On the report from Hong Kong of possible reinfection of COVID-19

[kadiot]

TL;DR Summary

The report says that they were able to sequence two different viruses, months apart from the same patient.

However, it is unclear if the patient got clinically sick both times, had mild or severe disease, had comorbid illnesses, was immunocompromised, or received immunosuppressive drugs.

We have good data that infection produces neutralizing antibodies and T-cell immunity in most patients. We do not know if the immunity lingers for life or a few months, just like in cold viruses. In addition, we do not know if a second infection is worse, milder or even subclinical.

What is the latest reasonable conclusion that we can draw from the available data about reinfection?

 

[atyy]

In the news reports I have seen, the second infection is thus far asymptomatic. If the case is representative (which remains to be seen), it may mean that although we may get a vaccine that reduces the severity of the disease, the vaccine may not be able to significantly reduce transmission.
https://www.scmp.com/news/hong-kong...51/hong-kongs-third-wave-losing-momentum-city

 

[Secan]

In the news reports I have seen, the second infection is thus far asymptomatic. If the case is representative (which remains to be seen), it may mean that although we may get a vaccine that reduces the severity of the disease, the vaccine may not be able to significantly reduce transmission.
https://www.scmp.com/news/hong-kong...51/hong-kongs-third-wave-losing-momentum-city

The common colds have 160 strains. What is the maximum strains possible for covid?

 

[kadiot]

In the news reports I have seen, the second infection is thus far asymptomatic. If the case is representative (which remains to be seen), it may mean that although we may get a vaccine that reduces the severity of the disease, the vaccine may not be able to significantly reduce transmission.
https://www.scmp.com/news/hong-kong...51/hong-kongs-third-wave-losing-momentum-city

Yes. The 2nd infection was asymptomatic, and there was a rise in the antibody titers after it had previously dropped off. What this sounds like is that the titers of antibodies naturally went down but went back up in response to reexposure. It was not fast enough to completely eliminate the virus, but it seemed to have neutralized it enough to prevent clinical disease. We need the viral load/Ct value to figure out whether it got to the level of becoming contagious, and how fast the virus was cleared. This can happen with other diseases as well and points to an effective immune response.

By the way, I've heard some anecdotal cases here in Singapore of re-infection which were worse than their (HK) first and eventually died. I just don't know the full details. I'm sure you know better because you are following the Covid-19 situation in Singapore.

 

[kadiot]

In the news reports I have seen, the second infection is thus far asymptomatic. If the case is representative (which remains to be seen), it may mean that although we may get a vaccine that reduces the severity of the disease, the vaccine may not be able to significantly reduce transmission.
https://www.scmp.com/news/hong-kong...51/hong-kongs-third-wave-losing-momentum-city

Yes. The 2nd infection was asymptomatic, and there was a rise in the antibody titers after it had previously dropped off. What this sounds like is that the titers of antibodies naturally went down but went back up in response to reexposure. It was not fast enough to completely eliminate the virus, but it seemed to have neutralized it enough to prevent clinical disease. We need the viral load/Ct value to figure out whether it got to the level of becoming contagious, and how fast the virus was cleared. This can happen with other diseases as well and points to an effective immune response.

By the way, I've heard some anecdotal cases here in Singapore of re-infection which were worse than their (HK) first and eventually died. I just don't know the full details. I'm sure you know better because you are following the Covid-19 situation in Singapore.

 

[DaveE]

Seriously? n=1. Reported in the press, not even a pre-print. We can do better people.

edit: we can conclude that we need more data.

