Vogt-Koyanagi-Harada (VKH) disease is an autoimmune disorder characterized by bilateral, chronic, and diffuse granulomatous panuveitis.1,2 The classic course of VKH is divided into four distinct stages: prodromal, acute uveitic, convalescent, and chronic recurrent. In the acute stage, ocular manifestations of VKH include diffuse choroiditis and exudative retinal detachment. An early, fast, and accurate diagnosis is crucial for prompt and effective systemic corticosteroid treatment,3 after which retinal detachments may subside and the uveitis usually becomes quiescent.4 A slowly progressive fundus depigmentation has been shown in chronic VKH patients, which may occur even when there are no detectable clinical signs of active VKH.5,6
One problem that can confound the management of patients with suspected VKH is difficulty in distinguishing VKH from treatment-naive central serous chorioretinopathy (CSC), particularly when VKH patients present with macular detachment and choroidal thickening, but without any signs of inflammation in the anterior segment,7 or if patients with CSC manifest exudative bullous retinal detachments.8–10 Accurate distinction is of paramount importance, as the treatments for VKH and CSC are completely different and even contrary, since corticosteroid treatment is the standard for VKH but is to be avoided for CSC. As a result, misdiagnosis and delayed or improper treatment can cause severe vision loss and a poor prognosis.9 Fluorescein angiography (FFA) is currently the gold standard for distinguishing VKH from CSC, but its use may be delayed due to limited medical facilities in some hospitals. Additionally, some patients cannot tolerate FFA because of hypersensitivity to the contrast medium or poor physical status. Finally, the FFA characteristics of multiple areas of leakage at the level of the retinal pigment epithelium (RPE) can overlap between the two conditions. Therefore, other tools and techniques for distinguishing VKH from CSC would appear to be of value.
Because a key feature of VKH is choroidal inflammation,3,11 identifying and monitoring choroidal changes is important for VKH diagnosis and assessment of treatment efficacy. Previous studies have shown that the choroids of patients with acute VKH and patients with CSC are thicker than those of normal controls and patients with chronic recurrent VKH.12–14 Thus, it is hard to distinguish VKH from CSC based on changes in choroidal thickness (CT) alone. An easier and noninvasive alternative to angiography for assessing the status of the choroid is enhanced depth imaging spectral-domain optical coherence tomography (EDI SD-OCT).15
To increase the sensitivity and specificity of EDI SD-OCT in the diagnosis and monitoring of VKH, the changes in the choroid need to be characterized in greater detail. The purpose of this study is to use EDI SD-OCT to better define the choroidal features of patients with VKH at varying stages of disease and contrast these features to those of patients with CSC. To the best of our knowledge, these morphologic changes in the choroids of patients with VKH have not been reported previously.
Patients and Methods
All study procedures adhered to the tenets of the Declaration of Helsinki and were approved by the investigational review board of Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China.
This study was a case control study. VKH and CSC patients who were diagnosed and followed at the Zhongshan Ophthalmic Center from July 2012 to July 2014, and normal participants with similar age and refractive error were also enrolled in this study. Individuals with other ocular diseases that could affect image quality, such as severe cataract, or any other retinal or choroidal disease that could affect the choroid were excluded from this study. The diagnosis of VKH was made in accordance with the revised diagnostic criteria for VKH established by the first international workshop.16 Patients were considered to be in the convalescent stage of VKH when they were free of intraocular inflammation for more than 3 months. Each patient underwent a comprehensive ophthalmic examination, including measurement of best-corrected visual acuity (BCVA), slit-lamp biomicroscopy with a non-contact lens, FFA, and EDI SD-OCT. The refractive error of VKH and CSC patients was measured after the retina was reattached. Patients with acute VKH received oral corticosteroid treatment (1.0 mg/kg to 1.2 mg/kg) upon diagnosis, and treatment was then adjusted in accordance with published standard guidelines.4,17 All CSC patients underwent an initial period of observation, and a few patients with persistent fluid received focal Argon laser treatment; none were treated with photodynamic therapy.
