Article Text

Research paper
Nystagmus in SCA territory cerebellar infarction: pattern and a possible mechanism
  1. Hyung Lee1,2,
  2. Hyun-Ah Kim1,2
  1. 1Department of Neurology, Keimyung University School of Medicine, Daegu, South Korea
  2. 2Department of Neurology, Brain Research Institute, Keimyung University School of Medicine, Daegu, South Korea
  1. Correspondence to Dr Hyun-Ah Kim, Department of Neurology, Brain Research Institute, Keimyung University School of Medicine, 194 Dongsan dong, Daegu 700-712, South Korea; kha0206{at}dsmc.or.kr

Abstract

Objectives To investigate the frequency and pattern of nystagmus associated with isolated cerebellar infarction in the territory of the superior cerebellar artery (SCA) and to discuss its possible mechanism.

Methods We identified 41 consecutive patients with isolated SCA territory cerebellar infarction diagnosed by MRI. Each patient completed a standardised dizziness questionnaire and underwent neurotological evaluations. Eye movements were recorded using 3-dimensional video-oculography during the acute period.

Results Approximately half (19/41) of the patients experienced true vertigo early in the course of the SCA distribution infarct. 11 (27%) of the 41 patients showed spontaneous nystagmus (SN) or direction changed bidirectional gaze-evoked nystagmus (GEN). SN was observed in 10 patients (24%) and the horizontal component of SN was predominant in most case (80%, 8/10) and always beat towards the lesion side. Direction changed bidirectional GEN was observed in five patients and was mostly (4/5) accompanied by SN. Lesion subtraction analyses revealed that damage to the rostral anterior cerebellum including the ala of the central lobule and part of the quadrangular lobule was more frequent in patients with nystagmus compared to patients without nystagmus (9/11, 82% vs 11/30, 37%) (p=0.015). In most (82%, 9/11) patients with SN or GEN, the nystagmus subsided within 1 week after hospitalisation.

Discussion Vertigo and nystagmus in SCA territory cerebellar infarction are more common than previously thought. Ipsilesional SN may result from damage to the anterior lobe of the cerebellum, which transmits the vestibular output to the fastigial nucleus.

  • Vertigo
  • Cerebellar Disease
  • Cerebrovascular Disease

Statistics from Altmetric.com

Introduction

Isolated cerebellar infarction in the territory of the superior cerebellar artery (SCA) is one of the most common ischaemic stroke syndromes in the posterior circulation.1 Limb ataxia, gait disturbance and dysarthria are common signs in patients with SCA cerebellar infarction, whereas vestibular dysfunction with vertigo and/or nystagmus is usually considered to be a less common symptom because the rostral cerebellum supplied by the SCA does not have significant vestibular connections.1–3 Terao et al reported patients with SCA territory cerebellar infarction presenting as cerebellar symptoms and signs. In their report, none of the patients showed vestibular dysfunctions such as vertigo or nystagmus.3 Traditionally, the vestibulo-ocular portion of the cerebellum is primarily located in the flocculonodular lobes, which are supplied by branches of the anterior inferior cerebellar artery and posterior inferior cerebellar artery (PICA).1 ,2 ,4

Several studies of SCA territory cerebellar infarction have described an uncommon clinical presentation with vertigo or nystagmus.2 ,5–8 In 1998, Stangel et al6 described eight patients with bilateral SCA territory infarction: three of the 8 (38%) patients had true vertigo as part of their stroke symptoms. However, their patients also had brainstem signs such as gaze palsy, weakness or sensory changes.6 Furthermore, some patients (3/8) were diagnosed by brain CT, which is less accurate in detecting acute ischaemic lesions within the posterior fossa.9 In 2006, Lee et al8 reported that three of the 14 patients with isolated cerebellar infarction in the territory of the medial branch of the SCA diagnosed by brain MRI showed true vertigo. However, the report did not perform eye movement recording to analyse nystagmus, which is a representative sign of vestibular dysfunction. To the best of our knowledge, there has been no systematic study that carefully examines the nature of nystagmus associated with isolated SCA territory cerebellar infarction. Thus, we sought to assess the frequency and the pattern of nystagmus in SCA territory cerebellar infarction, and to identify the structure involved in the generation of nystagmus.

