Cystoid Macular Degeneration in Exudative Age-Related

Macular Degeneration

GIUSEPPE QUERQUES, FLORENCE COSCAS, RAIMONDO FORTE, NATHALIE MASSAMBA,

MARGARET STERKERS, AND ERIC H. SOUIED

N
● PURPOSE: To investigate the prevalence and clinical EOVASCULAR AGE-RELATED MACULAR DEGENER-
significance of cystoid macular degeneration in eyes that ation (AMD) causes severe and irreversible vi-
underwent intravitreal ranibizumab injections for exuda- sion loss if untreated, with loss of 3 lines of visual
tive age-related macular degeneration. acuity by 1 year and 4 lines of visual acuity by 2 years.1
● DESIGN: Retrospective, interventional case series. Ranibizumab (Lucentis; Genentech, Inc, South San Fran-
● METHODS: We reviewed the charts of 56 consecutive cisco, California, USA) is a recombinant, humanized,
patients (19 male, 37 female; mean age ⴞ standard monoclonal antibody antigen-binding fragment that neu-
deviation, 80.81 ⴞ 4.8 years) with exudative age-related tralizes all biologically active forms of vascular endothelial
macular degeneration who received the last intravitreal growth factor A.2 Intravitreal ranibizumab was the first
ranibizumab injection at least 6 months before and were therapy for exudative AMD to show, in 2 phase III clinical
judged to have a fibroatrophic scar without signs of studies, an improvement in mean visual acuity.3,4 These
progression by fluorescein angiography or spectral-do- results were obtained by a fixed-dosing regimen of ranibi-
main optical coherence tomography. Main outcome mea- zumab 0.5 mg or 0.3 mg monthly injected over 24 and 12
sures were the estimated prevalence and clinical months, respectively.3,4
significance of cystoid macular degeneration. The Prospective OCT Imaging of Patients with Neo-
● RESULTS: Twenty-two eyes showed various combina- vascular AMD Treated with intraOcular Ranibizumab
tions of degenerative pseudocysts, whereas 34 eyes did study demonstrated a variable dosing regimen, mainly
not show any pseudocysts. The 95% confidence interval based on optical coherence tomography (OCT) findings,
for the prevalence estimate was 36.98% to 41.02%. to be as effective as a fixed-dosing regimen in improving
Degenerative pseudocysts appeared square-shaped, did visual acuity and OCT findings with a far lower number of
not change their overall appearance over time, and were injections over 24 months.5,6 It is noteworthy that retreat-
located just below the internal limiting membrane in 11 ment with ranibizumab (0.5 mg/0.05 mL) was performed if
eyes (50%), in the inner nuclear layer in 16 eyes any qualitative increase in the amount of fluid was de-
(72.7%), in the outer nuclear layer in 8 eyes (36.3%), tected using OCT (during the second year of the study).6
and in all the retinal layers in 6 eyes (27.2%). Best- This highlights how meaningful, to date, is the OCT
corrected visual acuity improved in eyes with and with- evaluation of intraretinal or subretinal fluid accumulation
out degenerative pseudocysts and decreased significantly in making the retreatment decision on an as-needed basis.
in eyes with degenerative pseudocysts (P ⴝ .03). Mean Recently, Cohen and associates, using spectral-domain
central macular thickness decreased significantly (P < (SD) OCT, reported on a characteristic and relatively
.001) to 324.1 ␮m and to 328.2 ␮m in eyes and without frequent retinal finding of atrophic AMD that they called
degenerative pseudocysts, respectively. retinal pseudocysts.7 In their series, retinal pseudocysts
● CONCLUSIONS: Cystoid macular degeneration repre-
appeared as optically empty spaces, frequently located in
sents a well-distinguished clinical entity that may be the inner nuclear layer. The authors proposed the term
detected in exudative age-related macular degeneration pseudocyst rather than the term cyst because these retinal
eyes showing a posttreatment fibroatrophic scar and lesions had no obvious wall. Interestingly, fluorescein
should not be considered as a manifestation of choroidal angiography (FA), which was performed in 80% of eyes,
neovascularization activity. (Am J Ophthalmol 2011; showed no evidence of choroidal neovascularization
152:100 –107. © 2011 by Elsevier Inc. All rights (CNV), and there was no macular edema in any of the eyes
reserved.)
with pseudocysts. They proposed that retinal pseudocysts
may correspond to Müller cell degeneration (degenerative
Accepted for publication Jan 13, 2011. pseudocysts), and not to exudation (exudative cysts), and
From the Department of Ophthalmology, Centre Hospitalier Inter-
communal de Creteil, University of Paris XII, Creteil, France (G.Q., F.C., thus their presence should not require prompt treatment.
R.F., N.M., M.S., E.H.S.). Iida and associates recently proposed the term cystoid
Inquiries to Giuseppe Querques, Department of Ophthalmology, Uni- macular degeneration to describe a similar entity in central
versity of Paris XII, Centre Hospitalier Intercommunal de Creteil, 40
Avenue de Verdun, 94000 Creteil, France; e-mail: giuseppe.querques@ serous chorioretinopathy.8 Cystoid macular degeneration
hotmail.it was characterized by intraretinal cystoid spaces (degener-

