Acta Otolaryngol 2001; Suppl 545: 170–173

Otolith Function after Acoustic Neuroma Surgery

FLORIS L. WUYTS1, ANJA VAN DER STAPPEN1, MIEKE HOPPENBROUWERS1, DIRK VANDYCK2, ROBERT H. SCHOR3, JOSEPH M. FURMAN3 and PAUL H. VAN DE HEYNING1

From the 1ENT Department, Uni×ersity Hospital of Antwerp, Uni×ersity of Antwerp, 2Vision Lab, Uni×ersity of Antwerp, Antwerp,Belgium and 3Department of Otolaryngology, Uni×ersity of Pittsburgh, Pittsburgh, Pennsyl×ania, USA

Wuyts FL, Van der Stappen A, Hoppenbrouwers M, Van Dyck D, Schor RH, Furman JM, Van de Heyning PH. Otolithfunction after acoustic neuroma surgery. Acta Otolaryngol 2001; Suppl 545: 170–173.

The response of the vestibular system after acoustic neuroma surgery was investigated in nine patients. The otolith systemwas studied by means of ocular counterrolling, assessed by video oculography. Horizontal vestibulo-ocular re� ex (VOR)function was tested by the sinusoidal harmonic acceleration test using electronystagmography. The results were comparedwith those obtained from a normal control population. The response to slow rotation tests was symmetric, but the gainwas signi� cantly reduced when compared to the normal population. Phase lag was signi� cantly increased. No differencein ocular torsion was observed with latero� exion of the head to the ipsilateral side in comparison with latero� exion to thecontralateral side. Moreover, the overall ocular counterrolling was well within normal limits. We conclude that thesemicircular canal response differs from the otolith response. The component of the torsional VOR mediated by otolithstimulation appears to be more robust than the horizontal VOR mediated mainly by the horizontal semicircular canalsystem. Ocular counterrolling induced by latero� exion does not reveal abnormalities in patients with surgically producedunilateral peripheral loss. Key words: ocular counterrolling, otolith, ×estibulo-ocular re� ex, ×ideo oculography.

INTRODUCTION

Disequilibrium and vertigo contribute substantiallyto the impairment and handicap of patients followingsurgery for an acoustic neuroma. This justi� es inves-tigation of the vestibular system with electronystag-mography, with the main purpose of assessing thecentral compensation process of the horizontalvestibulo-ocular re� ex (VOR). Usually, this isachieved with low-frequency rotary stimulation andvelocity step tests.

Only a few studies to date have reported on theutricular response in patients with unilateral vestibu-lar deafferentiation (1, 2). Clarke and Engelhorn (1)showed that with the latero� exion test, which is thesimplest test for otolith response measurements, noasymmetry was revealed. Using a speci� c eccentricrotation paradigm, however, with the axis of rotationpositioned through either the left or right labyrinth, aprominent asymmetric response was demonstrated (1,3).

This study correlates the � ndings from classicalvestibular rotation tests with those obtained from thelatero� exion test in order to investigate a possiblerelationship between the horizontal VOR and theotolith response.

MATERIALS AND METHODS

A group of nine patients (� ve male, four female)treated for acoustic neuroma at the ENT Departmentof the Antwerp University Hospital were investigatedduring the follow-up period after the surgery. Onaverage, the vestibular investigation was performed 6

months after surgery. The average age of the patientswas 45 years (range 22–72 years). Vestibular re-sponses were investigated using classical electronys-tagmography (ENG) and the more recentlydeveloped technique of video-oculography. The ENGtest battery consisted of the recording of spontaneousnystagmus, followed by tests for gaze-evoked nystag-mus, saccades, optokinetic nystagmus and smoothpursuit. Subsequently, the horizontal rotational VORwas assessed using a rotation test with a Servomedchair, which turned sinusoidally with a maximumvelocity of 50°:s and with a frequency of 0.05 Hz.This VOR test was performed in total darkness, whilethe patient performed a task to ensure alertness. Thehead velocity was measured with an angular ratesensor (ARS C152-1A; Watson Ind. Inc.) that wasplaced on the patient’s head by means of a Velcrostrap. The ratio of the eye velocity (response) to thehead velocity (stimulus) de� ned the gain, whereas thetime delay between the response and stimulus de� nedthe phase. Directional preponderance was calculatedas the difference between left and right slow compo-nent velocities divided by the sum of both slowcomponent velocities multiplied by 100.

