Examine eye movements

As with all aspects of the neurological examination, important clues come from a thorough and appropriate history. In relation to eye movement disorders the patients may be complaining of double vision, in which case they should be asked whether it is constant or intermittent; does it occur, or is it maximal, in certain directions of gaze; what is the relationship of one image with the other; and have they tried covering one eye and did that relieve the symptom? A less frequently reported symptom is oscillopsia, an illusion of movement of stationary objects, when enquiries need to be made whether the movement is horizontal or vertical, and does it become maximally apparent in certain positions of gaze, as for example in downbeat nystagmus when the oscillopsia is maximal on down gaze.



The first part of any examination is observation and again this may provide clues to the diagnosis. Look out for abnormal head postures, turns or tilts, the latter typically occurring in an isolated trochlear nerve palsy; abnormal patterns of eye–head coordination, such as the head thrusts seen in ocular motor apraxia; abnormalities of the eyelids such as ptosis in an oculomotor palsy and myasthenia gravis, or retraction as is seen in thyroid ophthalmopathy and progressive supranuclear palsy.

STATIC EYE MOVEMENTS
We are now ready to examine the eye movements, hopefully with the benefit of some clues to the diagnosis already obtained from the history and simple observation of the patient.
Ocular misalignmentFirst ask the patient to look at a distant object and observe any obvious ocular misalignment or abnormal, spontaneous eye movements. If one eye is deviated inward relative to the other this is referred to as an esotropia, whereas if there is an outward deviation this is called an exotropia. If there is a vertical misalignment it is described in relation to the laterality of the higher eye (for example, a right hypertropia indicates that the right eye is higher than the left).
Monocular or binocular diplopia?If the patient is complaining of diplopia it is essential to next differentiate a disparity in retinal stimulation between the two eyes (binocular diplopia), from the rarer form of diplopia which is present in one eye only (monocular diplopia). This is simply done by covering one eye and then the other to see if the diplopia disappears or persists in one eye only. Monocular diplopia occurs, with few exceptions, when there are abnormalities of the ocular refractive surfaces and media, producing multiple overlapping images on the retina. It is usually abolished or improved by getting the patient to observe a target through a pinhole. The commonest cause is myopic astigmatism, but monocular diplopia may occur with early cataracts, especially under conditions of dim illumination. Other causes include abnormalities of the cornea and iris, foreign bodies in the aqueous or vitreous humour, and retinal disease. Rarely, when monocular diplopia occurs in both eyes after each is covered in turn, it may be due to occipital cortex pathology, or there can be multiple images (cerebral polyopia). If the monocular diplopia does not disappear or improve with a pinhole it is either due to one of the ocular refractive problems described above, or it is psychogenic in origin, the cerebral causes being exceedingly rare.
