Skip to main content

Thoughts on Antibody Persistence and the Pandemic

The fundamental challenge of dealing with this pandemic is uncertainty. There are just too many important things that we don’t know – and what’s more, they aren’t easy to find out, either. We’ve seen that with the waves of optimism and pessimism about various therapies, just to pick one. As everyone has been learning (if they didn’t know it before!), getting solid clinical data is not easy. The good numbers don’t come quickly and they don’t come cheaply, and data that are fast and cheap tend not to be any good.

That uncertainty extends deep into the biology of the disease. Let’s illustrate that with a look at a paper that’s come out recently that has shaken a lot of people up. It’s a valuable look at a controversial topic: people who have definitely been infected with the coronavirus, but who are asymptomatic. There are very important questions about this situation, among them how infectious such patients are and for how long, what might be different in their immune response as compared to people who become more overtly ill, and how the longer term effects on immunity might differ as well. And not least: how many such asymptomatic patients are there in general, and are they more common among some parts of the population than others? All of these questions are very important to our understanding of the pandemic and our responses to it, and none of them are as well-worked-out as we need them to be.

This group (from China) looked at 37 patients who tested positive (by RT-PCR) for coronavirus infection, but who had not displayed symptoms before testing (and continued to show none during subsequent monitoring), and compared them to the more well-studied symptomatic patients. One thing that stands out from a public health perspective is that the asymptomatic group had a much longer period of viral RNA shedding. That is not what one would have wanted to hear, since these are the people who are walking around without realizing that they’re sick at all.

But it’s also important to realize that this report, as with so many reports of viral shedding, it looking at it by detection of viral RNA. That does not mean that there is infectious virus present. (This distinction has confused a great number of reports about the persistence of the virus on surfaces as well). We still don’t know how long infected patients (symptomatic and asymptomatic) are shedding infectious virus – those are far harder data points to obtain than just detecting viral RNA. I should note that this goes for the recent reports, which were all over the media, about the possibility of spreading coronavirus through flushing toilets. Now, I think that closing the toilet lid is a pretty good precaution, because it’s an easy enough thing to do. But the spread of infectious virus through toilet aerosols has not been proven, to the best of my knowledge. You can find viral RNA in sewage, and monitoring it could be a useful public health measure in water-treatment plants. But infectious virus is something else again.

Now, when you hear of patients who are infected but are showing no symptoms, you could have a mental picture of a stronger immune response that’s keeping the virus knocked down more. But it’s clearly not that simple: this paper shows that the asymptomatic patients had a weaker antibody response that tended to disappear during the convalescent phase. That’s the result that’s gotten a lot of attention – as it should – although it’s important to not run with it too quickly. Remember, there is more than one type of immune response – you have T cells pitching in as well. That response is one of the great unanswered questions of the epidemic, from what I can see. We already have data about some of the T cell responses in symptomatic infected patients as compared to uninfected ones, and one of the most intriguing things about that earlier work was the finding that 40 to 60% of those uninfected people had CD4+ T cells that appear to already recognize the new coronavirus. The speculation is that this could be immune memory of past infections with other coronaviruses (such as the ones that can cause common-cold illnesses), and that they are cross-reacting with the current pathogen. (Update: from the comments, this work that raises the possibility that the T-cell response is from zoonotic coronavirus infections that we hadn’t appreciated). That’s very much in contrast to the antibody situation in uninfected patients – it seems clear that the human population had raised no circulating antibodies to the new coronavirus before the advent of this pathogen. Prior infection with other coronaviruses does nothing for you on the antibody side.

And that’s the missing piece from this new work. It is possible that these asymptomatic patients are showing a weak antibody response because they already had a more robust T-cell response ready to go. Now, this hypothesis could be wrong, but it has to be tested. In the same way, it’s possible that there is a percentage of the population that, because of their T-cell profile, are more likely to have only such an asymptomatic infection or perhaps to not become infected at all. That’s a very important thing to clear up, but it’s unfortunately a lot more labor-intensive to profile CD4+ and CD8+ T cells in people than it is to profile their antibody responses. This applies to the vaccines under development as well: T cell effects are an important component of vaccine efficacy, but there’s a lot that we don’t understand about how various vaccines raise such responses. There have been many reports of recovered patients who never seem to have raised much of an antibody response at all, and it’s not yet clear how they cleared the virus: was it the innate immune response that did the job before the adaptive system kicked in, particularly in younger patients? Was it the T-cell response instead? We just don’t know yet.

So my advice is not to panic, but not to be complacent, either. The complexities of the immune system mean that we have a whole range of possible situations in how this pandemic is unfolding. At the most optimistic end, it is possible that a larger percentage of the population than we realize might already be protected (to some degree) from the coronavirus. Unfortunately, it’s also possible that almost everyone is, in fact, still vulnerable and that we just haven’t seen the virus run through most of the population yet. Everyone will have seen the various population surveys with antibody testing that have suggested, in most cases, that a rather small percentage of people have been exposed. Think of the various ways you could get such a result: (1) it’s just what it looks like, and most people are unprotected because they have so far been unexposed. (2) the antibody results are what they look like – low exposure – but people’s T-cell responses mean that there are actually more people protected than we realize. (3) the antibody results are deceiving, because (as this latest paper seems to show) the antibody response fades over time, meaning that more people have been exposed than it looks like. And that means you can split that into (3a) the antibody response fades, but the T-cell response is still protective and (3b) the antibody response fades and so does the T-cell response. That last one is not a happy possibility.

