Cognition and the limbic system in early Parkinson’s

AIM

• To investigate Diffusion Tensor Imaging (DTI) values from limbic system regions of interest in early Parkinson's disease (PD) as possible biomarkers for subsequent followup.

• To ascertain whether the limbic system, as quantified by diffusion MRI, is associated with cognitive decline in PD.

INTRODUCTION
A vital goal in Parkinson's disease (PD) research is the identification of a reliable imaging biomarker of neurodegenerative changes that facilitates diagnosis and effective treatment planning. It is now recognized that moderate to severe cognitive dysfunction is a common development that significantly adds to the overall burden experienced by PD patients.1 Studies using pathology,2 radiotracer imaging,3 structural MRI,4 and diffusion MRI5 have identified various limbic system regions involved in the progression of PD. Hence a Region-of-Interest (ROI) analysis focussed on key regions of the limbic system may provide a suitable biomarker for cognitive decline in PD. DTI's sensitivity to underlying microscopic structural changes may be particularly suited in this context.

METHODS
Participants
Forty-eight early PD participants (≤5 years since symptom onset; mean age±SD: 65.5±10.0 years; 35 male) and 28 healthy controls matched for mean age, gender, and education (mean age±SD: 68.2±9.4 years; 19 male) completed a measure of global cognitive status (the Montreal Cognitive Assessment, MoCA) and MRI scans. PD participants also completed Unified Parkinson's Disease Rating Scale (UPDRS) to assess motor impairment.

Data Acquisition
DTI data was acquired on a 3T GE HDx scanner using a diffusion-weighted spin echo EPI sequence (TE=75.5 ms and TR=13 s) with diffusion weighting in 28 uniformly distributed directions (b=1000 s/mm2) and four acquisitions without diffusion weighting. In-house software fit a tensor to motion-corrected volumes to generate Mean Diffusivity (MD) and Fractional Anisotropy (FA) maps for each participant. These images were preprocessed using statistical parametric mapping software SPM5 and custom Matlab scripts. The structural image was coregistered to the MD, FA, and T2-weighted b0 images, segmented, and normalized to a probabilistic elderly brain atlas.6 An isotropic 8mm smoothing kernel was applied to the modulated, normalized segmented grey, white, and CSF tissue maps. These segmentation maps were then used as priors to segment and normalize the T2-weighted image to its corresponding T1-weighted structural image. The resulting parameters were used to normalize the MD and FA maps to eliminate EPI distortion.

Data Analysis
Region-of-interest (ROI) analysis was used to investigate structural changes in selected grey matter regions of the limbic system, specifically the limbic thalamus (t), basal forebrain/septal region (bs), hippocampus (h), parahippocampus (ph), amygdala (am), anterior cingulate (ac), and posterior cingulate (pc) (Figure 1). The mean FA and MD value was extracted from each region; t-tests were used to determine whether FA and MD values significantly differed between early PD and control groups. Effect sizes were also calculated. Correlations between neuropsychological tests and DTI values were also assessed in only the PD group;p<0.05 Bonferroni-corrected for multiple comparisons was considered significant.

Figure 1: The limbic system regions displayed on the mean normalized structural brain
created from both PD and Control data: Cyan - Posterior Cingulate; Yellow - Hippocampus;
White - Parahippocampus; Green - Limbic Thalamus; Red - Amygdala; Pink - Basal
Forebrain/Septal Region; Blue - Anterior Cingulate.

RESULTS
Demographics: Table 1 summarizes gender, age, education, MoCA , and UPDRS part 3 (motor score) data for all participants..
Imaging: After correcting for multiple comparisons, neither FA nor MD significantly differentiated between the early PD group and controls across any of the limbic areas. Correlation analysis between DTI values and UPDRS scores (in the PD group) failed to reach significance in any of the limbic regions. However, MD values for these regions as a collective system showed a significant correlation with cognition as measured by MoCA in the patient group (r = -0.69, p < 0.001, df = 42; Figure 2). Individual regions within the limbic system also exhibited significant correlations between MD and MoCA (PD group only): ac, bs, h, pc, and t (-0.68<r<-0.54). FA and MoCA did not correlate significantly in the PD group. The effect size between PD and controls as measured by Cohen's d: d>0.3 in bs, h (FA) and ac, pc (MD).

