Abstract
Article abstract-We measured torsional eye movements induced by sinusoidal rotation or static tilt of the head in roll while viewing a far or near target in 4 patients with skew deviation due to brainstem lesions, 4 patients with spasmodic torticollis (ST), 2 patients with unilateral eighth nerve section (VIIIS), and 10 normal subjects. Torsional nystagmus was present in all 4 patients with skew deviation. In subjects and patients, responses to both sinusoidal and static roll were larger while viewing the far target, consistent with factors dictated by geometry. Response gains to sinusoidal roll were abnormal in 3 patients with skew (increased in one, decreased in two), abnormal in 3 with ST (increased in 1, decreased in 2), and in abnormal both VIIIS patients (decreased). Greater abnormalities were evident in 3 skew patients while rolling away from the side of their brainstem lesions and in both VIIIS patients while rolling toward their lesioned ears. There were similar but less pronounced changes during static head roll. We conclude that patients with skew, ST, and VIIIS may all have abnormal ocular counter-rolling that is more evident during dynamic testing while viewing a far target. Such abnormalities endure because of the limited influence exerted by vision on torsional eye movements.
NEUROLOGY 1997;48: 506-514
The short latency and sensitivity of the vestibular responses derive, in part, from the resting discharge of the vestibular nerves and the "push-pull" organization of the two sides of the vestibular system. [1] This organization, however, also predisposes to developing an imbalance if the inputs from one labyrinth, or its central connections, are affected by disease. Traditionally, clinicians have looked for vertigo, nystagmus, past-pointing, and a tendency to fall to one side as the manifestations of imbalance of inputs from the labyrinthine semicircular canals. [2] Yet, disease may also involve the otolithic organs, resulting in various degrees of the ocular tilt reaction-skew deviation, ocular cyclorotation, head tilt, and deviation of the subjective visual vertical (SVV). [3] In such cases, however, the relative contributions of the otolithic versus vertical canal imbalance to the symptoms is unclear. Recent studies have expanded our understanding of skew deviation, tilts of the head, and SVV, [3] but few attempts have been made to measure torsional eye rotations induced by head rotation in the frontal plane in such patients.* Several investigators used the magnetic search coil technique to define the properties of the torsional vestibulo-ocular reflex (VOR) in roll in normal subjects. [4-10] We attempted to apply this technique to three clinical syndromes, all of which can affect head position in roll: patients with skew deviation due to brainstem lesions, patients with spasmodic torticollis, and patients with long-standing, unilateral, eighth nerve section. Preliminary results were published as an abstract. [11]
*Such ear-to-shoulder rotations are referred to as head rotations in roll, and the torsional eye rotations that they induce are referred to as ocular counter-rolling. Note that we define torsional rotations from the point of view of the subject so that, for example, clockwise means extorsion of the right eye and intorsion of the left eye, with the upper poles of both eyes rotating to the subject's right. Similarly, we define head roll as clockwise when the right ear is lower.
Methods.
Patients and control subjects.
Patients with skew deviation.
Clinical details are summarized in Table 1. Static cyclotorsion was measured from the optic discfovea angle on fundus photographs. [12]
Table 1. Summary of clinical findings in patients with skew deviation
Patient 1 was a 72-year-old man with metastatic prostate cancer, whom we studied about 2 weeks after he presented complaining of progressive unsteadiness of gait. General examination was notable for some mental slowing and gait ataxia. He had a full right ocular tilt reaction (OTR) with left hypertropia and right head tilt. In addition, he had slow vertical saccades and difficulty sustaining upward or downward gaze. Vertical smooth pursuit and the visually enhanced VOR (VVOR) were judged to be normal for age, as were all horizontal movements. There was no internuclear ophthalmoplegia (INO). MRI demonstrated changes consistent with tumor in the left upper brainstem and thalamus. The patient gradually deteriorated and died within 2 months; autopsy was refused. The clinical diagnosis was of metastases to the brainstem, but no primary tumor other than prostate was identified.
Patient 2 was a 42-year-old man who had suffered a subarachnoid hemorrhage and undergone surgical treatment of a basilar aneurysm 2 months before recording. MRI demonstrated infarction of the left midbrain, rostral to the oculomotor nucleus. His main disability was unsteadiness and double vision. He had a left hypertropia of 8 degrees that did not change with head tilt. There was little spontaneous head tilt or cyclorotation of the ocular axis. He had nearly a full range of ocular movement, but vertical saccades were slow. Spontaneous torsional nystagmus was evident, with quick phases beating clockwise.