Normal, healthy volunteers were recruited from the hospital staff and postgraduate students. All had a BCVA greater than 1.0, no visual complaints, and no history of eye disease.
Choroidal Imaging and Measurement
EDI OCT imaging for choroidal assessments was obtained using a Spectralis HRA+OCT (Heidelberg Engineering, Heidelberg, Germany). A central 6-mm (1,024 A-scans) horizontal raster was obtained through foveal center in EDI mode with 100 times averaging using the automatic eye tracking system.
The subfoveal CT was vertically measured from the outer highly reflective layer corresponding to the RPE/Bruch's reflective complex to the hyperreflective margin of the choroid-scleral interface.
The number of round or oval hyperreflective profiles with hyporeflective cores in the mid- and outer choroid and the definition/demarcation of the outline of the hyperreflective profiles, subretinal septa, and choroidal folds were evaluated by two independent graders who were masked as to the group assignment.
Acute-stage VKH and CSC patients were scanned on two occasions, once at baseline and once after treatment (at 2 weeks for VKH patients, at the time of retinal reattachment for CSC patients). The age-matched healthy volunteers were only scanned on one occasion.
Data were analyzed using the SPSS statistical software (SPSS 16.0; SPSS, Chicago, IL). A Kruskal-Wallis ranks sum test was used to compare age and spherical equivalent differences between patient and volunteer (control) groups. A Chi-square test was used to compare the effect of gender among all participants. Wilcoxon ranks tests were used to compare between-group age and gender, respectively. A P value less than .05 was considered statistically significant.
Twenty-four patients (14 female, 10 male) with treatment-naïve VKH in the acute stage (48 eyes), 30 patients (nine female, 21 male) with VKH in the convalescent stage (60 eyes), and 24 patients (all male) with treatment-naive acute CSC (24 eyes) were enrolled in this study. In addition, 54 healthy volunteers (20 female, 34 male) participated as normal controls (108 eyes). Table 1 summarizes the demographic characteristics of the participants. Mean age (years ± standard deviation [SE]) for each group was 37.03 years ± 13.18 years (control participants); 38.92 years ± 6.37 years (CSC patients); 38.13 years ± 16.48 years (acute-stage VKH patients); and 41.40 years ± 11.36 years (convalescent-stage VKH patients). There were no significant differences in age (P = .389) or refractive error (P = .174) between groups.
Study Subject Profiles
Characteristic Choroidal Changes
CT changes in the different groups are shown in Table 2. The subfoveal CT (mean ± SE) for each group was as follows: control participants = 275.66 µm ± 73.13 µm; CSC patients = 395.69 µm ± 96.70 µm (baseline) and 345.25 µm ± 100.41 µm (after treatment); acute-stage VKH patients = 731.85 µm ± 228.77 µm (baseline) and 464.80 µm ± 214.86 µm (after 2-week corticosteroid treatment); and convalescent-stage VKH patients = 264.66 µm ± 127.72 µm. The mean subfoveal CT of CSC patients and acute-stage VKH, both at baseline and after treatment, were significantly thicker than that of control participants (P < .001 for each). In contrast, the mean subfoveal CT of convalescent-stage VKH patients tended to be thinner than that of control participants, though this was not statistically significant (P = .145). In the CSC group, the mean subfoveal CT of CSC patients significantly decreased after retinal detachment was resolved (P = .047). The mean subfoveal CT of acute-stage VKH patients decreased markedly after corticosteroid treatment (P = .002).
Changes in Choroidal Thickness on EDI SD-OCT Between Groups
Figure 1 (top left) and Figure 2 illustrate OCT B-scan images of control participants: the thin layer of low-to-medium reflectivity in the innermost aspect of choroid is believed to represent the choriocapillaris; the thicker layer of round or oval hyperreflective profiles with hyporeflective cores in the mid-choroid is presumed to be Sattler's layer; and the thickest layer of round or oval hyperreflective profiles with hyporeflective cores in the outer choroid represents Haller's layer. Images of control participants' choroids showed clear outlines of vessels in Sattler's and Haller's layers.
