different clinical pheno-types (e.g. secondary progressive MS or primary pro-gressive MS) or to other mood disturbances that have been described in MS patients (e.g. bipolar disor-der). Additional studies are necessary to address these important open issues.

ConclusionIn summary, this study provides a novel set of findings that significantly enhanced our knowledge of the neu-ral mechanisms underlying MD in MS. Specifically, we found that MD in MS was associated with aberrant information routing within key limbic networks, and at the interface between limbic and motor systems, a group of regions implicated in translating emotions into survival-oriented behaviors.

AcknowledgementsThe authors are grateful to the MS patients and con-trols who kindly participated in the experiment.

Conflict of interestThe authors declare that there is no conflict of interest.

FundingThis work was supported by the National Research Council.

NoteAuthors Nigro and Passamonti contributed equally to this work.

References 1. Janardhan V and Bakshi R. Quality of life in patients

with multiple sclerosis: The impact of fatigue and depression. J Neurol Sci 2002; 205: 51–58.

2. Gobbi C, Rocca MA, Riccitelli G, et al. Influence of the topography of brain damage on depression and fatigue in patients with multiple sclerosis. Mult Scler 2014; 20: 192–201.

3. Fassbender K, Schmidt R, Mossner R, et al. Mood disorders and dysfunction of the hypothalamic-pituitary-adrenal axis in multiple sclerosis: Association with cerebral inflammation. Arch Neurol 1998; 55: 66–72.

4. Feinstein A, Roy P, Lobaugh N, et al. Structural brain abnormalities in multiple sclerosis patients with major depression. Neurology 2004; 62: 586–590.

5. Feinstein A, O’Connor P, Akbar N, et al. Diffusion tensor imaging abnormalities in depressed multiple sclerosis patients. Mult Scler 2010; 16: 189–196.

6. Passamonti L, Cerasa A, Liguori M, et al. Neurobiological mechanisms underlying emotional processing in relapsing–remitting multiple sclerosis. Brain 2009; 132: 3380–3391.

7. Shu N, Liu Y, Li K, et al. Diffusion tensor tractography reveals disrupted topological efficiency in white matter structural networks in multiple sclerosis. Cereb Cortex 2011; 21: 2565–2577.

8. Li Y, Jewells V, Kim M, et al. Diffusion tensor imaging based network analysis detects alterations of neuroconnectivity in patients with clinically early relapsing–remitting multiple sclerosis. Hum Brain Mapp 2013; 34: 3376–3391.

9. Price JL and Drevets WC. Neural circuits underlying the pathophysiology of mood disorders. Trends Cogn Sci 2012; 16: 61–71.

10. Dedovic K, Duchesne A, Andrews J, et al. The brain and the stress axis: The neural correlates of cortisol regulation in response to stress. Neuroimage 2009; 47: 864–871.

11. Gold SM, Kern KC, O’Connor MF, et al. Smaller cornu ammonis 2–3/dentate gyrus volumes and elevated cortisol in multiple sclerosis patients with depressive symptoms. Biol Psychiatry 2010; 68: 553–559.

12. Steinberg M. Interviewers guide to the structured clinical interview for DSM-IV disorders (SCID). Washington, DC: American Psychiatric Press, 1994.

13. Rubinov M and Sporns O. Complex network measures of brain connectivity: Uses and interpretations. Neuroimage 2010; 52: 1059–1069.





14. Bakshi R, Shaikh ZA, Miletich RS, et al. Fatigue in multiple sclerosis and its relationship to depression and neurologic disability. Mult Scler 2000; 6: 181–155.

15. Kroencke DC, Lynch SG and Denney DR. Fatigue in multiple sclerosis: relationship to depression, disability, and disease pattern. Mult Scler 2000; 6: 131–136.

16. Giordano A, Granella F, Lugaresi A, et al. Anxiety and depression in multiple sclerosis patients around diagnosis. J Neurol Sci 2011; 307: 86–91.

17. Rocca MA, Tortorella P, Ceccarelli A, et al. The ‘mirror-neuron system’ in MS: A 3 tesla fMRI study. Neurology 2008; 70: 255–262.

18. Davidson RJ, Pizzagalli D, Nitschke JB, et al. Depression: Perspectives from affective neuroscience. Annu Rev Psychol 2002; 53: 545–574.

19. Barbas H. Connections underlying the synthesis of cognition, memory and emotion in primate prefrontal cortices. Brain Res Bull 2000; 52: 319–330.

20. Phillips ML, Ladouceur CD and Drevets WC. A neural model of voluntary and automatic emotion regulation: Implications for understanding the pathophysiology and neurodevelopment of bipolar disorder. Mol Psychiatry 2008; 13: 829, 833–857.

21. Kim MJ, Loucks RA, Palmer AL, et al. The structural and functional connectivity of the amygdala: From normal emotion to pathological anxiety. Behav Brain Res 2011; 223: 403–410.

22. Passingham RE, Bengtsson SL and Lau HC. Medial frontal cortex: From self-generated action to reflection on one’s own performance. Trends Cogn Sci 2010; 14: 16–21.

23. Grezes J, Valabregue R, Gholipour B, et al. A direct amygdala-motor pathway for emotional displays to influence action: A diffusion tensor imaging study. Hum Brain Mapp 2014; 35: 5974–5983.

24. Avendano C, Price JL and Amaral DG. Evidence for an amygdaloid projection to premotor cortex, but not

to motor cortex in the monkey. Brain Res 1983; 264: 111–117.

25. Applegate CD, Kapp BS, Underwood MD, et al. Autonomic and somatomotor effects of amygdala central N stimulation in awake rabbits. Physiol Behav 1983; 31: 353–360.

26. Pichon S, De Gelder B and Grezes J. Emotional modulation of visual and motor areas by dynamic body expressions of anger. Soc Neurosci 2008; 3: 199–212.

27. Gold SM, O’Connor MF, Gill R, et al. Detection of altered hippocampal morphology in multiple sclerosis-associated depression using automated surface mesh modeling. Hum Brain Mapp 2014; 35: 30–37.

28. Berg D, Supprian T, Thomae J, et al. Lesion pattern in patients with multiple sclerosis and depression. Mult Scler 2000; 6: 156–162.

29. Feinstein A, Magalhaes S, Richard JF, et al. The link between multiple sclerosis and depression. Nat Rev Neurol 2014; 10: 507–517.

30. Patten SB, Beck CA, Williams JV, et al. Major depression in multiple sclerosis: A population-based perspective. Neurology 2003; 61: 1524–1527.

31. Dahl OP, Stordal E, Lydersen S, et al. Anxiety and depression in multiple sclerosis. A comparative population-based study in Nord-Trondelag county, Norway. Mult Scler 2009; 15: 1495–1501.

32. Wood B, Van der Mei IA, Ponsonby AL, et al. Prevalence and concurrence of anxiety, depression and fatigue over time in multiple sclerosis. Mult Scler 2013; 19: 217–224.

33. Korgaonkar MS, Fornito A, Williams LM, et al. Abnormal structural networks characterize major depressive disorder: A connectome analysis. Biol Psychiatry 2014; 76: 567–574.

34. Descoteaux M, Deriche R, Knosche TR, et al. Deterministic and probabilistic tractography based on complex fibre orientation distributions. IEEE Trans Med Imag 2009; 28: 269–286.

Visit SAGE journals online http://msj.sagepub.com

SAGE journals


http://msj.sagepub.com


Structural 'connectomic' alterations in the limbic system of multiple sclerosis patients with major...

Documents

Transcript of Structural 'connectomic' alterations in the limbic system of multiple sclerosis patients with major...