Evaluation of the Retina, Choroid and Optic Disc Vascular Structures In Individuals With a History of Covid-19

Introduction In this study, we aimed to detect changes in posterior segment structures and posterior segment vascular density caused by COVID-19 using the optical coherence tomography angiography (OCTA) device. Material and Method The study included 20 eyes of 20 patients without any systemic or eye disease, who were followed up at the Ophthalmology Clinic of Health Sciences University Antalya Training and Research Hospital. The OCTA images of these individuals taken before COVID-19 and at six months after recovery were examined. Results The mean choriocapillaris blood �ow value of the patients was 2.00 ± 0.13 mm 2 before COVID-19 and 2.08 ± 0.23 mm 2 after the disease, the mean subfoveal choroidal thickness value was 247.33 ± 7.65 μ m before the disease and 273.08 ± 4.92 μ m after the disease, indicating a statistically signi�cant difference (p = 0.003, p = 0.001, respectively). The mean retinal nerve �ber layer values before and after COVID-19 were 119.33 ± 3.88 and 117.50 ± 3.92 μ m, respectively, and there was a statistically signi�cant decrease (p < 0.001).


Introduction
In this study, we aimed to detect changes in posterior segment structures and posterior segment vascular density caused by COVID-19 using the optical coherence tomography angiography (OCTA) device.

Material and Method
The study included 20 eyes of 20 patients without any systemic or eye disease, who were followed up at the Ophthalmology Clinic of Health Sciences University Antalya Training and Research Hospital.The Introduction Coronaviruses, belonging to the RNA virus family, can cause infections in both humans and animals.Coronavirus disease is an infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).The disease caused by this virus was named COVID-19, where "CO" stands for corona, "VI" for virus, "D" for disease, and "19" indicates the year in which it occurred.COVID-19 is a systemic disease affecting the whole body and manifesting with the general ndings of fever, cough, fatigue, headache, sore throat, loss of taste, loss of smell, and low back pain [1,2].In this disease, ocular ndings can also be seen and sometimes even present as initial ndings.Ocular ndings associated with COVID-19 include eye watering, itching, foreign body sensation, double vision, blurred vision, limitation of eye movements, conjunctivitis, Guillain-Barré syndrome, chemosis, lagophthalmia and ischemic optic neuropathy [3,4].
Optical coherence tomography angiography (OCTA) is a non-invasive imaging modality that has recently been used to visualize retinal vascular blood ow.OCTA can measure both the super cial and deep vessel densities in the macular capillary plexus by detecting motion contrast in blood ow [5].In previous studies, retinal and choroidal vascular density values in patients with COVID-19 were compared with control groups using OCTA, and statistically signi cant results were obtained [6,7].The current study aimed to compare the retinal, choroidal and optic disc vascular density values, choroidal thickness, and retinal nerve ber layer (RNFL) of the same individuals before and after COVID-19 using the OCTA device.We conducted this study with the hypothesis that changes in the posterior segment structures and posterior segment vascular density caused by COVID-19 can be detected using the OCTA device.

Patient Selection
The study included 20 eyes of 20 individuals without any systemic or eye disease, who were followed up at the Ophthalmology Clinic of Health Sciences University Antalya Training and Research Hospital.These individuals consisted of healthcare workers who had been selected as a control group for another academic study before the COVID-19 pandemic and later contracted the virus but recovered without the need for intensive care.The complete ophthalmologic examinations and OCTA ndings of these patients before COVID-19 and at six-month after recovery from the disease were recorded from their les.This study was carried out between 01.01.20 -01.01.21.None of the patients in this study received a coronavirus related vaccine.
The exclusion criteria were determined as follows: any eye pathology (e.g., glaucoma, uveitis, diabetic retinopathy, amblyopia, and epiretinal membrane), myopia greater than -6 diopters, axial length greater than 26 mm, history of eye surgery other than uncomplicated cataract surgery, and any systemic disease.
The demographic characteristics of the patients (age, gender, medications, systemic disease, etc.) and examination ndings were available in patient les.A full ophthalmological examination and OCTA were performed before COVID-19 and during follow-up at six months after the disease.The ndings of the fundus examinations of the patients were evaluated by biomicroscopy.The OCTA images of the patients were examined.A spectral-domain OCTA device (AngioVue; Optovue, Inc, Fremont, CA) was used to obtain OCTA images.