If we take the commonest clinical scenario of binocular diplopia, a systematic approach to evaluation is required. As well as determining the nature of the separation of the two images and the direction of maximal separation, enquiries about any family history of strabismus (squint), or a childhood history of orthoptic treatment should be made.
We next need to find out which muscle is weak. To do this the range and speed of movement of the two eyes are examined, first with ductions (one eye viewing) and then with versions (both eyes viewing) in the nine cardinal positions of gaze (up and left, up, up and right, left, straight ahead (primary position), right, down left, down, down right). Any limitation of ocular movement when examining ductions is recorded as a percentage of the normal range and if there are any differences on versions these are noted.
If the eyes are clearly misaligned at rest or on movement it should be ascertained at an early stage if one is dealing with a non-comitant or comitant strabismus; the degree of misalignment varies with gaze position in the first, but not in the second:
Non-comitance suggests a recent paretic or restrictive aetiology (occurring when an ocular muscle is unable to move the globe because its antagonistic muscle fails to relax—for example in thyroid ophthalmopathy, when the inferior rectus may become fibrosed and restricts upward gaze).
Comitance is characteristic of childhood strabismus, and diplopia in such circumstances is usually due to decompensation of a long-standing phoria (a deviation of the visual axes which occurs only when binocularity is interrupted), normally kept in check by fusional mechanisms (a latent deviation). This may sometimes be observed after relatively minor head trauma.
The term tropia refers to a deviation of the visual axes when both eyes are viewing, which is not kept in check by fusion and is present at all times (a manifest deviation).
To sort out these aspects of ocular misalignment the cover/uncover and alternate cover tests are used:
In the cover/uncover test, used to detect tropias, the patient, wearing appropriate refractive correction, is asked to fixate a distant target (for example, a letter on the Snellen chart) with the eyes in the primary position (fig 1). One eye is covered and any immediate movement of the uncovered eye is observed. Cover tests rely on the fact that foveation occurs in an eye that is forced to fixate; if the retinal image was not directed on to the fovea before the eye took up fixation, what is known as a movement of redress will be noted as the uncovered eye now fixates the target. This gives an indication of the degree of misalignment of the visual axes when binocular viewing took place. If the uncovered eye moves to take up fixation a manifest deviation is present (a tropia). Inward movement of the uncovered eye indicates an exotropia (the eye had previously been outwardly deviated), and an outward movement an esotropia (the eye had previously been inwardly deviated). A vertical deviation may be either a hypotropia or a hypertropia, depending on whether the eye moves up or down, respectively. It should be noted that the convention is that if there is a vertical deviation of the eyes, the higher of the two is referred to as hypertropic/hyperphoric, regardless of which eye is at fault.