Methods

Between January 2006 and September 2011, we identified 41 consecutive patients with unilateral isolated (without brainstem involvement) SCA territory cerebellar infarction from the acute stroke registry at Keimyung University Dongsan Medical Center. In all patients, MRIs including diffusion-weighted images (DWI) and magnetic resonance angiography were performed shortly after the onset of symptoms (mean interval: 13.5 h, range 1.5–73.0 h). The cerebellar territory of the SCA was determined according to the magnetic resonance (MR)-anatomic template, a previously validated MR-anatomic template for the delineation of the arterial territory.10 Because of the possible supply of the SCA hemispheric territory by the PICA, we considered a diagnosis of cerebellar infarct in the SCA only when the infarct spared the nodulus, and dorsomedial part of the caudal cerebellar hemisphere, which are classically supplied by the PICA.9 We also excluded patients with lesions in the middle cerebellar peduncle or brainstem. The vascular territory was determined by consensus between two neurologists (HL, H-AK) who independently reviewed the MRIs.

Each patient completed a standard dizziness questionnaire and underwent a neurotological evaluation, including a thorough history and neurotological examination performed by an experienced neurotologist (HL) within 3 h after patient's arrival at hospital. The standard dizziness questionnaire included detailed descriptions of vestibular symptoms such as the type (ie, vertigo or non-vertiginous dizziness) and duration of dizziness. Vertigo was defined as a spinning illusion of the environment. The usefulness of our dizziness questionnaire was validated in our previous studies.11–13

Eye movements were recorded using 3-dimensional video-oculography (SMI, Teltow, Germany, resolution of 0.1°, sampling rate of 60 Hz) during the acute period (mean interval from onset: 22 h, range 4–105 h). Digitised eye position data were analysed using MATLAB software. Spontaneous nystagmus (SN) was recorded with and without fixation. The intensity of SN was determined using the mean slow phase velocity (SPV, °/s) of the nystagmus without fixation for 10 s. Gaze-evoked nystagmus (GEN) was induced by horizontal (±30°) target displacement. Head shaking nystagmus (HSN) was induced by rotating the patient's head rapidly back and forth about the yaw axis at roughly 2 Hz and 30° in amplitude for 15 s. A head impulse test was also carried out manually by the rapid rotation of the head approximately 20° in amplitude in the yaw axis. The detailed testing techniques have been previously published.14

To identify the structures involved in the generation of nystagmus, the overlapped lesions of the nystagmus group were subtracted from those of the comparison (ie, non-nystagmus) group and vice versa, revealing a percentage overlay plot. We used DWIs obtained within the first 48 h after stroke onset and fluid-attenuated inversion recovery sequences when imaging was conducted 48 h or later for overlay technique and volume measurement. We combined the MRI images of the nystagmus group and non-nystagmus group by flipping the image of the lesions of the left brain-damaged subjects from the left to the right side. Thus all MRI images were viewed anatomically from the right side of the figure, which corresponded to the right side of the brain. Using MRIcro software (http://www.mricro.com), the lesions were mapped on slices of a T1-weighted template MRI scan from the Montreal Neurological Institute (http://www.bic.mni.mcgill.ca/cgi/icbm_view).15 ,16 We also measured the lesion volume using the Medical Image Processing, Analysis, and Visualization (MIPAV) software as previously described.17 Statistical analysis was conducted using SPSS V.18.0 for Windows.

All experiments complied with the tenets of the Declaration of Helsinki and the study protocol was reviewed and approved by our institutional review board.

