Journal of Neuro-Ophthalmology, December 2001, Volume 21, Issue 4

Vertical Saccades in Superior Oblique Palsy and Browns Syndrome

OBJECTIVE: To compare saccadic dynamics in superior oblique palsy and Brown's syndrome. METHODS: Vertical saccades in adduction and in abduction were studied in two subjects with superior oblique palsy and one with Brown's syndrome. Using large numbers of centrifugal saccades over a wide range of amplitudes, we measured peak velocity, duration, and the peak velocity/mean velocity ratio (PV/MV) as a function of saccadic amplitude. We compared vertical saccades in 30 degrees of abduction with those in 30 degrees of adduction. RESULTS: Superior oblique palsy caused a 15-18% reduction in peak velocities in adduction compared with abduction. Saccadic duration was also increased in adduction, with the result that there was no net change in the PV/MV ratio. In the patient with Brown's syndrome, velocities and durations of upward saccades were similar in abduction and adduction, but the PV/MV ratio was significantly elevated in adduction. We also observed an unusual high-speed lateral 'snap' of about 5 degrees that frequently interrupted vertical saccades in the midline but not elsewhere. CONCLUSION: Both paresis and restriction of the superior oblique alter vertical saccades. The effects of restriction on saccadic dynamics are distinct from the effects of paresis.

Journal of Neuro- Ophthalmology 21( 4): 250- 255, 2001. • 2001 Lippincott Williams & Wilkins, Inc., Philadelphia Original Contribution Vertical Saccades in Superior Oblique Palsy and Brown's Syndrome Jason J. S. Barton, MD, PhD, FRCPC, and James M. Intriligator, PhD Objective: To compare saccadic dynamics in superior oblique palsy and Brown's syndrome. Methods: Vertical saccades in adduction and in abduction were studied in two subjects with superior oblique palsy and one with Brown's syndrome. Using large numbers of centrifu­gal saccades over a wide range of amplitudes, we measured peak velocity, duration, and the peak velocity/ mean velocity ratio ( PV/ MV) as a function of saccadic amplitude. We com­pared vertical saccades in 30 degrees of abduction with those in 30 degrees of adduction. Results: Superior oblique palsy caused a 15- 18% reduction in peak velocities in adduction compared with abduction. Sac­cadic duration was also increased in adduction, with the result that there was no net change in the PV/ MV ratio. In the patient with Brown's syndrome, velocities and durations of upward saccades were similar in abduction and adduction, but the PV/ MV ratio was significantly elevated in adduction. We also observed an unusual high- speed lateral ' snap' of about 5 de­grees that frequently interrupted vertical saccades in the mid­line but not elsewhere. Conclusion: Both paresis and restriction of the superior ob­lique alter vertical saccades. The effects of restriction on sac­cadic dynamics are distinct from the effects of paresis. Key Words: Superior oblique- Brown's syndrome- Saccades- Restriction. The contributions of the oblique muscles of the eye to rotation of the globe are complex. Though their primary actions are cyclotorsion of the globe, ( excyclotorsion in the case of the inferior oblique and incyclotorsion for the superior oblique) ( 1), they also have secondary vertical actions, elevation for the inferior oblique and depression From the Human Vision and Eye Movement Laboratory, Depart­ments of Neurology and Ophthalmology, Beth Israel Deaconess Medical Center, Harvard Medical School; Department of Biomedical Engineering, Boston University, Boston, Massachusetts. Address correspondence and reprint requests to lason I. S. Barton, Neurology, KS 452, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, Boston, MA 02215; E- mail: jbarton@ caregroup. harvard. edu for the superior oblique. The magnitudes of their primary and secondary actions are dependent upon horizontal di­rection of gaze: theoretically, their torsional effect should be maximal at 60 degrees of abduction, and their vertical contribution maximal when the eye is adducted 30 de­grees ( 1). This variation can be understood from consid­ering the effective line of pull of these muscles on the eye, which is towards the anteromedial corner of the orbit, where the origin of the inferior oblique and the trochlea are located. The effects of superior oblique paresis on static eye position are predictable from the three actions of the muscle ( 2): excyclotorsion, hypertropia of the affected eye, which is worse in adduction and ipsilateral head tilt, and a small esotropia often worse in depression ( V-pattern). Conversely, superior oblique restriction, also known as Brown's syndrome, limits the ability