Examples of enhanced depth imaging spectral-domain optical coherence tomography choroidal images for each study group. Black double-headed arrows indicate the diameter of vessels in Haller's layer; black arrowheads identify vessels in Sattler's layer; long black arrows indicate the thickness of Haller's layer; short black arrows indicate the thickness of Sattler's layer. (Top left) The boundary of the choroid and the outline of choroidal vessels were clearly delineated in control participants. (Top right) In acute-stage Vogt-Koyanagi-Harada (VKH) patients, the choroid lost its normal anatomical structure. (Middle left) Following 2 weeks of corticosteroid treatment for acute-stage VKH patients, the exudative retinal detachment was relieved, the choroid was thinner, and the outline of the vessels was more distinct. (Middle right) Patients with convalescent VKH presented with thinner choroids and fewer hyperreflective profiles with hyporeflective cores than did acute-stage VKH patients. (Bottom left, bottom right) Central serous chorioretinopathy patients' choroids were thinner after treatment, but the anatomical structure of the choroids exhibited little change.
An enhanced depth imaging spectral-domain optical coherence tomography choroidal image of a control participant. There were a number of hyperreflective profiles with hyporeflective cores in Haller's layer (long black arrows) and Sattler's layer (short black arrows). The black double-headed arrow shows the diameter of the vessels in Haller's layer, and the black triangle shows the hyperreflective profile with hyporeflective core in Sattler's layer.
In contrast, the choroidal images of acute-stage VKH patients showed vessels without clear boundaries, and a significant reduction in the number of hyperreflective profiles with hyporeflective cores (Figure 1, top right; Figure 3, top). Of note, the numbers of these structures (presumed vessels) increased slightly after treatment (Figure 1, middle left; Figure 3, middle). In addition, choroidal folds were found in some eyes of acute-stage VKH patients. Hyperreflective dots or membranous structures (subretinal septae) were observed in some eyes with VKH, which were believed to represent fibrous material in these exudative detachments (Figure 3, Top). The number of hyperreflective profiles with hyporeflective cores in choroids of convalescent-stage VKH patients was also markedly fewer than that of control participants (Figure 1, middle right; Figure 3; bottom).
Enhanced depth imaging spectral-domain optical coherence tomography choroidal images of Vogt-Koyanagi-Harada (VKH) patients in acute and convalescent stages. (A) At the beginning of the VKH course, the vessel boundaries of the choroid were fuzzy, and the hyperreflective profiles could not be observed. Choroidal folds (black stars) and subretinal septa (white stars) were present. (B) The number of hyperreflective profiles progressively increased, and the choroidal folds were resolved after oral corticosteroid treatment. (C) The choroid with less hyperreflective profiles of a patient with VKH in convalescent stage.
In contrast, the number of hyperreflective profiles with hyporeflective cores in CSC patients was similar to that of control participants and did not change over time (Figure 1, bottom; Figure 4).
Enhanced depth imaging spectral-domain optical coherence tomography choroidal images of a patient with central serous chorioretinopathy. The number of hyperreflective profiles with hyporeflective cores did not change before or after treatment.
The choroidal morphological changes in each group are shown in Table 3. All the acute VKH patients (48/48, 100%) had fewer hyperreflective dots, and even after the retina reattached, most of these patients (40/48; 83.33%) still demonstrated a lower number of hyperreflective dots. As for CSC patients, the number of hyperreflective dots was similar before and after retina reattachment. Choroidal folds were found in six eyes (12.5%) of acute-stage VKH patients, and these folds flattened after treatment.
Morphologic Features of the Subretinal Space and Choroid on EDI SD-OCT
Reproducibility of Choroidal Assessments
A high level of intergrader reproducibility was observed for the choroidal metrics, including an unweighted Kappa of 0.941 for numbers of hyperreflective profiles and the demarcation of these profiles.
In this study, we utilized EDI SD-OCT techniques to evaluate and compare the morphologic features of the choroid in patients with VKH and CSC. The mean subfoveal CT of both acute-stage VKH patients and CSC patients in our study was quite similar with previous studies.12,13 However, we found markedly fewer hyperreflective profiles with hyporeflective cores (presumed vessels) in the choroid in acute and chronic/convalescent VKH compared to CSC and controls in this study.
