Multifocal Electroretinograms in Cases of Central Areolar Choroidal Dystrophy

Abstract

purpose. To study multifocal electroretinograms (mfERG) in patients with early-stage central alveolar choroidal dystrophy (CACD) with well-demarcated atrophic areas.

methods. Eight eyes of eight patients with CACD (ages, 47–67 years) and 20 normal control subjects were examined. The first- and second-order kernels (K1 and K2) were extracted from the responses elicited by 61 standard hexagonal elements of a visual response imaging system. The amplitudes and peak times of the focal responses at various retinal eccentricities were studied.

results. The amplitudes of K1 were reduced in the visibly atrophic areas, and they were also decreased in areas with no visible atrophy. The peak time was slightly delayed in many loci, but the delay was not as long as that in congenital stationary night blindness or diabetic retinopathy. The amplitude of K2 was very small in the central and peripheral areas, but the K2/K1 ratio in both areas was not significantly reduced, compared with that in normal subjects.

conclusions. Although the atrophic area was ophthalmoscopically well demarcated in patients with CACD, the abnormality of retinal function extended beyond the borders of the ophthalmoscopic and angiographic lesions. The retinal dysfunction outside the atrophic areas suggests a centrifugal progression of the disease, and abnormal K2 and K1 with preserved K2/K1 ratio are consistent with a presynaptic mechanism for the retinal dysfunction in this disease.

Central areolar choroidal dystrophy (CACD), originally reported by Nettleship in 1884, ¹ is a rare macular dystrophy that is characterized by the development of fine, mottled, depigmented retinal pigment epithelium (RPE) in the macula. The round or oval macular lesion shows a distinct boundary and sometimes includes small areas with normal appearance inside the lesion. Although dominant and recessive inheritances have been reported, most cases are sporadic. ² The associated findings include a central and/or paracentral scotoma and normal peripheral visual fields. Fluorescein and indocyanine angiograms reveal RPE atrophy and various degrees of loss of the choriocapillaris. Full-field electroretinograms (ERGs) are normal in the early stages, but may become abnormal in the advanced stages. ^{3 4}  

The technique for recording multifocal ERGs (mfERGs), based on an m-sequence stimulation, allows a rapid recording of focal ERGs from many retinal locations in the posterior pole of the eye. ^{5 6} Recently, cellular contributions to human mfERG have been extensively investigated, ^{7 8} and several studies have shown the usefulness of this technique for evaluating the distribution of cone dysfunction in retinal diseases. ^{9 10 11} It has been reported that the amplitude and the peak time of the first-order kernel (K1) of the mfERGs can be used to monitor retinal function in different kinds of retinal dystrophies. ^{9 12 13}  

The study of K1 has been especially useful in eyes with macular diseases, although the correspondence between the areas of retinal dysfunction and the areas with depressed K1 is not always high. Although an abnormal K1 is localized only in the central area in eyes with a macular hole ¹⁴ and North Carolina macular dystrophy (Paliga J, Szlyk JP, Seiple W, ARVO Abstract 386, 2001) the retinal dysfunction extends over the central lesion in the early stages of central serous chorioretinopathy (CSR) ¹⁵ and in occult macular dystrophy (OMD). ¹⁶  

We extracted the first- and second-order kernels (K1 and K2) of the mfERG from eight eyes of eight patients with CACD (not in the advanced stage) to determine the correspondence between the areas of retinal dysfunction and the lesions that were visible ophthalmoscopically. Because the visible lesion in CACD has a distinct border, image processing allowed us to make an accurate comparison between the responses inside and outside the lesion. 

Subjects and Methods

Subjects

Eight eyes of eight patients with CACD, ages 41 to 71 years (mean, 56.2), were recruited from the Department of Ophthalmology, School of Medicine, Nippon University. The diagnosis of CACD was made by the characteristic findings in color fundus photographs, fluorescein and indocyanine green angiograms, and other clinical findings (the onset of symptoms, slow progression, no history of chloroquine, normal dark adaptation, normal color vision, and normal peripheral field). Patients with secondary atrophy, such as age-related macular degeneration, were excluded. Stargardt disease was excluded by its earlier onset and typical angiographic findings (silent choroid). The inclusion criteria were good visual acuity (better than 0.7), normal electro-oculogram, normal peripheral field (Goldmann or Humphrey perimetry; Zeiss-Humphrey Systems, Dublin, CA), no significant cataract, no other ocular disease, no choroidal sclerosis, and good foveal fixation. The patients with disease at advanced stages were not included, because they could not fixate during the recordings. All patients had well-demarcated RPE and/or choriocapillaris (CC) atrophy. A summary of the clinical characteristics of the patients is provided in Table 1 . Figure 1 contains the fundus photographs of the patients. 

