Ocular torsion and perceived vertical in oculomotor, trochlear and abducens nerve palsies

Transcription

1 Brain (1993), 116, Ocular torsion and perceived vertical in oculomotor, trochlear and abducens nerve palsies Marianne Dieterich and Thomas Brandt Department of Neurology, Klinikum Grosshadern, Ludwig-Maximilkms-UniversitOt MQnchen, MUnchen, Germany SUMMARY Ocular torsion (OT) and subjective visual vertical (SW) were determined in acute and chronic oculomotor (n = 6), trochlear (n = 21) and abducens (n = 7) palsies separately for each eye in the primary position with the head upright. Ocular torsion measured by fundus photographs was not only within normal range in all abducens palsies, but unexpectedly also in 68% of third and fourth nerve palsies which involve oblique eye muscles. Pathological OT, when measurable, was slight (2 8 ), monocular and occurred either in the paretic or in the nonparetic eye. Subjective visual vertical tilts were more frequent (67% of third and fourth nerve palsies) although mostly small in amplitude (1 6 ). They were confined either to the paretic or the nonparetic eye depending on the duration of the palsy. Determinations of SW were always normal under binocular viewing conditions. The dissociated occurrence of OT and S W tilts in the paretic or the nonparetic eye was dependent on the acuteness of the palsy and reflected sensory and/or motor compensation mechanisms. Third and fourth nerve palsies cause only minor and unpredictable monocular OT and S W tilts as distinct from the frequent binocular and conjugate tilts seen in patients with acute unilateral brainstem lesions. INTRODUCTION The eyes not only move in horizontal and vertical directions but also rotate about the line of sight, i.e. the roll plane. The significance of the latter has recently been elaborated for topographical diagnosis in neurology. Direction-specific deviations of the position of the eye in roll, i.e. ocular torsion (OT), and the subjective visual vertical (SWO are sensitive clinical signs in acute unilateral brainstem lesions (Brandt and Dieterich, 1992; Dieterich and Brandt, 1993). A lesion caudal to the upper pons causes ipsiversive OT (upper pole of the eye tilted ipsilaterally) of one or both eyes with concurrent ipsiversive tilts of S W adjustments. A lesion rostral to this pontine level causes contraversive tilts of OT and SW. Peripheral third and fourth nerve palsies can also induce OT in the resting position of the eye. Misleading combinations of peripheral ocular motor and central vestibular pathway lesions on OT and S W may occur in midbrain tegmentum lesions Correspondence to: Dr Marianne Dieterich, Department of Neurology, Klinilcuin Grosshadern, Ludwig-Maximilians- Universitfit MQnchen, Marchioninistrasse 15, HD MOnchea, Germany. Oxford University Press 1993

2 10% M. DIETERICH AND TH. BRANDT that involve the oculomotor or trochlear nuclei or fascicles (Dieterich and Brandt, 1993). Dissociated or monocular deviations of OT and SVV may then result. Determination of the differential effects of extra-ocular eye muscle pareses on eye position in roll and its perceptual consequences for perceived vertical are prerequisites for appropriate localization of a lesion within the ocular motor system, infra- or supranuclear. The following study was therefore conducted as an essential control for our earlier investigations of OT and S W in unilateral brainstem lesions. The major questions were: (i) Do acquired third, fourth or sixth nerve palsies cause significant OT of the eye? (ii) Do they induce a monocular or binocular tilt of perceived visual vertical? (iii) Do direction and amplitude match if OT and SW tilt manifest in the same eye? (iv) Do infranuclear ocular motor disorders cause characteristic OT and SVV tilts useful for differentiation from supranuclear ocular motor disorders? PATIENTS AND METHODS Patients Thirty-four patients(14 women, age range years, mean = 48 years; 20 men, age range years, mean = 41 years) with the diagnosis of an extra-ocular eye muscle paresis were included in the study. Twentyone patients suffered from trochlear palsies (five women, mean = 46 years; 16 men, mean = 41 years), six patients suffered from oculomotor palsies (four women, mean = 45 years; two men, mean = 51 years), and seven patients had abducens palsies (five women, mean = 51 years; two men, mean = 31 years). Onset of the palsy varied between 3 days and 4 years (mean = 216 days) prior to the evaluation. In the subgroup of 16 patients with an acute onset of the disease, the mean of the onset of the palsy was 7.6 days. Seventeen of the 21 fourth nerve palsies were located unilaterally (11 right, six left), three bilaterally asymmetric and one bilaterally. All six third nerve palsies were unilateral (three left, three right). Six of the seven sixth nerve palsies were unilateral (four left, two right), one was bilateral. Aetiology of the palsies was mostly traumatic (n = 15), followed by ischaemic (n = 7), inflammatory (n = 4), compressive by a small cerebellopontine angle tumour (n = 3) without brainstem compression, abscess (n = 2) and aneurysm (n = 1), or unknown (n = 2). None of the patients had concurrent brainstem or cerebellar signs. Patients with supranuclear ocular motor disorders, such as skew deviation (identified by die clinical syndrome and neuroimaging), were excluded. Furthermore, patients who were taking pharmacological medications (analgesic, anticonvulsant, sedative) or who had ocular motor deficits from earlier central disorders, or diplopia prior to the onset of disease were also excluded. History taking made a confusion with strabismus or a decompensated childhood hyperphoria unlikely. Orthoptic examination Orthoptic examination included the measurement of head tilt (by optical methods using a protractor), determination of die width of the lid fissure (ptosis?), pupillary examination (light reactions, convergence test, accommodation test, determination of pupil diameters), measurements of muscle balance in primary position and different fields of gaze (differences in primary and secondary deviation and variation of squint angle with different fields of gaze). Vertical divergence (hypertropia/hypotropia) of the eyes was determined with the prism cover test, first with the head in the tilted position and secondly with the head upright, while viewing a target (at a distance of 5 m) binocularly with gaze straight ahead. Furthermore, vertical divergence was measured with Bielschowsky's head tilt test (45 tilt to left and right). Ocular torsion Ocular torsion, in degrees, was determined as the mean from four to six fundus photographs taken with the head upright in the sitting position. The head was also adjusted to true upright in patients with compensatory head tilt (e.g. fourth nerve palsy). Fundus photographs were taken for both eyes separately during monocular fixation of a central target and with pupil dilation by a mydriatic drug (tropicamide). The viewing target consisted of a dot (1 in diameter) presented at the lower end of a vertical bar. The position of the eye in

3 OCULAR TORSION 1097 roll plane was determined as the angle between a straight line through the papilla and fovea (papilla fovea meridian) and the horizontal line (for details on the methods, see our previous article, Dieterich and Brandt, 1993). With this method, in 80 normal subjects, the position for both eyes in roll plane is a slight excyclotropia (i.e. counter-clockwise rotation of the right eye, clockwise rotation of the left eye, from the viewpoint of the examiner). The right eye shows an excyclotropia of 4.9 ±2.9 (mean ±1 SD). The left eye shows an excyclotropia of 5.7 ±2.9. The normal range of OT (mean ±2 SDs) was -1 to +11 for the right eye and 0 to for the left eye (Dieterich and Brandt, 1993). Our normal values agree with similar data obtained by Bixenman and von Noorden (1982) and Herzau and Joos (1983) in smaller samples. Subjective visual vertical For determination of S W, subjects sat with their head fixed in the upright position by means of a bite board or a band fastened around the occiput, and looked into a hemispherical dome 60 cm in diameter. [Measures were taken with the head upright and not in different head positions in roll because of the bias by physiological Aubert- and Muller-phenomena with considerable interindividual variations (Dichgans et a]., 1974) and for better comparison between our group of normal subjects and patients with brainstem infarctions.] The surface of the dome extended to the limits of the observer's visual field, and was covered with a random pattern of coloured dots and contained no clues to gravitational orientation. The centre of the dome was fixed to the shaft of a DC torque motor; 30 cm in front of the observer was a circular target of 14 visual angle with a straight line through the