Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study

Parkinsonism and Related Disorders xxx (2018) 1e7

Contents lists available at ScienceDirect

Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis

Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study Junyi Li a, 1, Yongsheng Yuan a, 1, Min Wang b, Jiejin Zhang a, Li Zhang a, Siming Jiang a, Xixi Wang a, Jian Ding a, Kezhong Zhang a, * a b

Department of Neurology, The First Affiliated Hospital of Nanjing Medical University, No. 300 Guangzhou Road, Nanjing 210029, China Department of Radiology, The First Affiliated Hospital of Nanjing Medical University, No. 300 Guangzhou Road, Nanjing 210029, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 May 2017 Received in revised form 18 February 2018 Accepted 14 March 2018

Introduction: Freezing of gait is a common complaint in patients with Parkinson's disease (PD). However, the neural bases of freezing of gait in PD remain uncertain. Existing studies on PD patients with freezing of gait (PD-FOGþ) have reported damage of the corpus callosum, the largest commissural bundle of the brain. Thus, in this study we explored homotopic connectivity to investigate FOG-related interehemispheric alterations Methods: A total of 21 PD-FOG þ patients, 33 PD patients without freezing of gait (PD-FOG-), and 24 matched healthy controls were recruited. All PD patients were evaluated via the FOG questionnaire (FOGQ) and all subjects had a resting state functional magnetic resonance imaging (rs-fMRI) scan. The pattern of the homotopic connectivity was measured with the voxel-mirrored homotopic connectivity (VMHC) approach. Result: The PD-FOG þ patients showed decreased VMHC values in the inferior parietal lobe (IPL) compared to both PD-FOG-patients and healthy controls. In PD-FOG þ patients, the mean VMHC values in the IPL were negatively correlated with the FOGQ scores. Receiver operating characteristic curves analyses revealed that the VMHC in the IPL had discriminatory function distinguishing PDFOG þ patients from PD-FOG-patients or healthy controls. Conclusion: Decreased VMHC values of PD-FOG þ patients relative to PD-FOG- and healthy controls in IPL maybe a unique feature for PD-FOGþ and it may have the ability to separate PD-FOG þ patients from PD-FOG- and healthy controls. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Parkinson's disease Resting-state fMRI Voxel-mirrored homotopic connectivity Freezing of gait

1. Introduction Freezing of gait (FOG), which affects approximately 50% of people with Parkinson disease (PD), is a disabling phenomenon that seriously affects the quality of life of PD patients [1]. It is defined as “a brief, episodic absence or marked reduction of forward progression of the feet despite the intention to walk” and is often characterized by the episodic feeling of feet “glued” to the floor [2]. Medical and rehabilitation treatment may improve FOG but not to the same extent as other PD symptoms [3]. Although FOG has catastrophic consequences for the PD patients, its pathophysiological mechanisms are still not fully

* Corresponding author. E-mail address: [email protected] (K. Zhang). 1 The first two authors contributed equally to this work.

understood. Early studies had shown that impaired control of rhythmicity, bilateral incoordination, and gait asymmetry were important aspects of freezing [4e6]. Similarly, some studies had reported that the poor coordination between the legs or gait cycle disorders were related to FOG [7]. Recently, several studies have found damage of the corpus callosum in PD patients with FOG using structural and functional connectivity [8,9] It is well known that the corpus callosum (CC) is the main collection of white matter bundles connecting both hemispheres so that both sides of the body can be coordinated. Its structural damage may affect the functional coordination between the cerebral hemispheres. Thus, it is reasonable to expect that the deficits of hemispheric interactions play a key role in the pathophysiology of FOG in PD. Therefore, it would be meaningful to examine the inter-hemispheric coordination in PD. Resting-state fMRI (rs-fMRI), which captures the patterns of coherent spontaneous fluctuations of blood oxygen level

https://doi.org/10.1016/j.parkreldis.2018.03.015 1353-8020/© 2018 Elsevier Ltd. All rights reserved.

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015

2

J. Li et al. / Parkinsonism and Related Disorders xxx (2018) 1e7

dependent (BOLD) signals [10] during rest, can be used to measure the inter-hemispheric coordination. Functional homotopy, defined as the high degree of synchrony in spontaneous activity between geometrically corresponding inter-hemispheric regions, has been suggested to be a key characteristic of the brain's intrinsic functional architecture [11,12]. Thus, homotopic resting-state functional connectivity (RSFC) may be a sensitive index for detecting the PDrelated inter-hemispheric coordination alterations. Here, we examined homotopic RSFC in PD FOG þ patients using a recently validated approach named “voxel-mirrored homotopic connectivity (VMHC) [13].” Different strengths of VMHC between different symmetric regions could represent different characteristics of hemispheric specialization in the information processing, sensory integration, and motor coordination [12]. Using the VMHC method, abnormal homotopic RSFC has been demonstrated in PD, Luo et al. found that PD patients exhibited significantly lower VMHC in putamen and cortical regions associated with sensory processing and motor control, relative to healthy subject [14]. Hu et al. found that, compared to normal control subjects and subjects with akineticrigid PD, tremor-dominant PD exhibited significantly lower VMHC values in the posterior lobe of the cerebellum and akinetic-rigid PD exhibited lower VMHC values in the precentral gyrus compared with normal control subjects [15]. In this study, we hypothesized that the differences of functional coordination between left and right brain may be involved in the pathogenesis of FOG in PD patients, which would be reflected by reduced RSFC in PD FOG þ patients. Moreover, given the importance of bilateral hemispheric coordination for motor functions, we also expected that VMHC measures would be clinically relevant.

