OBJECTIVES Clinical signs of acute peripheral vestibulopathy (APV) were repeatedly reported with pontine lesions. The clinical relevance of such a mechanism is not known, as most studies were biased by patients with additional clinical signs of brainstem dysfunction.
METHODS Masseter reflex (MassR), blink reflex (BlinkR), brainstem auditory evoked potentials (BAEPs), and DC electro-oculography (EOG) were tested in 232 consecutive patients with clinical signs of unilateral APV.
RESULTS Forty five of the 232 patients (19.4%) had at least one electrophysiological abnormality suggesting pontine dysfunction mainly due to possible vertebrobasilar ischaemia (22 patients) and multiple sclerosis (eight patients). MassR abnormalities were seen in 24 patients, and EOG abnormalities of saccades and following eye movements occurred in 22 patients. Three patients had BlinkR-R1 abnormalities, and one had delayed BAEP waves IV and V. Clinical improvement was almost always (32 of 34 re-examined patients) associated with improvement or normalisation of at least one electrophysiological abnormality. Brain MRI was done in 25 of the 44 patients and confirmed pontine lesions in six (two infarcts, three inflammations, one tumour).
CONCLUSIONS Pontine dysfunction was suggested in 45 of 232 consecutive patients with clinical signs of APV on the basis of abnormal electrophysiological findings, and was mainly attributed to brainstem ischaemia and multiple sclerosis. The frequency of pontine lesions mimicking APV is underestimated if based on MRI established lesions only.
- acute peripheral vestibulopathy
- vestibular neuritis
- brainstem lesions
- masseter reflex
Statistics from Altmetric.com
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.
- acute peripheral vestibulopathy
- vestibular neuritis
- brainstem lesions
- masseter reflex
Acute peripheral vestibulopathy (APV), also called “vestibular neuritis”, “vestibular neuronitis”, “labyrinthitis”, or “neurolabyrinthitis”, is a clinical syndrome characterised by vertigo and vomiting, falls, and past pointing towards the affected side; horizontal-rotatory, direction fixed spontaneous nystagmus towards the unaffected side; and unilateral canal paresis with normal hearing.1-3 Although the aetiology remains uncertain in most patients, it is usually attributed to viral inflammation or ischaemia of the extra-axial vestibular portion of the eighth nerve, or the labyrinth, or both.1-3 There are, however, several studies indicating pontine dysfunction as a cause of APV. These studies were based on abnormal saccadic or pursuit eye movements,4-6 stapedius, or masseter reflex abnormalities,7 8 and MRI established lesions involving the region of the intrapontine vestibular nerve, the medial and lateral vestibular nuclei, or both.9-11 Some of these reports, however, were biased by the inclusion of patients with additional clinical signs of brainstem dysfunction.4-6 8 9 11 We report our 13 year experience based on 45 out of 232 consecutive patients with APV (without additional clinical signs of brainstem disease), who had electrophysiological abnormalities indicating responsible pontine lesions.
Patients and methods
DC electro-oculography (EOG), masseter reflex (MassR), blink reflex (BlinkR), and brainstem auditory evoked potentials (BAEPs) were examined routinely in 232 unselected patients with clinical signs of unilateral APV seen over a 13 year period. Diagnosis of APV was based on (a) sustained vertigo (at least for 24 hours) with nausea, or vomiting, or both; (b) horizontal-rotatory, direction fixed spontaneous nystagmus enhanced or only visible by Frenzel’s lenses or during EOG recordings with closed eyes; (c) past pointing, or falling, or both to the direction opposite to nystagmus fast phases; (d) clinically unimpaired hearing function on the affected side; (e) unresponsiveness or hyporesponsiveness of one horizontal semicircular canal during bithermal caloric testing, and (f) unimpaired fixation suppression of the vestibular nystagmus during bithermal caloric testing (maximum slow phase velocity with open eyes/maximum slow phase velocity with closed eyes <0.7). Clinical and electrophysiological examinations were done within 7 and 10 days, respectively after onset of vertigo.