Full-field scotopic and photopic ERGs were normal, except for a slightly delayed peak time of photopic b-wave in patient (P)7. K1 and K2 were extracted from the mfERG recorded from one randomly selected eye of each patient. 

Twenty control subjects ranging in age from 40 to 66 years (mean, 55.2) and with no known ocular abnormalities were also tested. Informed consent was obtained from all subjects before their participation. All experiments were performed in compliance with the tenets of the Declaration of Helsinki. 

Multifocal ERGs

Stimulation.

Multifocal stimulation and analyses were performed with the visual response imaging system (VERIS Science 4.1; Electro-Diagnostic Imaging, San Mateo, CA; and Meiyo Corp., Aichi, Japan). The stimulus was displayed on a high luminance monochrome CRT with P4 (white) phosphor. A stimulus array of 61 densely packed hexagons covered the central 50°. To achieve approximately equal signal-to-noise ratios at all locations, the size of the hexagons was appropriately scaled with eccentricity. ⁴ Within each frame of the CRT, each stimulus hexagon either flashed at an intensity of 2.67 cd · s/m² or remained dark (below 0.01 cd · s/m²) and modulated at an m-sequence (2¹⁵ − 1 steps). A frame rate of 75 per second resulted in a net record length of 7 minutes 17 seconds. For the patient’s comfort, the recordings were acquired in eight segments of approximately 27 seconds’ duration. 

A camera/refractor (Electro-Diagnostic Imaging) was used to refract the subject’s vision and to monitor eye position, the alignment of the contact lens electrode on the pupil, and fixation stability during recording. If fixation was lost or misalignment occurred, the segment was interrupted, discarded, and rerecorded. ¹⁷  

Recording and Analysis.

Retinal responses were recorded with a Burian-Allen bipolar contact lens electrode, amplified (50,000 gain) and band-pass filtered between 10 and 300 Hz. A notch filter was not used. K1 and K2 were extracted with a fast m-transform algorithm. A single cycle (1 iteration) of artifact removal ⁵ was applied to each record to minimize the effects of occasional blinks or slight eye movements. Spatial averaging was not used. 

We measured the implicit times and the amplitudes of the largest peaks of K1 and K2 responses, as shown in Figure 2A . The hexagons were divided into five concentric rings, with the single hexagon in the center being ring 1 (Fig. 2B) . For analysis, the individual responses in each ring were summated, and these summed responses for each ring were compared between patients and control subjects by the Mann-Whitney U-test. 

Image Processing for Assessment of Lesion.

In all patients, the CACD lesion (RPE atrophy and/or loss of the choriocapillaris) had a distinct border. Using flipped photographs, we could identify the relationship between each hexagon and the visible atrophy. The fundus photograph of P4, flipped to correspond to the visual field, is shown in Figure 3 . 

In this patient, the central elements were well within the visible atrophy, and the elements around it fell in both atrophic and nonatrophic areas. The more peripheral elements were outside the visible atrophic area. 

Results

First-Order Kernel

The 61 K1s extracted from the local responses from a normal subject and from eight patients with CACD are shown in Figure 4 . The responses in the central areas were attenuated. The amplitudes (Fig. 5A) and peak times (Fig. 5B) of the responses from the five rings are plotted in Figure 5 . 

The open symbols represent the responses from the areas without visible atrophy, the closed symbols represent those from the areas within the visible atrophy, and the gray symbols represent those from the areas overlapping atrophic and atrophy-free areas. The bars represent the normal range (mean ± SD) of the age-matched control. In all the responses, the amplitude was reduced regardless of whether the response originated from an area with or without visible atrophy. In some of the responses, the peak time was delayed, but in others it was normal. 