2. Materials and methods 2.1. Participants We investigated 54 right-handed patients with a diagnosis of PD according to the UK Parkinson's Disease Society Brain Bank criteria for idiopathic PD [16]. Exclusion criteria comprised: red flags suggestive of atypical parkinsonism, atypical parkinsonism suggestive of multiple system atrophy, corticobasal degeneration, progressive supranuclear palsy, severe tremor, significant comorbidities affecting gait (e.g. acute illness, visual disturbances, orthopedic disease, musculoskeletal disorders, or history of stroke), absence of proper medical treatment, resistance to dopaminergic drugs, significant cognitive dysfunction (mini-mental state exam (MMSE) score <24), contraindications to MRI, such as claustrophobia, metallic implants, or devices in the body. Patients were classified as exhibiting FOG (FOGþ) based on the following two conditions that had all to be fulfilled: (1) score 1 point on item 3 of the FOG questionnaire (FOGQ) [17], (2) the recognition of typical FOG in the patient's experience when this was identified and described to him or her by a physician. Patients not fulfilling any one of the above conditions were classified as not exhibiting FOG (FOG). Clinical tests and MRI scans were performed in the morning OFF medication, after at least 12 h withdrawal from antiparkinsonian medications to mitigate the pharmacological effects on neural activity. Additionally, a group of 24 gender- and agematched healthy controls (HCs), without neurological and psychological disturbances or imaging abnormalities were also recruited from local individuals who volunteered to participate in scientific studies. This study was approved by the ethics committee of the First Affiliated Hospital of Nanjing Medical University, and all participants gave their informed written consent before beginning the experiment.

2.2. Clinical assessment Motor symptoms and PD severity were evaluated via the motor component of the Unified Parkinson's Disease Rating Scale (UPDRSIII). The severity of FOG was evaluated using the FOGQ, a six-item scale (range 0e24) composed of four items assessing FOG severity and two items testing gait difficulties in general. Subjects also perform the Tinetti Mobility Test [18] and timed up and go (TUG) [19] to assess balance and gait. The MMSE [20] was administered to screen the global cognitive function and executive function. In addition, Levodopa equivalent daily dose (LEDD) was calculated according to the established methods [21]. 2.3. Image acquisition MRI scannings were performed with a 3.0 T Siemens MAGNETOM Verio whole-body MRI system (Siemens Medical Solutions, Germany) equipped with eight-channel, phase-array head coils. Tight foam padding was used to minimize head movement, and ear-plugs were used to reduce noise. Subjects were instructed to remain motionless, close their eyes, remain awake, and not to think about anything in particular. Three-dimensional T1-weighted anatomical images were acquired using the following volumetric 3D magnetization-prepared rapid gradient-echo (MP-RAGE) sequence (repetition time [TR] ¼ 1900 ms, echo time [TE] ¼ 2.95 ms, flip angle [FA] ¼ 9 , slice thickness ¼ 1 mm, slices ¼ 160, field of view [FOV] ¼ 230  230 mm2, matrix size ¼ 256  256 and voxel size ¼ 1  1  1 mm3). Resting-state functional images were collected using an echo-planar imaging (EPI) sequence (TR ¼ 2000 ms, TE ¼ 21 ms, FA ¼ 90 , FOV ¼ 256  256 mm2, in-plane matrix ¼ 64  64, slices ¼ 35, slice thickness ¼ 3 mm, no slice gap, voxel size ¼ 3  3  3 mm3, total volumes ¼ 240). DTI images were acquired using spin echo planar imaging sequence with the following parameters: TR ¼ 9800 ms, TE ¼ 95 ms, FOV ¼ 256  256 mm2, NEX ¼ 1, matrix ¼ 128  128, slice thickness ¼ 2 mm, and slice gap ¼ 0 mm. Diffusion gradients were applied in 30 non-collinear directions with a b factor of 1000 s/mm2 after an acquisition without diffusion weighting (b ¼ 0 s/mm2) for reference. 2.4. Preprocessing of fMRI data analysis Rs-fMRI data preprocessing was then conducted by SPM8 software package (http://www.fil.ion.ucl.ac.uk/spm/), REST (http:// restfmri.net/forum/rest) and Data Processing Assistant for Resting-State fMRI (DPARSF). Briefly, the preprocessing steps included the following steps: (1) removal of first 10 time points; (2) correction for differences in the image acquisition time between slices; (3) six parameter rigid body spatial transformation to correct for head motion during data acquisition; (4) coregistration of the T1 image to the mean EPI scans; (5) grey and white matter segmentation using “New Segment” and spatial normalization of the structural image to a standard template (Montreal Neurological Institute) by DARTEL “normalization”; (6) spatial normalization of the EPI images using the normalization parameters estimated in the previous preprocessing step and resampling to 3  3  3 mm3; (7) spatial smoothing with a 6 mm full width half maximum Gaussian kernel; (8) temporally bandpass filtering (0.01e0.08 Hz) and linearly detrended removal; (9) regressing eight nuisance covariates, including the white matter signal, the cerebral spinal fluid signal, and six head motion parameters, to remove the possible variances from time course of each voxel. According to the record of head motions within each fMRI run, all participants had less than 2.0 mm maximum displacement in the x, y, or z plane and less than 2.0 of angular rotation about each

