Abstract
Purpose:
The purpose of this article is to investigate the differences in intrachoroidal structures between eyes with idiopathic central serous chorioretinopathy (CSC) and those with steroid-induced CSC.
Methods:
Thirty eyes of 30 patients with idiopathic CSC and 17 eyes of 15 patients with steroid-induced CSC were studied. Cross-sectional swept-source optical coherence tomography images of eyes with idiopathic and steroid-induced CSC were retrospectively analyzed by the manual layer delineation technique and by the binarization technique.
Results:
The mean subfoveal choroidal thickness (SCT) was not significantly different between the eyes with idiopathic CSC (408.0 ± 106.6 µm) and those with steroid-induced (389.9 ± 105.4 µm; P = .67) CSC. Manual layer delineation showed that the mean ratio of the large choroidal vessel layer thickness to the SCT was significantly higher in eyes with idiopathic CSC (0.874 ± 0.070) than in eyes with steroid-induced CSC (0.828 ± 0.083; P = .02). The binarization method showed that the mean ratio of the luminal areas to the choroidal areas centered at the fovea was significantly higher in eyes with idiopathic CSC (0.747 ± 0.064) than in eyes with steroid-induced CSC (0.701 ± 0.046; P = .01).
Conclusion:
The intrachoroidal structures in steroid-induced CSC were different from those in idiopathic CSC. These findings suggest different pathophysiologic mechanisms may be involved in the pathogenesis of these 2 entities.
Keywords
Introduction
Central serous chorioretinopathy (CSC) is a common ocular disease characterized by serous retinal detachments and/or pigment epithelial detachments, and it occurs most frequently in young and middle-aged men. CSC is characterized by dilation of the middle and large choroidal vessels, and multifocal areas of choroidal vascular hyperpermeability are seen by indocyanine green angiography (ICGA). 1 -4 The pathogenesis of CSC has not been determined, but choroidal circulatory disturbances are thought to play a key role.
CSC is associated with many risk factors including type A personality, stress, pregnancy, cardiovascular diseases, high blood pressure, and hypercortisolism. 5 -7 Patients on corticosteroid therapy are known to be at higher risk for developing CSC. 6,8 Corticosteroids administered orally and intravenously, 6,8 by inhalation, by topical intranasal or dermal application, 6 epidurally, intra-articularly, and periocularly have been reported to induce CSC regardless of dose. 9
The clinical features of steroid-induced CSC are different from those of idiopathic CSC. Specifically, steroid-induced CSC tends to have a lower male preponderance and more bilaterality, and to be more frequently associated with atypical fundus manifestations compared to idiopathic CSC. 10,11 Thus, there appear to be different pathophysiologic mechanisms involved in idiopathic and steroid-induced CSC.
Recent advancements in optical coherence tomography (OCT) have enabled clinicians to obtain cross-sectional images of the choroid using enhanced-depth imaging (EDI)-OCT 12 and swept-source (SS)-OCT. 13 Previous studies reported that eyes with CSC had a thicker choroid than that of normal eyes. 14 However, the eyes with CSC secondary to corticosteroid use were reported to have a thinner choroid than those with idiopathic CSC. 15 Recent studies have attempted to view the morphological features of the choroidal vasculature in eyes with CSC using OCT. 16 -20 We have reported changes in the intrachoroidal structures after half-dose verteporfin photodynamic therapy for CSC analyzed by a binarization technique to quantify the luminal and stromal areas of the choroid and a manual delineation technique to measure the thickness of the intrachoroidal layers. 20 To date, however, there is no report comparing intrachoroidal structures between idiopathic CSC and steroid-induced CSC.
Thus, the purpose of this study is to determine whether there are differences in not only the choroidal thickness but also in the intrachoroidal structures between eyes with idiopathic CSC and those with steroid-induced CSC.
Methods
This is a retrospective study of 47 eyes of 45 patients with CSC examined and treated at the Macula Services of the Tokyo Women’s Medical University. All 45 patients underwent comprehensive ophthalmic examinations including measurements of the refractive error (spherical equivalent) and decimal best-corrected visual acuity (BCVA) determined by Landolt C charts. The participants also had fundus examination by slit-lamp biomicroscopy with and without a contact lens, fluorescein angiography (FA), and ICGA using a confocal scanning laser ophthalmoscopy (HRA-2; Heidelberg Engineering GmbH, Heidelberg, Germany), and SS-OCT (DRI-OCT; Topcon Corporation, Tokyo, Japan).
