676 BritishlournalofOphthalmology, 1990,74,676-680 Pupillary responses in amblyopia Br J Ophthalmol: first published as 10.1136/bjo.74.11.676 on 1 November 1990. Downloaded from Alison Y Firth Abstract light was then alternatively switched from one Relative afferent pupillary defects (RAPD) eye to the other, giving a period of stimulation of were detected in 32*3% ofpatients with ambly- 1 to 2 seconds, and the initial pupillary con- opia by a modification ofthe swinging flashlight striction was observed. The light was then left in test and the synoptophore. After consideration front ofeach eye for a count of3 and the pupillary ofvarious clinical investigations the significant escape noted. factors identified in patients showing a RAPD If a pupillary defect was observed, a neutral were: anisometropia, early age of onset where density filter was placed in the arm of the strabismus was present, level of visual acuity synoptophore in front of the eye without the following treatment, longer period ofocclusion defect. In practice it was not found possible to therapy. These points bear similarities to the quantify the defect to within 0-1 log unit as has results of pattern electroretinograms (PERG) previously been reported,5 but merely to confirm in amblyopes, and the possibility of the its presence. Where no defect was initially causative defect being at ganglion cell level is apparent, a 0-3 log unit NDF was placed in discussed. The effect of occlusion treatment either arm in turn to produce a difference in cannot be predicted from the presence or response. In some cases this revealed a subtle absence of a RAPD. defect, as the pupillary response was still present but to a lesser extent in one eye while completely absent in the other. During the examination An afferent or relative afferent pupillary defect refractive correction was worn, the interpupillary has been reported to be present in between 9% distance corrected, and the tubes set at the and 93% of amblyopes.'' The main criticism of objective angle of deviation by rotating each arm these findings is that poor fixation in the ambly- equally, thus ensuring similar stimulation in opic eye may result in the light stimulus striking either eye. The patient fixed simultaneous different retinal areas.7 macular perception slides throughout. Routine orthoptic examination does not Where possible this test was performed blind, include a test capable of detecting small afferent but casual observation - as in cases of obvious or relative afferent pupillary defects. When anisometropia or strabismus, wearing of attempting to assess the latter, the swinging occlusion, comment on suppression - meant that flashlight test89 is used, but this has drawbacks the amblyopic eye was known to the examiner in for the examination of The child will some cases. children. http://bjo.bmj.com/ often look at the light causing constriction of the After the assessment of the pupillary reaction pupil due to the near reflex, which obscures the on the synoptophore the density of any response to light,'0 or the reaction may be suppression present was measured at the objec- blocked in excitable children." In addition other tive angle by dimming the rheostat in front ofthe disadvantages to the test include: confusion due non-suppressing eye until the 'suppressed' to hippus7 (one pupil being observed on the image could be seen and the rheostat number upswing and the other on the downswing), noted. The and were subjective objective angles on September 23, 2021 by guest. Protected copyright. presence of anisocoria,9 and the danger of using compared to elicit the type of retinal correspon- too bright a light, as an after image can keep the dence. pupils small and so prevent the pupillary Visual acuity was assessed (with the patient escape." 12 wearing refractive correction) by a linear test Pupillomotor changes have also been reported (Snellen or Snellen with key card). Some patients in suppression.'3 had their acuity tested with single optotypes The implication of the presence of a relative (Sheridan Gardiner), fixation by the Visuscope afferentpupillarydefectin diagnosis and manage- and contrast sensitivity by the American Optical mentofamblyopia has not beenfullydetermined. System (which is based on Arden's gratings). The purpose ofthis study was to discover factors The size ofpupilswasmeasured, as anyanisocoria common to amblyopes who display a defect with or more than 2 mm could make the testing of a view to ascertaining whether assessment of the pupillary responses inaccurate.9 pupillary response is of clinical value during the Further details were taken from the hospital treatment of amblyopia. records. These included: age at date of test, type of amblyopia, age at onset of strabismus, visual acuity prior to occlusion therapy (and test used), Methods and patients age at first occlusion, types of occlusion therapy Welsh School of Orthoptics, University To observe any