Best Shot

People are looking at many potential therapies for Wuflu, as they should.   However, the chance that any particular existing drug will be effective against coronavirus is not very high.

I know one approach for which success is actually likely ( > 50% chance of success) : serum therapies, also known as  passive antibody therapy.  that approach has worked on a very wide spectrum of infectious diseases, such as pneumococcal pneumonia,  group A streptococci, hepatitis B, measles,  RSV,  and Ebola (far from a complete list).   It’s being tried right now, with encouraging results.



This entry was posted in Uncategorized. Bookmark the permalink.

56 Responses to Best Shot

  1. Magnus says:

    Here is a Dr. talking about his company doing exactly that, sounds promising. (Talk abou the therapy starts about midway in the interview).

    • minosp says:

      That dude is talking about monoclonal antibodies. Gregg’s talking about the antibodies (polyclonal) generated in an infected person.

      • 415 reasons says:

        The eventual monoclonal antibody products will very likely simulate convalescent antibodies mixtures. Regeron’s Ebola cure is a cocktail of three different monoclonal antibodies, and they are probably taking the same tack against COVID.

        One question to me is if we are so inept that we muddle through to the summer (when the regeneron cocktail will supposedly be ready for mass testing) with many many cases walking around, is there any chance we can manufacture enough of this stuff to change the outcome of the pandemic.

        Regeneron’s CEO talked about having 100,000 doses of the stuff like that was a lot. If it’s still spreading everywhere slowly in summer and nobody is able to, say, take a crowded subway to go to work and the serum is effective as a prophylactic, then we’re going to want far more than 100,000 doses. Fortunately that’s only off by 3 or 4 orders of magnitude.

        One last question is in a world where there’s enough antibody preparation to use widely as PREP, do people who get exposed to the virus eventually make their own antibody response? You’d think they might, since neutralized virus covered in the monoclonal antibody probably gets gobbled up by Fc receptor expressing macrophages and dendritic cells that should trigger an endogenous B cell response.

        If you had monoclonal antibodies by summer, AND you had a big lift to manufacture hundreds of times more doses than they are suggesting they will, you could see a path back to normalcy. While we wait for the vaccine, everyone gets a monthly dose of anti-COVID antibody cocktail. That would make it really stop spreading, people would stop getting sick, and waiting the 12 or 18 months for a typical vaccine testing cycle becomes a lot more feasible.

        • mapman says:

          Yeah, 100,000 dose injections is a lot! Just rough numbers: Need at least 0.01 mg/ml of blood concentration of neutralizing antibody. Let’s say an average human has 5,000 ml of blood. That’s 50 mg per injection – same order as ~ 100 mg for anti-cancer therapies. So 100,000 doses is 10 tons of antibodies (and 500 tons of vials containing them). Recombinant monoclonal antibodies are VERY expensive to make. Just about everything about the process is expensive AND it actually takes appreciable physical space. 100,000,000 doses is doable but would require massive scale up in manufacturing capacity for the entire supply pipeline – something that cannot be done overnight (and might bankrupt us all because it’s so bloody expensive)

          • 415 reasons says:

            I think you missed a decimal point somewhere. If one dose is 50 mg, then 1000 doses is 50 grams, and 100,000 doses is 5 kilograms. Did you mean 100 million doses would be more than a ton? Because that is correct.

            And, yes, it’s a heavy lift. Hopefully life will be tolerable without 100 million doses. But say at scale it costs $10/dose in bovine serum and glass vials. 100 million doses would cost… $1 billion dollars. Say I’m wrong on the cost by a factor of 10. It’s 1/200 of the cost of the stimulus bill they just passed.

            If it’s the right thing to do it’s the right thing to do.

            • mapman says:

              Yes. Thank you! Mg vs g, only three log units off. 10 kg vs 10 T 😦
              But the price and capacity points remain.

            • mapman says:

              BTW, $10/dose is never going to work. Assume maximal yields: 100 mg/L or 1 dose/L. 1 L of medium is $100 in “retail” today (fairly massive volumes already) – OK, assume 10X less for bulk operations! Then these antibodies will need to be purified (easily 10X more expensive because scale up is limited and the cost of materials and labor is very high). All in all, $100 per dose an absolute minimum. Monoclonals for injection these days go at > $1,000/dose, which probably contains 50% profit.

