To evaluate the incidence of flat, irregular pigment epithelium detachments (PEDs) in chronic central serous chorioretinopathy (CSC) and to determine whether they are consistent with active choroidal neovascularization (CNV).
Retrospective case series.
Review of medical records of patients with chronic CSC who were examined in the Ophthalmology Department of Lariboisière Hospital between June 1, 2007 and May 31, 2013. Multimodal imaging of the fundus, including optical coherence tomography (OCT), fundus autofluorescence, and indocyanine green and fluorescein angiography, was available in most cases.
One hundred and ten patients with chronic CSC were identified. Fifty-three eyes of 38 patients showed flat irregular PED on macular OCT examination. Mean age was 58.6 ± 13.2 years. Twenty-eight patients (73.6%) patients were male. Fifteen patients (39.4%) had bilateral flat irregular PEDs. The mean follow-up duration was 14.6 years (range: 2-39 years). PEDs were suggestive of type 1 CNV in 10 eyes, but no other signs of AMD, specifically no drusen, were present. In the remaining 43 eyes, flat irregular PEDs were stable over time (mean follow-up duration: 15 years) with no evidence of active neovascularization.
Although the possible occurrence of type 1 CNV complicating the course of chronic CSC should not be ignored, all cases of flat irregular PED should not be mistaken for active CNV and systematically treated with anti-VEGF. Nevertheless, in some cases with worsened vision not responding to usual CSC therapy, anti-VEGF could be considered as a therapeutic test to rule out the presence of secondary CNV.
Retinal pigment epithelium (RPE) detachment of variable shape and size is a well-known feature of central serous chorioretinopathy (CSC). The routine use of spectral-domain optical coherence tomography (SD OCT) has shown that a pigment epithelium detachment (PED), regardless of its shape, is present in 70%-100% of cases. The shape of the PED is variable, ranging from the typical dome-shaped PED in acute CSC to the flat irregular PED, which is more likely observed in chronic CSC. Various terms have been used to describe these PEDs, including “RPE with an irregular wave” or “low to flat PED” with an irregular or dimpled surface. Similar PED profiles have also been documented in acute CSC and have been called “irregular RPE layer” or “RPE irregularities.” On the other hand, irregular PED is one of the most common OCT findings of type 1 choroidal neovascularization (CNV) in age-related macular degeneration (AMD). Type 1 CNV can occur, although rarely, in eyes with long-term CSC. We have quite frequently observed flat irregular PEDs in cases of chronic CSC with no evolution toward obvious CNV. We therefore assume that some of these flat irregular PEDs could be avascular. The aim of this retrospective study was to try to distinguish cases with avascular flat irregular PEDs from those in which PEDs were consistent with type 1 CNV.
Patients and Methods
We retrospectively reviewed the medical records of patients with chronic CSC who were examined in the Department of Ophthalmology of Lariboisière Hospital between June 1, 2007 and May 31, 2013. The Ethics Committee of the French Society of Ophthalmology approved this retrospective review of patient records (IRB #00008855). The research followed the tenets of the Declaration of Helsinki.
Patients were selected using the following diagnostic filters on the OCT database: (1) chronic CSC, (2) atypical CSC, or (3) diffuse retinal pigment epitheliopathy (DRPE). Only patients who had performed at least 1 fluorescein angiography (FA) and indocyanine green angiography (ICGA) during the follow-up were selected. According to the broad definition of chronic CSC, in all included patients, the disease onset occurred at least 6 months prior to the first considered examination, and they all had either persistent or recurrent serous retinal detachment (SRD) more than 6 months after the onset.
Color, blue reflectance, and red-free filter fundus photographs were analyzed to identify any indirect sign of CNV, such as retinal or subretinal hemorrhage and hard exudates, and to rule out the presence of macular drusen. According to the routine protocol used in our department, OCT examination was performed in both eyes of each patient using the SD OCT Spectralis (Heidelberg Engineering, Heidelberg, Germany) as follows: (1) 9-mm horizontal and vertical SD OCT scans were centered on the fovea; (2) a raster of 7 horizontal enhanced depth imaging (EDI) SD OCT lines within a 5 × 20-degree rectangle centered on the fovea was performed, each line being composed of 100 averaged scans; (3) a macular map was generated including 25 horizontal SD OCT lines, with a minimum of 9 averaged scans for each line, in a 30 × 30-degree square centered on the fovea. Tighter raster lines were performed on PEDs or RPE irregularities and, when the examination included an FA or ICGA, at the dye leakage, if any. Macular thickness map and horizontal raster scans were independently studied by 2 readers (R.H., A.G.). To establish the diagnosis, all B-scans of the macular map and all raster scans were reviewed. When patients had a follow-up longer than 5 years, OCT examinations performed on Cirrus SD OCT (Carl Zeiss Meditec, Dublin, California, USA) and on Stratus Time-Domain OCT (Carl Zeiss Meditec) were also reviewed.
