Stray Light–Induced Multifocal Electroretinograms

Abstract

purpose. To evaluate the characteristics of stray light–induced response in multifocal ERG (mfERG) elicited by the stimulus falling on the disc.

methods. A patient with an enlarged optic disc (4 × 4 disc diameters of disc of normal fellow eye) and four normal volunteers served as subjects. The mfERGs elicited by different stimulus intensities (0.67–4.67 cd-sec/m²) were recorded from the patient, and the mfERGs obtained with stimuli on the enlarged optic disc. For comparison, full-field pseudorandom ERGs (ffprERGs) were also recorded in all subjects. The first-order kernels (K1) and the second-order kernels (K2.1) were analyzed.

results. A small and delayed K1 was recorded on the enlarged disc, but K2.1 was flat on the disc at all intensities. The implicit time of K1 at lower intensities was longer than at higher intensities. ffprERGs at very low intensities in the patient and normal subjects were similar to the mfERG on the disc (delayed K1 associated with flat K2.1).

conclusions. The responses elicited by stimulating the disc were delayed in K1 and flat in K2.1. Because similar ffprERGs were observed at very low intensities, it is likely that an optic disc with high reflectance scattered the stimulus light to create a weak full-field stimulus. Thus, care must be taken when focal lesions are investigated with mfERGs.

In the multifocal electroretinogram (mfERG) technique, ¹ the stimulus is made up of densely packed hexagons, each of which illuminates a focal retinal region to elicit a local ERG response. However, the focal stimulus light is scattered in the media and reflected from the ocular surfaces and may evoke stray light responses. The effect of stray light from focal stimuli has been noted and investigated extensively in earlier work, ^{2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23} especially for the responses obtained by focal illumination of the optic disc. ^{3 4 5 6 8 9 10 11 12 13 14 15 16 17 18 20 22} Because there are no photoreceptors on the optic disc, the responses recorded from stimulating the optic disc (ERGs from the disc) must be attributed to stray light. ^{5 6 13}  

It has been acknowledged that mfERGs can be recorded from local stimuli on the disc, and some possible sources of the stray light has already been reported. ^{1 24 25 26} However, the exact contribution of stray light to mfERGs remains to be determined. ²⁵ We recorded mfERGs in an eye with an enlarged optic disc in a case of disc coloboma, and their characteristics were investigated and compared with the full-field pseudorandom ERGs (ffprERGs) ²⁷ recorded from control subjects. 

Subjects and Methods

Subjects

A 32-year-old woman who had an optic disc coloboma with an enlarged optic disc in the right eye was studied (Fig. 1A) . The diameter of the enlarged disc was four times that of the disc in the normal fellow eye. Although the blind spot was enlarged and threatened the central field (Fig. 1B) , there were no subjective symptoms. From previous pathologic reports on this anomaly, ²⁸ and the absolute scotoma observed in the visual field corresponding to the enlarged optic disc, no functioning retina was present in this lesion. The corrected visual acuity was 20/20 (1.0 × −3.0 D), and full-field standard ERGs were normal. No other abnormality has been found. The fellow eye was completely normal. 

Four normal volunteers, a woman (29 years old) and three men (28, 33, and 40 years old) also served as subjects. The tenets of the Declaration of Helsinki were followed, and informed consent was obtained from all subjects before testing. 

Multifocal Electroretinograms

mfERGs were recorded with a visual evoked response system (VERIS Science, ver. 4.0; Electro-Diagnostic Imaging, San Mateo, CA). ¹ The recordings were performed under ordinary room light with the pupil dilated maximally. The stimulus was displayed on a monochrome monitor with a P4 (white) phosphor. An array of 37 densely packed hexagons stimulated the central 40° of the visual field (stretch factor was 13.18). Typically, we used an m-sequence rate of 75 frames/sec and cycle of 2¹⁴ − 1 steps or longer. Four stimulus intensities (0.67, 1.33, 2.67, and 4.67 cd-sec/m², corresponding to static measurement on the monitor screen of 50, 100, 200, and 350 cd/m², respectively) were used. The region surrounding the stimulus geometry on the monitor screen was illuminated at half the stimulus’s luminance. ²⁹ Responses were recorded with a Burian-Allen bipolar contact lens electrode. A camera/refractor (Electro-Diagnostic Imaging) was used for refraction and to monitor eye position and fixation stability during the recordings. ³⁰  

The signals were amplified (100,000×), bandpass filtered (10–300 Hz at half-amplitude) and digitized with a 1200-Hz sampling frequency. The first-order kernel (K1) and the first slice of the second-order kernel (K2.1) of the mfERGs were analyzed. 

