<<

(1989) 3, 816--832

Treacher Collins Prize Essay The Significance of

NICHOLAS EVANS Norwich

Introduction combined. The range of forms it takes, and found the term v!to"[

Correspondence to: Mr. N Evans, FRCS, FCOphth. THE SIGNIFICANCE OF NYSTAGMUS 817 provides the major contribution to the control ing head movement), is converted by this of eye position, and eye position determines 'neural integrator' into positional infor­ visual input. Division of nystagmus into sen­ mation, which is fed into the oculomotor sory and motor types is an oversimplification, nucleii to produce eye movements. The inte­ and it is necessary to examine the interface grator is located in the brainstem reticular for­ between the sensory and motor visual mation near to the oculomotor nucleii, and systems, nystagmus having consequences for has been placed in either the prepositus and both. The component eye movements of nys­ vestibular nucleiV or in the burst cells of the tagmus will be described, and the nature of PPRF themselves.4 The output of such an normal steady fixation and its control identi­ integrator intrinsically depends on its func­ fied. The circumstances in which nystagmus tional integrity, and it can be analysed accord­ occurs will then be considered, particularly its ing to theories of system-control. variation in relation to different aetiology. Leak and gain are integrator properties of Information will be drawn from neurophysi­ particular importance in relation to nystag­ ological and neuroanatomic descriptions of mus. Leak denotes a loss of signal in the trans­ normal eye movements and pathways, clinical formation from velocity to position, and the reports of nystagmus and visual disorders, output of a leaky integrator shows decay­ and human and animal laboratory studies. In that is the eyes cannot hold an intended new drawing together information from these eccentric position, and drift back towards the sources to understand nystagmus, the func­ primary position. This decay shows an expo­ tion of the normal visual and visumotor nentially-decreasing time constant, the elastic system is also illuminated. recoil of the orbital tissues reducing progres­ sively as the eyes approach the midline. A Eye Movements and the Components of slow movement whose drift from fixation Nystagmus shows exponentially increasing velocity, on The oculomotor system has two kinds of the other hand, implies integrator instability active movement at its disposal, from which with high gain; that is the SEM generator all reflexand voluntary fixating,repositioning drives the eyes in the direction of the target, and following movements are synthesised: the under positive-feedback, at increasing veloc­ fast (FEM) and the slow (SEM) eye move­ ity, which can exceed that of the target. In ments (together with a distinct variant of the both these cases fixationrequires a correcting SEM system producing vergence move­ movement after the eyes have moved off-tar­ ments). FEMs, exemplified by , are get beyond a critical amount, and this is ballistic, that is the movement is prepro­ accomplished by a . If eccentric drift is grammed and not continuously monitored repeated, refoveating saccades will follow in a and controlled. Its peak velocity is propor­ cyclical pattern and nystagmus ensues. The tional to its amplitude, and vision is sup­ origin of nystagmus of any kind, then, lies pressed while it takes place. The SEM, with the slow . exemplifiedby the movement, Saccades are fed into the ocular motor is by contrast designed to maintain constant nucleii from the paramedian pontine reticular foveation of a target, and is under visually­ formation (PPRF) in the pons, whereas SEMs controlled monitoring. Its velocity is related are probably driven from the Purkinje cells in to the target velocity, up to a maximum. This the cerebellar .5 This difference relation between eye and target velocity reflects the greater coordination demanded describes the 'gain' characteristics of a supra­ by slow eye movements, which must take into nuclear unit which summates or 'integrates' account , head position and move­ all the inputs ·influencing eye movements ment. Since the two kinds of eye movement (frontal cortex, and secondary arise in separate anatomic locations, they can visual pathways, and pro­ be involved in disease processes independen­ prioceptors). Velocity-coded information tly of one another, and disorder in one which (from retinal image slip, burst cell-generated interferes with accurate foveation can be com­ saccades, and from vestibular neurons signall- pensated by a restoring action by the other. 818 NICHOLAS EVANS

Pathological nystagmus arises when these Steady fixationoccurs when eye-target rela­ events alternate in a cycle with appropriate tive movement is zero, and can be regarded as time-characteristics. a zero-velocity SEM, displaying the charac­ It may be stated as a general principle that teristic features of other SEMs. Smooth pur­ nystagmus of any sort arises in three circum­ suit or VOR intervenes without a break in stances: asymmetric input to the integrator, foveation if head or target should move. integrator output which is mismatched with Steady fixation requires that the FEM gener­ target movement by inappropriately high ators on each side are silent, equally sup­ gain, or integrator leak preventing a fixated pressed by pause cell discharge. The target from being held. In addition peripheral flocculus, which maintains the eye-field disorder (of motor nerve or extraocular relationship and continuous foveation, has muscle) gives rise to nystagmus if the output inputs from the visual and prefrontal cortex, of the integrator is mismatched with the res­ superior colliculus, substantia nigra and ves­ ponse of muscle yokes, normal gain being tibular nucleii, and, by intracerebellar con­ insufficient to maintain steady fixation in the nection, somatic and extraocular muscle face of progressive effector decay. These proprioceptors. Of all of these, only the classes of nystagmus are examined below in vision-mediated response is a closed loop relation to their clinical setting, and the pat­ system (in which the sensor and effector are terns of their component movements shown the same), the others being open loops, cap­ by eye movement recording. able of less precise control. Because the SEM Vertical eye movements and field organis­ system is the controller of steady fixation, ation differ from those in the horizontal plane SEM disorder or lateral bias result in fixation in their combination, rather than separation, instability, producing nystagmus if accom­ of lateralised elements of visual field,and ana­ panied by refixation saccades. Conditions tomically in their more rostral location in the producing nystagmus may be located in the midbrain. There is no evidence that they dif­ SEM generator itself, or more remotely in its fer neurophysiologically from horizontal descending inputs. Thus the physiological movements, though of course they are nystagmus elicited in response to an imposed involved in differentfocal lesions in the brain­ lateral field bias (optokinetic nystagmus, stem, and vertical nystagmus occurs in clinical OKN) reflects an asymmetric descending circumstances separate from horizontal visual input to the integrator, while nystagmus nystagmus. associated with acute unilateral vestibular disease is caused by unequal vestibular input, Steady Fixation and that associated with and Active SEMs are designed to eliminate move­ albinism may reflect deficiency of closed-loop ment of the eyes relative to their target. The visual feedback associated with defective target may move relative to the observer and image transmission. In the case of albinism a the world (signalled by retinal slip), or the further part may be played by the disordered observer move relative to the target and the separation of retinal images into hemifields world (signalled by vestibular input). In their caused by the abnormal crossing at the pure form these SEMs are then respectively chiasm.6 It is possible that a similar mech­ optokinetic smooth pursuit (eyes moving with anism operates in idiopathic congenital nys­ respect to earth), or a component of the ves­ tagmus,7 but evidence is difficult to collect, tibulo-ocular reflex (eye stationary with res­ and the idea remains speculative. Lateralised pect to earth). In both of these the eyes move lesions of the labyrinths, and those involving in the orbits to maintain a stable relationship the visual and accessory visual pathways in the with the visual field, eye movement velocity brainstem, and cerebellar disease, all produce being governed by the relative movement of nystagmus by impairing the steady fixation target and subject. Both require constant function of the SEM system. visual and vestibular input to the SEM gener­ Both input and integrator are adaptable to ator, the vestibular input being symmetric in altered visual, labyrinthine and neurological the absence of subject movement. conditions, and the visual input is dominant, THE SIGNIFICANCE OF NYSTAGMUS 819

