The Brückner Test Revisited

Core Messages

■The Brückner test is useful to detect various amblyogenic disorders. After a short training, every physician can perform the test.

■The test as originally described consists of four elements to observe: (1) the position of the first Purkinje images (corneal light reflexes), (2) the fundus red reflex in the pupil, (3) pupillary light reflexes, and (4) any movement of the eyes when illumination alters from one eye to the other.

■Asymmetry in corneal light reflexes on both eyes may indicate strabismus. However, small deviations are not reliably detected, and asymmetry can also be caused by di erent angle kappa in both eyes.

■Performance of the red reflex test requires a direct ophthalmoscope. Substitution by an otoscope, indirect ophthalmoscope, or any other light source causes loss of test validity.

■The red reflex test allows for detection of refractive error, strabismus and organic disorders such as opacities of the optic media and distinct pathologies of the fundus.

■Media opacity is easily detected at a test distance of 0.3 m and less, examining each eye separately.

Any optically relevant opacity will be apparent by a shadow in the red reflex.

■Detection of refractive error can be improved by extending the test distance up to 4m and observing the brightness of the red reflex in both eyes simultaneously. While usually at a distance of 1m, the red reflex is brighter in the more ametropic eye, the reflex in this eye becomes increasingly darker with increasing test distance. With increasing test distance, myopia and hypermetropia, which are not compensated by accommodation, cause significant dimming, and anisometropia causes increasing asymmetry.

■The test sensitivity to detect microstrabismus by asymmetric fundus red reflex is low.

■Testing pupillary light reflexes is recommendable to assess visual a erence, pupillomotor e erence and pupil responsiveness. It is hardly suitable to diagnose or exclude amblyopia and amblyogenic disorders.

■Testing for fixation movements caused by switching illumination from one eye to the other is similar to the cover test. Data on diagnostic validity of this procedure are lacking.

Amblyopia is estimated to a ect approximately 2–5% of the population in Western countries and is a significant preventable cause of vision loss in children and adults [1–8]. Amblyogenic risk factors include ptosis, media opacity, fundus pathologies, strabismus and refractive error [9–11]. When these risk factors are detected at an early age, amblyopia can be prevented or minimized more e ectively [3, 12–14]. One significant limiting factor of most amblyopia screening programs is the reliance on the subjective responses of the child being tested.

Early detection of amblyopia and amblyogenic factors requires objective methods that are independent of any verbal response of the child. Refractive error and strabismus are the most frequent causes of amblyopia. So, methods are necessary that indicate ametropia and strabismus with a high sensitivity and specificity. Refractometry or retinoscopy in cycloplegia is the most reliable way to detect and measure ametropia in childhood. However, this requires experience of the examiner and the possibility to perform both cycloplegia and measurement. These

conditions as well as parental readiness are often lacking. Non-cycloplegic photorefractive screening is not a tantamount substitute of refractometry in cycloplegia [15, 16]. Besides, the technical equipment is relatively expensive,

9and therefore hardly any paediatrician or general practitioner performs photorefractometry. Even the Brückner test is not routinely used by paediatricians, although preconditions for performance are ideal and the test is recommended for paediatric screening examinations in Germany [17]. The Brückner test is a readily available screening tool that can be used with newborns, infants and preverbal children by non-ophthalmologists [18, 19]. The test requires not more than a direct ophthalmoscope and only few seconds for performance.

9.1.2Brückner’s Original Description

In 1962, Roland Brückner (1912–1996), an ophthalmologist in Basel, Switzerland, reported on ‘Exact strabismus diagnostic in ½- to 3-year-old children by a simple procedure, the “transillumination test”’ [18]. Brückner illuminated both pupils from a distance of 1 m and assessed the following criteria:

■Position of first Purkinje images relative to the pupil

■Colour of the fundus red reflex in the pupil

■Size and constriction of the pupils

■Eye movements with and illumination of the pupils

Assessment of the first two criteria requires simultaneous illumination of both eyes, whereas assessment of the following two criteria requires alternate illumination. Three years later, Brückner added an article on ‘Practical exercises with the transillumination test for early diagnosis of strabismus’, emphasizing the essential component of the test, which is the assessment of the red reflex of the fundus when the pupil is lighted and viewed with a direct ophthalmoscope [19]. This particular component was new concerning strabismus diagnostic and in the aftermath called Brückner test in the closer sense. It has also been called the Brückner reflex [3, 20].