The RPE profile was then classified into 2 different types: classical round PED and flat irregular PED (ie, PED with dimpled or irregular surface on OCT B-scan). In the flat irregular PED type, the reflectivity of the sub-RPE space was also characterized as completely, partially, or not “optically filled.” The space under the RPE was considered as “not optically filled” when its reflectivity was similar to that of the vitreous cavity.
All numeric data were recorded from the Spectralis OCT. We recorded the central macular thickness (CMT) calculated by the software as the mean value of the central subfield, with a diameter of 1 mm. We measured the foveal thickness using calipers from the inner portion of the hyperreflective line corresponding to the RPE to the inner surface of the retina. The subfoveal choroidal thickness was also measured manually, using calipers, on central horizontal raster EDI scans, from the outer portion of the RPE line to the inner surface of the sclera. The subfoveal choroidal area on the same scan was assessed by drawing a line around the edges of the choroid. Within the choroid, the diameter of the largest choroidal vessel was measured using calipers by 2 different operators (R.H., A.G.).
ICGA was performed in all patients using Heidelberg Spectralis, allowing concomitant OCT scanning. Infracyanine green (Laboratoires SERB, Paris, France) was given intravenously at a dose of 25 mg. The morphology and diameter of choroidal veins were assessed in the early phase of the examination. Hyperfluorescence and leakage were investigated in the areas of irregular RPE, in the early (<5 min), middle (15 min), and late (30 min) phases of the examination.
Finally, the following criteria were retained for avascular flat irregular PEDs in the context of chronic CSC: (1) OCT-documented SRD lasting more than 6 months, (2) multiple RPE impairments within and outside the area of subretinal fluid, (3) distinct or indistinct focal or multifocal leaks on FA within the SRD, (4) presence of dilated choroidal veins and late hyperfluorescent leakage in the intervascular choroidal tissue on ICGA, (5) choroidal thickening on EDI SD OCT (>350 μm), (6) an optically empty space under the elevated RPE, and (7) no sign of obvious CNV, including hemorrhages or hard exudates.
The diagnosis of type 1 CNV was based on multimodal imaging including variable associated features such as: (1) subretinal or sub-RPE hemorrhage, (2) hard exudates, (3) evidence of fibrovascular PED, (4) ill-defined leakage on FA, (5) abnormal vascular branching network on early phases of ICGA, and (6) late-staining plaque at the late phase of ICGA. Polypoidal vasculopathy could be associated with type 1 CNV when hyperfluorescent aneurysmal lesions appeared during ICGA dye transit followed by a washout during the late phase.
Owing to the CSC-specific context, neither SRD nor cystoid macular edema was selected as a criterion for the presence of CNV. The PED characteristics on OCT were also discussed.
Data were collected and analyzed using Stata 1998. A P value <.05 was considered as statistically significant. Mann-Whitney U test was used for comparison of unpaired groups. The χ 2 test was used to compare qualitative parameters of each group.
One hundred and thirteen patients were identified with chronic CSC, atypical CSC, or DRPE in our OCT database. Three patients were excluded because of an intercurrent ophthalmic disease (diabetic retinopathy and/or diabetic macular edema). The remaining 110 patients had chronic CSC in 1 or both eyes. In total, 172 eyes were affected, including 53 eyes of 38 patients with flat irregular PED.
Patient characteristics for these 38 patients are summarized in Table 1 . Twenty-eight patients (73.6%) were male. Fifteen patients (39.4%) had bilateral flat irregular PED. Patients’ mean age was 58.6 ± 13.2 years (range: 18–76 years) at the time of the study, with only 1 case of a patient for whom the onset was before the age of 34. The mean follow-up duration was 14.6 years (range: 2–39). Owing to the long-term course of chronic CSC, 7 patients were not able to date precisely their disease onset. The mean age at onset (48.8 ± 15 years; range: 13–72) was therefore calculated using the data available for the remaining 31 patients.
