593

Ann. N.Y. Acad. Sci. 1039: 593596 (2005). 2005 New York Academy of Sciences.doi: 10.1196/annals.1325.075

Studies of Eccentric Gaze StabilityEffects of Pitch Head Position on HorizontalGaze-Holding in Patients with Cerebellar DiseaseELIZABETH WILSON,a KAREN SNG,a JEFFREY T. SOMERS,b,cMILLARD F. RESCHKE,b AND R. JOHN LEIGHaaDepartment of Biomedical Engineering and Neurology Service,Veterans Affairs Medical Center, Case Western Reserve University,Cleveland, Ohio 44106, USAbNeurosciences Laboratories, Johnson Space Center, NASA, Houston,Texas 77058, USAcWyle Laboratories, Houston, Texas 77058, USA

ABSTRACT: We studied the effects of change of head position in the pitch planeon the ability to hold horizontal eccentric gaze in five patients with cerebellar dis-orders. All patient showed a change in the time constant of horizontal centripetaldrift when the head was pitched forward or back. This result suggests thatotolithic inputs can influence the neural integrator for horizontal eye movementsand indicates the value of testing horizontal gaze holding in different headpositions.

KEYWORDS: nystagmus; neural integrator; otoliths; cerebellum

A routine part of the cerebellar examination includes testing for horizontal nystag-mus while the patient attempts to sustain an eccentric horizontal gaze position. Thebrain-stem mechanism (ocular-motor neural integrator) that enables the eyes to beheld at an eccentric horizontal position is partly governed by the cerebellum.1 A lab-oratory measure of eccentric gaze-holding function is the time constant of centripe-tal eye drift in darkness (c). In normals, c ranges from 20 to 70 s,14 but isdecreased in cerebellar patients, who often cannot sustain their eyes in an eccentrichorizontal position, even when a fixation target is visible, and show gaze-evokednystagmus.1 We investigated whether repositioning the head in the pitch plane (backor forward) affected horizontal gaze-holding ability in patients with cerebellardisorders.

We studied five patients (two female, age range 31 to 54 years); their diagnoseswere Chiari malformation (2), spinocerebellar ataxia type 8, episodic ataxia type 2,and postresection of meningioma of the fourth ventricle. All had clinical and radio-

Address for correspondence: R. John Leigh, M.D., Neurology Service (127W), VeteransAffairs Medical Center, 10701 East Boulevard, Cleveland, OH 44106-1702. Voice: 216-844-3190; fax: 216-231-3461.

rjl4@case.edu

594 ANNALS NEW YORK ACADEMY OF SCIENCES

logical evidence of involvement of the vestibular cerebellum. Findings were com-pared with data from 33 age-matched normals.2 All patients gave informed consentin accordance with the Institutional Review Board of the Cleveland Veterans AffairsMedical Center. We measured horizontal and vertical gaze and head position usingthe magnetic search-coil technique.1 Gaze stability was measured during three headorientations: erect, pitched back (chin up) 20, and pitched forward (chin down) 20.Each trial began as subjects attempted to fix upon a visual target (laser spot at 1.2 m)centered in the horizontal plane. During trials when patients heads were pitchedback or forward, the laser spot was displaced 20 up or down, respectively, so that itcorresponded closely to the central orbital position at the beginning of each trial. Af-ter 5 s, the target jumped to either right or left 25, and then flashed (10-ms duration)every 2 s for 20 s. The visual target then returned to center position and continued toflash for a further 15 s; it was illuminated continuously at the end of the trial. Hori-zontal gaze holding was tested to the right and left for each of the three head posi-tions (total of six trials). In two patients we ran trials with the head erect twice (firstand last) to compare results before and after change of head position. Using an in-teractive method, we measured the time constant of centripetal eye drift (c) fromlinear regression of plots of the natural logarithm transformation of eye position andtime. We present values of 1/c rather than c because the former shows a more con-tinuous distribution, especially for the cases of a nearly perfect neural integrator thatcould be either just barely leaky, or just barely unstable (corresponding to values of1/c close to zero). We calculated 1/c for the first 2 s and last 2 s of gaze holding(the latter providing an index of adaptation and rebound phenomena). We madepaired comparisons of the values 1/c for each gaze direction (right or left) for eachof the three head positions for each patient using a paired t-test. We also looked forany effects of horizontal gaze holding on vertical eye position.