This phenomenon appeared to be consistent across all of our subjects and may represent a characteristic finding of VKH that could potentially aid in confirming the diagnosis in questionable cases. However, although this finding is not evident in patients with CSC or in normal subjects, the true specificity of the finding for VKH is unknown. For example, the same finding might be present in other choroidal inflammatory (eg, posterior scleritis, sympathetic ophthalmia) of infiltrative (eg, lymphoma) disorders. In our review of images from previous publications regarding the choroid in VKH, these changes can also be seen, but they were not described in those papers as the focus was on alteration of CT. Changes in CT in various stages of VKH have been well-described in the literature, with the choroid being massively thickened in the acute phase with a reduction in this thickening with treatment, and eventually a thinning in the chronic phase.12,13 It is interesting that despite changes in CT during the various stages of disease, this finding of reduced vascular profiles was consistently observed across all stages.
Classically by histology, the choroidal vasculature is divided into three strata: the Haller's layer of large vessels, the Sattler's layer of medium-sized vessels, and the inner layer of choroidal capillaries (choriocapillaris).14 In EDI SD-OCT B-scan images, the innermost portion of the choroid was relatively hyporeflective, and posterior to the innermost choroid, small hyperreflective structures were observed. The small oval hyperreflective structures gradually became larger in the outer choroid. Based on comparison to histology, one could theorize that the hyperreflective profiles in EDI SD-OCT choroid images represented the vessel wells, and the hyporeflective dots represented the vessel lumens. In our study, patients with CSC had a thicker Haller's layer with larger hyperreflective profiles with hyporeflective cores than control participants, which is consistent with the results of previous studies.14
The OCT features of the choroid in VKH may be better understood by considering histopathological findings. Histopathological studies showed that polymorphonuclear neutrophils and macrophages infiltrate into the thickened choroid and are accompanied by edema and vessel dilatation with a marked disappearance of choroidal melanocytes during acute VKH.11 The interstitial swelling could potentially constrict the vessels, and the exudation may also alter the reflectance of the vessel wells and lumens.18 This could explain the numerous iso-bright reflective dots in the choroidal images of acute-stage VKH patients in our study. We also speculate that the increased permeability of choroid vessels and the inflammatory exudates and infiltration caused by inflammation might account for the increased subfoveal CT, the markedly decreased number of hyperreflective dots, and the disrupted choroidal morphology in patients with acute-stage VKH.
In patients with chronic VKH, fuzzy vessels, late-diffuse hyperfluorescence, and dark dots are observed in late-phase frames of indocyanine-green angiography.19 These signs represent the sequelae of disease-related choroidal inflammation, although no other clinically detectable disease activity can be observed. Histopathological correlation studies have demonstrated the presence of numerous areas of small, depigmented peripheral choroidal lesions in the chronic phase that feature atrophy of the overlying choriocapillaris, the RPE, and the outer retina.11 These small atrophic lesions, presumably caused by the persistent inflammation, may account for the thinning of choroid and the decreased number of hyperreflective profiles in convalescent-stage VKH patients in our study. However, we found no changes in these reflective profiles in the choroid of patients with CSC. This may be because choroidal vessel hyperpermeability and increased hydrostatic pressure are the principal pathophysiological mechanisms responsible for the development of CSC.20,21
Patients with acute VKH may present with choroid folds, where the choroid expands toward the more malleable retina instead of the sclera due to choroidal congestion and exudation.22 However, choroid folds are not a specific clinical feature of VKH disease and can also be found in patients with acquired hyperopia, trauma, or scleritis.23 In our study, choroid folds were observed in some VKH patients but not in any of the CSC patients. Therefore, if present, choroidal folds could also potentially help distinguish between VKH and CSC.
