Retinal Adaptability Loss in Serous Retinal Detachment with Central Serous Chorioretinopathy

Abstract

Purpose.: To investigate the functional characteristics of the detached retina on a serous retinal detachment (SRD) in eyes with central serous chorioretinopathy (CSC) with spared visual acuity.

Methods.: Multifocal electroretinograms (mfERGs) were recorded with a long recording time of 14 minutes, 34 seconds, to obtain accurate measurement of the second-order kernel (K2.1), an index of functional adaptability of the retina, from seven eyes with CSC (visual acuity, ≥1.0). The first-order kernel (K1) and the K2.1, elicited by stimulating the area of the SRD, were compared with those from the corresponding areas in eyes of 15 age-matched volunteers (controls) and in 6 eyes of patients with diabetic retinopathy (DR) that have been reported to have a K2.1 attenuation.

Results.: K2.1 was essentially flat in the SRD eye. The K2.1 amplitude and log-scaled amplitude ratio of K2.1 to K1 (K2.1/K1) were severely reduced (to <95% confidence interval [CI] of control levels) in all eyes. The value of K2.1/K1 of the SRD was less than that in any of the control and DR eyes. K1 was moderately reduced but was not smaller than the 95% CI of control eyes. The mfERGs from the area without the SRD and those from the fellow eyes did not differ significantly from those in control eyes.

Conclusions.: A possible cause of the flat K2.1 observed on the SRD is the separation of the sensory retina. A substantial disparity between the recovery of cones and rods could contribute to the loss of retinal adaptability, resulting in the flat K2.1 as well as the unique visual impairments in CSC eyes.

Central serous chorioretinopathy (CSC) is characterized by a serous retinal detachment (SRD) caused by leakage of fluid from the retinal pigment epithelium (RPE). ^1–3 CSC can cause a mild to moderate visual reduction; however, in recent-onset cases, visual acuity is often fully spared. ¹ In these cases, the primary symptom of the patient is a transient entopic image of the SRD, ² which is reported as a dark spot with prolonged after-images in the center of the visual field. ^1–3 The entopic image is seen soon after awakening in the morning and reappears after blinking. ² The dark spot becomes brighter for only a few seconds when the visual field is partially covered. ⁴ Although the area of the SRD can be detected by perimetry with a flickering stimulus ⁵ or by photostress recovery tests, ⁶ it is difficult to detect by conventional static perimetry. ^5,7 The exact mechanism for the entopic image of the SRD in eyes with a CSC with spared visual acuity has not been determined. ^1,2  

The purpose of this study was to assess the visual function of the detached retina in the area of the SRD in eyes with CSC. In our investigation, we examined the higher order kernels of the multifocal electroretinograms (mfERGs). Although mfERGs have been recorded from five or more eyes with CSC in earlier studies, ^8–15 the higher order kernels of the mfERG have not been extracted from the area of the SRD. In most studies, the first-order kernel (K1) of the mfERG, which represents mainly the local cone ERG, was analyzed. The higher order kernels, however, are indices of the electroretinographic nonlinearities (i.e., how the mfERG responses are influenced by their adaptation to previous flashes). ^16,17 The first slice of the second-order kernel (K2.1) is the largest component of the higher order kernels and reflects the effect of an immediately preceding flash on the response to the after-flash. An attenuation of K2.1 relative to K1 (i.e., a reduction of the K2.1/K1 amplitude ratio) has been reported for specific retinal diseases, ^18–25 (e.g., diabetic retinopathy [DR]). ^19,23–25  

We studied patients with CSC, whose visual acuity was ≥1.0, and recorded mfERGs by using a long recording time to obtain an accurate analysis of K2.1. K2.1 and K1 of an mfERG elicited by stimulating the area of the SRD were compared with those recorded from the corresponding retinal areas of healthy controls and eyes with DR. 

