Prevalence of Angle Closure Glaucoma in Vogt-Koyanagi-Harada Disease

Background: This study aims to determine the prevalence, clinical characteristics, and mechanisms of secondary glaucoma in Vogt-Koyanagi-Harada (VKH) disease. We performed a retrospective, observational, and longitudinal study in VKH patients seen from 2001 to 2019. Demographic data, disease stage, glaucoma development, visual outcome, and management were analyzed. Results: A total of 305 eyes from 155 VKH patients, including 114 (74.4%) females and 39 (25.6%) males with a mean age of 38.8± 12.8 years, were analyzed. The median follow-up time was 22 months (range 7-60 months). Secondary glaucoma developed in 67 (22.0%) eyes, most of which had chronic recurrent VKH disease at presentation 43 eyes (62.4 %). Angle-closure was the most observed mechanism of glaucoma in 55 (82.1%) eyes, as shown by the development of peripheral anterior synechiae in 58 (86.6%) eyes and posterior synechiae in 51 eyes (76.1%). The combination of pupillary block and posterior synechiae resulted in iris bombé in 17 (25.4%) eyes with glaucoma. Forty-one (61.2%) eyes with glaucoma required either a laser or surgical procedure to control the intraocular pressure. At the last visit, visual acuity was signicantly worse in eyes with glaucoma (p<0.001). Conclusions: As evidenced by the development of anterior and posterior synechiae, the angle-closure disease is a signicant cause of secondary glaucoma in eyes with VKH. Eyes with glaucoma are more likely to present in the chronic recurrent stage of the disease. Most of the eyes with glaucoma required a surgical procedure to control intraocular pressure.

likely to present in the chronic recurrent stage of the disease. Most of the eyes with glaucoma required a surgical procedure to control intraocular pressure.

Background
Vogt-Koyanagi-Harada (VKH) is an autoimmune disease that affects the skin, auditory system, nervous system (meninges), and eyes [1,2]. VKH is characterized by a sudden onset bilateral granulomatous panuveitis during the acute stage of the disease, accompanied by diffuse choroidal stroma in ammation, and exudative retinal detachment [2,3]. Secondary glaucoma is a common and potentially blinding complication of VKH. The combination of open and closed angle mechanisms in the pathogenesis of glaucoma contributes to the development of ocular hypertension and characteristic optic disc neuropathy [4]. The compromised out ow of the trabecular meshwork by in ammatory cells and steroid-induced ocular hypertension are described as two of the most common open-angle glaucoma mechanisms in VKH eyes [5]. However, there is growing evidence that secondary angle-closure disease, characterized by the development of peripheral anterior and posterior synechiae with secondary iris bombé, plays a major role in the development of glaucoma [6]. Few authors have addressed angle-closure disease as a signi cant contributor to glaucoma development in VKH eyes [4,[6][7][8][9].
This study aims to determine the prevalence, clinical characteristics, and mechanisms of secondary glaucoma in a Mexican-mestizo population of VKH patients.