9.2Corneal Light Reflexes (First Purkinje Images)

Assessment of the first Purkinje images in the two eyes allows for more exact strabismus diagnostic than mere assessment of the position of the cornea within the palpebral fissure. The latter depends on the configuration of

the lids and the root of the nose. In infants and toddlers, as well as in Asians, epicanthus which is nasally covering the lid fissure can be suggestive of esotropia.

9.2.1Physiology

Purkinje described that when the eye is being illuminated by an examination light, reflexes appear from the corneal surface, the corneal endothelium, and both the anterior and posterior surface of the lens. The first Purkinje image coming from the corneal tear film is brightest. Usually it appears slightly nasally of the centre of the cornea and the pupil, when the eye is fixating a light source which is held directly below the pupil of the observer. Slight eccentricity of the corneal light reflex is caused by the di erence between the visual line and the pupillary axis, the angle k, which is similar to the angle g [21]. When the eye turns in a distinct direction, the position of the corneal light reflex relative to the pupil will shift to the opposite direction. Conjugate gaze movements induce parallel shift of the images in both eyes. This causes asymmetry in the two images, if their positions were symmetric at first. For instance, right gaze induces nasal shift of the image in the right eye and temporal shift of the image in the left eye. The same will happen, when the light source is moved to the righthand side from the observer’s point of view or when the observer assesses the image position from left-hand side beside the light source. Non-conjugate eye movements or manifest strabismus cause a non-parallel shift or position, respectively, of the images on both eyes. For instance, when the left eye fixates the light and the right eye is esotropic, then the first Purkinje image on the right eye will be temporally dislocated. So, this method in principle allows for detection of strabismus.

The idea to measure squint angles by using corneal light reflexes arose at the end of the nineteenth century [22, 23]. Hirschberg assumed that 1-mm shift of the corneal light reflex corresponded to an angle of 7° by which the eye is turned [22]. At the end of the twentieth century, empiric studies proved that within the range of small and moderate deviation the correct ratio is 12°/mm [10, 24, 25]. Nevertheless, up to the twenty-first century, the wrong ratio of 7°/mm is still wide spread. Recognition of asymmetry in the Purkinje images can be improved by evaluating photographs [26]. In laboratory trials, photographic Hirschberg testing was e ective in approximately 80% of cases in detecting a deviating eye in strabismus of about 5 prism dioptres [27]. Regarding more accurate diagnostic, the alteration of relative position of the first and the fourth Purkinje images due to deviation of the visual axis have

been studied [11, 28–32]. However, the fourth Purkinje image is not visible clearly enough by performing the Brückner test.

9.2.2Performance

Assessment of the corneal light reflex for symmetry on both eyes requires a small light source, which must be fixated by the patient. To avoid glaring the patient, the light should not be too bright. The observer compares the position of the corneal reflex images in the two eyes in relation to the pupils. Physiologically, the images appear approximately 0.5 mm nasal to the centre of the pupil. The eccentricity depends on the individual angle k. The images may be better visible when the observer looks above the ophthalmoscope. Then the pupils appear black and there is more luminance contrast of the images. If the iris is dark brown with low contrast to the black pupil, looking through the ophthalmoscope is advantageous. Favourite test distances are around 0.5 m. Closer test distance may cause defence in children and also adequate convergence might not be warranted. Larger distance makes it di cult to detect small asymmetry.

9.2.3Shortcomings and Pitfalls

Fig. 9.1 Corneal light reflexes in a 12-month-old girl. In this case, asymmetry of the corneal light reflex between both eyes is caused by flashlight position beside the objective of the camera. So, the image of the flashlight on the right eye is more and the image on the left eye is less nasally decentred. At 9 o’clock in front of both pupils, images of a window

Summary for the Clinician

■Evaluating the corneal light reflexes in both eyes for symmetry allows to detect manifest strabismus and to estimate its size. Exclusion of strabismus is impossible because slight asymmetry corresponding to small squint angle can hardly be recognized and asymmetry in the angles k in both eyes can both, simulate or mask strabismus. Bias occurs when the patient fixates a point beside the examination light or when the light is not on the examiner’s visual line.

9.3Fundus Red Reflex (Brückner Reflex)

False-negative findings are likely in case of small squint angle. Since misalignment of 6° corresponds to not more than 0.5 mm asymmetry in the position of the corneal light reflexes, it is evident that small angle strabismus can hardly be identified by this method. Asymmetry in the angle k between both eyes can veil strabismus.