as can be demonstrated by arranging experi­ The significance of an individual case of nys­ mental conditions such that visual and ves­ tagmus lies in the relation of its features to tibular demands conflict. Such circumstances their clinical setting. It can provide infor­ include the cancellation of vestibular-driven mation of diagnostic, localising and thera­ nystagmus following rotation, by visual fix­ peutic value, and is useful in the assessment of ation when rotation stops. This phenomenon visual function. It is important to bear in mind has clinical value in vision testing in the very the implications of nystagmus on vision in young infant, in whom persistence of the each case, since it is at this level that the dis­ VOR after rotation ceases indicates serious order has its impact on the subject. , the infant with normal vision suppressing VOR-mediated nystagmus Assessment and Classification of Nystagmus after one or two beats at the end of rotation. Nystagmus was formerly divided into pendu­ lar and jerk types, reflecting the relative Disconjugate eye movements symmetry or asymmetry of its vectors on the Because the postchiasmic sensory system is one hand, and the limitations of clinical obser­ organised according to field,and the supranu­ vation on the other. It is generally recognised clear eye movement centres produce conju­ now that such a classification has little diag­ gate deviations of both eyes, it might be nostic or nosologic value, and electronic tech­ supposed that the visual input is entirely field­ niques of eye recording have made possible coded, and that the motor system produces more detailed classifications according to equal movements of both eyes, in accordance wave form, which carry implications of mech­ with Hering's law of equal innervation. It is anism. Nevertheless, few centres are able to surprising, then, that asymmetric nystagmus undertake , and the is found in association with disorder in both initial assessment of a patient with nystagmus the sensory and the motor ends of the visu­ is clinical. A classificationaccording to clinical motor system. Violations of Hering's law have criteria is therefore presented, in Tables I been recorded8 as falling into two categories: and II. dynamic violations, which occur in the The examination of a patient with nystag­ absence of pathology, especially in fatigued mus must be organised to reveal as much eyes, and static violations, which only occur in information relating to the nystagmus as poss­ the presence of pathology in the motor con­ ible. It is not appropriate here to discuss trol pathways. Asymmetric nystagmus is an aspects of the general medical, neurological example of Hering violation in static oculo­ and ophthalmic history and examination motor conditions, and is considered in more which may be involved, and the discussion detail later. will be confined to consideration of the clini­ Nystagmus, then, signifies that the motor cal features of the nystagmus. control system is unable to maintain steady fixationon account of sustained asymmetry of History its inputs, integrator dysfunction, or failure of It is of fundamental importance to establish adaptation, fixation being restored with a the onset of nystagmus, so that acquired FEM when the visual input has sensed the slip forms may be distinguished from congenital.9 away. Similarly, defective visual input may The congenital status of nystagmus is a preclude stable fixation, if the image it pre­ particularly important point to establish when sents to the motor system is imprecisely presentation involves neurological signs indi­ organised, or possibly if the two hemifields cating possible acquired pathology. Many are incompletely coordinated, preventing the subjects with congenital nystagmus (eN) have integrator from accurately 'locking on' to an family histories suggesting autosomal domi­ internal representation of the world outside. nant or sex linked inheritance. Similarly, clues Having established the neural basis of suggesting other disease in the anterior visual steady fixation,and thence the background to pathway, or albinism, should be sought. The nystagmus, the clinical circumstances in effect of nystagmus on the subject's vision is which specific types arise can be considered. the principal source of symptoms, and oscil- 820 NICHOLAS EVANS

Table I Congenital nystagmus (onset within first lopsia or other visual change, and its variation 4 months) with time, head movement, gaze position, ill­ umination and fixation should be noted. Clinical Features These variables all have effects upon specific uniplanar horizontal almost always sites in the motor pathway, and in charac­ decrease on convergence voluntary lid closure terising the nystagmus may help localise its dark source. gaze to null zone mental distraction Clinical Examination increase on fixation Clinical examination reveals the 'direction' of mental concentration abnormal OKN ("reversal") nystagmus (conventionally the direction of its pursuit fast phase), and gives a subjective idea of its waveform exponentially increasing slow relative velocity and amplitude (together phase termed 'intensity') in different positions of associatcd abnormal head posture abnormal head movements gaze. This examination provides information which is often sufficient to characterise the without associated visual pathwaydisorder (idiopathic) Manifest (MN) pattern of nystagmus and suggest its neuro­ Latent (LN) revealed on occlusion pathologic basis, and it is easily recorded. The acuity drop on occlusion vectors are recorded in the primary position, often decreased if optical fogging in the six cardinal positions, and in midline replaces occlusion fast phase of occluded eye always elevation and depression, amplitude being towards fixating eye. represented by the number of shafts to the Manifest arrow or its thickness, and frequency of oscil­ Latent (MLN) manifest without occlusion lation by the number of tails (Fig. 1). This on occlusion covered eye simple standard notation permits ready nystagmus toward seeing eye. amblyopic eye commonly recording of the features of nystagmus, for nystagmus towards better eye-? comparison and future reference, as well as simulation of occlusion compelling the examining clinician to study small amplitude manifest the pattern of oscillation carefully. Additional component note should be made of any variation on con­ associated with other visual pathway disorders: vergence, fixationand lid closure, both volun­ albinism, achromatopsia, , hypoplasia, tapetoretinal degenerations, tary and passive, and of any associated head retinal dystrophy and detachment, glioma. movements. The response to optokinetic test­ onset delayed until secondary ing, rotation and vestibular stimulation, and development FEM and SEM function (to commanded sac­ absent in cortical blindness unstructured "roving" movements imply profound cades and on smooth pursuit) should be exam­ visual disability ined. Periodic Alternating Nystagmus (PAN) higher frequencies suggest relatively better visual may have cycles with periods lasting several function minutes, a fact that should be remembered in unilateral optic nerve/chiasm pathology associated cases demonstrating variability of vector. with lateralised nystagmus specifically localised intracranial lesions associated Finally, the importance of a full eye and with characteristic nystagmus patterns, cf neurological examination should not be acquired nystagmus forgotten.

