Corneal and anterior chamber morphology in patients with nonınfectious ıntraocular ınflammation

Submitted for publication: February 22, 2019 Accepted for publication: March 27, 2020 Funding: This study received no specific financial support. Disclosure of potential conflicts of interest: None of the authors have any potential conflicts of interest to disclose. Corresponding author: Ebru Nevin Çetin. E-mail: cetin.ebru@gmail.com; ecetin@pau.edu.tr Approved by the following research ethics committee: Pamukkale University (# 60116787-020/49878). ABSTRACT | Purpose: To evaluate the corneal and anterior chamber morphology in phakic eyes with noninfectious in­ traocular inflammation. Methods: This study included 59 eyes with active uveitis, 62 with inactive uveitis, and 95 healthy eyes. Corneal endothelial cell density, hexagonal cell ratio, coefficient of variation (CV), corneal thickness and volume, maximum keratometry, and anterior chamber volume and depth (ACD) measurements were performed using a specular microscope and Pentacam HR. Results: The mean duration of uveitis was 24.6 ± 40.5 (0­180) months. The mean number of uveitis attacks was 2.8 ± 3.0 (1­20). Coefficient of variation was significantly higher in the active uveitis group compared with inactive uveitis group (p=0.017, Post Hoc Tukey). Anterior segment parameters other than coefficient of variation were not significantly different between active/inactive uveitis and control groups (p>0.05). Multiple linear regression analysis showed that coefficient of variation was greater in active uveitis compared with inactive uveitis after adjusting for the duration of uveitis, type of uveitis, having a rheumatologic disease, and having immunosuppressive treatment (p=0.003). The duration of uveitis and number of attacks were not significantly correlated with ocular parameters (p>0.05, Spearman’s correlation). The difference in parameters was not significant based on uveitis type (p>0.05). Conclusions: Coefficient of variation was higher in eyes with active uveitis than that in eyes with inactive uveitis, whereas corneal endothelial cell density and anterior chamber morphology did not significantly differ between active/inactive uveitis and control groups.


INTRODUCTION
The healthy corneal endothelium is crucial to maintain the optical transparency of the cornea. The corneal endothelium requires lack of proliferation, and its loss is mainly accomplished by enlargement and spread of neighboring cells to cover any defective area, resulting in increased cellular pleomorphism, and decreased number of hexagonal cells (1) . Therefore, the variation in cell size, endothelial cell density (ECD), and number of hexagonal cells are parameters of healthy corneal endo thelium. Aging, trauma, drugs, contact lens, glaucoma, diabetes, and intraocular surgery are several factors that influenced corneal endothelium (2) .
Chronic and severe intraocular inflammation are often associated with ocular complications, such as glaucoma or cataract (3,4) . Therefore, patients with uveitic eyes are highly at risk of requiring intraocular surgery. In case of corneal endothelial damage, intraocular surgery might be even more challenging in uveitic eyes com bined with posterior synechia, small pupil, and fragile zonules (3) . In previous studies, anterior segment inflam mation was associated with lower ECD and variation of morphologic features including cell size and hexagonal cell ratio (58) . Another study showed that lower mean corneal hysteresis and corneal resistance factor found in anterior uveitis (9) . Some investigators also reported corneal thickness (CT) alterations that tended to increase during an attack and decrease with the treatment for anterior uveitis (1013) . Laboratory studies suggested that intraocular inflammatory reactions mediated by in flammatory cytokines in the aqueous humor might be responsible for corneal alterations in uveitic eyes (1416) .
Although previous studies revealed some alterations in the corneal endothelium and CT, small sample sizes, lack of appropriate control groups, or discrimination of active/inactive uveitis cause difficulties in making clear comments on the effects of intraocular inflammation on corneal structures (57) . This study aimed to assess the corneal ECD and morphological features as well as other anterior segment parameters including corneal volume and thickness, maximum keratometry, anterior chamber depth (ACD), and anterior chamber volume (ACV) as mea sured by Pentacam HR in patients with noninfec tious intraocular inflammation. To our knowledge, this is the largest study evaluating the effects of active/previous anterior segment inflammation on corneal and anterior chamber morphology in phakic eyes.

METHODS
This prospective crosssectional study was conducted at a tertiary setting during the approval of the Institutio nal Review Board and adhered to the principles of the Declaration of Helsinki. Data at the time of enrollment to the study was compared between the study and control groups. All participants provided written informed con sent. The study group consisted of patients with nonin fectious intraocular inflammation, whereas the control group included healthy participants. The study was group further divided into two groups based on uveitis activity. All participants underwent a comprehensive ophthalmic examination, including anterior and posterior segment examination and intraocular pressure mea surements. Anterior chamber cells and uveitis category were scored based on definitions published by the Standardization of Uveitis Nomenclature Working Group (17) . The time since uveietis diagnosis, the coexistent of systemic in flammatory diseases, number of previous uveitis attacks, and medications (topical/intraocular/systemic steroids, antimetabolites, and biological agents) were also re corded. Exclusion criteria were having ocular diseases that affect the cornea and anterior segment (corneal pathology, infectious keratitis/conjunctivitis, etc.), his tory of ocular trauma, intraocular surgery, and contact lens wear; and having poor quality images on specular microscope and Pentacam HR.
A noncontact specular microscope was used (CEM 530, Nidek, Japan) for corneal endothelium assessment. The device provides central, paracentral, and peripheral views with graphics and colorcoded cell images. Corneal ECD, hexagonal cell ratio, and coefficient of variation (CV) data were used for analysis. CV indicated the stan dard deviation of cell area per mean cell area.
Pentacam HR (Oculus, Wetzlar, Germany) provides a threedimensional model of the anterior segment, ele vation maps of the anterior and posterior corneal sur faces, anterior and posterior corneal power calcula tions, pachymetric and biometric measurements of the anterior segment using a rotating Scheimpflug camera (180°), and monochromatic slit light source combined with a static camera. Only scans with quality specifica tion examination of "OK" were considered for analysis. Central CT and volume, maximum keratometry, ACV, and ACD measurements were used for analysis.
In the control group, measurements of only one eye per patient were considered for analysis as findings of the right and left eyes were almost identical. In patients with bilateral uveitis, both eyes were analyzed since each uveitis attack may have different severities and influence of attack may vary between the eyes. Statistical analysis was performed by SPSS statistical software (SPSS 11.0.0 for MS Windows; SPSS Inc., Chicago, IL). KolmogorovSmirnov test was used to determine the data distribution. Parametric tests were used to compare homogenously distributed data, whereas nonparametric tests were used to compare nonhomogenously distri buted data. Categorical variables were compared using the χ2test. A p value <0.05 was considered significant.
The duration of uveitis and number of attacks were not significantly correlated with ocular parameters (p>0.05, Spearman's correlation). The difference in parameters was not significant based on the uveitis type (p>0.05).