100 © 2011 BY ELSEVIER INC. ALL RIGHTS RESERVED. 0002-9394/$36.00

doi:10.1016/j.ajo.2011.01.027
ative pseudocysts) with the absence of any intraretinal Spectralis HRA-OCT, Heidelberg Engineering, Heidel-
hyperfluorescence or leakage on FA. In the past, the term berg, Germany).
cystoid macular degeneration has been used to describe Initial treatment consisted of 1 intravitreal injection of
degeneration of the retina resulting in cystic spaces within 0.5 mg/0.05 mL ranibizumab, administered by retinal
the retina, or cystic degeneration (degenerative pseudo- specialists according to the usual technique.8 Briefly, each
cysts).9 –11 It seems reasonable that degenerative pseudo- eye was prepped with 5% povidone–iodine solution and
cysts (cystoid macular degeneration) may be detected not draped, and a lid speculum was placed. Eyes were anesthe-
only in patients affected by atrophic AMD,7 but also in tized with topical anesthetic, and ranibizumab was injected
patients who were undergoing intravitreal ranibizumab intravitreally through the inferotemporal pars plana 3.5 or
injections for exudative AMD and who had a posttreat- 4 mm posterior to the limbus for pseudophakic or phakic
ment fibroatrophic scar. Therefore, even in exudative patients, respectively. After the injection, patients were
AMD, degenerative pseudocysts (cystoid macular degener- instructed to administer topical antibiotics for 4 days.12
ation) should be recognized and distinguished from cystoid After the initial treatment, each patient underwent a
macular edema (CME), because their presence should not detailed medical and ocular history monthly, as well as
require prompt treatment. assessment of BCVA using ETDRS charts and ophthalmic
Our aim was to investigate the prevalence and clinical examination, which included slit-lamp biomicroscopy,
significance of cystoid macular degeneration and the dis- fundus biomicroscopy, OCT, and FA with or without
tribution of degenerative pseudocysts in the retinal layers ICGA. FA with or without ICGA was performed at the
of eyes that underwent last intravitreal ranibizumab injec- discretion of the examiner and not at every postinjection
tion for exudative AMD at least 6 months before and that evaluation. However, as a rule, FA was performed in all
were judged to have a fibroatrophic scar without signs of patients when deciding to discontinue the treatment.
progression by FA or SD OCT. After the initial treatment, intravitreal ranibizumab
injection was discontinued in eyes that showed absence of
leakage from the CNV on FA, a dry macula (absence of
subretinal fluid and CME) on OCT, or both. Eyes that
METHODS showed a persistent leakage from the CNV on FA, fluid in
the macula, or both received an additional injection 1
WE REVIEWED THE CHARTS OF ALL CONSECUTIVE PATIENTS month after each previous injection. Fibrovascular pig-
with exudative AMD who underwent intravitreal ranibi- ment epithelium detachment without increase in size was
zumab injections at the University Eye Clinic of Creteil at not considered to be a sign of exudation.
least 24 months before and who did not receive any Patients underwent repeat injections when there was a
intravitreal ranibizumab injection within the previous 6 recurrence in CNV activity defined as either a decrease of
months. In case of bilateral involvement, only 1 eye was BCVA (at least 1 ETDRS line) associated with persistent
included randomly in the study. leakage from the CNV or 1 of the following features: (1)
Inclusion criteria were age 50 years or older; best- an increase of intraretinal fluid on OCT (increase in CMT
corrected visual acuity (BCVA), as evaluated by Early by more than 100 ␮m); (2) a recurrence of subretinal fluid
Treatment Diabetic Retinopathy Study (ETDRS) charts, on OCT in a previously dry lesion; (3) a recurrence of
of 20/400 or better at presentation; and the presence of CME on OCT in a previously dry lesion; or (4) new
subfoveal CNV secondary to AMD (occult CNV [type 1], macular hemorrhage or new area of CNV. FA and ICGA
classic CNV [type 2], and minimally classic CNV [type 1 results were evaluated together with OCT results to assess
plus type 2]), as evaluated by fundus biomicroscopy, FA, the accuracy of other clinical data, as well as in case of
and indocyanine green angiography (ICGA), showing, at doubtful findings.
presentation, subretinal fluid, CME, or 1-mm central To assess the estimated prevalence and clinical signifi-
macular thickness (CMT) of at least 250 ␮m on OCT. cance of cystoid macular degeneration, we analyzed exu-
Excluded were patients with fewer than 24 months follow- dative AMD patients who received the last intravitreal
up, with previous treatment for exudative AMD in the ranibizumab injection at least 6 months before and were
study eye in the previous 6 months, previous surgery in the judged to be stable on the basis of BCVA (⫾ 1 line/5
study eye in the previous 6 months, refractive error of more letters change), FA with or without ICGA (absence of
than ⫺6 diopters, CNV attributable to causes other than leakage), and OCT (fibroatrophic scar without signs of
AMD, or active intraocular inflammation in the study eye. progression). Multiple horizontal and vertical raster SD
Each patient underwent a baseline assessment of OCT scans were obtained by Spectralis SD OCT. In all
BCVA measured at 4 m with standard ETDRS charts these patients, SD OCT examinations were monthly or
and an ophthalmic examination that included slit-lamp bimonthly and were performed after discontinuation of
biomicroscopy, fundus biomicroscopy, FA, ICGA, and intravitreal ranibizumab injection for at least 6 months.
OCT (Stratus OCT, Carl Zeiss-Meditec, Dublin, Cali- The search for retinal degenerative pseudocysts was per-
fornia, USA; Cirrus HD-OCT, Carl Zeiss-Meditec; formed independently by 3 of the authors (G.Q., F.C., and