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015

J. Li et al. / Parkinsonism and Related Disorders xxx (2018) 1e7

axis. We also calculated the mean framewise displacement (FD) [22,23] for each group and there was no difference in mean head motion (P > 0.05). 2.5. Voxel-mirrored homotopic connectivity VMHC assumes symmetric morphology between hemispheres. To account for differences in the geometric configuration of the cerebral hemispheres, we firstly averaged the normalized T1 images of all subjects to create a mean normalized T1 image. This image was then averaged with its left-right mirrored version to generate a group-specific symmetrical T1 template. After that, the individual T1 images in MNI space were nonlinearly registered to the symmetrical T1 template and those transformations were applied to the above processed functional data. The VMHC computation was performed with software REST. For each participant, the homotopic RSFC was computed as the Pearson correlation coefficient between each voxel's residual time series and that of its symmetrical inter-hemispheric counterpart. Correlation values were then Fisher z-transformed to improve the normality. The resultant values were referred to as the VMHC and were applied for the group comparisons. An analysis of covariance (ANCOVA) was performed to identify brain areas with significant differences in VMHC among the three groups with age, gender, education and mean FD as covariates (voxel-level p < 0.05, cluster size > 282 voxels, corresponding to a corrected p < 0.01 as determined by AlphaSim correction). These areas were then extracted as a mask. Then, two-sample post hoc t tests were performed within this mask, with age, gender, education and mean FD as covariates, to detect significant differences between groups (voxel level p < 0.01, cluster size > 6 voxels, corresponding to a corrected p < 0.01 as determined by AlphaSim correction) (http://afni.nimh.nih.gov/pub/dist/doc/manual/ AlphaSim.pdf). 2.6. Statistical analysis of demographic and clinical data Demographic and clinical characteristics of all subjects in different groups were compared by the one-way anova analysis (ANOVA) and Post-Hoc t-test, Chi-square test, Independent samples T test, or ManneWhitney U test, as appropriate. Brain regions exhibiting significant differences between the PD-FOGþ and PD-

3

FOG-groups were selected as ROIs. We computed Pearson correlation coefficients between the extracted mean VMHC values within these ROIs and the FOGQ scores of PD-FOG þ patients. Furthermore, the mean VMHC values extracted within these ROIs were further used for receiver operating characteristic curves (ROC) analyses [24]. The sensitivity, specificity, correct classification, and area under the curve (AUC) [25] were reported for the cut-off values. A cut-off was selected using Youden index and minimum distance from the coordinate (0,1) on the ROC curve methods. The higher AUC reveals a better discriminating ability of findings. Statistical analyses were performed with SPSS 20.0 statistical analysis software (SPSS Inc. Chicago, IL, USA). Significance threshold was set to p ¼ 0.05. 3. Result 3.1. Demographic and clinical characteristics The participants included 21 PD-FOG þ patients, 33 PD-FOGpatients and 24 HCs subjects. The demographic and clinical data are summarized in Table 1 and in Table 4. No significant differences were detected among the groups for age, gender and education levels. There was no significant difference between the FOGþ and FOG patients groups in terms of disease duration, UPDRS-III, LEDD and MMSE. As expected, the FOGQ score (p < 0.001) significantly differed between the two PD groups. Further motor examinations, as measured by the TUG test (p ¼ 0.04), the Tinetti Balance (p ¼ 0.002) and the Tinetti Gait (p ¼ 0.04), indicated that PDFOG þ patients performed significantly worse than PD-FOGpatients. 3.2. Voxel-mirrored homotopic connectivity An ANCOVA revealed significant differences in VMHC among PD-FOGþ, PD-FOG- AND HCs groups, followed by a two-sample post hoc t-test to detect VMHC differences between each pair of the HCs, PD-FOGþ and PD-FOG-groups. Compared to PD-FOG-, PDFOG þ had decreased VMHC in the inferior parietal lobe (IPL)(Fig. 1A); Compared to HCs, PD-FOG-had decreased VMHC in the precentral gyrus and the postcentral gyrus (Fig. 1C); VMHC in the IPL and the precentral gyrus was notably decreased in PDFOG þ patients compared with HCs (Fig. 1B; Table 2). In addition,