You can extend this thinking to the effects of the coming vaccines. The protective effects of the antibody response versus the T-cell response are going to be important to figure out, of course. And keep in mind, the timelines that we’re seeing do not allow for time to see how quickly either of these responses might fade. So we could be in a situation where people would have to get re-vaccinated with a different agent after a few months when some of this comes into better focus. There’s no particular reason why the first workable vaccine to make it through trials should be the best long-term answer, but we also don’t know how much these things are going to vary, either.

So we need to take this new paper and its results about antibody persistence seriously, but we also need to keep our eye on the T-cell situation as well. We’re not going to know the whole picture until we know more about both of those. And in the end, we’re not going to know about the protective effects of recovery from a coronavirus infection, or the protective effects of being vaccinated against it, until enough time has passed in a large population of both cohorts. You’d want a faster, less difficult readout for those things, but I’m just not sure that there is one.

91 comments on “Thoughts on Antibody Persistence and the Pandemic”

  1. Derek, you’re missing a “not”:

    This group (from China) looked at 37 patients who tested positive (by RT-PCR) for coronavirus infection, but who had *NOT* displayed symptoms before testing (and continued to show none during subsequent monitoring)…

    1. Derek Lowe says:

      Jeez, I sure am. Gotta whack the forebrain into not doing that sort of thing.

      1. Sue Fics says:

        Please, whatever you do, don’t whack your forebrain! World needs you as you are!

        1. Miles says:

          I think the virus whacks the forebrain for you…seen the study about invasion through the nose and olefactory bulb into the forebrain…

          1. loupgarous says:

            Reverse axonal transport up the olfactory bulb has been studied off and on for decades. In his book on prions/slow encephalopathies Deadly Feasts, Richard Rhodes repeates an anecdote Carleton Gajdusek told about having terminal cancer patients volunteer to snort atropine to dry out their nasal passages, then inhale fine gold dust through their noses… on autopsy the gold dust was found to have traveled up through the olfactory bulb in those patients… so all sorts of things can travel that way, if they’re the right size.

      2. loupgarous says:

        Derek, I find a good, strong medium or dark roast breakfast blend, 2 cups, repeat prn helps me with typos.
        Then again, I grew up in the marshes outside New Orleans drinking that. It’s the only way I know…. 🙂
        But when I try to use HTML, all bets are off.

        1. Derek Lowe says:

          Might be my problem, because I don’t even drink coffee. My wife can’t figure out how I walk around on two legs without it.

          1. Md Saiful Islam Roney says:

            I don’t drink coffee, tea. I’m an Aquaman.

        2. still here says:

          There cannot be that many coonass chemists outside of the oil industry. I’m from Thibodaux.

          1. loupgarous says:

            Minored in chemistry at LSU, finished up in tech writing at La Tech. I’m from St. Charles Parish, not far from Thibodaux.

  2. Name * says:

    “This group (from China) looked at 37 patients who tested positive (by RT-PCR) for coronavirus infection, but who had displayed symptoms before testing”

    Should probably be “who had NOT displayed symptoms”

    1. David Young MD says:

      I noticed that too

  3. charlie says:

    Any thoughts on whether vaccines (mostly Chinese) using the original Chinese virus virus will work on the mutated European virus?

    Dumb question. They pulled these patients from people living with another confirmed case. Does it stand to reason that if I lived with a person who had an active infection, and they did a nasal swab — they would find viral material but in reality no infection?

    1. Patrick says:


      The virus has *barely* changed globally, so the idea that there’s a European version/strain and a Chinese one (Or an American one or … etc) is wrong. Our ability to track viral lineages is based around precise observation of minor variations, not meaningful differences between the virus across regions. It is highly likely that a vaccine developed anywhere would work anywhere else, because as yet there is no evidence the virus is meaningfully different.

  4. Barry says:

    Culturing nasal swabs on a lawn of human (HeLa?) cells in culture is slow and expensive, but it would distinguish the persistence of viral RNA (in neutralized virions) from an active infection.

    1. Kevin Law says:

      Damn! That Henrietta Lacks just keeps on giving and giving.

    2. Anon says:

      Yes, that’s basically how it’s done. Plate a bunch of easy-to-culture but infectible cells (Vero monkey cells are typically used, instead of HeLa), then add dilutions of your viral sample, and then count the number of “plaques” formed when an infectious virus kills a cell, spreads to its neighbors, and kills them too.

      In principle, any old cell culture lab could do this all day long, if nobody cared about biosafety. But since this kind of assay inherently produces a lot of highly concentrated and definitely infectious virus in the process, it has to be done using biosafety level 3 containment. And right now, all the BSL3 labs and personnel are already running well beyond their normal capacity, trying to research the coronavirus and discover potential treatments.

