Choroidal Thickening in Patients with Coronavirus Disease – 2019

Mojtaba Abrishami Mashhad University of Medical Sciences Zahra Emamverdian Mashhad University of Medical Sciences Ramin Daneshvar (  radaneshvar@gmail.com ) Mashhad University of Medical Sciences Khatam-al-Anbia Hospital https://orcid.org/0000-0002-08840907 Neda Saeedian Mashhad University of Medical Sciences Fariba Tohidinezhad Mashhad University of Medical Sciences Mohsen Seddigh-Shamsi Mashhad University of Medical Sciences Mehdi Mazloumi Wills Eye Hospital Saeid Eslami Mashhad University of Medical Sciences


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a highly contagious and pathogenic human coronavirus. The virus can cause a spectrum of diseases from mild asymptomatic cases to severe, fatal conditions with multiorgan failure. Although the majority of cases have a good prognosis, the number of cases and the death toll is still rising sharply worldwide. [1] The rst cluster of cases was reported by China health authority from Wuhan city, Hubei province, China to the World Health Organization (WHO) on December 31, 2019. Thereafter, the WHO declared the COVID-19 outbreak as a Public Health Emergency of International Concern (PHEIC) on January 30, 2020, and recognized it as a global pandemic on March 11, 2020. [2,3] Coronaviruses have already been shown to cause ocular infection in mammals and it seems reasonable to anticipate ocular manifestations with SARS-CoV-2 infection; [4] however, no signi cant ocular nding was previously reported in the Middle East Respiratory Syndrome Coronavirus (MERS-CoV) and SARS-CoV infections. [5] There are few publications on ocular manifestations of COVID-19 and the majority of them originate in Eastern Asia. [4,[6][7][8][9]. Most frequent ocular ndings were in the anterior segment and included conjunctival hyperemia, chemosis, epiphora, and increased ocular secretions. [6] We recently published an article on ocular ndings in hospitalized patients with COVID-19, and found conjunctival congestion and hyperemia as the most frequent anterior segment ndings, followed by chemosis. [10] Ocular ndings were observed in more than half of our cases. [10] Recently, Casagrande and coworkers detected SARS-CoV-2 viral RNA in retinal biopsy specimens of 3 patients out of 14 cases deceased with COVID-19 and underwent autopsy. [11] Furthermore, Marinho and associates presented SD-OCT ndings in 12 patients with COVID-19 and reported hyperre ective lesions at the level of ganglion cell and inner plexiform layers in all cases. [12] However, the signi cance of their ndings was questioned in a later editorial by Vavvas and coworkers who suggest that the ndings could be normal retinal vessels. [13] As vascular involvement is a well-identi ed aspect of COVID-19 and peripapillary vascular changes have been reported recently, it is not surprising to expect chorioretinal involvement in these patients. [14] Hereby, we presented detailed SD-OCT ndings of the macular region in a prospective cohort of 67 eyes of 34 patients recovered from COVID-19. We aimed to investigate the choroidal changes and the chronological relationship between SD-OCT ndings and the disease course.

Study Participants
A cross-sectional study was conducted in Imam Reza General Hospital, a tertiary referral center, devoted to COVID-19 patients in Northeast Iran. Patients recovered from COVID-19 in this center were included in the study. All included subjects had a de nite diagnosis of COVID-19, con rmed by a positive test result of real-time, reverse transcription-polymerase chain reaction (RT-PCR) of nasopharyngeal swab sample, and had at least a two-week recovery period. Patients with a past medical history of diabetes mellitus, systemic hypertension, or dementia, and those with a history of any intraocular surgery, were excluded. nding and measure the subfoveal choroidal thickness (SFCT). Any disagreement was resolved by discussion and consensus of all examiners. The SFCT was measured using the 'Crossline Images'. Brie y, each horizontal or vertical foveal scan of the device is an average of at least twenty B-scan of the macular area with a width of 10 mm; these horizontal and vertical 'crosslines' are centered at the fovea and perpendicular to each other. To measure the SFCT, we used the 'Caliper tool' within OptoVue machine software (version 2018.0.0.14 ) to determine the minimum distance between the outer aspect of Bruch's membrane (identi ed as a hyperre ective line) and the border of the sclera (i.e., the innermost hyperre ective line of the chorioscleral interface). The measurement was repeated in both horizontal and vertical scans and the average was considered as the SFCT.

Statistical Analysis
Statistical analysis was performed using the SPSS program for Windows, version 20 (IBM SPSS Statistics, IBM Corporation, Chicago, IL, USA). Descriptive statistics were used to describe the main variables by providing central tendency measures for continuous variables and frequencies for categorical variables. The normal distribution of variables was evaluated through the Shapiro-Wilk test and normality plots. One sample t-test was used to compare the mean of SFCT with the previously reported mean SFCT in the same background population. [15] Independent sample t-test and Fisher's exact test were respectively used to compare continuous and categorical variables between any two subgroups de ned within the study population. Pearson correlation test was used to nd any possible correlation between continuous variables. For all tests, a p-value less than or equal to 0.05 was considered statistically signi cant.