of the eye to elevate in adduction, sometimes with hypotropia of the affected eye in primary position. The impact of oblique muscle dysfunction on ocular motor dynamics is less well known. In particular, their impact upon vertical saccades is controversial. Early ex­perimental studies with novocaine injections found a small reduction in saccadic velocities ( 3). An electro-oculographic study appeared to confirm this by finding a mean 25% reduction of downward saccades in adduction in 8 patients with superior oblique palsy ( 4). However, another electro- oculographic study failed to find a dif­ference in 18 patients ( 5). Citing potential difficulties in measuring vertical eye movements with electro-oculography, a more detailed study used the magnetic search coil technique to study 10 patients with superior oblique palsy and found no difference between the pa­tient and the control groups ( 6). Conflicting findings were also found in a few patients studied before and after superior oblique tenotomy ( 4,5). In interpreting these findings, it is important to note that the two negative studies reported primarily group means and group statistics. Given that the dominant de­pressor in all positions of gaze is the inferior rectus, and 250 VERTICAL SACCADES IN SUPERIOR OBLIQUE PALSY AND BROWN'S SYNDROME 251 that the theoretical maximal contribution of the superior oblique to depression is 18%( 7), the effect on downward saccadic velocity of even complete paresis of this muscle is bound to be small. A group study is likely to include patients with palsies of varying severity, and with the wide variability in normal saccadic metrics, it is hardly surprising that small group analyses should fail to find a difference. While these negative studies ( 5,6) suggest that vertical saccadic velocity is not a sensitive test for superior oblique palsy, a group analysis cannot answer the more specific question of whether the superior ob­lique makes any contribution to vertical saccades. To address this, only patients with severe superior oblique palsy should be studied individually. In support, we note that both negative studies mentioned a few individuals whose downward velocities were impaired. We performed detailed saccadic evaluation on two subjects with superior oblique palsy to address the ques­tion of the contribution of this muscle to vertical sac­cades. In addition, we studied one subject with a con­genital Brown's syndrome, for which there is even less data on saccadic behavior. METHODS Subjects MR is a 51- year- old woman noted to have a right superior oblique palsy, presumed congenital given her denial of diplopia. She had a 4- diopter right hypertropia in primary position, which increased to more than 25 diopters in left gaze and tolO diopters with right head tilt. There was excyclotropia of 5 degrees. Her vertical fu-sional amplitudes were large at 14 diopters. Other ocular motility was normal. MS is a 40- year- old man who sustained a severe closed head injury and a right fourth nerve palsy in a motor vehicle accident 8 months prior to his visit. In primary position he had a 10 diopter right hypertropia and 10- diopter esotropia. The hypertropia increased to 25 diopters in left gaze and to 18 diopters in right head tilt. In downgaze, esotropia increased to 18 diopters and right hypertropia to 20 diopters. He did not have a left hyper­tropia in any position. He had 5 degrees of relative ex­cyclotropia. He also had frequent square wave jerks and diffusely reduced pursuit gain, possibly on the basis of inattention. He had mild left hemiparesis and left infraor­bital numbness. ES is a 26- year- old medical student with left Brown's syndrome diagnosed in childhood. She had moderately limited elevation of the OS, worse in adduction. She was orthotropic in primary position, and had a small right hypertropia in upgaze in the midline, which in­creased during rightward upgaze, but which switched to a small left hypertropia in leftward upgaze, suggesting bilateral restriction. Two additional control subjects, age 31 and 42, with clinically normal eye movements were studied. All subjects gave informed consent according to a 500- 2; 4oo-velocity O o i •^ 200- Q* 100- 0- 1 1 T yS^ T^ J. - e- in abduction • in adduction 1 , , ' 150 amplitude (°) amplitude (°) amplitude (") FIG. 1. Control subjects: Downward saccadic metrics. Left graphs show peak velocity, center graphs show duration, and right graphs show PV/ MV ratios. Top graphs are from control subject 1, bottom graphs are from control subject 2. All are plotted against amplitude. Saccades in an abducted position are shown as clear circles, saccades in an adducted position as black circles. There are no consistent differences in any of the three relationships in these controls. ( As in all figures, error bars indicate one standard deviation, and asterisks indicate significant P values for individual bins: *. 