THE EFFECT OF PHOTOPIGMENT

BLEACHING ON FUNDUS
AUTOFLUORESCENCE IN ACUTE CENTRAL
SEROUS CHORIORETINOPATHY
KWANG-EON CHOI, MD, CHEOLMIN YUN, MD, PHD, YOUNG-HO KIM, MD,
SEONG-WOO KIM, MD, PHD, JAERYUNG OH, MD, PHD, KUHL HUH, MD, PHD

Purpose: To evaluate the effect of photobleaching on fundus autofluorescence (FAF)

images in acute central serous chorioretinopathy.
Methods: We obtained prephotobleaching and postphotobleaching images using an
Optomap 200Tx, and photobleaching was induced with a Heidelberg Retina Angiograph 2.
Degrees of photobleaching were assessed as grayscale values in Optomap images.
Concordances among the three kinds of images were analyzed. Hyper-AF lesions in
prephotobleaching images were classified as Type 1 (changed to normal-AF after
photobleaching) and Type 2 (unchanged after photobleaching). The FAF composite
patterns of central serous chorioretinopathy lesions were classified as diffuse or mottled.
Initial and final best-corrected visual acuity, central retinal thickness, and disease duration
were compared according to fovea FAF type.
Results: Forty-one eyes of 41 patients were analyzed. The lesion brightness of
postphotobleaching Optomap FAF showed greater concordance with Heidelberg Retina
Angiograph 2 FAF (94.74%) than the prephotobleaching Optomap FAF (80.49%). Eyes with
Type 1 fovea had greater initial and final best-corrected visual acuity (20/23 vs. 20/41, 20/
21 vs. 20/32, P , 0.0001, P = 0.001, respectively) and shorter disease duration (19.68 ±
12.98 vs. 51.55 ± 44.98 days, P = 0.043) than those with Type 2 fovea. However, eyes with
diffuse Type 2 fovea had only lower initial and final best-corrected visual acuity (20/23 vs.
20/45, 20/21 vs. 20/36, P , 0.0001, P , 0.0001, respectively) than those with Type 1 fovea.
Conclusion: Understanding the photobleaching effect is necessary for the accurate
interpretation of FAF images. Furthermore, comparing prephotobleaching and postphoto-
bleaching FAF images may be helpful for estimation of lesion status in central serous
chorioretinopathy.
RETINA 37:568–577, 2017

C entral serous chorioretinopathy (CSC) is a chorior-

etinal disorder accompanying serous detachment
of the neurosensory retina or pigment epithelial detach-
ment on the posterior pole, and typically affects young
to middle-aged men.1 Fluorescein angiography, indoc-
From the Department of Ophthalmology, Korea University
College of Medicine, Seoul, Korea.
J. Oh received a Grant from the Korean Ministry of Envi-
ronment through “The Environmental Health Action Program
(2012001350010).” The other authors have no financial/conflicting
interests to disclose.
Reprint requests: Seong-Woo Kim, MD, PhD, Department of
Ophthalmology, Korea University Ansan Hospital, 516 Gojan-
dong, Danwon-gu, Ansan-si, Kyung gi-do 425-707, Korea; e-mail:
ksw64723@korea.ac.kr
yanine green angiography, and spectral domain optical
coherence tomography (SD-OCT) are widely used for
evaluating eyes with CSC.2–4 Recently, fundus auto-
fluorescence (FAF) has also become a valuable modal-
ity for understanding disease status and prognosis in
CSC.5,6 Many reports have addressed FAF image char-
acteristics in CSC,1,5–9 and one study suggested that
FAF can predict best-corrected visual acuity (BCVA)
in eyes with CSC.10 Decreased photopigment density
in photoreceptors in CSC was previously demonstrated
by Burns et al11 and Ito et al.12 In addition, the change in
FAF during bleaching in CSC was first demonstrated by
Staurenghi et al.13 In CSC, because photoreceptors of the
acutely detached retina have less effective photopigment

568
 PHOTOBLEACHING AND FAF IN CSC  CHOI ET AL
for light absorption, the FAF inside the CSC lesion be-
comes brighter than the FAF outside the lesion (normal
retina) after dark adaptation.11–15
Recently, FAF images taken with two different
confocal scanning laser ophthalmoscopes (cSLO) in eyes
affected by CSC (Heidelberg Retina Angiograph 2
[HRA2]; Heidelberg Engineering, Heidelberg, Germany
vs. Optomap 200Tx [Optomap]; Optos, Dunfermline,
Scotland, United Kingdom) were compared.16 The two
types of FAF images of CSC showed similar lesion
composite patterns but different lesion brightness values.
The number of cases showing hyper FAF intensity inside
the lesion is significantly higher in Optomap images.16
Two possible explanations account for this difference.
One is that, there is a difference in the amounts of light
absorption and excitation caused by differences in wave-
length (488 vs. 532 nm).8,17,18 An alternative explanation
is photopigment bleaching. Differences in image acqui-
sition and wavelengths between the two cSLO machines
may result in different degrees of photopigment bleach-
ing. In a healthy retina, photopigment bleaching increases
AF intensity to a higher level.2
In this study, we primarily aimed to characterize how
photopigment bleaching affects differences in intensity
around CSC lesions in FAF images by comparing
images taken before and after photopigment bleaching.
We also compared basic characteristics of eyes accord-
ing to the hyper-AF change in the fovea before and after
photopigment bleaching.