 

[atyy]

By the way, I've heard some anecdotal cases here in Singapore of re-infection which were worse than their (HK) first and eventually died. I just don't know the full details. I'm sure you know better because you are following the Covid-19 situation in Singapore.

it is unclear whether to me (just as a reader of the news reports and government press releases) what happened. I hadn't heard the anecdote you mention, but it may refer to some cases reported in the news that had COVID-19 infection, were cleared and then died after that. Many of these deaths were classified as not being due to COVID-19 (ie. due instead to heart problems). COVID-19 may cause heart problems, so there is a possibility that some of these deaths were not correctly classified. However, at the statistical level, the error is thought not to impact the public health strategy because (i) the deaths assigned to heart problems were among a subpopulation, and the death rate due to heart problems in that subpopulation is roughly the same this year as last year, consistent with a negligible error in assigning the cause of death (ii) if we take say 50% of these deaths as misclassified, the death rate due to COVID-19 increases by about 0.5 per million.

 

[kadiot]

Seriously? n=1. Reported in the press, not even a pre-print. We can do better people.

edit: we can conclude that we need more data.

This NY Times article quotes one of the scientists saying there was no antibody response in the first infection, but I saw an IgG graph with some reponse. Really best to wait for the paper. Of note, NYT had access to the manuscript. The rest of us mortals have to wait for CID to spit it out. Yes we need as much data as we can analyze.

https://www.nytimes.com/2020/08/24/health/coronavirus-reinfection.html

 

[Ygggdrasil]

This NY Times article quotes one of the scientists saying there was no antibody response in the first infection, but I saw an IgG graph with some reponse. Really best to wait for the paper. Of note, NYT had access to the manuscript. The rest of us mortals have to wait for CID to spit it out. Yes we need as much data as we can analyze.

https://www.nytimes.com/2020/08/24/health/coronavirus-reinfection.html

I saw that a reporter on Twitter posted excerpts of the non-peer reviewed manuscript, though none of the figures with data are included:

 

[Ygggdrasil]

Here's my attempt to summarize what we know about immunity to COVID-19:

Evidence for immunity:

• Monkeys infected with SARS-CoV-2 could not be re-infected with the virus, suggesting that re-infection is a rare event (though what happens in monkeys may not always accurately reflect what happens in humans).
• Researchers have a case report suggesting that SARS-CoV-2 antibodies can protect against re-infection in humans. Researchers studied a fishing vessel where 104/122 individuals aboard the ship became infected and they had pre-departure and post-arrival antibody and RT-PCR testing results for the crew. Three crew members had neutralizing antibodies for the virus pre-departure, and none became infected or experienced symptoms during the outbreak. Obviously, this is a small n study, but does suggest that the presence of neutralizing antibodies can protect against infection.
• Although we cannot completely rule out the possibility, the mutation rate of the virus is slow enough that we should not expect that the virus will mutate to avoid immunity in the short term.

Evidence for re-infection:

• The aforementioned case report from Hong Kong shows that re-infection is possible. The main questions is, how common is re-infection?
• Studies of people who have recovered from SARS-CoV-2 suggest that, while many individuals have some neutralizing antibodies present in their blood, the levels are very low for ~ 1/3 of samples. If the presence of neutralizing antibodies protect against infection (as suggested above), this suggests that potentially 1/3 of infected individuals could be susceptible to re-infection. However, we don't yet know what levels of antibodies are necessary to prevent re-infection.
• Furthermore, studies on people who were infected by the similar SARS virus from the 2003 outbreak also suggests that levels of antibodies against the virus wane over the course of a few years (but again, the caveat is that we don't know how much antibody is necessary for immunity). This matches our experience with the four other endemic coronaviruses suggests that infection provides short term immunity that wanes over time. It is possible that immunity may wane more quickly for some in the population allowing short term re-infection.

An initial infection likely protects from severe disease later:

• The Hong Kong patient experienced much milder symptoms during the re-infection than during the first round of infection.
• Infection with SARS-CoV-2 produces a T-cell response, which along with antibody response, form the two major arms of the adaptive immune response. While neutralizing antibodies can directly bind to viral particles to prevent them from infecting cells, T-cells help to clear infected cells. Researchers see signs of memory T-cells that can recognize SARS-CoV-2 in many infected patients, even those who do not show signs of neutralizing antibodies.
• Researchers also see T-cells that react with SARS-CoV-2 in ~20-50% of samples taken from before the pandemic. These T-cells are likely from previous infection by the common cold-causing coronaviruses that are circulating in human populations. Researchers speculate that the presence of these cross-reactive T-cells could explain why some fraction of the population are asymptomatic or experience only very mild symptoms after SARS-CoV-2 infection (though there is not yet data to support this hypothesis).

