Abstract
We report here, a case of branch retinal artery occlusion (BRAO) in the left eye of a 76-year-old man that involved three arteries which was considered to be related to a COVID-19 infection due to high levels of blood cytokines and coagulation factors. Although the patient had hypertension and atherosclerosis, his hypertension had been well controlled for the past five years by regular antihypertensive medication. Twenty-five days after starting treatment with anti-inflammatory, anticoagulant and conservative therapy, the patient’s biomarkers of inflammation and coagulation returned to normal and his vision improved. However, some visual field defects remained and were probably a consequence of low oxygen saturation.
Background
Coronavirus disease of 2019 (COVID-19) was a global pandemic that not only affected the respiratory system, but also affected the gastrointestinal system, heart, blood vessels, and other systems. 1 Patients with severe disease often died due to acute respiratory failure or cardiovascular events. 1 Cytokine storm, vascular dysfunction, and thrombosis were important factors in the pathogenesis of COVID-19.1,2 Although the causal relationship is controversial, retinal artery occlusion has been described as a rare vascular disorder in patients with COVID-19. 3 The timely diagnosis of retinal artery occlusion, exploration of risk factors, and adoption of active and effective treatments will undoubtedly lead to good outcomes for affected patients. This article describes a patient with mild COVID-19 who developed branch retinal artery occlusion (BRAO) involving three arteries. The patient's visual function significantly improved following anti-inflammatory, antithrombotic and anticoagulant medications.
Case report
A 76-year-old man presented to our outpatient department on January 4, 2023, with sudden visual impairment and a dark shadow obstruction in his left eye that had persisted for two days. The patient reported that he had no prior history of diabetes or hyperlipidaemia, but had a five-year history of hypertension, which was well controlled with sustained-release nifedipine (30 mg once daily). He had developed a dry cough and fever two weeks previously on December 21, 2022, and a reverse transcription polymerase chain reaction (RT-PCR) test for COVID-19 had been positive. His symptoms abated three days later without drug treatment, and an RT-PCR test was negative nine days later on December 30, 2022. At the time of presentation to hospital, his temperature was 36.8°C, pulse ratewas 108 beats/minute, respiratory rate was 20 breaths/minute, blood pressure was 109/63 mmHg, and oxygen saturation was 94% on room air. His values for corrected-distance visual acuity (CDVA) were 6/20 and 20/20 for his left and right eyes, respectively. His intraocular pressures for left and right eye were 12.3 and 16.8 mmHg, respectively. With the exception of a grade III relative afferent pupillary defect in his left eye, anterior segment examination of both eyes was unremarkable. The lens opacity of both eyes was N1C2P2.
Fundus examination of the left eye showed an extremely fine arterial diameter, and grey-white oedema in the supranasal, infranasal, and supratemporal quadrants that involved the fovea (Figure 1(a)). Macular optical coherence tomography (OCT) showed diffuse signal enhancement in the inner layer of the foveal temporal side of the retina (Figure 1(b)). Macular en-face optical coherence tomography angiography (OCTA) showed the loss of superficial capillaries in the supranasal, infranasal, and supratemporal quadrants (Figure 1(c)). Visual field examination showed a severe defect, with a residual tubular visual field (Figure 2(a)). Fundus examination of the right eye was unremarkable.

Colour fundus images from optical coherence tomography (OCT), and optical coherence tomography angiography (OCTA) of the patient’s left eye at presentation (a–c) and 25 days after starting treatment (d–f). (a) Fundus image showing an extremely fine arterial diameter, with grey-white oedema in the supranasal, infranasal, and supratemporal quadrants that involved the fovea. (b) Macular OCT showing diffuse signal enhancement in the inner layer of the foveal temporal side retina. (c) En- face OCTA showing the loss of superficial capillaries in the supranasal, infranasal and supratemporal quadrants. (d) Fundus image taken 25 days after starting treatment showing the normal colour of the retina. (e) Macular OCT taken 25 days later showing atrophy and thinning of the foveal temporal side retina and (f) en-face OCTA taken 25 days later showing recovery of the superficial capillaries in the supranasal, infranasal and supratemporal quadrants.

Visual field of the patient’s left eye at presentation (a) and at 25 days after starting treatment (b). (a) Visual field examination showing a severe defect, with a residual tubular visual field and (b) substantial improvement in visual field 25 days after starting treatment, although there were residual defects.