01, **. 005, ***. 001 ,****. 0001, *****. 00001.) J Neuro- Ophthalmol, Vol. 21, No. 4, 2001 252 U. S. BARTON and J. M. INTRILIGATOR protocol approved by the hospital's committee on clini­cal investigations. Procedure and apparatus Eye movements were recorded with the magnetic search coil technique, using 3- foot search coils ( Crist Instruments, Baltimore, MD). Vertical and horizontal eye positions were sampled 500 times a second and ve­locity derived using a 7- point differentiation algorithm. The subjects sat in a chair facing a tangent screen upon which images were back- projected by an Eiki LC- 7000U liquid crystal display projector. Subjects with abnormali­ties viewed the screen with their unaffected eyes patched. Stimuli, data collection, and data analyses were per­formed with a Power Macintosh 9600/ 233 computer, us­ing the Vision Shell programming platform ( MicroML, St Hyacinthe, Quebec). The target was a small white annulus on a dark back­ground. This was positioned along the horizontal merid­ian, either at screen center, 30 degrees left, or 30 degrees right. After an unpredictable interval of up to 2 seconds, the target jumped vertically between 5 and 40 degrees, in 5- degree steps and in random order. Only one direction was tested at a time, downward for superior oblique pal­sies and upward for Brown's syndrome. The goal was to generate a large array of centrifugal saccades spanning as complete a range of amplitudes as possible. Saccades in the midline, in adduction and in abduction were per­formed in blocks. 120 trials were performed in each block, generating over 200 saccades for each of the three horizontal gaze positions. Analysis All prior studies compared upward to downward sac­cades in their primary analysis. However, given the anatomy of the superior oblique, it is more logical to compare centrifugal saccades in adduction, versus those in abduction. We divided downward centrifugal saccades into amplitude bins of 5 degrees. Upward centrifugal saccades were divided into 4- degree amplitude bins, be­cause of a smaller amplitude range. Our three primary variables were Peak velocity ( PV), duration ( D), and the peak velocity/ mean velocity ( PV/ MV) ratio. This ratio, also formulated as ( PVD)/ A, expresses the relation be­tween all three variables and is very uniform in normal subjects, around 1.6 ( 8). We plotted these variables as functions of amplitude. ANOVA was used for each sub­ject's parameters to determine if there was a significant effect of lateral position on vertical saccade performance. We also used t tests to compare data across bins between saccades in adduction and those in abduction. Because of the use of multiple t tests with the bins, we set a higher cut- off for significance, at P < 0.01. RESULTS In the two normal subjects, there was no significant difference between saccades in adduction and those in abduction in any variable, for upward or downward saccades. The only exception was that the one subject had slightly faster upward saccades in adduction. This lack of difference for velocity is consistent with prior reports ( 4,6) ( Fig. 1, 3). amplitude (°) 0 10 20 amplitude (°) amplitude (°) FIG. 2. Superior oblique palsy: downward saccadic metrics. Conventions as in Figure 1. Top graphs are from patient MR, bottom graphs are from patient MS. Both patients have a consistent reduction in peak velocity and increase in duration for saccades in an adducted position. / Neuro- Ophthalmol, Vol. 21, No. 4, 2001 VERTICAL SACCADES IN SUPERIOR OBLIQUE PALSY AND BROWN'S SYNDROME 253 > 200- 1- 0 5 10 15 amplitude (°) - r - FIG. 3. Controls: Upward saccadic metrics. Conventions as in Figure 1. Control subject 1 had slightly increased peak velocities in adduction than did control subject 2. No other relations were significantly different. Superior oblique paresis caused a significant reduction in peak velocity for downward saccades in adduction ( MR: P < 0.02, MS: P < 0.01) ( Fig. 2). In both MR and MS, this occurred mainly for saccades between 10 and 25 degrees in amplitude, where there was an average reduction of 18.6% in MR and 15% in MS. Saccadic duration was also significantly increased over this range ( MR: P < 0.06, MS: P < 0.03). As a result, the PV/ MV ratio was unaltered, indicating that the reduced force in generating peak velocity was reflected in a general and proportionate reduction in velocity across the en­tire saccade. The data for superior oblique restriction was very dif­ferent. Peak velocity did not differ between adducted and abducted positions for ES. However, in adduction the upward saccadic durations were increased in ES, leading to a striking elevation of the PV/ MV ratio ( P<. 