Vestibular compensation in acute unilateral medullary infarction

Transcription

1 Vestibular compensation in acute unilateral medullary infarction FDG-PET study Sandra Becker-Bense, MD Hans-Georg Buchholz, PhD Christoph Best, MD Mathias Schreckenberger, MD Peter Bartenstein, MD Marianne Dieterich, MD Correspondence to Dr. Becker-Bense: ABSTRACT Objective: The aim of this fluorodeoxyglucose (FDG)-PET study was to determine whether the activation pattern in patients with an acute unilateral central vestibular lesion (e.g., lesion of the vestibular nucleus) differs from that known in patients with an acute peripheral vestibular deficit. Methods: Twelve patients with circumscribed unilateral medullary brainstem infarctions (6 right, 6 left) causing acute vestibular imbalance underwent resting-state 18 F-FDG-PET. Regional cerebral glucose metabolism was measured twice without any stimulation and with eyes closed: in the acute phase after infarct onset on mean day 8 (range 4 12), and again 6 months later in 7 patients after recovery. Group subtraction analyses and comparisons with a dataset of 12 agematched controls were done with Statistic Parametric Mapping. Results: In the acute stage, the pattern of signal increases differed from that in peripheral vestibular lesions: whereas signals in the infratentorial areas in the contralateral medulla and cerebellum (peduncle, vermis, hemispheres) were increased, areas at the cortical level were largely spared. Signal decreases were found in similar sites in the visual cortex bilaterally. Conclusions: The current data provide evidence that the lesion site significantly modifies the glucose metabolism pattern in an acute vestibular lesion. Different compensation strategies seem to be apparent: after vestibular nucleus lesions, compensation occurs preferably in brainstemcerebellar loops; after peripheral lesions, it occurs at the cortical level. Neurology â 2013;80: GLOSSARY BA 5 Brodmann area; FDG 5 fluorodeoxyglucose; MT 5 middle temporal; PIVC 5 parieto-insular vestibular cortex; rcgm 5 regional cerebral glucose metabolism; SVV 5 subjective visual vertical. Supplemental data at Patients with an acute central vestibular lesion induced by a unilateral infarction of the vestibular nucleus in the pontomedullary brainstem present with vestibular symptoms in the (frontal) roll plane accompanied by unpleasant rotational vertigo and nausea. They spontaneously recover gradually over weeks to months. Until now it has been unclear which structures could be involved in the complex compensation of central vestibular tone imbalance. The multisensory vestibulo-cortical network has been identified in human imaging studies using vestibular stimulation 1 13 ; it is similar to that known from animal studies. 14,15 The core region of this bilateral temporoparietal circuit is the human homolog of the parieto-insular vestibular cortex (PIVC). 14 Simultaneously, during artificial vestibular stimulation, signal decreases have been detected in the visual cortex bilaterally. 5,10,12,13 These findings in healthy volunteers support the concept of a reciprocal inhibitory visual-vestibular interaction, which prevents a potential intersensory mismatch and allows adequate orientation in space and perception of motion. 16 According to this idea, the sensory modality of vision is suppressed when the sensory modality of the vestibular system predominates and vice versa. A fluorodeoxyglucose (FDG)-PET study in patients with an acute unilateral peripheral vestibular loss has supported this concept; these patients likewise showed signal increases in multisensory vestibular areas and decreases in the visual cortex bilaterally in the acute stage of disease. 17 From the Departments of Neurology (S.B.-B., M.D.) and Nuclear Medicine (P.B.), University of Munich; Departments of Nuclear Medicine (H.-G.B., M.S.) and Neurology (C.B.), Johannes Gutenberg-University Mainz; German Center for Vertigo and Balance Disorders (IFB LMU ) (S.B.-B., P.B., M.D.), University of Munich; and Munich Cluster for Systems Neurology (SyNergy) (M.D.), Munich, Germany. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article American Academy of Neurology 1103