    3. loupgarous says:

      Why’d you want to bring HeLa into your lab if you didn’t absolutely have to? As I understand it, the precautions needed to keep HeLa from spreading into other cultures are part of why working with it is so expensive.

  5. Maria says:

    Of all the human coronaviruses. What are the T-cell and antibody response longevity time frames?
    In other words, for each human coronavirus, how long does immunity last?

    1. Adrian says:

      Note that “T-cell and antibody response” does not imply immunity.

      What I’ve seen on studies on human corinaviruses, infection gives you immunity for a year or less from infection with the same virus.

      1. daksya says:

        I posted this earlier but it’s relevant here:

        An interesting preprint in this regard is at

        From the paper:

        “For the management of the current pandemic and for vaccine development against SARS-CoV-2, it is important to understand if acquired immunity will be long-lasting. Therefore, we tested if individuals who recovered from SARS 17 years ago still harbor memory T cells against SARS-CoV-1. Hence, their PBMC (n=15) were stimulated directly ex vivo with peptide pools covering SARS-CoV-1 NP (NP-1 and NP-2), NSP7 and NSP13 (Figure 3A). This revealed that 17 years after infection, those individuals still possess virus-specific memory T cells, and similar to COVID-19 recovered patients, we detected T cells reacting almost exclusively to NP and not to the NSPs (Figure 3B/C). Subsequently, we tested if the NP-specific T cells detected in SARS recovered patients could cross-react with SARS-CoV-2 NP peptides (aa identity = 94%). Indeed, although at lower frequency, T cells in all 23 individuals tested reacted to SARS-CoV-2 NP (Figure 3D, 4A). In order to test whether these T cells could expand after encounter with SARS-CoV-2 NP, their PBMC were stimulated in vitro with the whole battery of NP, NSP7 and NSP-13 peptides and the quantity of T cells responding to SARS-CoV-2 NP, NSP7 and NSP13 was analyzed after 10 days of cell culture. A clear and robust expansion of NP-specific T cells was detected in 7 out of 8 individuals tested (Figure 3E). Importantly, and in sharp contrast to the T cell response to NP peptides, we could not detect any T cells reacting to the peptide pools covering NSP13 and only 1 out of 8 reacted to NSP7, despite in vitro expansion. Thus, SARS-CoV-2 NP-specific cross-reactive T cells are part of the T cell repertoire of individuals with a history of SARS-CoV-1 infection and are able to robustly expand after encounter with SARS-CoV-2 NP peptides. These findings demonstrate that virus-specific memory T cells induced by beta coronanvirus infection are long-lasting, which supports the notion that COVID-19 patients would develop long-term T cell immunity. Furthermore, our findings also raise the intriguing possibility that infection with related viruses can also protect from or modify the pathology caused by SARS-CoV-2 infection.”

  6. Elane Imgoven says:

    Until we have more widespread testing (both for viral presence, and for antibodies), we won’t really know what the infection rates are, will we. I have just enough background in statistics to be aware of the pitfalls in population sampling.

    Thanks as always for an illuminating explanation of stuff that this mining engineer would otherwise not grasp.


    1. Adrian says:

      In New York City around 20% of the population had antibodies in late April. 20% is large enough to avoid some of the statistical pitfalls.

      The 1% fatality rate back from Wuhan times is still a surprisingly good way to estimate past infections, multiply the number of deaths in a region with 100 to get a reasonable rough estimate of the cumulative number of infections 2-3 weeks ago.

      For countries like the US or Canada this gives you the estimate that around 3% (US) or 2% (Canada) of the population already had a COVID-19 infection. This has to be taken with a grain of salt, but any hopes for numbers already > 5% wouldn’t sound realistic to me.

      1. confused says:

        I definitely wouldn’t count on higher numbers (in terms of making any decisions based on the hope that more people are protected).

        But there does seem to be some evidence of protection of some form other than antibodies — e.g. the comparative rarity of serious (or even symptomatic, maybe?) infection in children.

        And extrapolating based on a 1% fatality rate would require that the age profile of the infected population is the same. COVID fatality rates vary so sharply with age that this can make a huge difference (see Singapore, which has an anomalously low CFR since nearly all infections are in a worker population, therefore younger).

        1. Adrian says:

          My “reasonable rough estimate” does not apply in populations that exclude the COVID-19 risk groups because everyone is under 50 and not obese, like in your example or on Navy ships.

          1. confused says:


            But it’s not quite that simple… I mean, there are people over 50 in Singapore. It’s a matter of who gets infected, not just who is present in the population. And lots of things could affect that; actual government measures as in Singapore, but also individual behavior and pre-existing demographics (population density, age of working population, % of single-person households, % of multigenerational households, etc. etc.)

            I wouldn’t expect to see the more-than-order-of-magnitude difference between Singapore and places like Lombardy/NYC very often, but even in more “normal” situations (say Texas versus New York) the difference could be large enough to make extrapolating based on a fixed IFR problematic.

  7. Lorry Turner says:

    Thanks DL.
    To add another dimension: should we worry about the possibility that 65+ vaccine recipients are less likely to have the capacity to mount a good response to a vaccine? Or that mounting a good response to the vaccine correlates with existing immune response (ie. those who would have combatted the infection fairly well anyway). Guess the vaccine could still be beneficial in preventing transmission from younger/healthier individuals to others, but then how would you convince those to take it beyond relying on civic duty?

