Purpose To identify optical coherence tomography (OCT) findings associated with the leakage points in patients with central serous chorioretinopathy (CSC) to provide fluorescein angiography (FA)-free focal laser photocoagulation (FLP) of the leakage point.
Methods A retrospective study included 48 patients with CSC (48 eyes). Colocalisation of leakage points with pigment epithelial detachments (PEDs) and with areas of photoreceptor outer segments (PROS) layer thinning was evaluated with OCT. Using FA for each leakage point, the relationship to neurosensory detachment was evaluated with retro-mode confocal scanning laser ophthalmoscopy.
Results Coincidence with PED was found in 52 of 65 (80.0%) leakage points. The PROS thinning was found in 47 of 52 (90.4%) of the PEDs coincided with leakage point. The mean distance from the upper border of neurosensory detachment to the leakage point was 27.3%±13.0% of the vertical dimension of the neurosensory detachment.
Conclusion This study demonstrates that PEDs localised in the upper half of the neurosensory detachment area and associated with the PROS thinning area coincided with the leakage point in a significant number of patients with CSC. The patients with non-resolving CSC with a small single PED localising in the upper one-third to one-half of the neurosensory detachment area with an area of PROS thinning above this PED may undergo FA-free OCT-guided FLP treating whole PED.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
For decades, fluorescein angiography (FA) was the most important method to reveal central serous chorioretinopathy (CSC) as it was the only technique to clearly identify both central serous neurosensory detachment and leakage point, two key findings in CSC.1
At present, the sum of optical coherence tomography (OCT) findings allows confirming CSC without intravenous angiography in most cases.2 3 Nevertheless, identification of the leakage without FA or indocyanine green angiography is still considered impossible. At the same time, identification of the leakage point is a critical step for focal laser photocoagulation (FLP).
FLP of the leakage point is a relatively easy and effective treatment option for acute CSC4 if the duration of the disease exceeds 3 months without spontaneously resolving. FA-free identification of the leakage point could significantly reduce the need for FA in patients with CSC, and moreover, at least some patients with CSC could be managed without FA completely (especially in whom FA may be contraindicated). It also would increase the safety of CSC management since FA is an invasive procedure, which potentially associated with a variety of allergic reactions and more frequent mild adverse effects (ie, nausea and hives).
It is known that the leakage point is typically associated with changes of retinal pigment epithelium (RPE) (mostly with pigment epithelial detachment (PED)).3 5 6 However, the PED alone is not sufficient to indicate the leakage point. At the same time, the changes in outer retina were also noted in patients with CSC, but the relation of those (except for the outer retinal dipping) with leakage point was not studied.7 8 A number of studies described thickening, flaking, defects and hyper-reflective foci in photoreceptor outer segments (PROS) layer, but without analysing their relation to leakage point.8 9 We supposed that morphological changes in outer retina (first of all, in PROS layer) if associated with leakage could provide additional information to identify leakage point. In addition, the spatial distribution of leakage points has been reported to be non-random10, so could help facilitate identification of leakage point.
Therefore, the present work aims to identify and verify the leakage point using OCT findings (namely, PED and coexisting changes of PROS) given cumulative spatial distribution of leakage points to provide an FA-free FLP. In this study, we also estimated the proportion and characteristics of the patients with CSC who were suitable for FA-free FLP.
The study followed the ethical standards stated in the Declaration of Helsinki. Written informed consent was obtained from all patients after they were informed about the study procedures.
In this study, we retrospectively reviewed 48 eyes of 48 symptomatic patients with CSC with the presence of FA-confirmed leakage points. All patients were imaged using retro-mode scanning laser ophthalmoscopy (RM-SLO) and spectral domain OCT. After a standard ophthalmic examination, all patients underwent evaluation with OCT, RM-SLO and FA. Exclusion criteria were as follows: any previous treatment for CSC (including mineralocorticoid receptor antagonist, FLP, micropulse laser therapy (MLT) and photodynamic therapy (PDT) ; symptoms lasting less than 3 months or more than 12 months ; and signs of any other active retinal disease in the study eye (including choroidal neovascularis ation). In cases of bilateral CSC, only one eye was included in the study.