              • 415 reasons says:

                Yes I have no idea how much refining cell culture grade fetal bovine serum actually costs but I always assumed there was a huge markup since the raw material is waste from meat processing. But anyways say that it costs $100/dose, making 100 million doses would still “only” cost $10 billion, which is a drop in the bucket relative to the economic costs of this catastrophe. If having 100 million doses would really move the needle we should do it.

                Hopefully, by late summer when we would have such a monoclonal we will be in a setting of minimal disease with localized flare ups that get stomped down by non pharmaceutical interventions. At that point, having even a much smaller stockpile (5 million doses, perhaps) would be very useful, because you could administer post exposure prophylaxis to every traced contact of a confirmed case. If that worked well enough, perhaps it wouldn’t be necessary to lock a metro area down again when there was localized transmission chains.

  2. Gkai says:

    Reposting in the right topic: there are more and more mentions of plasma-transfusion therapy (plasma from recovered patient is injected into patients suffering from covid-19.
    Is it a one-to-one transfusion, one-to-many or many-to-one? I wonder because this would tell how scalable this would be, and how much potential for preferential treatment there is: many-to-one would make plasma rare and human-harvested, so if it works (it would be among the only real treatment available now – maybe the only one) I can see black market, corruption to get it first, and so on….another thing to consider : is this working better at early stage before life is threatened (hence a lot of people concerned, many for which treatment would help but be far from mandatory) or is as efficient when symptoms are bad, so could be used in intensive care?

    • 17c says:

      The only piece I’ve seen that mentioned that out of a New York study was that one recovered patient can provide enough plasma for three recipients. I don’t know enough about this to dispute that or conjecture they might be able to provide more later.

    • gkai says:

      Serum is the English word apparently, plasma is french…

    • 415 reasons says:

      It is one-to-a-few. I have seen quoted that a person typical donates about half a liter of plasma and that is good for 3-4 patients. So in a regime where R0 is ~0.95 it can be very helpful because there is enough for all the people who got it on the back half of the curve to get treated. If R0 is at all greater than 1 then it becomes little more than a stopgap for the most severe cases where people are on deaths door, because in a world of exponential spread the majority of cases are current cases and even at a 4:1 ratio there’s not enough old cases.

      Also actual convalescent plasma will never be abundant enough to use prophylactically in any widespread way, although it certainly seems like it would make sense for frontline health personnel.

      And finally they don’t quite know if the plasma works for severe patients yet. I think anecdotally it has activity. One interesting parallel is Ebola where remdesivir failed miserably but two different monoclonal antibody strategies basically cure the disease in all but the most desperate cases. To me it would not be unexpected for convalescent plasma and eventually monoclonal antibodies to be much more effective, and for antiviral drugs to need to be given quite early in the disease to see any marked activity.

      • Gkai says:

        So it could work as a large scale treatment… Interesting, as i do not see current Europe lockdown continue much longer. Mid may is a maximum imho, and not only for economical reasons. Amount of people feed up with the measures increase, at one time political support will vanish.
        Also curves hint more at a R0 slightly above one than below imho. Which is consistent with Wuhan after first set of measures. This means that we will probably reach saturation before vaccine, through a long flat peak

        • 415 reasons says:

          It can’t really work as a large scale treatment in the context of continued exponential growth. Maybe if R is just slightly higher than 1 then you can keep up with the growth because each person’s donation can treat multiple people, but I wouldn’t count on it. That also requires that every person who has recovered donates their plasma. Also I don’t know how many treatments are given (or would be given in an ideal world) to severely sick patients. Usually antibodies have a long half life in the bloodstream (on the order of weeks) so maybe only one. Even still, it’s hard to see this as a game changer. It may reduce morbidity and mortality at the margins but convalescent plasma by itself is unlikely to alter the overall trajectory of the epidemic.

          • Gkai says:

            If everybody who is sick needs medical treatment, and we were at the early stage of the exponential you are right. But it’s not the case, at least 50% of people who get it have no or mild symptoms and will not need medical attention. I think more probably 80%, with some of those not even realising they had wuflu. And we are not at the early stage, we are likely already at 5 to 25% percents here in Europe, depending on country… At this time, it’s more managing the peak /plateau than preventing the epidemy…
            i think that plasma /serum therapy is really interesting in this context, even with 1/4 th of people donating it would save many if not most of it’s really efficient. But we need serotological tests for this to be viable.