Figure 1 (A) The cover/uncover test, showing an
esophoria. Grey bars indicate the position of the eye when under cover. (i) At
rest the visual axes are aligned correctly. (ii) When the cover is placed before
the left eye, the eye no longer fixates and moves inwards. (iii) On removal of
the cover the eye moves outwards to take up fixation, indicating an esophoria.
(B) The alternate cover test, showing an esotropia. (i) At rest, with both eyes
viewing, there is a manifest inward deviation of the left eye. (ii) A cover
placed before the non-fixating left eye causes no movement. (iii) When the right
eye is occluded, the left eye is forced to fixate, and a movement of redress
occurs (primary deviation). The resulting additional innervation to the
contralateral muscle leads to deviation of the sound eye under the cover (the
secondary deviation). Note that the secondary deviation is greater than the
primary deviation. (iv) When the cover is transferred to the left eye both eyes
assume their original position.



If a tropia is present the examiner should next determine whether it is comitant or non-comitant by seeing if the magnitude of the deviation varies with the position of the eye.
If no tropia is present in the uncovered eye then the examiner’s attention moves to the covered eye. As the occluder is removed the eye is observed to determine whether it has moved to assume fixation just after it was uncovered, referred to as a latent deviation (a heterophoria). Depending on the direction of the deviation this may be classified as an exophoria, esophoria, hypophoria or a hyperphoria. The test is then repeated, and the same observations made while covering the other eye.
The alternate cover test is more dissociating than the cover/uncover test (because there is no possibility for binocular fusion which it effectively breaks) to show the maximal deviation—any tropia plus any latent phoria. While the patient fixates a target the occluder is quickly switched from eye to eye to prevent binocular viewing, allowing sufficient time for the eyes to settle in their new position after each transfer. The test should be performed in the nine cardinal positions of gaze to determine the direction of gaze that elicits the maximal shift of eye position on switching the occluder. While ensuring that the patient is never allowed to regain fixation during transfer of the occluder, the examiner notes the movement of the newly uncovered eye as the occluder is transferred from one eye to the other. Movement of the uncovered eye may indicate either a heterotropia or a heterophoria, but the alternate cover test will not differentiate between the two. The cover/uncover test should therefore be performed before the cover/uncover test, to determine if a tropia is present.
Having completed the cover tests, further examination of versions to determine the paretic muscle is best achieved using a bright pen torch with a red glass or piece of perspex (obtained from a pair of red-green glasses found in many toy shops!) placed in front of the right eye. The patient now sees a white and a red image if there is displacement of the images and can describe their location in relation to each other:
If the displacement is horizontal in the primary position then the position generating the maximum separation is next identified. With the knowledge that the outer of the two images derives from the paretic muscle and only the lateral rectus and medial rectus muscles are involved in horizontal movements, the paretic muscle can be determined.
If the patient complains of vertical or oblique diplopia ascertain whether the red image is higher or lower. The hypertropic eye always produces the lower image. Therefore, if the patient has a left hypertropia the involved muscles are either the elevators of the right eye (superior rectus and inferior oblique muscles), or the depressors of the left eye (inferior rectus and superior oblique muscles) and conversely for right hypertropia. The number of possibilities can be further reduced by ascertaining whether the misalignment is greater on right or left gaze—the recti muscles being responsible for vertical movements in the abducted position, and the obliques in the adducted position of gaze. Finally, determining which of these two muscles is paretic is decided by finding whether the deviation is maximal in up or down gaze.
If the patient complains of an oblique vertical separation of images then a simple test for weakness of the superior oblique muscle, as in a trochlear nerve palsy, is to hold a pencil or ruler horizontally in front of the patient and ask him or her to look at the middle of the object as it is slowly lowered. If the patient experiences vertical separation of the images, which are oblique to one another forming a V shape, the point of the V is directed towards the side of the weakened superior oblique muscle.
In some instances when performing these tests there may be no differential vertical deviation; this situation occurs with chronic palsies due to an adaptive phenomenon, termed "spread of comitance". To work out which vertical muscle is paretic the Bielschowsky head tilt test is performed; the vertical deviation is compared with the alternate cover test in right and left head tilt positions (fig 2). The degree of misalignment will increase when the head is tilted to the side of the paretic muscle if the ipsilateral intorters (superior oblique and superior rectus) are weak, and to the opposite side if the extorting muscles (inferior oblique and inferior rectus) are weak. In practice an increased misalignment on head tilt is usually indicative of an ipsilateral superior oblique palsy. The test is less often positive with palsies of the vertical recti or inferior oblique muscles. The explanation for the effect lies in the fact that a head tilt to either shoulder induces an ocular counter-rolling, which is mediated by the ipsilateral intorters (superior rectus and superior oblique) and the contralateral extorters (inferior rectus and inferior oblique). If, for example, the ipsilateral superior oblique is paretic, the superior rectus on the same side receives excessive innervation to intort the eye, and by virtue of its relatively unopposed primary action elevates the eye.

Figure 2 The Bielschowsky test for vertical diplopia, illustrated for a right
fourth (trochlear) nerve palsy. See text for full explanation.
If more than one extraocular muscle is paretic it may be difficult to determine which one is involved and a more sophisticated test may be required such as a Hess chart, usually performed by an orthoptist.