VOL. 152, NO. 1 CYSTOID MACULAR DEGENERATION AND EXUDATIVE AMD 101
 E.H.S.) by analyzing recorded horizontal and vertical
scans. In exudative AMD, as in atrophic AMD,7 there are
different kinds of degenerative pseudocysts: unique versus
multiple and superficial versus deep. However, despite
these differences, all degenerative pseudocysts typically
show a more-or-less square shape (at least 1 of their borders
appears concave or straight). The same 3 authors also
evaluated the location of degenerative pseudocysts: just
below the internal limiting membrane, in the inner nu-
clear layer, in the outer nuclear layer, and in all the retinal
layers. In rare cases when interpretation varied between
the 3 authors, the SD OCT scans were reviewed as a group
to reach consensus.
Also, we assessed the clinical significance of cystoid
macular degeneration in terms of outcomes by comparing
mean final BCVA and CMT, as well as mean final number
of injections between eyes judged to be stable with cystoid
macular degeneration and eyes judged to be stable without
cystoid macular degeneration. Finally, we investigated the
estimated prevalence of outer retinal tubulation (ORT) in
eyes with cystoid macular degeneration.13
Statistical calculations were performed using the Statis-
tical Package for Social Sciences (version 17.0; SPSS, Inc,
Chicago, Illinois, USA). The prevalence of cystoid mac-
ular degeneration was estimated by dividing the number of
stable eyes with cystoid macular degeneration by the total
number of stable eyes. Also, the prevalence of ORT was
estimated by dividing the number of stable eyes with ORT
by the total number of stable eyes. Comparisons of mean
BCVA (ETDRS letters), mean CMT, and mean number of
injections were performed using the paired t test. The odds
ratios (ORs) and 95% confidence intervals (CIs) were
calculated using a logistic regression model to assess
influence of cystoid macular degeneration on BCVA and
CMT. The 95% CIs of the prevalence estimates was
calculated. Correlation between development of ORT
FIGURE 1. Images from the left eye of an 87-year-old man and final BCVA was obtained with the Spearman rank
with exudative age-related macular degeneration. (Top left) correlation. The chosen level of statistical significance
Late-frame fluorescein angiogram (FA) and optical coherence was P ⬍ .05.
tomography (OCT) scan showing an active choroidal neovas-
cularization (CNV) characterized by fluorescein leakage and
exudative cysts at baseline. (Top right) After 3 monthly
intravitreal injection of ranibizumab were administered,, the
CNV was still active, but both FA and OCT showed a 5 further intravitreal injections of ranibizumab that the patient
reduction in fluorescein leakage and exudative cysts compared underwent over a further 12-month period, the patient underwent
with baseline. (Second row, top left) Late-frame FA showing the fifteenth and last intravitreal injections of ranibizumab in the
the CNV staining without leakage and (Second row, bottom left eye. He returned 2 weeks later for follow-up examination of
left) OCT scan showing a fibroatrophic scar and a degenerative his right eye, also affected with CNV and treated by intravitreal
pseudocyst located just below the internal limiting membrane injections of ranibizumab. Spectral-domain (SD) OCT also ob-
(ILM; dotted arrow) obtained after 9 intravitreal injections of tained in the left eye at that time revealing overall unchanged
ranibizumab that the patient underwent over a 12-month findings. (Bottom) Spectral-domain OCT scan showing a fibroa-
period. (Second row, top right) Late-frame FA showing the trophic scar after a further 12 months, without receiving any
CNV staining without leakage and (Second row, bottom right) intravitreal injection of ranibizumab. Note that the SD OCT scan
OCT scan still showing a fibroatrophic scar, a degenerative still shows almost the same degenerative pseudocyst located just
pseudocyst located just below the ILM (dotted arrow), and an below the ILM (dotted arrow). Also note the continued presence
of an ORT (arrowhead).