Table 1 Demographic and clinical characteristics of all subjects. Items

HCs (N ¼ 24) Mean ± SD

PD-FOGþ (N ¼ 21) Mean ± SD

PD-FOG- (N ¼ 33) Mean ± SD

P value

Age, years Education, years Gender, Female/Male Disease duration, years UPDRS-III LEDD, mg/day MMSE TUG, s Tinetti balance Tinetti gait FOGQ Mean FD

65.79 ± 4.18 10.71 ± 2.87 12/12 NA NA NA NA NA NA NA NA 0.137 ± 0.063

69.90 ± 5.70 12.33 ± 3.75 8/13 3.21 ± 2.30 24.52 ± 8.20 471.41 ± 292.55 28.10 ± 1.51 17.98 ± 7.28 11.05 ± 3.26 7.14 ± 2.97 10.90 ± 5.11 0.091 ± 0.052

65.64 ± 8.95 10.79 ± 2.90 13/20 3.79 ± 3.07 21.03 ± 10.30 377.14 ± 246.67 28.00 ± 1.89 13.15 ± 4.27 14.89 ± 1.76 10.24 ± 1.90 0.94 ± 1.48 0.101 ± 0.056

0.07a 0.15a 0.65b 0.08c 0.18c 0.09c 0.46c 0.04c 0.002c 0.04c <0.001c 0.06a

HCs, healthy controls; PD-FOGþ, Parkinson's disease with freezing of gait; PD-FOG-, Parkinson's disease without freezing of gait; UPDRS, Unified Parkinson's disease rating scale; LEDD, Levodopa equivalent daily dose; MMES, Mini Mental State Examination; TUG, timed up and go; FOGQ, FOG questionnaire; NA, not applicable; FD, framewise displacement. a P-values for differences between the three groups with regard to age, education, and the mean FD values, obtained using one-way analysis of variance (ANOVA). b P-value for gender distribution in the three groups, obtained using the chi-square test. c P-value for disease duration, UPDRS scores, l-dopa dose, MMSE scores, TUG, Tinetti balance scores, Tinetti gait score and FOGQ scores, obtained using the two-sample ttest.

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015

4

J. Li et al. / Parkinsonism and Related Disorders xxx (2018) 1e7

patients with PD develop FOG after the first 3 years of the disease. Although we conducted a follow-up of enrolled PD-FOG þ patients to re-diagnose, for the rigor of the experiment, we excluded patients with a disease duration of one year or less (i.e., patients 3, 5 and 9) and repeat the analyses again. The result was shown in the Supplementary Materials, Fig. S2 and Table 5, consistent with the previous results.

3.3. Correlation analysis Based on the VMHC results, Correlation analyses between FOGQ scores and mean VMHC values of brain regions showing significant differences between PD-FOGþ and PD-FOG-groups were examined for PD-FOG þ patients. FOGQ scores were inversely correlated with mean VMHC signals in the IPL in PD-FOG þ patients (r ¼ 0.437,

Fig. 1. VMHC differences among PD-FOGþ, PD-FOG- and HCs groups. (A): differences between PD-FOG þ patients and PD-FOG-patients; (B): differences between PDFOG þ patients and HCs group; (C): differences between PD-FOG-patients and HCs group; (D): The negative correlations between FOGQ scores and VMHC values in the IPL. The results were corrected by AlphaSim (p < 0.01, with a combined threshold of p < 0.01). VMHC, voxel-mirrored homotopic connectivity; PD-FOGþ, Parkinson's disease patients with freezing of gait; PD-FOG-, Parkinson's disease patients without freezing of gait; HCs, healthy controls; FOGQ, the FOG questionnaire; IPL, inferior parietal lobe; L, left; R, right.

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015

J. Li et al. / Parkinsonism and Related Disorders xxx (2018) 1e7 Table 2 VMHC differences among PD-FOG þ patients, PD-FOG-patients and HCs. Anatomical region(AAL)

Number of voxels

PD-FOG þ vs PD-FOGParietal_Inf PD-FOG þ vs HCs Parietal_Inf Precentral PD-FOG- vs HCs Postcentral Precentral

MNI coordinates of local maxima

5

Table 5 VMHC differences among PD-FOG þ patients, PD-FOG-patients and HCs.

T

Anatomical region(AAL)

x

y

z

72

±45

54

39

3.9852a

44 46

±42 ±33

60 18

51 48

3.9121b 4.5557b

7 22

±39 ±33

27 18

63 48

3.4829c 4.0302c

Number of voxels

PD-FOG þ vs PD-FOGParietal_Inf PD-FOG þ vs HCs Parietal_Inf Precentral PD-FOG- vs HCs Postcentral Precentral

VMHC, voxel-mirrored homotopic connectivity; HCs, healthy controls; PD-FOGþ, Parkinson's disease with freezing of gait; PD-FOG-, Parkinson's disease without freezing of gait. a Post hoc paired comparisons showed significant group differences between PDFOG þ versus PD-FOG-. b Post hoc paired comparisons showed significant group differences between PDFOG þ versus HCs. c Post hoc paired comparisons showed significant group differences between PDFOG-versus HCs.