Results

All patients had an isolated unilateral cerebellar (without brainstem involvement) infarction in the SCA territory, but two patients showed mild compression towards the superior cerebellar peduncle due to oedema. All of the patients with SCA territory cerebellar infarct had a normal head impulse test result to either side. Eleven (27%) of the 41 patients with isolated SCA territory cerebellar infarction showed SN (n=10, 24%) or direction changed bidirectional GEN (n=5, 12%). The SN was predominantly horizontal in most (n=8, 80%) patients and always beat towards the lesion side. The intensity of the ipsilesional SN was commonly attenuated by visual fixation (figure 1). Primary positional downbeat nystagmus was found in only two patients and was not attenuated by visual fixation. Eight patients had GEN—in five of these patients, the nystagmus beat in the direction of the gaze, that is, direction-changed bidirectional GEN. The other three patients showed nystagmus that remained in the same direction (that of the SN) on gaze to either side, that is, direction-fixed unidirectional GEN beating towards the side of the lesion (see video in the online supplementary materials). Direction changed bidirectional GEN was commonly accompanied by SN (80%, 4/5). Only one patient had direction changed bidirectional GEN in the absence of SN. HSN was found in six patients and was mostly down beating (n=4). HSN also occurred in four patients with no SN.

Figure 1

Three-dimensional video-oculography (SMI, Teltow, Germany, resolution of 0.1°, sampling rate of 60 Hz) shows a left beating spontaneous nystagmus (A). Interestingly, this patient was able to suppress his nystagmus during visual fixation (A). T2-weighted axial MRI shows an acute infarct in the left superior cerebellum including the ala of the central lobule in the same patient. RH, horizontal movement of the right eye; RV, vertical movement of the right eye; RT, torsional movement of the right eye.

Early in the course of the SCA territory infarction (ie, hyperacute period), approximately half (19/41, 47%) of the patients experienced true vertigo with a spinning sensation of the environment. Before the patient's arrival to the emergency room, however, vertigo spontaneously subsided in 10 patients. Only 22% (9/41) of the patients continued to have vertigo after their arrival to the emergency room. All patients with nystagmus complained of vertigo (n=7) or dizziness (n=4) at arrival. On the other hand, vertigo occurred in only two patients without nystagmus, and about 35% (n=10) of the patients reported neither vertigo nor non-vertiginous dizziness. Patients with nystagmus showed a higher incidence of nausea/vomiting (p=0.033) and tended to have shorter intervals from onset to examination compared to patients without nystagmus (p=0.037) (table 1). In most (82%, 9/11) patients with nystagmus, the nystagmus disappeared within 1 week after hospitalisation.

Table 1

Comparison of demographic and clinical data between patients with and without nystagmus

In both the nystagmus and non-nystagmus groups, the lesions included the central lobule, the ala of the central lobule, quadrangular lobule, simple lobule, superior semilunar lobule, culmen, declive and dentate nucleus. Figure 2A,B shows lesion density plots for each of the two groups. An overlay plot of the subtracted superimposed lesions of the nystagmus group minus the non-nystagmus group revealed the maximum overlap at the rostral anterior cerebellum including the ala of the central lobule and part of the quadrangular lobule (figure 2C). The ala of the central lobule and part of the quadrangular lobule were more frequently affected in the nystagmus group (9/11, 81.8%) compared to the non-nystagmus group (11/30, 36.7%) (p=0.015). There were no differences in lesion volume between the two groups (6.03 vs 3.99, p=0.103).