Ethical Considerations
The study protocol adhered to the tenets of the 1964 Declaration of Helsinki and its later amendments. All subjects participated in the study voluntarily at their convenience and provided written informed consent at the enrolment. The researchers had no superiority to the participants and worked in a separate institution. The study protocol was approved and supervised by the Regional Medical Ethics Committee at Mashhad University of Medical Sciences, Mashhad, Iran (IR.MUMS.REC.1399.104).

Results
Sixty-seven eyes of 34 patients were included in this study. The mean age of participants was 39.6 ± 1.7 years and eighteen (52.9%) were male. All included subjects had a positive RT-PCR test result of the nasopharyngeal swab sample in at least one examination. Fifteen cases were paramedics/ nurses and eight were physicians (six residents, two faculty members). Ten out of 34 patients were hospitalized, but none required intensive care. One hospitalized patient was monocular and lost his left eye because of an open globe injury in childhood.
Dilated choroidal vessels were observed in 55 eyes (82%) and retinal pigment epitheliopathy (n = 4 eyes, 6%) (Fig. 1). The SFCT (380.3 ± 12.40 µm) was signi cantly thicker than the previously reported SFCT in normal background population (310.7 ± 57.5 µm) (p < 0.001). [15] There was a signi cant positive correlation between the "Disease Duration" and the SFCT (Pearson correlation test, r = 0.48, p = 0.004). Nor "Disease Interval", neither "Recovered Period" demonstrated a statistically signi cant correlation with SFCT. Patients who had a "Recovered Period" of more than 30 days, were more likely to display hyperre ective lesions in choriocapillaris compared to the patients with a shorter period since recovery (Fisher's Exact Test, p = 0.01). No other difference was noted between these two groups in terms of abnormal SD-OCT ndings. No difference was noted in the frequency or type of abnormal SD-OCT ndings between hospitalized and outpatient cases.