Methods

The demographics of our subjects are summarized in Table 1. There were seven eyes of seven patients with unilateral CSC, and the visual acuity (VA) was ≥1.0 in all. All the patients were men whose ages ranged from 30 to 50 years. The interval between the first symptoms and the mfERG recordings ranged from 1 to 4 weeks. The sensitivity of the visual field tested in five eyes (patients CSC 2, 3, 5, 6, and 7) with a visual field perimeter (30-2 SITA program, Humphrey Field Analyzer; Carl Zeiss Meditec, Inc., Dublin, CA) was not reduced (shown in Table 1, Figs. 2, 3). 

Table 1.

 

Summary of Subjects

	Age/Sex/Eye	VA		OCT	FA	HFA*
CSC
1	50/M/R	1.0	5/5	Applied	Applied	—
2	47/M/L	1.2	5/4.2	Applied	Applied	Shown below
3	47/M/R	1.0	5/5	Applied	Applied	Shown in Figure 2
4	46/M/L	1.2	5/4.2	Applied	Applied	—
5	34/M/R	1.0	5/5	Applied	Applied	Shown in Figure 3
6	31/M/L	1.0	5/5	Applied	Applied	Shown below
7	30/M/R	1.2	5/4.2	Applied	Applied	Shown below
Age, mean ± SD	40.7 ± 8.6
Control healthy subjects
1	57/M/L	1.2	5/4.2	Applied
2	55/M/L	1.2	5/4.2	Applied
3	50/M/R	1.2	5/4.2	Applied
4	48/M/R	1.2	5/4.2
5	42/M/R	1.0	5/5	Applied
6	41/M/R	1.0	5/5	Applied
7	38/M/R	1.2	5/4.2
8	35/M/R	1.2	5/4.2
9	34/F/R	1.0	5/5
10	34/M/R	1.2	5/4.2
11	31/F/L	1.2	5/4.2
12	30/F/R	1.0	5/5	Applied
13	28/M/L	1.2	5/4.2	Applied
14	27/F/R	1.2	5/4.2
15	26/F/L	1.0	5/5
Age, mean ± SD	38.4 ± 10.1
DR
1	68/M/L	0.9	5/5.6	Applied	Applied
2	62/M/R	0.9	5/5.6	Applied	Applied
3	60/M/L	0.7	5/7.1	Applied	Applied
4	60/F/R	0.7	5/7.1	Applied	Applied
5	53/F/R	0.7	5/7.1	Applied	Applied
6	49/M/L	0.7	5/7.1	Applied	Applied
Age, mean ± SD	58.7 ± 8.7

 FA, fluorescein angiography.

*  Pattern deviation of visual field perimeter (Humphrey Field Analyzer, 30-2 program; Carl Zeiss Meditec, Inc., Dublin, CA).

Fifteen age-matched healthy volunteers (controls, CON) and six eyes of six patients with DR with slight macular edema causing minimal VA reduction (VA, 0.7–0.9) were also studied. The diagnoses of CSC and the macular edema in the DR eyes were confirmed by fluorescein angiography and cross-sectional retinal OCT images (OCT 3000; Carl Zeiss Meditec, Inc.). In control eyes, OCT was performed on seven eyes (subjects CON 1–3, 5, 6, 13, and 14), and fluorescein angiography was not performed. Fundus photographs and horizontal OCT images of one eye from each group are shown in Figure 1. 

Figure 1.

 

Fundus photographs and horizontal OCT images. Top: right eye of a 47-year-old man with CSC. VA, 1.0. Middle: right eye of a 50-year-old control subject. VA, 1.2. Bottom: right eye of a 53-year-old woman with DR. VA, 0.7.

None of the subjects had myopia ≥ −4.0 D, cataract, other media opacities, or ocular surgery. The procedures conformed to the tenets of the Declaration of Helsinki, and informed consent was obtained from all subjects before testing. 

mfERGs were recorded (VERIS, Science 5.1.12; Electro-Diagnostic Imaging [EDI], San Mateo, CA) with a stimulus displayed on a 33 × 24-cm monochrome cathode ray tube (CRT) with a P4 (white) phosphor. An array of 37 densely packed hexagons (stretch factor, 13.18) stimulated the central 40° of the visual field. ²⁶ A camera/refractor (EDI) was used for refraction and to monitor eye position and fixation during the recordings. ²⁶ An m-sequence rate of 75/second and repeated consecutive recordings of four cycles of 2¹⁴ − 1 steps resulted in a long recording time of 14 minutes, 34 seconds (net) to obtain a more accurate measurement of the higher order kernels. The stimulus intensity was 2.67 cd/m² corresponding to the static measurement on the CRT monitor screen of 200 cd/m². The black level was below 1 cd/m². The recordings were performed under regular room lights with the pupil maximally dilated. mfERGs were also recorded from the fellow eye with a conventional recording time of 3 minutes, 38 seconds (net). 