Diagnosis of a presumed vascular origin of the APV was based on (a) sudden onset; (b) presence of at least one relevant risk factor for the development of cerebrovascular diseases (diabetes, hypertension, previous strokes or transient ischaemic attacks, atrial fibrillation, heavy smoking, hypercholesterolaemia, signs of general arteriosclerosis); and (c) subsequent improvement or recovery. Multiple sclerosis was diagnosed according to the criteria given by Poser et al 12 and Paty et al.13
Electrophysiological abnormalities in favour of pontine lesions were found in 45 patients. None had additional abnormal clinical findings except for signs of a distal symmetric sensory neuropathy in three diabetic patients, reduced visual acuity in one eye from a previous optic neuritis in a patient with multiple sclerosis, and bilaterally increased deep tendon reflexes known from a previous myelitis in another patient with multiple sclerosis. In all patients, the vertigo improved during the first week. Thirty four patients were re-examined within 10 days to 10 weeks; all had improved from clinical signs of APV (spontaneous nystagmus, past pointing, tendency to fall). Brain MRI was done in 25 and CT in 14 of the 44 patients.
Horizontal eye movements were recorded separately for either eye with surface electrodes placed near the outer and inner canthi of the eyes (bandwidth 0–40 Hz), and documented by a linear ink jet writer or a polygraph (paper speed 10 and 100 mm/s).
Patients were asked to refixate between light spots projected at random intervals on a screen at a distance of 1.2 m. Mean peak velocities were evaluated graphically from the steepest part of the record (100 mm/s) from six 30° saccades from the primary position. (Normal range (50 healthy volunteers): 320 to 640°/s for 30° abduction saccades, 335 to 670°/s for 30° adduction saccades, interocular difference <35°/s.) Hypometric (undershooting) saccades were considered abnormal in patients with a constant dysmetria exceeding at least 20% of the amplitude of the saccade.
An alternating black and white stripe pattern (each stripe covered a visual angle of 7°) moved across the screen at 20, 40, and 60°/s. The gain (velocity of following eye movements/target velocity) was evaluated graphically from 10 successive nystagmic beats.(Normal gain values (50 healthy volunteers; target velocity: 60°/s): >0.7; right/left-difference <0.15.)
Stimulus was a sinusoidally swinging laser spot (0.3 Hz, peak to peak amplitude: 30°). Interruption by small saccades (=saccadic pursuit eye movements) was considered abnormal.
Bithermal caloric testing
This involved irrigation of either ear with 200 ml water within 30 seconds at 44°C and 30°C (intervals 5 minutes). Responses were recorded with closed eyes. Fixation suppression was assessed by eye opening when vestibular nystagmus was maximal. Maximum slow phase velocity of vestibular nystagmus was evaluated graphically from the recording to quantify the responses according to standard formulas. (see and Baloh and Honrubia14) For quantification of vestibular paresis, we subtracted the value and direction of spontaneous nystagmus from the patient’s warm and cold responses. (For example, patient 25 had a 24°/s spontaneous nystagmus to the left, which ceased for some seconds after right sided warm water irrigation, and increased to 69°/s after left sided warm water irrigation (fig1). This was interpreted as a 24°/s response to the right, and a 45°/s to the left.) Findings were considered abnormal for vestibular paresis >25%, and directional preponderance >30%.
MASS R AND BLINK R
Examinations were performed with commercially available systems (MassR: Toennies, Multiliner; BlinkR: Medelec MS 6). The MassR was elicited by a brisk tap on the patient’s jaw with a reflex hammer. The recording was triggered at the moment of the mechanical tap by a signal from a commercially available piezo-electric element mounted in the hammer. The reflex was recorded simultaneously on both sides with surface electrodes placed over the muscle belly 25 mm above the margin of the mandible (recording electrode), and over the zygoma at the lateral edge of the orbit (reference electrode) (bandwidth 20–2000 Hz). Latency refers to the onset of the EMG response, and mean latencies were calculated from 10 successive events. Normal MassR latencies are given in table 1. These latencies and their variances with follow up were repeatedly re-evaluated and confirmed in more than 200 other volunteers (unpublished data).15 16 Age related latencies of a recent study from our laboratory, however, were about 1.0 ms longer than our previously reported values (table1).17 This study was done with a new, self made recording system including a new reflex hammer with another piezo-electric element, which is the most probable expalanation of the longer latencies, as MassR latencies are essentially influenced by the delay of the trigger device.