MNI coordinates of local maxima

T

x

y

z

71

±48

54

36

3.9516a

43 56

±45 ±33

57 18

48 48

3.6846b 4.3082b

22 22

±39 ±33

27 18

63 48

3.4829c 4.0302c

VMHC, voxel-mirrored homotopic connectivity; HCs, healthy controls; PD-FOGþ, Parkinson's disease with freezing of gait; PD-FOG-, Parkinson's disease without freezing of gait. a Post hoc paired comparisons showed significant group differences between PDFOG þ versus PD-FOG-. b Post hoc paired comparisons showed significant group differences between PDFOG þ versus HCs. c Post hoc paired comparisons showed significant group differences between PDFOG-versus HCs.

Table 3 ROC analyses for differentiating different groups. Brain regions

AUC

Separating PD-FOGþ from PD-FOGIPL 0.821 Separating PD-FOGþ from HCs IPL 0.863 Separating PD-FOG- from HCs IPL 0.586

Cut-off point

Sensitivity

Specificity

p value

95% CI

0.3644a

61.90%(13/21)

90.91%(30/33)

<0.001

0.693e0.912

0.3644

61.90%(13/21)

100%(24/24)

<0.001

0.728e0.947

0.7405

36.36%(12/33)

100%(24/24)

0.321

0.429e0.731

VMHC, voxel-mirrored homotopic connectivity; AUC, area under the curve; HCs, healthy controls; PD-FOGþ, Parkinson's disease with freezing of gait; PD-FOG-, Parkinson's disease without freezing of gait. IPL, inferior parietal lobe. a By this cut-off point, the VMHC value of IPL could correctly classify 13 of 21 patients with PD-FOGþ and 30 of 33 patients with PD-FOG-, resulted in a sensitivity of 61.90% and a specificity of 90.91%. The meanings of other cut-off points were similar.

p < 0.05) (Fig. 1D), and the same as in all patients with PD (r ¼ 0.613, p < 0.001), indicating that with the deepening of the patient's FOG, the function coordination of IPL will be getting poorer. Since the IPL exhibited significant VMHC differences between

PD-FOGþ and PD-FOG-patient groups and between PD-FOGþ and HCs groups, it was selected as a region of interest (ROI). Mean VMHC values were extracted from this ROI for further ROC analyses. The results revealed that the area under the curve (AUC) of the IPL was 0.821 (95% confidence interval [CI]: 0.693e0.912, p < 0.001)

Table 4 The demographic and clinical characteristics of PD-FOG þ patients. PD-FOG þ patients

item 3 of the FOGQ

Gender, M/F

Age, year

Disease duration, year

FOG onset to PD onset, year

Follow-up time, year

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

4 4 3 3 3 3 3 3 3 3 3 2 1 1 1 1 1 1 1 1 3

M M F M M F M F M M F M F M M M F F M M F

64 75 75 69 77 67 68 70 75 76 66 75 76 70 64 70 69 60 62 79 61

3.00 5.00 0.50 9.00 1.00 4.00 3.00 3.00 0.58 3.25 2.00 3.00 1.50 2.00 5.00 4.00 9.00 3.00 2.00 1.50 2.00

2.5 4 0.5 7 1 3 2 3 0.5 3 2 2.5 1 2 4.5 3.5 8 2 2 1.5 1.5

2 2 3 2 3 3 1 2 3 1 3 2 3 2 1 3 1 2 2 3 3

PD-FOGþ: Parkinson's disease with freezing of gait; FOGQ: FOG questionnaire; M: male; F: female.