DYNAMIC EYE MOVEMENTS
We now turn to assessing the more dynamic aspects of ocular motility. First, however, ask the patient to fixate on a stationary target and look for nystagmus or extraneous saccades such as square wave jerks.
SaccadesNext assess the patient’s ability to perform voluntary saccades by asking him or her to look to the left and right, and up and down. Disturbances in the initiation of saccades may lead to a prolonged latency, or the addition of a head movement or blink to initiate the saccade, the latter in congenital or acquired oculomotor apraxia, and various degenerative conditions such as Huntington’s disease. It should also be possible to gauge the normality of the saccadic velocity.
Then instruct the patient to rapidly look at a suddenly appearing object, for example a rapidly lifted finger when placing your hands on either side of, and 40–50 cm away from, the patient’s face. This tests for reflexive saccades and in addition to assessing the saccadic latency and velocity the saccadic accuracy can now be evaluated—saccadic overshoot (hypermetria) or undershoot (hypometria).
While assessing saccadic function also look out for evidence of a partial internuclear ophthalmoplegia in which there is slowed adduction on the side of a lesion of the medial longitudinal fasciculus, with abducting nystagmus in the other eye. Sometimes this may be difficult to pick up so get the patient to make saccades between obliquely placed targets. Because the velocity is slowed in the horizontal and not the vertical plane, the resulting saccade is L shaped.
Predictive saccades are tested by alternately raising a finger of one hand and then the other in a predictable and regular pattern. The patient is asked to make saccades back and forth to the moving finger. Normally after a few saccades they anticipate the appearance of the stimulus and make a saccade in advance. Abnormalities are observed in neurodegenerative disorders such as Parkinson’s disease.
Any slowing of saccades can be accentuated by using an optokinetic striped drum, or measuring tape, when the repositioning saccades will appear clearly slowed.
Smooth pursuitSmooth pursuit can be tested by asking the patient to track a small moving target, such as the head of a hat pin, at a distance of about 1 metre, while keeping their head stationary. Both horizontal and vertical smooth pursuit should be assessed. The target should be moved initially at a slow uniform speed and the pursuit eye movements observed to determine whether they are smooth, or broken up by catch-up saccades. This is a non-specific sign when present in both directions—for example, it may be due to ageing or cerebellar disease—or it may indicate a focal posterior cortical lesion if only present in one direction, in which case the abnormal pursuit is in the direction of the lesion. The speed should be gradually increased, but at high velocities (>50 degrees per second) all smooth pursuit eye movements will be broken up by saccades even in normal people.
Despite their relevance for locomotion and social interaction in everyday situations, little is known about the cortical control of vertical saccades in humans. Results from mi-crostimulation studies indicate that both frontal eye fields (FEF) contribute to these eye movements. Patients with a damaged right FEF, who hardly made ver-tical saccades during visual exploration. This finding suggests that for the cortical control of exploratory vertical saccades, integrity of both FEF is indeed important.

PRACTICE POINTS
Eye movements should be examined in two stages, first when the eyes are static in different positions of gaze, and then dynamically.
Static eye movement examination, particularly when the patient complains of diplopia, requires the correct use and interpretation of the cover/uncover and the alternate cover tests.
Testing of dynamic eye movements involves examination of rapid conjugate eye movements, saccades, and smooth pursuit both horizontally and vertically.

Further reading
C Kennard. Normal and abnormal eye movements. In: Luxon L, ed. A textbook of audiological medicine: clinical aspects of hearing and balance. London: Martin Dunitz, 2003:781–96.
Leigh RJ, Zee DS. The neurology of eye movements. Fourth edition. Oxford University Press, 2006.
Shaunak S, O’Sullivan E, Kennard C. Eye movements. In: Hughes RAC, ed. Neurological investigations. London: BMJ Publishing Group, 1997:253–82.

Optokinetic nystagmus
The optokinetic system is often tested as part of the clinical examination using the striped optokinetic drum or a measuring tape moved in front of the eyes. However, because these occupy a small area of the visual field they in fact test smooth pursuit and not the optokinetic system. A full field revolving striped drum is required to elicit true optokinetic nystagmus. Reduced optokinetic nystagmus occurs in visual disorders, in pursuit system disorders, and in disorders of fast phases (saccades).

Vestibular system
If the vestibulo-ocular system is functioning, normally passive rotation of the patient’s head with the patient instructed to look straight ahead should result in a slow eye movement so that the eyes move in the opposite direction to that of the head movement. This is known as the doll’s head (oculocephalic) manoeuvre and should be performed both horizontally and vertically. This technique is not only valuable for assessing vestibular function (space does not allow a full coverage of its assessment), but in the current context for differentiating infranuclear and nuclear gaze palsies when the response is absent, from supranuclear gaze palsies in which a normal doll’s head response is present.

CONCLUSIONS
Examination of eye movements can usually be completed fairly quickly but all types of eye movement should be evaluated because they have different anatomical bases and therefore can provide clues to the possible location of lesions.