Methods
We conducted a retrospective, observational, and longitudinal single-center case series analysis of Vogt-Koyanagi-Harada (VKH) patients recruited from the Uveitis and In ammatory Eye Disease Service at Fundación de Asistencia Privada Conde de Valenciana, Mexico City, Mexico, between January 2001 and February 2019. Patients included in the study previously read and signed informed consent on using their clinical records data for research purposes. The study was approved by the Institutional Review Board and Ethics Committee following the tenets of the Declaration of Helsinki.
All patients included for analysis ful lled the revised diagnostic criteria for VKH disease and were followed for at least three months [10]. Other forms of uveitis were excluded based on history, laboratory investigation for infectious disease, and clinical examination. The patients´ medical records were entirely reviewed, and clinical information was collected from the rst and last visit, including demographics, stage of VKH disease at the rst visit, best-corrected visual acuity (BCVA), intraocular pressure (IOP), slitlamp biomicroscopic evaluation, lens status, optic nerve assessment, gonioscopy, and medical and surgical management. The mean change in BCVA, IOP, and optic nerve cup-to disk-ratio was recorded at the rst and last visit also.
The clinical features were used to categorize the stage of VKH disease. Initial-onset acute phase was de ned by the presence of the rst episode of bilateral granulomatous uveitis associated with areas of subretinal uid or serous retinal detachment and optic disc hyperemia and swelling. Patients with recurrent acute anterior uveitis or a distinctive depigmented "sunset glow fundus" with peripheral chorioretinal atrophic scars, loss of retinal pigment epithelium, or subretinal brosis were classi ed as having features of chronic recurrent disease. If no episodes of reactivation of anterior uveitis were documented, patients were classi ed in the convalescent phase.
Ocular hypertension (OHT) was de ned as an IOP elevation of more than 22 mmHg in two consecutive visits at least two weeks apart without characteristic glaucomatous optic disc neuropathy. Glaucoma was diagnosed when the cup-to-disc ratio was greater than 0.7, there was an asymmetry between the two eyes greater than 0.2, or a nerve ber layer defect was clinically observed. The iridocorneal angle was classi ed as closed if more than 180 degrees of posterior trabecular meshwork were not visible at gonioscopy without indentation. The presence of anterior peripheral or posterior synechia of more than 3 clock hours was documented. VKH eyes were divided into with or without glaucoma, and their demographic and clinical characteristics were compared in between.
Data were entered into an Excel spreadsheet (version 365, 2020, Microsoft Corporation, Redmond, WA) and analyzed using IBM Statistical Package for Social Sciences (SPSS) v.21 (IBM Inc., Armonk, NY, USA). Normality was assessed using the Kolmogorov-Smirnov test. Normally distributed variables were described with means and standard deviations, while non-normally distributed variables were described with medians and interquartile ranges (IQR). For the comparative analysis, normally distributed data will be examined with the parametric T-student test. The Mann-Whitney and Wilcoxon tests were used for the non-normally distributed data. The Pearson chi-square test was applied to analyze signi cance between variables. A p-value < 0.05 was considered statistically signi cant. BCVA was measured with the Snellen chart under standardized conditions, and it was converted into Logarithm of the Minimum Angle of Resolution (LogMAR) acuity for statistical purposes.
Clinical characteristics of VKH eyes with and without glaucoma Of the 305 eyes included, 67 (22%) developed secondary glaucoma during the follow-up time. We found no difference in age, gender, or follow-up time between groups. OHT developed in 63 (94%) and 62 (26.1%) of the eyes with and without glaucoma, respectively (p < 0.001). Four eyes with glaucoma developed hypotony after trabeculectomy. Glaucomatous eyes were more likely to present in the chronic recurrent stage at the initial visit, 64.2% (n = 43) of eyes compared to only 28.2% (n = 67) of eyes without glaucoma. The acute uveitic phase was more prevalent in the group without glaucoma, with 122 (51.3%) eyes presenting at this stage of disease.
Regarding the lens status, cataract development was more common (n = 36, 53.7%) in glaucomatous eyes, and only 9 (13.4%) eyes remained with a clear lens at the last visit (Table 1). The mean change in IOP, BCVA, and cup-to-disk ratio from the rst to last visit was compared between groups. (Table 2). In the glaucoma group, the median BCVA remained stable during the follow-up (1.8 LogMAR). There was, however, a signi cant change in the median IOP between the rst (17 mmHg, IQR 13-32) and last (14mmHg, IQR 10-21) evaluation. As expected, in eyes with secondary glaucoma, the mean cup-to-disk ratio (C/D) had a 0.4 increase from the rst (C/D = 0.3, IQR 0.3-0.6) to last visit (C/D = 0.7, IQR 0.5-0.9). However, in the non-glaucoma group, the median IOP and C/D remained unchanged, but there was a 0.3 LogMAR improvement in mean BCVA (p < 0.001). Anti-in ammatory management of VKH disease Nearly half of the glaucoma eyes (49.3%) continued to use topical corticosteroids at their last visit, and they were more likely to have been exposed to intravenous methylprednisolone 22 (32.8%). In contrast to the glaucoma group, only 62 (25.6%) of eyes without glaucoma were using topical corticosteroids at the time of the last evaluation and required less intravenous steroid therapy (n = 20, 8.4%). Both groups did not show a difference in the use of intravitreal (p = 0.388) or periocular (p = 0.968) corticosteroids during their follow-up time. The use of nonsteroidal immunosuppressive drugs during the disease course was not signi cantly different between groups (p = 0.668) ( Table 3).