From this evidence, I would hypothesize that re-infection is probably an uncommon event that could occur in individuals who do not produce a neutralizing antibody response after first infection or after antibody levels have waned over time. However, T-cell responses, which seem to be more common and longer lasting after infection, will likely protect against severe disease, so re-infection will likely cause mild or asymptomatic cases. However, if re-infected individuals remain infectious, this will cause issues with controlling the disease. It may be the case that SARS-CoV-2 stays with us forever continuing to circulate because some fraction of the population never develops neutralizing antibodies to give the sterilizing immunity needed to stop new infections. However, after enough people have been exposed to the virus (either through infection or vaccination), most people may have some amount of T-cell immunity to protect against severe symptoms, so it then the virus would just becomes like the other four endemic Coronavirus that mostly cause common cold-like symptoms.

STAT news also has a nice piece discussing some of the prospects for developing immunity to COVID-19: https://www.statnews.com/2020/08/25/four-scenarios-on-how-we-might-develop-immunity-to-covid-19/

 

[Tom.G]

Two more reports of re-infections, Belgium and the Netherlands.
https://www.businessinsider.com/2-n...on-cases-belgium-netherlands-hong-kong-2020-8

I saw a couple more short articles on it either yesterday or earlier today, but don't see them now. No real details in any of them.

 

[kadiot]

Two more reports of re-infections, Belgium and the Netherlands.
https://www.businessinsider.com/2-n...on-cases-belgium-netherlands-hong-kong-2020-8

I saw a couple more short articles on it either yesterday or earlier today, but don't see them now. No real details in any of them.

Yes. Even the very first case report in HK we have not yet seen the full circumtance. The case has been accepted by the journal Clinical Infectious Diseases but not yet published, meaning the data isn't yet available for full us to review.

 

[kadiot]

What is the difference between infection and disease? Those two are different concepts that can be confusing, IMO.

 

[jim mcnamara]

This thread is going downhill. There really is not enough information on what occurred, so we are starting to speculate. PF does not support speculation in the Science forums sections.

This report says that active clinically identified Covid infections can persist after the pneumonia phase:

https://www.sciencedirect.com/science/article/pii/S1201971220302794

NB: there are multiple patients in the case report.

So how do we see persistence versus reinfection? Answer: a good intervening test that shows no infection. Was there a test that proves this and with negative results? How do exclude we false negative results? They exist.

The worldwide number of cases per the JHU site is 23 MILLION + right now. With that many cases don't you think we are overreacting? We should have seen a lot of reinfection and some of it well documented by now, if it really was a clinical concern. One occurrence in 23 million, if correct, is not a worry. Yet. Let's wait before we decide.

 

[Rive]

We should have seen a lot of reinfection...

Well, I don't think so. The expected minimal time for the immunity to 'expire' is at least a few months: we need to seek for the reinfections among the numbers few months old. That's nowhere near close to that actual 23+ million.

Right now we have sporadic cases, but what we need is statistics. Long way to go.

Ps.: also, we need those reinfections (for statistics) among trackable cases: that means only developed countries, with adequate healthcare (and testing). I would be surprised if we could get real data before the end of autumn.

 

[atyy]

Well, I don't think so. The expected minimal time for the immunity to 'expire' is at least a few months: we need to seek for the reinfections among the numbers few months old. That's nowhere near close to that actual 23+ million.

Right now we have sporadic cases, but what we need is statistics. Long way to go.

Ps.: also, we need those reinfections (for statistics) among trackable cases: that means only developed countries, with adequate healthcare (and testing). I would be surprised if we could get real data before the end of autumn.

Also, if the reinfections are asymptomatic, they wouldn't be picked up except by screening - as was the case for the reported reinfection detected in Hong Kong.

