The fibre network contributing to hemiballism

Background: Hyperkinetic disorders (hemiballism, hemichorea, hemidystonia) are rare complications of ischemic basal ganglia lesions and manifest in about 1% of stroke patients (Ghika-Schmid et al., 1997). In the majority of cases, they resolve within weeks or months. It is interesting to note that no direct single-lesion-symptom mapping could be established for vascular hyperkinetic disorders (Postuma, 2003). Hemiballism was seen after lesions of the STN, GPi, thalamus, caudate and parietal lobe. This observation of various anatomical lesions together with the swift clinical recovery of symptoms generated our hypothesis that ischemic hyperkinetic disorders may be due to an affection of a fibre network rather than to an isolated nuclear lesion.

Objective: The aim of this study was to explore whether ischemic lesions producing hyperkinetic disorders are connected to different white matter tracts than lesions without hyperkinetic symptoms.

Methods: The archives of our hospital were screened for patients with hemiballism, hemichorea and hemidystonia after ischemic lesions to the thalamus, STN, pallidum and caudate. Exclusion criteria were territorial infarctions, no immediate temporal correlation between symptom and lesion (clinical onset within 30 days), a history of autoimmune/paraneoplastic diseases and missing CT/MRI scan. Nine patients presenting with hemiballism showed no exclusion criteria and were classified into the group of “symptomatic lesions” (symptomatic, i.e. producing hyperkinetic disorders) and 9 patients with ischemic basal ganglia lesions without hyperkinetic symptoms were chosen as asymptomatic (i.e. without hyperkinetic disorders) controls. Diffusion MR imaging (64 DIR, 2.4 mm) was obtained in 16 right-handed healthy controls. Symptomatic and asymptomatic lesions, as detected on patients' MR scan, were coregistered onto the healthy controls' FA-maps and probabilistic tractography initiated from the lesions. For each lesion, an average tractography map was obtained from the healthy controls and tractography patterns were compared for the symptomatic and asymptomatic group.

Results: Symptomatic lesions were primarily located within the GPi and thalamus, asymptomatic lesions were seen in the GPe and putamen; thus, symptomatic lesions presented more mesially than asymptomatic ones. Within symptomatic fibre tracts, connectivity was strongest for STN, GPi and caudate. Also, symptomatic fibres were seen to cross more to the contralateral hemisphere.

Conclusions: Our study provides evidence of a distinct fibre network associated with ischemic hyperkinetic disorders. Thus, we opt for a symptom-network mapping instead of a symptom-lesion mapping in hemiballism.

Figure 1 displays fibre tracts seeded from symptomatic (red) and asymptomatic (blue) lesions normalised to the MNI (Montreal Neurologic Institute) template. Fibres from symptomatic lesions connect stronger to the contralateral cerebellum, the contralateral hemisphere via corpus callosum and GPi.

Figure 2 displays statistical analysis of connectivity strength across basal ganglia subareas for symptomatic (red) and asymptomatic (blue) fibre tracts. Connectivity strength was quantified as mean intensity of voxels within a predefined anatomical mask. There were significant differences in fibre intensity of symptomatic tracts between mesial (GPi) and lateral (Putamen) basal ganglia regions, but not for asymptomatic tracts.

References:

1. Ghika-Schmid F, Ghika J, Regli F, Bogousslavsky J (1997) Hyperkinetic movement disorders during and after acute stroke: the Lausanne Stroke Registry. Journal of the neurological sciences 146 (2): 109 – 116.

2. Postuma RB, Lang AE (2003) Hemiballism: revisiting a classic disorder. Lancet neurology 2 (11): 661 – 668.