Table II Acquired nystagmus Eye Movement Recording, Waveforms The information gained from a clinical exam­ vestibular .labyrinthine ination is usually all that is available to the nuclear central gaze evoked clinician, and is generally sufficientfor clinical ataxic management. Objective electronic eye move­ toxic, drug induced and metabolic ment recording techniques, however, using mixed Bruns DC electro-oculography (EOG)lO and infra specific acquired nystagmus syndromes red recording techniquesll,12 have enabled THE SIGNIFICANCE OF NYSTAGMUS 821

gests that the motor output in nystagmus has a +< -< >-- r- r- r- symmetry which is variable. Electronic recording shows component Slow and Fast eye movements, corresponding to off-target +< -< >-- r- r- r- drift and refixation, and a pendular pattern implies approximate equality between these +< -< >-- r- r- r- opposing components at a given gaze position, the nystagmus often reverting to a a) congenital nystagmus b) vestibular nystagmus jerk pattern on gaze deviation. Truly sym­ metric pendular nystagmus, with a sinusoidal waveform, is described as one of the variants 0 '\- 1 -/ -r .,..- of congenital nystagmus by Dell'Osso,15 who stresses however that the majority of --< 0 >-- 0 >-- >-+ examples of nystagmus previously labelled as pendular were misnamed, objective recording 0 � >-.. confirming slow and fast components. True / T � pendular nystagmus can also occur in the presence of brainstem disease, commonly of c) gaze evoked nystagmus d) gaze paretic nystagmus demyelinating or cerebrovascular origin. Fig. 1 Clinical recording of nystagmus Gresty and coworkers have documented 32 (a) Congenital nystagmus. High frequency oscillation such cases using electrooculographic record­ confined to the horizontal plane, null zone' to left of primary position, increasing intensity on opposite gaze. ing,16 which are discussed more fully later. (b) Ve stibular nystagmus. Combination of horizontal Finally, in the non-directional 'roving' or and vertical vectors, waveform and intensity 'searching' nystagmus of those profoundly uninfluencedby gaze position. visually impaired from birth, the eye move­ (c) Gaze evoked nystagmus. Nystagmus absent in primary position, increasing on eccentric gaze, with fa st ments are not structured, and represent a phase directed towards gaze position. more complete failure to fixatethan the struc­ (d) Gaze paretic nystagmus. Resolving left gaze tured nystagmus, characterised by cyclical at­ paresis, producing nystagmus with maximal intensity tempted refixation, with which we are on left gaze. The fast phase beats toward the paretic presently concerned. side. With insights gained from the study of eye characteristic details of the components of movement traces, a more thorough under­ nystagmus to be studied. EOG recording is standing is possible of the interaction between relatively simple, using standard periocular Fast and Slow eye movements in nystagmus, skin electrodes as for conventional EOG,13 and the nature of the causal disorder can often but recording vertical, torsional and uniocular be deduced. The relationship of eye position movements is not satisfactory. For technical to fixation can also be studied,17 providing reasons, artefact can easily be introduced into clues as to how nystagmus arises, and indi­ the traces, produced by current supply, band­ cating the different ways in which the relation­ pass characteristics and recording mechanics. ship between the oculomotor system and the Infra red techniques, on the other hand, pro­ visual field is disturbed. duce a direct recording of eye movements, are The fast component represents an attempt suitable for recording uniocular and multi­ to refixate which may be more or less accu­ planar nystagmus, but require more subject rate, is generally linear in profile, and is pro­ cooperation and greater technical sophistica­ grammed with respect to the perceived visual tion than EOG. In animal experiments, field, predicted target motion, known eye finally, a search c'oil may be sutured to the position and predicted eye motion. Its signif­ , the electric current induced by eye icance lies in the accuracy with which it res­ movement within a magnetic field being tores foveation during the cycle of nystagmus. recorded directly.14 It may be regarded as the correcting response The historic division of nystagmus into pen­ in nystagmus. dular and jerk (or saw-tooth) patterns sug- The slow component, on the other hand, 822 NICHOLAS EVANS represents the primary driving force of nys­ tagmus and is highly variable. It may have a linear waveform, or one with an increasing or decreasing velocity. The different profiles of the slow phase imply respectively: p 1) Linear. Constant eccentric driving trend, not derived from the visual system. 2) Increasing. Inappropriately high gain in the vision-mediated control loop coordinating eye position and target. The slow component 5 sec may be interrupted by secondary saccades, Fig. 2 which increase available foveation time, implying a compensatory visual input into the Right Eye nystagmic mechanism in some types of con­ genital nystagmus. 3) Decreasing. Decay in the position signal left E e 15 from the integrator, allowing the eyes to drift y back towards the primary position under the Fig. 3 influence of mechanical forces, whose strength decreases exponentially with decreasing angle of eccentricity. Clinically these classes of slow component profile correspond characteristically with (1) nystagmus of vestibular origin, (2) congenital nystagmus, (3) and gaze-evoked or gaze-par­ etic nystagmus, and examples of eye move­ ment recording traces illustrating them are shown in Figures 2, 3 and 4. The relationship between target position and eye movement 20 deg I indicates the effect on visual function of nys­ tagmus, and the extent to which the FEM and 1 sec SEM systems are able to serve the require­ ments of vision, or are driven inappropriately. Fig. 4 Fig. 2, 3 and 4 Electro-oculographic records of Congenital Nystagmus nystagmus waveforms. Congenital nystagmus (CN) is usually first Eye movement recordings in vestibular, congenital and gaze evoked nystagmus, showing the pattern of slow noticed in infancy, but has a wide range of and fast eye movements, and the effect of gaze deviation presentation, ranging from two or three on the direction, frequency, amplitude and waveform. months until in some cases early adulthood.9 Convention: upward deflection corresponds to eye It may be seen alone, or in association with movement to right; downward deflection corresponds to eye movement to left. other disorders affecting the infant visual (Fig 2) Ve stibular nystagmus: Constant velocity slow system (Table I). This distinction forms the phase uninfluencedby gaze direction (R = right gaze, P basis of its classical division into sensory and = primary, L = left gaze). Right-beating nystagmus motor types, though there is no evidence to caused by left-sided vestibular pathology. 3) support a different final mechanism, and the (Fig Congenital nystagmus: Slow phase shows increasing velocity exponential drift away from fixation terms have little clinical value. (to the right), fast phase regaining fixation. Nystagmus Clinically €:N is confined to the horizontal is left-beating and conjugate, ie right and left eye plane regardless of vertical gaze, its waveform waveforms are identical and in phase. 