Progression of Corneal Endothelial Insult by Chronic Elevation of intraocular Pressure: A prospective Observational Comparative Study Among Primary Open Angle Glaucoma Patients


 Purpose: To evaluate specular microscopic changes of corneal endothelial cells (CEC) count and morphology in correlation to retinal nerve fiber layer RNFL changes detected by visual field (VF) and OCT in early and advanced primary open-angle glaucoma (POAG).Methods: A prospective observational comparative study was conducted on patients with POAG versus non-glaucomatous patients of the same age group. Specular microscopy, VF test, OCT scans of RNFL, and macular ganglion cell complex (GCC) were performed. Glaucoma group was further subdivided into early and advanced stages.Results: The study included 130 eyes of 130 subjects, 70 were eyes with POAG (glaucoma group), 60 were healthy eyes (control group). Both groups were comparable regarding mean age and sex.In the early glaucoma group, a significant negative correlation was found between the coefficient of variation (CV%) and superior PRNFL thickness (r=-0.5, p-value=0.018). A significant negative correlation was also found between percentage of cellular hexagonality (HEX%) and vertical cup/disc (C/D) ratio (r=-0.43, p-value= 0.035). A significant positive correlation was found between HEX% and (superior, inferior) PRNFL thickness (r=0.53, 0.5 and p-value= 0.008, 0.015) respectively.Mean CEC count was significantly lower in the advanced glaucoma group than in the control group (2958.7±371.2 vs 3085±172.5, p value=0.043). Conclusions: CEC is affected by chronic intraocular pressure (IOP) elevation just like the PRNFL and macular GCC. CEC morphology is insulted in the early stages of POAG while the count can withstand chronic IOP elevation till advanced stages were both are affected.


Introduction
One of the unique characters of the cornea is being clear and transparent, which is dependent on its stromal anatomical structure and the function of the most inner hexagonal endothelial cell layer 1, 2 .
Corneal endothelial cell (CEC) loss is irreversible and to maintain corneal transparency, preservation of these cells is a must 3 . Variable causes can insult corneal endothelium resulting in corneal decompensation and opaci cation like trauma, cataract surgery, and diseases such as diabetes and glaucoma [4][5][6] .
Glaucoma is a potentially blinding disease of variable types, primary or secondary which is associated with optic neuropathy and retinopathy in addition to its effect on corneal endothelial cell 7,8 . Endothelial cell loss in glaucoma can be attributed to the direct effect of compressing elevated intraocular pressure (IOP), the toxicity of medications, and the effect of surgeries and LASER procedures used to control IOP [9][10][11] .
Glaucoma and cataract are more prevalent and commonly coexist among the elderly population, in whom CEC evaluation is mandatory to check its ability to withstand the surgical trauma of cataract surgeries, in which endothelial cell loss is an inevitable consequence 12 , 13 .
The effect of acute IOP rise on corneal endothelium is documented in previous reports 14 , but little information is available about the effect of chronic elevation of IOP on corneal endothelial cells (count and morphology) in primary open angle glaucoma (POAG). In our study, we aimed to study CEC affection in early and advanced POAG and to nd out if there are any correlations between changes occurring in the retinal nerve ber layer (RNFL), macular ganglion cell complex (GCC) on one side, and changes of corneal endothelial cell (count and morphology) on the other side in glaucoma patients. To the best of our knowledge, such a study has not been done before.
Patient And Method

Study Design
A prospective observational comparative study was conducted on patients with POAG (controlled on medical treatment) versus non-glaucomatous patients of the same age group.

Patient Selection
After the approval of the Human Ethical Committee in Sohag faculty of Medicine; written consent was taken from all patients who participated in this study after explanation of the procedure and nature of the disease. The tenets of the Helsinki Declaration were followed.
The study included 130 eyes of 130 subjects, 70 were eyes with POAG (glaucoma group) , 60 were healthy eyes (control group). The right eye was included in the control and glaucoma groups. In the glaucoma group, IOP was controlled for at least 6 months with one or more of the following anti- borderline stage which is a stage de ned between stage 0 and one, and 5 consecutive stages from (1:5) according to damage severity. Classi cation also includes 3 types of defects (generalized, localized, and mixed). 15 In this study, we considered stages one and two GSS2 as an early stage (40 eyes) and groups three, four, and ve GSS2 an advanced group of glaucoma (30 eyes).
Sixty control subjects were the relatives of the patients and workers in Sohag ophthalmic investigations center of the same age group. The control group was examined two times for IOP measurement in two different occasions. They were also tested for VF, RNFL thickness to exclude normal tension glaucoma.
All eyes were subjected to VF examination using standard automated perimetry (SAP) Humphrey Field Analyzer (HFA) II 750 (Carl Zeiss Meditec Inc., Dublin, California) 30-2 test with Standard Swedish Interactive Thresholding Algorithm (SITA) Strategy. SAP tests were classi ed following Anderson and Patella criteria.
OCT of the optic disc (ONH/RNFL scan) and macular ganglion cell layer (GCC, ganglion cell complex segmentation) was performed using RTVue (Optovue Inc., Fermont, CA). Specular microscopy (SP-2000P; TOPCON, Tokyo, Japan). Images were captured from the central cornea (central measurement) and analyzed using IMAGEnet 2000 software (TOPCON). Endothelial (CED), (HEX%), and (CV%) were determined. All examinations were performed by the same examiner who was masked to the participant's condition. Multiple images are taken by a central view for each eye to obtain a clear image of a good quality of the central endothelium, in which the centers of at least 100 contiguous endothelial cells were marked using software available in the system with clear cell boundaries.