ENG was recorded and analyzed with an eight-channel PC-based system (Nystagliner Toennies). Themovements of each eye were recorded separately byplacing electrodes to the left and right of each eye.Two electrodes were placed above and below one eyeto monitor vertical eye movements as well as arti-facts. A calibration procedure was performed prior tospontaneous nystagmus detection, the saccade testand the rotation test.

© 2001 Taylor & Francis. ISSN 0001-6489

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y C

alif

orni

a St

ate

Uni

vers

ity N

orth

ridg

e on

10/

29/1

4Fo

r pe

rson

al u

se o

nly.

Otolith function after acoustic neuroma surgery 171

The other method used for eye movement record-ing was 3D video-oculography (VOG), developed bythe Universities of Maastricht, Rotterdam andAntwerp, based on the prototype by Kingma andco-workers (4, 5). The main component of the VOGsystem is a small infrared video camera placed on aspeci� c mount so that a frontal view of the eye isobtained (Figs. 1 and 2). The design of the VOGsystem permits a full visual � eld for the patient, whilethe camera-to-eye distance is large (\ 20 cm), whichreduces image deformation and improves imagesharpness upon lateral gaze.

The horizontal, vertical and torsional movementsof the eye are deduced by means of image processing.The horizontal and vertical coordinates of the eye arebased upon the � tting of a circle to the pupil. Thecalculation of ocular torsion is based on the identi� -cation of a speci� c detail in the iris structure, e.g. a

Fig. 3. Horizontal, vertical and torsional eye movements (°)when the head of a patient with right-sided acoustic neu-roma was rolled from the right-shoulder position to themid-position to the left-shoulder position. The lower traceindicates the amount of latero� exion (°). Note the dynamiccomponent (vertical canal system) during the roll, and thestatic component (utricular function) when the position ismaintained.

Fig. 1. Plan view of the 3D video goggles.

white or black spot. A small arc-shaped window (960–90°) is placed such that it overlays the spot. Theposition of this segment is � xed with reference to thecenter of the pupil. When the eye makes a rotationabout the roll axis, ocular torsion is calculated bymeans of cross- correlation of the gray scale informa-tion contained in the segment. This ability to measureocular torsion is the main advantage of 3D VOGover conventional ENG. It enables the assessment ofthe otolith system by means of otolith-mediated ocu-lar counterrolling.

The VOG test battery consists of a calibrationprocedure, followed by a latero� exion test. Duringthe test, which is performed in total darkness in aseparate room, the patient is asked, through an inter-com, to bend the head to the right shoulder and tokeep it in that position for : 15 s. Subsequently thepatient moves his:her head upright again, followed bybending the head to the left shoulder and � nally tothe upright position again. This sequence is thenrepeated but in the reverse order, i.e. the head is bent� rst to the left shoulder. The average ocular counter-rolling during left and right latero� exion is measured.A tilt-measuring device (ADS100A Gravitational In-clinometer; Watson Ind. Inc.) is placed on the pa-tient’s head to measure the amount of latero� exion.Fig. 3 shows the typical traces obtained during thisprocedure. All measurements, except the VOG cali-bration, were performed in total darkness, to avoidthe in� uence of visual input. Additionally, pilot ex-periments showed that in light conditions the pupildiameter can � uctuate greatly, which considerably

Fig. 2. Frontal view of the 3D video goggles. Behind thepatient, the image is projected as seen through one of thecameras.

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y C

alif

orni

a St

ate

Uni

vers

ity N

orth

ridg

e on

10/

29/1

4Fo

r pe

rson

al u

se o

nly.

F. L. Wuyts et al.172

hampers the reliability of the ocular torsion measure-ment, because the pupil can approach or even coverthe arc-shaped window used for the cross-correlation.Also, in darkness, full mytriasis makes the diameterof the pupil relatively stable, so that measurement ofthe torsion is stable.

