Mid-Phase Hyperfluorescent Plaques Seen on Indocyanine Green Angiography in Patients with Central Serous Chorioretinopathy

(1) Indocyanine green angiography (ICG-A) shows the presence of mid-phase hyperfluorescent area in central serous chorioretinopathy (CSCR). However, their exact meaning remains uncertain. (2) The clinical and multimodal imaging findings of 100 patients (133 eyes) with CSCR, including the enhanced-depth-imaging OCT (EDI-OCT), blue-light fundus autofluorescence (BAF), fluorescein and indocyanine green angiography (FA and ICG-A) findings were reviewed. Mid-phase hyperfluorescent plaques (MPHP) were defined as fairly well circumscribed hyperfluorescent regions during the midphase of the ICG-A. The association between MPHP and other clinical/imaging parameters was assessed using a multiple logistic regression analysis. (3) MPHP were detected in 59.4% of eyes with CSCR. The chronic form of the disease, the presence of irregular pigment epithelium detachments (PED) and the retinal pigment epithelium (RPE) changes seen on FA were associated with the presence of MPHP in the multivariate analysis (p = 0.015; p = 0.018 and p = 0.002; respectively). OCT showed RPE bulges or PED in 98.7% of areas with MPHP and BAF showed changes in 57.3% of areas with MPHP. (4) MPHP were associated with a chronic form of CSCR and colocated with PED or RPE bulges. MPHP should be recognized as a sign of early RPE dysfunction before it is detected with BAF.


Introduction
Central serous chorioretinopathy (CSCR) is a chorioretinal disease characterized by the presence of serous retinal detachments (SRD) associated with retinal pigment epithelium (RPE) detachments (PED) and an increased choroidal thickness, occurring more commonly in middle-aged men [1]. One of the characteristics of CSCR is its frequent association with dilated choroidal veins in the Haller layer beneath a thinning of the choriocapillaris and the Sattler's layer [2], likely to suggest a vein overload.
To date, several risk factors for CSCR have been identified. The most consistent risk factor is an exposure to glucocorticoids [3]. Several other predisposing factors or contributing factors have been identified as: psychological stress, anxiety and maladaptive personality [4,5], hypertension and cardiovascular disease, sleep disturbance and shift work [6], allergic disorders, pregnancy, peptic ulcer disease and also genetic risk factors [7].
Indocyanine green angiography (ICG-A) remains the gold standard imaging method to identify the dynamic choroidal vascular changes occurring in CSCR. Early-phase ICG-A shows a delayed choroidal filling and a dilation of choroidal veins, while geographic areas of hyperfluorescence are visible in the mid-phases and fade away in the late phases [8] (Supplementary Figure S1) [8,9]. They are not necessarily associated with observable pathologic manifestations, remain stable over years, independent from subretinal fluid and have not been observed only in the affected but also in the fellow unaffected eye of CSCR patients [8,10]. While it is widely accepted that hyperfluorescent areas during ICG-A result from dye leakage from leaky vessels, such as dilated veins or the choriocapillaris, this hyperpermeability has more than one clinically evident manifestation. Simple hyperpermeability may lead to increased dye transudation into the choroid. These areas have poorly defined outer borders, expand in size during the course of the angiogram, and produce faint hyperfluorescent regions that silhouette the larger choroidal vessels in the later phases of the angiogram [8]. But, there are other eyes that show more well-defined areas in which the underlying choroidal details are obscured. The regions appear few minutes after ICG injection and increase in intensity until 7 to 10 min and do not show expansion during the course of the angiogram. They have well defined borders and disappear at the very late phase of the angiogram (>20 min). Interestingly, the angiographic dynamics of these mid-phase hyperfluorescent plaques (MPHP) parallels the fate of ICG from the choroid into the RPE observed in post mortem eyes from monkeys that were enucleated at several intervals after ICG was injected intravenously [11]. At 15 min after injection, ICG was mostly found in the RPE cells and in the Bruch membrane, suggestive of an active transport of ICG from the choroidal endothelial cells to the RPE, which was further confirmed by in vitro study showing that RPE cells internalize ICG through active mechanisms [12].
In this context, the aim of this study was to investigate the multimodal imaging findings and clinical parameters associated with CSCR eyes and to evaluate the multimodal imaging findings of MPHP using enhanced-depth-imaging OCT (EDI-OCT), blue-light fundus autofluorescence (BAF), fluorescein angiography (FA) and ICG-A. The main conclusion of the study is that MPHP could be the clinical manifestation of RPE dysfunction and not only of vascular permeability.