Patient 3 was a 73-year-old woman whom we recorded 9 days after sudden onset of acute vertical diplopia. She spontaneously complained that her perception was tilted clockwise. She had a full right ocular tilt reaction with left hypertropia and right head tilt. There was a left INO and spontaneous nystagmus beating counter clockwise and upward. MRI failed to show any discrete cerebral lesion, although magnetic resonance angiography demonstrated bilateral carotid artery stenosis. The clinical diagnosis was brainstem infarction involving the left medial longitudinal fasciculus.
Patient 4 was a 46-year-old man recorded 2 weeks after he noted the onset of vertical diplopia and intermittent torsional oscillopsia. He had left hypertropia, modest clockwise deviations of the ocular axes, but no head tilt. There was a left INO and spontaneous nystagmus with quick phases beating clockwise and upward. Further testing revealed evidence of left optic neuropathy with mild impairment of color vision and a relative afferent pupillary defect. MRI was consistent with multiple sclerosis (MS).
Patients with spasmodic torticollis (ST).
Patients 5 to 8 (two women, two men) all had rightward head tilt; they were tested after treatment injection with botulinum A toxin, at which time their torticollis was minimized. The ages of the women were 40 years (with 3 years of torticollis) and 69 years (with 4 years' duration of torticollis). The men were aged 50 (with a 17 year duration) and 38 years (with a 7 year duration of torticollis). Patient 5 and patient 6 had complained of dizziness in the past, but rotational and caloric testing was normal. The other patients had no vestibular or other complaints except those referable to their torticollis. Neurologic examination in all ST patients was unremarkable, except for their dystonia.
Patients with unilateral eighth nerve section (VIIIS).
Patient 9 was a 46-year-old man who had undergone removal of a schwannoma of the left eighth nerve 3 years previously. He lost all left eighth nerve function, but left facial movements were almost normal. Patient 10 was a 62-year-old man who had undergone removal of a schwannoma of the right eighth nerve 17 years previously; after the surgery, he lost all right eighth and seventh nerve function. Subsequently, surgical transposition of his right hypoglossal nerve had been performed to reinnervate his right face. We were unable to detect evidence of brainstem or cerebellar dysfunction in either patient.
Ten control subjects (NS).
Six men and four women (age range, 24-48 years) were also studied. None had any vestibular symptoms or were taking medications. All subjects and patients gave informed consent approved by our Institutional Review Board.
Eye movement recording methods.
We measured torsional, vertical, and horizontal rotations of the head and gaze using the magnetic search coil technique, with 6-foot field coils (CNC Engineering, Seattle, WA). Eye and head coils were precalibrated on a protractor device. The system was 98.5% linear over an operating range of +/- 30 degrees in all three planes. Cross-talk between channels was always <4% and was <2% onto torsional signals from horizontal or vertical channels. SD of the system noise was <0.02 degrees. The translation artifact within the central 30-cm cube of the magnetic field, in which subjects' heads always remained, was <0.03 degrees/cm. Subjects wore a scleral search coil (Skalar, Delft, The Netherlands) on both eyes and on their foreheads to measure angular head position. The testing was performed with subjects and patients in a seated position, beginning with the head upright. No subject or patient wore spectacle corrections during testing; all were able to see the visual stimuli. Our strategy was to minimize horizontal and vertical eye rotations during head roll by avoiding horizontal and vertical head rotations; in this way, the direction of gaze was held close to the position in which the coils were calibrated. Using this approach, we minimized ocular torsion due to moving the eye to a tertiary position in the orbit (accounted for by Listing's law), which influences vestibular eye movements about half as much as saccadic movements. [10] During head roll, the median amplitude of horizontal head rotations was +/- 3 degrees (range +/- 1 to +/- 12 degrees) and of vertical rotations was +/- 2.5 degrees (range +/- 1 to +/- 5 degrees).
We carried out six experimental paradigms; each trial lasted 20 seconds. The frequency and amplitude of head movements were controlled manually by one of the investigators.
1. Position-step head displacements (peak velocity about 120 degrees/sec), at about 0.25 Hz in the roll plane while subjects viewed a target located at 7.2 m ("far target").