Methods
This study was approved by the Korea University
Ansan Hospital Institutional Review Board, Seoul,
Korea (IRB number: AS13122). All research protocols
and methods of data collection adhered to the tenets of
the Declaration of Helsinki.
This retrospective study included patients with acute
CSC treated at Korea University Ansan Hospital
between January 15, 2014 and November 30, 2014.
Acute CSC was diagnosed by fluorescein angiogra-
phy, indocyanine green angiography, and SD-OCT
when serous neurosensory retinal detachment was
present at the macula on SD-OCT, and fluorescein
leakage at the level of retinal pigment epithelium
(RPE) was present in fluorescein angiography. In
addition, the onset of symptoms was less than 6
months before presentation. Fundus autofluorescence
images were collected with both HRA2 and Optomap
before fluorescein angiography and indocyanine green
angiography images were obtained. If the 30° rectan-
gular photobleached area in the postphotobleaching
Optomap FAF image was clearly brighter than the area
569
outside the 30° area and the margin of the 30° square
was clearly identifiable, the image was considered
properly bleached (Figure 1). Eyes with other intraoc-
ular diseases such as age-related macular degeneration,
polypoidal choroidal vasculopathy, epiretinal mem-
brane, diabetic retinopathy, media opacity (cataract,
corneal, or vitreous opacity), or histories of any ocular
surgery except cataract surgery were excluded. In
addition, cases with long disease duration ($6
months) or any hypo-AF lesion in Optomap FAF im-
ages were excluded. Disease duration was defined as
duration from the onset of symptoms, such as visual
disturbance and metamorphopsia. Follow-up duration
was defined as duration from first visit to final visit in
persistent cases or to the visit that CSC resolution was
confirmed with SD-OCT.
Fundus Autofluorescence
Based on our previous study,16 the FAF examina-
tion protocol for patients with CSC used in this study
was developed in our clinic. Patients routinely under-
went both HRA2 and Optomap FAF imaging accord-
ing to the following three steps. Three kinds of FAF
images were serially acquired using Optomap and
HRA2. First, after waiting 30 minutes in a dark room
for full pupil dilation, a prephotobleaching Optomap
FAF image was acquired in 100°ResMax mode, which
acquired a single AF image with green light excitation
at 532 nm and emission at 570 nm to 780 nm. The
exposure time was 250 milliseconds to 300 millisec-
onds per scan. The 100°ResMax images cover an
external angle of 61° · 61° with a resolution of 11
mm/pixel. The intensity of the laser beam was 1.7 mW
to 1.8 mW. Second, after the Optomap FAF image was
acquired, HRA2 FAF images were taken (0.28 mW
beam intensity, high-speed mode; 30 · 30° field;
768 · 768 pixels; 8.9 frames/second). The image res-
olution was 10 mm/pixel, and the exposure time was
about 110 milliseconds. A wavelength of 488 nm (blue
light) was used for excitation, and emitted light was
detected above 500 nm with a barrier filter. To obtain
a high-quality mean FAF image, up to 100 images
were averaged using an automatic real-time algorithm.
Third, immediately after the HRA2 FAF images were
collected, a postphotobleaching Optomap FAF image
was acquired, again in 100°ResMax mode.
Optical Coherence Tomography
Spectral domain optical coherence tomography was
performed using the Spectral OCT/SLO system (Optos
OCT SLO; Optos). Line scans, raster scans, and
topographic mapping were performed on each eye.
The line scan was performed horizontally and vertically,
570 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2017  VOLUME 37  NUMBER 3

Fig. 1. Spectral domain optical

coherence tomography and
prephotobleaching and post-
photobleaching Optomap FAF
images of CSC. Spectral domain
optical coherence tomography
showing the detached neurosen-
sory retina (top). Prephotobleach-
ing Optomap FAF image showing
hyper-AF within the lesion (mid-
dle left). In the postphotobleaching
Optomap FAF image (middle
right), the hyper-AF within the
lesion was masked by an increased
30° square of peripheral AF,
which corresponded to the HRA2
image size. Cropped prephoto-
bleaching (bottom left) and post-
photobleaching (bottom middle)
Optomap FAF images, and 30°
field of an HRA2 FAF image
(bottom right). The arrow indicates
a Type 1 area, which was hyper-
AF in the prephotobleaching
Optomap FAF but normal-AF in
the postphotobleaching Optomap
FAF. The asterisk indicates a Type
2 area, which was hyper-AF in
both prephotobleaching and post-
photobleaching Optomap FAF.

and the image of the scan was created by averaging up
to twenty 9-mm long B-scans (1,024 A-scans per
B-scan). The raster scan was performed with 32 B-scans
(1,024 A-scans per B-scan) in a 9-mm · 9-mm area. The
topographic mapping consisted of 512 horizontal
A-scans · 128 vertical lines across a 5-mm · 5-mm area.
Central retinal thickness (CRT) was defined as the mean
distance between the internal limiting membrane and the
upper border of the retinal pigment epithelium in
a 1-mm-diameter foveal zone. Central retinal thickness
was generated automatically during OCT analysis and
then manually adjusted in cases of segmentation error.
Image Analysis
We compared the prephotopigment and postphoto-
pigment bleaching Optomap and HRA2 FAF images.
First, to determine the temporal change in AF intensity
by photopigment bleaching (photobleaching) in Opto-
map images, we calculated the grayscale values in the
prephotopigment and postphotopigment bleaching
Optomap FAF images using Image J software (1.48
version; National Institutes of Health, Bethesda, MD).