Signals were picked-up by a bipolar contact lens electrode (GoldLens; Diagnosys LLC, Littleton, MA), amplified by 10⁵×, band-pass filtered between 10 and 300 Hz (at half-amplitude), and digitized at a 1200-Hz sampling frequency. A single cycle of artifact removal (K1, 0–80 ms) without spatial averaging was performed. 

Results

The mfERGs recorded from a 47-year-old man (CSC 3) are shown in Figure 2. 

Figure 2.

 

mfERGs recorded from the right eye of a 47-year-old male patient with CSC. (A) Left: the stimulus pattern is superimposed on a fundus photograph that was flipped vertically; right: pattern deviation of the visual field perimeter. (B) Trace arrays of K1 (left) and K2.1 (right). mfERGs recorded from the area of the SRD (mfERG from SRD) and away from the SRD (mfERG from non-SRD) are based on a comparison between the stimulus pattern and the fundus photograph shown in (A, left). mfERGs from SRD and mfERGs from non-SRD are grouped and averaged independently as group averages shown in (C). (C) Bold traces are group averages of K1 (left) and K2.1 (right) of mfERGs from SRD (top) and non-SRD (bottom) defined in (B). The group averages of the 15 control subjects (CON, thin traces) and 6 DR eyes (dashed traces) are also aligned for comparison.

The stimulus pattern is superimposed on a fundus photograph to demonstrate the location of the stimuli on the retina (Fig. 2A, left). The fundus image was flipped vertically so that the mfERG-trace array corresponded to the stimulus geometry. The mfERGs elicited from the stimulus elements that fell on the area of the SRD (surrounded by dashed line) were summed and designated as mfERGs from the SRD. The mfERGs elicited by the stimulus elements that were completely separated from the SRD were designated as mfERGs from the non-SRD. The responses from the boundary regions were excluded from the analyses. 

The trace arrays of K1 (left column) and K2.1 (right column) are shown in Figure 2B. The local mfERGs from each group are summed and shown in Figure 2C on a response-density scale. Bold traces are group averages of K1 (left) and K2.1 (right) of the mfERGs from the SRD (top row) and of the mfERGs from the non-SRD area (bottom row). The group averages in the 15 control (CON, thin traces) and 6 DR (dashed traces) eyes are aligned for comparison. These mfERGs were recorded in areas corresponding to the mfERGs from the SRD and mfERG from the non-SRD areas of the CSC case, although none of them actually had an SRD. 

K2.1 on the SRD was essentially flat and thus significantly smaller than that of the control and DR eyes. The other higher order kernels (e.g., the second slice of the second-order kernel, K2.2, and the first slice of the third-order kernel, K3.1) were also flat (data not shown). On the other hand, K1 and K2.1 of the mfERGs recorded away from the SRD did not differ significantly from that of the control eyes, whereas those from the eyes with DR were also significantly smaller. 

The flat or essentially flat K2.1 and the reduced and delayed K1 from the area of the SRD were observed in all the CSC eyes. The mfERGs from CSC 5 are shown in Figure 3. 

Figure 3.

 

mfERGs recorded from the right eye of a 34-year-old patient with CSC. The layout is the same as in Figure 2.

K1 was reduced and delayed, and K2.1 from the area of the SRD was flat. K2.2 and K3.1 from the SRD were also flat (not shown). K1 and K2.1 recorded from the non-SRD area were not significantly different from that in the control eyes. 