Table 1: Group demographic and neuropsychological data (mean, SD)

                                 Gender,male (%)           Age               Education            MoCA             UPDRS3
Controls(n=28)          19 (67.9%)                    68.2±9.4         13.4±2.8             27.2±1.85          ‐‐
PD (n = 48)               35 (73%)                       65.5±10.0       12.9±3.1             25.4±3.2            31.7±15.1

The PD group had a significantly reduced mean global cognitive status score (MoCA) than that shown by the Control group (two-tailed t-test: p = 0.007). MoCA scores below 26 may be indicative of cognitive impairment.7

Figure 2: (a) Significant correlation between MD and MoCA in the limbic system as a whole
(Analysis includes only PD's, Controls shown on graph solely as visual aid: r = -0.69, p
<0.001, df = 42). As the MoCA score decreases, MD increases. (b) Correlation between MD
and MoCA in the limbic thalamus (PD only: r = -0.67, p <0.001, df = 42). Other individual
structures within the limbic system exhibited smaller but significant correlations with MoCA
scores.

DISCUSSION
Mean DTI measures in the limbic system of this sample of early PD composed of non-advanced age and relatively short disease duration, did not distinguish between PD and control. Medium effect sizes (d>0.3) in four regions imply a trend toward a difference in mean DTI values between PD and controls early in the disease progression. Nonetheless, significant correlations between MoCA and the MD measure of limbic system integrity suggested that this approach has some value in the context of cognitive decline even during the early stages of PD. Further studies including advanced PD will help to clarify the utility of DTI in the limbic system as a biomarker in PD.

CONCLUSION
The microstructural integrity of the limbic system in this early PD group was similar to that of healthy controls when measured by DTI, but even this early cohort suggests that subtle cognitive impairment is associated with degeneration in limbic structures, at least in terms of mean diffusivity. These results illustrate the potential of DTI biomarkers to evaluate and potentially track anatomical substrates of cognitive decline in PD.

ACKNOWLEDGMENTS
The authors gratefully acknowledge support funding from the Neurological Foundation of New Zealand and Canterbury Medical Research Foundation.

REFERENCES
1. Aarsland, D., Beyer, M. K., & Kurz, M. W. (2008). Dementia in Parkinson's disease. Current Opinion in Neurology December, 21(6), 676-82.
2. Rüb, U., Del Tredici, K., Schultz, C., Ghebremedhin, E., de Vos, R. A. I., Jansen Steur, E., & Braak, H. (2002). Parkinson's disease: the thalamic components of the limbic loop are severely impaired by [alpha]-synuclein immunopositive inclusion body pathology. Neurobiology of Aging, 23(2), 245-54. 3. Hosokai, Y., Nishio, Y., Hirayama, K., Takeda, A., Ishioka, T., Sawada, Y., Suzuki, K., Itoyama, Y., Takahashi, S., Fukuda, H., & Mori, E. (2009). Distinct patterns of regional cerebral glucose metabolism in Parkinson's disease with and without mild cognitive impairment. Movement Disorders, 24(6), 854-62. 4. Summerfield, C., Junque, C., Tolosa, E., Salgado-Pineda, P., Gomez-Anson, B., Marti, M. J., Pastor, P., Ramirez-Ruiz, B., & Mercader, J. (2005). Structural Brain Changes in Parkinson Disease With Dementia: A Voxel-Based Morphometry Study. Arch Neurol, 62(2), 281-5.
5. Karagulle Kendi, A. T., Lehericy, S., Luciana, M., Ugurbil, K., & Tuite, P. (2008). Altered Diffusion in the Frontal Lobe in Parkinson Disease. AJNR Am J Neuroradiol, 29(3), 501-5. 6. Lemaître, H., Crivello, F., Grassiot, B., Alpérovitch, A., Tzourio, C., & Mazoyer, B. (2005). Age- and sex-related effects on the neuroanatomy of healthy elderly. NeuroImage, 26(3), 900-11. 7. Nasreddine, Z. S., Phillips, N. A., Bédirian, V., Charbonneau, S., Whitehead, V., Collin, I., Cummings, J. L., & Chertkow, H. (2005). The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive Impairment. Journal of the American Geriatrics Society, 53(4), 695-9.

Tracy R. Melzer1,2, Richard Watts1,3, Michael R. MacAskill1,2, Ross Keenan4, Charlotte Graham1,2, Leslie Livingston1,2, John C. Dalrymple-Alford1,5, Tim J. Anderson1,2 1Van der Veer Institute for Parkinson's Disease and Brain Research, 2Department of Medicine, University of Otago, Christchurch, 3Department of Physics and Astronomy, University of Canterbury, Christchurch, 4Christchurch Radiology Group, 5Department of Psychology, University of Canterbury, Christchurch, New Zealand