Octa Evaluation
The OCTA measurements of the patients were performed on the 6x6 mm HD angio retina and 4.5x4.5 mm AngioDisc scales.Images with poor shooting quality due to motion and other artifacts were not included in the study.Only images with a signal quality higher than 8/10 were evaluated.The foveal vascular density was measured from a 1-mm circular region, parafoveal vascular density from a 1 mm x 3 mm circular region, perifoveal vascular density from a 1 mm x 3 mm circular region, and total vascular density from a 6 mm circular region in the center of the fovea, and all the values were recorded as percentages.These areas provided the density measurement of the super cial capillary area in automatic mode in cross-section.The ratio of the vascular image (white areas) in these regions to the whole area was used as percentage density.
The foveal avascular zone (FAZ) was automatically calculated by the device from as a measurement in mm 2 .Choriocapillaris blood ow (CBF) was automatically calculated as the amount of area with blood ow in the choriocapillaris layer with a central radius of 1 mm and an area of 3.142 mm 2 and was recorded in mm 2 [Figure 1, 2].Optic disc vascular density was obtained as vascular density measured from a 4.5 mm circular region centered on the optic disc and recorded as a percentage.RNFL was automatically measured and recorded by the device from a 3.4 mm scanning circle centered on the optic disc [Figure 3, 4].Retinal thicknesses were automatically measured between internal limiting membrane and retinal pigment epithelium in the foveal, parafoveal and perifoveal regions and recorded by the device.
Choroidal thickness measurements were performed from the subfoveal area by two different observers using the "Enhanced HD line" image.The average of the measurements of the two observers was calculated.The RPE-sclerachoroidal junction was taken as the boundary of the subfoveal choroidal thickness (SCT).

Statistical Analysis
The analysis of the data was performed using SPSS v. 24.0 software package.Descriptive statistics were given as percentages for categorical variables, and mean, standard deviation and median, minimum and maximum values for numerical variables.The conformity of the numerical variables to the normal distribution was evaluated with the Shapiro-Wilk test.Values were presented using mean and standard deviation.Pre-and post-disease measurements were compared using the Wilcoxon test.The results were evaluated at the 95% con dence interval, and p < 0.05 was considered statistically signi cant.