2. Position-step head displacements (peak velocity about 120 degrees/sec), at about 0.25 Hz in the roll plane while subjects viewed a target located at 20 cm ("near target").
3. Quasi-sinusoidal head rotations at 1.0 to 2.0 Hz (peak velocity about 100 degrees/sec) in the roll plane while subjects viewed a target located at 7.2 m.
4. Quasi-sinusoidal head rotations at 1.0 to 2.0 Hz (peak velocity about 100 degrees/sec) in the roll plane while subjects viewed a target located at 20 cm.
5. Quasi-sinusoidal head rotations at 1.0 to 2.0 Hz (peak velocity about 90 degrees/sec) in the yaw plane while subjects first viewed a target located at 20 cm (for 10 seconds) and then a target located at 7.2 m (for 10 seconds).
6. Quasi-sinusoidal head rotations at 1.0 to 2.0 Hz (peak velocity about 70 degrees/sec) in the pitch plane while subjects first viewed a target located at 20 cm (for 10 seconds) and then a target located at 7.2 m (for 10 seconds).
In addition, we measured horizontal and vertical saccades ranging 10 to 20 degrees and carried out a cover test to measure ocular alignment. Data were filtered (bandpass 0 to 90 Hz) before digitization at 200 Hz. The ocular responses to static head roll were measured after the eye had come to a resting position (Figure 1). Responses containing blinks or rapid drifts were not analyzed. Although we recorded movements of both eyes, torsional movements were conjugate in all patients and subjects, and data from the right eye channel were analyzed. We also measured the change in vertical alignment (right eye compared with left eye) during each static head roll. For data from quasi-sinusoidal head rotations (VVOR), we differentiated head and eye-in-head signals, desaccaded the data, [13] and subjected the remaining arrays (n > 400) to linear regression to measure the slope (gain) and its SE; we compared different gains using a t test. [14]
Figure 1. Representative responses to step head rotations in roll from a normal subject and three patients, during distant viewing. Responses were increased in patient 2 (with skew deviation) and decreased in patient 4 (with skew deviation) and patient 10 (with VIIIS). Note that quick phases occurred predominantly in the direction of ongoing torsional nystagmus in patient 4. Also note the different time scale for patient 4.
Measurement of SVV.
In the four skew patients (patients 1 through 4), we measured SVV monocularly, using double Maddox rods, while their heads were held erect. In patients 5 through 10 and in the normal subjects, we measured SVV binocularly using a visual hemisphere with a movable disc, similar to that described by Dieterich and Brandt, [12] except that subjects made six estimates with the head erect or tilted 45 degrees to the right or the left shoulder (18 measurements per individual). The patients with ST were able to position their heads similar to the control subjects.
Because many data were not normal in distribution, we present median values and interquartile ranges (25th to 75th percentiles) and used the Wilcoxon rank sum test for statistical comparisons, unless otherwise stated.
Results.
Fixation behavior.
With the head stationary and approximately erect and eyes close to primary position during fixation of the far target, all four patients with skew deviation showed torsional nystagmus. In patients 1, 2, and 4, the torsional quick phases beat clockwise; in patient 3, they beat counterclockwise. Mean slow-phase eye velocity was <1 degrees/sec in patient 1, 7 degrees/sec in patient 2, 1.5 degrees/sec in patient 3, and 2 degrees/sec in patient 4. This spontaneous nystagmus had variable, but lower-amplitude, components in the vertical plane (see Table 1) being hemisee-saw in patient 2. [15] The amplitude of the torsional nystagmus was similar during fixation of the far and near visual targets. In addition, patient 10 showed intermittent torsional nystagmus with clockwise quick phases during fixation; slow-phase velocity was <1 degrees/sec. No other patient or control subject showed torsional nystagmus, although bidirectional drifts were greater in this plane than in the horizontal or vertical directions, as previously noted. [5,16]
Static ocular counter-roll.
Examples of ocular counter-rolling in a subject and three patients are shown in Figure 1. The relationships between head roll and ocular counter-roll while viewing the far target are summarized in Table 2 and Figure 2. No subject or patient showed consistent asymmetries of responses to rightward versus leftward head roll (i.e., no significant difference in gain values), even though quick phases occurred predominantly in the direction of ongoing torsional nystagmus in patients 1 through 4 (e.g., patient 4 in Figure 1). The median gain of ocular counter-rolling in the control subjects was 0.24. Of the skew deviation patients, patient 2 showed increased gain, whereas patient 4 showed reduced gain; both were significantly different from the control subjects (p < 0.05). Of the ST patients, only patient 8 showed responses consistently different from control subjects, with a significantly increased gain (p < 0.001). Of the two VIIIS patients, both showed some reduction in gain without directional asymmetry, but the gain was significantly reduced only for patient 10 (p < 0.01).
