Retinal Nerve Fiber Layer, Ganglion Cell Complex and Choroidal Thickness Evaluation in Patients with Obstructive Sleep Apnea Syndrome

Purpose To evaluate the association of Obstructive Sleep Apnea Syndrome (OSAS) with Retinal Nerve Fiber Layer (RNFL), Ganglion Cell Complex (GCC) and choroidal tissue thickness (CCT). Methods: This cross-sectional study included 50 patients with OSAS and 50 controls between July 2014-January 2015). OSAS severity was classied by using the apnea-hypopnea index (AHI). RNFL thickness, GCC thickness and CCT analyses were performed by optic coherence tomography (OCT). Statistical analysis was performed using SPSS for Windows version 21 software. Results: The mean intraocular pressure (IOP) of the patients with OSAS 16.1±1.37, the cup/disc (C/D) ratio was 0.45±0.17 and vertical C/D ratio was 0.43±0.16. The control group mean IOP was 15.2±1.03, the C/D ratio was 0.36±0.20 and vertical C/D ratio was 0.34±0.19. These are all important measurements in glaucome and statisticallay signicant (p=0.00, 0.022, 0.012). Althought the differences between the mean values of all quadrants RNFL thickness were similarly lower in OSAS group, this signicance was not statistically different. Similarly the average GCC, minimum GCC and subfoveal CCT values between groups were not signicant (p=0.79, 0.53, 0.9). Conclusion: In summary we found that peripapillary RNFL, GCC, CCT did not change with OSAS patients or the severity of OSAS. Further more we found that IOP, C/D ratio and vertical C/D ratio of patients were all signicantly higher than control group. However multicenter longterm cohort studies are still needed to assess the denite changes of RNFL thickness, GCC and CCT in OSAS patients. inferior, superior and average RNFL thickness obtained. Data for RNFL thickness were acquired using the ‘Optic Disc Cube 200 × 200' The GCC mapping conducted based on macular protocol on fovea with a cube of × with automated measurement of the GC-IPL thickness. Choroidal thickness obtained by enhanced depth modality of OCT ( (cid:0) =840 nm, 26000 A scans/s, and 5 µm axial resolution). The choroidal thickness was shown as the vertical distance between the high-reective retinal pigment epithelium (RPE) and the choroid-sclera junction line. Diurnal variations may exist in the choroidal thickness


Introduction
Obstructive Sleep Apnea Syndrome (OSAS) is a breathing disorder which is related of oxygen desaturation during sleep. Especially it causes complete collapse of the pharyngeal airway or partial collapse repetitively during sleep and generally ends with reduction in oxygen desaturation. Overnight polisomnography (PSG) is the gold standart in diagnosis of OSAS. 1 The prevalence of OSAS is estimated to be 1% to 2% in men and 1.2% to 2.5% in women around the world. 2 The prevalence of OSAS in Turkey was reported as 1.8%. 3 The treatment of OSAS is continuous positive airway pressure (CPAP). OSAS causes hypoxia, hypercapnia and low oxygen saturation. The balance between vasodilatation and vasoconstriction in blood vessels is broken. [4][5][6] OSAS is also associated with ophthalmic disordes. There are many reports in the pathogenesis theories of vascular complications which can caused by that in ammatory processes, oxidative stress, endothelial dysfunction and vascular remodelling in ocular disorders 6,7 . These changes may distrupt the perfusion of the optic nerve head (ONH) and this result in ganglion cell loss. As a result, optic nerve may become more sensitive to increases in the intraocular pressure or optic nerve glaucomatous damage may develop even with normal intraocular pressure. The prevalence of glaucoma in the OSAS was found to be 5.9% to 27% by several studies 8-10 .
Spectral-domain optical coherence tomography (OCT) provides quantitative and reproducible measurements for investigating Retinal Nerve Fiber Layer (RNFL) thickness in glaucoma. In recent studies, a decreased RNFL thickness and increased ONH area-volume parameters measured with OCT have been reported in patients with OSAS 11,12 . The choroid is a vascular layer between the sclera and the retina. The choroid evaluation can be used for detecting of systemic vascular injury. New-generation spectral domain OCT with enhanced scanning speed shows us to see high-resolution images with more accurate measurements. In recent studies, OCT was an effective device for evaluating choroidal tissue thickness 13 . Previous studies also have been suggested that OSAS may also trigger the macular ganglion cell complex (GCC) loss. In recent years GCC can be measured by OCT. 11,14 .
OSAS have effects on RNFL, GCC, macula and choroidal tissue. But overall studies have con icting results. These studies especially focused on the prevalance of glaucoma or choroidal thickness of OSAS patients. In our study we evaluate all posterior ocular structures and whether there was a correlation between these structures and the clinical severity of OSAS.