        • j says:

          After spending two weeks of semi-voluntary quarantine, I understand that the population will not tolerate this state of affairs for long. We already had cases of shooting policemen and soldiers enforcing the quarantine. Yes, this may be a long flu season.

  3. TerrifiedInSweden says:

    Seems like there’s a unique business opportunity possible here for us Swedes!

    Our “unique” approach probably means we’ll burn through our population much faster than almost everyone else we’ll have a big supply of plasma when everybody else needs it (and our needy population won’t need it due to being safely dead). And since maybe 1%-10% of our population dies in a short time span with huge repercussions for infrastructure, social order and so on (for instance, we’re nowhere near self sufficient in food), we’ll surely need to trade using something just to keep society going.

    Additionally our well deserved reputation for being suicidally naive, trusting and trustworthy will serve us well in the upcoming international blood markets. Certainly the blood of Swedes will be seen as far more likely to be actual human blood with actual anti-bodies as opposed to say Chinese blood (probably will be mostly dog tbh).

    Only downside with this is that our leaders lack even the business savvy of a person with pretty much no business savvy. I look forward to us shipping out 7 million liters of life saving blood plasma in return for food and getting a bunch of shipping containers filled with foam pellets in return.

    • R49 says:

      “maybe 1%-10% of our population dies in a short time” -> do you think it is going to be that bad? Won’t the country switch to at least partial lockdowns over the next days or week(s)?


      • TerrifiedInSweden says:

        I don’t know. But Swedish government is incompetent and incapable of acknowledging error or take responsibility. They’ve failed plenty of prior crises of much smaller size. We have no reserves of any kind (food, medicine, equipment or fuel). And we also have no ability to mass mobilize. All that stuff have been systematically and deliberately dismantled over the last couple of decades. Neither public nor private organizations have any slack at all, everything depends on imports working smoothly.

        We should be at something like 2-4% infected now and I doubt we even are capable of long shutdowns. Doubling rate is at best 1.5 weeks. Hospital care couldn’t keep up even before the pandemic. So I don’t really see any way to avoid infection saturation * fatality rate absent care. And that comes out to 1% in a conservative estimate.

        I hope I am wrong.

  4. j says:

    Don’t exaggerate. The world remembers Sweden’s generosity, and in case of famine, Ethiopian planes will be dropping sacks of sorghum on your cities.

  5. j says:

    The fight for plasma has started. Let me paste from the Israeli Arutz Sheva:

    The Health Ministry has put a halt to Hadassah Hospital’s collection of blood plasma that would be used in the treatment of coronavirus patients. Plasma has already been collected from two Israelis who contracted the virus on the Japanese “corona ship” and have now recovered.

    Director General of Hadassah Professor Zev Rotstein reported to the Prime Minister prior to the Passover holiday on setting up a series of tests for finding coronavirus antibodies in the blood plasma of recovered COVID-19 patients; this plasma would be infused into other patients to help them fight the disease. According to a report from Yisrael Hayom, the Health Ministry has stated that since plasma is “a national resource” the Ministry alone should be involved in its therapeutic use.

  6. Smithie says:

    There’s huge techo-futurist potential in the immune system.

    Ought to be a way to take out a few of the plasma cells from someone who had the disease, and sequence their DNA, and then print the sequence out and implant it into a culture of cells, perhaps, one day, being able to make differentiated adaptive immune cells and inject them into the body. Your future vaccine might be memory cells.

  7. Bob says:

    “A general principle of passive antibody therapy is that it is more effective when used for prophylaxis than for treatment of disease.”

    • Gkai says:

      If true it’s very bad news : prophylaxis will not be doable even for vulnerable people only. So if it’s really effective preventively, but not when severe symptoms are there already, it will be used by few – Black market, nepotism, political favors and so on….if there is time enough, cause i think in 2 months max, epidemy will mostly have run it’s course in my country anyway.

      • Regret says:

        With limited vaccines, or in this case serum, avalible you can put the brakes on things by targeting highly social people with lots of contact with others. In the context of this disease though, that might amount to ‘the people most aggressively violating lockdown orders’.

        Would be politically difficult to reward them with preferential treatment, I imagine.

      • 415 reasons says:

        Post exposure prophylaxis of contacts of confirmed cases will probably be a major modality of therapeutic intervention in a few months when we’re trying to get back to normal but don’t yet have a vaccine.

        That is to say, if you get a big dose of monoclonal antibody 3 days after you get exposed and 2 days before you get symptoms, you might still get sick but I bet your odds of winding up in the hospital are way lower, the course of your illness is much shorter, and you shed a much lower amount of live virus.