CSC was diagnosed when subretinal fluid and/or pigment epithelial detachments associated with leaks from the retinal pigment epithelium (RPE) were present during FA. In addition to the inclusion criteria based on FA, dilated choroidal vessels and multifocal choroidal vascular hyperpermeability were also detected by ICGA in all cases.
For both idiopathic and steroid-induced CSC, the duration of symptoms was limited to ≤3 months to exclude more-chronic cases. We defined the cases using corticosteroid medications at the onset of symptoms as steroid-induced CSC.
SS-OCT was performed with 12-mm horizontal and vertical line scans through the foveal center, and up to 96 B-scan images were averaged to reduce the speckle noise. The horizontal OCT images were analyzed by 2 methods.
Choroidal Layer Analyses of OCT Images
The choroidal layers were analyzed by the methods used by Branchini et al. 16 We used the caliper function embedded in the SS-OCT to measure the size of the vessels and classified choroidal vessels measuring ≥100 μm transversely as large vessels. The subfoveal choroidal thickness (SCT) was defined as the distance between the hyperreflective line representing the Bruch membrane inferior to the RPE and the inner surface of the sclera.
The thickness of the large choroidal vessel layer (LCVL) was determined by drawing a line from the innermost point of the large choroidal vessel vertically to its intersection with the SCT measurement line. The thickness of the LCVL was defined as the distance from the intersection point on the SCT line to the inner surface of the sclera. The choriocapillaris-middle choroidal vessel layer (CC+MCVL) thickness was calculated by subtracting the distance of LCVL thickness from the SCT.
Binarization of Choroidal OCT Images
We used the binarization method of Sonoda and colleagues 17 to quantify the luminal and stromal areas of the choroid using ImageJ software (ImageJ version 1.47, National Institutes-of-Health, Bethesda, MD, USA; available at: http://imagej.nih.gov/ij/). Briefly, we analyzed a 3 mm–wide region that was 1.5 mm nasal and 1.5 mm temporal to the fovea.
The region-of-interest manager in the ImageJ software was used to set the vertical borders from the RPE to the inner surface of the sclera. Then, 3 choroidal vessels with lumens larger than 100 μm were selected with the oval selection tool, and the average reflectivity of these vessels was determined. The average brightness was set as the minimum value of the large vessels so they could be identified automatically.
The choroidal image was binarized by the Niblack method, and the image was converted to 8 bits and adjusted by the Niblack auto local threshold. The binarized image was converted to a red, green, and blue image again, and the luminal area was determined using the threshold tool.
After adding the distance information of each pixel, we calculated the luminal areas and the stromal areas automatically. The light pixels were defined as the stromal areas, and the dark pixels were defined as the luminal areas.
Statistical Analyses
The data were analyzed with frequency and descriptive statistics. Decimal BCVA was converted to the logarithm of the minimal angle of resolution units; χ2 tests were used for categorical analyses, and the Fisher exact tests were used when the expected cell count was less than 5. Mean values were compared by the Mann–Whitney U test. The data are expressed as the means ± SDs, and P < .05 was considered significant. All tests were 2 sided.
Results
Thirty eyes of 30 patients with idiopathic CSC (3 women and 27 men), and 17 eyes of 15 patients with steroid-induced CSC (8 women and 7 men) were studied. Based on the medical records, the primary causes for corticosteroid use in the 15 patients were post–kidney transplantation for 6, rheumatoid arthritis for 2, systemic lupus erythematosus for 2, membranoproliferative glomerulonephritis for 1, sarcoidosis for 1, sudden deafness for 1, peripheral stem cell transplantation for 1, and an undetermined reason for 1. Fourteen of the 15 patients took the systemic steroid orally, and 1 patient used steroid ointment for an undetermined reason. In the group of 14 patients, the number receiving each daily dose of steroid at the time of CSC onset was 1 for prednisolone 2 mg, 1 for prednisolone 8 mg, 3 for prednisolone 10 mg, 1 for prednisolone 19 mg, 1 for prednisolone 1000 mg, 1 for methylprednisolone 2 mg, 4 for methylprednisolone 4 mg, 1 for methylprednisolone 8 mg, and 1 for methylprednisolone 16 mg.
In terms of the characteristics of patients with idiopathic and steroid-induced CSC, the mean age was not significantly different between the 2 groups (49.7 ± 9.7 vs 48.8 ± 10.1, P = .75), but the mean refractive error (spherical equivalent: –0.3 ± 1.6 vs –2.2 ± 2.2, P = .007) and the men:women ratio (90% vs 47%, P = .01) were significantly higher in patients with idiopathic CSC than in those with steroid-induced CSC.