Ectopia and anomalies of the pupil have to be considered. False-positive finding of strabismus can be caused by parallel shift of the reflex images in the two eyes when the light is horizontally displaced. The light source must be exactly beneath (not beside!) the visual axis of the observer’s fixating eye. Severe bias occurs when the light is hold under one eye while the other eye is fixating: Taken the angle k were equal in both eyes, the interpupillary distance were 60 mm, and the examination distance were 0.5 m, then the resulting asymmetry would correspond to 12°. A similar mistake occurs by evaluating flashlight photographs, which were recorded with the flashlight beside the objective (Fig. 9.1). With the flashlight coaxially or above the objective, this bias can be avoided, but it cannot be assured that the child was really fixating the camera [33].

Performing the ‘transillumination’ test requires a direct ophthalmoscope. Looking through the ophthalmoscope, the examiner can see the patient’s pupil shining red, caused by the light reflected by the choroid and the retinal surface of the eye. The fundus reflex was also called Brückner reflex [3, 20]. Colour and brightness of the fundus reflex depend on brightness of the examination light, consistence and refractive quality of the optical media, pigmentation of the fundus and refractive state of the eye. Any opacity of the optic media causes an abnormally dark or lacking red reflex in the region of the opacity. Slight nuclear cataract may be visible by a darker ring, which is caused by the equator of the nucleus (Fig. 9.2). Posterior pole cataract causes a black shadow in the centre of the pupil. Frequently, a very small shadow is visible nasally belowthecentreof thepupilasthecorrelateof Mittendorf’s spot. With eye movement these shadows move to the opposite direction within the pupil while shadow caused by corneal opacity or anterior cataract will move to the same direction. An examiner who is familiar with the Brückner test will probably detect every optically relevant cataract, albeit we are not aware of any scientific study on

the sensitivity of the Brückner test to detect media opacity. Visualizing media opacity and pathologies of the fundus the Brückner reflex is extremely important for paediatricians, general practitioners and others who are not equipped to perform slitlamp biomicroscopy and indirect ophthalmoscopy. Abnormally, bright, white or dark fundus reflex can also be caused by the optic nerve head and by pathologies of the fundus, such as coloboma, retinoblastoma, toxoplasmosis scars and medullated nerve fibres.

When the patient takes up central fixation of the ophthalmoscope light, there is normally a constriction of the pupils and dimming of both fundus reflexes [18]. By interfering with this dimming phenomenon, manifest strabismus and anisometropia can produce asymmetry in the brightness and colour of the fundus reflexes in both eyes. Brückner stressed the point that strabismus could be reliably detected by this asymmetry. Traditionally, the deviated or more ametropic eye was described to have the brighter reflex [9, 18]. Regarding ametropia, however, examination distance is a decisive factor. At larger distance, the more ametropic eye yields the darker fundus reflex [34].

9.3.1Physiology

Examination of the fundus red reflex can roughly be compared with direct ophthalmoscopy performed at a large distance so that only very small part of the fundus is visible. Provided central fixation of the patient, the fundus red reflex represents the patient’s fovea. To explain the dimming of the red reflex when the patient takes up fixation, Brückner discussed various factors [18]. Pupillary constriction, di erent reflectivity of the central and peripheral retinal surface and accuracy of accommodation were assumed to be the major causes of dimming and change in colour [18, 35–37]. Backscattering of the light by the retinal nerve fibre layer proportional to the thickness of the layer and changes arising from variation in retinal pigment epithelium density, with the retina displaying the characteristics of a di use reflector, were further discussed but not as primary factors of dimming [35]. Mere pupillary constriction does not explain asymmetric dimming due to strabismus, but it may amplify e ects of defocus and retinal reflectivity. Brückner’s idea that di erence in reflectivity between the central and para-central or peripheral retinal surface contribute to

the dimming phenomenon was refreshed by Roe and Guyton who described specular reflection of the retina from the internal limiting membrane that changes slope with ocular rotation [35, 36]. The fundus reflex is not solely caused by reflection from the choroid and the retinal pigment epithelium but, to a minor part, also by reflection from the retinal surface. If significant light were reflected from the internal limiting membrane of the retina, the slope of the foveal pit would reflect enough light away from the pupil. Because this part of light would not be reflected back to the observer, the red reflex would appear darkened [35, 36]. Misalignment of one eye with light being reflected from the para-foveal retinal surface, which is rather perpendicular to the direction of the incoming light, increases coaxial reflection and thus the brightness of the fundus reflex (Fig. 9.3).