102 AMERICAN JOURNAL OF OPHTHALMOLOGY JULY 2011

 RESULTS
A TOTAL OF 56 EYES OF 56 PATIENTS (19 MALE, 37 FEMALE;
age range, 67 to 87 years; mean age ⫾ standard deviation,
80.81 ⫾ 4.8 years) who did not receive any intravitreal
ranibizumab injection at least 6 months previously and
who were judged to be stable on the basis of BCVA (⫾ 1
line/5 letters change), FA with or without ICGA (absence
of leakage), and OCT (fibroatrophic scar without signs of
progression) were included in the study. Overall, the mean ⫾
standard deviation follow-up period was 29.9 ⫾ 4.7
months, and the mean ⫾ standard deviation number of
intravitreal injections was 11.6 ⫾ 3.0. Thirty eyes had a
type 1 CNV, 13 eyes had a type 2 CNV, and 13 eyes had
a mixed type 1 plus type 2 CNV. All eyes showed presence
of exudative cysts in the macular area before treatment.
Among the 56 eyes judged to be stable, 22 eyes (39.3%)
showed various combinations of degenerative pseudocysts
(just below the internal limiting membrane, in the inner
nuclear layer, in the outer nuclear layer, and in all the
retinal layers) on the different SD OCT analyzed scans
(Figures 1– 4), whereas 34 eyes (60.7%) did not show any
degenerative pseudocysts (Table 1). The 95% CI for the
prevalence estimate for cystoid macular degeneration in
eyes judged to be stable was 36.98% to 41.02%.
There was no difference in CNV type distribution
between eyes with and without cystoid macular degenera- FIGURE 2. Images from the right eye of a 73-year-old man
tion (P ⬎ .05; Table 1). There was no difference in mean with exudative age-related macular degeneration. (Top left)
Fluorescein angiogram (FA) and (Top right) spectral-domain
age, gender, mean BCVA, and CMT at baseline between
optical coherence tomography (SD OCT) scan obtained at
eyes with and without cystoid macular degeneration (P ⬎ baseline showing a classic (type 2) choroidal neovascularization
.05; Table 1). Also, both eyes with or without cystoid and typical exudative retinal cysts (asterisks). (Middle left) FA
macular degeneration were followed up for a mean of 30 and SD OCT (Middle right) scan showing a fibroatrophic scar
months, underwent a mean of 12 intravitreal injections of (staining without leakage) in the same patient, 24 months later,
ranibizumab, and received the last treatment a mean of 9 after 11 intravitreal injections of ranibizumab. Note that the
months before inclusion in the current analysis (Table 1). SD OCT scan shows a degenerative pseudocyst (Middle right,
Mean CMT decreased significantly from baseline to the arrow) involving all the retinal layers. (Bottom left) FA and
last follow-up visit in eyes with and without cystoid (Bottom right) SD OCT scan still showing a fibroatrophic scar
(staining without leakage) in the same patient 12 months later,
macular degeneration (P ⬍ .001; Table 2), and there was
without receiving any further intravitreal injections of ranibi-
no difference in mean CMT between eyes with and zumab. Note that the SD OCT scan still shows almost the same
without cystoid macular degeneration at the last follow-up degenerative pseudocyst (Bottom right, arrow) involving all the
visit (P ⬍ .001). Interestingly, BCVA improved in eyes retinal layers.
without cystoid macular degeneration (yet not signifi-
cantly), whereas it decreased significantly in eyes with
cystoid macular degeneration (P ⫽ .03). A nonsignificant
correlation was found between duration of macular thick- but not CMT (OR, 0.6; 95% CI, 0.274 to 2.531; P ⫽ .1;
ening before treatment and final BCVA (Spearman ␳, Table 3).
⫺0.61; P ⫽ .06), whereas a negative yet not significant Overall, degenerative pseudocysts were located just
correlation was found between baseline CMT and final below the internal limiting membrane in 11 (50%) of 22
BCVA (Spearman ␳, ⫺0.18; P ⫽ .08; Table 3). Correla- eyes, in the inner nuclear layer in 16 (72.7%) of 22 eyes,
tion between mean CMT decrease and mean BCVA in the outer nuclear layer in 8 (36.3%) of 22 eyes, and in
improvement was statistically significant for eyes without all the retinal layers in 6 (27.2%) of 22 eyes. In all cases
cystoid macular degeneration (P ⫽ .02), but not for eyes included in our analysis, the overall appearance of degen-
with cystoid macular degeneration (P ⬎ .05). The pres- erative pseudocysts did not change over a mean of 9
ence of cystoid macular degeneration influenced final months from their detection (time of discontinuation of
BCVA (OR, 3.642; 95% CI, 1.028 to 5.289; P ⫽ .01), intravitreal ranibizumab injection).

