Background
Coronavirus disease 19 (COVID-19) caused by the SARS-CoV-2 virus primarily affects the respiratory system, most often in the form of severe bilateral pneumonia, but many organ systems can be affected (Guan et al., 2020). The SARS-CoV-2 virus enters the cell by binding to angiotensin convertase 2 (ACE2), which is located on the surface of host cells (Hoffmann, Kleine-Weber et al. 2020). ACE2 is present in many tissues, including the lung, kidney, heart, blood vessels, conjunctiva, cornea, and retina (de Figueiredo et al., 2020). In the retina, it is expressed in the vascular endothelium in Müller cells and in neurons of the inner nuclear layer (Piippo, Korhonen et al. 2018, Hoffmann, Kleine-Weber et al. 2020). By binding to ACE2, SARS-CoV-2 reduces its activity, which leads to the accumulation of angiotensin II in the serum and thus vasoconstriction, inflammation, fibrosis, endothelial dysfunction, microvascular thrombosis and the formation of reactive oxygen species (de Almeida-Pititto, Dualib et al. 2020). A pronounced inflammatory response with hyperproduction of inflammatory cytokines and chemokines (e.g. TNF-α, IL-1β and IL-6) can lead to a cytokine storm and multi-organ involvement (Guan, Ni et al. 2020). It is assumed that the combination of a pronounced inflammatory response and endothelial damage causes a hypercoagulable state that predisposes patients to thromboembolic complications, which could also be reflected in the retina. Dysregulation of the renin-angiotensin-aldosterone system and, thus the altered ratio between ACE and ACE2 are thought to influence the course, severity, and outcome of COVID-19. Therefore, ACE, ACE2, and AGTR2 gene polymorphisms could explain why some individuals have a more severe course of COVID-19 (Asselta, Paraboschi et al. 2020, Devaux, Rolain et al. 2020, Hou, Zhao et al. 2020, Zheng and Cao 2020, Abobaker, Nagib et al. 2021). The rs4646994 polymorphism of the ACE gene is a functional polymorphism found on intron 16 in the form of an insertion and/or deletion (I/D) of 289 base pairs (Alu sequences) (Zmorzynski, Szudy-Szczyrek et al. 2019). We decided to investigate this polymorphism because the results of previous research show that it could influence the severity of the disease, while carriers of the D/D genotype are said to have a higher risk of lung involvement and pulmonary embolism as part of SARS-CoV-2 infection in the European population (Howe, Achuthan et al. 2020). The rs2285666 polymorphism (8790 G/A) on the ACE2 gene is located at the beginning of intron 3. Although previous studies have not been able to demonstrate a connection between the presence of the rs2285666 polymorphism and a more severe course of COVID-19, we decided to investigate because of its presumed involvement in the pathogenesis of COVID-19. The rs1403543 polymorphism in the AGTR2 gene (1332A/G) is located in intron 1. We chose the rs1403543 polymorphism because previous studies have shown a possible association with a larger diameter of arterioles in the retina (Liu, Kuznetsova et al. 2011).
Purpose
The aim of this dissertation was to determine the effect of SARS-CoV-2 on the retina in patients with COVID-19, without associated diseases, using optical coherence tomography (OCT) and OCT angiography (OCTA). In addition, we aimed to investigate the influence of selected polymorphisms in the ACE, ACE2, and AGTR2 genes and their association with the presence of retinal findings.
Methods
Patients in the acute phase of COVID-19, hospitalized in the COVID-19 unit of the University medical center Ljubljana, and without associated diseases that could also be reflected on the retina were included in the study. We compared them with a control group of healthy, age-matched subjects. Patients with COVID-19 were divided into two groups according to the course of the disease in accordance with the established classification of the disease (i.e. milder and severe course). Asymptomatic patients and patients with a mild form of the disease who had symptoms and signs of COVID-19 without shortness of breath, dyspnea or abnormal lung imaging were included in the mild course group. The group with a severe course included patients who, according to clinical assessment or imaging tests, had a lower respiratory tract infection and needed oxygen treatment. Demographic data, laboratory results, and the course of treatment were obtained from the patient's documentation. In the first part of the study, we used multimodal imaging methods (photographs of the fundus of the eye, OCT, OCTA, Topcon DRI OCT Triton, Topcon, Inc, Tokyo, Japan) to determine the impact of COVID-19 on the retina and retinal microvessels. The eye with a better image quality index was included in the analysis. We evaluated the fundus phtographs for the presence of retinal hemorrhages, soft exudates, and dilated and tortuous retinal vessels on photographs of the fundus of the eye. The diameter of the retinal vessels was analyzed using the previously described method (Invernizzi et al., 2020), using the ARIA software (Automated Retinal Image Analyser, ARIA, V1-09-12-11), available on the open source platform MATLAB ( MathWorks Inc., Natick, MA, USA). To calculate the mean vein diameter (MVD) and the mean artery diameter (MAD), we used the diameters of the veins of the four main veins and the four main arteries. OCT and OCTA images were automatically segmented with built-in software (Topcon Corp., Tokyo, Japan). We evaluated the correctness of the segmentation in the images and adjusted it manually if necessary. The thickness of the RNFL layer and the GCL layer was evaluated in four quadrants: the upper quadrant (S), the lower quadrant (I), the nasal quadrant (N) and the temporal quadrant (T ) of the inner (3 mm) and outer ring (6 mm) of the early treatment diabetic retinopathy study (ETDRS). Images of the superficial capillary plexus (SCP) and deep capillary plexus (DCP) as well as the foveal avascular zone (FAZ) were analyzed using the open-source MATLAB platform (MathWorks Inc. , Natick, MA, USA) using a validated, previously published method (Díaz et al., 2019). We first determined the