Table 2. Gain values of static ocular counter-roll
Figure 2. Plots of changes in torsional eye position (ocular counter-roll) in response to static changes in head roll while viewing a far target. NS +/- 95% PI indicates mean and 95% prediction intervals for the population of data from normal subjects. Responses from patient 2 (fill circle) and patient 8 (fill triangle) are increased and from patient 4 (fill square) and patient 10 (open diamond) decreased significantly (p < 0.05) compared with control subjects.
While viewing the near target, the median gain of ocular counter-rolling in the control subjects was 0.18; this was significantly less (p < 0.001) than while viewing the far target. Of the skew deviation patients, two were similar to the normal subjects: patient 1 showed an increased and patient 4 a reduced gain; both were significantly different from the control subjects (p < 0.01). Of the ST patients, only patient 8 showed a gain that was significantly increased (p < 0.001). Of the two patients with unilateral vestibular loss, both showed some reduction in gain without directional asymmetry, but the gain was only significantly reduced for patient 10 (p < 0.001).
Dynamic ocular counter-roll (torsional VVOR).
The values of the gain of the torsional VVOR while viewing the far target are summarized in Figure 3. The distributions of gain values for control subjects are displayed as box plots; values for patients are shown as data points. The response is broken down into the direction of head movement (clockwise or counterclockwise, with respect to the subject) and the hemi-range of movement about the head-erect position (with right or left ear lower). There was no significant difference between gain values for direction or hemi-range of movement in normal subjects; the overall median gain was 0.82 (interquartile range 0.75 to 0.86). Of the patients with skew deviation (see Figure 3A), patients 2 through 4 showed gain values that lay outside 95% CIs for normal subjects, and all three showed asymmetries of gain values for the direction of head roll in both hemi-ranges of movement (p > 0.005). Patient 2 showed increased gain values (median 1.00); the larger values corresponding to clockwise head rotation partly reflected his resting torsional nystagmus with clockwise quick phases. In contrast, patient 3 (median 0.46) and patient 4 (median 0.27) showed decreased gain values; asymmetries of responses could have been partly ascribed to effects of spontaneous nystagmus in patient 3 but not in patient 4. Responses in patient 1 were at or below the 10th percentile for normal subjects but showed no asymmetry.
Figure 3. Box plots summarizing gain values of the torsional VVOR (dynamic counter-rolling) with sinusoidal head rotations. The data are displayed to show the median, 10th, 25th, 75th, and 90th percentiles (as indicated in A). Test condition refers to each direction and hemi-range of head roll about the head-erect position, during viewing of a distant target. CWM-clockwise motion; CCWM = counterclockwise motion; CWHR-clockwise hemi-range (right ear down); CCWHR = counterclockwise hemi-range (left ear down). (A) Patients with skew deviation; (B) patients with ST; (C) patients with VIIIS.
Three patients with ST (see Figure 3B) had gain values that lay outside 95% CIs for normal subjects. Gain values were increased in patient 8 (median 0.92) and decreased in patient 5 (median 0.57) and patient 6 (median 0.58). No patient with spasmodic torticollis showed spontaneous nystagmus. Both VIIIS patients showed reduced gain values; this decrease was greater in patient 10 (median 0.42) than in patient 9 (median 0.55). In both patients, there was a significantly smaller gain (p < 0.001) when the direction of head roll was toward the abnormal ear (i.e., counterclockwise for patient 9 and clockwise for patient 10). However, there was no significant difference between responses in the same direction but in different hemiranges of movement.
When normal subjects and patients viewed the near target, gain values of the torsional VVOR significantly decreased (p < 0.01) compared with viewing the far target; median was 0.74 (interquartile range 0.68 to 0.78). During near viewing, the differences between patients with skew deviation and normal subjects were still evident, with increased gain in patient 2 (median 0.87) and decreased gain in patient 3 (median 0.45) and patient 4 (0.05), but no patient with ST showed gain values consistently outside of the 95% confidence range for normal subjects. During near viewing, gain values in the two VIIIS patients were qualitatively similar to those during far viewing in terms of direction of head movement; median gain was 0.45 in patient 9 and 0.37 in patient 10.