VOL. 152, NO. 1 CYSTOID MACULAR DEGENERATION AND EXUDATIVE AMD 103
FIGURE 3. Images from the right eye of an 82-year-old woman
with exudative age-related macular degeneration. (Top left)
Fluorescein angiogram (FA) and (Top right) spectral-domain
optical coherence tomography (SD OCT) scan obtained at FIGURE 4. Images from the left eye of a 77-year-old woman
baseline showing an occult (type 1) choroidal neovasculariza- with exudative age-related macular degeneration. (Top left)
tion and subretinal fluid (asterisks). (Middle left) FA and Fluorescein angiogram (FA) and (Top right) spectral-domain
(Middle right) SD OCT scan showing a fibroatrophic scar optical coherence tomography (SD OCT) scan obtained at
(staining without leakage) in the same patient 24 months later, baseline showing a mixed classic plus occult (type 1 plus type 2)
after 12 intravitreal injections of ranibizumab. Note that the choroidal neovascularization and subretinal fluid (asterisks).
SD OCT scan shows a degenerative pseudocyst located both (Middle left) FA and (Middle right) SD OCT scan showing a
just below the internal limiting membrane (ILM; Middle left, fibroatrophic scar (staining without leakage) in the same patient
dotted arrow). Also note the presence of an outer retinal 15 months later, after 7 intravitreal injections of ranibizumab.
tubulation (ORT; Middle right, arrowhead). (Bottom left) FA Note that the SD OCT scan shows a degenerative pseudocyst
and (Bottom right) SD OCT scan still showing a fibroatrophic located within the inner nuclear layer (INL; Middle left, dotted
scar (staining without leakage) in the same patient 6 months arrow). Also note the presence of an outer retinal tubulation
later, without receiving any further intravitreal injection of (ORT; Middle right, arrowhead). (Bottom left) FA and (Bot-
ranibizumab. Note that the SD OCT scan still shows almost the tom right) SD OCT still showing a fibroatrophic scar (staining
same degenerative pseudocyst located both just below the ILM without leakage) in the same patient 12 months later, without
(Bottom right, dotted arrow) and the ORT (Bottom right, receiving any further intravitreal injections of ranibizumab.
arrowhead). Note that the SD OCT scan still shows almost the same
degenerative pseudocyst located within the INL (Bottom right,
dotted arrow) and the ORT (Bottom right, arrowhead).
Among the 56 eyes judged to be stable, 37 eyes (66.1%)
showed ORT (Table 1). Overall, prevalence of ORT was
follow-up visit, the mean CMT in eyes with cystoid
significantly greater than prevalence of cystoid macular
macular degeneration was 324.1 ␮m; thus, eyes showing
degeneration (P ⫽ .01). The 95% CI for the prevalence
cystoid macular degeneration did not actually present an
estimate for ORT was 64.22% to 68.37%. In 15 (68.2%) of
increased macular thickness on SD OCT compared with
22 eyes with cystoid macular degeneration and in 22
eyes with no known retinal disease (recently proposed to
(64.7%) of 34 eyes without cystoid macular degeneration,
be 315 ␮m).9 Of note, there was no difference in mean
we also found ORT (Table 1). No significant difference
CMT at last follow-up visit between eyes with and without
was present between the 2 groups regarding the presence of
cystoid macular degeneration (324.1 vs 328.2 ␮m; P ⬎
ORT (P ⬎ .05). The 95% CI for the prevalence estimate
.05).
for ORT in eyes with cystoid macular degeneration was
65.94% to 69.98%. There was no correlation between
presence of ORT and location of degenerative pseudocysts
(P ⬎ .05). A nonsignificant correlation was found between DISCUSSION
development of ORT and BCVA change during follow-up
(Spearman ␳, 0.167; P ⫽ .4). IN THIS STUDY, WE INVESTIGATED THE PRESENCE OF
Different from typical exudative cysts, degenerative cystoid macular degeneration in a subset of eyes with
pseudocysts showed, in all eyes, at least 1 of their borders exudative AMD that underwent last intravitreal ranibi-
to be concave or straight; thus, degenerative pseudocysts zumab injection at least 6 months before and that were
presented a square shape rather than the typical biconvex judged to have a fibroatrophic scar. We also estimated the
shape of exudative cysts (Figure 5). Moreover, at the last prevalence of degenerative pseudocysts and their distribu-

104 AMERICAN JOURNAL OF OPHTHALMOLOGY JULY 2011

 TABLE 1. Characteristics of 56 Eyes with Exudative Age-Related Macular Degeneration That Received the Last Intravitreal
Ranibizumab Injection at Least 6 Months Previously and That Were Judged to Be Stable

Cystoid Macular Degeneration No Cystoid Macular Degeneration P Value (between the 2 Groups)