|Number of patients (eyes)||38 (53)|
|Age (range), y||58.6 (18–76)|
|Age at onset, y (calculated with data of 32 patients)||48.8 (13–72)|
|Associated conditions (number of patients)|
|Systemic inflammatory disorder||5|
|Diabetes without retinopathy||2|
|History of corticotherapy, number (%) of patients||11 (28.9%)|
|CSC treatment used (number of eyes)|
|Central macular thickness (range), μm||310 (111–816)|
|Foveal thickness (range), μm||125.5 (40–544)|
|Macular volume (range), μm 3||8.5 (4.9–16.6)|
|Choroidal thickness (range), μm||458.5 (135–817)|
|Subfoveal choroidal area (range), μm 2||3.05 (0.96–5.17)|
Patients received various treatments during the course of their disease. Twenty-eight eyes underwent half-fluence photodynamic therapy (PDT), 18 eyes underwent laser photocoagulation, 14 eyes received at least 1 anti-VEGF intravitreal injection, and 7 patients received eplerenone 50 mg daily for several weeks. A single eye could have undergone repeated and/or several treatments among those above mentioned.
Serous Retinal Detachment and Choroidal Thickening
All 53 eyes with flat irregular PED had either an ongoing SRD or an OCT-documented SRD, which occurred at any time during the follow-up and always more than 6 months after the onset. In 45 eyes, the macula was affected by the SRD. In the remaining 8 eyes, SRDs spared the macula and were found because the fellow eye was symptomatic. Five eyes had yellow subretinal deposits within the SRD. Areas of hyper- or hypo-autofluorescence were seen within the SRD in all cases. Thirty-four eyes had some degree of hypo-autofluorescence in the macular area, which was suggestive of more or less significant RPE atrophy. Eight eyes showed hypo- or hyperfluorescent gravitating tracts.
The mean CMT was 310 μm (range: 111–816). The mean retinal thickness at the fovea was 125.5 μm (range: 40–544). The mean macular volume was 8.5 μm 3 (range: 4.9–15.6). The mean choroidal thickness was 458.5 μm (range: 135–817). Only 4 eyes had choroidal thickness less than 300 μm and each of them showed other evidence of CSC, including clinical course (spontaneous resolution of SRD), characteristic FA findings, and/or nonresponse to anti-VEGF therapy, making another diagnosis, such as AMD, unlikely.
Eighteen eyes showed a significant enlargement (>350 μm) of choroidal vessels.
Flat Irregular Pigment Epithelium Detachments
Flat irregular PEDs were diagnosed by analyzing multiple OCT scans. Flat irregular PEDs were located in the foveal area in 41 of 53 eyes and were extrafoveal in 13 of 53 eyes, including 2 eyes in which flat irregular PEDs were located in the vicinity of the temporal vessels. Three eyes had flat irregular PEDs in multiple extramacular locations and 1 eye showed flat irregular PEDs in both macular and extramacular locations.
Type 1 Choroidal Neovascularization
In 10 eyes, flat irregular PEDs were suggestive of type 1 CNV, although no other signs of AMD—especially no drusen—were present, as already shown by Fung and associates. The diagnosis of type 1 CNV was suspected because of the presence of subretinal or retinal hemorrhages and/or hard exudates (5 eyes) or a choroidal neovascular network visible on the early phases of scanning laser ophthalmoscopy (SLO) ICGA (6 eyes). In all cases, flat irregular PEDs had a hyperreflective content compatible, although not exclusive, with the presence of type 1 CNV. One eye had associated features typical of polypoid choroidal vasculopathy (PCV). Seven eyes had been previously treated with PDT and only 4 had a history of laser photocoagulation. Finally, 8 out of the 10 cases responded well to anti-VEGF treatment. In the remaining 2 eyes, a severe vision loss was present and anti-VEGF treatment had not been prescribed. Fundus, OCT, and angiography findings of these 10 patients are presented in Table 2 and compared with the remaining 43 patients with flat irregular PED.