asymmetry in pupillary response undergone, continuity of occlusion, best pre- Hospital of Wales, Heath the synoptophore was used with modification of vious visual acuity (if higher than on day of Park, Cardiff CF4 4XW the bright light source normally used for the testing), refractive correction, and fundus AY Firth production of after images. The light intensity examination. When any lesion of the fundus or Correspondence to: Alison Y Firth. was reduced by fitting neutral density filters media was present the patient was excluded. Accepted for publication (NDF) of 0-4 log units into the same rubber Since the study was of necessity conducted 7 June 1990 holder as each eye piece lens. The after image during normal clinical sessions, the selection of Pupillary responses in amblyopia 677 patients was haphazard. They were examined * RAPD Present Br J Ophthalmol: first published as 10.1136/bjo.74.11.676 on 1 November 1990. Downloaded from during their routine orthoptic examinations. o RAPD Absent Initially only patients with amblyopia were 601 (12) examined, but later all patients examined with no prior knowledge of whether or not amblyopia was present. No attempt was made to examine 50 patients under the age of 3 years. Nine children were followed up through occlusion therapy. A group of 25 children from a local junior ? 40 c0 school were used as controls. m9 z 30 Results Seventy six patients were examined with ages ranging from 3 years 2 months to 13 years 10 20 months. Sixty five had amblyopia, this being defined as any difference in linear visual acuity. The type ofamblyopia is shown in Table 1. 10 Of the patients with equal visual acuity four had previously had strabismic amblyopia which had responded to treatment, five had intermittent (0) or alternating deviations, and two had equal but Idiopathic Ansio Strab+Ansio Strab reduced visual acuity due to ametropic ambly- Type of Ambyopia opia. Figure 1 Relativefrequencies ofdifferent types of The pupillary responses in 72% of patients amblyopia: idiopathic, anisometropic (aniso), combination of strabismic and anisometropic (strab+aniso) and strabismic were examined blind. (strab) in patients with (shaded) and without (unshaded) a Of the 65 amblyopic patients 21 had a relative relative afferent pupilary defect (RAPD). Actual patient afferent pupillary defect in their amblyopic eye numbers shown in brackets. and two in their non-amblyopic eye. Of 25 controls tested a subtle defect was found in one of 26 without a defect were anisometropic, it was child. considered that this may have caused a bias. Contraction anisocoria is estimated to occur to Unfortunately the numbers were too small to an extent which is clinically visible in 5% of the analyse in the pure strabismic amblyopes, but in population." This may explain the finding of a a comparison of patients with anisometropia and defect in the control and non-amblyopic eyes. strabismus (Fig 2) the age at onset of the However, it could have been observer error. The strabismus was found by the Mann-Whitney U two patients with the defect in their non ambly- test to be significant at the p<005 level. opic eye were excluded from further consider- Various factors concerning occlusion were ation. considered: (a) age at first occlusion; (b) the http://bjo.bmj.com/ The type of amblyopia of the remaining 63 delay from onset of strabismus to start of patients was first considered (Fig 1). For statis- occlusion; (c) type and continuity (d) period of tical evaluation the idiopathic group was occlusion; and (e) time lapse from the last excluded because of the low number. By means occlusion to the date of testing. Of these, the of the likelihood ratio criterion the type of period ofocclusion (Fig 3) was the only significant amblyopia was shown to be significant at the p<0 02 level, and further grouping of patients Table 2 Presence or absence ofrelative afferent pupillary on September 23, 2021 by guest. Protected copyright. into those with or without anisometropia (Table defect (RAPD) in amblyopia involving anisometropia and 2) and those with or without strabismus (Table 3) amblyopia without anisometropia. Statistical evaluation showed only the difference in the former group given to be The amount significant (p<001). actual of Type ofamblyopia anisometropia, however, did not prove to be significant. Aniso and An accurate age at onset in those with strabis- Strab+Aniso Pure strab Total mus was given in 35 patients. The Mann- RAPD present 17 4 21 Whitney U test showed the age at onset to be at a RAPD absent 18 22 40 significantly younger age in those with a relative Total 35 26 61 afferent pupillary defect (p=0 0294). However, df= as eight out of the nine patients with a pupillary XL'=7-7297883, 1, p<OOl (2 tailed). defect also had anisometropia and only seven out Table 3 Presence or absence ofrelative afferent pupillary defect (RAPD) in amblyopia involving strabismus and amblyopia without strabismus.