The article by Araujo-Silva et al1 in this issue of JAMA Ophthalmology provides important evidence that SARS-CoV-2 can be detected in the retina of patients who have died of severe COVID-19 infection. There have been numerous reports describing ophthalmic disorders in patients infected with COVID-19 that are thought to be associated with the virus. These include conjunctivitis, keratoconjunctivitis, blepharitis, uveitis, retinal vascular occlusion, optic neuritis, and cranial nerve dysfunction. In some of these studies, COVID-19 has been identified in ocular tissues, particularly the conjunctiva, typically using reverse-transcription polymerase chain reaction testing.2 Some of these reports infer that COVID-19 is the putative causal agent of the ocular symptoms, based on a temporal association with infection. Retinal vascular occlusions might be explained by the propensity of COVID-19 to disrupt vascular endothelial cell function resulting in thromboembolic events, a common problem in patients with severe COVID-19. The true incidence of ocular disorders in patients severely ill with COVID-19 remains unknown, since many of these patients are too ill or die of their infections and the highly contagious nature of the virus may limit ophthalmic examinations.
Severe acute respiratory syndrome coronavirus 2 is a positive-stranded RNA virus with a capsid surrounded by an outer envelope. The capsid is formed by the nucleocapsid protein. The envelope contains the well-known spike glycoprotein, the major constituent of most current vaccines against SARS-CoV-2. Antibody generated to the spike glycoprotein in people who have been immunized blocks attachment of the virus and thereby prevents infection.
A virus closely related to SARS-CoV-2 is mouse hepatitis virus (MHV), a coronavirus that infects rodents. This virus was discovered more than 70 years ago and therefore has been extensively investigated, including detailed analyses by electron microscopy.3 The MHV studies provide an extra degree of rigor by which to compare the electron microscopy studies of SARS-CoV-2 described in the current report.1 The early MHV studies demonstrated that prototypical coronavirus particles were easily seen within cytoplasmic vacuoles of the infected cell. The number of viral particles within each vacuole varied from a few to many.
In another recent electron microscopic investigation of SARS-CoV-2, Caldas et al4 examined infected cell cultures (rather than human tissues) and found very similar examples of virus-filled cytoplasmic vacuoles, as previously seen in MHV-infected cells.3 Caldas et al4 also were able to closely examine the endoplasmic reticulum and Golgi bodies for sites of early virus assembly. They observed virus-budding sites on smooth vacuoles, as well as an accumulation of electron-dense material within the inner face of newly formed double-membraned vesicles. Finally, viral particles appeared to be carried within the secretory pathway to the outer cell membrane, where exocytosis occurred.
In the current article, Araujo-Silva et al1 performed a similar imaging analysis of enucleated human eyes from patients who had died of COVID-19. Retinal sections were processed for electron microscopy. The electron micrographs are illustrated in Figures 2 and 3.1 The resolution is slightly less than that seen in a prior study of infected cells,4 an expected result because of the difficulty of preparing human tissues, such as those from the eye, for examination by microscopy. Nevertheless, the appearance of approximately 10 particles with all the features of SARS-CoV-2 particles in Figure 3 in the study by Araujo-Silva et al1 is particularly convincing because of the very close resemblance to SARS-CoV-2 particles and MHV particles grown in cell culture.3,4 The current study identifies viral particles in both retinal tissue and retinal capillary endothelial cells.
The current study1 also examined both retinal and choroidal tissue with immunocytochemical analysis using a combination of fluorescent-tagged antibodies directed against various SARS-CoV-2 antigens. Using this technique, the authors1 identified viral antigens in multiple layers of the neurosensory retina, as well as the retinal pigment epithelium and choroid. The images provide convincing evidence that these antigens are present in the retina and choroid. However, neither the ultrastructural evaluation nor the immunofluorescent studies are indicative that the presence of the virus is associated with any local disease process. The authors1 did not examine the eyes with routine light microscopic techniques to screen for abnormalities in these tissues, such as the presence of inflammation. Only macroscopic examinations were performed.
One patient in the study1 was noted to have bilateral intraocular (subretinal and vitreous) hemorrhage prior to death. However, the methods of this study1 do not provide evidence that these hemorrhages were directly caused by the virus. Intraocular hemorrhage is not uncommon in patients who are critically ill and receiving systemic anticoagulation. The presence of viral particles in the capillary endothelium cannot be equated with endothelial dysfunction resulting in hemorrhage.
This article1 demonstrates convincing evidence of SARS-Co-V in multiple layers of the retina and the choroid using both immunofluorescent studies directed against specific COVID-19 antigens and transmission electron microscopy. This is an important finding that suggests that COVID-19 becomes disseminated throughout the body in patients with severe infections. How the virus enters retinal cells is not well understood. COVID-19 is known to bind to angiotensin-converting enzyme 2 receptors in other tissues, such as the lung. There is evidence5 that some retinal cells may express these receptors. There is also evidence that COVID-19 is neurotrophic and has the capacity to infect astrocytes and microglia in the central nervous system.6 This may predispose the retina to invasion by virus. Whether the presence of the virus results in localized disease in intraocular tissues in most cases remains unknown. The study by Arauyo-Silva et al1 is a starting point. However, further study is needed to answer this question as we begin to unravel the complex mechanisms of COVID-19 infection. We congratulate the authors on this important study.1
Corresponding Author: Nasreen A. Syed, MD, Department of Pathology, University of Iowa, 233 Medical Research Center, Iowa City, IA 52242-1182 (nasreen-syed@uiowa.edu).
Published Online: July 29, 2021. doi:10.1001/jamaophthalmol.2021.2806
Conflict of Interest Disclosures: Dr Grose is funded by National Institutes of Health grant AI 153817 outside the submitted work. Dr Syed reported receiving unrestricted funding to the Department of Ophthalmology and Visual Sciences, University of Iowa, from Research to Prevent Blindness.