  8. NM says:

    “We already have data about some of the T cell responses in symptomatic infected patients as compared to uninfected ones, and one of the most intriguing things about that earlier work was the finding that 40 to 60% of those uninfected people had CD4+ T cells that appear to already recognize the new coronavirus.”

    I’m not sure why this hasn’t gotten more press since its discovery. In theory this could change _everything_ about the way the disease has been modeled.

    1. nobody says:

      Cross-immunity from other coronaviruses also opens up the prospect of using minimally hazardous (e.g. common-cold level danger) coronaviruses as a stopgap vaccine.

      1. Adrian says:

        This is old news, and no one is assuming that there would be immunity against COVID-19 after infection with a different coronavirus.

        The idea has been around for months that recent infection with a different coronavirus might increase the chances that the infection is mild if catching COVID-19, but the effect seems small enough that this is hard to prove (or disprove) and barely relevant in practice.

  9. Philip says:

    At the most optimistic end, it is possible that a larger percentage of the population that we realize might already be protected (to some degree) from the coronavirus.

    Please delete.

    Thanks for the work that goes into this blog.

  10. Steve Scott says:

    Many concerns have been raised over a vaccine’s ability to protect the elderly with their less efficient immune systems. It has been suggested that some vaccines may not work well for 65+ and others might, eliciting a stronger immune response. But the best ones for seniors might not be the first to be released.

  11. Philip says:

    I think the timeline for any vaccine based on a modified active virus (SARS-COV2 or any other virus modified to present SARS-COV2 proteins) is going to run into big delays. I can see how these vaccines could offer short term protection and very little if any long term protection. Showing protection for a few months is very different than showing protection for 12.

  12. Lane Simonian says:

    To go progressively further out on a limb. Nuclear Factor-kappa Beta hyper-activation is likely critical for the overimmune responses to coronaviruses.

    By extension the use of effective NF-kB inhibitors should be explored with more urgency than currently seems to be the case.

    Most of those individuals with co-morbidities (diabetes, hypetension, Alzheimer’s disease, etc.) have high activation of NF-kB to begin with so they are more likely to be susceptible to the virus and to its damaging effects.

    NF-kB activation via oxidative stress and inflammation may inhibit T cell activation. This may explain in part why NF-kB inhibition also appears to limit viral loads.

    P.S. my caption test was two + 2=
    Wish it were that easy.

  13. Simon Auclair the Great and Terrible says:

    Bali is still having great resistance and no one knows why.

    1. Big Mystery says:

      The secret is to count covid cases as dengue cases.

    2. Adrian says:

      They might just have had few initial cases when they went into lockdown.

      In Europe most countries went into lockdown at the same time, and the different numbers of infections and deaths in these countries can be largely explained by the number of cases at that point (in Italy COVID-19 has spread for months, in other countries there were very few cases).

  14. luysii says:

    “40 to 60% of those uninfected people had CD4+ T cells that appear to already recognize the new coronavirus. The speculation is that this could be immune memory of past infections with other coronaviruses (such as the ones that can cause common-cold illnesses), and that they are cross-reacting with the current pathogen”.

    Well this supports an idea of mine of a month ago. Stop the pandemic in its tracks by giving everyone in the world an infection with the 4 coronaviruses known to cause the common cold. People reading that post were worried about something called immune enhancement as seen with Dengue, but I’m not.

    For details please see

    1. Andrew P says:

      Do you really think the 4 known common cold coronaviruses are the only ones circulating? I can’t believe that. Lots of new viruses must have come from China over the years. As long as there isn’t a lot of death, nobody will notice. I mean, how often is the virus that causes a cold, cough, or sinus infection sequenced in a lab? Unless the person is serious enough to be hospitalized, the probability is next to nil.

      1. loupgarous says:

        Among the coronaviruses found in homes is feline coronavirus. From the wikipedia article:

        “Feline coronavirus is typically shed in feces by healthy cats and transmitted by the fecal-oral route to other cats. In environments with multiple cats, the transmission rate is much higher compared to single-cat environments. The virus is insignificant until mutations cause the virus to be transformed from FECV to FIPV. FIPV causes feline infectious peritonitis, for which treatment is generally symptomatic and palliative only. The drug GS-441524 shows promise as an antiviral treatment for FIP, but at the moment it’s only available on the black market and still requires further research.”

        There are also infectious coronaviruses affecting pigs and dogs in the same species Alphacoronavirus 1
        as feline coronavirus. While we think SARS_CoV2 was originally spread zoonotically from an animal host in China to humans, elsewhere the virus jumps from humans to pet cats and other animals (including mink herds in the Netherlands).

        No news on whether feline or canine coronaviruses spread zoonotically to humans, even asymptomatically. Some researchers have found antibodies to coronaviruses which are not the four known to cause “common colds”. More work’s needed to see what coronavirus(es), exactly, those antibodies are associated with.