Full-Field Pseudorandom ERGs

We used a built-in LED contact lens electrode (H2000; Kyoto Contact Lens, Kyoto, Japan) and an LED driver (CLS-10; Mayo, Aichi, Japan) with the visual evoked response system (VERIS; Electro-Diagnostic Imaging). ²⁷ The built-in LED was driven at 75 frames/sec m-sequence stimuli to give 1-ms flashes (up to 10.0 cd-sec/m²: 10,000 cd/m² × 1 ms). The reference electrode was attached on the ipsilateral temple. The collected signal was directly processed by the system with the same analysis condition as mfERG, except the steps of the m-sequence cycle (2¹³ −1) and the gain of the amplifier (20,000×). The contact lens illuminated a whole retina equally with LEDs inside and obtained pseudorandom ERGs from all subjects. The details of this technique has been reported. ²⁷ K1 and K2.1 were obtained from all subjects. 

Results

For the eye with the disc coloboma, the 23 temporal hexagons were imaged off of the optic disc but at least four nasal hexagons fell entirely on the optic disc (Fig. 2A) . We grouped and summed the responses from the 23 temporal hexagons and the 4 hexagons within the disc as retinal ERGs and disc ERGs, respectively (Fig. 2B) . The fundus in the left eye was normal, we assigned a symmetric region to retinal ERGs as the retinal ERGs in the fellow eye in the left eye (Fig. 2C) . 

The trace arrays (2.67 and 0.67 cd-sec/m²) are shown in Figure 3 . The retinal ERGs were almost as large as the corresponding focal ERGs recorded from the same regions of the fellow eye without a coloboma (Fig. 3B) . In contrast, the disc ERGs were reduced (K1, left column) and the K2.1 was flat (Fig. 3A , right column). 

The intensity-response series of the response density in nanovolts per degree squared of the retinal and disc ERGs are compared in Figure 4 . The K1 of disc ERGs were smaller and slightly delayed compared to K1 of the retinal ERGs. The second negative peak (approximately 42 ms) and the following wavelets (later than 50 ms) were clearly observed in the retinal ERGs but were missing in the disc ERGs. The K2.1s of the disc ERGs were flat for all parameters of the stimulus. The second-order kernel second slice and the third-order kernel first slice were also extinguished (not shown). 

The response densities (K1: amplitude between first negative and first positive peak, K2.1: between first positive and first negative peak) and implicit times (latency of first positive peak in K1) are plotted as a function of the stimulus intensity in Figure 5 . The closed and open symbols indicate the values of the retinal ERGs and disc ERGs, respectively. The K2.1s of the disc ERGs were below noise level and are shown in gray. Although retinal ERGs and retinal ERGs in the fellow eye showed some interocular difference, probably at least partially due to the refractive error ³¹ (right eye −3 D; left eye 0 D), the intensity-response characteristics of both in the patient matched the mfERGs reported in 20 normal eyes (K1 and K2.1). ³² The response density of K1 of the disc ERGs increased with increasing stimulus intensity, and, at its maximum, it was approximately 50% that of the retinal ERGs, with a stimulus intensity of 4.67 cd-sec/m² (Fig. 5 , top). 

The implicit times of K1 of the retinal ERGs were not greatly changed at all intensities, but that for the disc ERGs were markedly delayed at the lowest intensities (Fig. 5 , middle). The increase in the amplitude of K2.1 of the retinal ERGs was similar to that of K1 for the retinal ERGs (Fig. 5 , bottom), although the K2.1 of the disc was flat at all intensities. 

The ffprERGs were recorded through a monopolar built-in LED contact lens referred to the ipsilateral temple. Our preliminary recording showed that a monopolar Burian-Allen electrode referred to the temple induced almost identical waveforms to those elicited in bipolar recording, except the scale was approximately 20% larger. The ffprERGs recorded in the patient showed that both K1 and K2.1 increased proportionally as the stimulus intensity rose, for intensities between 0.1 to 10.0 cd-sec/m². However, for intensities less than 0.1 cd-sec/m², only a K1 was detected. The waveforms of ffprERGs elicited by dim flashes on weak background illumination (see legend for detail) are shown in Figure 6 (top). They mimicked the disc ERGs shown in Figure 4 (bottom) very well, as shown by the absence of the second negative peak and wavelets, delay of the positive K1 peak with dim stimuli, and flat K2.1 response. These particular waveforms were also obtained from all control subjects in the ffprERG with minor intersubject variability (Fig. 6) . 

Discussion

Rationale for Recording mfERG from Enlarged Disc

The optic disc coloboma ²⁸ consists of the enlarged optic disc and the coloboma surrounding it. It is often difficult to differentiate between them. Unlike the optic disc, the region of the coloboma has an intercalary membrane that extends over the choroidal defect, ²⁸ and, in our case, the intercalary membrane did not appear to contribute to mfERGs under our stimulus conditions. 