Our study does have some limitations including, most notably, its retrospective design. Prospective, longitudinal studies with larger cohorts are needed to better understand the significance of the changes in the hyperreflective profiles over the course of the disease and in response to treatment. Another limitation of our study is that the assessment of the choroidal morphologic changes is largely qualitative. We did not specifically quantify the mean diameter of the vascular profiles or their numbers per unit volume. As choroidal segmentation techniques improve, possibly with the incorporation of OCT angiography, such a precise quantification may be possible in the future.
In summary, we observed a reduction and loss of demarcation of vascular profiles in the choroid of patients with VKH but not in patients with CSC. These findings may be useful in the differential diagnosis, treatment selection, and monitoring of patients with VKH.
- Sheu SJ. Update on uveomeningoencephalitides. Curr Opin Neurol. 2005;18(3):323–329. doi:10.1097/01.wco.0000169753.31321.4e [CrossRef]
- Damico FM1, Cunha-Neto E, Goldberg AC, et al. T-cell recognition and cytokine profile induced by melanocyte epitopes in patients with HLA-DRB1*0405-positive and -negative Vogt-Koyanagi-Harada uveitis. Invest Ophthalmol Vis Sci. 2005;46(7):2465–2471. doi:10.1167/iovs.04-1273 [CrossRef]
- Fang W, Yang P. Vogt-koyanagi-harada syndrome. Curr Eye Res. 2008;33(7):517–523. doi:10.1080/02713680802233968 [CrossRef]
- Peizeng Yang YR, Bing Li, Fang Wang, Fang W, Meng Q, Kijlstra A. Clinical characteristics of Vogt-Koyanagi-Harada syndrome in Chinese patients. Ophthalmology. 2007;114(3):606–614. doi:10.1016/j.ophtha.2006.07.040 [CrossRef]
- Bouchenaki N, Herbort CP. The contribution of indocyanine green angiography to the appraisal and management of Vogt-Koyanagi-Harada disease. Ophthalmology. 2001;108(1):54–64. doi:10.1016/S0161-6420(00)00428-0 [CrossRef]
- da Silva FT, Hirata CE, Sakata VM, et al. Indocyanine green angiography findings in patients with long-standing Vogt-Koyanagi-Harada disease: a cross-sectional study. BMC Ophthalmol. 2012;12:40. doi:10.1186/1471-2415-12-40 [CrossRef]
- Khairallah M, Kahloun R, Tugal-Tutkun I. Central serous chorioretinopathy, corticosteroids, and uveitis. Ocul Immunol Inflamm. 2012;20(2):76–85. doi:10.3109/09273948.2011.650776 [CrossRef]
- Kunavisarut P, Pathanapitoon K, van Schooneveld M, Rothova A. Chronic central serous chorioretinopathy associated with serous retinal detachment in a series of Asian patients. Ocul Immunol Inflamm. 2009;17(4):269–277. doi:10.1080/09273940802702579 [CrossRef]
- Carvalho-Recchia CA, Yannuzzi LA, Negrão S, et al. Corticosteroids and central serous chorioretinopathy. Ophthalmology. 2002;109(10):1834–1837. doi:10.1016/S0161-6420(02)01117-X [CrossRef]
- Mastropasqua L, Di Antonio L, Toto L, Mastropasqua A, Di Iorio A, Carpineto P. Central serous chorioretinopathy treated with navigated retinal laser photocoagulation: visual acuity and retinal sensitivity. Ophthalmic Surg Lasers Imaging Retina. 2015;46(3):349–354. doi:10.3928/23258160-20150323-09 [CrossRef]
- Rao NA. Pathology of Vogt–Koyanagi–Harada disease. Int Ophthalmol. 2007;27(2–3):81–85. doi:10.1007/s10792-006-9029-2 [CrossRef]
- Nakayama K, Keino H, Okada AA, et al. Enhanced depth imaging optical coherence tomography of the choroid in Vogt–Koyanagi–Harada disease. Retina. 2012;32(10):2061–2069. doi:10.1097/IAE.0b013e318256205a [CrossRef]
- Nakai K, Gomi F, Ikuno Y, et al. Choroidal observations in Vogt-Koyanagi-Harada disease using high-penetration optical coherence tomography. Graefes Arch Clin Exp Ophthalmol. 2012;250(7):1089–1095. doi:10.1007/s00417-011-1910-7 [CrossRef]
- Yang L, Jonas JB, Wei W. Choroidal vessel diameter in central serous chorioretinopathy. Acta Ophthalmol. 2013;91(5):e358–362. doi:10.1111/aos.12059 [CrossRef]
- Fujiwara A, Shiragami C, Shirakata Y, Manabe S, Izumibata S, Shiraga F. Enhanced depth imaging spectral-domain optical coherence tomography of subfoveal choroidal thickness in normal Japanese eyes. Jpn J Ophthalmol. 2012;56(3):230–235. doi:10.1007/s10384-012-0128-5 [CrossRef]
- Read RW, Holland GN, Rao NA, et al. Revised diagnostic criteria for Vogt-Koyanagi-Harada disease: report of an international committee on nomenclature. Am J Ophthalmol. 2001;131(5):647–652. doi:10.1016/S0002-9394(01)00925-4 [CrossRef]
- Peizeng Y. Clinical Uvetitis. 2nd ed. Beijing, China: People's Medical Publishing House; 2004.
