Impaired cognitive flexibility and disrupted cognitive cerebellum in degenerative cerebellar ataxias

Abstract

There is a clinically unmet need for a neuropsychological tool that reflects the pathophysiology of cognitive dysfunction in cerebellar degeneration. We investigated cognitive flexibility in degenerative cerebellar ataxia patients and aim to identify the pathophysiological correlates of cognitive dysfunction in relation to cerebellar cognitive circuits. We prospectively enrolled degenerative cerebellar ataxia patients with age-matched healthy controls who underwent 3 T 3D and resting-state functional MRI. All 56 participants were evaluated with the Scale for Assessment and Rating of Ataxia and neuropsychological tests including the Wisconsin Card Sorting Test, Trail Making Test, Montreal Cognitive Assessment and Mini-Mental State Examination. From MRI scans, we analysed the correlation of whole-brain volume and cortico-cerebellar functional connectivity with the Wisconsin Card Sorting Test performances. A total of 52 participants (29 ataxia patients and 23 healthy controls) were enrolled in this study. The Wisconsin Card Sorting Test scores (total error percentage, perseverative error percentage, non-perseverative error percentage and categories completed), Trail Making Test A and Montreal Cognitive Assessment were significantly impaired in ataxia patients (P < 0.05) compared to age-matched healthy controls. The Wisconsin Card Sorting Test error scores showed a significant correlation with the ataxia score (P < 0.05) controlling for age and sex. In volumetric analysis, the cerebellar right crus I, II, VIIb and VIII atrophy correlated with non-perseverative error percentage in the ataxia group. In functional connectivity analysis, the connectivity between crus I, II and VIIb of the cerebellum and bilateral superior parietal and superior temporal gyrus was significantly altered in ataxia patients. The functional connectivity between left crus II and VIIb of the cerebellum and dorsolateral prefrontal and superior frontal/parietal cortices showed a positive correlation with perseverative error percentage. The connectivity between left crus VIIb and pontine nucleus/middle cerebellar peduncle showed a significant negative correlation with non-perseverative error percentage in the ataxia group. The impaired cognitive flexibility represented by the Wisconsin Card Sorting Test was significantly impaired in degenerative cerebellar ataxia patients and correlated with disease severity. The Wisconsin Card Sorting Test performance reflects hypoactivity of the cognitive cerebellum and disrupted cortico-cerebellar connectivity in non-demented patients with degenerative cerebellar ataxia.

Keywords: Wisconsin Card Sorting Test; clinical biomarker; cognitive cerebellum; cognitive flexibility; degenerative cerebellar ataxia.

Graphical Abstract

Figure 1

Correlation of SARA scores with cognitive profiles in CA patients. Scatter plots of the SARA score with cognitive profiles of MMSE (A), MoCA (B), TMT-A (C), TMT-B (D) and WCST scores (E-H) in the SCA group. The correlation was analysed with partial correlation with age and sex as cofactors. R- and P-values are described in each figure. SARA, Scale for Assessment and Rating of Ataxia; TMT, Trail Making Test; MoCA, Montreal Cognitive Assessment; MMSE, Mini-Mental State Examination; WCST, Wisconsin Card Sorting Test; CA, cerebellar ataxia.

Figure 2

Comparison of whole-brain and cerebellar volume between CA patients and HCs. (A, B) Comparison of whole-brain volume (A) and cerebellar volume (B) between the CA group and HCs. The comparison was performed with a voxel-wise independent t-test. Multiple corrections were done with the GRF method (voxel P < 0.001, cluster P < 0.01 for whole-brain VBM and voxel P < 0.05, cluster P < 0.05 for SUIT analysis). The reference bar in A and B represent the T-value. L, left; R, right; VBM, voxel-based morphometry; CA, cerebellar ataxia.

Figure 3

Significant correlation of cerebellar volume with WCST score in CA patients. Cerebellar regional correlation with non-perseverative error percentage in WCST. Multiple corrections were done with the GRF method (voxel P < 0.05, cluster P < 0.05). Colour bar represents T-value. CA, cerebellar ataxia; WCST, Wisconsin Card Sorting Test; SUIT: Spatially Unbiased Infratentorial Template; VBM, voxel-based morphometry; L, left; R, right.

Figure 4

Correlation of FC of cognitive cerebellum with the performance of WCST in the CA group. Seed-based whole-brain functional connectomics in the CA group. Multiple corrections were done with the GRF method (voxel P < 0.005, cluster P < 0.01). Colour bar represents T-value. (A) FC correlation with total error percentage with left crus II as seed. (B) FC correlation with non-perseverative error percentage with left crus VIIb as seed. (C) FC correlation with perseverative error percentage with left crus II (left) and VIIb (right) as seeds. Colour bar represents T-value. L, left; R, right.