With regard to laboratory tests, the following blood parameters were elevated: interleukin-6 (IL-6; 7.8 pg/ml; normal: <7 pg/ml); C-reactive protein (CRP; 30 mg/l; normal: 0–10 mg/l); amyloid protein A (73 mg/l; normal: ≤10 mg/l); erythrocyte sedimentation rate (ESR; 42 mm/h; normal: 0–15 mm/h); D-dimer (1.04 mg/l; normal: 0–0.55 mg/l); fibrinogen (5.1 g/l; normal: 2.0–4.0 g/l); fibrinogen degradation products (FDP; 14 mg/l; normal: 0–5 mg/l). However, the following parameters were decreased: haemoglobin (Hb, 126 g/l; normal: 130–175 g/l); total protein (59 g/l; normal: 65–85 g/l); albumin (36 g/l; normal: 40–55 g/l); red blood cell (RBC) count (3.83 × 109/l; normal: 4.3–5.8 × 109/l). Lipid and homocysteine levels were normal.
Colour duplex ultrasonography showed atherosclerosis with plaques in bilateral lower limb arteries, carotid arteries, and abdominal aorta, but there was no evidence of venous thrombosis. Chest computed tomography (CT) showed an increased texture in both lungs, with scattered patchy ground glass shadows, particularly in the subpleural area (Figure 3).

Chest computed tomography (CT) scan showing an increased texture in both lungs, with scattered patchy ground glass shadows.
The patient was diagnosed with BRAO in his left eye associated with mild COVID-19, and was admitted to hospital for anterior chamber puncture, supplemental oxygen (4 l/min), vasodilatation with a retrobulbar injection of scopolamine hydrobromide (10 mg once daily) and intravenous infusion of isosorbide nitrate (2 mg/h). He was also given anti-inflammatory treatment (i.e., oral prednisone, 30 mg once daily) and anticoagulation (i.e., oral aspirin, 0.1 g once daily and subcutaneous injection of nadroparin calcium, 4100 U/0.4 ml every 12 h).
Nine days after starting treatment, his IL-6 (4.5 pg/ml) and amyloid protein A (<5 mg/l) had normalized, and his CRP (8.6 mg/l), ESR (22 mm/h), D-dimer (0.63 mg/l), fibrinogen (4.4 g/l) and FDP (8.4 mg/l) levels had decreased. Consequently, his treatment was changed to oral aspirin, 0.1 g once daily. After 25 days, the CDVA of his left eye was 20/20, and the retinal colour had returned to normal (Figure 1(d)). Macular OCT showed atrophy and thinning of the foveal temporal side of the retina (Figure 1(e)). En-face OCTA showed the superficial capillaries in the supranasal, infranasal, and supratemporal quadrants had recovered (Figure 1(f)). Although the patient’s visual field had improved substantially, some visual field defects remained (Figure 2(b)). However, his laboratory tests had returned to normal. The patient’s visual acuity remained stable over the following 12 months.
The reporting of this study conforms to CARE guidelines. 4 The patient provided consent for publication of his anonymised data. Publication of this case report was approved by the Ethics Committee of Shanghai Xuhui Central Hospital.
Discussion
Retinal artery occlusions are rare, but may be a manifestation of COVID-19 infection, especially in patients at risk of systemic hypercoagulability and thromboembolism. Indeed, inflammation, hypercoagulability, and thrombosis are important risk factors for retinal vascular occlusion. However, while many case histories and series have reported the occurrence of retinal vascular occlusive events as a consequence of COVID-19 infection, the causal relationship has not been proven.3,5–7 Interestingly, in one study, a 65-year-old woman developed BRAO five weeks following a relatively mild bout of COVID-19 had only elevated levels of D-dimer. 8 She had no other systemic disorders that might cause retinal artery occlusion.
Fundus examinations of patients hospitalized for COVID-19 have shown signs of retinal vascular injury, including cotton wool spots, retinal haemorrhage, venous vessel tortuosity, and dilatation. 9 Moreover, the degree of injury was related to the severity of the COVID-19 infection. In addition, a study using OCTA analysis, showed that retinal vascular densities were lower in patients with COVID-19 compared with patients without COVID-19. 10 Although it is possible that the occurrence of retinal vascular occlusion with COVID-19 infection is coincidental, an association seems a reasonable hypothesis considering the prothrombotic state induced by SARS-CoV-2 infection.