Second-Order Kernel

The 61 K2s extracted from the local responses of a normal subject and from eight patients with CACD are shown in Figure 6 . The K2 responses in the central areas, overlapping the atrophic area, appear to be deteriorated. Because the amplitude of the K2 responses was very small in the central areas, we averaged the responses in R1 to R4 and compared them with those in normal subjects. The amplitudes (Fig. 7A) and peak times (Fig. 7B) of the responses from R1 to R4 and R5 are plotted in Figure 7 . 

The amplitudes were significantly reduced in both R1 to R4 and R5, and the peak time was delayed in R5. The ratio of the amplitude of K2 to K1, a way to compare the two kernels, ^{18 19} in R1 to R4 and R5 are shown in Figure 7C . In both areas, the ratios were not significantly reduced compared with those of normal subjects. 

Discussion

K1 in CACD

An abnormal K1 in advanced CACD has been reported, ^{20 21 22} but K1 in early CACD has not been studied. In the present study on early CACD, the K1 from the central and paracentral atrophic area were reduced significantly (Fig. 3) . However, K1 was also depressed in the area outside the lesion where no fundus abnormality was visible ophthalmoscopically or angiographically (Fig. 4) . The marked reduction in the amplitude of K1 in the atrophic area can be explained by the loss of cone cells, RPE, and/or choriocapillaris as has been reported in previous publications. ^{23 24 25} However, the reduced K1 in areas without visible ophthalmoscopic atrophy cannot be easily explained, although a similar phenomenon was reported in the early stages of CSR ¹⁵ and OMD. ¹⁶ Both of these diseases also have distinctly localized central retinal lesions, and the K1s outside the central lesion are also abnormal. In CSR, K1 is reduced in amplitude, with a delayed peak time outside the central detachment, whereas the psychophysical sensitivity is normal outside the lesion. In eyes with OMD, there is a delay of the peak time outside the lesion that is not detectable by psychophysical testing, and the implicit times and amplitudes of K1 outside the atrophy in these eyes are also delayed and reduced. 

These two reports have another finding in common with the present study. The conventional full-field ERGs are normal in these three macular diseases. These findings suggest that the reduction of K1 outside the lesion does not extend over the entire retina. Because the multifocal ERGs were recorded from the central 50° of the posterior pole, only 23% of all the cone cells were stimulated. ²⁶ Therefore, normal full-field ERGs can be obtained when the retina peripheral to the central 50° is normal. 

The question then arises as to how the area outside the lesion is affected. In CSR, it has been suggested that the primary lesion is not the central detachment, but that diffuse pathologic changes in the choroids and/or retinal pigment epithelium lead to ERG abnormalities outside the detachment. ¹⁵ In OMD, the mechanism for the delayed ERG outside the lesion has not been determined. ¹⁶ CACD is a progressive disease, and it has been reported that the primary lesion is in the RPE in the atrophic area. ^{23 24 25} Therefore, it is likely that abnormal K1 outside the lesion indicates an early involvement of the retina and/or the RPE outside the visible atrophic region and may indicate a future progression of the disease. As Hood et al. ¹³ suggested, abnormal local ERG timing in the presence of normal visual field sensitivity may be an early sign of local retinal damage in retinal diseases. 

The peak time of K1 was slightly delayed in most rings (Fig. 4) , but not so much as in congenital stationary night blindness or branch retinal artery occlusion (BRAO; mean delay, 1.2 ms in CACD; 3.2 ms in CSNB; ¹⁷ 4.0 msec in BRAO ²⁷ ). 

K2 in CACD

Our results indicate that K2 was also severely affected in the central and paracentral areas (Fig. 5) . As with K1, the amplitude of K2 was depressed, and the peak time was delayed in the area outside the lesion (Fig. 6) . However, the ratio of K2 to K1 was not significant reduced in both the atrophic area and the atrophy-free area. The reduced ratio of K2 to K1 associated with delayed K1 has been reported in CSNB, ¹⁶ diabetic retinopathy, ⁹ and BRAO, ²⁷ suggesting some defect in the adaptation mechanism of the retinal outer plexiform layer. ²⁶ However, we found a normal K2-to-K1 ratio associated with a slight delay in CACD, which was probably caused not by abnormal adaptation mechanism but by dysfunction of the cone cells. This is in accord with the fact that the primary lesion in this disease is in the RPE. ^{23 24 25} Therefore, our results are consistent with early changes of retinal function in this disease that are presynaptic to the bipolar cells. Local lesions of all retinal area could give similar findings, but it is unlikely in this disease, according to previous reports. ^{23 24 25}  