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015

6

J. Li et al. / Parkinsonism and Related Disorders xxx (2018) 1e7

when separating PD-FOG þ patients from PD-FOG-patients and 0.863 (95% CI: 0.728e0.947, p < 0.001) when separating PDFOG þ patients from HCs group. Further analyses showed that the specificity for separating PD-FOG þ patients from PD-FOG- and HCs groups were relatively high, up to 90.91% (cut-off point is 0.3644) and 100% (cut-off point is 0.3644) respectively. (Table 3; Supplementary Materials, Fig. S1). 4. Discussion In this study, we used the VMHC method to compare interhemispheric RSFC changes between PD-FOG þ patients, PD-FOGpatients and HCs. As far as we know, this is the first article to study the functional connectivity between the hemispheres of the brain in the PD patients with FOG. Our primary finding was that the PD-FOG þ patients showed significantly reduced VMHC values in the IPL compared to both PD-FOG- and HCs groups, the mean VMHC values of this brain region were negatively correlated with FOGQ scores and the further ROC analyses indicated that decreased VMHC values in the IPL could be used to differentiate the PDFOG þ patients from PD-FOG- and HCs groups. These results are consistent with previous studies about FOG. Herman et al. found that gray matter (GM) atrophy in the parietal lobe, angular gyrus and caudate are related to FOG [26]. Tessitore et al. found significant GM atrophy in the left cuneus, precuneus, lingual gyrus, and posterior cingulate cortex [27]. Kostic et al. observed atrophy of the left inferior frontal gyrus, left precentral gyrus, and left inferior parietal gyrus [28]. Canu et al. found PDFOG þ patients showed white matter (WM) damage of the pedunculopontine tract (PPT), corpus callosum, corticospinal tract, cingulum, superior longitudinal fasciculus, and WM underneath the primary motor, premotor, prefrontal, orbitofrontal, and inferior parietal cortices, bilaterally [8]. The fMRI study did not observe any FOG correlated brain regions in frontal and cingulate cortex. This may result from different statistical approaches and imaging techniques. FOG most frequently occurs at turns or during the initiation of walking [29], during turns, the actual movement of each leg is significantly different, which is a challenge for the control of leg movement and the bilateral coordination and could contribute to FOG [7]. We found that function coordination of the IPL reducing in PD-FOG þ patients, which may be due to gray matter volume atrophy in the left IPL and white matter damage underneath the bilateral IPL in PD-FOG þ patients [26,28], proved this view. The IPL is activated during action execution and is involved in the sensory integration of perceptual spatiotemporal information [30,31]. The functional deficits of IPL may lead to loss of control of the gait [28], and a decrease in functional coordination of bilateral IPL may result in an uncoordinated control of the gait. When this complex action occurs, this incongruity will be more obvious, which may lead to FOG. In addition, the existing studies have shown that the IPL plays an important role in more advanced cognitive functions such as attention, language, action processing and so on [32]. Interestingly, cognitive resources are used in order to maintain consistent and accurate alternations in left-right stepping in PD patients [33]. Our results suggest that differences in bilateral IPL function may lead to inconsistencies in the use of these cognitive resources by PD patients, resulting in inconsistencies in the maintenance of gait, which may lead to FOG. Furthermore, a dorsal pathway through the parietal cortex is involved in the integration of sensory and visual information with locomotion [34]. Posterior parietal cortex receives both visual information from occipito-parietal pathways, and motor-related information from somatosensory cortex. In addition, the posterior

parietal region has a projection to the frontal lobe. This seems to indicate that this fronto-parietal network functions is based on proprioceptive, visual, gaze, attentional, and other information to produce an output that reflects the selection, preparation, and execution of movements [35]. And FOG may be related to the functional barrier of the fronto-parietal network [36]. In our study, we found that the mean VMHC values of PD-FOG þ patients decreased in the IPL and the precentral gyrus, perhaps because the uncoordinated input from the bilateral IPL resulted in unbalanced output of bilateral motion function through the frontal lobe network pathway, eventually leading to FOG. We also determined the predictive performance of pretreatment VMHC in differentiating the PD-FOG þ patients from PD-FOG- and HCs groups. The results of ROC curve analyses indicated that VMHC changes in the IPL may have the ability to identify the PD-FOG þ patients from PD-FOG- or HCs group. Further analyses found that VMHC values in the IPL brain region were particularly specific to distinguish PD-FOG þ patients from PD-FOG- and HCs groups, which may be a help to reduce misdiagnosis. In addition, our study found that VMHC values in the IPL brain region were not distinguish PD-FOG-from HCs, which might suggest that VMHC values in the IPL brain region appeared to be a unique feature for PD-FOGþ. There are however some limitations of our study. First, the sample size of our study is relatively small, which may limit examinations of minor brain structures that correlate with FOG in addition to those identified in present study. Second, there are existing asymmetries in cortical structure. We attempted to mitigate these issues by using a symmetric template. Finally, the occurrence of FOG is episodic and unpredictable often not appearing during evaluations. We depended on self-report of FOG in our patients which could be affected by biased recall. In conclusion, the decreased VMHC values in the IPL may represent an imaging biomarker for PD-FOGþ and may have the ability to separate PD-FOG þ patients from PD-FOG- and HCs groups. The findings also suggest that uncoordinated control of the gait and using of cognitive resources, which may result from the uncoordinated function of bilateral IPL, may be involved in the pathophysiological mechanisms of PD-FOGþ.

Author contribution Junyi Li: conception of the study, acquisition of clinical data, statistical analysis, writing of the first draft. Yongsheng Yuan: conception of the study, acquisition of clinical data, writing of the first draft. Min Wang: acquisition of MRI data, statistical analysis. Jiejin Zhang: acquisition of clinical data, revising the manuscript for content. Li Zhang: acquisition of clinical data, revising the manuscript for content. Siming Jiang: acquisition of clinical data, revising the manuscript for content. Xixi Wang: acquisition of clinical data, revising the manuscript for content. Jian Ding: acquisition of clinical data, revising the manuscript for content. Kezhong Zhang: study concept or design, revising the manuscript for content, study supervision, obtaining funding.