The mean SCT was not significantly different between idiopathic and steroid-induced CSC eyes (408.0 ± 106.6 µm vs 389.9 ± 105.4 µm, P = .67; Figure 1). The mean LCVL thickness and the mean CC+MCVL thickness determined by the manual layer delineation technique did not differ between idiopathic and steroid-induced CSC eyes (359.3 ± 105.2 µm vs 321.8 ± 91.8 µm, P = .24; and 48.7 ± 23.2 µm vs 68.2 ± 37.7 µm, P = .09, respectively). However, the mean ratio of the LCVL thickness to the SCT was significantly greater in the idiopathic CSC than in steroid-induced CSC eyes (0.874 ± 0.070 vs 0.828 ± 0.08, P = .02; Figure 2).
Graph showing the mean subfoveal choroidal thickness (SCT), large choroidal vessel layer (LCVL) thickness, and choriocapillaris-medium choroidal vessel layer (CC+MCVL) thickness of idiopathic and steroid-induced central serous chorioretinopathy (CSC). There were no significant differences in thickness between idiopathic and steroid-induced CSC.
Graph showing the mean ratio of large choroidal vessel layer (LCVL) thickness to subfoveal choroidal thickness (SCT) in idiopathic and steroid-induced central serous chorioretinopathy (CSC). The mean ratio of LCVL thickness to SCT was significantly greater in idiopathic CSC (0.874 ± 0.070) than in steroid-induced CSC (0.828 ± 0.083) (*P = .02).
The mean luminal and stromal areas determined by the binarization method were not significantly different between idiopathic and steroid-induced CSC eyes (1.133 ± 0.300 mm2 vs 1.040 ± 0.270 mm2, P = .43; 0.861 ± 0.286 mm2 vs 0.738 ± 0.226 mm2, P = .22; and 0.272 ± 0.051 mm2 vs 0.302 ± 0.068 mm2, P = .19; Figure 3). However, the mean ratio of the luminal areas to the choroidal areas was significantly greater in the eyes with idiopathic CSC than in those with steroid-induced CSC (0.747 ± 0.064 vs 0.702 ± 0.046, P = .01; Figure 4).
Mean 3 mm–wide choroidal areas, mean luminal areas, and mean stromal areas in idiopathic and steroid-induced central serous chorioretinopathy (CSC). There were no significant differences between the areas.
Mean ratio of luminal areas to 3 mm–wide choroidal areas centered at the fovea in idiopathic and steroid-induced central serous chorioretinopathy (CSC). The mean ratio of luminal areas to choroidal areas was significantly greater in idiopathic CSC (0.747 ± 0.064) than in steroid-induced CSC (0.702 ± 0.046) (*P = .02).
A representative case of idiopathic CSC and steroid-induced CSC is shown in Figure 5.
(A, B) A 39-year-old man with idiopathic central serous chorioretinopathy (CSC), and (C, D) a 65-year-old man with steroid-induced CSC. (A) Horizontal optical coherence tomographic (OCT) image through the fovea, subfoveal choroidal thickness (SCT), large choroidal vessel layer (LCVL) thickness, choriocapillaris-medium vessel layer (CC+MCVL) thickness, and ratio of LCVL thickness to SCT were 390 μm, 355 μm, 35 μm, and 0.910, respectively. (B) In the binarized OCT image, the choroidal areas, luminal areas, stromal areas, and ratio of luminal areas to choroidal areas were 1.012 mm2, 0.735 mm2, 0.277 mm2, and 0.726, respectively. (C) Horizontal OCT image through the fovea, SCT, LCVL thickness, CC+MCVL thickness, and ratio of LCVL thickness to SCT were 228 μm, 190 μm, 38 μm, and 0.833, respectively. (D) In the binarized OCT image, the choroidal areas, luminal areas, stromal areas, and ratio of luminal areas to choroidal areas were 0.678 mm2, 0.448 mm2, 0.230 mm2, and 0.661, respectively.
Conclusions
Steroid-induced CSC has been reported to have clinical features different from those of idiopathic CSC. 10,11 Our results showed that the intrachoroidal structures of steroid-induced CSC were different from those of idiopathic CSC eyes. Specifically, manual analyses of the choroidal layers showed that the ratio of LCVL thickness to the SCT was greater in idiopathic CSC than in steroid-induced CSC. The binarization of the choroidal OCT images showed that the ratio of luminal areas to choroidal areas was greater in idiopathic CSC than in steroid-induced CSC. These results also indicated that the mean SCT was not significantly different between idiopathic and steroid-induced CSC.