Effect of head roll on skew deviation.
Static head roll produced only small changes of vertical alignment of the eyes, and no normal subject or patient, apart from those with skew deviation, reported diplopia during this testing. When normal subjects viewed the far target, the median (interquartile range) change in vertical alignment was 0.18 degrees (0.12 to 0.23) of depression of the right eye compared with the left eye, per degree of head roll to the right. Thus, for a 20 degrees clockwise head roll to the right, the right eye would be depressed about 3.6 degrees compared with the left. No patient, including those with resting skew deviations (patients 1 through 4), showed asymmetrical changes of vertical ocular alignment when they rolled the head to the right or the left. During viewing of the far target, patient 2 showed a significantly (p > 0.05) reduced response (median 0.07), and patient 4 showed an inverted response (median -0.02), as if he was viewing a near target (see below).
When normal subjects viewed the near target, the median (interquartile range) change in vertical alignment was 0.09 degrees (0.02 to 0.14) of increased elevation of the right eye compared with the left eye, per degree of head roll to the right. One normal subject did not show this pattern but responded with an inverted response (median -0.09), as if he was viewing a far target. During viewing of the near target, patient 6 (median -0.05) also showed an inverted response, as if she was viewing a far target.
Horizontal and vertical VVOR.
The results are summarized in Table 3. For the skew patients, patient 1 showed low gain values, whereas patient 3 showed an increased gain of her horizontal VVOR at far (measured from movements of her right eye) and patient 4 showed no increase in his VVOR gain during near-viewing in either plane. Of the patients with ST, patients 5 and 6 both showed low values for their horizontal VVOR, regardless of viewing conditions; no patient showed an asymmetry of responses. Patients 9 and 10 showed low gain values for both their horizontal and vertical VVOR, during viewing of far and near targets.
Table 3. Gain values of horizontal and vertical VVOR
Subjective visual vertical.
Results from the patients with skew deviation, measured using Maddox rods for each eye, are summarized in Table 1; all showed a deviation of 4 degrees or more in at least one eye. Results from patients with ST, measured binocularly using a hemisphere and movable disc, were not significantly different from normal subjects. The range of medians for ST patients was from -2.5 to +0.75 degrees. The 95% CI from normal subjects was 0.0 +/- 4.75 degrees. Binocular SVV for patient 9 showed a median deviation to the left of 3.5 degrees and for patient 10 showed a median deviation of SVV to the right of 1.75 degrees; neither fell outside the 95% CI for normal subjects.
Discussion.
We investigated whether ocular counter-rolling, induced by either static or sinusoidal head rotation in roll, was abnormal in patients with skew deviation, ST, or unilateral vestibular nerve section. We found abnormalities in all three groups of patients, with either increases or decreases of the ocular counter-rolling responses to static or sinusoidal head roll. The abnormalities were most marked while viewing distant targets and during sinusoidal rotation. Before attempting to explain these findings, we summarize normal properties of torsional eye movements, because they differ substantially from properties of horizontal and vertical movements.
Distinctive properties of normal torsional eye movements in normal subjects.