Two sample areas inside and outside the superonasal
corner of the 30° rectangular photobleached area were
selected for evaluation of gray values. These two areas
corresponded to prephotopigment and postphotopig-
ment bleaching Optomap FAF images. The sample
area was 1° square in size (256 · 256 dot pixels in
Image J software) in the superonasal quadrant from the
optic disc to avoid CSC lesions and neighborhood
vascular structures (Figure 2). The ratio (grayscale
value inside/grayscale value outside the 30° rectangu-
lar area) of postphotobleaching Optomap FAF image
was compared with that of the prephotobleaching Op-
tomap FAF image.
Second, to evaluate the concordances of lesion
characteristics among the three type of image, FAF
 PHOTOBLEACHING AND FAF IN CSC  CHOI ET AL
brightness and the lesion composite pattern were
analyzed. Fundus autofluorescence brightness was
classified as normal-AF or hyper-AF. Normal-AF was
AF of similar intensity to that of the normal attached
retina, for which lesions could not be delineated from
the normal attached retina in the FAF image. Physio-
logic hypo-AF due to foveal macular pigment was
considered normal-AF, especially in HRA2 images.
Hyper-AF was defined as AF brighter than that of the
attached retinal area. Hyper-AF was further divided into
Types 1 and 2. Type 1 was defined as an area showing
a hyper-AF lesion in prephotobleaching images that was
no longer hyper-AF after photobleaching. Type 2 was
defined as an area showing hyper-AF in both prephoto-
bleaching and postphotobleaching images (Figure 3).
The lesion composite pattern was classified as diffuse
or mottled type. Before image analysis, Photoshop 7.0
(Adobe Systems Co, San Jose, CA) was used to crop
the area outside the 30° field of view corresponding to
that of the HRA2 image to minimize bias because of
differences in the field of view between HRA2 and
Optomap FAF images. The fovea was identified on
571
Fig. 2. The grayscale values of
2 sample areas inside and out-
side the superonasal corner of
a 30° rectangular photobleached
area were calculated in pre-
photopigment and postphotopig-
ment bleaching Optomap FAF
images. A. Superonasal 1° area
inside the 30° rectangular area in
prephotobleaching (left) and
postphotobleaching (right) Opto-
map FAF image. B. Superonasal
1° area outside the 30° rectan-
gular area prephotobleaching
(left) and postphotobleaching
(right) Optomap FAF image. The
grayscale value ratio (grayscale
value inside/grayscale value out-
side the 30° rectangular area) of
the postphotobleaching Optomap
FAF image was higher than that
of the prephotobleaching Opto-
map FAF image (83.227/73.824
vs. 76.617/72.754).
the cropped Optomap FAF image by comparing the
image with the corresponding HRA2 FAF image and
matching large retinal vasculature of each FAF image.
In cases of disagreement, both observers made a final
decision through consensus. The interobserver correla-
tion coefficient (kappa value [k]) was calculated for the
two physicians (C.K.E. and K.S.W.) who evaluated
images.
Third, to compare BCVA, CRT, and disease
duration according to FAF type in the fovea, the
FAF type of the fovea was also classified. If the
detached fovea was not directly adjacent to the normal
attached retina, fovea AF type was defined by
comparing its AF brightness with that of the adjacent
general background AF within the CSC lesion. If the
foveal AF was homogenously bright similar to the
background AF within the lesion, it was defined as the
same type as the CSC AF Type. If the foveal AF was
brighter than that of the adjacent general background
AF within the CSC lesion, it was defined as a Type 2
fovea (Figure 4). The foveal composite pattern was
also classified as diffuse or mottled type.
572 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2017  VOLUME 37  NUMBER 3
Fig. 3. Three cases of each hyper-AF type. A. Type 1 image. SD-OCT
showing subretinal fluid and retinal pigment epithelial detachment in
CSC (top). Hyper-AF area in the cropped 30° field of a prephoto-
bleaching Optomap FAF image (bottom left) was not present in the
postphotobleaching Optomap FAF image (bottom middle). The diffuse
hyper-AF area did not appear in the HRA FAF image, except for the
portion corresponding to pigment epithelial detachment (bottom right).
B. Diffuse pattern Type 2 image. Spectral domain optical coherence
tomography showing subretinal fluid in CSC (top). Diffuse pattern
hyper-AF area was manifest both in the cropped 30° field of pre-
photobleaching (bottom left) and postphotobleaching (bottom middle)
Optomap FAF images. The same pattern hyper-AF area was also noted
in the HRA FAF image (bottom right). C. Mottled pattern Type 2
image. Spectral domain optical coherence tomography showing
subretinal fluid and subretinal deposits in CSC (top). Mottled pattern
hyper-AF area was manifest both in the cropped 30° field of pre-
photobleaching (bottom left) and postphotobleaching (bottom middle)
Optomap FAF images. The same pattern hyper-AF area was also noted
in the HRA FAF image (bottom right).
Statistics
Statistical analyses were performed using IBM
SPSS for Windows version 20 (IBM Co, Somers,
NY). Best-corrected visual acuity was converted to
the logarithm of the minimal angle of resolution
(logMAR) for statistical analysis. The Wilcoxon
signed-rank test, Mann–Whitney U test, and Pearson
chi-square test were used as appropriate. P values ,
0.05 were considered statistically significant.