No. 22 34
Age (yrs) 80.81 ⫾ 4.8 ( range, 67 to 87) 81.29 ⫾ 3.9 ( range, 72 to 88) .5
BCVA (ETDRS letters) 43.86 ⫾ 8. 43.08 ⫾ 7.1 .5
Duration of macular thickness (mos) 1.3 ⫾ 0.6 1.2 ⫾ 0.8 .09
CMT (␮m) 437.81 ⫾ 69.3 442.5 ⫾ 53.9 .7
CNV .5
Type 1 12 18
Type 2 5 8
Type 1 ⫹ 2 5 8
ORT (%) 15 (68.2) 22 (64.7) .3
Follow-up (mos) 29.86 ⫾ 5.2 29.92 ⫾ 4.4 .8
Injections (no.) 11.63 ⫾ 3.3 11.58 ⫾ 2.8 .7
Time to last treatment (mos) 9.00 ⫾ 2.4 9.08 ⫾ 2.0 .8

BCVA ⫽ best-corrected visual acuity; CMT ⫽ central macular thickness; CNV ⫽ choroidal neovascularization; ETDRS ⫽ Early Treatment
Diabetic Retinopathy Study; mos ⫽ months; ORT ⫽ outer retinal tubulation; yrs ⫽ years.
Data are presented as mean ⫾ standard deviation or number, unless otherwise indicated.

TABLE 2. Best-Corrected Visual Acuity and Central Macular Thickness at Baseline and at Last Visit in the 2 Groups of Exudative
Age-Related Macular Degeneration Eyes (with and without Cystoid Macular Degeneration)

BCVA ETDRS Letters CMT (␮m)

No. Baseline Last Visit P Value Baseline Last Visit P Value

Cystoid macular degeneration 22 43.86 ⫾ 8 39.09 ⫾ 8.4 .03 437.81 ⫾ 69.3 324.13 ⫾ 42.1 ⬍.001
No cystoid macular degeneration 34 43.08 ⫾ 7.1 45.0 ⫾ 13.4 .3 442.5 ⫾ 53.9 328.2 ⫾ 31.4 ⬍.001

BCVA ⫽ best-corrected visual acuity; CMT ⫽ central macular thickness; ETDRS ⫽ Early Treatment Diabetic Retinopathy Study.

TABLE 3. Logistic Regression Showing Influence of Cystoid Macular Degeneration on Best-Corrected Visual Acuity and Central
Macular Thickness and Correlation of Baseline Central Macular Thickness and Duration of Macular Thickening with Final
Best-Corrected Visual Acuity

Odds Ratio (95% CI) of Influence of Odds Ratio (95% CI) of Influence of Correlation between Duration of Macular
Cystoid Macular Degeneration on Correlation between Baseline CMT Cystoid Macular Degeneration on Thickening before Treatment and Final
BCVA and Final BCVA CMT BCVA

3.642 (1.028 to 5.289), P ⫽ .01 Spearman ␳, –0.18; P ⫽ .08 0.6 (0.274 to 2.531), P ⫽ .1 Spearman ␳, –0.61; P ⫽ .05

BCVA ⫽ best-corrected visual acuity; CI ⫽ confidence interval; CMT ⫽ central macular thickness.

tion within the retinal layers. Identification of degenera- known to be present in the vitreous cavity, but the cystoid
tive pseudocysts was achieved through a retrospective spaces did not show any change in size or appearance after
analysis of multiple horizontal and vertical SD OCT scans treatment.
acquired during clinical practice. Unlike exudative cysts that typically show a biconvex
The recognition of cystoid macular degeneration as shape, degenerative pseudocysts present a peculiarly square
distinct from CME seems of great importance because its shape (Figure 5). Moreover, eyes with and without degen-
presence at least in some exudative AMD eyes that were erative pseudocysts showed similar mean CMT values;
undergoing treatment with intravitreal injections of these values should not be considered as increased com-
ranibizumab may not be considered as a manifestation of pared with that of eyes with no known retinal disease.14
CNV activity. It is noteworthy that, when SD OCT Interestingly, all cases included in our analysis did not
examination was performed 1 to 2 weeks after intravitreal show changes in the overall appearance of degenerative
ranibizumab injection (Figure 1), active levels of drug were pseudocysts at least 6 months after discontinuation of