 

[atyy]

So how do we see persistence versus reinfection? Answer: a good intervening test that shows no infection. Was there a test that proves this and with negative results? How do exclude we false negative results? They exist.

In the Hong Kong reinfection report they happened to have the full sequence for both putative infections, and the genetic sequence of the virus was different.

 

[jim mcnamara]

Can someone please cite a reasonable paper from a scientific source? Thanks.

 

[Ygggdrasil]

What is the difference between infection and disease? Those two are different concepts that can be confusing, IMO.

From the Mayo Clinic:

There's a difference between infection and disease. Infection, often the first step, occurs when bacteria, viruses or other microbes that cause disease enter your body and begin to multiply. Disease occurs when the cells in your body are damaged — as a result of the infection — and signs and symptoms of an illness appear.

https://www.mayoclinic.org/diseases-conditions/infectious-diseases/in-depth/germs/art-20045289

Infection is characterized by the presence of a foreign organism growing within us, whereas disease is characterized by symptoms. Infections can lead to disease, but this does not always need to be the case. For example, our guts are infected with many commensal bacteria that make up our gut flora, but these bacteria normally do not cause disease (though they can under certain circumstances, for example, if they escape the gut in immunocompromised people).

A good example of where we differentiate between infection and disease is in HIV/AIDS. HIV is the virus that causes the disease AIDS. If left untreated, the HIV infection will deplete the body of immune cells to cause the immunodeficiency disease known as AIDS. However, many people are infected with HIV but do not have AIDS because they are able to control the infection through anti-retroviral drugs.

In the context of coronavirus, SARS-CoV-2 is the virus and COVID-19 is the disease. We know that individuals can be infected with the SARS-CoV-2 without displaying the symptoms of COVID-19 (the asymptomatic cases). The hope would be that even if prior infection with the virus/vaccination is not able to prevent infection with SARS-CoV-2, that it would at least prevent progression of the infection to severe disease with life-threatening symptoms.

 

[Ygggdrasil]

Can someone please cite a reasonable paper from a scientific source? Thanks.

The news was initially released only through a press release from Hong Kong University, though it has been carried by reputable sources of scientific news, for example:
https://www.statnews.com/2020/08/24...tion-documented-in-hong-kong-researchers-say/
https://www.sciencemag.org/news/202...ic-virus-twice-study-suggests-no-reason-panic

[edit]The paper has now been published online in the journal, Clinical Infectious Diseases: https://academic.oup.com/cid/advance-article/doi/10.1093/cid/ciaa1275/5897019

COVID-19 re-infection by a phylogenetically distinct SARS-coronavirus-2 strain confirmed by whole genome sequencing
To et al. Clin Infect Dis, ciaa1275
Published 25 Aug 2020

Abstract:

Background
Waning immunity occurs in patients who have recovered from COVID-19. However, it remains unclear whether true re-infection occurs.

Methods
Whole genome sequencing was performed directly on respiratory specimens collected during two episodes of COVID-19 in a patient. Comparative genome analysis was conducted to differentiate re-infection from persistent viral shedding. Laboratory results, including RT-PCR Ct values and serum SARS-CoV-2 IgG, were analyzed.

Results
The second episode of asymptomatic infection occurred 142 days after the first symptomatic episode in an apparently immunocompetent patient. During the second episode, there was serological evidence of elevated C-reactive protein and SARS-CoV-2 IgG seroconversion. Viral genomes from first and second episodes belong to different clades/lineages. Compared to viral genomes in GISAID, the first virus genome has a stop codon at position 64 of orf8 leading to a truncation of 58 amino acids, and was phylogenetically closely related to strains collected in March/April 2020, while the second virus genome was closely related to strains collected in July/August 2020. Another 23 nucleotide and 13 amino acid differences located in 9 different proteins, including positions of B and T cell epitopes, were found between viruses from the first and second episodes.