All functional classes of eye movements show some differences in the torsional plane. During fixation, gaze is less stable in the torsional than in the horizontal or vertical planes; torsional drifts are greater by a factor of four than in the other planes. [5,16] Visually induced torsional responses, such as optokinetic nystagmus, are weak. [6] The gain of the torsional VOR, even with visual enhancement, is never high enough to compensate for natural head movements, [5-10,17] and dynamic enhancement of the VOR ("velocity storage" [18]) is absent. [19] The torsional VOR can be suppressed, however, probably by both visual and nonvisual mechanisms. [6,17,20] The ability to hold the eye in an eccentric torsional position ("neural integrator" function) is poorly developed, so that the eye drifts back to its resting position. [7] Torsional saccades cannot be made voluntarily, without extensive practice. [21] These distinctive properties of torsional eye movements may be related to the geometry of the eyes, which dictates a different set of visual demands from horizontal and vertical movements. Thus, if the eye drifts in the torsional plane, images are not displaced from the fovea; only in the periphery of the retina, where photoreceptor density is lower, is there an appreciable displacement of images. The same is true during head roll: in the absence of compensatory eye movements, only in the retinal periphery will there be appreciable image slip. Thus, from a visual standpoint, a modest torsional VOR is probably all that is required to maintain visual acuity and lessen image slip in the periphery. Furthermore, perceptual mechanisms involved in the processing of visual information appear to be better suited for tolerating image slip in the torsional than in the horizontal and vertical planes. Thus, although the stability of gaze in the torsional direction is much less constant than in the horizontal or vertical directions, vision remains clear and stable. One exception to this apparent laxness of control in the torsional plane concerns cyclovergence (torsional vergence); for sustained disparities between two textured patterns, the gain of the cyclovergence response is closer to 1.0, probably to ensure normal depth perception. [22]
Geometric factors also dictate that the gain of ocular counter-rolling will be reduced during viewing of a near target, due to the medial position of the eyes during convergence. [23] This is the opposite of the case for horizontal or vertical head rotations, for which the VVOR gain is increased at near. [24] In our normal subjects, the gain of ocular counter-rolling (torsional VVOR) was decreased during viewing of the near target for both static and dynamic head roll. Furthermore, geometric factors dictate vertical adjustments of eye position during head roll. If a subject fixes on a target at optical infinity, no vertical rotation of the eyes is necessary to compensate for head roll. Additionally, if the subject fixes on a target lying at a near point during head roll, no vertical rotation of the eyes is necessary provided the target lies in the subject's midsagittal plane. In this latter case, the eyes must rotate vertically in the orbits to compensate for vertical linear displacement (translation) but must also rotate in the opposite direction due to the fact that the eyes are turned in [23]; these two factors cancel out and no vertical rotation of the eyes is required. We found that changes in vertical alignment in normal subjects were small and somewhat idiosyncratic. In general, during viewing of the distant target, the change in vertical alignment was as if the eye moved down on the side to which the head was rolled. During viewing of the near target, the change in vertical alignment was as if the eye moved up on the side to which the head was rolled.
Abnormalities in patients with skew deviation.
All four patients with skew deviation showed torsional nystagmus that was not always obvious clinically. Recognition of such nystagmus is not easy, even with the ophthalmoscope, and it was probably missed in previously reported cases of skew deviation or ocular tilt reaction. Although there was a right OTR in all four patients, quick phases of torsional nystagmus were directed counterclockwise in patient 3 but clockwise in the other three patients. The direction of nystagmus is not readily explained, and systematic study of this is needed. For example, lesions of the interstitial nucleus of Cajal (INC) and rostral interstitial nucleus of the medial longitudinal fasciculus (riMLF) may both cause tonic torsional deviation that is contralesional, but associated torsional nystagmus has fast phases that are ipsilesional with INC lesions (i.e., compensatory to the tonic torsion) but contralesional with riMLF lesions. [25]
Three of four patients with skew deviation showed abnormalities of ocular counter-rolling, and dynamic responses in the fourth (patient 1) were consistently at or below the 10th percentile for normal subjects (see Figure 3A). These changes were more evident while viewing the far target. In patient 2, the resting nystagmus could partly, but not completely, explain increased gain values. However, in patients 1, 3, and 4, the nystagmus could not be held accountable for the low gain values. Responses were asymmetrical in patients 2 through 4, with abnormal gains (increased in patient 2 and decreased in patients 3 and 4) consistently occurring with clockwise head movements, irrespective of the hemi-range of motion. The absence of a clear dichotomy between abnormalities of static and dynamic responses correlates with the demonstration of convergence of inputs from the otoliths and semicircular canals in vestibular nuclei, under the influence of the cerebellum. [26,27] Nevertheless, the more prominent gain changes during sinusoidal testing and the presence of nystagmus are typical of a canal imbalance, as is dependence of the response asymmetry on the direction of head motion.
In contrast, the vertical misalignment in patients with skew deviation changed only slightly more than in control subjects with head roll except for patient 4, who responded as if he was fixing on the near target while he was viewing the target at 7.2 m. Patient 4 also showed an esophoria and failure to increase the gain of horizontal or vertical vestibulo-ocular responses while viewing the near target. In patients 1, 2, and 4, the gain of the horizontal and vertical VVOR was slightly lower than in control subjects. In patient 3 (measured in her right eye), horizontal values were increased, perhaps reflecting plastic-adaptive changes that may occur in the eye contralateral to the side of INO. [28]
Abnormalities in patients with spasmodic torticollis.