4) and intensity changing with horizontal gaze. (Fig Gaze-evoked nystagmus: Decreasing velocity slow phase away from fixation (towards primary Its uniplanality contrasts with most forms of position), fast phase towards direction of gaze. Right­ acquired nystagmus, which may superficially beating nystagmus on right gaze, left-beating resemble CN, but often show vertical or tor- nystagmus on left gaze. THE SIGNIFICANCE OF NYSTAGMUS 823 sional components in positions involving a than a theoretical construction), or its input, vertical gaze vector. There are exceptions to is ultimately responsible, cannot be inferred this horizontal predilection of CN,1 8,19 but from examination of eye movement traces. these cases are very unusual, and are The similarities between idiopathic and asso­ generally associated with ocular or cerebellar ciated CN suggests that defective visually­ disorder. Intrafamilial studies20,21 show that based monitoring of eye movement position there is often a genetic background to CN, plays a part in both, but speculation that sub­ though waveform and clinical manifestation jects with idiopathic CN may have an occult are variable amongst related subjects. visual sensory disorder, for example subclini­ Electronic eye movement recording tech­ cal albinism,7 have not been conclusive.23,6 niques have demonstrated three different CN occurs in manifest (MN), manifest­ classes of waveform in CN, showing pendular, latent (MLN) and latent (LN) forms,24 LN unidirectional jerk and bidirectional jerk pat­ and MLN being associated with . terns. These have been further subdivided In LN it seems that the change from binocular into a total of twelve different individual (through the 'midline cyclopean eye') to uni­ waveforms according to the relationship of ocular viewing conditions, on occlusion, pro­ the FEM and SEM with foveation, 15,22 These duces fixation ambiguity, central projection waveform types are shared by subjects with being an unreconciled conflict between the and without associated visual pathway dis­ uniocular single viewing fovea and the bin­ order,17 and differentiation between idio­ ocular 'midline eye', in the presence of het­ pathic and associated CN cannot be made on erotropia. , moreover, can the evidence of electronystagmography. simulate occlusion, producing the small Furthermore it is typical of CN that the wave­ amplitude manifest nystagmus in MLN, form changes readily from one type to whose fast phase is directed towards the non­ another with variation in gaze position and amblyopic eye. Defective pursuit and associ­ fixation, Common to all CN waveforms is a ated OKN asymmetry are features of strabis­ characteristic pattern of exponentially mic amblyopia25 as well as of CN, The increasing velocity deflection away from fix­ recognition that MLN has features of both ation, interrupted by a linear correcting sac­ MN and LN suggests a fundamental relation­ cade. This basic pattern may be complicated ship between all three, and implies that a to a variable degree by superimposed secon­ defect in the coordination of the projections dary movements, including small breaking and fields of the two eyes may contribute to and secondary refoveating sacca des , and in the lateral instability. The different OKN res­ some cases the identification of waveform ponses to temporal-nasal and nasal-temporal requires electronic differentiation of position­ stimulation contrast the influence on oculo­ time into velocity-time traces, The signifi­ motor control of the uncrossed nasal fields, cance of the waveforms recorded by electro­ which are wholly binocular, with the crossed nystagmography is that they suggest the kind temporal fields, which are partly uniocular, of neural dysfunction underlying the nystag­ and it has been suggested that congenital mus, and describe the relationship between absence of fusion is the reason that MLN is eye position and foveation. seen in children who have been functionally The division into twelve different wave­ uniocular from birth.26 Field separation forms may have limited practical clinical sig­ abnormalities are known to occur in albinism, nificance, but the constant finding of a slow and the difficulties of registration of the two component with exponentially increasing fields in any disorder involving congenital velocity indicates a common mechanism. The central visual defect (eg achromatopsia) are increasing velocity is caused by high gain evident. It should be remembered, however, instability in the integrator,-that is the SEM that binocular visual function is not incom­ generator drives the eyes eccentrically, under patible with idiopathic CN on grounds of positive feedback, at faster than the target insufficient acuity alone, since acuity as high velocity. Whether the integrator itself (which as 1.2 is possible in CN.27 Such acuity implies is perhaps no more a neuroanatomic locus target foveation lasting at least 100 msec each 824 NICHOLAS EVANS cycle.28 High acuity therefore requires a high Optican and Zee31 have developed a math­ degree of coordination between target and ematical model which explains eN in terms of fixation,which does not always occur in eN .17 an inappropriately high positive feedback sig­ Optokinetic Nystagmus (OKN) abnormal­ nal of eye velocity to the integrator. This ele­ ities have long been recognised in eN, impli­ gant model works whether the destabilising cating an SEM defect in its origin. Recent eye velocity signal is derived from extraocular work suggests that the OKN abnormality is muscle proprioceptors, or internally, from not simply an inversion, or an algebraic sum­ information on eye velocity within the inte­ mation of the optokinetic and congenital com­ grator itself. It also satisfactorily explains the ponents. The variation in the dynamics and many different waveforms seen in the same waveform of eN with gaze position and on subject during different visual conditions. pursuit has already been described. Dell'Osso Visual input may then interact with the inte­ suggests28 that the SEM involved in following grator to produce nystagmus in one of two an OKN stimulus causes a change in the ways. Either a normal input is processed position of the null point, thus altering the through a faulty integrator, producing an direction, and sometimes the waveform, of unstable output, or the tuning of the inte­ the underlying eN, so that OKN may appear grator is defective because of imperfect adap­ to be reversed-though in fact it is being tation in the presence of degraded visual shifted. Kestenbaum , and the use of input. This explanation of the mechanism of vergence- and version-inducing prisms, eN allows as partly correct many previous similarly have their modifying effect on eN by theories, and emphasises that both motor and a manipulation of the null point. The alternat­ sensory factors are simultaneously involved in ing direction of Periodic Alternating Nystag­ eN, since the integrator receives inter­ mus (PAN), and the similar, but more labile , dependent input from both. Similarly the changes in vector seen in 'bias reversal 17 clinical and electronic characteristics of idio­ probably represent another manifestation of pathic and associated eN being indistinguish­ variable null zone effect, on the one hand with able, it is likely that an identical mechanism, stable, and the other with unstable, period. such as this integrator abnormality, operates Kommerelf9 has suggested that an OKN in both. disorder itself may be the cause of CN. In this The capacity for exquisite accuracy of sac­ extension of earlier theories that eN is the cadic control in eN, which permits accurate result of high gain SEM disorder (subjects momentary foveation during nystagmus, in with eN are able to track targets with higher velocity than can normals), he proposed that spite of constant movement of eyes, target the defect involves failure of coordination and sometimes head, appears to rule out a between eye position and retinal slip of a primary saccadic defect as the cause of eN. moving image, i.e. that the defect is in the Primary cause and its secondary effect may, OKN system, involving a mismatch between however, be difficult to distinguish, as was motor output and visual input. This is postu­ pointed out in a former theory suggesting that lated to be due to a developmental failure of a saccade deficit produces eN, the observed the optokinetic control loop, with partial fail­ high gain in the SEM system being a compen­ ure of development producing LN. The sating response. 