Klinische Neurophysiologie 2014; 45 - V10
DOI: 10.1055/s-0034-1371189

Imaging the vestibular system after acute unilateral midbrain infarction – A combined lesion-behaviour mapping and FDG-PET study

S Becker-Bense 1, 2, HG Buchholz 3, B Baier 4, M Schreckenberger 3, P Bartenstein 2, 5, 6, M Dieterich 1, 2, 6
  • 1LMU München, Department of Neurology, München, Deutschland
  • 2LMU München, German Center for Vertigo and Balance Disorders (DSGZ), München, Deutschland
  • 3University Mainz, Department of Nuclear Medicine, Mainz, Deutschland
  • 4University Mainz, Department of Neurology, Mainz, Deutschland
  • 5LMU München, Department of Nuclear Medicine, München, Deutschland
  • 6München Cluster of Systems Neurology – SyNergy, München, Deutschland

Introduction:

The aim of this study was twofold, to determine 1) relevant vestibular midbrain structures by voxelwise lesion-behavior mapping (VLBM), and 2) whether PET activation pattern in acute midbrain lesions affecting ascending vestibular tracts differs from that found in patients with acute medullary [1] or peripheral vestibular lesions [2].

Methods:

Subjective visual vertical (SVV) was tested in 17 patients with acute unilateral mesothalamic strokes. On the basis of evident SVV tilt (n = 8), patients were devided in two groups before VLBM. Patients with SVV tilt (n = 8) underwent resting state FDG-PET twice: in the acute phase (A), and 6 months later after recovery in five patients (B). Group subtraction analyses and comparisons with a dataset of healthy controls were done with Statistic Parametric Mapping.

Results:

VLBM showed that voxels affecting paramedian meso-diencephalic regions, particularly Interstitial Nucleus of Cajal (INC) and medial longitudinal fasciculus (riMLF) were damaged up to 70% more frequently in patients with SVV tilt. The contrast of PET A vs. B gave bilateral signal differences mainly in limbic and inferior temporal cortex areas partly merging into the very inferior insula region, but no clusters in known higher cortical vestibular areas. Comparison of PET A with controls revealed glucose increases in the cerebellum, whereas at cortical level bilateral signal decreases dominated.

Conclusions:

The INC and riMLF represent a midbrain vestibular pathway to the cortex. If this ”gateway“ is damaged, activations within the known cortical vestibular network are missing. Compensation seems to occur preferably in brainstem-cerebellar loops and via functional “deactivation“ of cortical structures bilaterally.