RESULTS

Fig. 4 shows the average ocular torsion resultingfrom latero� exion to the ipsilateral and contralateralsides of the operated lesion in the patient group.There was no statistically signi� cant difference ob-served between latero� exion to the ipsilateral andcontralateral sides. Therefore, the ipsi- and contralat-eral data were pooled for comparison with the nor-mal data, obtained from a group of 39 healthy

subjects who underwent exactly the same procedurein the ENT Department of the Antwerp UniversityHospital. From this comparison we conclude that theaverage ocular torsion upon latero� exion in theacoustic neuroma group did not differ from that inthe control group.

The average directional preponderance from therotation test for the horizontal VOR assessment inthe acoustic neuroma group was 17% (SD ¾9.4%).This is clearly below the limit for abnormality, whichwas determined in our control population as 22%(95% prediction interval, i.e. mean92SD) (6, 7). Fig.5 shows the gain and phase for both acoustic neu-roma patients and healthy subjects. Student’s t-testindicated that the gain was signi� cantly reduced (pB0.01) and the phase signi� cantly increased (pB0.001)for the acoustic neuroma patients compared to thehealthy group.

DISCUSSION

The assessment of vestibular function in patients whosuffer from vertigo is of major importance during thediagnostic process, but also during follow-up andrecovery of the patient. In most cases it enables theclinician to localize the side of the lesion, or todetermine whether the vertigo originates peripherallyor centrally. In the speci� c etiology of acoustic neu-roma, where the diagnosis is primarily made bymeans of imaging techniques (MRI, CT), the vestibu-lar examination can assess whether the tumor impairsthe vestibular function totally, partially or not at all.If the tumor has no effect on the vestibular functionthen it merely interferes with the cochlear part of thevestibulo-cochlear nerve bundle. This knowledgehelps the surgeon to predict the postoperative state ofthe patient in terms of vertigo. When the acousticneuroma has already destroyed most of the vestibularfunction on the lesioned side, the brain will have hadthe time to adapt to this slowly occurring damage.Therefore, in most of these cases compensation hastaken place quite satisfactorily. These patients willnot suffer much from dizziness prior to surgery, andit will be primarily their tinnitus or asymmetric hear-ing loss that will lead them to the ENT doctor.However, the classical vestibular function test con-sists mostly of the rotation test and the caloric testwhere each of the horizontal semicircular canal sys-tems is assessed separately. The results from thesetests are usually generalized to the entire peripheralvestibular system, but in some cases this may be toomuch of a simpli� cation. Indeed, it is not alwaysguaranteed that the tumor will affect the entirevestibular part of the vestibulo-cochlear nerve bun-dle. That is why in some cases where caloric are� exia

Fig. 4. Left: average ocular counterroll during latero� exiontowards the lesioned side and towards the contralateralside. Right: average ocular counterroll in the acoustic neu-roma group (Ocr AN), compared to that in the controlgroup (Ocr Norm).

Fig. 5. Rotational chair data at 0.05 Hz. Left: average gainof the acoustic neuroma group (AN) compared with that ofthe control group (Norm). The difference is signi� cant(p¾0.006). Right: average phase of the acoustic neuromagroup compared with that of the control group. The differ-ence is signi� cant (pB0.001).

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y C

alif

orni

a St

ate

Uni

vers

ity N

orth

ridg

e on

10/

29/1

4Fo

r pe

rson

al u

se o

nly.

Otolith function after acoustic neuroma surgery 173

is found these patients still complain of severe vertigoafter tumor resection.

When the acoustic neuroma does not hampervestibular function at all, the abrupt removal ofvestibular function as a result of tumor resection willinduce a period of signi� cant vertigo. This may beanticipated by means of an intensive rehabilitationprogram starting prior to surgery, so that post-surgerythe patient is able to cope better with the acute vertigo.Vestibular function assessment performed after acous-tic neuroma surgery provides information about thevestibular recovery.