This might also explain the lack of dimming in newborns and young infants as a consequence of development of the foveal pit.While most infants 8 months of age and older show dimming of the fundus reflexes in both eyes occurring with central fixation, neonates and most infants younger than 2 months of age do not show dimming of the fundus reflex with fixation and between 2 and 8 months of age up to 28% of infants have asymmetric dimming of the fundus reflexes in the two eyes [9]. So, in newborns and young infants, asymmetry may represent a normal stage of development and symmetry does not exclude strabismus.

Another mechanism might be o -axis aberration resulting in poor image formation on the retina. Roe and Guyton believed the fundus reflex would appear darker in an eye that is fixating and focusing on the ophthalmoscope light because the light source in the ophthalmoscope and its retinal image are conjugate to one another.

If an eye is deviated, o -axis optical aberrations will decrease the conjugacy of the ophthalmoscope light and the retina. If the fovea is not exactly conjugate to the light source, the light from the retina spills passed the light source into the examiner’s eye, increasing the brightness of the red reflex [35, 36]. This hypothesis might fit with the observation that – at the traditional examination distance of 1 m – the fundus red reflex in the (more) ametropic eye is usually brighter compared to an emmetropic eye. The hypothesis corresponds to the assumption that accuracy of accommodation is one reason of dimming.

Foveal dimming of the red reflex allows for sensitive discrimination between subsequent central and eccentric illumination of the same eye. Dimming occurred in 97.2% of trials with fixation of the light compared with fixation of a target between 2.5 and 10° beside the light, regardless of the angle of eccentricity. This rate did not decrease when the pupil was dilated by mydriatic eye drops (Gräf et al., MS in preparation). However, the static inter-ocular difference in the reflexes due to strabismus was less apparent. In young adults, simulated esotropia with squint angles up to 5° was detected in not more than 62%. The deviated eye was identified by the brighter red reflex in 48%. Esotropia of 7.5 and 10° was detected in 85 and 97% with identification of the deviated eye in 75 and 86% (Table 9.1). To achieve these rates, very discreet red reflex asymmetry was considered. The rate of false-positive findings was 36% (Gräf et al., MS in preparation). These results confirm prior findings [38]. When esotropia of, for example, 8 prism dioptres was simulated by fixating a near target, not more than two thirds of strabismus conditions were detected [27]. One might argue that these were only laboratory studies, but an increase in sensitivity and specificity in young children compared with highly cooperative

adults is rather unlikely. Strabismus detection will hardly improve by extending the test distance, except indirectly, by detection of anisometropia which frequently accompanies esotropia [35]. There might be a chance to improve test sensitivity and specificity by using a short-pass filter that blocks the reflexes coming from the retinal pigment epithelium and the choroid and thus augments asymmetry caused by asymmetric light reflection from the internal limiting membrane.

Considering optical basics, examination distance must be an essential factor influencing the red reflex in case of refractive error. Uncorrected ametropia causes defocus of the retinal image of the light source. On the way back to the observer, this image is projected through the pupil. A myopic eye focuses the light beams at the far point of the eye. Beyond the far point, the light bundle is divergent. In case of hypermetropia, which is not compensated by accommodation, the light beams depart the eye as a primarily divergent bundle. With increasing

distance between the observer and the patient, the portion of the reflected light bundle reaching the observer’s pupil decreases. So, when the observer moves backwards, the brighter reflex, which at a distance of 1 m, usually corresponds to the (more) ametropic eye, becomes darker (Fig. 9.4) [34]. The test sensitivity to detect unilateral refractive error by the weak reflex in the ametropic eye at a test distance of 4 m is better compared with the traditional test at a distance of 1 m or less [30]. Using a direct ophthalmoscope, unilateral myopia of 1–4 diopters was detected in 60–82% of trials at 1 m but in 100% of trials at 4 m (Table 9.2). Unilateral hypermetropia of 1–4 diopters was detected in 34–80% of trials at 1 m but in 52–98% of trials at 4 m. Compared with experts, results of students were weaker at 1 m but equivalent at 4 m [34]. The low rate of false-positive findings shows that rather discreet asymmetry was not considered pathologic in that study, in contrast to the study on simulated strabismus, (Fig. 9.5).