Ethics Statement
The study was approved by the Ethics Committee of the French Society of Ophthalmology (IRB 00008855 Société Française d'Ophtalmologie IRB#1). The study adhered to the tenets of the Declaration of Helsinki (1964). Signed informed consent was obtained from all subjects.

Study Design
This was a retrospective study conducted in the departments of ophthalmology of Hotel Dieu and Cochin Hospitals, Paris, between 2012 and 2020.

Study Patients
The medical records and imaging findings of consecutive patients with central serous chorioretinopathy (CSCR) imaged with indocyanine green angiography (ICG-A) were retrospectively reviewed. CSCR was diagnosed by two retinal experts (EB, FBC). Exclusion criteria were patients with drusen within 30 degrees, dome-shaped macula, vitreomacular traction syndrome, any history of uveitis, other retinal diseases, or poor image quality.

Study Protocol
The medical data collected included age, gender, history of corticosteroid intake, shift work, previous CSCR treatments and the best-corrected visual acuity (BCVA) converted into LogMAR scale. The clinical form of the disease was also recorded. Chronic CSCR was defined by the presence of a persistent serous retinal detachment for at least 6 months. In other cases, CSCR was classified as acute/recurrent.

Image Analyses
The choroidal thickness was manually measured on the EDI horizontal B-scans passing through the fovea as previously described [13].
The presence of pigment epithelium detachment (PED) was assessed on the SD-OCT B scan. A bulge of the RPE was defined as a protrusion of the RPE without the visualization of the Bruch's membrane as previously described [14]. The PED was defined by an elevation of the RPE with a distinct visualization of the Bruch membrane. The form of PED was also analysed and classified as dome-shaped or irregular PED defined by an irregular elevation of the RPE [15].
The presence of hyper or hypo autofluorescent area and gravitational tracks was assessed on BAF. The presence of CSCR focal leak with a smokestack and ink-blot pattern was assessed on FA. The presence of dilated choroidal veins (pachyvessels) was assessed on early-phase ICG-A, and the presence of mid-phase hyperfluorescent plaques (MPHP) on ICG-A was recorded.
All imaging analyzes were performed by two trained retinal specialists (JP and MZ). Disagreements were resolved by a third retinal specialist (FBC).

Statistical Analysis
Statistical analyzes were performed using XLstat software (version 2020; Addinsoft Paris, France), and R version 3.6.3 (The R Foundation for Statistical Computing). Descriptive data are presented as the mean ± standard deviation for quantitative variables and as counts and percentages for categorical variables. A univariate logistic regression was performed to assess the association between the outcome variable MPHP and the independent variables. Correlations between both eyes were eliminated using the marginal generalized estimating equations (GEE) model. Factors showing significant associations in the univariate analysis (p < 0.10) were included in the multivariate regression model using a stepwise backward variable elimination. Adjusted odds ratios and their 95% confidence intervals were calculated for the factors included in the multivariate logistic regression models. All p values were 2-sided and p values ≤ 0.05 were considered statistically significant.

Clinical and Imaging Findings of Patients with and without Mid-Phase Hyperfluorescent Plaques
The comparison of the characteristics of patients with and without MPHP is summarized in Table 2. The rate of MPHP was significantly higher in eyes with a chronic form of the disease (66 eyes (83.5%) versus 17 eyes (31.5%); p < 0.001). In addition, the rate of irregular PED was higher in eyes with MPHP (40 eyes (51%) versus 10 eyes (18.5%); p < 0.001). On BAF, more eyes with MPHP had BAF abnormalities and gravitational tracks (p < 0.001 and p = 0.001, respectively). On early-phase ICG-A, a foveal dilated vein was more frequently visualized in eyes with MPHP (p = 0.001).

Factors Associated with the Presence of Mid-Phase Hyperfluorescent Plaques
In the multiple logistic and LASSO regression analysis, a chronic form of the disease, the presence of irregular PED, the RPE changes seen on FA and a foveal dilated vein seen on ICG-A were associated with MPHP in CSCR patients (p = 0.015; p = 0.018; p = 0.002 and p = 0.036; respectively Table 3).