centre mounted on a coaxial shaft connected to a DC servo-motor. The central test edge had to be adjusted to the vertical by the subject, using a potentiometer. The output of the potentiometer was recorded on a strip chart recorder. Visual vertical was determined by means of 10 adjustments of the target disc from a random offset position to the subjective vertical with the hemispherical dome stationary. Under these conditions, the normal range of the visual vertical is ±2.5 under all three conditions (median ±2 SEM: -1.5 to +2.1 binocularly, -2.3 to +2.5 monocularly right eye, -3.7 to +1.1 monocularly left eye; for details, see our previous article, Dieterich and Brandt, 1993). There was no significant difference for binocular or monocular stimulation (Spearman's rank correlation coefficient: P < 0.001, r s = 0.49, n = 110). Static SW was measured binocularly and monocularly for both eyes for comparison with OT of each eye. RESULTS Pathological OT and SVV tilts were found in third and fourth nerve palsies, but not in the seven patients suffering from sixth nerve palsies. However, only seven of 22 patients (32%) with third and fourth nerve palsies exhibited pathological OT which usually occurred monocularly. In three patients the paretic eye was affected but surprisingly in four the nonparetic eye was involved (Table 1; Fig. 1). Slight OT (incyclotropia in third or TABLE 1. FREQUENCY OF SVV TILT AND OCULAR MOTOR PALSIES Peripheral ocular motor palsy Third nerve Fourth nerve Sixth nerve Total Percentages Third + fourth Percentages n Number Uniku Bilat Binocular conjugate 0/21 on 0/34 0% 0/27 0% or r uu Paretic 4/6 7/21 0/7 11/34 32* 11/27 41% Nonparetic 2/6 5/21 0/7 7/34 21% 7/27 26% Binocular conjugate 0/16 0/28 0% 0/22 0% Paretic 3/16 3/28 11% 3/22 14% Nonparetic 1/6 3/16 4/28 14% 4/22 18% Unilat. = number of unilateral palsies; bilat. the same direction when tested monocularly. number of bilateral palsies; binocular conjugate = tilt of both eyes in

4 1098 M. DIETERICH AND TH. BRANDT paretic eye nonparetic eye SW - Tilt excyclotropia binocular SW SW-Tilt SW-Tilt - 7 FIG. 1. Schematic representation of the eye position in rou (fundus photographs and adjustments of SW) in three patients with fourth nerve palsy of the right eye. In the acute case (day 7; lop) ocular torsion (OT) was within normal range of excyclotropia of 7-8. Subjective visual vertical was only tilted with monocular adjustment by the paretic eye (19 ). The two chronic cases (year 2, middle; year 3, bottom) caused different abnormalities: pathological excyclotropia to 17 of the paretic eye with normal adjustment of SW but contralateral SW tilt to 8 in the nonparetic eye which showed normal position in roll (middle). The third patient (bottom) showed only SW tilt for the nonparetic eye without pathologica] OT of both eyes. Binocular measures of SW were always within normal range. These three cases demonstrate the separation of OT and SW tilt in extra-ocular eye muscle paresis and their dependence on the time course of the palsy.

5 OCULAR TORSION 1099 excyclotropia in fourth nerve palsies) may go undetected in our study because of the conservative definition of a normal range from 0 to 11 excyclotropia. Pathological OT was 6 8 beyond normal range for the paretic eye (n = 3) and 1 3 for the nonparetic eye (n = 4). Pathological SVV tilts were more frequent. Eighteen (67%) of the 27 patients with third and fourth nerve palsies showed S W tilts when tested monocularly, 11 (41%) affecting the paretic eye, seven (26%) the nonparetic eye (Table 1). In all patients with SW tilts confined to the paretic eye the palsy was acute lasting between 3 days and 2 months, whereas in S W tilts confined to the nonparetic eye palsy was chronic lasting between 3 months and 3 years. Subjective visual vertical tilts were not observed when determined under binocular viewing conditions. Net tilt angles of S W were 2.45 ± 1.4 (mean ± SD; n = 11) beyond normal range for the paretic eye and 2.9 ±2.0 (n = 7) for the nonparetic eye. In an exceptional patient a higher SW tilt angle of 16.5 was found. It should be noted that SW tilts and pathological OT did not occur simultaneously within the same eye and that the direction of S W tilt was not predictable by the expected or measured position of the eye in roll (in eight of 10 patients direction of SW tilt confined to the paretic eye corresponded to the expected paretic OT). The amount