J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.50.4.411 on 1 April 1987. Downloaded from

Journal of Neurology, Neurosurgery, and Psychiatry 1987;50:411-415

A new optical treatment for oscillopsia

DAVID RUSHTON,* NEIL COXt From the Medical Research Council Neurological Prosthesis Unit,* and Kings College Hospital Department of Ophthalmology,t London, UK

SUMMARY A simple optical device (spectacles plus contact ) enabling viewing of the real world with either partial or almost-complete retinal image stabilisation has been tested in patients with oscillopsia caused by . The device gave a useful improvement in vision in two of 14 patients. Reasons for success and failure were clear and are discussed. Obvious contraindications include severe optic atrophy, titubation and dementia. Net benefit is also unlikely if there is a good null point or area to the nystagmus, or if acuity (corrected but unstabilised) is 6/9 or better. It is not suitable for the treatment of oscillopsia caused by failure of the vestibulo-ocular reflex.

Some, but not all, patients with nystagmus experience spectacle lens; so whichever way the eye looks, a ray a sensation of continuous oscillation of the visual from that object is traversing the optic axis of the eye, world (oscillopsia), which can result in a marked loss and reaches the same retinal point. The high-minus

of acuity owing to continuous retinal image-slip. contact lens is then needed to restore image focus to Protected by copyright. Oscillopsia of this sort is not uncommon in multiple the . sclerosis.' Up to now, no effective treatment has been available for oscillopsia, with rare exceptions. Good Nystagmus, oscillopsia and central cancellation results have been reported in periodic alternating nys- Some patients with nystagmus may report only slight, tagmus, using baclofen.2 Occasional patients treated intermittent or no oscillopsia, so evidently the brain with isoniazid for rubral tremor in multiple sclerosis3 can under some circumstances "cancel out" the reti- have also reported some improvement. Surgical treat- nal image-slip, with the effect that the visual world ment by extraocular myotomy or recession is destruc- remains steady. This may be done by mechanisms tive and often ineffective;4 it is not often used. akin to those that preserve the stability of the seen Treatment of oscillopsia using retrobulbar injections world during saccades and the fast phase of phys- of botulinus toxin has also been described,5 but this procedure has obvious dangers. We report results in an unselected group of patients S referred with oscillopsia, using a recently-developed method6 for stabilising the retinal image, using a h z _ ~ NOX9" B http://jnnp.bmj.com/ wearable device (fig 1) which allows viewing of the a2 real world. It comprises a high-minus contact lens b F -I X A s- I (fig 2a) combined with a high-plus spectacle lens a 1I C--- (fig 2b). Depending on the power of the lenses, this combination has a greater or lesser effect in reducing retinal image-slip as the eye moves. The simplest way to visualise the principle of the device is to consider a Fig I Diagrammatic representation ofthe optical method lens whose is such that (if the for obtaining retinal image stabilisation (RIS), Forfull RIS, plus spectacle power on September 30, 2021 by guest. eye's own optics are disregarded for a moment) a dis- the spectacle lens S is placed with its secondaryfocalpoint F' tant object on the lens axis would be imaged at the at the centre ofrotation ofthe eye. When viewing a distant centre of rotation of the eye. Then, all rays from that object AB, the lens would then converge all rays (such as bl, the b2) from the point ofinterest B on its optic axis, so that they object are radial to the eye after passing through are radial to the . Then, when the eye is turned, rays to traverse the axis the so Institute of from B will continue optic of eye, Address for reprint requests: Dr DN Rushton, to view that same in The Psychiatry, De Crespigny Park, London SE5 8AF, UK. that it continues point space. strongly divergent contact lens C is now required to diverge Received 4 February 1986 and in revised form 10 June 1986. the rays sufficiently so that the imnage isformed at the retina, Accepted 19 June 1986 rather than in the vitreous. 411 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.50.4.411 on 1 April 1987. Downloaded from

412 Rushton, Cox iological nystagmus. Many observers believe that this lopsia, such as a raised perceptual threshold, or normal "central cancellation" mechanism involves reduced sensitivity.8 9 outflow monitoring of oculomotor signals (ref 7, Patients whose oscillopsia amplitude is less than pp 237-41). In many other patients with nystagmus would be predicted from the amplitude of nystagmus an imperfect attempt at central cancellation may then require only partial retinal image stabilisation occur, so that the amplitude of oscillopsia is smaller (RIS) in order to eliminate oscillopsia by supple- than would be predicted from the amplitude of the menting their central cancellation. In them it is neces- nystagmus. This phenomenon (oscillopsia smaller sary to measure the angular size ("N") of the than retinal slip) has been well described and recorded nystagmus, and the angular size ("O") of the oscil- in oscillopsia caused by a deficiency of the vestibulo- lopsia and compare the two to see how much RIS is ocular reflex (VOR),8 and in downbeat nystagmus,9 needed. The necessary observations of 0 and N can but it may be more general.10 Besides cancellation, be done semi-quantitatively without special equip- there are other possible mechanisms for reduced oscil- ment, so as to calculate O/N approximately for each patient (see Methods). The amount of supplementary RIS required to abolish the remaining oscillopsia is then given directly by O/N; RIS is on a scale running from 0 (normal; no stabilisation) to 1 (retinal image fully stabilised). There are several assumptions implicit in this argu- ment. They are firstly that O/N is relatively constant l | S. j j ll. | --- with time and eye-position for a given patient. Sec- 101I |#|s''l ondly, it is assumed that 0 is caused by the retinal ...... l.. | W. Z. image-slip . .S resulting from N, so that it can be elimi- . S5

nated by reducing the image-slip. Thirdly, it isProtected by copyright. assumed that any cancellation signal is not greatly w.W.Z ,W .lw xvX. phase-shifted in relation to N. Fourthly, it is assumed that supplementary RIS will not itself alter the amount of central cancellation. These assumptions seem to be roughly borne out in practice. 11_..# ka. Prescribing the correct amount ofstabilisation In theory, the amount of RIS applied should be O/N. In practice, owing to difficulties of lens manufacture, the choice available to our patients was limited to three values, approximately 0 3, 0 5 and 0-82. It is well known that perfect stabilisation (RIS = 1'0) of the retinal image leads to intermittent visual fading.11 This problem does not arise with the RIS values used Fig 2 (a) Contact lens. This example is-58 D, giving RIS here. The nomogram (fig 3) shows the relation = 0-82. It is made with a planefront surface, which is easy to between contact lens power (Pc), spectacle lens power http://jnnp.bmj.com/ form, and is made ofconventional polymethyl methacrylate (Ps), separation of the principal planes of the lenses (.u = 1J5). It weighs 55 mg. Ifthefront surface is made to (Ds-c), and RIS, and it is used for prescnrbing the the largest radius available on a conventional contact lens correct lenses. Given values for any two of the four maker's radius lathe, then the power is about -28 D (RIS = variables, the other two can 0 5 approx). For this practical reason, lenses between -30 be read off the nomo- and -55 D have not been available to us. In order to obtain gram. RIS = 1, the contact lens po