VOL. 152, NO. 1 CYSTOID MACULAR DEGENERATION AND EXUDATIVE AMD 105
 Cohen and associates and others, degenerative pseudocysts
may correspond to Müller cell degeneration that has been
suspected to occur in various degenerative retinal disorders
such as tamoxifen retinopathy or group 2A idiopathic
juxtafoveolar retinal telangiectasis.7,15–18 Degenerative
pseudocysts may correspond to an intermediate stage of the
atrophic process involving the retina, characterized by the
loss of Müller cells. Therefore, degenerative pseudocysts
may develop before the thinning of the neurosensory
retina.
The design of the current study does not allow for
understanding why, among the eyes judged to be stable,
cystoid macular degeneration developed in some, whereas
in others it did not. Future studies specifically should
address this issue.
FIGURE 5. Diagram showing degenerative pseudocysts with at Zweifel and associates recently reported a novel OCT
least one concave or straight border, giving a squared aspect to finding occurring in a variety of advanced degenerative
the lesion, whereas exudative cysts have round borders, giving retinal disorders that they called outer retinal tubulation.13
the typical biconvex aspect to the lesion. ILM ⴝ internal ORT are branching tubular structures located in the outer
limiting membrane; INL ⴝ inner nuclear layer; ONL ⴝ outer
nuclear layer of the retina and appear as round or ovoid
nuclear layer.
hyporeflective spaces with hyperreflective borders on OCT
sections. In the current series, 37 of 56 eyes showed ORT,
and thus, prevalence of ORT was significantly greater than
intravitreal ranibizumab injection. In case of the cystoid prevalence of cystoid macular degeneration.
spaces being the result of vascular leakage, retinal thick- ORT originally was described as a common findings in
ening (outside normal ranges) is to be expected, as well as exudative AMD eyes that were undergoing treatment with
changes in size and overall appearance over time after the intravitreal injections of ranibizumab or bevacizumab, often
treatment has been discontinued. Therefore, our data overlying retinal pigment epithelium damage or subretinal
suggest that the cystoid spaces may reflect a tissue break- fibrosis, or in eyes with atrophic disorders or subretinal
down, rather than vascular leakage. fibrosis. It is hypothesized that with resolution of fluid, the
There were no differences between eyes with and degenerating photoreceptors become arranged in a tubular
without cystoid macular degeneration with respect to fashion.13 In our series, prevalence of ORT was similar in
mean age, gender, mean BCVA, and CMT at baseline, as both eyes with or without cystoid macular degeneration.
well as mean follow-up, number of intravitreal injections These data further support our speculation on the role of
of ranibizumab, and time from administration of the last degenerative retinal process (secondary to posttreatment
treatment. Interestingly, although mean CMT decreased fibroatrophic scar development in exudative AMD eyes) in
significantly from baseline to the last follow-up visit, in the development of cystoid macular degeneration, possibly
eyes both with and without cystoid macular degeneration, before the thinning of the neurosensory retina.
eyes with cystoid macular degeneration showed a signifi- Our study has several limitations. First, it is a retrospec-
cant decrease in mean BCVA. Moreover, although there tive analysis of a subset of eyes judged to be stable, and we
was a significant correlation between mean CMT decrease cannot exclude a selection bias. Second, multiple horizon-
and mean BCVA improvement in eyes without cystoid tal and vertical SD OCT scans were acquired during
macular degeneration, we did not find a similar correlation clinical practice, without a well-defined acquisition proto-
in eyes with cystoid macular degeneration. Based on these col, and our analysis relied on the recorded scans; there-
data, it seems that presence of cystoid macular degenera- fore, some degenerative pseudocysts may have been missed
tion may negatively predict long-term functional outcome. because of the relatively limited number of analyzed scans,
We speculate that a progressive retinal degeneration ultimately resulting in an overall underestimation of de-
(including photoreceptor degeneration) may be associated generative pseudocysts.
with the development of a fibroatrophic scar in patients In the era of anti–vascular endothelial growth factor
undergoing treatment with intravitreal injections of treatments, recognizing OCT features like degenerative
ranibizumab for exudative AMD. This may be responsible pseudocysts should not be considered as a manifestation of
for cystoid macular degeneration development and, possi- CNV activity in exudative AMD. It is of great importance,
bly, for negative functional outcome (decrease in BCVA) especially because, to date, the variable dosing regimen is
despite morphologic improvement (CMT decrease and mainly based on OCT findings. Future prospective studies are
fibroatrophic scar development without signs of CNV necessary to improve our understanding of the clinical signif-
activity over at least 6 months). In fact, as proposed by icance of cystoid macular degeneration in exudative AMD.

106 AMERICAN JOURNAL OF OPHTHALMOLOGY JULY 2011

THE AUTHORS INDICATE NO FINANCIAL SUPPORT OR FINANCIAL CONFLICT OF INTEREST. INVOLVED IN DESIGN AND
conduct of study (G.Q., E.H.S.); Collection, management, analysis (G.Q., F.C., R.F., N.M., M.S., E.H.S.) and interpretation (G.Q., F.C., E.H.S.) of
data; Manuscript preparation (G.Q., E.H.S.), review, and approval (G.Q., E.H.S.). Written informed consent according to the tenets of the Declaration
of Helsinki was obtained from each patient enrolled. University of Paris XII Institutional Review Board approval was obtained. This study was performed
in accordance with the ethical standards set forth in the 1964 Declaration of Helsinki.