2 Our aim was to determine whether this intersensory interaction also occurs in patients with a vestibular tone imbalance due to an acute central vestibular disorder or whether the potential compensatory strategies differ between peripheral and central vestibular disorders. METHODS Patients and clinical examination. Twelve patients (10 males, 2 females; mean age 68.3 years) with acute lateral medullary infarction (6 right-sided, 6 left-sided) were recruited from the interdisciplinary stroke unit of the Johannes Gutenberg-University of Mainz between 2002 and Diagnosis was based on key symptoms, a careful neurologic and neuro-otologic examination including neuro-orthoptic assessment (e.g., Frenzel glasses, fundus photography, and testing for the subjective visual vertical [SVV]) (for methods, see references 18 and 19), and the exclusion of other neurologic disorders. Neurologic and neuro-otologic examinations were performed twice: once during the acute stage and a second time 6 months later during the follow-up, after clinical recovery. The diagnosis of acute infarction was confirmed by T2-weighted and diffusion-weighted MRI sequences of the brainstem in a 1.5T clinical MRI scanner (Siemens, Munich, Germany) within 2 days after symptom onset. The MRI scans were independently analyzed by 2 neuroradiologists. Standard protocol approvals, registrations, and patient consents. The study was conducted according to regulations of the Helsinki Declaration and was approved by the local Ethics Committee and the Radiation Protection Authorities (Bfs). All patients gave informed written consent. PET measures, data acquisition, image reconstruction, and statistical analysis. Each patient underwent 2 18 F-FDG-PET scans in an ECAT Exact PET Scanner (Siemens/CTI, Knoxville, TN) while they lay supine in a quiet, darkened room with eyes closed and without any artificial stimulation: 1) during the acute stage of brainstem infarction (mean at day 8 after symptom onset, range 4 12 days), and 2) 6 months later after clinical recovery. The interval between symptom onset and first PET scan varied for logistical reasons (e.g., availability of the tracer, scanning time, transport logistics). Parametric z score images were calculated. Statistical data analysis was performed with Statistical Parametric Mapping software (SPM 99; Wellcome Department of Cognitive Neurology, London, UK). To have all infarctions on the same side, the data of patients with left-sided brainstem infarction were flipped before normalization to the right. For details of the standardized PET data acquisition, image reconstruction, and statistical analysis, see reference 17. Furthermore, patient PET data were compared with those of 12 sex- and age-matched healthy controls (mean age 67.1 years, no history or signs of former vestibular dysfunction) who had been scanned under equal conditions (standardized procedure) by means of a 2-sample t test. This was useful, because no PET scan before the infarction was available for the patients. A statistical threshold of p, and a minimal cluster size.10 voxels were always applied. Because of partial flipping of the patient data, the terms ipsilateral or contralateral to the lesion side are used throughout the article. The single components of the brainstem-cerebellar loop such as deep cerebellar nuclei, flocculus or fastigial nucleus, are not elaborated because the spatial resolution in PET was limited. RESULTS Patient data. None of the patients had a history of previous CNS disorders or relevant cochlear or vestibular disorders, and none took vestibular sedatives. More detailed patient data, in particular cardiovascular risk factors and the most probable etiology of infarction revealed by the standardized interdisciplinary setup, are given in table e-1 on the Neurology Web site at The laterality quotient for handedness according to the 10-item inventory of the Edinburgh test was 1100 in 8 and 180 in 4 patients; thus, the patients were strongly right-handed. 