Purpose
To report the presence of pseudocysts in retinal layers of eyes with geographic atrophy (GA) attributable to age-related macular degeneration (AMD) and to estimate their prevalence.
Design
Retrospective study.
Methods
setting: Clinical practice. patients: Consecutive patients with GA, assessed by spectral-domain optical coherence tomography (OCT). main outcome measures: Assessment of pseudocyst prevalence in retinal layers. Statistical analysis by the χ 2 test, Fisher exact test, Mann-Whitney U test, and Cramer test, performed to explore links between the presence of pseudocysts and demographic features and/or pattern of atrophy, ie, horseshoe, homogeneous area, homogeneous area or patchy atrophy with foveal sparing, and patchy atrophy.
Results
Eighty-eight eyes of 68 GA patients aged between 61 and 94 years (mean: 79.8) were examined. In 20 patients, GA was bilateral. Twenty-four eyes (27.2%) exhibited pseudocysts corresponding to small cystoid spaces frequently located in the inner nuclear layer of the retina. There was no macular edema. Fluorescein angiography, performed in 71 eyes (80%), ruled out possible choroidal neovascularization. No correlation was found between 1) patients’ age ( P = .7) or gender ( P > .99) and the presence of pseudocysts or 2) patterns of atrophy (Cramer test: V = 0.183) and the presence of pseudocysts.
Conclusions
Pseudocysts seemed to be frequent in atrophic AMD. They might correspond to Müller cell degeneration, as suspected in other degenerative retinal disorders like tamoxifen retinopathy or group 2A idiopathic juxtafoveolar retinal telangiectasis. The present results indicate that pseudocysts are frequently seen on OCT in eyes with atrophic AMD and their presence should not be considered as a manifestation of neovascular AMD that requires prompt treatment.
Geographic atrophy (GA) constitutes the end stages of the dry type of age-related macular degeneration (AMD) and often causes severe loss of vision. About 1 million individuals in the US population exhibit GA in at least 1 eye. Imaging of GA was formerly limited to color photography, but the recent introduction of fundus autofluorescence imaging provided a very useful tool for GA imaging, and also for the anticipation of GA progression by imaging the autofluorescent areas at the margin of GA that eventually evolve into atrophy of the retinal pigment epithelium (RPE).
Tomographic retinal imaging is provided by optical coherence tomography (OCT), currently used as a routine tool because it is a noninvasive means of assessing the posterior pole in patients with macular disease. OCT allows the analysis of retinal morphology and the assessment of structural changes within the retina. The introduction of Fourier (ie, spectral-domain) analysis has made high resolution and fast scanning possible. Spectral-domain OCT (SD-OCT) has improved the imaging of numerous macular diseases, including non-neovascular AMD. In several studies of eyes with GA, SD-OCT showed the atrophic areas as thin and hyperreflective. This optical hyperreflectivity, which originates in the choroids, was interpreted as the increased penetration of light through the atrophic RPE. In these studies, SD-OCT also showed marked structural alterations of the outer retinal layers, with the complete loss of clear boundaries between the layers. Some of these studies also described the presence of hyperreflective clumps in different retinal layers, and of segmented plaques in the outer band.
We observed an additional peculiar finding, ie, the presence of small retinal pseudocysts, in several eyes with GA. We decided to perform a retrospective analysis of patients presenting with GA documented by SD-OCT.
Methods
At the Centre Ophtalmologique d’Imagerie et de Laser, a private-practice tertiary care retinal center in Paris, France, the charts and images of 68 consecutive patients with a diagnosis of GA in at least 1 eye were retrospectively reviewed over a period of 3 months.
Inclusion criteria were dry age-related macular degeneration with GA of at least half a disk area, imaged both by color fundus photography and SD-OCT. Some patients had also been imaged by fundus autofluorescence and fluorescein angiography (FA). Exclusion criteria were atrophy attributable to pathologic myopia or to any condition other than AMD.
The charts of the patients included were analyzed for gender, age, and the laterality of the eye studied.
Fundus photographs were obtained using a Topcon fundus camera (Topcon, Tokyo, Japan) coupled with digitized LediOPH software (Lheritier, Saint-Ouen L’Aumône, France). Atrophy was classified into the following types or patterns, according to the shape of the atrophic area(s): homogeneous, corresponding to a single roundish area of GA; homogeneous with foveal sparing, corresponding to a doughnut-shaped area of GA; horseshoe, corresponding to a single incomplete area with foveal sparing; and patchy, corresponding to 2 or more clearly separated GA areas. This classification was established by 2 separate retinal specialists (S.Y.C. and S.N.-B.). In case of disagreement, discussion between the 2 readers allowed final classification of the case.
SD-OCT scans were obtained by Cirrus SD-OCT (Zeiss, Jena, Germany). The search for cysts was performed by one of the authors (L.D.), by defilading the recorded square of the scans. Horizontal and vertical scans passing through the foveola were printed. Scans of pseudocystic or suspicious areas, when present, were also printed. Prints were analyzed by 2 independent retinal specialists (S.Y.C. and S.N.-B.) to explore the possible presence of pseudocysts, estimate their prevalence, and specify their location in the retinal layers. If the readers disagreed on these issues, the case concerned was considered to exhibit no pseudocysts. When fluorescein angiography had been performed, images were also reviewed to verify that there was no sign of choroidal neovascularization (CNV). It was verified that the included cases presented no sign of CNV on fundus examination records, fundus photographs, and FA (hemorrhage, hard exudates, serous retinal detachment, pin-points, any fluorescein leakage).
The study did not focus on other OCT findings typical of GA, and assessment of pseudocyst prevalence in the retinal layers was the main outcome measure. Statistical analysis using the χ 2 test, Fisher exact test, Mann-Whitney U test, and Cramer test was performed in a search for links between the presence of pseudocysts and 1) demographic features, and/or 2) patterns of atrophy, ie, horseshoe, homogeneous area, homogeneous area with foveal sparing, and patchy atrophy.
Consent to perform the study was obtained from the Institutional Review Board (IRB) and Ethics Committee of the Société Française d’Ophtalmologie (French Society of Ophthalmology).
Results
Eighty-eight eyes of 68 patients were included. Patients comprised 50 women (73%) and 18 men. They were aged from 61 to 94 years (mean age: 79.8 ± 7.9). Twenty patients had bilateral GA. The right eye was studied in 50 cases, and the left eye in 38.
Twenty-four of the 88 eyes (27.2%) exhibited pseudocysts corresponding to cystoid spaces ( Figures 1–3 ). The 95% confidence interval for the prevalence estimate for pseudocysts is [17.72%–36.28%]. The pseudocysts were located just below the internal limiting membrane in 1 eye (4.1%), in the inner nuclear layer of the retina in 10 eyes (41.6%), in the outer nuclear layer in 3 (12.5 %), in the deep outer layers in 7 (29.1%), and in all the retinal layers ( Figure 4 ) in 3 (12.5%). Note that precise location of the pseudocysts was sometimes difficult to identify when the retina was thin. For example, in 4 of the 7 eyes with deep pseudocysts, the cysts were probably located in the inner layers, but seemed to be close to the choroids, because the outer layers of the retina had disappeared. There was no macular edema in any of the eyes with pseudocysts. More precisely, there was no difference between the retinal thickness in the areas with pseudocysts and retinal thickness in the adjacent areas. It was verified that there was no vitreoretinal traction. FA, performed in 71 eyes (80%), showed no evidence of CNV.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