Management of secondary glaucoma
Topical hypotensive drugs, either alone or in combination with glaucoma surgery, were the mainstay of treatment for OHT. Regarding medical treatment, a signi cant proportion of eyes (n = 48, 71.6%) required at least three medications to achieve a successful IOP control between 6-18 mmHg (Table 4). In 18 (7.6%) eyes without glaucoma, a period of transient OHT related to steroids and in ammatory relapses was detected and treated with either temporary IOP-lowering medication and laser or surgical peripheral iridotomy when necessary. Forty-one (61.2%) eyes with glaucoma required an IOP-lowering procedure.

Discussion
Multiple mechanisms have been implicated in secondary glaucoma development in eyes with VKH disease. According to our ndings, the angle-closure disease appears to be a signi cant contributor in combination with open-angle mechanisms for the development of OHT/glaucoma. We found a signi cantly higher prevalence of anterior (86.6%) and posterior (76.1%) synechiae in eyes with glaucoma during the follow-up period. Our ndings are consistent with recent studies that found posterior and anterior synechiae in 64% of glaucoma eyes in a smaller population (n = 28) of VKH patients [6]. However, earlier studies report a lower prevalence, with only 25% (34/136) of VKH eyes exhibiting this iris complication [14]. The value of posterior synechiae as a predictor of VKH-related ocular complications, including glaucoma, was reported in a retrospective analysis of 87 VKH patients treated with high-dose systemic corticosteroids [11]. The relatively high prevalence of anterior and posterior synechiae in our population suggests that VKH recurrent episodes of anterior segment in ammation may be a key mechanism in the development of glaucoma. This condition is supported by the fact that 64.2% of the eyes with glaucoma had chronic disease at presentation, probably due to a late diagnosis or inadequate control of in ammation. Our ndings emphasize the importance of performing gonioscopy during routine evaluations to detect early signs of angle-closure disease.
The prevalence of glaucoma in our study (n = 67 eyes, 22%) is consistent with previous studies that have reported a rate of glaucoma in VKH eyes ranging from 2.6-45%, depending on the year and population studied [4,8,[12][13][14]. However, compared to other studies, the prevalence of angle-closure glaucoma (82.1%) was higher. In a large series of 448 eyes of 224 VKH patients of India, Pandey et al. reported that angle-closure mechanisms were responsible for glaucoma development in only 29.6% of cases [4]. This nding is similar to one of the rst studies of secondary glaucoma in VKH, which found angle-closure, mainly due to pupillary block in 43.7% of the eyes in a Hispanic population [8]. Both studies failed to provide a clear de nition of angle-closure disease. Since we de ned angle-closure as more than 180 degrees of irido-trabecular contact on gonioscopy, we probably found a higher prevalence of secondary angle closure in our study. Our ndings are consistent with Yang et al., who discovered the angle-closure disease in 50.6% of OHT/glaucoma cases in a large study of 695 VKH eyes from China. Among the mechanisms described by these authors were pupillary-block arising from complete iris posterior synechiae (28.9%), extensive peripheral anterior synechiae (10.8%), and acute angle-closure glaucoma at onset (10.9%) [15].
Other in ammatory mechanisms, such as ciliary body swelling and effusion, have been implicated in uveitic glaucoma, in addition to a pupillary block from posterior synechiae, as demonstrated by ultrabiomicroscopic ndings [16,17]. In our study, 8 of the 17 glaucoma-affected eyes had acute angleclosure and an iris bombé con guration, requiring immediate iridectomy. It is likely that besides pupillary block, ciliary body edema with anterior rotation of the ciliary processes resulted in a shallow anterior chamber and acute angle-closure due to poorly controlled in ammation [16][17][18]. However, in the acute angle-closure, no ultrabiomicroscopy was performed.
At the time of presentation, 35.8% of eyes had acute VKH disease, and 64.2% had chronic recurrent VKH disease (p < 0.001). When considering eyes that had angle-closure glaucoma, 69.1% had the chronic VKH disease associated with recurring episodes of anterior segment in ammation. The chronic recurrent form is signi cantly associated with secondary glaucoma [6,12,14,16,19,20]. Arevalo et al. reported that 58.4% and 20.7% of the eyes that developed glaucoma had the chronic recurrent and acute form of VKH disease, respectively (p ≤ 0.0001) [12]. Abu El-Asrar et al. also reported a statistically signi cant association between glaucoma, cataract, or subretinal neovascular formation with the chronic recurrent form of VKH [11].