20,21 MRI gave no evidence of prior silent stroke. All patients exhibited typical vestibular signs in the (frontal) roll plane in the acute stage: transient rotatory vertigo with vomiting (n 5 12), spontaneous nystagmus (n 5 8), and head and/or body tilt (n 5 12). Examination also revealed signs of a tonic vestibular disorder in most patients (9 skew deviation; 9 ocular torsion, 5.4 mean), and ipsiversive tilts of SVV in all 12 (7.7 mean). Seven patients showed a complete ipsilateral ocular tilt reaction with head tilt, skew deviation, and ocular torsion. Depending on the extent of the unilateral medullary infarction in the actual MRI (9 isolated posterolateral medullary, 3 with circumscribed expansion in the caudal pons, no cerebellar infarction), patients showed additional clinical signs. Six months after symptom onset, 7 patients agreed to a second clinical examination and FDG-PET. Dropouts for the second PET scanning were due to 2 deaths (1 cardiac infarction, 1 pulmonary infection), 1 distant relocation, and 2 severe disabilities (1 heart failure, 1 vertebral body fracture). At this stage, none of the remaining 7 patients showed any significant spontaneous nystagmus, skew deviation, or tilts of SVV. The 2 patient groups (5 with 1 PET, 7 with 2 PET images) showed no significant differences in mean age, clinical and neuro-otologic signs at symptom onset, or days until first PET. PET subtraction analysis: Acute stage vs recovery 6 months later. The contrast between the PET during the acute stage and the control scan 6 months later showed mainly bilateral cerebellar signal changes (figure 1A, table 1). In the first scan, regional cerebral glucose metabolism (rcgm) was significantly increased in the lateral medulla and the inferior and middle cerebellar peduncles contralateral to the lesion side, as well as in the vermis (predominantly pyramid and tonsil) and both cerebellar hemispheres, both tonsils, and dentate nuclei (ipsilateral. contralateral to the lesion). It was not clear whether parts of the deep cerebellar nuclei such as the fastigial nucleus were included. The flocculus and the (para)hippocampal area showed predominant ipsilateral glucose hypermetabolism. Cortical rcgm increases were also seen in the medial hypothalamic/thalamic area bilaterally (probably due to vegetative side effects), the anterior cingulate 1104 Neurology 80 March 19, 2013

3 Figure 1 FDG-PET statistical group analysis of patients with vestibular tone imbalance due to acute medullary infarction (A) The PET scan at the acute stage vs a second scan 6 months later after recovery mainly showed cerebellar signal differences due to glucose hypermetabolism in the acute stage. (B) The inverse contrast revealed widespread bilateral signal changes in primary and secondary visual cortex areas (Brodmann areas [BAs] 17 19, MT/V5 BA 19/37, and upper occipital cortex BA 19/39) as well as in temporoparietal areas (superior temporal gyrus, inferior parietal lobule, BA 39/40). These signal changes were caused by signal decreases in the acute stage. (C) Categorical comparison of the PET scans of 12 agematched healthy controls with those of the 12 patients with vestibular tone imbalance due to acute medullary infarctions. Significant differences (probably caused by signal decreases below the normal level in the acute stage of disease) were seen not only in visual, but also in multisensory (vestibular) temporoparietal (e.g., posterior insula, anterior and posterior cingulate gyrus, inferior parietal lobule), ocular motor (cortical eye fields), and the parahippocampal areas bilaterally. The arrow indicates the lesion side. FDG 5 fluorodeoxyglucose. gyrus, and a small area in the medial insula contralaterally. No signal increases were observed in the well-known multisensory vestibular cortical areas (e.g., posterior end of the insula and adjacent retroinsular regions, superior temporal gyrus, inferior parietal lobule, vestibular thalamus, anterior insula, inferior frontal gyrus, and anterior cingulate gyrus), even at the lowered threshold of p, PET subtraction