Figure 2

Overlapping MRI lesions in patients with (A) or without (B) nystagmus. Left-sided lesions are flipped to the right side. The number of overlapping lesions is illustrated by different colours that code for increasing frequencies, which range from violet (n=1) to red (n=11) in the nystagmus group and n=30 in non-nystagmus group). Overlay plot of the subtracted superimposed lesions of patients with nystagmus minus the comparison (non-nystagmus) group. (C) The percentage of overlapping lesions of the group with nystagmus after subtraction of the comparison group is illustrated by five different colours, where dark red represents a difference of 1–20% and white–yellow represents a difference of 81–100%. Each colour represents increments of 20%. The regions are coloured from dark blue (difference of 1–20%) to light blue (difference of 81–100%), indicating regions that were damaged more frequently in patients without nystagmus. The Talairach z-coordinates of each transverse slice are given. This figure illustrates that the rostral anterior cerebellum is the anatomical area related to the nystagmus, which includes the ala of the central lobule. This figure is only reproduced in colour in the online version.

Because 23 (56%) patients had an obvious cardiac source of emboli such as atrial fibrillation or valvular heart disease and no definite arterial occlusion or stenosis within the vertebrobasilar system on brain MR angiography, the cause of the infarction was most likely cardiac emboli in these patients. Local atherosclerosis with stenosis of the basilar or vertebral arteries was identified in 11 patients. Forty-six per cent (5/11) of patients with nystagmus had an obvious cardiac source of emboli.

Discussion

In the present study, approximately 30% of patients with isolated SCA cerebellar infarction showed SN that mostly beat towards the lesion side or direction changed bidirectional GEN in the acute period. The occurrence of nystagmus in isolated SCA cerebellar infarction is associated with lesions in the rostral anterior cerebellum including the ala of the central lobule and part of the quadrangular lobule, which are located around the upper pons.

Unlike a previous study reported by Lee et al,8 in which only 14% of patients had vertigo, our patient had a relatively high incidence of vertigo and nystagmus. We used a standard dizziness questionnaire to determine the type of dizziness and recorded eye movements during the acute period in all patients. This may explain the relatively high incidence of vestibular symptoms and signs in our study.

It was surprising that approximately half of the patients complained of vertigo early in the course of the infarction, suggesting that vestibular imbalance commonly occurred in the SCA territory infarction. It has been found that efferent fibres from not only the nodulus and uvular but also the lateral part of the anterior lobe project to the ipsilateral vestibular nuclei in animal studies, although efferent projections from the anterior lobe developed less densely compared to projections that emanated from the nodulus and uvula.18 These cerebellar efferents were also connected to the fastigial nucleus bilaterally.18 Electrophysiological studies have demonstrated excitatory projections from the unilateral fastigial nucleus to the ipsilateral medial vestibular nucleus (figure 3A).19 Thus, the fastigial nucleus may contribute to the maintenance of a balanced resting discharge between bilateral vestibular nuclei.20–22 In an animal study, lesions in the fastigial nucleus caused nystagmus, in which the slow phase of the nystagmus predominantly occurred in the horizontal plane and beat towards the side of the lesion (ipsilateral drift).22 Contralesional SN in unilateral fastigial nucleus lesions may be explained by hypoactivity in ipsilateral vestibular nuclei due to a loss of normal excitation from the damaged fastigial nucleus (figure 3B). We speculated that the mechanism of spontaneous ipsilesional nystagmus in patients with lesions in the rostral anterior cerebellum may have involved an increased tonic activity of ipsilesional medial vestibular nucleus neurons caused by disinhibition (ie, hyperexcitation) of the ipsilesional fastigial nucleus due to the loss of Purkinje fibres mediated cerebello-fatigial inhibition (figure 3C). According to experimental animal studies, nystagmus following unilateral lesions of the fastigial nucleus disappeared within 1 day23 or completely suppressed by vision.22 This may explain why nystagmus has not been noted as a prominent feature in cases of discrete lesions of the fastigial nucleus in the monkey as well as in lesions of the SCA territory cerebellum in humans.