Results
Subject Characteristics
Initially, the study sample included 70 eyes of 70
patients. We excluded 11 cases of extrafoveal CSC and
18 cases of chronic CSC. Finally, 41 fovea-involving
CSC cases were included. The numbers of right and left
eyes were 20 (48.8%) and 21 (51.2%), respectively, and
the male to female ratio was 30 (73.2%) to 11 (26.8%).
The mean age of patients was 42.80 ± 9.28 years. Mean
disease duration was 36.78 ± 37.37 days, and mean
follow-up duration was 5.66 ± 3.28 months. At the
initial visit, BCVA was 20/32, and BCVA at last visit
was 20/23. The mean spherical equivalent was 20.65 ±
1.72, and the mean cylinder was 20.81 ± 0.61. There
were no significant differences in mean spherical equiv-
alent or mean cylinder between eyes with Type 1 fovea
and those with Type 2 fovea (P = 0.313, 0.155, respec-
tively, Mann–Whitney U test). Mean CRT at the initial
visit was 430.95 ± 136.79 mm on SD-OCT.
Assessment of Degree of Photobleaching
The ratio (grayscale value inside/grayscale value
outside the 30° rectangular area) of postphotobleach-
ing Optomap FAF images was significantly higher
than that of prephotobleaching Optomap FAF images
(1.34 ± 0.05 vs. 1.06 ± 0.04, P , 0.0001, Wilcoxon
signed-rank test).
Lesion Brightness and Composite Pattern in
Fundus Autofluorescence Images
There were no normal-AF images and 41 hyper-AF
images among the prephotobleaching Optomap FAF
images. Among the HRA2 FAF images, 8 were hypo-
or normal-AF, and 33 were hyper-AF. Among the
postphotobleaching Optomap FAF images, 11 were
normal-AF and 30 were hyper-AF. Lesion brightness
was more similar between the postphotobleaching
Optomap FAF images and HRA2 FAF images than
between the prephotobleaching Optomap FAF images
and HRA2 FAF images. The concordance of lesion
brightness was 94.74% for the postphotobleaching
Optomap FAF images and HRA2 FAF images but
only 73.17% for the prephotobleaching Optomap FAF
images and HRA2 FAF images. The concordance
between prephotobleaching and postphotobleaching
Optomap FAF images was 80.49% (Table 1). The
 PHOTOBLEACHING AND FAF IN CSC  CHOI ET AL 573

Fig. 4. A case of heterogeneous hyper-AF in lesion brightness (combined Type 1 and Type 2). Spectral domain optical coherence tomography showing
SRF and some irregular retinal pigment epithelial changes in CSC (top). Cropped 30° field of a prephotobleaching Optomap FAF image (bottom left)
showing a hyper-AF area (asterisk) (Type 1 AF). Foveal hyper-AF lesion (arrow head) (Type 2 AF) was still present in the postphotobleaching
Optomap FAF image (bottom middle). A donut-shaped hyper-AF lesion was noted around the macula in the HRA FAF image (bottom right).

interobserver correlation coefficients (k) were 1.000,
0.689, and 0.819 in prephotobleaching and postphoto-
bleaching Optomap FAF images and HRA2 FAF im-
ages, respectively. Of 41 cases, 17 eyes showed
homogenous AF in lesion brightness (8 cases with
hyper-AF Type 1 and 9 cases with Type 2), and 24
cases were heterogeneous in lesion brightness (com-
bined hyper-AF Type 1 and Type 2) (Figure 4). There
was 100% concordance in the analysis of composite
pattern (Table 1). The numbers of diffuse and mottled
patterns were 32 and 9, respectively.
Relationships Between Patient Characteristics and
Optomap FAF Type in the Fovea
In the foveal area, there were 19 eyes classified as
Type 1 and 22 eyes as Type 2. And the numbers of
diffuse and mottled pattern in fovea were 33 and 8,
respectively. All 8 mottled AF eyes were in eyes with
Type 2 fovea. There were significant differences in
BCVA and disease duration between eyes with Type 1
and Type 2 fovea. Eyes with Type 1 fovea had greater
visual acuity (20/23 vs. 20/45, P , 0.0001, Mann–
Whitney U test) and shorter disease duration (19.68 ±
12.98 vs. 51.55 ± 44.98 days, P = 0.043, Mann–Whit-
ney U test) than those with Type 2 fovea. There were
also significant differences in final visual acuity between
eyes with Type 1 and Type 2 fovea (20/21 vs. 20/32,
P = 0.001, Mann–Whitney U test). There were 3 unre-
solved CSC cases in eyes with Type 2 fovea (2 in diffuse
type, 1 in mottled type). Except for these 3 cases, final
BCVA of eyes with Type 2 fovea also showed worse
visual acuity than those with Type 1 fovea after CSC
resolution (20/33 vs. 20/21, P = 0.020, Mann–Whitney

Table 1. Characteristics of FAF Images for Prephotobleaching and Postphotobleaching Optomap and HRA2 in CSC
Prephotobleaching Optomap HRA2 Postphotobleaching Optomap
Lesion brightness* (normal/hyper) 0/41 11/30 8/33
Lesion composite pattern† (diffuse/ 32/9 32/9 32/9
mottled)
*Concordance for lesion brightness: prephotobleaching Optomap versus HRA2 = 73.17%; HRA2 versus postphotobleaching Optomap =
94.74%; and prephotobleaching Optomap versus postphotobleaching Optomap = 80.49%.
†Concordance and Kappa value for lesion composite pattern: 100% and 1.0 for all 3 kinds of FAF images.