    • Craken says:

      Thanks for the link. The article presented more question than answers, but gave some general context of what to expect. The main unanswered questions:

      How many survivors have significant antibodies?
      How many can be induced to donate plasma?
      How effective is it prophylactically?
      How effective is it as a treatment at different stages of the disease?
      What is the most efficient use of the limited serum available given that larger doses may be needed as one moves from prophylactic to early infection to late infection?
      Does serum treatment prevent development of immunity?
      What it the donor/recipient ratio? Relatedly, how predictable is the course of this disease–can some be denied serum with very high confidence of a good outcome anyway?
      What other animals can produce useful serum and how soon?
      Are there significant risks to recipients of this serum?
      Would a mixture of antibodies from multiple survivors be more effective than antibodies from one survivor?

      • gcochran9 says:

        a. Probably almost all
        b. almost all, without a kind word.
        c. pretty effective on other infectious diseases
        d. pretty effective early, less so later. Can be combined with other drugs if we have any
        e. possibly people doing important stuff who are also highly exposed.
        f. Don’t think so.
        g. one to a few. perhaps?
        h. humanized domestic animals.. pigs?
        i. with human serum you must check for other possible pathogens.
        j. dunno.

        • gothamette says:

          This is such a good idea it will never be implemented in the US (on a large scale).

          We have somehow developed an immunity to good ideas in this country. Stupidity is our real virus, followed by sclerosis.

      • gothamette says:

        “How many can be induced to donate plasma?”

        There have been several interviews on local NYC tv of survivors saying they are desperate to donate blood if that will help. There is no way for them to do so. The system is overwhelmed – exactly what the “alarmists” warned of.

  8. saintonge235 says:

    I’m surprised so few people know of what used to be THE main infection fighter: horse serum. Inject the virus into horses, let them develop antibodies, take some of their blood, remove the red cells, transfuse.

    This really ought to be revived.

  9. kpkinsunnyphiladelphia says:

    Serum therapies are promising, but there are fairly substantial logistical hurdles, the main one being disease prevalence against a suitably sized donor pool and the time/expense/coordination required to executed the apheresis. Serum therapy works great from a logistical point of view for the occasional, indeed rare, poisonous spider and snake bites, but for 4000 cases in an urban hospital?

    Wanna get a sense of how non-simple it is? Read this — scroll down to the collection discussion:

    And the FDA is not being overly bureacratic here.

    The better solution is to develop a monoclonal antibody or a cocktail of them. You can make a ton of it easily and deliver it easily. And a properly designed one works.

    It’s one of the reasons, among others, that Humira is a multibillion dollar products.

  10. Hans says:

    Have you seen the cohort analysis on Remdesivir?

    It is encouraging. For the invasively intubated subgroup, only 18% had died at follow up, and 57% had been extubated (the remainder were still on vent). If the various reports from China, NYC, etc. are accurate that 80% or more die after being placed on a vent, this would seem to indicate extraordinary effect, particularly since anti-virals usually require early administration to impact outcomes.

    • R49 says:

      Thanks for the pointer and link. At least in the paper, they try to use very guarded wordings. Still, “clinical improvement was observed in 36 of 53 patients (68%)” sounds promising!


      • Анисимов Дмитрий says:

        “clinical improvement was observed in 36 of 53 patients (68%)”

        …was this a placebo-controlled study?
        …they would still write this same 68% improvement even if other 32% were dead.

    • nllssn says:

      I think hydorchloroquine is both antiviral and anti-inflammatory. That may be why it’s a good drug for COVID-19 (if indeed it is a good drug for COVID-19, I know Greg is skeptical).

      • luisman says:

        Some reports say it only really works with zinc, as zinc in the cell prohibits replication of the RNA. Zinc as a supplement may also be a good prophylaxis.

  11. Only slightly off topic, any thoughts on Paul Romer’s plan?

  12. John Massey says:

    Convalescent Plasma to Treat COVID-19: Chinese Strategy and Experiences.

  13. nllssn says:

    Does bacteria make antibodies? Can we infect a huge vat of bacteria with SARS-COV-2 and come back later and harvest the antibodies?

  14. Rob says:

    Could we immunize horses with a veterinary vaccine (lower regulatory burden) and then use horse serum in people? Are there enough horses left to make a difference? Is that too many moving parts?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s