SARS-CoV-2 reinfection: What does it mean?

Not only the popular press but even some of my distinguished science colleagues are claiming that the recent report of reinfection of a COVID-19 patient is the end of the world. Nothing could be farther from the truth.

First, the facts. The patient is a healthy Hong Kong male who was diagnosed by PCR with SARS-CoV-2 infection on 26 March 2020, after developing cough, fever, sore throat and headache. He was hospitalized for two weeks and discharged on 14 April after two consecutive negative SARS-CoV-2 PCR tests done 24 h apart.

Upon returning to Hong Kong from Spain on 15 August 2020, the patient was tested by PCR for SARS-CoV-2 and found to be positive. He was hospitalized but remained asymptomatic. Oropharyngeal viral loads gradually decreased during his hospital stay. Patient was IgG negative for SARS-CoV-2 nucleoprotein 10 days after symptom onset for the first infection, and 1 day after hospitalization for the second episode. However a serum specimen taken 5 days after the second hospitalization was positive. I conclude that the first IgG test was likely negative because it was too soon after infection; an IgM test should have been done. It is very likely that the patient eventually became IgG positive after the first infection. However, levels of IgG appear to have decreased substantially in the ensuing months, allowing the second infection. IgG positivity on day 5 after the second infection is a classic memory response.

Whole genome sequence analysis of viruses isolated during the patient’s first and second infections clearly revealed that he was infected with two different isolates.

I am convinced that this patient was infected twice by two different isolates of SARS-CoV-2. The second infection is likely a consequence of waning anti-viral IgG antibodies. However, patient T cells were not studied. Such an analysis would have been useful because if virus-specific T cells had been found, it would have suggested why he did not develop disease upon reinfection.

This re-infection is the first of which I am convinced; the others are anecdotal and not supported by laboratory evidence. That gives us one reinfection with SARS-CoV-2 in the nearly 25 million that have been detected so far. To be fair, there are likely other reinfections, and as time passes and antibody levels wane, there will be more. Should we worry?

As long as reinfections with SARS-CoV-2 do not cause disease, I am not worried. Could this patient transmit virus? He was clearly shedding virus the second time; whether it was enough to transmit is not known as PCR Ct values are not given in the manuscript. But if reinfections remain asymptomatic, it doesn’t matter if these infections transmit. Let’s say in the coming months, as antibodies to SARS-CoV-2 wane in the population, more and more reinfections occur. If none or few of them are symptomatic, why do we care about them? If SARS-CoV-2 continues to circulate and causes little disease, it would have little impact.

Reinfection with SARS-CoV-2 in the absence of symptoms is reminiscent of the four seasonal coronaviruses. These viruses circulate widely and infect nearly everyone. Antibody levels appear to wane after each infection. Reinfections occur, but they are generally mild.

As the human population approaches 90% infection with SARS-CoV-2, we will likely see a pattern of waning immunity and reinfection in the absence of symptoms. As I’ve said before, SARS-CoV-2 will become the fifth common cold coronavirus.

Some have suggested that this reinfection story bodes ill for SARS-CoV-2 vaccines. This conclusion is not necessarily true. First, we don’t yet know what kind of immunity the experimental vaccines will produce. There are a few possibilities. The vaccines might not work at all to prevent infection or disease. In this case we would have to depend on other approaches to limit mortality (Mina-style susceptibility testing, quarantine, perhaps antivirals). At the either extreme, some vaccines might prevent both infection and disease. I find this scenario highly unlikely, because not even natural infection can do that. Some of the vaccines in development might produce immunity approximating that of a natural infection – that is, it wanes soon after vaccination. Reinfection will likely occur, but in the absence of disease. The virus will continue to circulate, even in a population that has been widely vaccinated. This scenario will make the vaccine unnecessary at some point in the future.

The reinfected COVID-19 patient simply reinforces what we already knew about the pattern of infection caused by common cold coronaviruses: immunity wanes after infection, reinfection occurs, and there no disease.

Note added 9/1/20: A second reinfected patient has been identified in Nevada.