Cortical dysfunction in patients with Huntington's disease during working memory performance

Abstract

Previous functional neuroimaging studies on executive function suggested multiple functionally aberrant cortical regions in patients with Huntington's disease (HD). However, little is known about the neural mechanisms of working memory (WM) function in this patient population. The objective of this study was to investigate the functional neuroanatomy of WM in HD patients. We used event-related functional magnetic resonance imaging and a parametric verbal WM task to investigate cerebral function during WM performance in 16 healthy control subjects and 12 mild to moderate stage HD patients. We excluded incorrectly performed trials to control for potential accuracy-related activation confounds. Voxel-based morphometry (VBM) was used to control for confounding cortical and subcortical atrophy. We found that HD patients were slower and less accurate than healthy controls across all WM load levels. In addition, HD patients showed lower activation in the left dorso- and ventrolateral prefrontal cortex, the left inferior parietal cortex, the left putamen, and the right cerebellum at high WM load levels only. VBM revealed gray matter differences in the bilateral caudate nucleus and the thalamus, as well as in inferior parietal and right lateral prefrontal regions. However, volumetric abnormalities in the patient group did not affect the activation differences obtained during WM task performance. These findings demonstrate that WM-related functional abnormalities in HD patients involve distinct WM network nodes associated with cognitive control and subvocal rehearsal. Moreover, aberrant cortical function in HD patients may occur in brain regions, which are relatively well preserved in terms of brain atrophy.

Figure 1

fMRI activation paradigm, shown for a trial of load level 2: In the stimulus period, three capital gray letters appeared on a black screen (1,500 ms). One, two, or three of these letters would then become highlighted at the end of the stimulus period (500 ms). Subjects were instructed that, during the subsequent delay period (6,000 ms), they were to focus only on those letters that were highlighted and to memorize the letters that followed them in the alphabet (manipulated set). In the probe period (2,000 ms), a lower case letter was presented, and subjects had to indicate whether this letter was or was not part of the manipulated set. The control condition displayed three gray Xs and required a stereotype button press in response to the presentation of a small x during the probe period. In this example, the probe (t) was part of the manipulated set (starting from the highlighted letters S and G).

Figure 2

Behavioral performance during the fMRI task. Upper Panel: Average reaction times across all load levels (±standard error), shown for healthy controls and HD patients. Lower Panel: Average accuracy (±standard error) across all load levels. * indicates significant differences at P < 0.05.

Figure 3

Brain regions in which healthy controls showed relatively more cerebral activation compared with HD patients. Results of the 2nd level between‐group ANCOVA (load‐by‐group contrasts at load levels 2 and 3, voxel level P < 0.001 uncorrected, P < 0.05 corrected for spatial extent).

Figure 4

Mean activation effects (estimated beta parameters) in the left middle frontal gyrus, the left inferior frontal gyrus, the left inferior parietal cortex, the left putamen, and the right cerebellum. The β parameters were extracted from the between‐group ANCOVA (significant group differences between healthy controls and HD patients; center of local maximum at the most significant voxel; P < 0.001 at the voxel level, P < 0.05 corrected for spatial extent).

Figure 5

Gray matter volume reductions in HD patients displayed on the glass brain implemented in SPM2, maximum‐intensity‐projected (MIP) regions; results of the ANCOVA, P < 0.001 (familywise error corrected for multiple comparisons).