DISCUSSION
In this study, CV was found to be higher in the active uveitis group than that in the inactive uveitis group after adjusting the duration and type of uveitis, presence a rheumatologic disease, and presence immunosup pressive treatment. The difference in anterior segment parameters was not significant between the uveitis and control groups. To our knowledge, this is the largest study evaluating the corneal and anterior chamber mor phology in phakic uveitic eyes.
The corneal endothelium in uveitic eyes may be prone to damage in several ways. Proinflammatory cytokines and chemokines had been detected within the ocular fluids or tissues in uveitic eyes (14,15) . They were suggested to be responsible for tissue damage and ocular compli cations in uveitis (16) . In an endotoxininduced uveitis model, investigators observed that corneal edema and significant corneal thickening in the cornea compared with the control group and showed that corneal struc ture could be altered during inflammation (18) . Corneal endothelial damage may be clinically sig nificant in uveitis in two ways. First, uveitic eyes often require intraocular surgeries due to secondary ocular complications such as cataract and glaucoma (3,4) . As in traocular surgery itself is a trauma for endothelium, any additional endothelial damage may further affect the postoperative corneal transparency and visual outcome. Second, CT alterations may affect intraocular pressure measurements and may serve as a confounding factor during the followup of uveitis patients with glauco ma (19) . Therefore, the effect of intraocular inflammation corneal parameters should be clearly understood.
Previous studies focusing on specular microscopic findings in patients with uveitis reported ECD altera tions in a small series (5,6) . In recent years, Alfawaz et al. compared corneal endothelium in 84 eyes with active or previous anterior segment intraocular inflammation with a historical population of normal, agematched participants as the primary control group (7) . The authors reported that ECD was lower in eyes with uveitis than in the control eyes, whereas CT was similar between the two groups. Lack of elimination of pseudophakic eyes and a relatively small number of patients in subgroups cause difficulties comparing findings with other studies. In another recent study, Guclu et al. assessed the effects of previous inflammation on corneal ECD and morpho logy (8) . Investigators found a significant decrease in ECD and hexagonal cell ratio, whereas increased CV in the eyes with inactive uveitis was compared to that in con trol patients. CT was not significantly different between eyes with inactive uveitis and controls.
In this study, CV was significantly higher in active uveitis than that in inactive uveitis group; however, it was not significantly different between inactive uveitis and controls, suggesting that CV alterations, which might reflect the variation in individual cell area caused by stress during active inflammation, were reversible. ECD and hexagonal cell ratio were similar between active, inactive uveitis, and control groups in this study. Pre vious literature reported that increased CV might occur without ECD changes in certain stressful conditions (20) . Persistence of CV alterations in inactive uveitis may be suggested by Guclu et al., which might be associated with the duration of previous active inflammation; however, further studies are needed for further clarification (8) .
CT was shown to increase during an acute uveitis attack (1013,21,22) . However, a significant difference was observed in CT between the groups, irrespective of activity; similar to Alfawaz et al. and Guclu et al.'s studies (7,8) . Since some patients were diagnosed and started treatment before the referral in our hospital, our examination might not be at the most initial period of the attack and therefore might not reflect the effect of the maximum inflammation level on the anterior segment. The difference in findings of these studies may be related to different timing of examination as well as differences in sample sizes ranging from 13 eyes to 121 eyes as in our study.
One of the limitations in this study is the lack of laser flare photometry measurements. The second limitation is that the number of attacks was recorded based on the information provided by patients without previous records in our hospital. Third, some patients presented to our clinic undergoing treatment alreadly; therefore, measurements were not performed on the maximum inflammation level in the active uveitis group. Fourth, participants in this study were patients undergoing uveitis treatment and under followup; therefore, findings did not reflect the chronic results and untreated intraocular inflammation. Final limitation is that 16% of eyes in our series had intermediate or posterior uveitis that may de monstrate minimal or no cells in the anterior chamber.
In conclusion, we found that CV was significantly higher in eyes with active uveitis compared to eyes with inactive uveitis; however, ECD, hexagonal cell ratio, and anterior chamber parameters were similar between uveitic eyes and controls, regardless of uveitis activity. Our findings suggest that some corneal morphological changes appear with intraocular inflammation in the active phase; however, uveitic eyes during the inactive phase are not likely to show significant and permanent anterior segment alterations.