Spectral domain OCT
The spectral domain OCT system used was the RTVue-100 (Optovue, Fremont, California, USA). In addition to enhanced-depth retinal cross-sectional images (16 384 A-scans, 6 mm) centred on the fovea, three-dimensional macular scans (141 B-scans each consisting of 385 A-scans, 7 mm) were obtained.
Using OCT scans, (1) the height of the serous neurosensory detachment, (2) the number of PEDs, (3) the presence of PROS layer thinning above the PED coinciding with leakage point, (4) the greatest diameter of the PROS thinning area and (5) the OCT findings in the outer retinal layers and in the subretinal fluid at the leakage point were evaluated in each case (table 1).
The PROS thickness was measured on cross-sectional scans using built-in calliper tool of the OCT system software by single observer (DSM). The PED coinciding with the leaking point with overlying PROS thinning was identified on cross-sectional scans by comparing (1) an exported en face image superimposed on FA and (2) an en face image with associated cross-sectional scan image in OCT software. The PROS thickness was measured above the PED coinciding with a leakage point, and the mean of two measurements (figure 1A) was compared with the mean PROS thickness outside the PED. The PROS thickness outside the PED was calculated as the mean of four measurements in the area of the retina outside the PED where PROS thinning was visually undetectable (figure 1B). PROS thinning above the PED was considered to be present if the difference between the PROS thickness above the PED coincided with a leakage point and the PROS thickness outside the PED was more than 20 µm (more than 2 SD of the PROS thickness outside the PED).
The greatest diameter of the PROS thinning area was measured using a retinal cross-section image selected after the complete evaluation of the three-dimensional macular scan (figure 1C).
Confocal scanning laser ophthalmoscopy and FA
RM-SLO and FA were performed using confocal scanning laser ophthalmoscope F-10 (Nidek, Gamagori, Japan). Only images with a full zone of neurosensory detachment and optic nerve head visualisation were selected for processing. Since F-10 allows to visualise the area 40°×60° (including a full area of neurosensory detachment and optic nerve in each case), only one image from each modality was needed for superimposition, and no changes in the image size were produced during superimposition of those images. Evaluation of the neurosensory detachment included (1) the shape (round (having equal diametric measurements) or irregular (having not equal diametric measurements)), (2) the vertical dimension and (3) the area in categories (small: less than 2 optic disc diameters; medium: from 2–4 optic disc diameters; large: more than 4 optic disc diameters).
All patients underwent a standard FA with an intravenous injection of 5 mL of a 10% sodium fluorescein solution (Novartis International AG, Basel, Switzerland).
To display the leakage point in relation to the area of the neurosensory detachment and the PED, RM-SLO, OCT and FA images were superimposed in the image processor Adobe Photoshop CS2 (Adobe Systems, San Jose, California, USA) by aligning the retinal blood vessels. Evaluation of the leakage points was done using ImageJ software (NIH, Bethesda, Maryland, USA). Each leakage point was evaluated for: (1) the percentage distance from the upper border of the neurosensory detachment to the leakage point (calculated as a ratio of the distance from the upper border of the neurosensory detachment to the leakage point and the greatest vertical diameter of the neurosensory detachment; figure 1D,E), (2) the location of the leakage point relative to the horizontal line centred on the fovea (on the line or above vs below the line; figure 1E) and (3) the spatial overlap between the leakage points and the PEDs (this was analysed after superimposing en face OCT images with FA images; figure 1F,G). If multiple leaks were present in one eye, each leakage point was evaluated independently from the other. The cumulative spatial distribution of the leakage points within the posterior eye pole was evaluated by superimposition of all RM-SLO images (from all included eyes) with the leakage point shown on a single randomly selected RM-SLO image of the patient with CSC by aligning the centre of the fovea and optic nerve head. Images of the contralateral eyes were reflected in the horizontal dimension (figure 2).