Statistics:
Software version (26.0) (SPSS, Inc., Chicago, IL, USA) was used. All participants were chosen by using the systematic random sample technique from the attendees who ful lled the inclusion criteria.
The data were tested for normality using the Kolmogorov-Smirnov test and for homogeneity variances before further statistical analysis.
Differences between groups for continuous measures were analyzed by an independent sample t-test and the chi-squared test for categorical measures. Z test was used to compare 2 proportions.
Pearson's correlation coe cients (r) were used to assess the correlations between variables of corneal parameters, VF, and OCT disc parameters. A p-value ≤ 0.05 was considered statistically signi cant.

Results
The study included 130 eyes of 130 subjects, 70 were eyes with POAG (40 eyes were early glaucoma group and 30 eyes were advanced glaucoma group), 60 were healthy eyes (control group), who were the relatives of the patients and workers in Sohag ophthalmic investigations center of the same age group.
Positive correlations were found between HEX% and (superior, inferior) PRNFL thickness in early glaucoma (r=0.53,0.5 and p-value=0.008,0.015) respectively. (table 3 and   In the advanced glaucoma group; correlations couldn't be found between RNFL changes and corneal endothelial cell changes.

Discussion
The study aims to evaluate and compare corneal endothelial cell changes (count and morphology) in response to chronic IOP elevation in early vs advanced POAG. We also aimed to assess the correlation between corneal endothelial cell changes and RNFL changes as detected by VF and OCT.
To the best of our knowledge, such a study had not been done before. Previous studies discussed corneal endothelial changes in correlation to acute IOP elevation in ACG, the use of anti-glaucoma medications, and anti-glaucoma surgical procedures 18-20 .
Both groups (Control and glaucoma groups) were comparable regarding mean age (58.5±9.42 vs 59.5 ±8.4 years, p-value=0.5) to exclude the age as confounding factor, since POAG is usually seen at older age asscociated with simultaneous age-related endothelial cell loss.
The study revealed a signi cant negative correlation between CV% and superior PRNFL thickness and also between HEX% and vertical C/D, which means that the progressive thinning of the superior PRNFL which is 1stly affected by glaucoma 21  This study revealed early affection of CV% and HEX% but not CEC density, which occurred in the interval period of chronic IOP elevation before the diagnosis of glaucoma has been made. Yu ZY et al reported that elevated IOP cause destruction of the HEX% and the damage of the cells causes further deformation and loss, but he didn't correlate these changes of corneal endothelial cells to PRNFL thickness nor macular GCC thickness 23 Morphological changes detected in early glaucoma with controlled IOP could be attributed only to IOP changes as eyes with previous surgical and LASER procedures-known to affect corneal endothelium in other studies were excluded [24][25][26] . The effect of topical antiglaucoma medication on corneal endothelium is negligible according to Kwou-Yeung Wu et al, 19 .
The study revealed a signi cant reduction of CEC density in advanced glaucoma when compared to the control group. The signi cantly lower CEC density in the advanced glaucoma group is consistent with the literature proved low CEC density among patients with glaucoma, ocular hypertensive glaucoma, and normal-tension glaucoma. Early reduction of CEC density in patients with ACG was reported by Sushma Verma et al which is attributed to the different nature and course of ACG 3 , 17, 27 .
In the advanced glaucoma group; correlations couldn't be found between RNFL changes and corneal endothelial cell changes, it could be due to the oor effect of marked thinning of the RNFL to a certain level that makes the correlations di cult to appear 28, 29 .
From this study, we can conclude that corneal endothelial cells are insulted in POAG with corneal cell morphology affected in early stages of glaucoma while the cell count can withstand chronic IOP elevation till advanced stages were both morphology and count are affected.
This conclusion could be helpful in surgical concerns, especially that most glaucomatous patients would need cataract surgery in different periods of their lives 12,30 .
Patients with cataract and early controlled POAG in whom specular microscopy is di cult to be done as in bedridden and intellectual disability paients, patients with neurological diseases associated with head nodding and severe blepharospasm) cataract surgeries could be done safely relying on the fact that (CEC) density remains within the normal range in early controlled POAG.
However, there were some limitations in the study as it did not include another group of ACG to compare and correlate to control and POAG groups.
Further studies should be done including a larger number of eyes and different types of glaucoma to understand the effect of different types of glaucoma on corneal endothelial cells. Figure 1 Correlation between (HEX%) percentage of hexagonality and superior (RNFL) peripapillary retinal nerve ber layer. Correlation between (HEX%) percentage of hexagonality and inferior (RNFL) peripapillary retinal nerve ber layer.