  15. Daniel Barkalow says:

    There’s also the possibility that they saw longer viral shedding just due to starting the clock sooner. From the cases that became symptomatic while being studied (and were then excluded), we can presume that some people in this population entered the study up to 17d before they would have started showing symptoms if they were going to, while studies of symptomatic patients don’t enroll people 17d before they are symptomatic patients. I don’t see anything to suggest that they tried to correct for this (or how you would, aside from getting a big population of infected people, studying them all, and splitting them into symptomatic and asymptomatic groups to compare retrospectively).

  16. BigSky says:

    The exclusion of IgA and emphasis (and this reported subsequent “loss over time”) in blood prick IgG really doesn’t add much to the field does it???
    The unusual focus on a blood Ig that is incidental to a respiratory infection to the exclusion of a respiratory Ig integral to the entire dance points to the importance/ease of blood vs. sputum recovery and reagents to report IgG vs IgA measurement.
    Has anyone seen good data on the IgA levels in various patients and infections over time ’cause that would be interesting to read.

    1. Barry says:

      THIS ^

    2. RA says:

      Just saw an interesting pre-print that addresses IgA!

      “In asymptomatic patients, we show an early burst of IgA that may rapidly and effectively eliminate the virus in the respiratory mucosa and prevent a full adaptive immune reaction. Accordingly, IgG is produced in low amounts and becomes undetectable after 6-8 weeks. An intermediate response is observed in patients with mild disease without the early IgA peak and characterized by a late, but modest IgG and IgA production, suggesting a minor adaptive immune response. Further studies are necessary to establish whether specific memory persist and for how long after asymptomatic and mild disease. “

      1. eub says:

        That’s very interesting!

        Not to be alarmist, but before taking a vaccine that raising IgG and not IgA, I’d very much like to know what the course is of SARS-CoV-2 in that immunological situation. Hopefully a cold with no systemic involvement.

        1. RA says:

          Yes, I agree. My guess is that rasing IgG without IgA will lead to most being protected from severe invasive respiratory disease, but not mild infections….but if those mild respiratory infections flare, will you still see the weird, likely post-infectious inflammatory sequelae in a subset of patients…i.e. MIS-C in kids? Hopefully raising IgG will protect from that, but I don’t know if we understand enough to say that. And from a broader population health point of view, it seems like the current vaccine approaches will allow the vaccinated to spread the virus. Imagine telling people who get the vaccine that they still need to wear masks and socially distance to protect others…letdown of the century! Is anybody looking at any inhaled or other approaches to raise IgA or innate defenses in the respiratory mucosa? This is the “frontline” of viral entry into the body and it seems like there is little attention to strategies to fortify it!

          1. Barry says:

            TB is not covid 19, of course. But we know that in case, the BCG vaccine elicits IgG without much IgA and although it protects from miliary TB( infection be the same organism of non-lung tissues) it fails utterly to protect from infection on the lung surface (the common infection)
            By the time we have an effective covid 19 vaccine, we may know a lot more about how to make an effective TB vaccine

  17. nobody says:

    (danger: layman speculation below)

    The recent Italian retrospective wastewater PCR testing results point to the emergence of SARS-CoV-2 in parts of Italy in mid-December. Given how quickly SARS-CoV-2 spreads in the population, it’s very hard to square a December introduction with the very low levels of antibodies detected in population-level sampling in Italy. It seems to me that either our understanding of R0 is vastly wrong, or antibodies fade so quickly that many historical infections cannot be detected by antibody tests and we have no idea how many people have been infected.

    I have no explanation for why it took several months for the Italian outbreak to go from levels of population infection large enough to be detectable in wastewater to causing a spike in excess mortality. There’s no obvious way to reconcile these data points.

    1. Adrian says:

      If unmitigated R0 in Italy was at or slightly below 2 the data would match nicely.
      Such an R0 value would be within the range of previous estimates.

      1. Marko says:

        Exactly. The sky-high ROs early on were mainly a function of testing coming on board to finally detect transmission that had already been ongoing for weeks or months.

        Additionally , apparent high ROs are largely driven by cluster events ( nursing homes , meat packing plants , etc. ) , and early transmission in many countries could have been largely non-cluster in nature , for example if the first people infected were at some international event and then carried the virus to scattered places around the globe , only later to find their way to other cluster hotspots.

        We don’t even know for sure that patient zero was in China , much less Wuhan , and we seem to have constructed roadblocks to discover the answer to this important question :

  18. Marko says:

    Do we know whether antibody-dependent cytotoxicity requires COV2-specific anti-spike/RBD antibodies? If not, could this be the basis of protection against serious COV2 infection and disease provided by antibodies to other coronaviruses?

  19. Valdis Andersons says:

    According to this pre-print, the cross reactivity is not from other HCoVs but from some other zoonotic CoV(s):

    If I understood the paper right, then they basically went T cell hunting and found some. Then they tried to map them back to the pathogens they’re for and came up empty when comparing with the 4 Common Cold ones. They did not investigate what the animal CoVs could be, just noted that there could be a number of them that match the T cell targets.

    Another intersting bit was that the pre-existing T cells (from 2 years ago) had a potentially better target recognition against SARS-cov-2 than the actual T cells from recovered SARS-cov-2 patients.