32 increase in intensity of eN during willed The relationship between fovea and target visual intention or effort, and its decrease not only determines the effect of nystagmus during voluntary (but not passive) lid on vision, but also helps explain functional closure30 and optical fogging (+ 30d) or in differences between OKN and eN. The darkness,1O confirms that it arises because of relationship between eye position and fixation disorder of a process in which vision plays a has been investigated using low light video­ fundamental part. Its modulation by lateral photography through an indirect ophthalmo­ gaze, and the reduction in its intensity during scope, to monitor and correlate the precise convergence, indicates a disorder involving foveal position with the target.17 Optimal the version system, and by implication the lat­ foveation strategy in eN requires that the flat erally-separated visual field. portion of the slow phase coincides with target THE SIGNIFICANCE OF NYSTAGMUS 825 foveation. These conditions were found in nystagmus. Metabolic disorders associated approximately 50% of subjects studied, the with nystagmus include Wernicke's encepha­ remainder failing to coordinate target fovea­ lopathy34 and organophosphate poisoning,35 tion with that part of the waveform during although conjugate gaze paresis seems to be a which eye movement is least. This variable more widespread effect on the oculomotor relationship between SEM and foveation system36 than nystagmus, perhaps reflecting accounts for the variability in in the symmetric effect on the eNS of metabolic subjects with eN, and also indicates that the disease. SEM is not a visualising movement as it is in OKN, but that the source of the instability Peripheral nystagmus itself is closely related to SEM abnormality. The labyrinths normally supply equal tonic innervation to the vestibular nucleii, the Acquired Nystagmus (Table II) balanced output of which, in the cerebellar Acquired nystagmus signifies acquired disor­ flocculus, maintains gaze stability. Lateral der in any of the centres or pathways control­ bias of vestibular output, caused by head ling the final input into the oculomotor rotation, causes reduction in the SEM gener­ nucleii. Such disorder results in nystagmus, ated on that side, the relatively unopposed rather than another supranuclear disorder contralateral SEM generator producing an (for example sustained deviation of conjugate opposite slow movement which maintains fix­ gaze), if inappropriate directional SEM out­ ation. This is the vestibulo-ocular reflex put is cyclically restored by FEM compensa­ (VOR), which is normally modulated by con­ tion. These centres and pathways may current input from the visual system via the conveniently be divided into those producing: corticobulbar and colliculobulbar pathways, a) asymmetric tonic input (derived from the and stabilises foveation in the face of head labyrinthine and visual systems), into the movements. Acute unilateral labyrinthine neural integrator. Nystagmus is characte­ disorder simulates a VOR, the reduced tonic ristically present in the primary position. output of the affected labyrinth leading to an b) defective output from the neural integra­ SEM towards the affected side, driven by tor (driving the SEM system). Nystagmus unopposed tone from the normal labyrinth. is characteristically maximal on eccentric This is cyclically interrupted by a refoveating gaze. saccade arising from the ipsilateral pontine Nystagmus produced by these two mech­ gaze centres. Peripheral vestibular nystagmus anisms is described respectively as peripheral therefore beats away from the side of the and central. They are distinguished by slow lesion, and has linear (constant velocity) slow phases which have respectively linear and phases. The intensity of peripheral vestibular exponentially decreasing time courses, nystagmus is reduced by fixation, as visual reflecting the different mechanisms which feedback (via the optokinetic system) over­ produce them. rides the VOR instability. It is characteristic of peripheral vestibular disease that adapta­ Aetiological considerations tion and recovery occur within a few weeks,37 Acquired nystagmus occurs in association the nystagmus generally resolving to leave with brainstem, cerebellar or vestibular only abnormal VOR gain responses after 3 disease, which the character of the nystagmus months, in spite of persistent deficient ves­ may help to localise. The underlying pathol­ tibular output. Since the disorder does not ogy may be focal (glioma, vascular tumour, include the SEM and FEM generators, pur­ metastasis) or more diffuse (demyelination, suits and sacca des are normal, and since the ischaemia, metabolic). eNS depressing drugs defect does not include the visual pathway, are associated with nystagmus, and it is of the pattern of nystagmus is unaffectedby gaze interest to correlate the effect of diazepam on position. the integrator, reported as causing a reduction It is instructive to compare and contrast the of VOR gain and increase in its time con­ features of vestibular and congenital nystag­ stant? with its clinical effect of gaze-evoked mus. eN is characteristically horizontally uni- 826 NICHOLAS EVANS planar, as has been described, reflecting the nerve,41 and can be congenital. The hetero­ organisation of the visual field about a hori­ geneous neural background to see-saw nys­ zontal symmetry. Similarly the plane of per­ tagmus confirms that an extended fieldof the ipheral vestibular nystagmus corresponds CNS may be involved in the response, and with the plane of response of the affected ves­ that the local ising value of such specific nys­ tibular end organs. It is usually predominantly tagmus forms can be limited. Their signif­ horizontal, with rotatory or vertical compon­ icance lies rather in the comparisons which ents. Pure rotatory or vertical nystagmus can­ they suggest with physiological eye-move­ not have a peripheral vestibular origin38 ment responses and other forms of nystag­ because of the labyrinthine geometry. mus, and in the insight they provide about Changes in vestibular nystagmus in response normal control of eye position and to vestibular stimulation39 are analogous to movement. changes seen in CN in response to manip­ ulation of visual input. Though vestibular nys­ Central nystagmus tagmus is more complex than CN (involving 3 The second kind of nystagmus associated with planes), the response to reinforcement of the acquired neurological disease occurs as a vector in one plane by rotation or caloric result of failure of the SEM system to sustain a stimulation (termed 'geo-fugal enhance­ constant tonic output to the motor nucleii. It ment') is comparable with reinforcement of is generally caused by loss ('leak') of tonic, or the vector in CN by lateral gaze, pursuit or 'step' signal, from the neural integrator, but optokinetic stimulation. Though the cause of may be simulated by motor neuropathy, neu­ fixation .instability is different in vestibular romuscular junction disorder (e.g. myasthe­ and congenital nystagmus, both occur on a nia) or extraocular muscle disease, which plane whose framework is definedby the sen­ cause progressive attenuation of effector res­ sory input (vestibular and visual) which ponse. The basic defect is, then, a neural out­ governs the nystagmus, and each responds put which is progressively inadequate to analogously to lateral bias in its sensory input. sustain the intended eye position, permitting Lesions affecting the central vestibular drifts toward the midline which are of pathway produce nystagmus with similar decreasing velocity, interrupted by refoveat­ characteristics, though pure vertical and rota­ ing saccades. The paradox in this waveform is tory patterns may be seen, and saccades and that conditions for satisfactory foveation pursuits may be impaired, reflecting the improve progressively as eye position higher level of involvement, and greater neu­ becomes increasingly off-target (though of ral organisation, as well as involvement of course the visual intention lies in the saccade, adjacent brainstem functions. not the SEM). Just as peripheral vestibular nystagmus in This form of nystagmus is termed 'gaze­ the horizontal plane can be compared with an evoked' (GEN), a specific type being gaze inappropriately activated VOR, certain idio­ paretic (GPN), which is directional, and seen syncratic forms of nystagmus can be under­ transiently during recovery of a paresis of con­ stood as specific VOR patterns activated jugate gaze. GEN is maximal on eccentric pathologically. See-saw nystagmus (elevation gaze, and GPN is maximal towards the and intorsion of one eye accompanied by affected side, decreasing on contralateral depression and extorsion of the other) is a gaze. The fast phase beats in the direction of cyclical repetition of the ocular response to gaze, in an attempt to restore fixation,and the head tilt, analogous to artificial horizon intensity increases with gaze towards the fast instrumentation in an aircraft, and while it has phase (Alexander's Law). True central GEN a complex pattern, it can be understood as the occurs as a result of functional disorder in the inappropriate activation of a physiological pontine gaze centre or cerebellar neurons, mechanism. It may be acquired with midbrain which may be associated with demyelination, or parasellar pathology, can be simulated ischaemia or compression. Drug-induced nys­ experimentally by stimulation of the inter­ tagmus (especially caused by anticonvulsants stitial nucleus of Cajal40 or the ultricular and sedatives) generally takes the form of THE SIGNIFICANCE OF NYSTAGMUS 827