The BlinkR was elicited by stimulating the supraorbital nerve on either side using rectangular stimuli (duration 0.1 ms; current 25 mA; intervals 10 s). Responses of the orbicularis oculi muscles were recorded simultaneously with surface electrodes inferior to the lower lid halfway between the inner and outer edge of the orbit (recording electrode) and at the lateral edge of the orbit (reference electrode) (bandwidth 20–2000 Hz). Mean latencies of the R1 and R2 components were calculated from five successive events. Age related normal latencies of the R1 component (BlinkR-R1), which were repeatedly re-evaluated and confirmed in more than 100 other volunteers (Hopfet al 15 and unpublished data), are given in table 1.
Criteria of MassR and BlinkR-R1abnormalities18-21 were: (1) unilateral or bilateral delay outside the age related mean + 2.5 SD, (2) unilateral or bilateral loss (for MassR also partial loss—that is, loss of more than four responses out of 10 trials); (3) right/left differences outside the age related mean + 2.5 SD. Criteria for improvement were (1) reappearance of completely or partially abolished responses and (2) shortening of latency by ⩾0.8 ms (MassR) or ⩾1.3 ms (BlinkR-R1).
Rarefaction and condensation clicks of 70 dB above individual hearing threshold levels were delivered monaurally with a repetition rate of 10/s. Responses were recorded with surface electrodes over the mastoid (recording electrode) and the vertex (reference electrode). Mean latencies were determined from 1000 averaged responses. Unilateral or bilateral delay of wave III, IV, or V, or wave III to V interpeak latency outside the age related mean + 2.5 SD were interpreted in favour of a brainstem lesion only if present after stimulation with both rarefaction and condensation clicks.
MRI was carried out with 1.0 (Siemens Magnetom) or 1.5 Tesla (Philips S) superconducting systems before and after intravenous gadolinium (T1 weighted (repetition time: 500–750 ms, echo time: 20–50 ms) and T2 weighted (repetition time: 1800–2080 ms, echo time: 80–100 ms)). Slice thickness was between 3 and 7 mm.
CT was performed with an EMI 1010 or an Siemens Somatom ARP with slice thicknesses between 3 and 7 mm.
Based on the above criteria, an ischaemic cause for the APV was diagnosed in 22 patients, 15 with one and seven with more than one relevant risk factor for the development of cerebrovascular disease (mainly hypertension:12 patients, type II diabetes four patients). An ischaemic origin was evident in another patient (12) undergoing surgery because of hemifacial spasm: coagulation of a small basilar artery branch to the pons was followed by ipsilateral APV. In eight patients with definite multiple sclerosis, clinical signs of an APV were the only indication of the first (four patients), second (three patients), or third (one patient) exacerbation. Another two patients had an undiagnosed inflammatory disorder (CSF 14 and 18 white cells/μl without oligoclonal banding), and one had an ipsilateral pontomedullary tumour. In the remaining nine patients, the aetiology was undetermined.
Abnormal electrophysiological findings are given in detail in table 2. Bithermal caloric testing documented an unresponsiveness in 12 patients, and a hyporesponsiveness in 33 patients (between 26% and 93%) of one horizontal semicircular canal. At the time of bithermal caloric testing, which was done within 1 week after the clinical examination, spontaneous nystagmus was still present in 28 patients. It was unimpaired by warm water irrigation of the affected side in six patients, decreased, or ceased, but was not reversed in 13 patients (fig 1), and was reversed in nine patients (table2).
Unilateral MassR or BlinkR-R1 or BAEP abnormalities were almost always (in 20 of 21 patients) on the side of the APV (figs2 and 3) (table 2). Slowed abduction or adduction saccades on one eye, and slowed lateral saccades or impaired following eye movements were also almost (17 of 22 patients) on the side or directed to the side of the APV (figs 4 and 5) (table 2). Clinical improvement was almost always (32 of 34 re-examined patients) associated with improvement or normalisation of at least one electrophysiological abnormality (figs 2,3, and 5) (table 2). The remaining two patients (8 and 11) had an ipsilateral MassR abnormality, which was normal at re-examination, but shortening of latency was not significant (0.6 and 0.3 ms, table2).