The characteristics of eyes with a scleral defect are similar to those in our case. Miyake et al. ²³ reported that focal light stimulations (approximately 30 cd/m²) of a macular coloboma produce no focal ERGs. Therefore, our patient with a large optic disc coloboma, normal full-field ERGs, and normal visual acuity, is a good model for investigating the effects of stray light on the mfERGs. 

Nature of the ERGs from the Disc

In the mfERGs in normal eyes, responses have been recorded from stimuli falling on the optic disc, although they are smaller and often delayed compared with the surrounding mfERGs. It has been argued that the responses are present because the images of the hexagons do not fall completely on the optic disc. ^{1 25 33} This suggestion may explain the smaller amplitude of the response from the disc, but not the delayed peak time. When Sutter and Tran ¹ introduced the mfERG technique, they suggested that light-scattering may contribute to the responses. Hood ²⁵ and Hood et al. ²⁴ were also aware of the stray light effect. However, the difficulty in mapping the stimulus hexagons on an optic disc has hindered confirmation of this hypothesis. ^{1 25}  

Our results from an enlarged disc, on which several hexagons were completely imaged, provided evidence of the source of the ERGs in the disc. Because of its high reflectance, the white optic disc strongly reflects the stimulus light to other regions of the retina. ^{8 9 10 25 26} The disc in the patient was remarkably large and pale so that the pure stray light effect was visible; however, any region of ocular fundus reflects the stimulus light to some extent. 

Two components of stray light have been noted. ^{5 13} Besides the back-scattered (reflected) light ^{8 9 10 25 26} described earlier, any optics of the eyeball scatters the stimulus light and generates local scattered (defocused) light ^{5 13} in various ways. Although in many studies ^{2 3 4 6 7 11 12 14 15 16 17 18 19 20 21 23} the two components were not discriminated, we assume that the response from the optic disc ^{3 4 5 6 8 9 11 13 14 15 18} was mainly caused by back-scattered (reflected). Only 23% of the cones are present in the central 50° of the retina, ³⁴ and so the remaining 77% of cones can contribute only to the stray light–induced responses. Although the stray light is weak compared with the stimulus itself, it covers a broad area. 

A close resemblance of the disc ERG to the ffprERGs supports this explanation (Fig. 4 , bottom; Fig. 6 ). ffprERGs (K1 and K2.1) in the same eye elicited by a stimulus intensity of 0.01 cd-sec/m² were very similar to the disc ERGs at an intensity of 4.67 cd-sec/m², and ffprERGs at an intensity of 0.005 cd-sec/m² were very similar to the ERGs in disc at an intensity of 0.67 cd-sec/m². In all likelihood, the light reflected by the enlarged disc created a weak full-field stimulus to elicit the disc ERGs. 

Characteristics of Stray Light–Induced Responses

It was interesting that the K2.1 of the disc ERGs was flat. A flat K2.1 yields a particular waveform of K1, because the late period of the K1 response reflects the overlapped higher-order kernels. ^{25 35 36 37} It was also interesting that dim full-field pseudorandom stimuli did not elicit K2.1s (Fig. 6) . Our results indicate that K2.1 had different intensity-response characteristics than K1 at low intensities. Thus, it appears that dim lights, such as stray light, do not affect the local adaptation level of the retina sufficiently to produce higher-order kernels. 

Consequently, we can assume that the K2.1 responses are not affected by the stray light effect. These findings suggest that we must be careful about reflections from focal lesions in mfERG analyses. The relative reduction of K2.1 to K1 or delayed K1 with reduced K2 is thought to indicate an abnormal adaptation mechanism. ^{27 38 39 40 41 42} Clinical studies ^{38 39 41} have attributed the reduction of K2.1 to inner retinal layer dysfunction in patients with glaucoma, ³⁸ diabetic retinopathy, ³⁹ or branch retinal artery occlusion. ⁴¹ Neurotransmitter experiments ²⁷ have shown that K2.1 includes mainly the components originating from the inner retinal layers in rabbits. The K2.1-to-K1 ratio was tested to diagnose glaucoma ⁴⁰ and congenital stationary night blindness. ⁴² Our results indicate that these findings could be influenced by the responses elicited by stray light, especially in the lesions with high reflection. 

In addition, we can assume K2.1 components are practically stray light effect free, giving the K2.1 analysis an advantage in localizing lesions. 