    1. Derek Lowe says:

      Now, that’s interesting. Just added a link inside the body of the post – thanks!

      1. Jeffrey L Bell says:

        Shades of CowPox.

  20. JasonP says:

    >>>You can find viral RNA in sewage, and monitoring it could be a useful public health measure in water-treatment plants. <<<


    Early on, my county did this, bought testing of the sewage to see if the virus was more widespread then the meager testing showed. But now the state is on a push to TEST, TEST, TEST and if they use all of the kits they bought they will have tested 25% of the population!

    Too many scientific geeks around here, everything seems to be data driven.

  21. matt says:

    Don’t know about spreading infectious virus, but the spread of bacterial pathogens in infectious form to all the surfaces in a bathroom by aerosol is well demonstrated. Not DNA/RNA, infectious material to plates which showed colony growth. Not only that, but for pathogens that managed to form a biofilm in the toilet, multiple flushes did not eliminate the tendency to spread infectious particles.

    Proving that that particular method was what spread an actual infection from human to human is much, much harder, but the possibility is certainly there.

    1. Barry says:

      The oral-fecal route is critical for a bunch of diseases (polio, cholera, typhoid…) Sanitary sewers, potable water, and hand-washing remain crucial for the Public Health. But in the case of Covid19, infection seems to be overwhelmingly air-borne respiratory-tract to respiratory tract, unmediated.

  22. FlushingMeadows says:

    Re. spreading coronavirus through toilet flushing and subsequent aerosol dispersion of infected fecal particles. This was shown for the original flavour SARS CoV at the Amoy Gardens apartment complex in Hong Kong:

    An unusual plumbing configuration was the culprit, but it’s definitely possible …

  23. RA says:

    Thanks for an interesting post. I do wonder whether the concern about these weak circulating IgG responses in asymptomatics is as bad as some are speculating and this is more an issue of inadequate measurement of the immune response. If they remained asymptomatic the entire time, their immune system collectively (i.e. innate, t-cell, IgA, IgM, IgG, etc) must be doing something right…it’s just not doing something right that we are measuring when we focus almost exclusively on circulating IgG! If it were shown that these asymptomatic positives with declining IgG levels got re-exposed and then became symptomatic or seriously ill, then I would worry! The abstract says “These data suggest that asymptomatic individuals had a weaker immune response to SARS-CoV-2 infection.” Um, how can you knock their “immune response” as “weak” when they stayed asymptomatic! Would we not all want that immune response? Is IgG level synonymous with the entire immune response? Don’t forget to actually look at the patient, not just one of their test results!!!

    The prevailing paradigm is that a robust IgG response is “good,” the marker of a strong, protected immune response to COVID-19. But maybe in some people, the other parts of their immune systems make a big, lasting IgG response unnecessary….and that is not a bad thing if they stay asymptomatic!!!!

    On the flip side, I wonder if in some patients, especially those who get severely ill, a robust IgG response should not be knee-jerk celebrated as “good” but rather viewed in context as the result of the failure of the non-IgG immune response that made all that IgG production necessary. After all, the severe forms of the disease are associated with an out-of-control immune response!!! “I went to the ICU and all I got were these lousy (yet Robust!) antibodies!”

    I wonder if the fact that vaccines are not producing sterilizing immunity in monkeys goes along with view. Maybe an intramuscular vaccine will produce strong IgG responses that help prevent severe invasive disease…but won’t stop recipients from getting a mild respiratory tract infection, perhaps asymptomatic. So, just like a mild resp infection in someone with strong non-igG immune response produces little IgG, producing strong IgG through IM vaccination will do little to stop those mild resp infections. As some have noted, perhaps you need a way to increase mucosal IgA (something inhaled) in order to sterilize. But since that is not going to happen, we could have a situation where people who get vaccinated are mostly protected from severe disease…but those who refuse vaccines or don’t mount an adequate response to them might face continued risk because even the vaccinated and partially protected will be able to spread the virus to them…especially if they shed as long as those in this study!

  24. Mason says:

    It seems like prisons such as Marion and Pickaway in Ohio would be ideal research populations since most of the inmates there tested positive in April. Why didn’t all test positive? How many have antibodies? What are the true asymptomatic and fatality rates?
    Have researchers been able to study these groups?

    1. Barry says:

      Inmates as research populations are ethically problematic, because consent is problematic; they are powerless.

      1. loupgarous says:

        That would be the case for research in which various treatments would be studied and compared, but mere compilation of biostatistics surrounding COVID-19 in the prison setting, in which inmates’ risk of morbidity or mortality didn’t vary whether they took part in the research or not… it’s hard to see an ethical conundrum there.

        All you’d be doing is tabulating prisoners’ previous experience with COVID-19, and existing antibody titers, not varying their treatment or other risk factors for the disease.

        1. Barry says:

          It’s my understanding that even a blood-draw constitutes a “search” and requires consent or a warrant

          1. loupgarous says:

            I assume that any reputable CRO or researcher would obey the Helsinki Declarations on biomedical research and obtain an informed consent from all subjects, incarcerated or not.