GEN, the drugs' membrane-active effect origin, the MLF providing interneuron con­ probably underlying the leak in integrator nection between the VI and III nucleii, and function. contrasts with the uniocular nystagmus associ­ ated with lateralised sensory lesions (of the Mixed central/peripheral nystagmus optic nerve or chiasm). MLF involvement, Cerebellopontine tumours may produce a usually associated with demyelination, pro­ combination of central and vestibular nystag­ duces in the abducting eye a nystagmus which mus, by causing pontine compression simul­ is essentially of the high-gain instability type, taneously with direct or the SEM system increasing its gain in an effort end-organ damage. On gaze toward the lesion to overcome the internuclear adduction par­ the features are those of GEN (maximum esis of the adducting eye. intensity of nystagmus, the fast phase beating toward the lesion, and an exponentially Specific Nystagmus Types decreasing waveform). On opposite gaze the Acquired nystagmus usually corresponds vestibular (tonic imbalance) type of nystag­ broadly with the patterns described above, its mus predominates (unidirectional, no form and associated clinical features suggest­ increase on lateral gaze, suppression by fix­ ing its causal pathology. Certain idiosyncratic at ion , and a linear slow phase). This picture is forms of nystagmus, showing highly charac­ known as Bruns nystagmus,42 and elegantly teristic clinical patterns, have acquired demonstrates the separate mechanisms descriptive names. Some of thes.e imply well­ involved in the two classes of acquired defined CNS localisation, allowing aetiologic nystagmus. diagnosis on clinical examination alone. Localisation is not always so accurate, how­ Asymmetric (ataxic or dissociated) ever, and their characteristic patterns can nystagmus often be correlated only loosely with under­ Asymmetric and uniocular nystagmus are lying neural mechanism disorder. Though the found in association with disorders of both the idiosyncracy of their patterns is intriguing, sensory and motor parts of the visuaUoculo­ their study does not contribute greatly to our motor system. general understanding of nystagmus and its Uniocular nystagmus is found in associ­ mechanisms, and they will be mentioned ation with uniocular pathology in infants, briefly. particularly optic nerve glioma,43 and asym­ metric nystagmus with greater amplitude on Acquired pendular nystagmus the more affected side occurs in optic chiasm A distinct form of pendular nystagmus occur­ glioma.44 Marked asymmetry, and sometimes ring with brainstem disease, usually of unilaterality of nystagmus, is likewise a fea­ demyelinating or cerebrovascular origin, has ture of spasmus nutans, though optic nerve been documented by objective eye movement disease should be excluded by computerised recording. 16 The nystagmus differs from pen­ tomography.45 The supranuclear centres gen­ dular CN in that its waveform is independent erating eye movements seem to be capable of of gaze position. It is variably uniocular or initiating different movements of the two eyes binocular, conjugate or dysjugate, and occurs independently, in violation of Hering's law. in a combination of one or more of the three More surprisingly, sensory information rela­ planes of ocular rotation. Concurrent body ting to eye side as well as field side must be is common, and oscillopsia invariable. preserved beyond the chiasm. The signifi­ Saccadic and pursuit movements, and optoki­ cance of this is considered in the discussion of netic and vestibulo-ocular reflexes may all be congenital nystagmus and amblyopia. normal. This form of nystagmus is held to Nystagmus involving the abducting eye represent a distinct entity on account of its selectively, in association with adduction par­ unusual constellation of features, and particu­ esis, is a feature of lesions involving the larly on account of the narrow band of fre­ medial longitudinal fasciculus (MLF). This quencies of its oscillation, the constant unilateral form of nystagmus is truly of motor vergence deficiency, and internuclear oph- 828 NICHOLAS EVANS thalmoplegia (INO) with which it is associ­ Spasmus nutans is a self-limiting disorder of ated. It is suggested that the underlying lesion infants in which nystagmus coexists with an involves cells in close relation to the oculo­ abnormal head posture (AHP) and abnormal motor nucleii and the vergence system. head movements, in the absence of other neurological or ocular disorder. The nystag­ Vertical nystagmus mus may be conjugate or dysjugate, and is Downbeat and upbeat nystagmus have been sometimes confined to one eye. If dysjugate, classed with vestibular nystagmus as caused the two eyes may share a common pattern by tone imbalance46 because of the linear time with phase-difference, or the two patterns constant of their slow phases. The causal may be dissimilar. 48 The head movements in defect is obscure, but probably involves spasmus nutans are of some interest. Pre­ defective balance in the vertical pursuit gener­ viously thought to be the convergence of the ators. The defective horizontal pursuits seen nystagmus movements, producing a net zero in horizontal nystagmus, and its modulation movement of the eyes, it is now recognised by horizontal gaze movements, have parallels that head movement reduces the nystagmus in vertical nystagmus: the slow phase in down­ by suppression, rather than by neutralising it beat nystagmus is dependent on vertical head algebraically. This suppression, which is a position, and vertical pursuit movements are strategy to improve vision, is held to be impaired.47 In convergence-retraction nystag­ mediated by vestibular stimulation.49 The mus, attempted downward vertical pursuit, reduction in the intensity of nystagmus in best elicited in response to a downward­ spasmus nutans on vestibular stimulation is moving optokinetic stimulus, causes extra­ analogous to the supression of vestibular nys­ ocular muscle co-conduction and produces tagmus by visual stimuli. cyclical retraction and convergence. It occurs in association with midbrain disorder, as does Physiological Nystagmus upbeat nystagmus, and the association of con­ Development and adaptive aspects vergence-retraction with a downward optoki­ Physiological nystagmus may be produced in netic response perhaps reflects the anatomic response to visual (OKN) or vestibular proximity of the neural substrates of vertical (V OR) stimulation. It maximises foveation movement and convergence. Downbeat nys­ time when the relative rotation of eyes, head tagmus, however, is associated with cranio­ and field would otherwise reduce visual cervical junction and cerebellar acuity. The forms of pathological nystagmus abnormalities, especially platybasia and the described above need to be considered in the Arnold-. The difference light of this capability of oculomotor system to between other types of vertical nystagmus, produce oscillating eye movements. The pres­ which are located with vertical gaze in the ent discussion of physiological nystagmus will midbrain, and the lower brainstem location of be confined to certain experimental and downbeat nystagmus, is not clear. developmental aspects of OKN and VOR, which are relevant to an understanding of the See-saw and Periodic alternating nystagmus, interaction of its components. and Spasmus nutans The maintains a static See-saw nystagmus has already been men­ relationship between fovea and target during tioned as corresponding with a VOR in the movement of the target, using smooth pursuit sagittal plane of rotation, and may be caused to lock foveation for as long as the ocular by inappropriate input to the SEM centres excursion permits, and saccade to recapture a from the central vestibular system. Periodic subsequent phase of fixation. The control alternating nystagmus has also been system is very precise, and being a closed loop described, and the variation with time of its (sensor and effector are the same) is readily waveform and direction was shown to corre­ adaptable. The critical function in the opto­ spond with a cyclical movement of the null kinetic response is its gain, or the relationship zone, which is itself influenced by the SEM between the velocity of eye and target. system. Experiments investigating the plasticity of THE SIGNIFICANCE OF NYSTAGMUS 829 this gain in the OKN and VOR systems in response takes 5-6 months to develop, prob­ monkeys, involving manipulation of retinal ably in parallel with the development of image motion (in an ), ves­ binocular single vision, and shows tem­ tibular input (by unilateral labyrinthectomy) poral-nasal asymmetry during its develop­ and visual memory (by occipital iobectomy) ,50 ment.57,58 The delay in development of the indicate that the occipital lobe participates in smooth pursuit system in infants, during adaptation (gain adjustment) to fast target foveal development, may be another reason movement, though at slow speeds subcortical why the onset of congenital nystagmus is pathways may suffice. Vestibular nystagmus delayed. Following movements are achieved induced by unilateral labyrinthectomy spon­ in the interim by a combination of head move­ taneously disappeared independently of ments and small saccades,59 and the accuracy visual or cortical function. This adaptation of following improves significantly over the appears to be a local property of the ves­ first two months. 