Results
The mean age of the 20 participants, twelve female and eight male, was calculated as 36.70 ± 1.81 years.None of the patients had pneumonia, and their bleeding pro le values (active partial thromboplastin time, prothrombin time, INR, D-dimer, and brinogen) were normal.The oxygen saturation of all the patients varied between 94 and 100%, and none required oxygen therapy.All the patients received favipiravir and anticoagulant therapy.None of the patients required hospitalization or received steroid therapy.The demographic characteristics of the patients are given in Table 1.Total number of patients 20 Burning and stinging were present in six of the 20 patients (30%), redness in four (20%), itching in ten (50%), eye pain in twelve (60%), and visual impairment in four (20%) patients, while four patients did not have any ocular symptoms.Epiphora, conjunctivitis and neuroophthalmologic symptoms were not detected in any of the patients.[15].In COVID-19 cases, Guillain-Barré syndrome manifesting with diplopia, third or sixth nerve involvement, and nystagmus, as well as its eye component, Miller Fisher syndrome, have also been described [16,17].However, although literature studies have shown that coronaviruses cause anterior uveitis, retinitis, vasculitis and choroiditis in animals, these ndings have not been found in humans [18].
The etiopathogenesis of ocular ndings seen in COVID-19 has been investigated in various studies but has not yet been fully elucidated.Considering the pathogenesis of conjunctivitis symptoms, it is thought that the virus uses the ocular surface as an entry point through the mediation of angiotensin-converting enzyme 2 receptors [19].Peripheral nerve damage, demyelination and in ammation in diplopia are held responsible for visual impairment [20].In Guillain-Barré and Miller Fisher syndromes, the autoimmune mechanism has been suggested to be effective against Schwann cells and myelin antigens [16,17,20].
Savastano et al. compared the OCTA ndings of 80 patients with COVID-19 to 30 healthy controls.The authors reported the mean RNFL value as 94.09 ± 10.77 µm for the COVID-19 group and 96.50 ± 7.78 µm for the control group.Although this difference was not statistically signi cant, the RNFL value of the COVID-19 control was found to be lower than the control group (p = 0.26) [21].In the current study, the mean RNFL was 119.33 ± 3.88 µm before COVID-19 and 117.50 ± 3.92 µm after the disease.The decrease in the RNFL value after the disease was statistically signi cant (p < 0.001).We consider that the reason for this is in ammation and retinal nerve ber damage, as also described in previous studies [20].
Various OCTA studies have been conducted in relation to COVID-19 and its vascular etiopathogenesis.Abrishami et al. compared the OCTA ndings of 30 patients with COVID-19 (nine requiring hospitalized treatment) obtained at two weeks after recovery to 23 healthy individuals.The total, parafoveal and foveal vascular density values of SCP and DCP were found to be statistically signi cantly lower in the COVID-19 group (p = 0.001, p = 0.002, p = 0.004, p = 0.001, p = 0.011, and p = 0.003, respectively [6]. Similarly, Cennoma et al. evaluated the OCTA values at six months after recovery in 40 patients who had developed pneumonia due to COVID-19 and compared them to 40 healthy individuals.The authors reported that the total vascular densities of the optic disc and SCP and the total, parafoveal and foveal vascular densities of DCP were statistically signi cantly lower in the COVID-19 group compared to the control group (p < 0.001, p = 0.038, p = 0.029, p = 0.016, and p = 0.027, respectively). 7In another study, Savastano et al. found the radial peripapillary plexus perfusion density to be statistically signi cantly lower in individuals with COVID-19 compared to the control group, and this difference statistically signi cant (p = 0.041) [21].
In a study evaluating OCTA ndings in 27 patients with a history of COVID-19 at one week after recovery from the disease compared to a healthy control group, Turker et al. observed a statistically signi cant decrease in the parafoveal vascular density values of SCP and DCP in the COVID-19 group (p < 0.05).However, the foveal vascular density did not statistically signi cant differ between the two groups.In the same study, the CBF value was reported to be 2.15 ± 0.23 mm 2 for the COVID-19 group and 2.08 ± 0.11 mm 2 for the control group, indicating a statistically signi cant difference (p = 0.042) [22].In our study, when we compared the retinal and optic disc vascular density values of the same patients before and after COVID-19, we found no statistically signi cant results.While the mean CBF value of the patients was 2.00 ± 0.13 mm 2 before COVID-19, it was 2.08 ± 0.23 mm 2 after the disease.When the pre-and postdisease CBF values, there was a statistically signi cant increase in the post-disease CBF value (p = 0.003).
In an autopsy performed on a person who died due to COVID-19, SARS-CoV-2 ribonucleic acid was detected in the retina [23].Since there are ACE2 receptors in the retina and choroid, these tissues also have binding sites for the SARS-CoV-2 virus [6].Studies have shown that COVID-19 causes vascular endothelial damage and vascular complications.Although endothelial damage, tissue edema, in ammatory cell migration, intravascular coagulation cascade, and endotheliitis are considered to be involved in vascular pathogenesis, the exact mechanism of this cascade remains unknown [24][25][26].
In ammation and increased vascular permeability caused by COVID-19 are held responsible for the increase in CBF [22,27].We consider that despite the increase in the CBF value, the vascular density values of the retina and optic disc were not altered by the disease due to the autoregulation mechanism in the retina [6].In addition, the retina contains a blood retinal barrier, but there is no such barrier in the choroid.Therefore, while in ammation and increased vascular permeability cause an increase in CBF, no change is observed in retinal vascular density values.It has been suggested that individuals that have had severe COVID-19 can develop hypoxia, ischemia, and microvascular damage in the retina [6, 22,28].
The absence of signi cant changes before and after COVID-19 in our study despite the decrease in retinal vascular density values in the literature may be related to the patients in previous studies suffering from a more severe form the disease requiring hospitalization.
Savastano et al. found the mean SCT value as 310.463 ± 81.60 µm in the patients with COVID-19 and 293.5 ± 86.56 µm in the control group.Although not statistically signi cant, the mean SCT value was higher in individuals who had a history of COVID-19 (p = 0.34) [21].In our study, the mean SCT value was measured as 247.33 ± 7.65 µm before the disease and 273.08 ± 4.92 µm after the disease, and this increase was found to be statistically signi cant (p = 0.001).
Tan et al. listed the causes of an altered choroidal thickness as age-related changes, gender differences, sympathetic system activity, metabolic diseases (diabetes mellitus, hypercholesterolemia, etc.), changes due to choroidal hyperperfusion (hypertension), systemic in ammatory conditions (ankylosing spondylitis, Raynaud's phenomenon, Vogt-Koyanagi-Harada disease, Behçet's disease, and sarcoidosis), cerebral vasculature changes (Alzheimer's disease), microarteriolar occlusions (sickle cell disease), choroidal tumors and choroidal metastases [29].We also consider that the increase in the choroidal thickness of our patients was caused by in ammation, increased vascular permeability, and increased CBF.
Considering the limitations of this study, it can be stated that the major limitation was the small number of patients.Another limitation is that none of the patients in this study required intensive care during active infection, and none had a severe form of the disease.Although the patients in our study did not experience severe COVID-19, we found changes in their CBF, SCT and RNFL values compared to the predisease period.The retina, choroid and optic disc may be affected more in severe COVID-19, and this may be re ected in results in the presence of larger data.We consider that our study will contribute to the literature by guiding further research to be conducted with larger samples.