Clinical and Imaging Findings of Patients with and without Mid-Phase Hyperfluorescent Plaques
The comparison of the characteristics of patients with and without MPHP is summarized in Table 2. The rate of MPHP was significantly higher in eyes with a chronic form of the disease (66 eyes (83.5%) versus 17 eyes (31.5%); p < 0.001). In addition, the rate of irregular PED was higher in eyes with MPHP (40 eyes (51%) versus 10 eyes (18.5%); p < 0.001). On BAF, more eyes with MPHP had BAF abnormalities and gravitational tracks (p < 0.001 and p = 0.001, respectively). On early-phase ICG-A, a foveal dilated vein was more frequently visualized in eyes with MPHP (p = 0.001).

Multimodal Imaging Analysis of the Areas with Mid-Phase Hyperfluorescent Plaques
A total of 249 MPHP were detected in 79 eyes and analyzed by multimodal imaging (Table 4). On early phase ICG-A, available for 132 MPHP, a delayed choroidal filling was detected in 77.3% of areas with MPHP. Dilated choroidal veins were observed in 42.8% of areas with MPHP (data available for 243 MPHP). On FA, a focal leak was detected in 25 eyes with MPHP and it was located in areas with MPHP in 20 eyes (80%). BAF abnormalities were detected in the area of MPHP in 57.3% of the cases (data available for 141 MPHP) (Figures 1 and 2). For 152 MPHP, the SD-OCT B-scans passing through the plaque were available. RPE bulges or PED were detected in 98.7% of cases while pachyvessels were observed in 57.2% of areas with MPHP ( Figure 2).

Evolution of MPHP
During a mean follow-up of 1.9 ± 1.4 years, another ICGA was available in 43 eyes. We found a stability of the number and size of MPHP in 33 eyes (76.7%). In 8 eyes (18.6%), we found a resolution of MPHP. In these eyes, RPE became atrophic, as shown in Supplementary Figure S2. In 2 eyes, we observed the emergence of new MPHP.

Evolution of MPHP
During a mean follow-up of 1.9 ± 1.4 years, another ICGA was available in 43 eyes. We found a stability of the number and size of MPHP in 33 eyes (76.7%). In 8 eyes (18.6%), we found a resolution of MPHP. In these eyes, RPE became atrophic, as shown in Supplementary Figure S2. In 2 eyes, we observed the emergence of new MPHP.