60 tibular-cerebellar axis, though in more physiological circumstances the dominant Summary optokinetic influencedoubtless plays a signifi­ Nystagmus occurs in a very wide range of cir­ cant role in adaptation. Accurate VOR gain cumstances, each type showing characteristic adjustment is vision-dependent, being clinical, pathological and electrophysiological delayed until vision is restored, and slowly features, and analogies between them can be lost after subsequent occipital lobectomy. The identified by comparing and contrasting nys­ circumstances of these experiments are tagmus of different kinds. The effec. of unphysiological, and extrapolation of the altered visual and vestibular conditions on results to the normal human must be guarded. nystagmus, and the features of its waveform, They do, however, suggest how vision, the indicate the relationship between eye move­ VOR and the optokinetic response are ments and vision, and the influence of visual related, emphasising the dominant influence and vestibular input in stabilising steady fix­ of visual input in the accurate control of reflex ation. Ultimately, the significance of nystag­ ocular movement. Previous experiments had mus is that it indicates the state of the suggested that extracortical pathways can sus­ mechanisms underlying this stabilisation: in tain an attenuated OKN,51 and it seems that physiological nystagmus they are operating , the proposed 'two visual systems 52 may pro­ successfully, and in pathological nystagmus ject a dual input, derived from the cortical and they are disturbed. More than this, investiga­ extra cortical systems, to the motor control tion of nystagmus has shown that the visual centres. system is not divided in a clear-cut way into The extent to which oculomotor reflexes, sensory and motor poles, but that between including nystagmus, require a functional them there exists a neural region where a geniculostriate system is unresolved. Clinical 'copy' of the visual world is matched with a observation suggests that the onset of programme of potential eye movements, and congenital nystagmus of any aetiology is where sensorimotor information exists indi­ delayed until the geniculostriate system is visibly. Long feedback loops, involving occip­ functional, and that cortically blind infants ital cortex and extraocular muscle fail to develop nystagmus, in contrast to those proprioceptors, and short ones within the with anterior visual pathway pathology. 53 On and integrator, emphasise the the other hand OKN has been reported in great precision involved in eye movement (older) infants with cortical blindness.54 The control, enabling the visual cortex to make V.OR is developed before geniculostriate optimal use of the resolution capabilities of maturity, and the one month infant can main­ the fovea. tain fixationby a combination of eye and head Nystagmus always reflectsan asymmetry in movements and VOR.55 The infant's VOR is the output of the eye movement generators, less accurate than that of adults,56 probably and it has been shown that the inappropriate reflecting the lesser influence of their imma­ movement which is responsible for patho­ ture optokinetic response. The optokinetic logical nystagmus is the slow movement. This 830 NICHOLAS EVANS may arise because of an intrinsic defect in that producing an 82% reduction in amplitude, part of the generator called the neural inte­ and a 34% reduction in frequency.62 grator, or because of 'tonic imbalance' in its Nystagmus is, in conclusion, a protean sign. input. It may signify profound visual disability, but is Nystagmus occurring with identifiable compatible with high visual acuity. It was acquired central nervous pathology can, to formerly classifiedas of motor or sensory orig­ some extent, be understood mechanistically, in, but derives from a part of the brain where but idiopathic congenital nystagmus poses the two functons are brought together into greater difficulties. Analysis of its waveform one. It may signify serious progressive central suggests that an intrinsic fault in the integrator nervous disease, but may also occur tran­ can explain the clinical and electrophysiolog­ siently in infancy, along with head-nodding, ical findings in CN, but the cause of the high and resolve spontaneously. It may be gain instability in the integrator remains to be explained in terms of neurophysiology, neu­ explained. The integrator is adaptable. or roanatomy, mechanics, mathematics, or self-tuning, adjusting its output by visual feed­ cybernetic systems analysis, or it may defy the back. Circumstantial evidence suggests that best efforts of explanation. It will always pro­ the original disorder in idiopathic CN may vide a great clinical challenge, and the multi­ occur higher than the integrator, de tuning it tude of related factors it involves will by conveying an inappropriately organised guarantee that whoever studies nystagmus visual input. In particular if the organisation will learn about much else besides in the of the visual system into fieldsis defective, the process. gaze generators, whose output is orientated according to field, will have a less accurate I am grateful to the MRC Neuro- Unit at 'copy' of the world from which to formulate the Institute of , and to Dr. RV Abadi, their movements. Defective binocular func­ Department of Optometry and Vision Sciences, tion is a feature of latent and manifest latent University of Manchester Institute of Science and Technology, for kindly providing the traces of nys­ nystagmus, as well as those examples of CN tagmus recordings reproduced in this paper, and to associated with other visual pathway path­ Beverley Sabberton-Coe for their preparation. ology, and the different motor responses to nasal and temporal field stimulation suggest References that registration of the two hemifields may J Good: The Study of Ill, pub unknown, play a role in motor coordination. There quoted OED. 2 Cogan DG: Congenital Nystagmus. Can J Ophthal­ remain many more mysteries than answers in mol 1967, 2: 4-10. congenital nystagmus. 3 Cannon SC and Robinson DA: The final common Comparisons between different nystagmus integrator is in the prepositus and vestibular types may be made at many levels, indicating nucleii. In Adaptive processes in visual and oculo­ motor systems. Proc Asolimar conference 1985. that some basic features are common. The Eds Keller EL and Zee DS, Pergamon Press relationship between the plane of oscillation Oxford. 307-311. and the plane of the eyes and the labyrinths, 4 Eckmiller 13-:The transition between promotor eye and their reinforcement in eccentric rotation velocity signals and oculomotor eye position sig­ of gaze, invites comparison between acquired nals in primate brainstem neurons during pursuit. in Adaptive Processes in Visual and Oculomotor vestibular and congenital nystagmus. Vestib­ Systems. Proc Asolimar conference 1985. Eds ular nystagmus is reduced or cancelled by Keller EL and Zee DS, Pergamon Press Oxford. visual fixation, and nystagmus in spasmus 301-5. nutans is cancelled by vestibular stimulation. 5 Miles FA and Fuller JH: Visual tracking and the pri­ mate flocculus . Science 1975, 189: 1000. A method of controlling CN involving feed­ 6 The VEP and misrouted pathways in human albi­ back of information via the VIII nerve has nism. In Cracco RQ, Wollner IB Eds. Frontiers of 1 been proposed by Abadi and co-workers ,6 Clinical Neuroscience Series. Evoked Potentials. using an audible signal whose frequency New York. Alan Liss Inc. 1986: 211-16. 7 depends on the intensity of the nystagmus, as Simon JW, Kandel GL, Krohel GB, Nelsen PT: Albinotic characteristics in congenital nystagmus. feedback. Though this approach has not Am J Ophthalmol 1984, 97: 320-7. found widespread use, it has been reported as 8 Bahill AT, Ciuffreda KJ, Kenyon R, Stark L. THE SIGNIFICANCE OF NYSTAGMUS 831