Conclusion
This is the rst study in the literature to evaluate the pre-and post-disease changes in the vascular values of the retina, choroid and optic disc in people that recovered from COVID-19.We consider that in individuals with a history of COVID-19, in ammation and increased vascular permeability are effective in increased CBF and SCT values compared to the pre-disease period.In addition, in ammation and destruction of the retinal nerve ber may be responsible for the decrease in RNFL in these patients.

Declarations
Funding No funding was received for conducting this study.
Con ict of interest All authors declare no funding was received and no con ict of interest/competing interests. Figures

Figure 1 Pre
Figure 1

Table 1
Demographic characteristics of the patients participating in the study

Table 2
presents the comparison of the super cial capillary plexus (SCP) and deep capillary plexus (DCP) vascular density values measured by OCTA before and after COVID-19.No statistically signi cant difference was found in the comparison of the pre-and post-disease vascular density values of SCP and DCP.

Table 2
Comparison of the super cial and deep capillary plexus vascular densities before and after COVID-19 SD: Standard deviation; Min: Minimum; Max: Maximum; SCP: Super cial capillary plexus; DCP: Deep capillary plexus

Table 3
shows the comparison of the FAZ, CBF, SCT and retinal thickness values measured using OCTA before and after COVID-19.No statistically signi cant difference was found in the comparison of the preand post-disease FAZ and retinal thickness values .The mean CBF value of the patients was 2.00 ± 0.13 mm 2 before COVID-19 and 2.08 ± 0.23 mm 2 after the disease.When the pre-and post-disease CBF values

Table 3
Comparison of foveal avascular zone, choriocapillaris blood ow, subfoveal choroidal thickness and total retinal thickness before and after COVID-19

Table 4
[14]ents the comparison of the optic disc vascular density and RNFL values measured before and after COVID-19.No statistically signi cant difference was found in the comparison of the optic disc vascular density values before and after COVID-19.While the mean RNFL value was 119.33 ± 3.88 µm before COVID-19, it was measured as 117.50 ± 3.92 µm after the disease, indicating a statistically signi cant decrease (p < 0.001).4%) of 90 patients with COVID-19[14]while Mao et al. reported visual loss in three (1.4%) of 214 patients SD: Standard deviation; Min: Minimum; Max: Maximum Discussion COVID-19 is a systemic disease that affects the whole body, especially the respiratory tract.The main symptoms of COVID-19 are fever, cough, sore throat, and loss of smell and taste.Ocular symptoms can also be seen and sometimes even present as initial ndings [8].Bertoli et al. reported conjunctivitis, Kawasaki disease, Guillain-Barré syndrome, polyneuritis, encephalopathy, dry eye, exposure keratopathy, chemosis, infectious keratitis, ischemic optic neuropathy, and Horner syndrome in patients with COVID-19 [4].Wu et al. found ocular symptoms, including conjunctivitis, conjunctival hyperemia, Marinho et al. examined 12 patients with COVID-19 and found soft exudate in the retina, retinal hemorrhage, and hyperre ectivity in ganglion cells and inner plexiform on optical coherence tomography [12].In another study, Virgo et al. observed acute macular neuroretinopathy in a patient with COVID-19 and paracentral moderate acute retinopathy in another case [13].Tostmann et al. described eye pain in 31 (34.