Conflicts of interest The authors declare that they have no conflict of interest.

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015

J. Li et al. / Parkinsonism and Related Disorders xxx (2018) 1e7

Acknowledgment This work was supported by the National Natural Science Foundation of China (No.81671258), the Science and Technology Project of Jiangsu Provincial Commission of Health and Family Planning (No.H201602), the Natural Science Foundation of Jiangsu Province (No.BK20141494), the Jiangsu Provincial Personnel Department “the Great of Six Talented Man Peak” Project (No.2014WSN-013), the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), and the Science and Technology Project of Jiangsu Bureau of Traditional Chinese Medicine (No.YB2015163).

[14]

[15]

[16]

[17] [18]

[19]

Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.parkreldis.2018.03.015

[20]

[21]

References [22] [1] M. Macht, Y. Kaussner, J.C. Moller, K. Stiasny-Kolster, K.M. Eggert, H.P. Kruger, H. Ellgring, Predictors of freezing in Parkinson's disease: a survey of 6,620 patients, Mov. Disord. 22 (7) (2007) 953e956. [2] J.G. Nutt, B.R. Bloem, N. Giladi, M. Hallett, F.B. Horak, A. Nieuwboer, Freezing of gait: moving forward on a mysterious clinical phenomenon, Lancet Neurol. 10 (8) (2011) 734e744. [3] N. Giladi, Medical treatment of freezing of gait, Mov. Disord. 23 (Suppl 2) (2008) S482eS488. [4] M. Plotnik, N. Giladi, Y. Balash, C. Peretz, J.M. Hausdorff, Is freezing of gait in Parkinson's disease related to asymmetric motor function? Ann. Neurol. 57 (5) (2005) 656e663. [5] M. Plotnik, N. Giladi, J.M. Hausdorff, Bilateral coordination of walking and freezing of gait in Parkinson's disease, Eur. J. Neurosci. 27 (8) (2008) 1999e2006. [6] M. Plotnik, N. Giladi, J.M. Hausdorff, Is freezing of gait in Parkinson's disease a result of multiple gait impairments? Implications for treatment, Parkinsons Dis 2012 (2012) 459321. [7] D.S. Peterson, M. Plotnik, J.M. Hausdorff, G.M. Earhart, Evidence for a relationship between bilateral coordination during complex gait tasks and freezing of gait in Parkinson's disease, Parkinsonism Relat Disord 18 (9) (2012) 1022e1026. [8] E. Canu, F. Agosta, E. Sarasso, M.A. Volonte, S. Basaia, T. Stojkovic, E. Stefanova, G. Comi, A. Falini, V.S. Kostic, R. Gatti, M. Filippi, Brain structural and functional connectivity in Parkinson's disease with freezing of gait, Hum. Brain Mapp. 36 (12) (2015) 5064e5078. [9] M. Wang, S. Jiang, Y. Yuan, L. Zhang, J. Ding, J. Wang, J. Zhang, K. Zhang, J. Wang, Alterations of functional and structural connectivity of freezing of gait in Parkinson's disease, J. Neurol. 263 (8) (2016) 1583e1592. [10] M.D. Fox, M.E. Raichle, Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging, Nat. Rev. Neurosci. 8 (9) (2007) 700e711. [11] R. Salvador, J. Suckling, M.R. Coleman, J.D. Pickard, D. Menon, E. Bullmore, Neurophysiological architecture of functional magnetic resonance images of human brain, Cerebr. Cortex 15 (9) (2005) 1332e1342. [12] D.E. Stark, D.S. Margulies, Z.E. Shehzad, P. Reiss, A.M. Kelly, L.Q. Uddin, D.G. Gee, A.K. Roy, M.T. Banich, F.X. Castellanos, M.P. Milham, Regional variation in interhemispheric coordination of intrinsic hemodynamic fluctuations, J. Neurosci. 28 (51) (2008) 13754e13764. [13] X.N. Zuo, C. Kelly, A. Di Martino, M. Mennes, D.S. Margulies, S. Bangaru, R. Grzadzinski, A.C. Evans, Y.F. Zang, F.X. Castellanos, M.P. Milham, Growing

[23] [24] [25] [26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34] [35]

[36]