of SW tilts and OT was not related to the grade of the paresis. Oculomotor palsies Of six patients with unilateral third nerve palsies five had normal eye position in roll (no OT) but none had normal S W adjustments when tested monocularly (Table 1; Figs 2, 3). Four of the six monocular S W tilts affected the paretic eye (net tilt angles, 1.1, 1.6, 6.0 and 16.5 ; mean = 8.9 ) and two the nonparetic eye (net tilt angles, 0.8 and 3.2 ). One would expect SW tilt to occur as a sensory consequence of paretic cyclodeviation, in this case incyclotropia. The direction of SW tilt, however, corresponded to expected incyclorotation of the paretic eye only in four of the six patients, but in two (one with a complete paresis) the paretic eye monitored a 'paradoxical SVV tilt' to the opposite side. In another patient SW tilt was measured in the paretic eye whereas slight OT of 1 2 was found in the nonparetic eye. In Table 2 different subjective and objective measures of cyclodeviation are compared in a patient with a right third nerve palsy lasting 7 days. The data demonstrate the poor correlation of tilt angles when determined with different methods. Trochlear palsies In six of 16 measured patients (38%) pathological OT was found monocularly (three paretic eye: 6-8 ; three nonparetic eye: l -3 ) (Table 1). Independent of the eye involved (paretic or nonparetic) excyclotropia was the pathological position. Twelve out of 21 patients (57%) exhibited monocular S W tilts (33% confined to the paretic eye, 24% to the nonparetic eye; Figs 1, 3). Net tilt angles of SVV were 2.25 ±1.4 (range ) for the paretic eye and 3.2 ±2.0 (range ) for the nonparetic eye (Fig. 2). All SW tilts determined for the paretic eye (n = 7) corresponded in direction to expected excyclotropia, whereas in four of the five SVV tilts determined for the nonparetic eye the direction was opposite. One patient with an acute (sixth day) bilateral palsy had SW tilts for both eyes in opposite directions. Adjustment of SW was normal with binocular viewing.

6 1100 M. DIETERICH AND TH. BRANDT Nonparetic Eye Paretic Eye D I D 2 f IVth nerve Illrd nerve Illrd nerve IVth nerve FIG. 2. Net tilt angles of pathological OT (open squares) and S W tilt (filled circles) in 14 of 21 patients with trochlear palsy and in six patients with oculomotor palsy determined as the means of four to six fundus photographs or 10 adjustments of SVV separately for the paretic and the nonparetic eye. Means (± 1 SD) of SW tilts in trochlear palsies are indicated by the asterisk. Ocular torsion and SW tilt in degrees are moderate between 1 and 8 and not only occur in the paretic eye but also in the nonparetic eye. SW tilts mostly corresponded to the direction of ocular torsion. Vertical divergence could be determined in 18 out of 21 patients and varied from 1 to 5 in the primary position of gaze. Bielschowsky head tilt test was positive in 19 patients, 17 of whom showed an abnormal head position. Abducens palsies None of the seven patients with monocular (n = 6) or bilateral (n = 1) sixth nerve palsies exhibited pathological OT or SVV tilts (Table 1). Tilt angles of SVV were within normal range, for the paretic eye 2.0 ±0.7 (mean ±SD), and for the nonparetic eye 1.0 ±0.7. DISCUSSION Pathological OT and SVV tilts were only found in third and fourth nerve palsies which induced a cyclodeviation by a paresis of a cyclovertical extraocular muscle or muscle group (incyclorotation in third nerve palsy, excyclorotation in fourth nerve palsy in the primary position). They were not found in the seven patients presenting with sixth nerve palsies. Ocular torsion in third and fourth nerve palsy With respect to the eye position in roll plane the three major findings were: pathological OT occurred in only seven of 22 patients (32%); it was always monocular and could affect either the paretic or the nonparetic eye; objective measurements of OT by fundus