Discussion
From 1965 onwards, a primitive RPE dysfunction has been suggested in the pathogenesis of CSCR, supported by the presence of the focal leaks from the RPE [1]. An involvement of the choroidal vasculature has been identified on ICG-A, that showed delayed choroidal arterial filling, choroidal venous dilatation and focal hyperfluorescent area [1,8,10,16]. According to Hayashi et al., as the delayed filling was associated with RPE degeneration, circulatory insufficiency has been proposed to cause the epitheliopathy [17]. The areas of ICG-A hyperfluorescence, called MPHP, seen on mid-phase ICG-A [18] at a time when the surrounding fluorescence decreases have been attributed to choroidal vessels permeability. Such regions of hyperfluorescence are reported in more than 90% of CSCR cases in most studies [10,19,20] and a risk site for fluorescein leakage and SRD [10,16,18]. In our cohort of CSCR patients, 62.4% of patients had a chronic form of the disease, and MPHP were found in almost 60% of eyes, suggesting that they are different from previously described area of hyperfluorescence and that they may characterize a specific type of patients.
In this study, MPHP were associated with the chronicity and with multiple signs suggestive of a chronic form of the disease such as irregular PED [15,21] and a subfoveal choroidal dilated vein potentially suggesting abnormal venous drainage and anastomosis [22,23]. In addition, when a focal leak was present, it was found within the plaque in 80% of the cases, in line with previous studies and suggesting that the RPE barrier could be weaker in these area.
Since MPHP were also observed in contralateral unaffected eyes, they could be a prognosis factor rather than a consequence of the chronicity, but further prospective studies are needed to confirm this assumption. Indeed, In our study, we found MPHP in 35.3% contralateral unaffected eyes and we observed the occurrence of subretinal detachment in one contralateral eye. However, a longer follow-up is mandatory to evaluate more accurately the incidence of subretinal detachment according to the presence of MPHP.
Interestingly, in areas with MPHP, pachyvessels were seen on early ICG-A in 42.8% of cases and on SD-OCT in 57.2% of cases while PED or RPE bulges were found in 98.7% of areas with MPHP. In other words, areas with MPHP were more frequently associated with abnormalities of the RPE layer on SD-OCT than with the presence of dilated veins. There was also a significant association between the areas of delayed ICG filling and the areas of MPHP, which could lead to hypoxic suffering of the RPE in these specific areas. Altogether, these observations indicate that MPHP occurred in areas of diseased RPE, even if 43% of areas with MPHP had a normal aspect on BAF, suggesting an "early" RPE dysfunction.
The mechanisms of increased fluorescence of ICG in the choroid and RPE after systemic injections are multiple. Since more than 95% of the systemically injected ICG binds to lipoproteins, ICG is assume to follow the transport mechanisms of these molecules in the choroid/RPE/retina complex [24]. The choriocapillaris is poorly permeable to circulating endogenous proteins despite its fenestration [25,26] that are highly VEGF-dependent [27]. In CSCR, no increase in VEGF levels have been measured [28,29] suggesting that increased passage through fenestrations is unlikely. Other mechanisms related to active transports rather than passive diffusion could be deregulated. The delayed visualization of ICG in MPHP (few minutes after ICG injection) correlates well with the active transports from vascular endothelial cells to the RPE cells, related to the Na+/K+ ATPase activity [12] or to caveolins [30], that have been shown to control the transcellular movement of albumin and lipoproteins from the vessels towards the RPE and photoreceptor cells [30]. Excessive passage of proteins in the outer retina through RPE would increase the interstitial pressure and explain the smokestack appearance of fluorescein leaking through the RPE following an oncotic gradient. Recently, Sakurada et al. [31] have reported an association between the presence of choroidal caverns and the presence of hyperfluorescent area on mid-phase ICG-A. Since caverns could correspond to lipoproteins and lipids, it could be assumed that ICG could also accumulate in these deposits [31], also reflecting abnormal transports between the choroid and the RPE since lipoproteins physiologically transfer beta carotene or xanthophylls from the choroid to the retina [32].
We assumed that MPHP could result from an excessive accumulation of ICG (and other macromolecules such as proteins and lipoproteins) either within the abnormal RPE, or in sub RPE deposits ( Figure 3) and could thus be an early indicator of epitheliopathy, visible before any change is observed on BAF as autofluorescent changes were observed in only 57% of areas with MPHP. This finding is in line with cases in which MPHP tend to disappear when RPE hypoautofluorescence progresses after years of evolution, suggesting mainly a RPE loss (Supplementary Figure S2). We assumed that in CSCR, MHPH could indicate an over transcellular transport of proteins or lipoproteins in diseased RPE.

Conclusions
CSCR is a multifactorial disease with a complex pathogenesis and we propose to add an altered macromolecular transport from choroidal endothelial cells to the RPE/Bruch's membrane, shown by the presence of MPHP on ICG-A, to the multiple pathogenic mechanisms. The presence of MPHP may indicate an "hyperporosity" of the RPE and/or Bruch's membrane and be an early sign of epitheliopathy. Prospective longitudinal studies are needed to explore the prognosis value of MPHP in diseased and contralateral eyes The limitations of this study are its retrospective design and the fact that corresponding SD-OCT sections were not acquired for all MPHP areas. In addition, wide-field ICG-A was not performed and it could have provided a dynamic and more complete analysis of the choroidal circulation.

Conclusions
CSCR is a multifactorial disease with a complex pathogenesis and we propose to add an altered macromolecular transport from choroidal endothelial cells to the RPE/Bruch's membrane, shown by the presence of MPHP on ICG-A, to the multiple pathogenic mechanisms. The presence of MPHP may indicate an "hyperporosity" of the RPE and/or Bruch's membrane and be an early sign of epitheliopathy. Prospective longitudinal studies are needed to explore the prognosis value of MPHP in diseased and contralateral eyes of CSCR patients.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/jcm10194525/s1, Figure S1: A-C: Indocyanine angiography (ICG-A) and OCT-angiography (OCT-A) of a patient with central serous chorioretinopathy (CSCR). A-B: ICG-A shows mid-phase hyperfluorescent plaques (MPHP) that become isofluorescent during late-phase associated with hypofluorescent spots in around two-third of plaques. C: OCT-A does not show macular neovascularization (MNV). D-F: ICG-A and OCT-A of a CSCR patient associated with macular neovascularization (MNV). D-E: ICG-A shows a macular hyperfluorescent area (arrow, D) during mid-phase that increases in size and intensity during late phase (arrow, E). F: OCT-A shows macular neovascularization (arrow). Figure   Informed Consent Statement: Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patients to publish this paper. Data Availability Statement: Data available on request from the corresponding author. The data are not public available due to privacy and ethical.