- Hirose S, Saito W, Yoshida K, et al. Elevated choroidal blood flow velocity during systemic corticosteroid therapy in Vogt-Koyanagi-Harada disease. Acta Ophthalmol. 2008;86(8):902–907. doi:10.1111/j.1755-3768.2008.01384.x [CrossRef]
- Herbort CP, Mantovani A, Bouchenaki N. Indocyanine green angiography in Vogt-Koyanagi-Harada disease: angiographic signs and utility in patient follow-up. Int Ophthalmol. 2007;27(2–3):173–182. doi:10.1007/s10792-007-9060-y [CrossRef]
- Iida T, Kishi S, Hagimura N, Shimizu K. Persistent and bilateral choroidal vascular abnormalities in central serous chorioretinopathy. Retina. 1999;19(6):508–512. doi:10.1097/00006982-199911000-00005 [CrossRef]
- Spaide RF, Goldbaum M, Wong DW, Tang KC, Iida T. Serous detachment of the retina. Retina. 2003;23(6):820–846; quiz 895–826. doi:10.1097/00006982-200312000-00013 [CrossRef]
- Tanigawa M, Ochiai H, Tsukahara Y, Ochiai Y, Yamanaka H. Choroidal folds in acute-stage vogt-koyanagi-harada disease patients with relatively short axial length. Case Rep Ophthalmol. 2012;3(1):38–45. doi:10.1159/000336451 [CrossRef]
- Lavinsky J, Lavinsky D, Lavinsky F, Frutuoso A. Acquired choroidal folds: a sign of idiopathic intracranial hypertension. Graefes Arch Clin Exp Ophthalmol. 2007;245(6):883–888. doi:10.1007/s00417-006-0455-7 [CrossRef]
Study Subject Profiles
|Normal||CSC||Acute VKH||Convalescent VKH||P Value|
|Age||40.00 ± 14.50||44.30 ± 5.30||38.12 ± 16.47||43.67 ± 12.68||.389|
|RE (D)||−1.12 ± 1.78||−0.53 ± 1.38||−0.16 ± 1.54||−0.73 ± 2.13||.174|
Changes in Choroidal Thickness on EDI SD-OCT Between Groups
|Mean CT (µm)|
|Baseline||After Retinal Detachment||P Value|
|Controls||275.66 ± 73.13|
|CSC||395.69 ± 96.70||345.25 ± 100.41||.047|
|Acute VKH||731.85 ± 228.77||464.80 ± 214.86||.002|
|Convalescent VKH||264.66 ± 127.72|
Morphologic Features of the Subretinal Space and Choroid on EDI SD-OCT
|Number of Eyes With Decrease in Hyperreflective Dots Versus Normal Controls||Choroid Folds||Subretinal Septa|
|Baseline||After Retinal Detachment|
|Acute VKH||48 (100%)||40 (83.33%)||6 (12.5%)||20 (41.7%)|
|Controls||0 (0.0%)||0 (0%)|
|Convalescent VKH||52 (86.67%)|