MRI was carried out in 25 of the 45 patients and confirmed ipsilateral pontine lesions in six patients. The lesions mainly involved the region of the intrapontine vestibular nerve in five patients—three with multiple sclerosis (figs 1, 6, and 7), one with ischaemia (fig 8), and one with pontomedullary tumour—and the region of the vestibular nuclei in one patient with ischaemia (fig 4). Another patient (23) with clinical signs of a left sided APV had a right sided space occupying cerebellar infarction with pontine compression, but no pontine or medullary lesion. Another two patients with risk factors for cerebrovascular diseases had hyperintense bilateral pontine lesions, which correspond to “pontine ischaemic rarefaction” shown to be a common finding in such patients.22 23 The bilaterality of these lesions and lack of gadolinium enhancement prevent definite conclusions on their functional relevance. Small vascular lesions of the supratentorial white matter were seen in all patients with risk factors for cerebrovascular diseases, and multiple high signal lesions of the periventricular white matter in all patients with multiple sclerosis. CT was performed in 14 patients and failed to disclose any brainstem lesion.
Forty five of 232 consecutive patients with clinical signs of unilateral APV had abnormal electrophysiological findings, which cannot be explained by lesions of the labyrinth or the extra-axial vestibular nerve. Abnormal saccadic and pursuit eye movements are well known indicators of CNS involvement,24 25 although abnormal pursuit eye movements are of little topodiagnostic value.25 The central MassR loop follows the entire length of the trigeminal mesencephalic tract and nucleus, closely approaching the region of the vestibular nuclei at the level of the trigeminal nerve motor nucleus (fig 9).26-28 The BlinkR-R1 is mediated between the lateral midpons and the facial nucleus via a trigeminofacial connection, which runs likely ventrolateral to the medial longitudinal fasciculus (MLF).21 27 28 Provided that trigeminal nerve functions are intact as in all our patients (a normal corneal reflex, normal trigeminal sensory function, no masseter paresis), MassR abnormalities indicate ipsilateral brainstem lesions between the level of the trigeminal motor nucleus and the oculomotor nucleus, and BlinkR-R1 abnormalities indicate lesions between the lateral midpons and the medial caudal pons.15 26 28-33
In a given setting, latencies of the MassR do not show much variation,15 18 29 and do not seem to be significantly affected by electrode placement.17 Our average MassR and BlinkR-R1 latencies and our criteria of abnormal responses are in accord with previous reports from other laboratories.18-21 29 34 Only Yates and Brown reported longer MassR latencies,30 but they used a different triggering mechanism which may account for the difference, as MassR latencies are essentially influenced by the delay of the trigger device.15 35 Very short MassR latencies (<6.0 ms) were occasionly found in previous studies,15 18 even by Yates and Brown,30 and in some of our patients, and may reflect normal variations. Unlike latencies, MassR amplitudes are much less useful interpretive tools because of much greater variation. Typical values range between 0.5 and 3 mV, but can even be below 0.5 mV and may reach 7 mV. Amplitudes also depend on the degree of masseter muscle preinnervation and the strength of the stimulus.35 36Therefore, we do not consider amplitudes as a reliable indicator of an abnormal MassR in our study. As shown in patients with internuclear ophthalmoplegia, demyelinating and vascular lesions, which were the most common causes of APV in our study, caused similar MassR abnormalities; the average of the delayed responses were only slightly longer in patients with multiple sclerosis, and the average amplitudes were about 10% smaller in patients with lacunar infarctions.15 As long as ischaemic lesions do not cause relevant axonal degeneration, delayed latencies with relative preservation of amplitude are explicable and the most likely explanation of such MassR abnormalities in patients with mesencephalic and pontine infarctions.15 31
The electrophysiological abnormalities of our patients may be attributed to (a) pre-existing brainstem lesions, (b) disseminated brainstem lesions, or (c) single common lesions involving the intrapontine vestibular nerve, the medial or lateral vestibular nucleus, or some combination of these, and the central loop of the MassR or BlinkR, or the neural networks generating saccadic and pursuit eye movements, or some combination of these. Pre-existing lesions are unlikely, as most electrophysiological abnormalities had improved at re-examination, which suggests acute lesions. Moreover, unilateral MassR, BlinkR-R1, or BAEP abnormalities were almost always (19 of 20 patients) on the side of the APV, as were slowed abduction or adduction saccades (seven of eight patients), or slowed horizontal saccades ( six of seven patients). This can hardly be explained by chance and indicates an association with the APV, which was also suggested by the associated improvement of APV and electrophysiological abnormalities seen in more than 90% of re-examined patients. These findings strongly indicate a single common or simultaneously formed disseminated lesions, but disseminated lesions are not required to explain the electrophysiological abnormalities in most of our patients.