02) at almost all amplitudes ( Fig. 4). ES's data also revealed an unusual abnormality in sac­cadic trajectory, specific for saccades in the midline ( Fig. 5). Centrifugal upward saccades, and to a lesser extent, centripetal upward and downward saccades, were often transiently interrupted by a slight reversal and simulta­neous lateral shift of the eye, before resumption of their trajectory in a horizontally displaced position. Upward saccades were always shifted laterally, whereas down­ward saccades were always shifted medially. Prior to this dramatic shift, upward saccades were also found to have drifted slightly laterally. The peak velocities of the lateral component of these disruptive shifts were very rapid, 2: 200- velocity 3100- cu o- l ES rl/ i * & ^ rr. - e- in abduction -#- in adduction 1 r - 1 8 12 16 amplitude (°) 20 12 amplitude (°) 12 16 amplitude (°) FIG. 4. Brown's syndrome: upward saccadic metrics. Conventions as in Figure 1. There are trends to slightly faster and longer saccades in adduction. The net result is a highly significant increase in the PV/ MV ratio. J Neuro- Ophthalmol, Vol. 21, No. 4, 2001 254 U. S. BARTON and J. M. INTRILIGATOR 820 870 920 time ( ms) 970 7 - 6 - 5 - 4 - 3 - 2 - 1 0 1 horizontal eye position (°) i o - 16- £ 1 4 - | 12- g. 10- S1 8" CO /- " O 0 - t 2 4- 2^ "•• start 1111| 1111| i ii i| 11 end i i] i i n | I I i i| in D M 1800 1850 time ( ms) 1900 - 3 - 2 - 1 0 1 2 3 horizontal eye position (°) FIG. 5. Brown's syndrome: the lateral flip effect. Top left graph shows horizontal and vertical positions plotted against time for an upward saccade; bottom left graph shows the same for a downward saccade. Top right graph shows plots of horizontal versus vertical eye position for the same upward saccade, revealing the spatial trajectory of the eye. The bottom right graph shows the same for the same downward saccades. ranging from 400 to 700 degrees per second ( two to four times the normal velocity of about 2007s for 5- degree saccades ( 9). These transient shifts were not tied to a specific orbital position. DISCUSSION Our data show that oblique muscle dysfunction can affect vertical saccadic velocities. Moreover, the effects of paresis are distinct from the effects of restriction. Su­perior oblique paresis reduced downward saccadic ve­locities. Upward peak velocities were not affected by restriction. In the patient with congenital Brown's syn­drome, restriction caused a disproportionate increase in upward saccadic duration, and hence a lowering of mean velocity relative to peak velocity. A decrease in saccadic peak velocity is typical of many types of ocular motor paresis, both neuropathic and myopathic ( 10,11). The theoretical maximum contribu­tion of the superior oblique to vertical eye movement is 18% ( 7), which should be achieved when the eye is adducted 30 degrees ( 1). Our finding that patients with severe superior oblique palsy have a 15- 18% difference in the peak velocity of downward saccades performed in 30 degrees of abduction versus adduction fits this esti­mate precisely. There are few studies of the effects of restrictive ocu­lar motor disease on saccadic dynamics. Most investiga­tors have examined Graves' ophthalmopathy and have reported normal saccadic peak velocities, except in se­vere disease ( 12- 15). One study did find reduced sac­cadic peak velocities ( 16). Graves' disease is compli­cated, however, by the fact that it may affect extraocular muscle function with a combination of restriction and paresis ( 17). Nevertheless, most studies are consistent with our finding of normal peak velocities in Brown's syndrome. Whereas some studies also comment on sac­cadic duration in Graves' disease, none relates duration to peak velocity as we did with the PV/ MV ratio. The PV/ MV ratio found in ES indicates that restriction caused a subtle but disproportionate increase in duration, / Neuro- Ophthalmol, Vol. 21, No. 4, 2001 VERTICAL SACCADES IN SUPERIOR OBLIQUE PALSY AND BROWN'S SYNDROME 255 implying, for a given amplitude and peak velocity, a decrease in mean velocity. This finding has implications for our concepts of muscle restriction. One type of restriction may be a rigid obstruction to eye movement beyond a certain orbital position. One might expect that, after relatively unhin­dered motion to that point, a saccade would then be abruptly terminated. Peak velocity, which is usually achieved early in the course of a saccade ( 8), may not be affected, but amplitude and duration would be relatively reduced. Supernormal peak velocities