In our study, baseline IOP was signi cantly higher in glaucoma eyes (p < 0.001). Sixty-three of the eyes (94%) with glaucoma developed OHT, compared to only 26.1% in the non-glaucoma group. In VKH eyes, IOP elevation may result mainly from two mechanisms. First, topical corticosteroid use is a well-known risk factor for IOP elevation. Studies report that 13%-62% of eyes with acute and chronic uveitis of any etiology are steroid-responders [21][22][23]. During the acute stage of VKH, eyes usually have a low IOP secondary to low aqueous production by the ciliary body, which may protect from OHT development.
However, anterior segment in ammation may damage the trabecular meshwork either by accumulation of in ammatory cells, posterior synechiae formation, and/or appositional closure [5]. After the in ammation subsides and aqueous production restores, the damaged angle leads to a compromised aqueous out ow with subsequent IOP elevation [4].
Regarding treatment for VKH disease, 33 eyes (49.3%) and 61 eyes (25.6%) of glaucoma and nonglaucoma group, respectively, were on topical corticosteroids at the last visit (p < 0.001). This difference might be explained by the fact that most eyes in the glaucoma group also have the chronic recurrent form of VKH disease. In such a form, sustained in ammatory control is harder to achieve [6]. Statistical signi cance was also reached in the increased use of intravenous corticosteroids in the glaucoma group (32.8% vs. 8.4%). In our study, 212 eyes (69.5%) from both groups required management with IMT to control intraocular in ammation, with no statistically signi cant difference between groups. The use of IMT has been widely recognized as effective in improving visual outcomes and reducing complications arising from VKH disease, such as cataracts, glaucoma, and/or subretinal neovascular membranes [11,24,25]. Despite the latter, prospective controlled studies are required to determine the role of IMT in glaucoma development in eyes with VKH.
Although most eyes with and without glaucoma (64.2% vs. 96.6%, p < 0.001) met our de nition of IOP control, most eyes in the glaucoma group required laser and/or surgical management (61.2% vs. 9.7%, p < 0.001). Ahmed valve placement was the surgery of choice for glaucoma treatment in 17 eyes (36.2%). Previous evidence supported this nding that both Ahmed valve and trabeculectomy with mitomycin-C achieved adequate IOP control in patients with uveitic glaucoma; however, the cumulative success rate at 1-year favored signi cantly the Ahmed valve group [26]. Regarding other tube implants, a study performed by Chow et al. reported no difference in IOP reduction between trabeculectomy, Ahmed, and Baerveldt implant; however, postoperative hypotony was signi cantly higher in the trabeculectomy group [27]. In our study, hypotony was reported in 5 (71.4%) of the trabeculectomy eyes, with persistent choroidal detachment resulting in two eyes with no light perception. There were no cases of hypotony identi ed in the Ahmed valve group.
Limitations of our study include its retrospective design. Also, since many patients were referred and previously managed, data of their initial presentation and the exact moment of glaucoma development could not be obtained. Our clinical service is a tertiary referral center, and the referral bias poses some limitations on the extrapolation of the results. The relatively large sample size provides a more accurate assessment of glaucoma prevalence and mechanisms in a population of Mexican VKH patients over 18 years.

Conclusion
Secondary glaucoma is a common and sight-threatening complication in the eyes with VKH disease. Our ndings suggest that angle-closure disease, as evidenced by the high prevalence of peripheral anterior and posterior synechiae, is a common mechanism of IOP elevation and glaucoma development in VKH eyes. The chronic recurrent form of VKH disease is signi cantly associated with glaucoma. The clinical observations highlight the importance of adequate in ammatory control and early detection of angleclosure disease with gonioscopy as an aid in the prevention of secondary glaucoma in VKH disease.

Declarations
Ethics approval and consent to participate: The study was approved, prior to initiation, by the Institutional Review Board and Ethics Committee following the tenets of the Declaration of Helsinki. Written informed consent was obtained from each study participant, and from the guardian of underage (< 18 years) particpants prior to clinical examination and data collection. Information was kept con dential throughout the study.

Consent for publication:
Not applicable Availability of data and materials: Results obtained in this study were generated from data collected and analyzed based on the stated methods section. Since all data is already found in the manuscript, there are no supplementary les.