Affect of Stray Light–Induced Responses mfERGs in Normal Retina

Our results suggested that lesions with high reflectance may elicit stray light–induced responses. The question then arises as to whether the mfERGs in the normal retina may contain the stray light–induced responses. It has been suggested that lowering the luminance reduces the stray light effect, ^{14 15 22 25} yet it has also been reported that the contribution of stray light with dim flashes is larger in the rod mfERGs. ²⁴ The amplitude of K1 and K2.1 in the pseudorandom ERGs obtained by full-field stimuli, which elicit no stray light effect, ^{43 44 45} increases equally with the increase in intensity with the stimulus intensity more than 0.1 cd-sec/m². Therefore, the ratio of K2.1 to K1 was nearly constant. For the mfERGs in our patient, the amplitude of K1 and K2.1 increased equally between 0.67 and 2.67 cd-sec/m² (Fig. 5 , top, bottom). If K1 of the retinal ERGs contained substantial stray light–induced responses, the amplitude of K1 increased more than that of stray light–free K2.1. Therefore, the intensities of 1.33 and 2.67 cd-sec/m² (corresponding to 100 and 200 cd/m², respectively) recommended in the International Society for Clinical Electrophysiology of Vision (ISCEV) guidelines ²⁹ seem reasonable to investigate focal lesions with a normal appearance. 

In some reports on focal ERG, ^{2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23} an ERG response was recorded in the disc ^{3 4 5 6 8 9 11 13 14 15 18} , but in others, no response was observed from the disc ^{10 11 12 14 15 16 17 20 22} or a coloboma. ²³ It has been emphasized that the background illumination suppressed the stray light effect. ^{10 11 14 15 17 20 22 23} Whether the ERG is recorded on the disc or not depends on the combination of the intensity of focal stimulus and background illumination ^{11 14 15 22} and the signal-to-noise ratio. The mfERG is extracted by the cross-correlation process ¹ corresponding to temporal averaging of thousands of presentations (a cycle of 2¹⁴ −1 steps generates 8191 flashes), so that its signal-to-noise ratio is high enough to visualize very tiny responses. This may result in small but certain responses observed on disc even with illumination in the surrounding region on the monitor and ordinary room light. 

In summary, our results indicate that the mfERGs recorded from an enlarged disc were delayed for K1 and flat for K2.1, and these responses were elicited by stray light from a wide area. Therefore, we should be aware of those effects when we analyze responses from a focal lesion in mfERGs. 

 Supported by Grants 13877293, 13470372 and 14571692 from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

 Submitted for publication May 31, 2002; revised August 26, 2002; accepted September 20, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

(A) Appearance of the posterior pole of the right eye of the patient, a 34-year-old woman, showing disc coloboma with a remarkably enlarged optic disc (four times the diameter of the disc in the normal fellow eye). (B) Goldmann perimetry of the eye. The blind spot was enlarged and threatened foveal vision.

Figure 2.

 

(A) Stimulus geometry superimposed on the fundus image. The image has been flipped vertically to correspond to the field view of mfERGs. (B) Twenty-three hexagons were projected off of the optic disc and at least four hexagons were projected entirely on the optic disc and, if any, coloboma. The mfERGs from the 23 hexagons were grouped and averaged and referred to as the retinal ERGs; the mfERGs from the four hexagons were grouped and referred to as the disc ERGs. The trace arrays were recorded with a 2.67-cd-sec/m² (200-cd/m²) stimulus intensity. (C) In the left eye, the mfERG from a symmetric region of retinal ERGs in the right eye was assigned the retinal ERGs in the fellow eye.

Figure 3.

 

(A) Trace arrays of mfERGs showing the first-order kernel (K1: left column) and the second-order kernel first slice (K2.1: right column). Top and bottom rows were obtained with 2.67 and 0.67-cd-sec/m² stimulus intensities, respectively. Shaded regions: retinal ERGs and disc ERGs assigned in Figure 2 . (B) The left eye.

Figure 4.

 

Group-averaged waveforms scaled in response density, comparing the retinal ERGs and the disc ERGs extracted from mfERGs, as shown in Figure 2 . Both K1 and K2.1 of the retinal ERGs increased with higher stimulus intensities (top). In contrast, the disc ERGs had a particular K1 waveform. Its implicit time was delayed at low intensities, and K2.1 was flat (bottom). The absolute scale is also shown only for the disc ERGs, to compare the ffprERG shown in Figure 6 .

Figure 5.

 

The amplitudes and the implicit times extracted from retinal and disc ERGs are plotted as a function of the stimulus intensity. Top: the amplitude (scaled in response density) of K1. Middle: The implicit times of K1. Bottom: the amplitude (scaled in response density) of K2.1.

Figure 6.

 

ffprERG obtained from an eye with disc coloboma and control eyes. All traces were obtained from dim pseudorandom flashes on weak background illumination that was a half of stimulus luminance. For example, 0.01-cd-sec/m² flashes were presented with a 10-cd/m² × 1-ms flash on a 5-cd/m² background illumination. In keys, 40y and similar entries represent the age of the subject.

The authors thank Erich E. Sutter who provided the basic inspiration for this research, Donald C. Hood who gave us kind suggestions, Marcus A. Bearse, Jr, for continuous encouragement, and the patient for her willing cooperation.