574 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2017  VOLUME 37  NUMBER 3
U test). There were no differences in age, CRT, or
follow-up duration between eyes with Type 1 and Type
2 fovea (Table 2). In subgroup analyses, there were no
significant differences in age, CRT, or follow-up duration
between eyes with diffuse Type 2 fovea and those with
mottled Type 2 fovea. Eyes with mottled Type 2 fovea
had lower visual acuity than those with Type 1 fovea (20/
34 vs. 20/23, P = 0.025, Mann–Whitney U test) and
longer disease duration (82.43 ± 42.79 vs. 19.68 ±
12.98 days, P = 0.025, Mann–Whitney U test). Eyes with
Type 1 fovea had better visual acuity at the initial visit
than those with diffuse Type 2 fovea (20/23 vs. 20/45,
P , 0.0001, Mann–Whitney U test), and had also better
visual acuity at the last visit than those with diffuse Type
2 fovea (20/21 vs. 20/36, P , 0.0001, Mann–Whitney U
test). There were 2 unresolved CSC cases in eyes with
diffuse Type 2 fovea. Except for these 2 cases, the final
BCVA of resolved eyes with diffuse Type 2 fovea also
indicated worse visual acuity than those of resolved eyes
with Type 1 fovea (20/37 vs. 20/21, P = 0.020, Mann–
Whitney U test). However, there was no difference in
disease duration between eyes with Type 1 and diffuse
Type 2 fovea (19.68 ± 12.98 vs. 37.13 ± 39.37 days, P =
0.212, Mann–Whitney U test). There were also no differ-
ences in age, CRT, or follow-up duration between eyes
with Type 1 and diffuse Type 2 fovea (Table 3).
Discussion
In this study, postphotobleaching Optomap FAF
images were more similar to HRA2 FAF images than
to prephotobleaching Optomap FAF images. The 488-
nm short wavelength blue light of HRA2 was used to
capture SW-FAF images and to induce photopigment
bleaching, as Theelen et al19 demonstrated that retinal
photopigments can be bleached with HRA2. Photopig-
ments in the rods and cones produce visual system
Table 2. Comparison of Basic Characteristics According to Hyper-AF Type in the Fovea
Type 1 (n = 19)
Age, years 40.58 ± 6.10
Sex, male/female 14/5
Initial BCVA 20/23 (0.06 ± 0.07 logMAR)
Final BCVA§ 20/21 (0.03 ± 0.06 logMAR)
Follow-up duration, months¶ 4.47 ± 1.81
CRT, mm 308.79 ± 78.23
Disease duration, days 19.68 ± 12.98
Results are presented as mean ± SD.
*There were 3 unresolved CSC cases in eyes with Type 2 fovea.
†Mann–Whitney U test.
‡Chi-square test.
§Final logMAR BCVA in resolved eyes (Type 1 vs. Type 2 = 20/21 [0.03 ± 0.06 logMAR] vs. 20/33 [0.22 ± 0.25 logMAR], P = 0.020,
Mann–Whitney U test).
¶Follow-up duration in resolved eyes (Type 1 vs. Type 2 = 4.47 ± 1.81 vs. 6.21 ± 3.66 months, P = 0.212, Mann–Whitney U test).
activity through the absorption of photons. In photo-
pigment bleaching, a molecule of rhodopsin absorbs
one quanta of light, and then is bleached. The bleached
molecule is not capable of capturing another quantum.
Photopigment bleaching within the outer segments
involves the dissociation of rhodopsin into opsin and
11-cis retinal, followed by isomerization to all-trans
retinal.20–22 Through photo-isomerization, retinal pho-
topigments lose their original absorption properties, and
the optical density of the photoreceptors is altered.23
Considering only the wavelength differences
between 2 kinds of cSLOs, a 532-nm green laser beam
produced more intense bleaching than a 488-nm blue
laser beam because of more complete absorption by
retinal photoreceptors. However, red/green-sensitive
cone cells are more abundant than blue-sensitive cone
cells among all retinal photoreceptors.24 Green light is
therefore much more effective for total retinal bleaching
than blue light, although irradiation with a 488-nm blue
laser beam can bleach visual pigments of the outer seg-
ments not only in short-wavelength-sensitive cones and
rods but also in middle- and long-wavelength-sensitive
cones to a certain extent.19,25,26 With regard to scan
time, although the Optomap requires 250 milliseconds
to 300 milliseconds to cover a 61° · 61° field, HRA2
requires 110 milliseconds to cover a 30° · 30° field. In
terms of light exposure time for a single scan, Optomap
takes less time than HRA2 for the same size area. Fur-
thermore, cSLO serially scans the retina point by point
and does not take the whole retina simultaneously, as
does a conventional fundus camera. Therefore, although
cSLO requires hundreds of milliseconds to take a fundus
image, the light exposure time for each point on the
retina is far less than that needed to take an entire
image, unlike a conventional fundus camera. If we pos-
tulate that the whole scanned retina was continuously
exposed to light during the 0.3 seconds of acquisition
time required by Optomap, as with the conventional
Type 2* (n = 22) P
44.73 ± 11.13 0.081†
16/6 0.945‡
20/41 (0.31 ± 0.21 logMAR) ,0.0001†
20/32 (0.21 ± 0.23 logMAR) 0.001†
6.68 ± 3.91 0.100†
474.28 ± 161.71 0.073†
51.55 ± 44.98 0.043†
 PHOTOBLEACHING AND FAF IN CSC  CHOI ET AL
Table 3. The Comparison of Type 1 and Diffuse Type 2 Hyper-AF in Fovea
Type 1 (n = 19)
Age 40.58 ± 6.10
Sex, male/female 14/5
Initial BCVA 20/23 (0.06 ± 0.07 logMAR)
Final BCVA§ 20/21 (0.03 ± 0.06 logMAR)
Follow-up duration, months¶ 4.47 ± 1.81
CRT, mm 308.79 ± 78.23
Disease duration, days 19.68 ± 12.98
*There were 2 unresolved CSC cases in eyes with diffuse Type 2 fovea.