analysis: Recovery vs acute stage. The inverse contrast (figure 1B, table 1) revealed a widespread signal cluster in the visual cortex bilaterally (Brodmann areas [BAs] 17 19) including the motion-sensitive area middle temporal (MT)/V5 (BA 19/37) and merging into the superior occipital gyrus (BA 39) and the precuneus (BA 7/19). Additional signal differences were seen in multisensory vestibular areas, i.e., in parts of the temporoparietal cortex (superior/medial temporal gyrus, BA 22/39; inferior/superior parietal lobule, BA 39/40). These signal changes occurred bilaterally, partly integrated in the large cluster of the visual cortex. Comparisons of the patient datasets with those of age-matched healthy controls suggested that these signal changes were caused by signal decreases in the acute Neurology 80 March 19,

4 Table 1 Areas showing significant differences in glucose metabolism in the statistical subtraction analyses thresholded at p uncorrected: (A) medullary patients: PET acute stage vs 6 months later at recovery; (B) medullary patients: PET at recovery 6 months later vs acute stage; and (C) healthy controls vs patients: PET at acute stage (A) Medullary patients: PET acute stage vs 6 months later at recovery Area R/L X Y Z t Value Cluster Medulla/cerebellum contralesional L 2,395 Lateral medullary brainstem ,048 Inferior/superior semilunar lobe Biventer lobule, vermis (tonsil, pyramid), dentate, inferior/middle cerebellar peduncle Cerebellum ipsilesional R 1,558 Inferior/superior semilunar lobe ,379 Biventer lobule, tonsil, dentate Anterior cingulate gyrus Midline (Hypo)thalamic area Midline (Para)hippocampal area, flocculus R (Para)hippocampal area, flocculus L Medial insula L (B) Medullary patients: PET at recovery vs acute stage Area R/L BA X Y Z t Value Cluster Occipitotemporal cortex L/R ,787 Visual cortex L/R MT/V5, inferior temporal gyrus L/R 19/37 Superior occipital gyrus, precuneus L/R 39 Superior/medial temporal gyrus L/R 22/39 Inferior/superior parietal lobule L/R 39/40 Middle frontal gyrus, anterior insula R 6/ Paracentral lobule L (C) Healthy controls vs patients: PET at acute stage Area R/L BA X Y Z t Value Cluster Visual cortex R/L Lingual gyrus, cuneus Precuneus/posterior cingulum L 7/ Precuneus R Posterior cingulum Anterior cingulum 32/ Inferior parietal lobule R/L 39/ Posterior insula L PIVC Precentral/middle frontal gyrus R/L 6/ Inferior/middle frontal gyrus R/L 44/ , Inferior temporal gyrus R Abbreviations: BA = Brodmann area; MT 5 middle temporal; PIVC 5 parieto-insular vestibular cortex. stage, and not by signal increases above the normal level in the course of the disease. Small clusters were furthermore located in the paracentral lobule (BA 6) contralaterally, and in the middle frontal gyrus/anterior insula (BA 6/47) ipsilaterally to the brainstem lesion. Significant signal differences in the cerebellum and the multisensory core region in the posterior insula (PIVC) were not found, even at a lowered threshold of p, Comparison of patient data with those of healthy controls. PET in the acute stage: Comparison of healthy controls vs patients PET A. Significant signal differences were found bilaterally in the visual cortex (BA 17/ 18/19), and upper occipital areas (precuneus, BA 7), as well as in multisensory (vestibular) temporoparietal areas, e.g., the inferior parietal lobule bilaterally (BA 40), the anterior (BA 32), and posterior cingulate gyri (BA 31), and the contralesional posterior insula (PIVC) (figure 1C, table 1). Bilateral signal differences of the cortical eye fields were seen bilaterally in the middle frontal/precentral gyrus (frontal eye field, BA 6/4), and partly in the prefrontal cortex in the inferior/middle frontal gyrus (BA 44/46/10). The parietal eye fields were integrated in the signal clusters of the posterior parietal lobuli (BA 40). There was an additional ipsilesional signal difference in the inferior/middle temporal gyrus (BA 20). All these signal clusters are attributable to rcgm decreases in the acute stage of brainstem infarction. Comparison of patients PET A vs healthy