Methods
This cross-sectional study included 50 patients with OSAS and 50 age-sex matched controls between July 2014-January 2015. The study was directed in according with the Declaration of Helsinki for biomedical research and the study protocol was approved by the local ethics committee of Mustafa Kemal University (18/08/2014/152). All participants were given informed consent.
Diagnoses of OSAS was based on overnight PSG (Compumedics E Series 44 channel polysomnograph, Profusion PSG3 Software (Abbotsford, VIC, Australia)). OSAS severity was classi ed by using the apneahypopnea index (AHI), which was de ned as the number of apnea and hypopnea per hour of total sleep time. Apneas were full cessation of air ow more than 10s. Hypopneas were de ned as reduction of >50% in air ow signal with a fall of ≥3% in oxygen saturation or an arousal. Apneas and hypopneas were considered during the sleep time and AHI ≥5 were considered as OSAS and those with AHI 5-15 were regarded as having mild, AHI 15-30 to have moderate and AHI >30 to have severe OSAS.
All patients underwent a detailed ophthalmic examination. Visual acuity testing, refraction assessment, biomicroscopic investigation, intraocular pressure measurement with Goldman aplanation tonometry and fundus examination were all done. All participants were examined with OCT device (Cirrus HD-OCT 4000 (Carl Zeiss Meditec Inc., Dublin, CA, USA)). Temporal, nasal, inferior, superior and average RNFL thickness values were obtained. Data for RNFL thickness were acquired using the 'Optic Disc Cube 200 × 200' protocol. The GCC mapping was conducted based on macular protocol centered on fovea with a cube of 512 × 128 with automated measurement of the GC-IPL thickness. Choroidal thickness obtained by enhanced depth modality of OCT ( =840 nm, 26000 A scans/s, and 5 µm axial resolution). The choroidal thickness was shown as the vertical distance between the high-re ective retinal pigment epithelium (RPE) edge and the choroid-sclera junction line. Diurnal variations may exist in the choroidal thickness measurements 15 ; therefore, all of the measurements were performed between 09:00 and 12:00 with same ophtalmologist (E.E).
The exclusion criterias from the study were patients with a history of ocular surgery, ocular trauma, any anterior or posterior segment disease, all glaucomas even normotensive glaucoma and ocular hypertension. Systemic hypertension and diabetes mellitus can affect the optic nerve and choroidal blood ow so we also exclude the patients who have these diseases.
Statistical analysis was performed using SPSS for Windows version 21 software (SPSS Inc, Chicago, IL, USA). The descriptive statistics for continuous variables were expressed as mean ± standart deviation and the signi cance of the difference between the mean values of the groups was evaluated using the Stundent's t-test. A p value less than 0.05 was considered statistically signi cant. Comparison between mean values of groups was evaluated by Pearson's correlation tests. Pearson's correlation e cient was used to measure strength and direction of the linear relationship between two variables.

Results
Fifty patients with OSAS (3 mild, 18 moderate, and 29 severe) and fty volunteers were studied. There was no statistically signi cant difference in age, gender, visual acuity and refractive values between the groups. Baseline demographic characteristics of patients are shown in Table 1.
The mean RNFL thickness of the patients with OSAS were lower than the control group. However, this difference was not statistically signi cant. Figure 1 shows the distribution of average RNFL thickness in patients and control group.
In addition, we found that the differences between the mean values of all quadrants RNFL thickness were similarly lower in OSAS group than in control groups. But this signi cance was not statistically different. Table 2 shows the mean RNFL thickness ( ) values in the groups. According to severity of disease there was no correlation was found between mean RNFL values and AHI ( = 0.86, = -0.02).
The RNFL thickness and GCC measurements were not statistically signi cant. The mean IOP, C/D ratio and vertical C/D ratio of the OSAS group signi cantly higher thought us there might be glaucome insidance higher in OSAS group. Table 3 shows optic nerve-related indicators in obstructive sleep apnea syndrome (OSAS) and control groups The subfoveal choroidal thickness of OSAS patients was 301.3±56.1 µm and the control group was 305.3±44.0 µm. There was not signi cant correlation was found between choroidal thickness and severity of disease according to AHI (p=0.9 r=-0.01). Figure 2 shows the distribution of choroidal thickness in patients with OSAS and control group.