center of the FAZ and then superimposed the ETDRS circle on the center of the FAZ and calculated the capillary vessel density in the four quadrants within a 3 mm circle around the center of the FAZ. Each quadrant was binarized separately using the Otsu method (Nicolò et al., 2017). The density of the capillary vessel was expressed in percentages derived from the ratio between the total area of the vessel (white pixels) and the total area of the analyzed region (number of pixels in the quadrant) using the method described by Nicoló et al. (Nicolò et al., 2017). In the quantitative analysis, we used the mean densities of SCP and DCP veins. In the second part of the research, we used multimodal imaging methods to determine the impact of the long-term consequences of COVID-19 on the retina and retinal microvessels, comparing the retinal changes in the acute phase of COVID-19 and one year after discharge from the hospital, and analyzing the images as described. In the third part of the research, we analyzed the influence of selected polymorphisms in the genes for ACE (rs4646994), ACE2 (rs2285666) and AGTR2 (rs1403543) on the severity of the course of COVID-19 and their association with the presence of COVID-19 retinopathy. DNA was isolated in the Research Laboratory for Histology and Genetics of Atherosclerosis and Microvascular Diseases of the Institute of Histology and Embryology, Faculty of Medicine, University of Ljubljana. DNA isolation from 200 μl of blood was performed using the QIAamp DNA Blood Mini Kit (250) (Qiagen GmbH, Hilden, Germany). Selected polymorphisms were determined using the polymerase chain reaction (PCR) method in real-time. Statistical analysis was performed with R statistical software (version 4.1.3, R Development Core Team, Vienna, Austria). We applied the appropriate correction for multiple testing (Benjamini-Hochberg procedure) and considered a p-value of less than 0.05 (p < 0.05) as statistically significant.
Results
In the first part of the study, we included 75 patients in the acute phase of COVID-19 without associated diseases and compared them with 101 subjects from an age-matched control group. The following changes were seen on fundus images: retinal hemorrhages in 5 patients (8.5%), soft exudates in 3 patients (4.0%), and dilated veins in 50 patients (66.6%) patients. Retinal changes were more frequent in the group of patients with severe disease (p = 0.006). Patients with a more severe course of the disease also had a statistically significantly increased diameter of retinal veins (Coef. = 19.28; 95% CI: 7.34–31.23; p = 0.002) and retinal arteries (Coef. = 11.07; 95 %CI: 0.84–21.67; p = 0.044). Vascular thickness was not associated with any of the variables (age, gender, oxygen therapy, LDH, ferritin). Patients with a more severe course of the disease had a significantly thicker retinal nerve fiber layer (RNFL) measured by OCT in the upper and lower quadrants, parafoveolarly (S: p = 0.046; I: p = 0.016) and perifoveolarly (S: p = 0.026; I: p = 0.014). In addition, they also had a significantly thicker ganglion cell layer in the outer temporal quadrant (p = 0.038). There were no differences in the density of the superficial or deep capillary plexus measured by OCTA between the groups of patients with COVID-19 and healthy subjects. In addition, there were no differences in the size or shape of the foveal avascular zone.
In the second part of the research, we included 30 patients in the acute phase of COVID-19, aged 18-65 years, without known associated systemic diseases. Using multimodal imaging methods (photographs of the fundus of the eye, OCT and OCTA, Topcon DRI OCT Triton, Topcon, Inc, Tokyo, Japan), we compared retinal changes in the acute phase of COVID-19 and one year after hospital discharge. The mean diameter of the veins decreased from 134.8 μm in the acute phase to 112.4 μm after one year of follow-up (p < 0.001). One year later, there was also significant RNFL thinning in the lower quadrant of the inner ETDRS ring (p = 0.047) and in the lower quadrant (p < 0.001), nasal quadrant (p < 0.001), and upper quadrant (p < 0.001) of the outer ETDRS ring. There were no statistically significant differences in vessel density of the superficial capillary plexus and deep capillary plexus between the two groups or in the FAZ area.
In the third part of the research, we included 69 hospitalized patients and compared them with 96 healthy controls. According to the course of the disease, the group of patients with COVID-19 was divided into a group with a milder course of the disease (N = 12) and a group with a more severe course of the disease (N = 57). Patients were divided into two groups based on the presence of COVID-19 retinopathy (Yes: N = 50 and No: N = 19). The presence of the AGTR2 rs1403543-AA genotype was associated with a 3.8-fold increased risk of COVID-19 retinopathy in males (p = 0.05). Genotype frequencies in the other selected polymorphisms were not associated with the severity of the course of COVID-19 or the presence of COVID-19 retinopathy.
Conclusions
Our study has shown that retinal changes are more common in patients with a more severe course of COVID-19. Dilatation of the retinal veins and arteries and thickening of the RNFL and GCL in the acute phase of COVID-19 could become biomarkers of systemic angiopathy in patients with a severe course of COVID-19. By analyzing the association of retinal changes with selected polymorphisms, we proved that the risk of retinal involvement is higher in men with the AGTR2 rs1403543-AA genotype. The presence of the AGTR2 rs1403543-AA genotype in men could therefore represent a genetic risk factor for the presence of COVID-19 retinopathy. We, therefore, identified four biomarkers: i) dilatation of retinal vessels, ii) thickening of the RNFL layer, iii) thickening of the GCL layer, and iv) presence of the AGTR2 rs1403543-AA genotype in men. They could be used to predict a more severe course of the disease and a potentially higher risk of systemic angiopathy in the acute phase of COVID-19. Further longitudinal research will be needed to assess potential long-term complications.
|