Clinical signs of APV with ipsilateral vestibular paresis can occur with brainstem lesions involving the medial vestibular nucleus, lateral vestibular nucleus, and the intrapontine vestibular fascicle.8-11 37 Such lesions are usually associated with additional clinical evidence of brainstem dysfunction such as long tract signs or internuclear ophthalmoplegia,8 9 11 37but may also cause isolated APV.8 Parts of the paramedian pontine reticular formation (PPRF), the intrapontine abducens nerve and nucleus, the MLF, and the central MassR loop are adjacent to the proximal intrapontine vestibular nerve and the lateral and medial vestibular nucleus (figure 9), the last receiving direct afferents from the horizontal semicircular canal.38 This region is mainly supplied by branches of the basilar artery.39-41Ischaemia (or inflammation) within this region may variably impair these structures causing APV by involvement of the intrapontine vestibular nerve and/or the medial and/or lateral vestibular nucleus,8 9 11 37 ipsiversively slowed horizontal saccades by involvement of the abducens nucleus and/or the PPRF,42 43 ipsilaterally slowed abduction or adduction saccades by involvement of the intrapontine abducens nerve or the MLF,44 45 and ipsilateral MassR abnormalities by affection of the central MassR loop.8 15 26 28-30 Such a mechanism was obvious in the patient with APV and ipsilaterally abnormal MassR after coagulation of a small basilar artery branch to the pons, and is the most likely explanation in patients with clinical signs of APV and ipsilateral MassR abnormalities (16 patients), slowed ipsiversive saccades, (seven patients), and ipsilaterally slowed abduction and adduction saccades (four and three patients, respectively) (table 2). Penetrating basilar artery branches often show asymmetric terminations,39-41 which may explain bilateral MassR abnormalities (four patients) by asymmetric bilateral ischaemic lesions.
Smooth pursuit eye movements are generated by a neural network, which includes afferent and efferent cerebellar connections (from the flocculus, paraflocculus, dorsal parts of the vermis), dorsolateral pontine nuclei, vestibular nuclei, nucleus prepositus hypoglossi, ocular motor nuclei, and connections between these structures.25 46 47 Following eye movements during optokinetic nystagmus (slow phases) are mainly generated by the same premotor structures as smooth pursuit eye movements,25 46 47 and the smooth pursuit component mainly determines optokinetic nystagmus in humans.47Pontine lesions may be followed by impaired pursuit eye movements25 48 and are a plausible explanation of saccadic pursuit eye movements and reduced velocities of following eye movements during optokinetic nystagmus as seen in five and two patients, respectively (table 2). Although saccadic pursuit eye movements are often associated with impaired fixation suppression of vestibular nystagmus38 49 such lesions do not necessarily affect fixation suppression. A responsible lesion in the region of the intrapontine vestibular nerve may affect parts of the smooth pursuit system—for example, connections between the dorsolateral pontine nuclei or the nucleus prepositus hypoglossi and the vestibular nuclei. Such lesions may spare efferent cerebellar connections from the flocculus and paraflocculus mediating fixation suppression of vestibular nystagmus, which descend to the vestibular nuclei.38 49
Disseminated lesions have to be considered in some patients. Two patients had contralaterally slowed abduction saccades, which were attributed to lesions of the intrapontine abducens nerve opposite to the side of the APV. A single lesion involving the intrapontine vestibular nerve, the medial vestibular nucleus or both on one side and the intrapontine abducens nerve on the other should also involve the intrapontine abducens nerve on the side of the APV. But signs of an ipsilateral sixth nerve involvement were absent in these patients, which makes a single lesion unlikely. The simultaneous improvement of clinical signs of APV and contralaterally