controls. This inverse contrast revealed only 2 significant clusters due to rcgm increases in the acute stage: 1 within the pontine brainstem above the lesion side, possibly due to crossing brainstem fibers or to a partial volume effect beside the lesion, and 1 in the ipsilesional anterior insula (inferior frontal gyrus, BA 46). PET at recovery: Comparison of healthy controls vs patients PET B. This comparison showed signals in both cerebellar hemispheres and (para)limbic lobes due to rcgm decreases in the course of the disease. Cortical signal differences were not evident. Comparison of patients PET B vs healthy controls. This contrast revealed signal differences in the course of the disease only in the medial temporal gyrus (BA 6, frontal eye field) bilaterally, the postcentral gyrus (BA 43), and the parieto-occipital cortex (precuneus, superior parietal gyrus, superior occipital gyrus, inferior parietal lobule, BA 7/40/19), mainly contralateral to the lesion side. DISCUSSION Patients with acute unilateral lesions of the vestibular nucleus due to a medullary brainstem infarction showed a novel pattern of rcgm changes: there was a significant hypermetabolism in the contralateral medulla (probably comprising the contralateral vestibular nucleus), the contralateral middle and inferior cerebellar peduncles, vermal 1106 Neurology 80 March 19, 2013

5 Figure 2 Schematic drawings of regional cerebral glucose metabolism increases ( activation pattern ) (A) In healthy volunteers during right-sided vestibular caloric stimulation. (B) In patients with a right-sided peripheral vestibular lesion due to vestibular neuritis. (C) In patients with a circumscribed infarction of the vestibular nucleus in the medullary brainstem (all lesions flipped to the right). (A and B according to references 7 and 17, C according to the current study.) structures (predominantly pyramid and tonsil),and both cerebellar hemispheres. These data suggest that central vestibular compensation and recovery processes occurred after medullary infarction mainly in brainstem-cerebellar loops. Thus, the cerebellum s inhibitory control on the vestibular nuclei, which is mediated by vestibulocerebellar networks, 22 appears to be modulated. This type of modulation seems to occur preferably in central vestibular disorders. In contrast, an earlier PET study found that patients with acute unilateral peripheral vestibular lesions due to right-sided vestibular neuritis exhibited a significant rcgm increase especially in ipsilateral multisensory vestibular cortical and subcortical areas (e.g., posterior insula, thalamus, anterior cingulate gyrus, hippocampus), but not in cerebellar regions. 17 Predominant involvement of cortical structures in central compensation of peripheral deficits has also been previously reported in voxelbased morphometry in patients after vestibular neuritis (atrophy in the superior temporal gyrus and left hippocampus, intensity increases in the medial vestibular nuclei, right gracile nucleus, and pontine commissural fibers). 23 The simultaneous occurrence of signal decreases within areas of the visual cortex bilaterally including themotion-sensitiveareamt/v5andinadditionin some multisensory vestibular areas such as the inferior parietal lobule and superior temporal gyrus during the acute stage also does not fit the classic visual-vestibular interaction pattern found earlier in healthy volunteers during artificial vestibular stimulation (for review, see reference 15) or in patients with acute peripheral vestibular failure 17 : all multisensory vestibular areas typically showed signal increases. 15,17 Thus, the lesion site appears to significantly modify the activation pattern, not only at the brainstem-cerebellar level but also, to a lesser extent, at the cortical level where signal increases turn into decreases. This change seems logical because the ascending vestibular pathways and the thalamocortical network are bilateral, and a peripheral vestibular lesion leaves all mediating pathways and centers intact. Early compensation of a peripheral vestibular lesion might occur particularly at the brainstem level (figure 2B) via the commissural fibers between the vestibular nuclei. Although the mediating structures are all intact, they are now predominantly in use on the healthy side (not bilaterally) and thus mediate the signal via the contralateral pathways to the network in the contralateral cortical hemisphere. The result is an imbalance of activity at the cortical level as was documented by an earlier PET study (figure 2B). 