Dynamic and static violations of Hering's law of 27 Kommerell G, Horn R, Bach M: Motion equal innervation. Am J Optom and Physiol in congenital nystagmus. In: Adaptive processes Optics 1976, 53: 786-96. in visual and oculomotor systems. Proc Asolimar 9 Gresty M, Page N, Barratt H: The differential dia­ conference, Eds Keller EL and Zee DS, Oxford: gnosis of congenital nystagmus. J Neurol, Neu­ Pergamon Press 1985, pp 485-91. rosurg Psychiatry 1984, 47: 936-42. 28 Dell'Osso LF. Congenital Nystagmus: Basic 10 Yee RD, Wong EK, Baloh RW, Honrubia V: A aspects. In: Functional Basis of Ocular Motility study of congenital nystagmus: Waveforms. Neu­ Disorders. Proc Int Symposium , Stockholm. Eds: rology 1976, 26: 326-33. Lennerstrand G, Zee DS and Keller EL, Oxford: 11 Cook G and Stark L: The Movement Pergamon Press 1981. pp 129-38. Mechanism; experiments, modelling and model­ 29 Kommerell G and Medhorn E. Is an optokinetic testing. Arch OphthalmoI 19.68 , 79: 428-36. defect the cause of congenital and latent nystag­ 1 2 Cook G, Stark L, Zuber BL: Horizontal eye move­ mus? In: Functional Basis of Ocular Motility Dis­ ments studied with the on-line computer. Arch orders. Proc lnt Symposium, Stockholm. Eds: Ophthalmol 1966, 76: 589-95. Lennerstrand G, Zee DS, Keller EL, Oxford: L1 Dix MR , Hallpike CS , Hood JD. Electronystag­ Pergamon Press 1981. pp 159-70. mography and its uses in the study of spontaneous 311 Shibasaki H, Yamashita Y, Motomura S: Suppres­ nystagmus. Tr ans Ophthalmol Soc UK 1963, 83: sion of congenital nystagmus. J Neural, Neuro­ 531-57. surg and Psychiat 1978, 41: 1078-83. 14 Fuchs AF and Robinson DA: A method for measur­ 31 Optican L and Zee DS: A hypothetical explanation ing horizontal and vertical eye movement chron­ of congenital nystagmus. Bioi Cy bern. 1984, 50: ically in the monkey. J Appl Physiol 1966, 2: 119-34. 1068-70. 32 Dell'Osso LF, Gauthier G, Liberman G, Stark L. 15 Dell'Osso LF and Daroff RB: Congenital nystagmus Eye movement recording as a diagnostic tool in a waveforms and foveation strategy. Doc Ophthal­ case of congenital nystagmus. Am J Optom and mol 1975, 39: 155-82. Arch Am Acad Optom. 1972, 49: 3-13. 10 Gresty MA, Ell 11 and Findley LJ: Acquired pen­ 33 Blair SM and Gavin M: Modifications of vestibulo­ dular nystagmus: its characteristics, localising ocular reflex induced by diazepam. Arch Otolar­ value and pathophysiology. J Neural, Neurosurg yngol 1979, 105: 698-701. and Psychiatry 1982, 45: 431-9. 34 Zak TA and D'Ambrosio FA: Nutritional nystagmus 17 Abadi RV and Dickinson CM: Waveform charac­ in infants. J Ped Ophthalmol Strabismus 1985, 140-2 . teristics in congenital nystagmus. Doc Ophthal­ 224: mol 1986, 64: 153-67 . 35 Jay WM, Marcus RW, Jay MS: Primary position 18 Hoyt CS and Gelbart SS: Vertical Nystagmus in upbeat nystagmus with organophosphate poison­ ing. Ped Ophthalmol Strabismus 1982, 19: infants with congenital ocular abnormalities. J 318-9. Ophthalmic Paediatrics Genetics 1984, 4: 155-62. 30 Cogan DG, Chu FC , Reingold MA, Barranger J: 19 Sogg RL and Hoyt WF: Intermittent vertical nystag­ Ocular motor signs in some metabolic diseases. mus in a Father and Son. Arch Ophthalmol 1962, 68: 113-5 . Arch Ophthalmol. 1981, 99: 1802-8. 37 Baloh RW, Honrubia V, Yee RD, Langhofer L, 211 Dell'Osso LF, Flynn JT, Daroff RB: Hereditary Minser K. Recovery from unilateral peripheral Congenital Nystagmus: An Intrafamilial Study. vestibular lesions. In: Adaptive processes in Arch Ophthalmol 1974, 92: 366-74. visual and oculomotor systems. Proc Asolimar 21 Spooner SN, Bateman JB, Yee RD: Congenital Nys­ conference 1985. Eds Keller EL and Zee DS, Per­ tagmus in identical twins: Discordant Features. J gamon Press Oxford. pp 349-56. Pediat Ophthalmol Strabism 1986, 23: 115-9. " Van Vliet AGM: Nystagmus. Doc Ophthalmol 1982, 22 Dell'Osso LF: Functional Definitions and classifi­ 52: 435-46. cations of congenital nystagmus waveforms. Oph­ 39 Gresty MA, Barratt H, Bronstein A, Page N: Clini­ thalmol Digest 1976, 38: 19-27 . cal Aspects of -oculomotor relationships. 23 Meienberg Hemphill G, Rosenberg M, Hoyt 0, In: Adaptive processes in visual and oculomotor WF: Visually evoked response asymmetries in a systems. Proc Asolimar conference 1985. Eds family with congenital nystagmus. Possible evi­ Keller EL and Zee DS, Pergamon Press Oxford. dence of abnormal visual projections. Arch Neu­ pp 357-{)5 . 40 rol 1980, 37: 697-8. Westheimer G and Blair SM: The ocular tilt reac­ 24 Dell'Osso LF, Schmidt D, Daroff RB: Latent, Mani­ tion-a brainstem oculomotor routine. Invest fest Latent and Manifest Nystagmus. Arch Oph­ Ophthalmol 1975 , 14: 883-93. thalmol 1979, 97: 1877-85 . 4, Suzuki Jr, Tokumasu K, Goto K: Eye movements 25 Tychsen L, Hurtig RR, Scott WE: Pursuit is from single utricular nerve stimulation in the cat . impaired but the vestibulo-ocular reflexis normal Acta Otolaryngol 1969, 68: 350-{)2. in infantile strabismus. Arch Ophthalmol 1985, 42 Baloh RW, Konrad HR, Dirks D, Honrubia V: 103: 536-9. Cerebellopontine angle tumours: results of quan­ 26 Harcourt B: Aetiology, classification and clinical titative vestibulo-ocular testing. Arch Neurol. characteristics of . In infancy, in Strabis­ 1976, 33: 507-12. mus and Amblyopia. Wenner- Gren International 43 Farmer J and Hoyt CS: Monocular nystagmus in Symposium Series, 49: London Macmillan. 1988, infancy and early childhood. Am J Ophthalmol 23-35. 1984, 98: 504-9. 832 NICHOLAS EVANS

44Lavery MA, O'Neill JF, Chu F, Martyn U: 53 Fielder AR and Evans NM: Is the Geniculostriate Acquired nystagmus in early childhood: a pre­ System a Prerequisite for Nystagmus? Eye, 1988, senting sign of intracranial tumour. Ophthal­ 2: 628-35. mology 1984, 91: 425-35. 54 Van Hof-Duin J and Mohn G: Optokinetic and spon­ 45 Koenig SB, Naidich TP, Zaparackas Z: Optic glioma taneous nystagmus in children with neurological masquerading as spasmus nutans. J Ped Ophthal­ disorders. Behavioural Brain Research 1983, 10: mol Strabismus 1982, 19: 20-4. 163-75. 46 DaroffRa nd Dell'Osso LF: Nystagmus-a Contem­ 55 Regal DM, Ashmead DH, Salapetek P: The coordi­ porary Approach. In: Topics in Neuro-Ophthal­ nation of eye and head movements during early mology, Baltimore. ed Thompson HS. Williams infancy: a selective review. Behavioural Brain and Wilkins. 1979, 286-97. Research 1983, 10: 125-32. 47 Halmagyi GM, Rudge P, Gresty MA , Sanders MD: 56 Cordero L, Clarke DL, Urrutia JG: Postrotatory Downbeating nystagmus. A review of 62 cases. nystagmus in the full-term and premature infant . Arch Neural 1983, 40: 777-83. Int J Pediatric Otorhinolaryn 1983, 5: 47-57. 48 Weissman BM, Dell'Osso LF, Abel LA, Leigh RJ: 57 Naegele JR and Held R: The postnatal development Spasmus nutans. A quantitative prospective of monocular optokinetic nystagmus in infants. study. Arch Ophthalmol 1987, 105: 525-8. Vision Res 1982, 22: 341---{j. 49 Gresty MA and Hamalgyi GM: Head nodding asso­ 58 Schor CM, Narayan V, Westall C: Postnatal ciated with idiopathic congenital nystagmus. In: development of optokinetic aftcr nystagmus in Cohen B (ed): Vestibular and Oculomotor Physi­ human infants. Vision Research 1983 23: 1643-7. ology: International meeting of the Barany 59 Roucoux A, Culee C, Roucoux M: Development of Society. New York, New York Academy of fixation and pursuit eye movements in human Sciences, 1981. 614--8. infants . Behavioural Brain Research 1983, 10: 50 Fetter M, Zee DS, Proctor LR: Effect of lack of 133-9. vision and of occipital lobectomy upon recovery 60 Ruff HA , Lawson KR, Kurtzeberg D, McCarton­ from unilateral labyrinthectomy in Rhesus mon­ Daum C, Vaughan HG: Visual following of key. J Neurophysiol. 1988, 59: 394-407. moving objects by full-term and pre-term infants. 51 Wood CC, Spear PD, Braun JJ: Direction-specific J Pediatric Psychology 1982, 7: 375-86. 1 deficits in horizontal optokinetic nystagmus fol­ 6 Abadi RV , Carden D, Simpson J: A new treatment lowing removal of visual cortex in the cat. Brain for congenital nystagmus. Br J Ophthalmol 1980, Res. 1973, 60: 231-7. 64: 2---{). 52 Schneider GE: Two visual systems. Brain mechan­ 62 Ciuffreda KJ , Goldrich SG, Neary C: Use of eye isms for localisation and discrimination are dis­ movement auditory feedback in the control of sociated by tectal and cortical lesions. Science nystagmus. Am J Optom Physiolog Optics 1982, 1969, 163: 895-902. 59: 396-409.