Each patient showed a change of horizontal gaze holding (manifest by a changein 1/c) in at least one direction (right or left) when values with the head erect werecompared with the pitched back or pitched forward positions. An example is shownin FIGURE 1A and 1B. As a group, there was a significant effect of pitching the headback on the initial values of 1/c, (P < .05), but not of pitching the head forward, norfor final values (FIG. 1C). Paired comparisons of the values of 1/c during initial andfinal 2-s epochs were not significantly different for the group as a whole, reflectingidiosyncratic changes, with some patients showing decreased or increased drift at theend of the eccentric gaze-holding period. Although nystagmus was encountered ineach patient after the eyes returned to center, it was not classically rebound, beingpredominantly unilateral (e.g., right-beating nystagmus following rightward or left-ward gaze holding). Comparison of gaze holding during the two trials with headerect (one before pitching the head forward and back, the second afterwards) in twopatients showed no significant differences. Change of head position in the pitchplane also variably affected vertical nystagmus; for example, one Chiari patientshowed upbeat nystagmus that increased in lateral gaze except in left gaze, duringflexing the head forward, when downbeat nystagmus occurred.

In summary, we found that all of our cerebellar patients showed a change in theirhorizontal gaze-holding ability when their head was pitched forward or backward(especially the latter). These findings provide evidence that otolithic inputs can in-fluence the neural integrator for horizontal eye movements, although neck proprio-ception may have contributed. Changes in head position also affected vertical

595WILSON et al.: STUDIES OF ECCENTRIC GAZE STABILITY

FIGURE 1. See following page for legend.

596 ANNALS NEW YORK ACADEMY OF SCIENCES

nystagmus during eccentric horizontal gaze in each patient. Our findings suggestthat examination of eye movements during positional testing should include lateralgaze positions, during which gaze-evoked nystagmus, with vertical components,may become clinically manifest.

ACKNOWLEDGMENTS

This work was supported by NASA/NSBRI NA00208, NIH grant EY06717, theOffice of Research and Development, Medical Research Service, Department ofVeterans Affairs, and the Evenor Armington Fund (to R.J.L.). We are grateful toYanning Han, Arun Kumar, and Victor Lin for technical assistance, and to Drs. HenryKaminski, Robert Tomsak, and Bernd Remler for referring patients.

REFERENCES

1. LEIGH, R.J. & D.S. ZEE. 1999 The neurology of eye movements, 3rd ed. Oxford Univer-sity Press. New York.

2. RESCHKE, M. et al. 2004. Studies of the ability to hold the eye in eccentric gaze: mea-surements in normal subjects with the head erect. Vision Res. Submitted.

3. ROBINSON, D.A. et al. 1984. Alexanders law: its behavior and origin in the human ves-tibulo-ocular reflex. Ann. Neurol. 16: 714722.

4. EIZENMAN, M., P. CHENG, J.A. SHARPE & R.C. FRECKER. 1990. End-point nystagmusand ocular drift: an experimental and theoretical study. Vision Res. 30: 863877.

FIGURE 1. (A) Example of effects of change of head position in the pitch plane onhorizontal gaze holding in a patient with spinocerebellar ataxia type 8. Upward deflectionsindicate rightward or upward eye movements. In the top panel, with head erect, transientgaze-evoked nystagmus is followed by small saccadic intrusions. When the eye returns tocenter, there is rebound nystagmus (left-beating). (B) With head pitched back, horizontalgaze-evoked nystagmus persists throughout the period of eccentric gaze holding, with nowell-developed rebound nystagmus. Low-amplitude downbeat nystagmus is evidentthroughout. (C) Paired comparisons of initial and final 1/c values for head erect versus headpitched back (circles), and head erect versus head pitched forward (squares). A 1/c valueof 0.0 corresponds to a perfect integrator; positive values indicate a leaky integrator, andnegative values indicate an unstable integrator. If change of head position had no effect on1/c compared with values with the head erect, data points would lie on the diagonal. Foreach patient, values during right and left eccentric gaze holding are shown. As a group, pitch-ing the head back produced significantly greater initial values of 1/c (leakier integrator)which mainly lie above the diagonal. The area enclosed by the dotted lines at the bottom leftcorner corresponds to 90th percentiles for initial 1/c values for 33 normal subjects with theirheads erect.2