The areas of the SRD and non-SRD differed among the eyes with a CSC, and so the comparisons to control and DR eyes had to be performed on a case-by-case basis. The density-scaled amplitudes of K1 and K2.1, and the log-scaled amplitude ratio of K2.1 to K1 (log K2.1/K1) are plotted in Figure 4. The measurement of the amplitudes is shown in the bottom panel. The thick horizontal line for each set of values in each column represents the value of each CSC eye. The open and closed circles represent the responses recorded from the corresponding areas of the control and DR eyes, respectively. The thin horizontal lines among the open and closed circles are the mean values of the control and DR eyes, respectively. An asterisk indicates that the value of the CSC case is less than the 95% confidence interval [CI] (<5%) of the control or DR eyes. 

Figure 4.

 

The density-scaled amplitudes of K1 and K2.1 (measurement of amplitudes is shown in the bottom panel), and the log-scaled amplitude ratio of K2.1 to K1 (log K2.1/K1). Thick horizontal lines: the value of each CSC eye; open and closed circles: the corresponding responses from control and DR eyes, respectively. Thin horizontal lines among the open and closed circles are the mean values of control and DR eyes, respectively. *The value in the CSC eye is less than the 95% CI (<5%) of the control or DR eye.

The K1 amplitude from the area of the SRD (top row) was not smaller than the 95% CI of control eyes in all cases. However, the K2.1 amplitude recorded from the area of the SRD (second row) was severely reduced to less than the 95% CI (<5%) of that in control eyes in all cases. In three cases (CSC 1, 3, and 5), the K2.1 amplitude was less than the 95% CI of the DR eyes. The reduction of log K2.1/K1 (third row) emerged even more clearly; K2.1/K1 in the CSC eye was less than in any control or DR eyes, without exception. In contrast, all the K1, K2.1, and log K2.1/K1 from the non-SRD area fell within the 95% CI of controls (bottom). The mfERGs obtained from the fellow eyes without an SRD of CSC were not altered (not shown). 

Discussion

Although the mfERGs in eyes with CSC have been investigated, ^8–15 most of these studies focused only on K1. ^8–14 A more recent study ¹⁵ examined K2.1, but the investigators did not discriminate between the mfERGs recorded from the area of the SRD and from the areas without the SRD. Our results showed that the higher order kernels including K2.1 were flat, or essentially flat, at the site of the SRD, whereas K1 was moderately reduced and prolonged at the site of the SRD. In contrast, the mfERGs recorded from the retina away from the SRD were not significantly different from that in control eyes. 

It has been reported that a CSC can affect the retina away from the SRD and even the fellow eyes without an obvious retinal detachment. ⁸ We could not confirm these reports in our cases, but it may be due to the different areas analyzed, as we focused on detached retinas on a case-by-case basis. 

We selected subjects with visual acuities ≥1.0, whereas earlier studies included patients with reduced vision. This reduced vision can be caused by retinal atrophy or other pathologic changes secondary to prolonged SRD, ^1–3 which in turn could affect different components of the mfERGs. 

The higher order kernels are indices of ERG adaptability, and K2.1 is the largest component. A flat K2.1 indicates that the retinal response to a flash does not recover rapidly (i.e., the response to the flash is not modified by the immediately preceding flash). ^16,17 K1 was moderately reduced and delayed at the site of the SRD, as reported earlier. ^8–15 This reduction and delay could be due to the flat K2.1 because the late period of the K1 response reflects the overlapping higher order kernels. ^16,17,26  

A reduction of K2.1 is consistent with the findings reported by Miyake et al., ²⁷ who found that the oscillatory potentials (OPs) were more reduced than the a- and b-waves of the focal macular ERGs elicited by a 10° stimulus. This size was approximately the size of an SRD in CSC eyes with good VA. The OPs of the photopic ERGs are the components most sensitive to the state of adaptation, ²⁸ and they are thought to arise from the neural elements as K2.1. ²⁹  