|A. Fundus||No CNV (43)||CNV (10)||χ²|
|Retinal atrophy||27||7||0.18 ( P = .66)|
|Hemorrhage||0||2||12.7 ( P = .002) a|
|Exudates||0||3||16.9 ( P < .0001) a|
|Mirocystic retinal edema||1||6||21.2 ( P < .001) a|
|Choroidal thickness (μm)||461.5||405.5||Z = –0.68 ( P = .49) (Mann-Whitney U test)|
|Sub-retinal deposit||17||2||0.58 ( P = .44)|
|Sub-RPE material (filled sub-RPE space)||26||10||5.82 ( P = .016) a|
|Neovascular choroidal network||1||6||23.5 ( P < .001) a|
|Late hyperfluorescence||26||8||1.34 ( P = .25)|
|Smokestack or punctiform leakage||17||4||0.0007 ( P = .97)|
|B. Flat Irregular RPE Area|
|Fovea (inside the 1 mm ETDRS area)||31||10||3.6 ( P = .057)|
|Extra-foveal (outside the 1 mm ETDRS area)||13||0||4 ( P = .045) a|
Presumed Avascular Flat Irregular Pigment Epithelium Detachments
Avascular flat irregular PEDs were characterized by the absence of angiographic evidence of vascular network on FA or ICGA, the absence of hard exudates or hemorrhages, and very few changes during the follow-up period. Forty-three cases were diagnosed with avascular flat irregular PED. Nineteen had been followed for less than 1 year while the remaining 24 had been followed for 1–15 years. No obvious evolution toward CNV was observed in any case. The maximum flat irregular PED height was 255 μm. Its content was hyporeflective in 18 cases, hyperreflective in 23 cases, and mixed in 2 cases. The hyperreflectivity corresponded to yellow or pigmented subepithelial deposits or fibrosis in 17 eyes. The SRD attributed to the chronic CSC waxed and waned without any change in the flat irregular PED morphology. Twenty eyes had previously been treated with half-fluence PDT and 14 had a history of laser photocoagulation.
Anti–vascular Endothelial Growth Factor Treatment
Fourteen eyes had received anti-VEGF treatment, including 8 of the 10 eyes with CNV (the remaining 2 eyes were not treated because of too-poor vision). In these 8 eyes, a significant improvement was observed in both SRD and visual acuity (VA). Table 3 presents the features of anti-VEGF-treated CNV eyes.
|Eye||Duration of Anti VEGF Treatment (Number of Injections)||SRD During Treatment||Status of Anti-VEGF Treatment and Reason for Discontinuation|
|1||6 years (>20)||Resolved||Ongoing|
|2||4 months (3)||Improved||Discontinued; no improvement in VA|
|3||4 months (3)||Resolved||Discontinued; no SRD recurrence|
|4||2 years (9)||Resolved||Ongoing|
|5||2.5 years (8)||Resolved||Discontinued; no SRD recurrence|
|6||4 months (3)||Improved||Discontinued; no improvement in VA|
|7||2 years (>20)||Resolved||Ongoing|
|8||2 years (>20)||Resolved||Ongoing|
However, at some point during the course of the disease, because of the chronicity of the SRD and a slow decrease in vision, 4 other cases were treated with anti-VEGF by the treating ophthalmologist although the diagnosis of CNV was questionable. Neither SRD nor vision improved and anti-VEGF was then discontinued after 3–9 intravitreal injections.
A 52-year-old man presented in 1995 with moderately decreased vision in the left eye (OS). The VA was 20/20 in the right eye (OD) and 20/30 OS. Fundus examination showed some small pigmentary changes OS. There was a shallow SRD in the macula and FA showed an ill-defined area of moderate leakage in the macular area. An ICGA was performed and revealed dilated choroidal veins and bilateral late staining of choroidal plaques ( Figure 1 ). No treatment was administered. The OS visual acuity worsened in 1998. FA showed a small leakage close to an already known pigment clumping. An OCT was performed (OCT 1; Carl Zeiss Meditec, Dublin, California, USA) that showed a shallow foveal detachment. Subretinal fluid waxed and waned over the next 8 years. In 2002, an OCT (Stratus; Carl Zeiss Meditec) showed the persistence of subretinal fluid and a very shallow and slightly irregular PED, which remained stable until 2006 ( Figure 1 ). The VA stabilized at 20/30. In 2007, the VA decreased to 20/60 and ICGA revealed a remarkable choroidal neovascular network in the macula, which corresponded to a flat irregular PED suggestive of type 1 CNV on SD OCT scans ( Figure 2 ). The subfoveal choroid was thickened at 336 μm in 2008. The CNV is partially controlled with repeated intravitreal injections of anti-VEGF since then.
A 45-year-old man had experienced decreased vision OD for 5 years and was diagnosed with chronic CSC. The VA was 20/50. Fundus examination showed some macular pigmentary changes and a shallow exudative retinal detachment in the lower part of the macula. FA showed an ill-defined area of hyperfluorescence, which was stable until the late phases with some pinpoints. ICGA showed some dilated choroidal veins crossing the macular area, late pinpoints, and hypofluorescence of the center of the lesion surrounded by some choroidal staining on the late phases. OCT showed an area of subretinal fluid in the lower part of the macula, combined with a flat irregular PED. The subfoveal choroid was thickened at 336 μm. Over a 4-year follow-up, the amount of subretinal fluid varied. The flat irregular PED slightly enlarged but the VA remained stable at 20/30 ( Figure 3 ).