Figure 3

Schematic explanation of spontaneous ipsilesional nystagmus in an isolated superior cerebellar artery territory cerebellar infarction. (A) The normal circuit with inhibitory projection from the anterior lobe of the cerebellar cortex to the fastigial nucleus and excitatory projections from the fastigial nucleus to the vestibular nucleus. (B) Lesions of the fastigial nucleus resulted in a decrease in tonic resting activity in the ipsilesional vestibular nucleus due to a loss of excitatory efferent from the fastigial nucleus and consequently to the contralateral beating spontaneous nystagmus (SN). (C) Lesions of the anterior lobe resulted in increased tonic resting activity in the ipsilesional vestibular nucleus due to a loss of inhibitory efferents from the anterior lobe and consequently to the ipsilateral beating SN. AL, anterior lobe; FN, fastigial nucleus; VN, vestibular nucleus.

We should also consider the possibility of transient labyrinthine ischaemia, which may cause transient nystagmus and vertigo. Indeed, cardiac emboli are considered as a leading cause of SCA territory cerebellar infarction. In our study, approximately 60% of patients with nystagmus had an obvious cardiac source of emboli as expected. In the case of transient labyrinthine ischaemia, the head impulse test can be normal.

In our patients, perverted downbeating HSN was found in four patients. Although perverted HSN is commonly observed in lesions involving the vestibulocerebellum,24 ,25 which has a strong interconnection with the vestibular nucleus, the anterior cerebellum also has projections to the vestibular nucleus via the fastigial nucleus.18 Thus, lesions involving this area may inappropriately transfer the activity of the horizontal vestibulo-ocular reflex pathway elicited by horizontal headshaking to the vertical one (cross-coupling), resulting in perverted HSN.

The caudal cerebellar vermis and caudal medial cerebellar hemisphere, which are constantly supplied by the PICA, are important sites for the generation of direction changed bidirectional GEN that is usually attributed to the dysfunction of gaze-holding mechanism.26 ,27 Downbeat nystagmus has mostly been ascribed to lesion in the vestibulocerebellum. Our patients had no lesions in these areas on brain MRI. We speculated that the functional disturbance was due to transient ischaemia (without infarction) of other nearby cerebellar structures, including the PICA territory cerebellum which may explain the direction changed bidirectional GEN or downbeat nystagmus. The rapid disappearance of nystagmus, a rich network of vascular anastomoses within the cerebellum,28 an embolic mechanism in SCA cerebellar infarction,29 ,30 and the fact that nystagmus was more likely found in patients who had shorter intervals from onset to examination support our hypothesis, although we did not obtain any data directly supporting this assumption.

Interestingly, patients with SN commonly showed a marked attenuation of their nystagmus by visual fixation, which is generally considered as a common sign of acute peripheral vestibulopathy. Our data suggest that suppression of nystagmus by visual fixation did not guarantee the presence of peripheral vestibulopathy. Importantly, despite the suppression of nystagmus by visual fixation, the intact head impulse test meant that a peripheral vestibular cause was an untenable diagnosis.

Our study has some limitations. First, because structural MRI cannot fully evaluate the functional extent of a lesion, we could not be certain that the cerebellar areas supplied by arteries other than the SCA were functionally intact. Second, the number of patients was relatively small, and in particular, the small number of patients with nystagmus might have lessened the significance of the extracted lesions in these patients. Thus, the possible contribution of the rostral part of the anterior lobe related to nystagmus should be investigated again with a large sample size, especially in the hyperacute period. Finally, because our study focused on nystagmus, further studies, including a detailed analysis of oculomotor dysfunction, are required to assess the pattern of abnormalities in saccade and smooth pursuit in SCA territory cerebellar infarction.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors H-AK conducted the design and conceptualisation of the study, interpretation of the data, and drafting and revising the manuscript. HL wrote the manuscript and analysed and interpreted the data.

  • Funding This research was supported by the Scholar Research Grant of Keimyung University in 2012.

  • Competing interests HL serves on the editorial boards of Research in Vestibular Science, Frontiers in Neuro-otology and Current Medical Imaging Review.

  • Patient consent Obtained.

  • Ethics approval Keimyung University Dongsan Medical Center.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.