The possibility to perform FA-free FLP was evaluated in all patients included in the study. FA-free FLP was considered possible in patients with non-resolved CSC associated with one or two PEDs (<500 µm), if at least one of which was found in the upper half area of the neurosensory detachment and was colocalised with the leakage point and the area of PROS thinning. FA-free FLP was also considered possible in patients with specific changes of the outer retina (outer retina dipping), RPE (ie, microrip of the RPE) or subretinal fluid (hyporeflective subretinal lucency), all of which allows clear identification of leakage points.
Statistica V.10.0 (StatSoft, Tulsa, Oklahoma, USA) was used for statistical analysis. The results are expressed as mean±SD for continuous variables. For statistical analysis, Snellen visual acuity was transformed into a logarithm of the minimal angle of resolution visual acuity. For comparing continuous parametric variables (the location of the leakage point among patients with different sizes and shapes of neurosensory detachment as well as the PROS length within and outside the leakage point), one-way analysis of variance test was used. All P values <0.05 were considered statistically significant.
A total of 48 eyes of 48 patients were included in the study (table 2) and demonstrated 65 PEDs and 56 leakage points.
Out of 48 eyes, 32 (66.7%) had 1 PED. Six (12.5%) had PEDs, 3 (6.3%) had 3 PEDs, 3 (6.3%) had 4 PEDs and 4 (8.3%) had no PEDs. The mean number of PEDs was 1.35±0.96. Out of 48 eyes (91.7%), 44 had 1 leakage point, 1 (2.1%) had 2 leakage points, 2 (4.2%) had 3 leakage points and 1 (2.1%) had 4 leakage points. The mean number of leakage points was 1.17±0.6.
The coincidence of leakage points with PED was found in 44 of 48 eyes (91.7%); of these, 32 eyes (72.7%) had a single PED. Therefore, all eyes with a single PED demonstrated coincidence of leakage point and PED. In the group of 12 eyes (25.0%) that had two or more PEDs, 20 of 33 (60.6%) PEDs coincided with a leakage point and 8 of 12 eyes (66.7%) had a single leakage point.Out of 48 eyes (8.3%), four with non-coincidence of leakage point with PED showed a complete absence of PED. In general, 52 of 65 (80.0%) PEDs coincided with the leakage points.
The majority of leakage points were found at 27.3%±13.0% distance from the upper border of the neurosensory detachment area. An irregular shape of the neurosensory detachment area was associated with a decrease in the percentage distance of the leakage point from the upper border of the neurosensory detachment area, whereas a round shape of the neurosensory detachment area was associated with an increase in the percentage distance of the leakage point from the upper border of the neurosensory detachment area. The percentage distance of the leakage point from the upper border of the neurosensory detachment area in eyes with an irregular neurosensory detachment area versus eyes with a round neurosensory detachment area was 22.9±11.3% versus 39.3±9.6% (P<0.001), respectively (figure 3A). No statistically significant differences were seen in the percentage distance of the leakage point from the upper border of the neurosensory detachment area among eyes with large, medium or small neurosensory detachments (figure 3B). However, eyes with large neurosensory detachment areas demonstrated a lower percentage distance of the leakage point from the upper border of the neurosensory detachment area compared with the eyes with medium or small neurosensory detachment areas.
Interestingly, the majority of the leakage points (51 of 56, 91.1%) were localised in the upper part of the posterior eye pole above or on the horizontal line which crosses the fovea (figure 4). Further, a majority of the leakage points (54 of 56, 96.4%) were localised in the upper half of the neurosensory detachment area.
The PROS length tended to be lower in the area above the PED which coincided with the leakage point (22.5±4.5 µm), compared with the area outside of the leakage point (64.0±9.4 µm) (P<0.001). PROS thinning was found in 47 of 52 (90.4%) of the PEDs coinciding with leakage points and in 40 of 44 eyes (90.1%) that had at least one PED coinciding with a leakage point. PROS thinning was not found in the area of the detached retina above the PED not associated with a leakage point (figure 5). The mean greatest diameter of PROS thinning area was 1174.7±878.5 µm (range, 326–3290 µm).