Conclusions
Epidemiological, clinical, serological and genomic analyses confirmed that the patient had re-infection instead of persistent viral shedding from first infection. Our results suggest SARS-CoV-2 may continue to circulate among the human populations despite herd immunity due to natural infection or vaccination. Further studies of patients with re-infection will shed light on protective correlates important for vaccine design.

 

[Ygggdrasil]

Another case report of reinfection in a 25-year-old from Reno, NV was posted in a non-peer-reviewed pre-print from the medical journal, The Lancet:

Genomic Evidence for a Case of Reinfection with SARS-CoV-2
https://ssrn.com/abstract=3681489

Abstract:

The degree of protective immunity conferred by infection with SARS-CoV-2 is currently unknown. As such, the possibility of reinfection with this virus is not well understood. Herein, we describe the data from an investigation of two instances of SARS-CoV-2 infection in the same individual. Through nucleic acid sequence analysis, the viruses associated with each instance of infection were found to possesses a degree of genetic discordance that cannot be explained reasonably through short-term in vivo evolution. We conclude that it is possible for humans to become infected multiple times by SARS-CoV-2, but the generalizability of this finding is not known.

So how do we see persistence versus reinfection? Answer: a good intervening test that shows no infection. Was there a test that proves this and with negative results? How do exclude we false negative results? They exist.

This article addresses this in two ways. First, the patient had two negative tests for the virus between the two infections:

In April, 2020, a twenty-five year old resident of Reno, NV tested positive for SARS-CoV-2 through a community-based testing event held by the Washoe County Health District (collection date: 4/18/2020). The patient indicated symptoms consistent with viral infection (sore throat, cough, headache, nausea, diarrhea; onset: 3/25/20). During isolation, the patient indicated resolution of symptoms (4/27/20). The patient was subsequently tested by two nucleic acid amplification tests and was found negative for the presence of SARS-CoV-2 RNA for specimens collected on 5/9/2020 (by transcription-mediated amplification (TMA)) and again on 5/26/2020 (by real-time PCR (RT-PCR)). The patient continued to feel well until 5/28/20. On 5/31/20, the patient sought care with self-reported fevers, headache, dizziness, cough, nausea, and diarrhea. A chest x-ray was performed and he was discharged home. Five days later, on 6/5/20, the patient presented to a family care doctor and was found to be hypoxic and was instructed to go to the emergency department after provision of oxygen. The patient was hospitalized that day and was assessed for SARS-CoV-2 infection by RT-PCR testing. The patient required ongoing oxygen support and reported symptoms that included myalgia, cough and shortness of breath. A chest x-ray was performed on 6/5/20 and compared to that of 5/31/20 with the development of new patchy bilateral interstitial opacities suggestive of a viral or atypical pneumonia. RT-PCR results were positive for the presence of SARS-CoV-2. On 6/6/20, the patient was tested for IgG/IgM for SARS-CoV-2 and was positive.

and, second, sequencing of viruses isolated from the two infections indicate that they are distinct and not related to each other.

Worryingly, unlike the report from Hong Kong, the second infection was not asymptomatic and required hospitalization + oxygen support. The patient was not on any immunosuppresants or had any other condition that would be expected to more easily allow re-infection.

Of course, while these reports show that short term re-infection is possible, we still need statistics to determine the prevalence of re-infection. Given the number of infected people, it is likely that re-infection leading to sever disease (as in this case) is rare, though it is difficult to determine how prevalent re-infections that are asymptomatic would be.

 

[jim mcnamara]

MedCram discussion of reinfection.
Per NYT article:
The RNA analysis for both infections found the first and second infections to vary by 24 nucleotides. Watch the video. Pay particular attention the comments near the end by Michael Mina, an immunologist on why reinfection is not a black and white disaster/non-disaster thing. Generally it works out better for the patient.

 

[Laroxe]

The common colds have 160 strains. What is the maximum strains possible for covid?