REFERENCES atrophic tracts secondary to central serous pigment

epitheliopathy. Ophthalmology 1984;91(12):1554 –1572.
1. Wong TY, Chakravarthy U, Klein R, el al. The natural 10. Green WR. Retinal detachment. In: Spencer WH, ed.
history and prognosis of neovascular age-related macular Ophthalmic Pathology. Philadelphia: WB Saunders, 1996:
degeneration: a systematic review of the literature and 960 –964.
meta-analysis. Ophthalmology 2008;115(9):116 –126. 11. Yannuzzi LA. A perspective on the treatment of aphakic
2. Chen Y, Wiesmann C, Fuh G, et al. Selection and analysis cystoid macular edema. Surv Ophthalmol 1984;28(Suppl):
of an optimized anti-VEGF antibody: crystal structure of an 540 –553.
affinity-matured Fab in complex with antigen. J Mol Biol 12. Querques G, Azrya S, Martinelli D, et al. Ranibizumab for
1999;293(4):865– 881. exudative age-related macular degeneration: 24-month out-
comes from a single-centre institutional setting. Br J Oph-
3. Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for
thalmol 2010;94(3):292–296.
neovascular age-related macular degeneration. N Engl J Med
13. Zweifel SA, Engelbert M, Laud K, Margolis R, Spaide RF,
2006;355(14):1419 –1431.
Freund KB. Outer retinal tubulation: a novel optical coher-
4. Brown DM, Kaiser PK, Michels M, et al. Ranibizumab vs
ence tomography finding. Arch Ophthalmol 2009;127(12):
verteporfin for neovascular age-related macular degenera-
1596 –1602.
tion. N Engl J Med 2006;355(14):1432–1444.
14. Grover S, Murthy RK, Brar VS, Chalam KV. Normative data
5. Fung AE, Lalwani GA, Rosenfeld PJ, et al. An optical
for macular thickness by high-definition spectral-domain
coherence tomography-guided, variable dosing regimen with optical coherence tomography (Spectralis). Am J Ophthal-
intravitreal ranibizumab (Lucentis) for neovascular age-re- mol 2009;148(2):266 –271.
lated macular degeneration. Am J Ophthalmol 2007;143(4): 15. Gualino V, Cohen SY, Delyfer MN, Sahel JA, Gaudric A.
566 –583. Optical coherence tomography findings in tamoxifen reti-
6. Lalwani GA, Rosenfeld PJ, Fung AE, et al. A variable-dosing nopathy. Am J Ophthalmol 2005;140(4):757–758.
regimen with intravitreal ranibizumab for neovascular age- 16. Yannuzzi LA, Bardal AM, Freund KB, Chen KJ, Eandi CM,
related macular degeneration: year 2 of the PrONTO Study. Blodi B. Idiopathic macular telangiectasia. Arch Ophthalmol
Am J Ophthalmol 2009;148(1):43–58. 2006;124(4):450 – 460.
7. Cohen SY, Dubois L, Nghiem-Buffet S, et al. Retinal 17. Gaudric A, Ducos de Lahitte G, Cohen SY, Massin P,
pseudocysts in age-related geographic atrophy. Am J Oph- Haouchine B. Optical coherence tomography in group 2A
thalmol 2010;150(2):211–217. idiopathic juxtafoveolar retinal telangiectasis. Arch Opthal-
8. Iida T, Yannuzzi LA, Spaide RF, Borodoker N, Carvalho CA, mol 2006;124(10):1410 –1419.
Negrao S. Cystoid macular degeneration in chronic central 18. Cohen SM, Cohen ML, El-Jabali F, Pautler SE. Optical
serous chorioretinopathy. Retina 2003(1);23:1–7. coherence tomography findings in nonproliferative group 2a
9. Yannuzzi LA, Shakin JL, Fisher YL, Altomonte MA. Periph- diopathic juxtafoveal retinal telangiectasis. Retina 2007;
eral retinal detachments and retinal pigment epithelial 27(1):59 – 66.

VOL. 152, NO. 1 CYSTOID MACULAR DEGENERATION AND EXUDATIVE AMD 107
 Biosketch
Giuseppe Querques, MD, PhD, is an Associate Professor of Ophthalmology at University Paris XII, and Consultant
Surgeon Ophthalmologist in the Department of Ophthalmology at Creteil University Eye Clinic in France. He earned his
MD and PhD degrees from the University of Foggia, Italy. Professor Querques has contributed to more than 70
peer-reviewed articles, mainly in the areas of imaging of AMD and hereditary retinal diseases. His area of research is
focusing on medical retina.

VOL. 152, NO. 1 CYSTOID MACULAR DEGENERATION AND EXUDATIVE AMD 107.e1
 Biosketch
Eric H. Souied, MD, PhD, is Professor and head of Department of Ophthalmology at Creteil University Eye Clinic in
France. He earned his MD and PhD degrees from the University of Paris XII and a post-doctoral fellowship at UCLA,
USA. Professor Souied has contributed to more than 100 peer-reviewed articles published in the areas of AMD and
ophthalmic genetics. His main area of research is focusing on the genetics, prevention and treatment of AMD.

107.e2 AMERICAN JOURNAL OF OPHTHALMOLOGY JULY 2011