17 By the activation of the vestibular system with a shift to the contralesional cortical hemisphere, the reciprocal inhibitory visual-vestibular interaction is preserved. 17 In contrast, a central lesion affects the vestibular nucleus of 1 side and causes much more acute damage Neurology 80 March 19,

6 (figure 2C): it interrupts the ascending pathways to the thalamus and cortex on the ipsilateral side as well as vestibular projections to the contralateral vestibular nucleus and the ipsilateral vestibulocerebellar structures. Consequently, an erroneous signal to the cortex causes a downregulation in parts of both cortical sensory systems, the visual (broad signal decreases bilaterally) more than the multisensory vestibular areas (signal decreases in the superior temporal gyrus and the inferior parietal lobule; only small signal increases in the contralateral insula). Thus, the pattern of visualvestibular interaction changes, and activating processes predominantly occur within brainstem-cerebellar loops of the contralateral healthy side close to the damaged vestibular nucleus. Both vestibular nuclei complexes are closely connected; they integrate semicircular canal and otolith information from the vestibular periphery and other sensory systems (somatosensory, optokinetic, visual, and neck proprioceptive) Integrative processes are partially achieved by the vestibulocerebellum and by the so-called oculomotor vermis for eye movements The fibers run from the vestibular nucleus via the cerebellar peduncle, the dentate nucleus to the cerebellar vermis, deep cerebellar nuclei, and to the flocculus and fastigial nucleus (monkeys 31,32 ). Because lesions of these structures lead to a gaze-evoked nystagmus, 29 which regularly occurs in patients with medullary infarctions, these regions seem to represent a major cerebellar part of the neural integrator control system for head, eye, and body movements. 33 Thus, our current data on metabolism increases in most of these cerebellar regions indicate that compensation in medullary lesions is mainly achieved by readjusting integration processes at the brainstem-cerebellar level mediated by the intact contralateral brainstem, cerebellar peduncle, and flocculus to the dentate nuclei, tonsils, and vermis bilaterally. The primary visual and motion-sensitive visual cortical areas were also downregulated in both cortical hemispheres to suppress blurred vision caused by oscillopsia due to the regularly occurring spontaneous and gaze-evoked nystagmus in the acute phase of disease. Our PET findings were obtained between days 4 and 12 (mean day 8). This interval corresponds to the time of the first PET scan examination of patients with acute peripheral vestibular lesions (mean day 6.6). 17 To date, our knowledge about the time courses of central processes in central vestibular lesions in animals and humans is quite limited. Human lesion studies have shown that the measurable effects of vestibular dysfunction due to an acute vestibular nucleus lesion decrease gradually over days to several weeks. 18,19 FDG-PET data after acute labyrinthectomy in rats 34 showed 4 successive central compensational phases (nuclear, commissural, proprioceptive, and cerebellar phase). The cerebellar phase in the rats was reached on mean day 8, which nicely fits with our study in humans (also on mean day 8). The significant hypermetabolism in the contralateral medulla (comprising the contralateral vestibular nucleus) and the cerebellum apparently indicate that the human vestibular commissural system is involved relatively early, on analogy to the compensatory stage at days 5 to 10 in rats. Six months later, the pattern in our study was totally different: cerebellar hemispheres showed rcgm decreases, whereas visual parieto-occipital areas showed increases within the contralateral hemisphere. This was probably attributable to processes of substitution in the later stage. It has been reported that focal cerebral ischemia may elicit an inflammatory response that results in a focal increase in FDG uptake induced by activated microglia and macrophages. 