"The Common Cold" really just describes a set of symptoms that can result from infection with lots of viruses from different families rather than strains of one particular pathogen. While some viruses can mutate fairly quickly, only a few of these mutations can change the nature of the infection. Covid 19 doesn't seem prone to frequent mutations with only one significant change identified and that is of questionable significance.
I'm not sure why a single case report of a reinfection is causing so much discussion really, it would be unusual not to see at least a few reinfections in most viral diseases and these can occur for all sorts of reasons.

 

[atyy]

Genomic Evidence for a Case of Reinfection with SARS-CoV-2
https://ssrn.com/abstract=3681489

The took care to ensure that their conclusions were supported by multiple lines of evidence (multiple negative swab tests, full sequencing), and also to avoid sequencing errors by using multiple methods and other precautions, eg. "all positions supported by fewer than four reads, whether reference or alternative, were replaced with Ns". But is there any possibility that some of the sequence differences were due to sequencing errors?

 

[256bits]

Following Hong Kong by a week, the first case of reinfection in the USA.
https://www.msn.com/en-ca/health/medical/lab-confirms-1st-coronavirus-reinfection-in-the-us/ar-BB18tWXp?OCID=ansmsnnews11
Pre-print, and not peer reviewed yet.
Second virus is a mutatation.

Hmm -9 months or so after the initial occurrences, and mutations?

 

[kadiot]

"I'm not sure why a single case report of a reinfection is causing so much discussion really, it would be unusual not to see at least a few reinfections in most viral diseases and these can occur for all sorts of reasons.

Because we do not know what implication this reinfection has for upcoming vaccines. We need to closely listen to what the science is telling us.

 

[Laroxe]

Well yes, but with Covid 19 its difficult to know if there are any implications beyond what we already know. Our immune response to this virus can be a bit confusing. When they were investigating possible cross immunity from other coronaviruses they identified around 40 antibodies that targeted different parts of the virus but the only one that had any impact on either preventing or modifying the illness was very specific to the spike protein. Of course we do produce a range of antibodies against most infective agents of varying degrees of effectiveness but I think its pretty unusual to require one very specific antibody and we already know that people vary in their response to natural infection.
All of the vaccines already target this protein, largely ignoring the rest of the viral coat and the theory at least is that the immunity produced will be far more specific than from natural infection. They also think that long lasting immunity will be highly dependent on cellular immunity T and B cells rather than the immediate circulating antibodies but these are more difficult to measure.
That's why I suggested that I didn't understand the excitement, I simply couldn't see what these cases would tell us when the disease doesn't appear to have listened to the science either. At least this time its done something predictable, but of course I will be interested to see what comes out of these cases, I'm certainly not dismissing the role of science.

 

[Ygggdrasil]

while these reports show that short term re-infection is possible, we still need statistics to determine the prevalence of re-infection. Given the number of infected people, it is likely that re-infection leading to sever disease (as in this case) is rare, though it is difficult to determine how prevalent re-infections that are asymptomatic would be.

We now have some statistics on the prevalence of re-infection by SARS-CoV-2. The SIREN study has been monitoring COVID-19 infections in hospital staff in the UK, identifying people who had previously been infected by the disease, then seeing how many would later become re-infected, and published some early results in a non-peer reviewed preprint on medRxiv:

Do antibody positive healthcare workers have lower SARS-CoV-2 infection rates than antibody negative healthcare workers? Large multi-centre prospective cohort study (the SIREN study), England: June to November 2020
https://www.medrxiv.org/content/10.1101/2021.01.13.21249642v1

Overall, they observed a 83% lower risk of infection among the previously infected group versus the control group who had not previously been infected, and the protection lasts at least five months.

Between 18 June and 09 November 2020, 44 reinfections (2 probable, 42 possible) were detected in the baseline positive cohort of 6,614 participants, collectively contributing 1,339,078 days of follow-up. This compares with 318 new PCR positive infections and 94 antibody seroconversions in the negative cohort of 14,173 participants, contributing 1,868,646 days of follow-up.

See also the press release from Public Health England: https://www.gov.uk/government/news/...but-people-may-still-carry-and-transmit-virus