7 OCULAR TORSION 1101 A Unilateral Illrd nerve palsy B Unilateral IVth nerve palsy Unilateral bralnstem lesion ato pontonwdudary FIG. 3. Determination of SW in unilateral oculomotor palsy (A, n = 6), trochlear palsy (B, n = 21) and in patients with acute unilateral brainstem infarction (c, n = 38) without concurrent peripheral ocular motor palsy. Values are presented for binocular and monocular viewing conditions of the ipsilateral (paretic) and contralateral (nonparetic) eye. With binocular vision SW was always within normal range (median ±2 SM) in third and fourth nerve palsies. Under monocular viewing conditions slight deviations of SW were determined in 18 out of 27 patients with third and fourth nerve palsies. Surprisingly abnormal values were found either in the paretic or the nonparetic eye, not in both eyes. With acute unilateral brainstem lesions S W tilts were obtained for both binocular and monocular viewing conditions. Thus, the fact that third and fourth nerve palsies may cause moderate monocular SW tilts enables differentiation from central ocular motor disorders. TABLE 2. DIFFERENT SUBJECTIVE AND OBJECTIVE MEASURES OF OT IN A PATIENT WITH A RIGHT THIRD NERVE PALSY LASTING 7 DAYS (PRIMARY POSITION OF GAZE) Methods Subjective visual vertical Binocular viewing Subjective cyclodeviation By Maddox double rod By Harms tangent scale Objective cyclodeviation By fundus photographs Right eye +19 (tijt to the right) -1.6 (normal) 4-5 right excyclodeviation 2 excyclodeviation 8 excyclotropia (normal) Left eye -2.0 (normal) 2 excyclodeviation 7 excyclotropia (normal) Patient no. 6 (aged 43 years): right mydriasis, right hypotropia of 3 in primary position, slight head turn to the left. photographs revealed 6-8 tilts of the paretic eye and smaller 1 3 tilts of the nonparetic eye. Because of the considerable physiological variation of normal eye position in roll and our conservative definitions of a normal range from 0 to 11 excyclotropia (mean ±2 SDs) minor paretic OT may go undetected. Repeated measurements during the course of a paresis will reveal this by slight rotations of some degree within normal range.

8 1102 M. DIETERICH AND TH. BRANDT The results on OT and its consequence for vision (rotated double images) are in good agreement with the knowledge of sensory and motor mechanisms compensating for cyclodeviation. Binocular mechanisms to compensate for OT include an abnormal head posture, cyclofusional adaptations, abnormal retinal correspondence and suppression, and reorientation of the spatial values of retinal meridians. Their particular contribution to adaptation differs. Cyclodeviation in humans is compensated primarily by sensory means because sensory cyclofusion is well developed in a normal person allowing fusion of cyclodisparity in either direction (Jampel et al., 1976; Guyton and von Noorden, 1978). Motor cyclofusion accounts for only a few degrees (Crone, 1975; Crone and Everhard-Halm, 1975). In greater cyclodeviations motor cyclofusion is divided between both eyes (Crone, 1975; Crone and Everhard-Halm, 1975; Hooten et al., 1979; Herzau and Joos, 1983; Hessel and Herzau, 1983/1984). Thus, motor compensation can account for the differential effects of third and fourth nerve palsies on OT as determined in our study: pathological OT is absent or smaller than expected in the paretic eye (ipsilateral motor compensation) or OT occurs only in the nonparetic eye but to a lesser degree (bilateral motor compensation). Sensorimotor adjustment is driven by ocular dominance with the tendency to adjust the dominant eye to normal position in all three planes, roll, pitch and yaw. It is recognized that there is no good correlation between subjective and objective measurements of cyclodeviation {see Table 2) which supports the findings of others who found the maximal difference to be up to ±10 (Bixenman and von Noorden, 1982; Kolling, 1986). Subjective visual vertical tilts in third and fourth nerve palsy Monocular and binocular determinations of the visual perception of verticality in degrees of S W tilts is a much more sensitive measure than OT (normal range: ± 2.5 ) and revealed the following (Fig. 3): S W was normal in all patients on binocular viewing; monocular SW tilts (18 out of 27 patients; 67%) were more frequent than pathological OT; they were confined to the paretic eye (41 %) when the onset of the palsy was 3 days to 2 months prior to the evaluation. When the palsy lasted longer than 3 months (up to 4 years), SW tilt was confined to the nonparetic eye. Subjective visual vertical tilts did not occur simultaneously under monocular viewing conditions using the eye that showed pathological OT. Monocular SW tilts and OT were not combined. Thus, as we have stated earlier, SW tilt does not simply reflect the perceptual correlate of eye rotation (Dieterich and Brandt, 1993). It is not the net position of the eye in roll that determines visual spatial coordinates because