This study reveals that the recovery or compensationis not identical for the semicircular canal and otolithsystems. At the low-frequency stimulus of 0.05 Hz, noasymmetry between leftward and rightward rotationswas revealed, but the gain and phase of the horizontalVOR were altered in comparison to the control popu-lation. Also, the utricular function showed no left–right asymmetry and did not show an overall decreasein amplitude when compared with values of oculartorsion obtained in a control population. The de-creased horizontal VOR gain indicates that the overallperformance of the vestibular function did not fullyrecover to a normal level after an average period of 6months. However, the compensation process was effec-tive in so far as the asymmetry reached normal levelsat low stimulation frequency. Nevertheless, the phasewas signi� cantly increased. The aforementioned differ-ence in recovery between the horizontal semicircularcanal and otolith systems suggests that ocular counter-rolling, and possibly the torsional VOR, is more robustthan the horizontal VOR. In other words, the detectionof gravity is much more ef� ciently preserved underconditions of partial failure of the vestibular systemthan the gaze stabilization system. Obviously, contin-uous detection of gravity is a necessary condition inorder to maintain stance and prevent falls. This can beexpected since the anatomical design of each separateutricle makes it a sensor for accelerations in multipledirections. Therefore, there is considerable redundancyin the peripheral linear acceleration transducers.

We are aware of the fact that during latero� exionthe gaze of the subjects can drift to either the left orright, as the latero� exion test is performed in totaldarkness. Such a lateral gaze might induce an addi-tional ocular counterrolling due to the image analysisprocessing and:or by the ocular motor system itself.However, as the same paradigm was used for both thesubjects and controls, we did not take this effect intoaccount.

The latero� exion paradigm tests the ocular counter-rolling as a reaction to tilt by both labyrinths. In casesof surgically con� rmed unilateral failure, a normalocular torsion was observed. Thus, it is highly ques-

tionable whether latero� exion will reveal any clinicallysigni� cant information in cases where both vestibularsystems are still working but to an unknown extent.Therefore, it is much more appropriate to apply theunilateral otolith function tests. For this purpose,new-generation rotary chair systems have beenequipped (1, 3) with a moving sled so that, duringrotation, subjects can be translated 3.9 cm along theinter-aural axis, corresponding to the distance betweenthe center of the head and the position of eachlabyrinth. Consequently, each utricle is placed throughthe axis of rotation and the contralateral otolith systemis then centrifuged. When the chair rotates at 400°:s anacceleration of 0.39 g is produced at the level of theutricle. Adding this to gravity, the utricle perceives atilt in the gravito-inertial acceleration vector of : 21°.This generates an ocular torsion of 2–3° amplitude,which can be measured by means of the recentavailability of 3D VOG systems. This unilateral otolithtest appears to be promising for the evaluation of theutricular function when ocular counterrolling is in-duced by latero� exion.

REFERENCES

1. Clarke AH, Engelhorn A. Unilateral testing of utricularfunction. Exp Brain Res 1998; 121: 457–64.

2. Lempert T, Gianna C, Brookes G, Bronstein A, GrestyM. Horizontal otolith-ocular responses in humans afterunilateral vestibular deafferentation.Exp Brain Res 1998;118: 533–40.

3. Wuyts FL, Van der Stappen A, Van de Heyning P.Unilateral otolith function testing. XXXIV Congres Soci-ete Internationald’otoneurologie,Marseille, France, Sep-tember 2000.

4. Kingma H, Gullikers H, de Jong I, Jongen R, DolmansM, Stegeman P. Real time binocular detection of horizon-tal vertical and torsional eye movements by an infra redvideo-eye tracker. Acta Otolaryngol Suppl (Stockh) 1995;520: 9–15.

5. Goddijn OJM, van Dun K. Trehalose metabolism inplants. Trends Plant Sci. 1999; 4: 315–9.

6. Van der Stappen A, Wuyts FL, Van de Heyning P.In� uence of head position on the vestibulo-ocular re� exduring rotational testing. Acta Otolaryngol (Stockh)1999; 119: 892–4.

7. Van der Stappen A, Wuyts FL, Van de Heyning PH.Computerized electronystagmography: normative datarevisited. Acta Otolaryngol 2000; 120: 724–30.

Address for correspondence:Floris L. WuytsENT DepartmentUniversity Hospital of AntwerpWilrijkstraat 10BE-2650 EdegemBelgiumTel.: »32 3 821 47 10Fax: »32 3 825 05 36E-mail: � oris.wuyts@ua.ac.be

Act

a O

tola

ryng

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y C

alif

orni

a St

ate

Uni

vers

ity N

orth

ridg

e on

10/

29/1

4Fo

r pe

rson

al u

se o

nly.