          2. Aidan chappuis says:

            I have a con-law background, not a bioethics one, but isn’t the definition of a search there (including blood draws) for fourth amendment purposes? I.E., police need a warrant for a blood draw for say a DUI. I didn’t know/don’t think the fourth amendment is that relevant to prison bioethics, and it doesn’t really apply at all in prisons anyway.

            Not commenting on the (lack of) ethics of U.S. prisons, just the law

        2. Charles H. says:

          Another problem is you’d be depending on the prison medical system, and in the US at least prisons have a very bad reputation when it comes to correctly identifying medical problems.

  25. White science says:

    That’s all very interesting but I would ask an expert idiot. There is a consensus here, mainly carved by the likes of cravatt at scripps.

  26. idiotraptor says:

    To date, discussion about the role of innate and humoral immune mechanisms have dominated the focus of primary anti-SARS-CoV2 immunity to the near exclusion of adaptive T -cell mediated responses. It’s been twenty-five years since I was actively involved in research studying anti-viral immunity to orthomyxo/paromyxovirus infection in mice and people. The experimentally-based paradigm for these and most other viral agents was that virus clearance and resolution of primary infection was principally T-cell mediated. Generation of virus-specific T cells always precedes production of protective antibody. Understanding the immune response to and the epidemiology of SARS-CoV2 in humans is complex and evolving. In vitro cellular immune assay and reagent development as well as patient lymphocyte procurement can require more time and technical complexity than development of antibody-based screens and assays. Still I find it curious that until recently comparatively little discussion of SARS-Cov2 specific T cell responses in clearing primary infection has been apparent. Others better read than myself please feel free to comment.

    Derek and readers, keep up the thoughtful and informative work.

  27. Lane Simonian says:

    Here is a simplified hypothesis: the coronavirus in severe cases disables t cells while overactivating other parts of the immune system (a cytokine storm).

  28. Daren Austin says:

    When you want perspective, you need to step back. Only 10 subjects, but TWENTY YEARS of coronavirus infection monitoring says that immunity is unlikely and reinfection occurs about every 30 months.

    1. Darby says:

      I’m confused by that study – they seem to have defined “reinfection” by a rise in antibody levels, plus something about influenza-like symptoms (including fever, which is rare in colds) in some. Wouldn’t a re-exposure cause a rise in circulating antibodies?

      1. Marko says:

        They measured virus-specific antibodies for each strain :

        “…We used a relatively small part of the N protein as the antigen in our tests, to most specifically identify infections by each virus individually. “

  29. Lane Simonian says:

    One more critical finding:

    “Additionally, upon discovering that COVID-19 patients had significantly higher concentrations of cytokines, The authors suggested that rather than directly attacking T cells, the COVID-19 virus triggers a cytokine storm, which then acts to drive down T cell numbers.”

  30. TallDave says:

    good point on uncertainty… remember the imminent ventilator crisis?

    outcomes highly contingent… effective treatments may (or may not) reduce vaccine urgency

    if we’re fortunate, by the time vaccines are available COVID death rates may be so close to nil that the major challenge is persuading enough to vaccinate

  31. catspyjamas says:

    Nice article, thank you . Given the polyspecifity of T cells why are you assuming that prior exposure to coronas is what’s needed. Maybe just a robust and diverse T cell repertoire does the trick. And would explain why the younger folks are generally more resistant

  32. confused says:

    Does anyone know whether this preprint makes sense?

    It’s a tiny sample size, of course, but if valid, it would at least be an “existence proof” (that people can be infected, clear the virus, and still test negative for antibodies).

    1. Marko says:

      I don’t think they proved the point. None of the “contact” patients (i.e. the group with the patients that presumably cleared the virus sans antibody response) tested positive by PCR. Only positive by “symptoms”.

      I’m not so sure that their T-cell assays adequately ruled out non-COV2 coronavirus reactivity , either.

      That said , I think it’s possible , if not likely , that some people clear a COV2 exposure before the development of a COV2-specific antibody response could be detected. To the extent that antibody is required , I think preexisting cross-reactive antibodies might do the job. I’d still expect the immune system to generate some level of COV2-specific antibody capability , however low.

  33. Jon says:

    A simple enough question for someone out there. Is the virus test specific for SARS-CoV-2, or does it return a positive result for any coronavirus infection? Presumably the antibody tests are also selective for the novel coronavirus?

  34. idiotraptor says:

    Presently there are multiple antibody tests available and being used as allowed under FDA EUA. I don’t know details, but I suspect they differ in sensitivity (limit of quantification) and possible their antigenic specificity as well.

  35. Erik Dienemann says:

    Just wanted to say thanks for an incredibly informative article.

  36. Erik Dienemann says:

    In case folks didn’t see this today – a preprint of an NIH paper entitled, “Serologic cross-reactivity of SARS-CoV-2 with endemic and seasonal Betacoronaviruses.” As far as I can tell (not an expert in this by any stretch) this is cross-reactivity of antibodies, not the T-cell cross reactivity most have been discussing.

    1. Daren Austin says:

      It does corroborate the earlier preprint I linked to. They found two instances of cross-reactive antibodies dating from testing pre-2000. I don’t think this should be viewed as surprising. Of course, does low antibody titre and cross reactivity generate any protection – or antibody dependent enhancement of infection? So much unknown.