of the poor sense of position. Paretic deviation of normal eye position due to lack of afferent extra-retinal information causes a dissociation of the subjective and objective visual directions in space (Brandt and Biichele, 1979; Brandt, 1984). Visual localization is usually calculated from both the position of the target on the retina and the awareness of the eye position in the head. Eye position information is available from the ocular motor system, with an accuracy of better than 0.5 of arc (Hansen and Skavenski, 1977), but the extra-retinal eye position information and visual localization of targets is poor during smooth pursuit (Festinger and Canon, 1965) and after saccades (Matin, 1972). Thus, perception may be influenced by the pattern of innervation of extra-ocular eye muscles, e.g. when attempting to compensate for paretic deviation by excess innervation of agonistic and inhibition of antagonist muscles (Dieterich and Brandt, 1987). In third and fourth nerve palsy sensorimotor compensation of the position in roll of the dominant

9 OCULAR TORSION 1103 eye would then cause normal OT but tilted S W (towards the compensated rather than the paretic roll). It could explain the seemingly paradoxical finding of OT and S W tilts occurring separately for the two eyes. These mechanisms are illustrated in the three patients in Fig. 1. Our findings on SW tilts are comparable to those known from the literature, based on another subjective measure for cyclodeviation, the double Maddox rod test. The double Maddox rod test showed 'subjective excyclotropia of the nonparetic eye in 15 out of 60 patients with unilateral superior oblique muscle palsy' (Guyton and von Noorden, 1978; Olivier and von Noorden, 1982). The result in the nonparetic eye was explained by a monocular sensorial adaptation to the cyclodeviation by means of a reordering of the spatial responses of retinal elements along new horizontal and vertical meridians (Graefe, 1898; Bielschowsky, 1935; Guyton and von Noorden, 1978; Olivier and von Noorden, 1982). This phenomenon was reported consistently only by those patients who habitually fixated with the (dominant) paretic eye (Rolling, 1982, 1984; Olivier and von Noorden, 1982; Herzau and Joos, 1983; Ruttum and von Noorden, 1983). Over a period of several weeks (up to 3 months) the subjective orientation of the paretic eye normalizes in the roll plane and pathological excyclotropia is transposed to the nonparetic eye. Ocular torsion and SW tilts in supranuclear and infranuclear ocular motor disorders Subjective visual vertical tilts and OT are sensitive signs for topographic diagnosis of acute unilateral brainstem lesions (Brandt and Dieterich, 1992; Dieterich and Brandt, 1993). Perceptual and eye tilt in central lesions are typically associated in direction (ipsilateral with pontomedullary, contralateral with pontomesencephalic lesions) with involvement of both eyes and under monocular and binocular viewing conditions (Fig. 3). As we have shown, the majority of peripheral lesions do not cause pathological OT and, if present, the net angles of rotation are relatively small. However, those which cause OT must be differentiated from central lesions. The following features may help to distiguish between central brainstem signs and OT and SW tilts secondary to peripheral fascicular or nerve lesions of the third and fourth nerve especially in disorders of the midbrain tegmentum. A peripheral third or fourth nerve palsy must be suspected as being the cause of pathological OT or SW tilt if: (i) OT is measurable in one eye only without concurrent S W tilt; (ii) S W tilt is only measurable under monocular and not under binocular viewing conditions. In the patient with an acute bilateral (symmetric) fourth nerve palsy, monocular measurements of S W showed tilts of similar degree for both eyes in opposite directions. Thus, even a bilateral infranuclear ocular motor disorder with cyclorotation can be differentiated from a supranuclear disorder. The latter is characterized by S W tilts under binocular viewing conditions and conjugate S W tilts under monocular viewing conditions, both tilted in the same direction. ACKNOWLEDGEMENTS We wish to thank Mrs S. Eithoff, Mrs C. Frenzel and Mrs B. Leikam for orthoptic assistance, and Mrs M. J. Finke for manuscript preparation. This work was supported by the Deutsche Forschungsgemeinschaft SFB 220, D6, and the Wilhelm-Sander-Stiftung.