Discussion
In this study, we evaluated the macular SD-OCT ndings in retinal and choroidal layers of patients recovered from COVID-19 and found some features of the pachychoroid disorder spectrum including pachyvessels and retinal pigment epitheliopathy along with signi cantly thicker SFCT in these patients.
Since Li Wenliang, a glaucoma surgeon in Wuhan, described emerging cases of severe pneumonia with unknown etiology, COVID-19 has received signi cant global attention, and various aspects of its epidemiology, pathophysiology, different kind of disease presentations, and potential treatment were widely investigated. [1] Worldwide, the total number of cases reached almost 75 million patients with near 1.7 million death as November 22, 2020. [16] Initial reports were mainly focused on life-threatening consequences including pneumonia, acute respiratory distress syndrome (ARDS), multiorgan failure, and septic shock. [17,18] Speci c organ involvement studies are being rapidly emerged by many investigators throughout the world. However, there is a paucity of data on ocular involvement, and most of them are limited to the external eye and conjunctival manifestations and tear lm viral load. [5,6,10,19] Marinho and associates were the rst investigators who reported abnormal retinal OCT ndings in 12 patients with COVID-10. [12] They noticed hyperre ective lesions at the level of GCL and IPL, but not at any other retinal or choroidal layers. The authors proposed that the GCL and IPL ndings might be correlated with the central nervous system manifestations that have been described in animal studies of COVID-19 and patients with neurological sequels of the disease. In a recent editorial, Vavvas and colleagues questioned the signi cance of OCT ndings in the Marinho et al report and proposed that the hyperre ective bands may be normal retinal vessels. [13] We observed some hyperre ective lesions in all retinal layers as reported by Marinho and colleagues; however, similar to Vavvas and colleagues, we believe that these hyperre ective bands are probable the retinal vessels and the prominence of these ndings in COVID-19 patients could be due to the vascular involvement (i.e., micro thrombosis and vasodilation) associated with this infection. Besides, we noted some features of the pachychoroid spectrum of disorders, including increased SFCT, presence of pachyvessels, and retinal pigment epitheliopathy. In a recent report, Savastano and associates highlighted peripapillary vascular changes in patients recovered from COVID-19, which indicate the involvement of ocular vasculature in this disease. [14] SARS-CoV-2 uses the angiotensin-converting enzyme (ACE) 2 receptor to enter the host cells and downregulates the ACE2. [20,21] The ACE2 is a major role player in the renin-angiotensin-aldosterone system (RAAS). This system has fundamental roles in several biological and pathological processes, including uid and electrolytes hemostasis, regulation of blood pressure, brosis, cell migration and proliferation, in ammation, neovascularization, and oxidative damage. [22][23][24] RAAS is a highly complex system with multiple components. The rst detected component was renin which was identi ed in 1898. [25] ACE2 was cloned in 2000 and has since been identi ed in multiple organs such as kidney, heart, brain, liver, intestine, and lung. [25,24] ACE2 is a potent regulator of RAAS, which is critical in maintaining the homeostasis of RAAS and counter-balancing the ACE. [26,27] ACE2 is also a facilitator of amino-acid transport in the kidney and intestine. [20] ACE2 expression is increased by aging, [24] smoking, [28], and several chronic diseases, [24] which partially explains higher COVID-19 frequency in these groups.
Besides the systemic RAAS, there are accumulating pieces of evidence that there is a local RAAS in the retina with components most frequently detected on retinal microvessels, glia (e.g. Müller cells), and neurons (ganglion cells). [29] Components of the RAAS have also been found in other ocular structures such as the uveal tract. [29] ACE2 was detected in aqueous humor, retina, and retinal pigmented epithelium (RPE) of humans. [30] It has also been identi ed in INL and Müller cells of rodent retina. [31] Notably, RAAS dysfunction has been contributed in several ocular pathologies, [32] including diabetic retinopathy, [33][34][35] central serous chorioretinopathy (CSC), [36,37] age-related macular degeneration, [38] and retinopathy of prematurity. [39,40] With this background in the mind, there are several hypothetical explanations for the SD-OCT ndings in our patients. Pulmonary dysfunction and multiorgan failure due to ischemia are fundamental ndings in COVID-19. [20] It is well expected that tissues with high oxygen demand have greater damage and the retinochoroidal ndings can be due to these ischemic events. Moreover, loss of ACE2 regulatory functions, result in increased RAAS pathologic process including in ammation, brosis, vascular damage, and leakage. It has been demonstrated that ACE2 dysfunction is accompanied by impaired migration and function of hematopoietic stem/progenitor cells, [35] retinal leukostasis, and upregulation of adhesion molecules. [29,40,39] As the SARS-CoV-2 RNA has been identi ed in the retinal layers of infected patients, [11] the source of retinal and choroidal vasculopathy can be the involvement of vascular endothelial cells by the virus. Müller cells and blood vessels are a potential target of the virus and their processes/branches cross all retinal layers which potentially explain the pan-retinal ndings in COVID-19. Also, it has been shown that the virus interacts with hemoglobin molecules and alter iron homeostasis. [41] Indeed, patients with COVID-19 can have hyper-ferritinemia and it has been suggested as an indicator of more severe disease. [41,42] Tissue damage by free iron and its deposition in retinal layers can be another explanation for our observations in macular SD-OCT. Finally, the retinal ndings can be attributed to the side effects of treatments. As an instance, ritonavir is one of the antiviral treatments suggested for COVID-19. [43,44] Retinal toxicity of ritonavir has been previously reported in a patient with human immune de ciency virus (HIV) infection under treatment with this drug. [45] The authors reported thickened and irregular RPE with loss of integrity in overlying inner-segment/outer-segment (IS/OS) junction at fovea along with ELM irregularity. [45] The drug has also been reported to cause bull's eye maculopathy. [46] In a recent report, Savastano and colleagues also reported a correlation between antiviral therapy and peripapillary vascular changes in COVID-19 patients.
We also noticed some features of pachychoroid, [47] including increased choroidal thickness, abnormal dilation of Haller's layer vessels, and irregular elevation of RPE in our patients. In a recent article by Savastano and colleagues, the authors reported a higher choroidal thickness in patients recovered from COVID-19; however, the difference was not statistically signi cant. [14] The average SFCT in our patients was almost 70 µm higher than the previously reported thickness in normal subjects with the same ethnic background. Rahman and coworkers reported that a change in SFCT more than 32 µm is more than interobserver variability and could be a true change; [48] however, they used the enhanced depth imaging (EDI) protocol on a different machine (Spectralis; Heidelberg Engineering, Heidelberg, Germany) to measure SFCT. [48] Interestingly, aldosterone contributes to the maintenance of normal retinal function by regulating ion/water channels in the Müller cells. [36,29] As mentioned earlier, CSC is a well-known example of pachychoroidal disease caused by a disturbance in RAAS function, [36] and the same pathogenic process can result in choroidal thickening in COVID-19 patients.
Our study had some limitations, including a small sample size and minimal diversity of COVID-19 severity, a short follow-up period, and a lack of a control group. In addition, our OCT machine had a limited resolution, especially for the evaluation of choroidal changes. Using swept-source OCT devices can potentially yield more useful information. Despite these, to the best of our knowledge, this is the rst report on remarkable choroidal thickening and involvement in COVID-19 patients.
In conclusion, we found that in COVID-19, patients can demonstrate choroidal changes that resemble those in the pachychoroid spectrum and choroidal vascular changes may be a prominent clinical nding in this infection. The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Funding:
The authors would like to acknowledge the nancial support of the Vice-Chancellor of Research, Mashhad University of Medical Sciences, Mashhad, Iran for this research project (Grant Number: 990069).

Consent for publication:
Informed consent for images and data publication and identifying clinical details was obtained from the patients.
Financial Disclosure: None of the authors has a con icting nancial or propriety interest to disclose.

Authors' contributions:
All authors contributed to conception and study design. All authors agreed to be accountable for all aspects of the work. All authors made critical revision and nal approval of the manuscript.