  37. Boris says:

    What I am about to share is personal experience, it’s not a scientific study, but I think sometimes eye-tests should also be considered.

    Me and a student of mine both had COVID-19 back in the middle of March. It is most likely that I infected him back then, even though we had very minimal contact in a closed space on the day when I started developing very mild symptoms. He tested negative around April 1st. I had more severe symptoms, and was in hospital as a precaution for 16 days. He is a teenager and only had mild symptoms – fever and cough for a couple of days – much like a common cold. He donated blood plasma two weeks ago, so his level of antibodies was still high. A week ago he was in close contact for a couple of days with a person who turned out to be positive for the virus (The person had the virus for sure when they met). My student has not had any symptoms, he has been in close contact with many people since then, and none of them are feeling unwell either.

    I know there is an incubation period, or that this person may not have been contagious at the time they met, but they did have more contact than when I possibly infected him a few months ago.

  38. Joshua Roopchand says:

    I think that it may be also a robust TLR (Toll like receptors) response which is clearing the virus without the asyptomatic patients serocoverting. T cell responses are also a possibbles infcetionsbut that may be delayed because thaere may not be as many memory T cells from previous coronaviru . Don’t forget thet TLR3, TLR 7 and I believe TLR 9 can shut down protein synthesis in virus infected cells as soon as the viral RNA is detected and the type 1 IFN system gets activated. Long story short, TLRs clear the virus, B cells dont get activated ano seroconversion after virus infection. Innate immunity wins again. There also could be RNA silencing as well.

  39. Marko says:

    Antibody Responses to SARS-CoV-2 at 8 Weeks Postinfection in Asymptomatic Patients

    Abstract : “We compared levels of severe acute respiratory syndrome coronavirus 2 neutralizing antibodies in recovery plasma from 7 completely asymptomatic coronavirus disease patients with those in symptomatic patients in South Korea. We found that serologic diagnostic testing was positive for 71% (5/7) of completely asymptomatic patients, but neutralizing antibody response occurred in all 7 patients.”

    The reported absence of antibody production in some patients may be due to time-of-sample issues or assay type ( i.e. conventional serology vs neutralization ).From the same report:

    “… The difference in results from our study compared with the previous study might be caused by differences in the timing of the test. In the previous study, (neutralizing) antibody tests were performed 2–3 weeks after symptom onset, whereas we tested 2 months after symptom onset or laboratory diagnosis. Seroconversion in asymptomatic patients might take longer.”

  40. Erik Dienemann says:

    New preprint with some more tantalizing T-cell cross-reactivity findings in unexposed donors, with this bit being especially intriguing with regard to “heterologous immunity” (exposure to virus A providing some native immunity to virus B). Really need to figure this stuff out.

    “In our study we identified and characterized the exact T-cell epitopes that govern SARS-CoV-2 cross-reactivity and proved similarity to human common cold coronaviruses regarding individual peptide sequences, physiochemical and HLA-binding properties30,31. Notably, we detected SARS-CoV-2 cross-reactive T cells in 81% of unexposed individuals. To determine if these T-cells indeed mediate heterologous immunity and whether this explains the relatively small proportion of severely ill or, even in general, infected patients during this pandemic32,33, a dedicated study using e.g. a matched case control, or retrospective cohort design applying our cross-reactive SARS-CoV-2 T-cell epitopes would be required.”*

    This related “viewpoint” article in the Journal of Clinical Investigation on innate immunity and priming it with various substrates was also interesting.

  41. Angie says:

    Hi Derek! You have a very calming, yet analytical and objective voice. I enjoyed reading this a lot. I’m a student and have been following the latest articles on the pandemic – so, understandably I’ve had difficulty finding solid ground. I’m never sure what to believe. I really appreciate this article because of the assurances (but not certainties) that you provide, while educating the reader about all possibilities. I finished reading with a “well if he’s not worried, I’m not worried” point of view, while still taking your advice of not being complacent either. Thanks for this! 🙂

  42. Riah says:

    Hi Derek
    Thanks for this very interesting info. Apologies I am coming in very late having just seen this blog. In response to your statement “That’s very much in contrast to the antibody situation in uninfected patients – it seems clear that the human population had raised no circulating antibodies to the new coronavirus before the advent of this pathogen. Prior infection with other coronaviruses does nothing for you on the antibody side” I have just dug up a pre-print which claims to find antibodies cross reacting with Cov2 (buried at top of pg 8): “Coronovirus Protective Immunity is Short Lasting”
    Would be interested to know what you make of this. Also what you think of this paper in general. Particularly re methodological rigour (which I’m not sure is great although this does not impact on the initial point above). Also, athough a 35 year time span and 2473 floow up months sounds impressive this only translates to less than 6 people, a very small sample and all male. I remember some other old studies showing that people do develop resistance to the same strain of cold coronoviruses which contradict this- will have to try and dig those out again -unless you have ready refs? Thanks

    1. Derek Lowe says:

      Thanks – I’ll have a look!

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.

This site uses Akismet to reduce spam. Learn how your comment data is processed.