†Mann–Whitney U test.
‡Chi-square test.
§Final logMAR BCVA in resolved eyes (Type 1 vs. diffuse Type 2 = 20/21 [0.03 ± 0.06 logMAR] vs. 20/37 [0.27 ± 0.25 logMAR], P = 0.020,
Mann–Whitney U test).
¶Follow-up duration in resolved eyes (Type 1 vs. Type 2 = 4.47 ± 1.81 vs. 5.00 ± 3.39 months, P = 0.734, Mann–Whitney U test).
fundus camera, at least 30% of the cone intensity may
be lost, as recently shown by Bedggood and Metha,27
although their experimental conditions differed from
those of this study. However, a single green laser
(532 nm) 1.7-mW scan was estimated to bleach approx-
imately 2% of the rhodopsin in the scan field.28
More importantly, because HRA2 acquires a refined
FAF images by averaging multiple serial images
(automatic real-time algorithm), significant photopig-
ment bleaching occurs throughout the retinal area
within a 30° field of view. In the human retina, the
photobleaching phenomenon is complete within 20
seconds to 30 seconds of exposure to HRA2 cSLO
light.29,30 Therefore, normal background FAF will be
higher after photopigment bleaching, similar to that
seen in detached lesions in CSC. In contrast, Optomap
acquires an image by scanning the fundus only once.
By comparing images before and after photopigment
bleaching, we were able to evaluate the relative hyper-
AF of the lesion. FAF allows topographic mapping of
lipofuscin distribution in the RPE cell monolayer, as
well as of other disease-associated fluorophores of the
outer retina and subretinal space.10,26,31 Excessive accu-
mulation of lipofuscin granules, mostly at the RPE level,
causes hyper-AF. Inversely, a change in the capacity of
light absorption also changes originally hypo-AF im-
ages to pseudo-hyper-AF images.18,32,33 In acute CSC,
there may be a relative loss of photopigment density in
outer segments because of anatomical photoreceptor
separation, which is similar to the photobleaching of
a photoreceptor caused by light exposure.14,15,18,34,35
Thus, the increased FAF of the normal retina after light
exposure could mask the hyper-AF of CSC lesions.2
Acute CSC cases with hyper-AF usually did not
present as purely homogeneous hyper-AF in our study.
Many cases of hyper-AF are heterogeneous rather than
homogeneous. Among these, eyes with Type 1 fovea
had higher BCVA and shorter disease duration than
575
Diffuse Type 2* (n = 14) P
44.80 ± 13.22 0.226†
9/5 0.561‡
20/45 (0.35 ± 0.18 logMAR) ,0.0001†
20/36 (0.26 ± 0.23 logMAR) ,0.0001†
5.42 ± 3.65 0.986†
496.57 ± 186.25 0.091†
37.13 ± 39.37 0.212†
those with Type 2 fovea. All mottled pattern hyper-AF
was included in Type 2 fovea cases. Mottled pattern
hyper-AF is a known chronic sign of CSC pathophys-
iology. With longer CSC disease duration, lipofuscin,
and other fluorophores accumulate, and FAF images
then show hyper-AF.1,8 The hyper-AF of lipofuscin-
like materials is brighter than that of the photopigment
bleaching area or subretinal fluid lesions.6 Deposits of
lipofuscin sometimes show a mottled pattern in FAF
images,36 which is associated with irreversible foveal
damage in CSC.3
After cases of mottled Type 2 fovea were excluded
from the present analysis, eyes with Type 1 fovea had
still better initial and final BCVA than did those with
diffuse Type 2 fovea. Type 1 was hyper-AF in the
unbleached state but lost its relative hyper fluorescence
after photobleaching. This type of hyper-AF might be
caused by a decrease in photopigment density or
photoreceptor loss in the presence of viable RPE before
the accumulation and generation of photoreceptor fluo-
rophores. Therefore, after exposure to repetitive Heidel-
berg 488-nm blue light (photobleaching effect), as the
normal retina outside the CSC lesion became hyper-AF,
the relative hyper-AF of the lesion disappeared. Type 1
fovea may be considered an early sign of CSC
pathophysiology, and may be positive predictors of
favorable visual prognosis. Type 2 fovea seemed to
represent true hyper-AF, demonstrating consistent hyper-
AF after exposure to 488-nm light, which could be
distinguished from pseudo–hyper-AF of the unbleached
retina. Type 2 hyper-AF might be caused by relative
increases in fluorescent material such as lipofuscin in
RPE or other fluorophores of the outer retina and sub-
retinal space, and may indicate irreversible damage.