Specific OCT findings in the outer retinal layers and the subretinal fluid associated with leakage points were found in eight (16.7 %) patients. Dipping of the outer retinal layers was found in six (12.5%) eyes. In one case (2.1%), en face imaging through the subretinal space revealed a smokestack lucency with the surrounding hyper-reflective material corresponding to the subretinal fibrin. Microrip of the RPE in the presence of a fragment of the RPE on the outer surface of the retina was found in one case (2.1%).
Finally, 26 of 48 (54.2%) patients met the criteria to be appropriate candidates for FA-free FLP.
In this study, the presence of both PEDs and PROS thinning coinciding with the leakage point was found in a significant number of eyes with CSC. This study also showed that in the majority of the eyes with CSC, the leakage points coincided with PEDs located in the upper half of the neurosensory detachment area. Taken together, these findings may allow us to plan and perform FLP based on OCT imaging alone without FA in a significant proportion of patients using as a target for laser shots the PED localised in the upper half of the neurosensory detachment area and associated with the PROS thinning area.
The presence of PEDs and the coincidence of PEDs with leakage points in most of the cases of acute CSC is well known. Both the presence of PEDs and the coincidence of PEDs with leakage points have been reported to range from 53% to 100%2 and from 32% to 71%,11 respectively. In our study, PEDs were present in 91.7% of the subjects with non-chronic CSC.
Thus, PED appears to be a significant finding for the precise localisation of presumed leakage points without FA. Nevertheless, this approach is limited by the number of cases with more than one PED as well as cases without PED (in total, 33.3% of patients in this study). Moreover, non-coincidence of leakage points with PED could not be ruled out. To improve accuracy in identification of presumed leakage points using OCT, additional topographic OCT findings verifying the focal leakage associated with particular PEDs should be identified.
In our study, an additional OCT finding allowing us to verify the presence of leakage points was PROS thinning in the area of PED, revealed in 89.3% of PEDs coinciding with leakage points. The location of the PED in the upper half of the neurosensory detachment area was another finding allowing us to presume a leakage point coinciding with this particular PED, which was found in 91.6% of cases. The descending movement of the subretinal fluid under the force of gravity could explain the prevalence of localisation of leakage points to the upper half of the neurosensory detachment area.
Analysis of the cumulative spatial distribution of leakage points showed that the majority of leakage points do not localise below the centre of the fovea. This finding can be explained by the anatomical structure of the choroid, which plays an important role in the pathophysiology of CSC.2 The superior area of the posterior pole has a greater baseline choroidal thickness12 and may have a predisposition to choroidal hyperpermeability and RPE decompensation. This finding is consistent with the data of Vukojević et al who reported that most leakage points (32.5%) were located in the upper nasal quadrant,10 but detailed spatial distribution was not described, as it was in our study.
In our study, the area of PROS thinning coincided with the leakage points in 88.9% of patients. Although an erosion of PROS just above the leakage point has been described previously,2 but the reasons for, prevalence and characteristics of this finding, were not analysed. PROS thinning may result from the washout of the PROS by an active flow through the leakage point. The size of the PROS thinning area is quite variable (most likely representing a wide range of duration and intensity of leakage) and this finding alone does not allow us to localise the leakage point with the precision needed for FLP. However, the PROS thinning area found above the PED can reliably prove the presence of leakage associated with this particular PED.
A number of patients included in this study demonstrated findings indicating the presence of leakage points, including (1) dipping of the outer retinal layers, (2) microrip of the RPE and (3) hyporeflective subretinal lucency, however, these findings were less prevalent than PROS thinning.