Three of four ST patients showed abnormalities, most evident for dynamic counter-roll during distant viewing. Patient 8 showed increased gain, whereas patients 5 and 6 showed decreased values. No patient showed torsional nystagmus that might account for these findings, and no directional asymmetries were evident. A prior study of ocular counter-rolling during very low-frequency rotation demonstrated "a lack of sustained eye torsion at the extreme positions, resulting in rolling of the eyes in the direction of head tilt rather than counter-rolling." [29] Other studies investigated horizontal vestibular responses in darkness and demonstrated asymmetries [30] and "hyperactivity" of responses. [31] Our measurements were of the VVOR in the horizontal and vertical planes. Two patients, 5 and 6, who had past vestibular symptoms, showed reduced gain values horizontally; both had normal vertical responses. Whether vestibular abnormalities are the root cause or an effect of ST (due, for example, to reduced neck motion) has been investigated but not settled, [31-33] and there might be a subgroup of patients in whom ST is precipitated by vestibular disease. [32] Our present study of a relatively small group of patients with ST suggests that abnormalities in the torsional plane may be quite common and easier to study than those in the horizontal and vertical planes. Furthermore, as discussed above, the visual system has limited ability to influence torsional eye movements, and abnormalities might be expected to be more apparent and enduring in this plane.
Abnormalities in patients with unilateral eight nerve section (VIIIS).
Both of our patients with long-standing VIIIS showed abnormalities of ocular counter-rolling that, as in other subjects, were more marked for distant viewing. Especially during dynamic testing (see Figure 3C), the decrease in gain was most prominent when the direction of head roll was toward the abnormal ear; this directional asymmetry was greater than in any other patient group. This asymmetry of torsional VOR gain in VIIIS patients was dependent on the direction of the dynamic movement but was independent of the positional hemi-range of head movement-to the right or left of the erect head position. Furthermore, asymmetries were not evident during static responses. Taken together, this suggests that the changes in ocular counter-rolling demonstrated in our VIIIS patients were due to an imbalance of canal rather than otolithic influences. Patients 9 and 10 also showed reduction of their horizontal and vertical VVOR. It has been shown that the gain of horizontal and vertical VOR, in response to high-acceleration stimuli, can be permanently reduced after unilateral eighth nerve section. [34,35] Because the abnormalities in patients 9 and 10 were more evident with sinusoidal than with static head rotations, measurement of the transient response to high-acceleration head rotation in roll should be the optimal way to demonstrate asymmetries of ocular counter-rolling. Unfortunately, we did not apply high-acceleration stimuli and thus cannot relate these studies directly to findings in our patients.
Potential role of measuring ocular counter-roll in neurologic patients.
Our present study indicates that measuring torsional eye movements, especially during sinusoidal head rotations, may provide a new method to study vestibular abnormalities. In the horizontal and vertical planes, the vestibulo-ocular reflexes can undergo extraordinary adaptive changes in response to visual stimuli, including inversion of responses. [1] In the torsional plane, visual inputs do not seem to figure so strongly with the possible exception of static cyclovergence. [22] Apparently because of a lack of visual adaptation in the torsional plane, asymmetries of ocular counter-rolling lasted years after eighth nerve section in patients 9 and 10. This may be contrasted with skew deviation, which rapidly resolves after eighth nerve section. [36,37] On the other hand, perceptual aspects of disturbed vestibular function in roll, such as SVV, are reported to improve, usually within a few months, [3,12,36] and were only abnormal in our patients with acute skew deviation. Static ocular torsion also shows some recovery, [36,37] unlike the persistent abnormalities of dynamic counter-rolling than we demonstrated.
We did not attempt to provide a complete description of abnormalities of torsional eye movements in these three disorders, in which abnormal ocular counter-rolling might be expected. However, we were surprised to find such frequent, distinct, and persistent changes; further studies of torsional eye movements seem warranted in these disorders and others, such as diseases of the cerebellum which may be important in controlling the torsional VOR. [27,38]
Acknowledgments
We are grateful to Scott H. Seidman, PhD, for clarifying the geometry concerning the vertical alignment of the eyes that is necessary during head roll and for critically reading the manuscript. We are also grateful for Dr. Robert A. Ratcheson for referring patient 9.
- Copyright 1997 by Advanstar Communications Inc.
REFERENCES
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.