Possible Problems in mfERG Recording

One of the important problems in recording and evaluating K1 in patients with macular diseases is the difficulty with fixation. In cases with dense central scotoma, poor fixation may affect the amplitude and/or peak time of K1. In our patients, however, the visual acuity was better than 0.7 in all eyes; none had a dense scotoma, and foveal fixation was identified before the recording. Further, a camera/refractor used for monitoring fixation at least partially resolved this problem. 

Another factor that must be considered is the effect of stray light. The atrophic area with a change of color reflects the stimulus light falling on the region to peripheral retina, and the responses from the peripheral retina may contaminate the response from within the lesion. Although our study did not include patients with reflective choroidal sclerosis, K1 from the atrophic area may have had a small contamination from the periphery. However, this would mean that the real response from the atrophic area may be more reduced, which does not influence our conclusions. The stray light could also influence retinal adaptation in the more peripheral areas and thus cause the differences observed in the peripheral mfERG of CACD and the control group. However, our recent report on stray light’s effect on the mfERG indicates that stray light from a large disc coloboma does not affect mfERGs from outside the coloboma. ²⁸  

In conclusion, K1 was extremely abnormal in the areas with visible atrophy, but it was also abnormal in areas without visible atrophy. This may indicate a centrifugal extension of retinal dysfunction. K2 was very abnormal in the atrophic area, but was also abnormal in the atrophy-free area. However, the ratio of K2 to K1 in the atrophy-free area was normal, suggesting that early retinal dysfunction is presynaptic. 

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

 Submitted for publication August 30, 2002; revised October 29, 2002; accepted November 6, 2002.

 Disclosure: K. Nagasaka, None; M. Horiguchi, None; Y. Shimada, None; M. Yuzawa, None

 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.

 

Table 1.

 

Characteristics of Patients with CACD

Patient	Age	Gender	Eye	Visual Acuity	Foveolar Involvement
1	56	M	R	1.2	Slight
2	71	F	R	0.7	—
3	63	M	L	1.2	—
4	46	M	L	1.5	—
5	55	F	R	1.2	—
6	41	F	R	1.5	—
7	51	F	R	1.0	—
8	67	M	L	0.7	Slight

Figure 1.

 

Fundus photography in eight eyes in the patients with CACD.

Figure 2.

 

(A) Method of measuring the amplitude and peak time of the positive component of K1. (B) Five ring groups in 61 hexagons.

Figure 3.

 

Fundus photograph of one patient with the 61 standard hexagon stimulus superimposed. The central element (R1) of the stimulus fell within the atrophy-free area and the paracentral ring (R2) fell in the atrophic areas. R3 involved both atrophic areas and atrophy-free area, and R4 and R5 fell on atrophy-free areas (see Fig. 2B for ring numbers).

Figure 5.

 

(A) The amplitude of the K1 responses from the five eccentric rings. The amplitude was reduced in all rings. Responses from the rings falling in the visibly atrophic area (•), in both atrophic and atrophy-free areas ( ), and in the atrophy-free areas (○). The vertical bars indicate mean ± SD of normal subjects. (B) The peak time of the K1 responses at the five eccentric rings. The peak time was delayed in all rings.

Figure 4.

 

Sixy-one K1 responses from the local responses of a normal subject and eight patients with CACD.

Figure 6.

 

Sixy-one K2 responses from the local responses of a normal subject and eight patients with CACD. The responses in the central areas (R1–R4) appear to have deteriorated.

Figure 7.

 

(A) The amplitudes of the K2 responses from the central rings (R1–R4) and peripheral ring (R5). The amplitudes were significantly reduced in both areas. (B) The peak time of the K2 responses from R1 to R4 and R5. The peak time was delayed in R5. (C) The K2-to-K1 ratio of the amplitudes in R1 to R4 and R5. The ratio was not reduced in both areas. The vertical bar indicates the mean ± SD of normal subjects.

The authors thank Donald C. Hood for his comments on this study and Mr. Nagasaka for technical support.