We propose that all of these findings, as well as the topography of the PED and RPOS thinning area, can be used to identify leakage points in patients with CSC. Based on the above considerations, a cohort of patients appropriated for FA-free FLP can be identified. Ideal candidates for this procedure are patients with acute non-resolving CSC with single PED that is less than 500 µm in diameter localising in the upper one-third to one-half of the neurosensory detachment area with an area of PROS thinning above this PED. In our study, the proportion of patients with acute CSC in which FA-free FLP may be performed based on these characteristics was as high as 54.2%. However, in situations, when leakage point cannot be predicted confidentially, FA-free FLP is not recommended.
Finally, steps of the OCT-based algorithm for the identification of presumed leakage points in patients with CSC include (1) the visualisation of all PEDs within the neurosensory detachment area; (2) the exclusion of the PEDs localised in the lower half of the neurosensory detachment area and (3) an analysis of the PED and PROS thinning area colocalisation; or identification of specific OCT findings associated with leakage points.
Except for intravenous contrast angiography, a number of different diagnostic tools have been described as also applying for evaluation of specific findings associated with leakage point. Standard autofluorescence (FAF) imaging6 13 as well as short-wavelength and near-infrared FAF typically could demonstrate a significantly decreased AF at the leakage point in the great majority of the acute CSC cases.6 14 However, these modalities also found spots that were not leaking on angiogram.6 RM-SLO enables evaluation of the topographic distribution of RPE abnormalities within neurosensory detachment area in acute CSC,15 but without differentiation between RPE protrusions and different subtypes of PEDs.15 Therefore, although all these modalities may identify areas which are suspicious for leakage, they cannot provide reliable guidance for FLP.
At the same time, simultaneous visualisation of the distribution of RPE irregularities (including PEDs) and neuroepithelial detachment is available with en face OCT.16 Moreover among the other most widely used imaging modalities, spectral domain OCT has the greatest axial resolution and can provide three-dimensional view. Although we could combine data from different modalities (eg, OCT and standard FAF) for identification of leakage point, OCT appears to be sufficient to be used alone.
Interestingly, the fact of decreased FAF at leakage point6 13 14 is consistent with the results of our study. Matsumoto et al earlier reported that elongated PROS in CSC contributes to increase of FAF17 so it is reasonable to expect a decreased FAF in the areas of PROS thinning which we described above leakage points in this study.
Visualisation of a leakage point or a site of diffuse leakage using FA (or indocyanine green angiography) is essential for treatment planning in all laser-based approaches to CSC treatment, including FLP, MLT and PDT.18–20 To this end, the FA-free OCT-guided approach can also be used for MLT or PDT because the precision of these procedures is inferior to that of FLP. Nevertheless, further studies should be conducted to verify this suggestion.
To the best of our knowledge, there have been no attempts to realise FA-free guided approach for FLP in CSC except for two papers.21 22 Two patients with CSC with allergy to fluorescein were successfully treated without FA by Boyko and Mal’tsev, using a small single PED as a target for laser.21 Narayanan et al reported a focal laser treatment in pregnant women using RPE microrip as an indicator of leakage point.22 Apart from those papers, our study provides a complex approach which might also make FA-free FLP more frequently used.
In conclusion, this study demonstrates that PEDs localised in the upper half of the neurosensory detachment area and associated with the PROS thinning area coincided with the leakage point in a significant number of patients with CSC. Therefore, patients with non-resolving CSC with a small single PED localised in the upper one-third to one-half of the neurosensory detachment area with an area of PROS thinning above this PED may undergo FA-free OCT-guided FLP treatment. Patients with the presence of focal dipping of the outer retinal layers, microrip of RPE or hyporeflective subretinal lucency also may be candidates for FA-free approach.
Contributors DSM, ANK and JC designed the study. DSM and ANK collected the data. DSM and JC analysed the data. DSM and JC drafted the manuscript. DSM, ANK and JC critically reviewed and approved the manuscript.
Competing interests None declared.
Patient consent Obtained.
Ethics approval The Ethics Committee of Military Medical Academy.
Provenance and peer review Not commissioned; externally peer reviewed.