Retinal imaging study diagnoses a case of COVID-19

Background: Hyper-reective lesions at the level of ganglion cell (GCL) and inner plexiform retinal layers (IPL) by Optical Coherence Tomography (OCT) and cotton wool spots in the examination of the eye fundus have recently been described as ndings in patients with COVID-19 infection. Case report: We report a case of a 42-year-old male anesthetist who treated COVID patients during the previous ve weeks and suddenly debuted with a temporal relative scotoma in his left eye (OS); three weeks before, he presented with ageusia for several days. Best corrected visual acuity was 20/20 for OS; no discromatopsy or afferent pupillary defect were present. Visual eld was performed, with no signicant ndings associated to the focal loss of sensitivity referred by the patient. The anterior segment was unremarkable on slit lamp examination in both eyes. Fundus examination of the left eye showed no signicant ndings. A placoid, hyperreective band at the level of GCL and IPL was visible in the temporal and nasal side of the fovea on OCT which spared the outer retina, at the time of diagnosis and at one month. A propharyngeal swab test for SARS-CoV-2 RNA, IgG and IgM ELISA determinations were performed. Real-time reverse-transcriptase polymerase chain reaction (RT ‐ PCR) was negative. ELISA testing and a third rapid antibody detection test performed 7 days after the onset of symptoms were positive.

segment was unremarkable on slit lamp examination in both eyes. Fundus examination of the left eye showed no signi cant ndings. A placoid, hyperre ective band at the level of GCL and IPL was visible in the temporal and nasal side of the fovea on OCT which spared the outer retina, at the time of diagnosis and at one month. A propharyngeal swab test for SARS-CoV-2 RNA, IgG and IgM ELISA determinations were performed. Real-time reverse-transcriptase polymerase chain reaction (RT-PCR) was negative. ELISA testing and a third rapid antibody detection test performed 7 days after the onset of symptoms were positive.
Conclusions: Ocular signs and symptoms in COVID cases are rarely reported, but may be underestimated, especially those that affect the retina and occur in asymptomatic or paucisymptomatic cases. We present the rst case of diagnosis of COVID-19 based on retinal ophthalmic examination.

Background
Coronavirus disease 2019 (COVID-19) is cause by Severe Acute Respiratory Syndrome-Coronavirus 2 (SARS-CoV-2). In humans, diseases of the coronavirus family range from mild common cold to more severe diseases such as Middle East respiratory syndrome (MERS) and SARS [1]. COVID-19 can cause pathological ophthalmologic involvement, initially associated only with conjunctivitis [2] with later descriptions of retinal aggression [3,4].

Case Presentation
We present the case of a 42-year-old male anesthetist who had been working with COVID patients during the previous ve weeks prior to onset, who debuted with a sudden temporal relative scotoma in the left eye (OS); three weeks before, he suffered from ageusia for several days. Best corrected visual acuity was 20/20 for OS, while no discromatopsy or afferent pupillary defect were present. Visual eld was performed, with no signi cant ndings associated to a focal loss of sensitivity referred by the patient.
The anterior segment and fundus examination were unremarkable in both eyes.
Swept Source optical coherence tomography (SS-OCT, Topcon Co., Tokyo, Japan) showed a hyperre ective band at the level of ganglion cell and inner plexiform layers, which spared the outer retina ( Figure 1a and b). Multimodal imaging showed neither hypo-nor hyper-auto uorescence in the area.
Fluorescein angiography showed no areas of leakage or vascular exudation in early or late times.
The patient did not report respiratory symptoms, fever or any other clinical symptoms typically described in COVID-19 cases, but he suffered from ageusia for several days, three weeks prior to the ophthalmologic onset. The thoracic computerized tomography did not show lesions compatible with those described in COVID-19 cases with respiratory involvement.
After identifying the aforementioned retinal lesions and considering the patient's high-risk profession with regards to COVID-19 exposure, a pharyngeal swab test for SARS-CoV-2 RNA and ELISA determination of Early clinical evidence suggests that cases of COVID-19 are frequently characterized by hyperin ammation, renin-angiotensin-aldosterone system imbalance, and a particular form of vasculopathy, thrombotic microangiopathy, and intravascular coagulopathy. In pauci-symptomatic or poorly clinical cases there are no conclusive studies [7].
To date, there is very limited evidence of the correlation between COVID-19 and the appearance of retinal lesions, presumably because there is a wide clinical variation in the presentation and severity of the disease, that may induce the appearance of different mor-phological patterns of retinal involvement. Marinho et al [3], for instance, describe the presence of hyper-re ective lesions at the level of ganglion cell and inner plexiform layers more prominently at the papillomacular bundle, but we must be extremely careful with these ndings because, as Vavvas DG et al [8] point out, OCT hyper-re ective bands in the inner retina and/or ganglion cell layer can confuse us with normal inner retinal vessels. Recently, Landecho [4] described a cotton wool spots in the examination of the eye fundus of the retina and, as corresponds in the B-scan optical coherence tomography, in ammation of the nerve ber layer of the retina appears, in 6 of 24 asymptomatic subjects fourteen days after hospital discharge for bilateral COVID-19 pneumonia. For this reason, we consider the study with multimodal imaging to be important, agreeing with these authors that we must check at least the near infrared re ectance record to con rm that the hyper-re ective bands do not represent normal vessels (Figure 1 and 2).
Exclusively from an ophthalmological point of view and given the potential implica-tions, COVID-19 infection should be excluded using all means available in cases showing these hyper-re ective lesions at the level of ganglion cell and inner plexiform layers in OCT imaging, to facilitate a timely diagnosis and intervention. Vascular occlusions described in COVID-19 cases might as well be the cause for these retinal ndings [7] or could possibly be associated with the neurological manifestations described in animal studies and in COVID-19 neurological events [9,10,11].
This case, and the papers presented by other authors [3,4,8] support a probable hypothesis that these retinal OCT ndings should be considered another sign of COVID-19 disease and the importance of retinal imaging study in these patients. Furthermore, as far as we know, our case is the rst case of COVID-19 diagnosed through an imaging study of the retina.   Figure 1 a and b: Swept Source optical coherence tomography (SS-OCT, Topcon Co., Tokyo, Japan) showed a hyperre ective band at the level of ganglion cell and inner plexiform retinal layers, which spared the outer retina. Green line where the B-scan of the OCT was acquired superimposed automatically by the acquisition instrument on an "en face" infrared fundus image. Figure 2 a and b: One month later, Swept-Source Optical Coherence Tomography shows a more prominent hyperre ective band at the level of ganglion cell and inner plexiform retinal layers. An "en face" infrared fundus image with green line where the B-scan of the OCT was acquired superimposed automatically by the acquisition instrument on.