35 This could result in an increase in FDG signal not due to increased neuronal activity in the area directly surrounding the infarction zone. Therefore, the increase in the PET signal in the lateral medulla observed in the first PET scan must be interpreted with care. Relevant effects on glucose metabolism outside the brainstem, e.g., in the cerebellum or cortex, however, seem unlikely. Earlier water-activation PET studies in infarctions affecting the vestibular nucleus showed normal activity in the ipsilesional hemisphere and decreased activity only in the contralateral vestibular cortex. 36 Thus, an unspecific or disconnecting effect of the small brainstem lesion on both hemispheres seems unlikely too. Until now, there has been no indication in the literature of unspecific effects of small brainstem infarctions on the uptake of FDG in widespread cortical and subcortical areas. Even if this was the case, the statistical normalization to the global mean ( proportional scaling ) performed in the current SPM analyses would remove or minimize such an effect. To our knowledge, the current PET study is the first imaging study to explore vestibular compensation processes of the human brain after a central lesion of the vestibular nucleus. Further imaging studies of patients at different time intervals after vestibular nucleus lesions will help elucidate the sequence of adaptive changes made in the human brain. AUTHOR CONTRIBUTIONS S. Becker-Bense designed and conceptualized the study, recruited the patients, oversaw the measurements, analyzed the data, designed the figures, and wrote the manuscript. H.-G. Buchholz performed the PET measurements and statistical analyses of the PET data. C. Best recruited the patients and coordinated the study. M. Scheckenberger and P. Bartenstein participated in the design of the PET protocol, evaluated the data, and edited the manuscript. M. Dieterich conceptualized and supervised the study, evaluated the data, and wrote the manuscript. ACKNOWLEDGMENT The authors thank Judy Benson for critically reading the manuscript, Anja Schroer for orthoptic assistance, and the PET staff for technical assistance Neurology 80 March 19, 2013

7 DISCLOSURE S. Becker-Bense, H.-G. Buchholz, and C. Best report no disclosures relevant to the manuscript. M. Scheckenberger has a funded scientific cooperation with Philips Healthcare, but it is not related to the topic of this study. P. Bartenstein reports no disclosures relevant to the manuscript. M. Dieterich received research support from the Deutsche Forschungsgemeinschaft. Go to Neurology.org for full disclosures. Received May 7, Accepted in final form December 5, REFERENCES 1. Bottini G, Sterzi R, Paulesu E, et al. Identification of the central vestibular projections in man: a positron emission tomography activation study. Exp Brain Res 1994;99: Bucher SF, Dieterich M, Wiesmann M, et al. Cerebral functional MRI of vestibular, auditory, and nociceptive areas during galvanic stimulation. Ann Neurol 1998;44: Lobel E, Kleine JF, Le Bihan D, Leroy-Willig A, Berthoz A. Functional MRI of galvanic vestibular stimulation. J Neurophysiol 1998;80: Brandt T, Dieterich M. The vestibular cortex: its locations, functions, and disorders. Ann NY Acad Sci 1999; 871: Bense S, Stephan T, Yousry TA, Brandt T, Dieterich M. Multisensory cortical signal increases and decreases during vestibular galvanic stimulation (fmri). J Neurophysiol 2001;85: Fasold O, von Brevern M, Kuhberg M, et al. Human vestibular cortex as identified with caloric stimulation in functional magnetic resonance imaging. Neuroimage 2002; 17: Dieterich