Re: Seah et al.: Assessing viral shedding and infectivity of tears in Coronavirus disease 2019 (COVID-19) patients (Ophthalmology. 2020;127:977-979)

To the Editor:

We would like to congratulate Seah et al¹ for the well-written article that evaluated 64 tear samples from 17 patients with coronavirus disease-2019 (COVID-19) and found that all tear samples tested negative for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), even for the 1 patient with ocular symptoms. They obtained a meaningful conclusion that the risk of ocular transmission of COVID-19 is low.

We applaud the authors for a major endeavor. However, we have different views on the role the ocular surface plays in COVID-19 transmission. Several potential limitations are worth discussing. First, the authors used a Schirmer strip to collect tears, which may not be reliable enough to test SARS-CoV-2. Although the Schirmer strip was previously validated in testing herpes simplex virus-1 from tears,² no evidence shows it also works for SARS-CoV-2. In recent studies in China,^{3 , 4} a conjunctival swab technique was used, and it obtained positive results. The authors only collected tears and may have missed the virus attached to the epithelium of the ocular surface or in conjunctival secretion, which may have led to incomplete results. Second, reverse transcriptase polymerase chain reaction may not be sensitive enough to detect small quantities of SARS-CoV-2 RNA. Therefore, negative test results may be false negatives and cannot exclude the presence of the virus. Multiple specimens are needed to increase sensitivity. Finally, even if the Schirmer strip and reverse transcriptase polymerase chain reaction tests were highly accurate, it still cannot be concluded that transmission through tears is likely to be low.

We think that 2 possible explanations can account for this. One is that the COVID-19 infection above the ocular surface may not express SARS-CoV-2 in the tears, or the concentration of SARS-CoV-2 may be low. The other is that the authors may have missed the window, because viral shedding in ocular tissue may only last for a short period. The authors did not mention the exact time of testing. Xia et al³ detected positive conjunctival swab samples at 3 days after the course of the disease, when the patient had no severe fever or respiratory symptoms. Owing to these limitations, the negative results must be interpreted with caution.

In all, we believe negative results do not conclusively show that the risk of ocular transmission for COVID-19 is low and even positive results cannot be understood as the risk is high. Until now, the issue has been controversial, but we suggest that the conjunctiva is another transmission route for COVID-19. In Deng’s study (Deng et al, 2020 Preprint, available from https://doi.org/10.1101/2020.03.13.990036), 2 rhesus macaques received 1 × 10⁶ 50% tissue-culture infectious doses of SARS-CoV-2 conjunctival inoculation, and the results showed the viral load was distributed in the whole body at 7 days after inoculation. However, this is a preprint article, and direct evidence for this conjecture is lacking. Nevertheless, we suggest that appropriate precautions are needed to prevent transmission through ocular tissues and secretions, especially for clinical staff. We hope convincing evidence from related animal experiments will be put forward soon.