The pure stray-light effect also causes delayed K1 associated with flat K2.1. ²⁶ When the stimulus light is imaged on a bright and enlarged optic disc, the reflected light can be equivalent to a weak full-field stimulus. ²⁶ Because the stray light illuminates the entire retina with a low intensity, a delayed K1 is elicited. K2.1 and higher order kernels have different intensity-response characteristics than K1 at low intensities, so that they cannot be evoked by the weak stimuli. Consequently, a delayed K1 associated with a flat K2.1 is observed at the site. The stray-light–induced mfERGs can be magnified in the dark-adapted retina or by strong stimulus lights. However, at least under the stimulus condition used in this study, the stray-light–induced K1 was estimated to be as large as one third the amplitude of a local response, even on the optic disc with the highest reflection. ²⁶ On the retina without high reflection, the stray-light effect became much smaller. ²⁶ Because the reflection of the detached retina in CSC differs little from that of the retina without an SRD, the stray-light effect hardly contributes to the moderately reduced K1 observed on SRD in CSC eyes. 

The reduction of the K2.1/K1 ratio or a delayed K1 with reduced K2.1 has also been reported in eyes with glaucoma, ^18,22 branch retinal artery occlusion, ²⁰ and congenital stationary night blindness. ²¹ All these diseases have been attributed to inner retinal layer dysfunction. ^{18–22,24,25} In agreement with this, pharmacologic agents that block inner retinal activity reduced K2.1 in animals. ^29–31 The inner retina is the most affected layer in early diabetes, ²⁵ and an abnormality in the adaptational process in the inner retina is assumed to be the cause of the K2.1 reduction in diabetic eyes. ¹⁹ The reduction of OPs in CSC eyes was also thought to arise from inner retinal dysfunction. ²⁷ However, it is important to note that K2.1 on the SRD in CSC eyes was essentially flat and much smaller than that in eyes in diabetes, which is one of the most common pathologic states that cause K2.1 attenuation. ^19,23–25 In most of the earlier studies, ^{18–22,24,25} different inner retinal diseases reduced K2.1 relatively, but did not flatten it. Greenstein et al. ²³ noted that a pharmacologic blockage of the inner retinal activity ^29–31 also reduced K2.1 but did not eliminate it. 

They reported that a flat K2.1 was associated with a broad-shaped and markedly delayed K1 from regions with poor visual field sensitivity. They reported a flat K2.1 in two of seven cases of retinitis pigmentosa, three of four cases of cone dystrophy, and two of five cases of DR. DR affects the inner retina as well as the middle and outer retina, and the functional alterations in retinitis pigmentosa and cone dystrophy have been attributed to damage and loss of the rod and cone photoreceptors. ²³ They proposed that a flat K2.1 can be caused by abnormal adaptation mechanisms due to damage to the outer plexiform layer. ^17,23 Their proposal is suggestive, but the characteristics of their cases ²³ with a flat K2.1 do not fully agree with those in our cases of CSC. Their K1s were more severely attenuated than ours. Their atypical mfERGs were obtained in part in retinal regions with poor visual field sensitivity. All our CSC eyes had good VA and normal visual fields. 

Finally, we suggest that the separation of the sensory retina from the RPE causes the flat K2.1 in eyes with CSC. The separation of the outer segments of photoreceptors from RPE by the subretinal fluid should slow down the exchange of all-trans and 11-cis retinal. The RPE–photoreceptor visual cycle serves mainly the rods, and cone function is supported by a separate visual cycle within the sensory retina ³² and is thus less affected by the separation from the RPE. Cone function partially compromised by the separation from RPE or the loss of the contribution of rods to photopic vision may alter the adaptational mechanisms causing the entopic image of the SRD. ² Delayed adaptation of the detached retina is also supported by the demonstration of elevated fatigability of the double-flash ERG in CSC eyes, ² whereas conventional single-flash ERGs are normal. In agreement with this, deeper scotomata are found with a flickering stimulus than with a steady state stimulus. ⁵ A substantial disparity between the recovery of cones and rods could contribute to the loss of retinal adaptability resulting in the flat higher order kernels as well as the unique visual impairments in CSC eyes. ^1–7  

Footnotes

 Disclosure: Y. Shimada, None; D. Imai, None; Y. Ota, None; K. Kanai, None; K. Mori, None; K. Murayama, None; S. Yoneya, None