would be the chief result. Because velocity would still be relatively high at the time of the sudden termination, it is also probable that mean velocity would remain relatively high, and the PV/ MV ratio would be unaltered or even lowered for a given amplitude. Also, small saccades may not be af­fected if they do not cause the eye to reach the limiting orbital position. On the other hand, with a more elastic restriction, passive resistive forces would gradually escalate during a saccade as the eye reached positions of greater eccen­tricity. Again, the early peak of velocity would not be affected, but the gradual braking effect of the elastic forces would reduce terminal velocity and prolong the final phase of the saccade. The result would be a reduc­tion in mean velocity relative to peak velocity, as we found in ES. Our patient with Brown's syndrome had unusual ' horizontal flips' during vertical saccades. These were lateral during upgaze and medial during downgaze, and were accompanied by transient interruptions in the ver­tical trajectory. The speed of the horizontal movements was too fast for voluntary saccades. These movements are consistent with a sudden passive release from resis­tance. This finding might be analogous to the vertical shifts in Duane's syndrome, where co- contraction of the horizontal recti causes the eye to ' flip' vertically during horizontal movements, mainly in a plane above the me­ridian ( 18,19). Here the restriction of the superior ob­lique combined with the contraction of the inferior ob­lique may generate a similar ' bridle effect' when the eye is moving in a vertical plane just lateral to the plane of maximal vertical function of the obliques. However, this would not explain the equally rapid medial devia­tions during downgaze. Another possibility is that an upward saccade stretches taut a short superior oblique tendon, until the tendon is nearly tangential to the globe and can slip around it with an abducting jerk. With a downward saccade from upgaze, the tendon would also have to slip back, with the eye rolling medially. Dynamic orbital MRI or ultrasound may help explore this issue in future patients. REFERENCES 1. Jampel R. Ocular torsion and the function of the vertical extraocu­lar muscles. Am J Ophthalmol 1975; 79: 292- 304. 2. Parks M. Isolated cyclovertical muscle palsy. Arch Ophthalmol 1958; 60: 1027- 1035. 3. Scott A. Strabismus: Muscle forces and innervations. In: Lenner-strand G, Bach- y- Rita P, ed. Basic mechanisms of ocular motility and their clinical implications. Elmsford, NY: Pergamon Press, 1975: 181- 91. 4. Rosenbaum A, Carlson M, Gaffney R. Vertical saccadic velocity determination in superior oblique palsy. Arch Ophthalmol 1977; 95: 821- 823. 5. Metz H. Saccadic velocity studies in superior oblique palsy. Arch Ophthalmol 1984; 102: 721- 722. 6. Stathacopoulos R, Yee R, Bateman J. Vertical saccades in superior oblique palsy. Invest Ophthalmol Visual Sci 1991; 32: 1938- 1943. 7. Boeder P. The co- operation of extraocular muscles. Am J Oph­thalmol 1961; 51: 469- 481. 8. van Opstal A, van Gisbergen J. Skewness of saccadic velocity profiles: a unifying parameter for normal and slow saccades. Vi­sion Res 1987; 27: 731- 745. 9. Becker W. Metrics. In: Wurtz R, Goldberg M, ed. The neurobiol­ogy of saccadic eye movements. Amsterdam: Elsevier, 1989: 13- 68. 10. Yee R, Whitcup S, Williams I, Baloh R, Honrubia V. Saccadic eye movements in myasthenia gravis. Ophthalmology 1987; 94: 219- 225. 11. Baloh R, Konrad H, Sills A, Honrubia V. The saccade velocity test. Neurology 1975; 25: 1071- 1076. 12. Metz H. Saccadic velocity studies in patients with endocrine ocular disease. Am J Ophthalmol 1977; 84: 695- 699. 13. Feldon S, Unsold R. Graves' ophthalmopathy evaluated by infra­red eye- movement recordings. Arch Ophthalmol 1982; 100: 324- 328. 14. Feldon S, Levin L, Saundra K, Liu M. Graves' ophthalmopathy: correlation of saccadic eye movements with age, presence of optic neuropathy, and extra- ocular muscle volume. Arch Ophthalmol 1990; 108: 1568- 1571. 15. Mauri L, Meienberg O, Roth E, Konig M. Evaluation of endocrine ophthalmopathy with saccadic eye movements. J Neurol 1984; 231: 182- 187. 16. Wouters R, van den Bosch W, Lemij H. Saccadic eye movements in Graves' disease. Invest Ophthalmol Visual Sci 1998; 39: 1544- 1550. 17. Hermann J. Paretic thyroid myopathy. Ophthalmology 1983; 89: 473- 478. 18. Rogers GL, Bremer DL. Surgical treatment of the upshoot and downshoot in Duanes retraction syndrome. Ann Ophthalmol 1984; 16( 9): 841- 844. 19. von Noorden GK, Murray E. Up- and downshoot in Duane's re­traction syndrome. J Pediatr Ophthalmol Strabismus 1986; 23( 5): 212- 215. J Neuro- Ophthalmol, Vol. 21, No. 4, 2001