Discussion
OSAS is a syndrome characterized by recurrent apnea or hypopnea episodes during sleep, and frequently decrease in blood oxygen saturation (SpO2). Optic disc damage can be occured by loss of ganglion cells. The hypoxia secondary to OSAS-induced imbalance between mediators; nitric oxide (NO) and endothelin can cause this ganglion cell loss. 16 High IOP is the most important risk factor for the glaucoma. The higher IOP causes a change in the ONH structures, especially makes RNFL thinning. A study by Karakucuk et al. 17 found that four patients have glaucome and the prevalence of glaucoma in OSAS patients was 12.9% (n=4/31). They also found a positive correlation between IOP and AHI. In our study we found that IOP values in OSAS group was signi cantly higher than in controls. IOP is important risk factor for glaucoma and glaucomatous ONH damage. At the same time there is no correlation between AHI groups.
Another importent parameter is increase vertical cup/disk (C/D) ratio in glaucomatous optic nerve. A few study were compared vertical C/D ratio between OSAS and control. Sergi et al 18 did not nd any difference in the cup/disk ratio between the OSAS patients and the control group. They found a signi cant correlation between AHI and the cup/disk ratio. 18 In contrast with their study, our study shows that the vertical cup/disk ratio signi cantly higher in OSAS group than in controls without OSAS.
In glaucoma, RNFL thickness decreases progressively. This thinning can be present in eyes of patients with glaucoma before detectable changes occur in the visual eld 19,20 . If a decrease in the RNFL can be detected, it gives an alert to the risk for developing glaucoma. Firstly Kargi et al. 21 were reported RNFL thickness decreases in patients with OSAS. Conversely, a study performed in Turkish population by Adam et al. 22 reported that there is no difference in RNFL thickness between healthy and OSAS subjects.
Another study by Ferrandez et al. 23 reported that RNFL thickness did not differ signi cantly between OSAS and control groups. Similarly in our study, we detected no correlation between OSAS patients and control group in RNFL parameters.
In addition to the measurement of peripapillary RNFL thickness by OCT, measurement of the macular ganglion cell complex thickness has been used in several ocular and neurological diseases, particularly glaucoma. Ferrandez et al 24 showed that neither ganglion cell layer (GCL) nor ganglion cell inner plexiform layer (GC-IPL) thickness were reduced in OSAS patients compared with healthy controls.
Conversely, Huseyinoglu et al. 25 found GCL thickness decreases in patients with severe OSAS compared to the control group. In addition, there was a signi cant negative correlation between AHI and GCL thickness. 24 Kara 26 reported that the GC-IPL thickness in patients with severe OSAS was signi cantly lower than that of controls. In our study we found no signi cant correlation in GCL thickness between OSAS and control group.
The retina is metabolically very active tissue in the human body, and it has a double blood supply from the central retinal artery and the choroid. Certain characteristics of the choroid have been studied histologically. Spaide et al. 27 described a "enhanced deep imaging" (EDI) technique to monitor the full thickness of the choroid and this optimize the parameters of OCT. Karalezli et al 28 showed that the median choroidal thickness was statistically thinner in the eye of patients with severe OSAS than that of the controls. On the other hand Tonini et al. 29 reported that choroidal vascular response to hypoxia does not affected in the patients with OSAS. In a study Yuvacı et al. 5 although differences in choroidal thickness were observed, no signi cant differences were found between the control group and the other OSAS groups. Similarly in our study although central choroidal thickness were thinner in OSAS group but datas were not statistically signi cant.
The different results in several studies may be because of the calibrations of measuring instruments, the duration of OSAS, the demographic characteristics of the participants, the vascular dysregulation present in patients with OSAS. In addition to these general factors, there are some limitations of our study and the major one is the limited sample of patients. A larger cohort study can gives us more meaningful analysis on the relation between the severity of OSAS and posterior ocular ndings. Another limitation was we were the exact timing of the development of OSAS was not known, which has a certain in uence on the degree of hypoxia. The results of our study may be compared after longterm follow up and after CPAP treatment.
In summary we found that peripapillary RNFL, GCC, central choroidal thickness did not change with OSAS patients or the severity of OSAS. Further more we found that IOP, C/D ratio and vertical C/D ratio of patients were all signi cantly higher than control group. These association between OSAS and glaucome supports the several previous studies. However multicenter longterm cohort studies are still needed to assess the de nite changes of RNFL thickness, GCC and CCT in OSAS patients.

Declarations
Con ict of interest: The authors declare no con ict of interest.    Figure 1 The distribution of average retinal nerve ber layer thickness (RNFL) thickness in obstructive sleep apnea syndrome (OSAS) and control groups.