There are several limitations to this study. First, the
design was retrospective in nature. Further studies
with prospective or cohort designs are needed. Second,
the degree of photopigment bleaching induced by
576 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES  2017  VOLUME 37  NUMBER 3
repetitive HRA2 FAF images differed between pa-
tients. Many factors seemed to influence photopigment
bleaching, including exposure time, presence of vita-
min E, cholesterol, and sphingolipids.37 There were no
significant differences in age or exposure time. How-
ever, we did not consider biologic factors such as
vitamin E, cholesterol, or sphingolipids. In addition,
40-second exposure to HRA2 may not be sufficient for
full bleaching.38 However, because the HRA2 FAF
images and postbleaching Optomap FAF images were
taken serially and without delay, the degree of photo-
pigment bleaching we observed was similar between
HRA2 FAF images and postbleaching Optomap FAF
images. In this study, we did not intend to acquire fully
bleached FAF images, but rather to collect postphoto-
bleaching Optomap FAF images at same level as the
HRA2 FAF images. Third, eyes with diffuse Type 2
fovea showed only lower visual acuity than those with
Type 1 fovea in our study, and there were no signif-
icant differences in disease duration between Type 1
and diffuse Type 2 fovea. These results may reflect the
small size of our data set. Further studies including
many cases of diffuse Type 2 fovea will be needed.
Fourth, the brightness of AF between the lesion and
normal attached retina was compared subjectively.
Furthermore, the fovea was not directly adjacent to
normal attached retina in many cases where foveal
AF could not be compared with normal attached ret-
ina. In such cases, foveal AF was not compared with
the distant attached normal retina but compared with
that of adjacent general background AF within the
CSC lesion. Therefore, observer bias may have inter-
fered with interpretation despite the high degrees of
concordance between the two readers. So, we attemp-
ted to minimize the potential interference on fovea AF
brightness (luminance) by AF brightness of the inter-
vening detached retina between fovea and attached
retina (Machband effect).39
It is true that in real clinical settings, most clinics
would not have both of the imaging devices (Optomap
and HRA2) used in this study. Nevertheless, it may be
possible to adapt the methods described in this study to
analyze patients examined using only the HRA2 cSLO
machine by comparing the FAF brightness of the CSC
lesion with that of the normal attached peripheral
retina after image averaging.
In conclusion, we demonstrated the effects of photo-
pigment bleaching on FAF images by comparing FAF
intensity between two different kinds of cSLO machines.
The nature of the hyper-AF area inside of the CSC
lesion was associated with disease duration and BCVA
through comparisons of FAF intensity of the CSC
lesion with that of normal attached retina before and
after photopigment bleaching. Understanding the
effects of photopigment bleaching in CSC is necessary
for the accurate interpretation of FAF images. Addi-
tional information drawn from FAF before and after
photobleaching may be useful for understanding disease
status in patients with acute CSC. Furthermore, photo-
bleaching status may be a useful prognostic factor to
determine visual prognosis after CSC resolution.
Key words: central serous chorioretinopathy, pho-
topigment bleaching, photobleaching, fundus auto-
fluorescence.
References
1. Dinc UA, Tatlipinar S, Yenerel M, et al. Fundus autofluores-
cence in acute and chronic central serous chorioretinopathy.
Clin Exp Optom 2011;94:452–457.
2. Pang CE, Shah VP, Sarraf D, Freund KB. Ultra-widefield
imaging with autofluorescence and indocyanine green angiog-
raphy in central serous chorioretinopathy. Am J Ophthalmol
2014;158:362–371.
3. Pryds A, Larsen M. Foveal function and thickness after verte-
porfin photodynamic therapy in central serous chorioretinop-
athy with hyperautofluorescent subretinal deposits. Retina
2013;33:128–135.
4. Quin G, Liew G, Ho IV, et al. Diagnosis and interventions for
central serous chorioretinopathy: review and update. Clin
Experiment Ophthalmol 2013;41:187–200.
5. Kim SK, Kim SW, Oh J, Huh K. Near-infrared and short-
wavelength autofluorescence in resolved central serous chorior-
etinopathy: association with outer retinal layer abnormalities.
Am J Ophthalmol 2013;156:157–164.
6. Ojima A, Iida T, Sekiryu T, et al. Photopigments in central
serous chorioretinopathy. Am J Ophthalmol 2011;151:940–952.
7. Eandi CM, Ober M, Iranmanesh R, et al. Acute central serous
chorioretinopathy and fundus autofluorescence. Retina 2005;
25:989–993.
8. Framme C, Walter A, Gabler B, et al. Fundus autofluorescence
in acute and chronic-recurrent central serous chorioretinopathy.
Acta Ophthalmol Scand 2005;83:161–167.
9. Kim SW, Oh J, Huh K. Correlations among various functional
and morphological tests in resolved central serous chorioretin-
opathy. Br J Ophthalmol 2012;96:350–355.
10. Imamura Y, Fujiwara T, Spaide RF. Fundus autofluorescence
and visual acuity in central serous chorioretinopathy. Ophthal-
mology 2011;118:700–705.
11. Burns SA, Elsner AE, Lobes LA Jr. Foveal cone photopigment
bleaching in central serous retinopathy. Appl Opt 1988;28:
1045–1049.
12. Ito Y, Horiguchi M, Miyake Y, Awaya S. Extrafoveal photo-
stress recovery testing with a scanning laser ophthalmoscope.
Jpn J Ophthalmol 1997;41:255–259.
13. Staurenghi G, Luiselli C, Levi G, et al. Retinal photopigment
density measured using autofluorescence imaging. Invest Oph-
thalmol Vis Sci 2004;45:2795.
