Longitudinal corpus callosum microstructural decline in early-stage Parkinson's disease in association with akinetic-rigid symptom severity

Abstract

Previous diffusion tensor imaging (DTI) studies of Parkinson's disease (PD) show reduced microstructural integrity of the corpus callosum (CC) relative to controls, although the characteristics of such callosal degradation remain poorly understood. Here, we utilized a longitudinal approach to identify microstructural decline in the entire volume of the CC and its functional subdivisions over 2 years and related the callosal changes to motor symptoms in early-stage PD. The study sample included 61 PD subjects (N = 61, aged 45-82, 38 M & 23 F, H&Y ≤ 2) from the Parkinson's Progressive Markers Initiative database (PPMI). Whole-brain voxel-wise results revealed significant fractional anisotropy (FA) and mean diffusivity (MD) changes in the CC, especially in the genu and splenium. Using individually drawn CC regions of interest (ROI), our analysis further revealed that almost all subdivisions of the CC show significant decline in FA to certain extents over the two-year timeframe. Additionally, FA seemed lower in the right hemisphere of the CC at both time-points, and callosal FA decline was associated with FA and MD decline in widespread cortical and subcortical areas. Notably, multiple regression analysis revealed that across-subject akinetic-rigid severity was negatively associated with callosal FA at baseline and 24 months follow-up, and the effect was strongest in the anterior portion of the CC. These results suggest that callosal microstructure alterations in the anterior CC may serve as a viable biomarker for akinetic-rigid symptomology and disease progression, even in early PD.

Fig. 1. Whole-brain voxel-wise analysis of callosal changes across 24 months.

a Whole-brain ANOVA data. Suprathreshold clusters show FA (red) and MD (blue) changes across the baseline, 12-month, and 24-month follow-up time-points (FDR p < 0.05). b Results from paired t tests comparing the baseline and 24-month time-points. Suprathreshold clusters show a within-subject microstructural decline (FDR p < 0.05). Red indicates areas where FA decreased over time, and blue indicates areas where MD increased over time.

Fig. 2. Longitudinal analysis of individually-based CC ROI’s.

a This bar chart shows the FA values for the full CC ROI and each callosal segment at the baseline (light gray) and 24-month (dark gray) follow-up time-points. Error bars displayed are 95% confidence intervals. b Averaged callosal volume indexed in number of voxels at the baseline and 24-month follow-up time-points. Voxels are 2 mm³. c Individual FA changes for the full corpus callosum from baseline to 24-month follow-up. Each line is a subject, with dashed and solid lines indicating increases and decreases, respectively. Annotations: *p < 0.05; **p < 01; ***p < 0.001; n.s. not significant.

Fig. 3. Lateralized and onset-based callosal microstructural changes.

a Change in FA of the left and the right hemisphere of the corpus callosum over two years. Each line indicates a different callosal segment. b Changes in FA in the left and right hemispheres of the corpus callosum for subjects with left and right symptom onset. I = prefrontal section; II = premotor section; III = motor section; IV = somatosensory section; V = temporal-parietal-occipital section.

Fig. 4. UPDRS-III total and akinetic-rigid scores are associated with FA of the full corpus callosum at both time points.

Scatter plots in the left show across-subject correlation between motor symptoms and baseline CC FA and those in the right show the correlation with 24-months FA. a The bivariate relationship between baseline and 24-month UPDRS-III with baseline CC FA. b The bivariate relationship between baseline and 24-month tremor UPDRS-III subscores with baseline CC FA. c The bivariate relationship between baseline and 24-month akinetic-rigid UPDRS-III subscores with baseline CC FA. Annotations: *p < 0.05; **p < 1.

Fig. 5. Whole-brain regression maps showing changes in callosal FA in association with widespread microstructure decline.

Red indicates areas where callosal FA decline was associated with FA decrease. Blue indicates areas where callosal FA decline was associated with MD increase.

Fig. 6. Segmentation of the corpus callosum.

a Corpus callosum segmentation scheme as in the study by Hofer and Frahm (2006). b An example of a fully segmented hand-drawn corpus callosum ROI in sagittal and axial view. The five callosal segments are illustrated in different colors. I = prefrontal section; II = premotor section; III = motor section; IV = somatosensory section; V = temporal-parietal-occipital section.