10 1104 M. DIETERICH AND TH. BRANDT REFERENCES BIELSCHOWSKY A (1935) Lectures on motor anomalies of the eyes. II. Paralysis of individual eye muscles. HI. Paralysis of conjugate movements of the eyes. IV. Functional neuroses; etiology, prognosis, and treatment of ocular paralysis. Archives of Ophthalmology, 13, 33-59, , BIXENMAN WW, VON NOORDEN GK (1982) Apparent foveal displacement in normal subjects and in cyclotropia. Ophthalmology, 89, BRANDT TH (1984) Visual vertigo and acrophobia. In: Vertigo. Edited by M. R. Dix and J. D. Hood. Chichester: John Wiley, pp BRANDT TH, BOCHELE W (1979) Ocular myasthenia: visual disturbance of posture and gait. Agressologie, 20, No. C, BRANDT TH, DIETERICH M (1992) Cyclorotation of the eyes and subjective visual vertical in vestibular brain stem lesions. Annals of the New York Academy of Sciences, 656, CRONE RA (1975) Optically induced eye torsion. II. Optostatic and optokinetic cycloversion. Albrecht von Graefe's Archiv fur KUnische und Experimentelle Ophthalmologie, 196, 1 7. CRONE RA, EVERHARD-HALM Y (1975) Optically induced eye torsion. I. Fusion cyclovergence. Albrecht von Graefe's Archiv far KUnische und Experimentelle Ophthalmologie, 195, DICHGANS J, DIENER HC, BRANDT TH (1974) Optokinetic-graviceptive interaction in different head positions. Ada Oto-laryngologica, 78, DIETERICH M, BRANDT TH (1987) Impaired motion perception in congenital nystagmus and acquired ocular motor palsy. Clinical Vision Sciences, 1, DIETERICH M, BRANDT TH (1993) Ocular torsion and tilt of subjective visual vertical are sensitive brainstem signs. Annals of Neurology, 33, FESTINGER L, CANON LK (1965) Information about spatial location based on knowledge about efference. Psychological Review, 72, GRAEFE A (1898) MotilitStsstorungen mit einleitender Darlegung der normalen Augenbewegungen. In: Handbuch der gesamten Augenheilkunde, Band 8, 1. Abt, Kap. XI. Second edition. Edited by A. Graefe and Th. Saemisch. Leipzig: Breitkopf und Hartel. GUYTON DL, VON NOORDEN GK (1978) Sensory adaptations to cyclodeviations. In: Strabismus. Edited by R. D. Reinecke. New York: Grune and Stratton, pp HANSEN RM, SKAVENSKI AA (1977) Accuracy of eye position information for motor control. Vision Research, 17, HERZAU V, Joos E (1983) Untersuchungen von Bewegungen und Stellungsfehlern um ihre sagittale Achse. Zeitschrift fur praktische Augenheilkunde, 4, HESSEL L, HERZAU V (1983/1984) Erfolgt die fusionale Zyklovergenz nach dem Hering' schen Gesetz? Orthoptik-Pleoptik, 11, HOOTEN K, MYERS E, WORRALL R, STARK L (1979) Cyclovergence: the motor response to cyclodisparity. Albrecht von Graefe's Archiv far KUnische und Experimentelle Ophthalmologie, 210, JAMPEL RS, STEARNS A, BUGOLA J (1976) Cyclophoria or cyclovergence: illusion or reality? In: Orthoprics: Past, Present, Future. Edited by S. Moore, J. Mein and L. Stockbridge. New York: Stratton, pp KOLUNG G (1982) Zur Praxis der Zyklodeviationsmessung. Orthoptik-Pleoptik, 10, KOLUNG G (1984) Das Verhaltnis zwischen sensorisch wahrgenommener und motorisch objektivierbarer Transposition der Zyldotropie bei M. obi. sup. Paresen. In: Pathophysiologie des Sehens: Grundlagenforschung und KUnik der visuellen Sensorik. BOcherei des Augenarztes, Volume 98. Edited by V. Herzau. Stuttgart: Ferdinand Enke, pp KOLUNG G (1986) Diagnostik und operative Korrektur von Vertikal- und Zyklodeviationen bei StSrungen schrsger Augenmuskeln. Dosis-Wirkungsbeziehung verschiedener Eingriffe. Habilitan'onsschrift, Giessen, pp. 8-11, 57-59, MATIN L (1972) Eye movements and perceived visual direction. In: Handbook of Sensory Physiology, Volume VII/4: Visual psychophysics. Edited by D. Jameson and L. M. Hurvich. Berlin: Springer-Verlag, pp OUVIER P, VON NOORDEN GK (1982) Excyclotropia of the nonparetic eye in unilateral superior oblique muscle paralysis. American Journal of Ophthalmology, 93, RUTTUM M, VON NOORDEN GK (1983) Adaptation to tilting of the visual environment in cyclotropia. American Journal of Ophthalmology, 96, Received January 6, Revised April 12, Accepted May 28, 1993