Honda and associates recently reported the mean SCT in steroid-induced CSC was significantly thinner than that in idiopathic CSC. 15 Although the SCT in eyes with steroid-induced CSC was thinner than that of idiopathic CSC in our cohort, the difference was not significant. The reasons why the differences did not reach statistical significance might be the limited sample size and the variable choroidal thickness with high variances.
Our results showed that the mean LCVL thickness and the mean CC+MCVL thickness were not significantly different between idiopathic and steroid-induced CSC eyes. However, the ratio of LCVL thickness to the SCT was significantly larger in idiopathic CSC than in steroid-induced CSC. It was reported that the thickness of the Haller layer and the ratio of the Haller layer thickness to SCT were greater in CSC eyes than in normal eyes while the Sattler layer did not differ significantly. 19 These findings suggest that dilation of choroidal vessels was observed mainly in the deeper layer of the choroid.
Our results indicated that the composition of choroidal layers in idiopathic CSC was more markedly altered compared to that in steroid-induced CSC eyes. Chung et al suggested that this was due to nonvascular smooth muscle cells (NVSMCs). 19 The sympathetic network innervates the NVSMCs, and its activation leads to vascular dilation. The density of NVSMCs is known to be present more in the Haller layer than in the choriocapillaris or the Sattler layer; thus, NVSMCs may be involved in the thickening of the deeper layer of the choroid.
In our study, the ratio of LCVL thickness to the SCT was significantly larger in idiopathic CSC than in steroid-induced CSC. Therefore, the sympathetic effect on the deeper choroidal layer may be weaker or the function of NVSMCs may be inactivated in steroid-induced CSC compared to idiopathic CSC.
Our results demonstrated that the mean luminal and stromal areas were not significantly different between idiopathic and steroid-induced CSC; however, the mean ratio of the luminal area to the overall choroidal area was larger in idiopathic CSC than in steroid-induced CSC. Sonoda and colleagues reported that the ratio of the luminal areas to the choroidal areas was significantly greater in CSC eyes than in normal control eyes. 21
Our results suggest that the size of the luminal areas in the choroid in steroid-induced CSC was closer to that in normal eyes than that of eyes with idiopathic CSC. The reason steroid-induced CSC develops under the relatively normal choroidal structures might be that the RPE is more damaged in steroid-induced CSC than in idiopathic CSC. It has been reported that epinephrine can damage the RPE cells, which ultimately leads to apoptosis in vitro. 22 Steroid also causes endothelial dysfunction in blood vessels that may cause hypoperfusion of the choriocapillaris and lead to RPE atrophy. 9
Recently, the relationship between the pathogenesis of CSC and vortex vein status has been reported. 23,24 Hiroe and Kishi 24 suggested that congestion of the vortex vein induces the dilation of outer choroidal vessels. That congestion may increase the permeability of the choriocapillaris and lead to serous retinal detachment.
In our study, the ratios of LCVL thickness to the SCT and the luminal area to the overall choroidal area were significantly greater in idiopathic CSC than in steroid-induced CSC. That fact indicates that the morphological changes in the outer choroidal vessels are less prominent in steroid-induced CSC than idiopathic CSC. Therefore, the vulnerability of the RPE seems to be a more important factor in the pathophysiologic mechanism of steroid-induced CSC than choroidal vascular congestion.
The men:women ratio and the mean refractive error were also significantly different between idiopathic and steroid-induced CSC. Sonoda and colleagues reported that sex and refractive error were not correlated with the ratio of the luminal to the choroidal areas. 18 The effect of these differences on our results may not be considerable.
There are several limitations to this retrospective study. First, the number of cases was limited, so the results cannot be generalized. Second, the study was limited to acute CSC cases. Further studies including more patients with long-term follow-up are needed. Third, we performed the choroidal structural analyses only at the foveal area. Last, the measurements of SCT and each choroidal layer were manually performed, and automated software will be required for more objective evaluations.
In conclusion, the intrachoroidal structures of steroid-induced CSC are different from those of idiopathic CSC. A different pathophysiologic mechanism may be involved in the development of these 2 conditions. Further studies are necessary to determine the clinical significance of the differences in the choroidal structures.
Footnotes
Ethical Approval
This research followed the tenets of the Declaration of Helsinki and was approved by the institutional review board of Tokyo Women’s Medical University (approval number 2636).
Statement of Informed Consent
The informed consent for the routine examination was verbally obtained, but the agreement for this study was not required since it was impossible to identify each patient.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported in part by grant no. 25670739 from the Ministry of Education, Culture, Sports, Science and Technology-Japan to Dr Koizumi.