slowed abduction saccades were interpreted in favour of disseminated brainstem lesions, which were reported in up to 25% of patients with vertebrobasilar ischaemia.50 A third patient had a space occupying cerebellar infarction opposite to the side of the APV. Clinical signs of acute APV including ipsilateral vestibular paresis can occur with ipsilateral cerebellar infarctions,51 52 and are attributed to labyrinthine ischaemia by circulatory arrest in the territory of the internal auditory artery, which is a branch of the anterior inferior cerebellar artery (AICA), or ischaemia in the territory of recurrent penetrating arteries off AICA, which supply the region of the intrapontine vestibular nerve.51 53 54 But APV occurred contralateral to the cerebellar infarction in our patient, indicating bilateral lesions, which were concluded from bilaterally delayed MassR latencies more pronounced on the side of the APV. This may reflect brainstem compression by the space occupying cerebellar infarction, or disseminated lesions, which were reported in up to 25% of patients with vertebrobasilar ischaemia.50 Another patient with right sided vestibular paresis and ipsilaterally slowed adduction saccades interpreted as internuclear ophthalmoplegia had MRI documented hyperintense lesion of the pons involving the region of the ipsilateral intrapontine vestibular nerve, but not the medial longitudinal fasciculus (fig 7). But a second demyelinating lesion or an extension of the documented lesion into the region of the medial longitudinal fasciculus were probably too small to be detected by MRI, as MRI may fail to disclose lesions of the medial longitudinal fasciculus in some patients with internuclear ophthalmoplegia.55-57 The simultaneous improvement of both clinical signs of APV and ipsilaterally slowed adduction saccades, indicated that both were caused by the same actual or simultaneously formed disseminated lesions.
Our findings add further evidence on the existence of acute APV due to pontine lesions, which were mainly suggested by abnormal brainstem reflexes and abnormal saccadic eye movements in 45 of 232 consecutive patients (19.4%). Vestibular paresis may be caused by pontine lesions affecting the intrapontine vestibular nerve fascicle or the medial and lateral vestibular nucleus.9-11 But normal fixation suppression of caloric nystagmus, which was seen in all our patients, indicated fascicular vestibular nerve lesions, as nuclear lesions usually show little or no inhibitory effect of visual fixation on caloric nystagmus.9 58 Previous studies on smaller groups of patients with APV reported higher frequencies of responsible pontine lesions (between 31.5% and 59.4%) based on abnormal saccadic, or pursuit eye movements, or prolonged BAEP latencies.4-6But these studies were biased by up to 20% of patients with additional clinical signs of brainstem dysfunction, which does not match the diagnosis of APV, thereby overestimating patients with central lesions. But even an incidence of 19.4% as in our study shows that pontine lesions are a common cause of APV, which is underestimated, if based on MRI established lesions only; MRI was performed in 25 of our patients with electrophysiological abnormalities and confirmed corresponding pontine lesions in six patients only, indicating that some brainstem lesions escape proof by MRI. This was also shown in a large series of patients with third and sixth nerve palsies due to brainstem lesions,31 33 59 and in up to 55.6% of patients with multiple sclerosis with BAEP and BlinkR abnormalities, or internuclear ophthalmoplegia with impaired convergence as an indicator of midbrain lesions, who showed no corresponding MRI established brainstem lesions.55 60-62
We thank Mrs Annette Möller for her excellent technical assistance, and are indebted to Professor P Stoeter (Department of Neuroradiology, University of Mainz) and to Dr A v Essen (Neuroradiologist, Koblenz) for their kind permission to publish MRI findings and to reproduce figs 1 and 6-8 (Professor P Stoeter), and fig 4 (Dr A v Essen).