7

together and growing apart: regional and sex differences in the lifespan developmental trajectories of functional homotopy, J. Neurosci. 30 (45) (2010) 15034e15043. C. Luo, X. Guo, W. Song, B. Zhao, B. Cao, J. Yang, Q. Gong, H.F. Shang, Decreased resting-state interhemispheric functional connectivity in Parkinson's disease, BioMed Res. Int. 2015 (2015) 692684. X. Hu, J. Zhang, X. Jiang, C. Zhou, L. Wei, X. Yin, Y. Wu, J. Li, Y. Zhang, J. Wang, Decreased interhemispheric functional connectivity in subtypes of Parkinson's disease, J. Neurol. 262 (3) (2015) 760e767. J. Joutsa, M. Gardberg, M. Roytta, V. Kaasinen, Diagnostic accuracy of parkinsonism syndromes by general neurologists, Parkinsonism Relat Disord 20 (8) (2014) 840e844. N. Giladi, Freezing of gait. Clinical overview, Adv. Neurol. 87 (2001) 191e197. D.A. Kegelmeyer, A.D. Kloos, K.M. Thomas, S.K. Kostyk, Reliability and validity of the Tinetti mobility test for individuals with Parkinson disease, Phys. Ther. 87 (10) (2007) 1369e1378. A. Shumway-Cook, S. Brauer, M. Woollacott, Predicting the probability for falls in community-dwelling older adults using the Timed up & Go Test, Phys. Ther. 80 (9) (2000) 896e903. M.F. Folstein, S.E. Folstein, P.R. McHugh, Mini-mental state". A practical method for grading the cognitive state of patients for the clinician, J. Psychiatr. Res. 12 (3) (1975) 189e198. C.L. Tomlinson, R. Stowe, S. Patel, C. Rick, R. Gray, C.E. Clarke, Systematic review of levodopa dose equivalency reporting in Parkinson's disease, Mov. Disord. 25 (15) (2010) 2649e2653. J.D. Power, K.A. Barnes, A.Z. Snyder, B.L. Schlaggar, S.E. Petersen, Spurious but systematic correlations in functional connectivity MRI networks arise from subject motion, Neuroimage 59 (3) (2012) 2142e2154. K.R. Van Dijk, M.R. Sabuncu, R.L. Buckner, The influence of head motion on intrinsic functional connectivity MRI, Neuroimage 59 (1) (2012) 431e438. J.A. Swets, ROC analysis applied to the evaluation of medical imaging techniques, Invest. Radiol. 14 (2) (1979) 109e121. J.A. Hanley, B.J. McNeil, The meaning and use of the area under a receiver operating characteristic (ROC) curve, Radiology 143 (1) (1982) 29e36. T. Herman, K. Rosenberg-Katz, Y. Jacob, N. Giladi, J.M. Hausdorff, Gray matter atrophy and freezing of gait in Parkinson's disease: is the evidence black-onwhite? Mov. Disord. 29 (1) (2014) 134e139. A. Tessitore, M. Amboni, G. Cirillo, D. Corbo, M. Picillo, A. Russo, C. Vitale, G. Santangelo, R. Erro, M. Cirillo, F. Esposito, P. Barone, G. Tedeschi, Regional gray matter atrophy in patients with Parkinson disease and freezing of gait, AJNR Am. J. Neuroradiol. 33 (2012) 1804e1809. V.S. Kostic, F. Agosta, M. Pievani, E. Stefanova, M. Jecmenica-Lukic, A. Scarale, V. Spica, M. Filippi, Pattern of brain tissue loss associated with freezing of gait in Parkinson disease, Neurology 78 (6) (2012) 409e416. J.D. Schaafsma, Y. Balash, T. Gurevich, A.L. Bartels, J.M. Hausdorff, N. Giladi, Characterization of freezing of gait subtypes and the response of each to levodopa in Parkinson's disease, Eur. J. Neurol. 10 (4) (2003) 391e398. A. Assmus, J.C. Marshall, A. Ritzl, J. Noth, K. Zilles, G.R. Fink, Left inferior parietal cortex integrates time and space during collision judgments, Neuroimage 20 (Suppl 1) (2003) S82eS88. S. Shimada, K. Hiraki, I. Oda, The parietal role in the sense of self-ownership with temporal discrepancy between visual and proprioceptive feedbacks, Neuroimage 24 (4) (2005) 1225e1232. S. Caspers, A. Schleicher, M. Bacha-Trams, N. Palomero-Gallagher, K. Amunts, K. Zilles, Organization of the human inferior parietal lobule based on receptor architectonics, Cerebr. Cortex 23 (3) (2013) 615e628. M. Plotnik, N. Giladi, J.M. Hausdorff, Bilateral coordination of gait and Parkinson's disease: the effects of dual tasking, J. Neurol. Neurosurg. Psychiatry 80 (3) (2009) 347e350. M.A. Goodale, A.D. Milner, Separate visual pathways for perception and action, Trends Neurosci. 15 (1) (1992) 20e25. S.P. Wise, D. Boussaoud, P.B. Johnson, R. Caminiti, Premotor and parietal cortex: corticocortical connectivity and combinatorial computations, Annu. Rev. Neurosci. 20 (1997) 25e42. A.L. Bartels, K.L. Leenders, Brain imaging in patients with freezing of gait, Mov. Disord. 23 (Suppl 2) (2008) S461eS467.

Please cite this article in press as: J. Li, et al., Decreased interhemispheric homotopic connectivity in Parkinson's disease patients with freezing of gait: A resting state fMRI study, Parkinsonism and Related Disorders (2018), https://doi.org/10.1016/j.parkreldis.2018.03.015