14. van Meel GJ, Smith VC, Pokorny J, van Norren D. Foveal
densitometry in central serous choroidopathy. Am J Ophthal-
mol 1985;98:359–368.
15. Freund KB, Mrejen S, Jung J, et al. Increased fundus auto-
fluorescence related to outer retinal disruption. JAMA Ophthal-
mol 2013;131:1645–1649.
16. Nam KT, Yun CM, Kim JT, et al. Central serous chorioretin-
opathy fundus autofluorescence comparison in two different
 PHOTOBLEACHING AND FAF IN CSC  CHOI ET AL
confocal scanning laser ophthalmoscopies. Graefes Arch Clin
Exp Ophthalmol 2015;253:2121–2127.
17. Deli A, Moetteli L, Ambresin A, Mantel I. Comparison of fun-
dus autofluorescence images acquired by the confocal scanning
laser ophthalmoscope (488 nm excitation) and the modified Top-
con fundus camera (580 nm excitation). Int Ophthalmol 2013;
33:635–643.
18. Spaide R. Autofluorescence from the outer retina and subreti-
nal space: hypothesis and review. Retina 2008;28:5–35.
19. Theelen T, Berendschot TT, Boon CJ, et al. Analysis of visual
pigment by fundus autofluorescence. Exp Eye Res 2008;86:
296–304.
20. Burkhardt DA. Light adaptation and photopigment bleaching
in cone photoreceptors in situ in the retina of the turtle.
J Neurosci 1994;14:1091–1105.
21. ST George RC. The interplay of light and heat in bleaching
rhodopsin. J Gen Physiol 1952;35:495–517.
22. Williams TP. Photoreversal of rhodopsin bleaching. J Gen
Physiol 1964;47;679–689.
23. Bui TV, Han Y, Radu RA, et al. Characterization of native
retinal fluorophores involved in biosynthesis of A2E and
lipofuscin-associated retinopathies. J Biol Chem 2006;281:
18112–18119.
24. Wikler KC, Rakic P. Distribution of photoreceptor subtypes in
the retinal of diurnal nocturnal primates. J Neurosci 1990;10:
3390–3401.
25. Hanazono G, Tsunoda K, Kazato Y, et al. Functional topogra-
phy of rod and cone photoreceptors in macaque retina deter-
mined by retinal densitometry. Invest Ophthalmol Vis Sci
2012;53:2796–2803.
26. Delori FC, Dorey CK, Staurenghi G, et al. In vivo fluorescence
of the ocular fundus exhibits retinal pigment epithelium lip-
ofuscin characteristics. Invest Ophthalmol Vis Sci 1995;36:
718–729.
27. Bedggood P, Metha A. Variability in bleach kinetics and
amount of photopigment between individual foveal cones.
Invest Ophthalmol Vis Sci 2012;53:3673–3681.
28. Margan JI, Pugh EN Jr. Scanning laser ophthalmoscope mea-
surement of local fundus reflectance and autofluorescence
577
changes arising from rhodopsin bleaching and regeneration.
Invest Ophthalmol Vis Sci 2013;54:2048–2059.
29. Charbel Issa P, Singh MS, Lipinski DM, et al. Optimization of
in vivo confocal autofluorescence imaging of the ocular fundus
in mice and its application to models of human retinal degen-
eration. Invest Ophthalmol Vis Sci 2012;53:1066–1075.
30. Delori F, Greenberg JP, Woods RL, et al. Quantitative meas-
urements of autofluorescence with the scanning laser ophthal-
moscope. Invest Ophthalmol Vis Sci 2011;52:9379–9390.
31. Schmitz-Valckenberg S, Holz FG, Bird AC, Spaide RF. Fun-
dus autofluorescence imaging: review and perspectives. Retina
2008;28:385–409.
32. Moreno-López M, Pérez-López M, Casas-Llera P, et al. Per-
sistent subretinal fluid due to central serous chorioretinopathy
after retinal detachment surgery. Clin Ophthalmol 2011;5:
1465–1467.
33. Oh J, Kim SW, Kwon SS, et al. Correlation of fundus auto-
fluorescence gray values with vision and microperimetry in
resolved central serous chorioretinopathy. Invest Ophthalmol
Vis Sci 2012;53:179–184.
34. Joseph A, Rahimy E, Freund KB, et al. Fundus autofluores-
cence and photoreceptor bleaching in multiple evanescent
white dot syndrome. Ophthalmic Surg Lasers Imaging Retina
2013;44:588–592.
35. Burns SA, Elsner AE, Lobes LA Jr. Foveal cone photopigment
bleaching in central serous retinopathy. Appl Opt 1988;27:
1045–1049.
36. Matsumoto H, Kishi S, Sato T, Mukai R. Fundus autofluor-
escence of elongated photoreceptor outer segments in cen-
tral serous chorioretinopathy. Am J Ophthalmol 2011;151:
617–623.
37. Yamamoto K, Zhou J, Hunter JJ, et al. Toward an understand-
ing of bisretinoid autofluorescence bleaching and recovery.
Invest Ophthalmol Vis Sci 2012;53:3536–3544.
38. Sekiryu T, Iida T, Maruko I, Horiguchi M. Clinical application
of autofluorescence densitometry with a scanning laser oph-
thalmoscope. Invest Ophthalmol Vis Sci 2009;50:2994–3002.
39. Lane EJ, Proto AV, Phillips TW. Mach bands and density
perception. Radiology 1976;121;9–17.