M, Bense S, Lutz S, et al. Dominance for vestibular cortical function in the non-dominant hemisphere. Cereb Cortex 2003;13: Emri M, Kisely M, Lengyel Z, et al. Cortical projection of peripheral vestibular signaling. J Neurophysiol 2003;89: Naito Y, Tateya I, Hirano S, et al. Cortical correlates of vestibulo-ocular reflex modulation: a PET study. Brain 2003;126: Stephan T, Deutschländer A, Nolte A, et al. Functional MRI of galvanic vestibular stimulation with alternating currents at different frequencies. Neuroimage 2005;26: MiyamatoT,FukushimaK,TakadaT,deWaeleC,VidalPP. Saccular stimulation of the human cortex: a functional magnetic resonance imaging study. Neurosci Lett 2007; 423: Janzen J, Schlindwein P, Bense S, et al. Neural correlates of hemispheric dominance and ipsilaterality within the vestibular system. Neuroimage 2008;42: Schlindwein P, Mueller M, Bauermann T, Brandt T, Stoeter P, Dieterich M. Cortical representation of saccular vestibular stimulation: VEMPs in fmri. Neuroimage 2008; 39: Guldin WO, Grüsser OJ. Is there a vestibular cortex? Trends Neurosci 1998;21: Dieterich M, Brandt T. Functional imaging of peripheral and central vestibular disorders. Brain 2008;131: Brandt T, Bartenstein P, Janek A, Dieterich M. Reciprocal inhibitory visual-vestibular interaction: visual motion stimulation deactivates the parieto-insular vestibular cortex. Brain 1998;121: Bense S, Bartenstein P, Lochmann M, Schlindwein P, Brandt T, Dieterich M. Metabolic changes in vestibular and visual cortices in acute vestibular neuritis. Ann Neurol 2004;56: Dieterich M, Brandt T. Wallenberg s syndrome: lateropulsion, cyclorotation and subjective visual vertical in thirtysix patients. Ann Neurol 1992;31: Dieterich M, Brandt T. Ocular torsion and tilt of the subjective visual vertical are sensitive brainstem signs. Ann Neurol 1993;33: Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971;9: Salmaso D, Longoni AM. Problems in the assessment of hand preference. Cortex 1985;21: Dutia MB. Mechanisms of vestibular compensation: recent advances. Curr Opin Otolaryngol Head Neck Surg 2010; 18: zu Eulenburg P, Stoeter P, Dieterich M. Voxel-based morphometry depicts central compensation after vestibular neuritis. Ann Neurol 2010;68: Barmack NH. Central vestibular system: vestibular nuclei and posterior cerebellum. Brain Res Bull 2003;60: Cullen KE. The vestibular system: multimodal integration and encoding of self-motion for motor control. Trends Neurosci 2012;35: Angelaki DE, Bush GA, Perachio AA. Two-dimensional spatio-temporal coding of linear acceleration in vestibular nuclei neurons. J Neurosci 1993;13: Robinson FR, Fuchs AF. The role of the cerebellum in voluntary eye movements. Annu Rev Neurosci 2001;24: Baier B, Stoeter P, Dieterich M. Anatomical correlates of ocular motor deficits in cerebellar lesions. Brain 2009;132: Baier B, Dieterich M. Incidence and anatomy of gazeevoked nystagmus in patients with cerebellar lesions. Neurology 2011;76: Leigh RJ, Zee DS. The Neurology of Eye Movements, 4th ed. New York: Oxford University Press; Angelaki DE, Hess BJ. The cerebellar nodulus and ventral uvula control the torsional vestibulo-ocular reflex. J Neurophysiol 1994;72: Angelaki DE, Hess BJ. Lesion of the nodulus and ventral uvula abolish steady-state off-vertical axis otolith response. J Neurophysiol 1995;73: Zee DS, Leifg RJ, Mathieu-Millaire F. Cerebellar control of ocular gaze stability. Ann Neurol 1980;7: Schniepp R, Zwergal A, La Fougere C, et al. Sequential micro-pet based brain imaging in the unilateral labyrinthectomy rat model: insights to compensatory brain plasticity. J Neurol 2009;256:S13 S14. Abstract. 35. Schroeter M, Dennin MA, Walberer M, et al. Neuroinflammation extends brain tissue at risk to vital peri-infarct tissue: a double tracer [11C]PK and [18F]FDG-PET study. J Cereb Blood Flow Metab 2009;29: Dieterich M, Bense S, Stephan T, Brandt T, Schwaiger M, Bartenstein P. Medial vestibular nucleus lesions in Wallenberg s syndrome cause decreased activity of the contralateral vestibular cortex. Ann NY Acad Sci 2005;1039: Neurology 80 March 19,