Retinal vascular injury may be related to downregulation of angiotensin converting enzyme 2 (ACE2) expression in the retina. The spike protein of SARS-CoV-2 binds to subdomain I of the ACE2 receptor, which has been detected in retinal tissue. 11 The decreased expression of ACE2 in the retina, together with inflammation and the procoagulant state secondary to COVID-19 infection, may trigger retinal artery thrombosis. For example, in an animal model of diabetic retinopathy, overexpression of ACE2 suppressed the upregulation of pro-inflammatory factors and adhesion molecules in the retinal circulation, whereas downregulation of ACE2 was related to dysfunction of the profibrotic and pro-inflammatory capillaries, and retinal nerve fibre layer infarction. 12
Two cohort studies, retrospectively examining large data sets, have investigated the incidence of retinal vascular occlusions in patients diagnosed with COVID-19 but the results were contradictory. 17 ,18 For example, in one study the electronic medical records of 285,759 patients before COVID-19 and 156,427 patients during COVID-19 were reviewed and researchers found that during COVID-19 the proportion of patients with newly diagnosed retinal artery or vein occlusions remained stable. 13 By contrast, a study that examined medical records of 432,515 patients with COVID-19, found that the incidence of new cases of retinal vein occlusion at six months following COVID-19 was greater than six months before the infection. 14 However, the incidence of retinal artery occlusion did not change.
Generally, in the non-inflammatory state, retinal vascular occlusion is caused primarily by atherosclerosis, especially as a result of emboli originating from plaques in the carotid artery. 15 Therefore, treatments aimed at reperfusion of occluded retinal arteries or reversal of retinal cell death have been suggested. 16 A comprehensive literature review conducted by the American Heart Association in 2021, suggested that treatment with intravenous tissue plasminogen activator may be an effective treatment for central retinal artery occlusion. 17 Conservative approaches, including anterior chamber paracentesis, ocular massage, topical intraocular pressure-lowering agents, sublingual isosorbide dinitrate, have also been used in an effort to restore vision. 16 The putative rationale behind these therapies is that modulation of intraocular pressure or vasodilatation of the retinal vasculature may dislodge the obstruction and allow the embolus to migrate peripherally. However, most evidence of the effectiveness of these interventions is based on data from retrospective small case series. In fact, there are no effective evidence-based forms of therapy for this condition. Theoretically, the sooner the retina is reperfused following an acute retinal artery occlusion, the better the chance of improving visual function. 16 Consequently, if the patient fails to seek medical attention in a timely manner, they will not regain functional visual acuity in the affected eye. 17 This situation may have been particularly relevant for some patients with COVID-19 infection who may have been able to undergo ophthalmic evaluations due to the severity of their disease or infection control procedures.
Although the patient reported here had hypertension and atherosclerosis with plaques in bilateral lower limb arteries, carotid arteries, and abdominal aorta, his hypertension had been well controlled for the past five years by regular antihypertensive medication. His visual symptoms appeared three days after a negative RT-PCR test for COVID-19, but his blood levels of IL-6, CRP, ESR, and amyloid A, were significantly elevated, indicative of systemic inflammation. The increase in his blood levels of D-dimer, fibrinogen, and FDP indicated a procoagulant state, consistent with the pathological changes that occur following COVID-19 infection. Nine days after starting treatment with anti-inflammatory, anticoagulant and conservative therapy, the patient’s biomarkers of inflammation and coagulation related to mild COVID-19 symptoms returned to normal and his vision improved. However, although reperfusion of retinal capillary was observed in this patient, many visual field defects remained and were probably due to low oxygen saturation; the retina is highly sensitive to a reduction in oxygen tension.
In conclusion, we describe a patient with BRAO in his left eye that involved three arteries that we considered related to his mild COVID-19 infection due to the presence of high blood levels of cytokines and coagulation parameters. Following anti-inflammatory, antithrombotic and anticoagulant medication, his visual outcome significantly improved. While the causal relationship between BRAO and COVID-19 infection requires further investigation, we suggest that the onset of BRAO in this patient may have been secondary to mild or asymptomatic COVID-19 infection because of associated pathological events, including a cytokine storm, vascular dysfunction, and thrombosis.
Footnotes
Acknowledgements
The authors wish to thank the patient who generously agreed to participate in this report.
Declaration of conflicting interests
The authors declare there are no conflicts of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
