4 ms·

This may be a stupid question. But can the antibodies be synthesized based on samples from recovered patients? If not, why?

jliptzin 7y ago
Definitely not a stupid question
felipemnoa 7y ago
Following your train of thought, could the synthesis be done in the same way that insulin is currently created? i.e. Have some germs synthesize the antibodies.
- chispamed 7y ago
  Yes, that’s a possibility and currently many monoclonal antibodies are produced that way (basically all drugs ending in “-mab”). However the process is costly and current capacities are not sufficient to supply a whole population regularly with antibodies which is why we still use some older drugs for certain diseases (think MTX for rheumatoid arthritis) and reserve monoclonal antibodies for treatment-resistant cases. I hope that it’ll be possible to protect high risk HCW and ICU patients through antibodies and thereby bring down Covid’s mortality soonish but we still need to buy some time until then and work on a vaccine to protect the population long-term. Right now what we need to do is stay at home and invest in medical care and drug development but with emphasis on the staying at home part because it’s the necessary condition for everything else to work.
  - felipemnoa 7y ago
    Thanks for your reply. Very informative.
ethbro 7y ago
Not stupid. NOTE: Not my field, so speaking off basic knowledge here. Happy to be corrected. Tl;dr: because biology at scale is complicated. Longer-form, because antibodies and immunological response are somewhat complicated. Knowing what an antibody looks like, how to produce it, and how to scale that production in a safe and reliable way are all very different things. Older article, but general concepts: http://neobiolab.com/research/in-vitro-and-in-vivo-production-of-antibodies http://neobiolab.com/research/in-vitro-and-in-vivo-productio... ... And because the in vivo methods we know of are ethically dicey. To rephrase it in rough terms -- creating custom cancers that produce the thing we want and giving them to mice, then harvesting concentrated antibodies from the tumors. See: https://en.wikipedia.org/wiki/Hybridoma_technology https://en.wikipedia.org/wiki/Hybridoma_technology
snowmaker 7y ago
Yes, and there are companies working on exactly this. Abcellera is furthest along: https://www.fiercebiotech.com/biotech/lilly-taps-abcellera-to-get-coronavirus-drug-into-clinic-4-months https://www.fiercebiotech.com/biotech/lilly-taps-abcellera-t...
boticus_prime 7y ago
https://www.fiercebiotech.com/research/fast-moving-regeneron-eyes-summer-clinical-trial-for-covid-19-antibody-cocktail-therapy https://www.fiercebiotech.com/research/fast-moving-regeneron... >Regeneron scientists have isolated hundreds of neutralizing antibodies against the SARS-CoV-2 virus from a humanized mouse model as well as from humans who have recovered from COVID-19, the company said Tuesday.
whoisjuan 7y ago
Yes. Synthetic antibodies are a thing, but they are very expensive and very hard to produce (from what I inderstand). Using animals when possible (for example horses, for the production of snake venom antidotes) seems more efficient in cost, volume and time. Take for instance this crisis. They just need the plasma of immune individuals to give temporary immunity to sick or at risk individuals. It doesn’t scale but it works. Some good basic reads: Antigens - https://en.m.wikipedia.org/wiki/Antigen https://en.m.wikipedia.org/wiki/Antigen Humoral Inmunity -https://en.m.wikipedia.org/wiki/Humoral_immunity https://en.m.wikipedia.org/wiki/Humoral_immunity Synthetic antibodies -https://en.m.wikipedia.org/wiki/Synthetic_antibody https://en.m.wikipedia.org/wiki/Synthetic_antibody Antibodies memetic - https://en.m.wikipedia.org/wiki/Antibody_mimetic https://en.m.wikipedia.org/wiki/Antibody_mimetic
andreareina 7y ago
Protein synthesis is complicated. Antivenom is still produced in the same way: inject an animal with small doses of venom, collect the antibodies from the plasma. [1] https://en.wikipedia.org/wiki/Antivenom https://en.wikipedia.org/wiki/Antivenom
ars 7y ago
That's basically what a https://en.wikipedia.org/wiki/Monoclonal_antibody https://en.wikipedia.org/wiki/Monoclonal_antibody is - and those are by far the most expensive drugs there are, with costs in the $100,000/person range.
- freepor 7y ago
  What would the costs look like if the market scaled up by 1000 or 10000x?
  - acqq 7y ago
    The cost comes from the nonscalability of both the production and the use.
    - jlokier 7y ago
      What are the main obstacles to scalability of production and of use? Could the former be amenable to treatment as an engineering problem? There are a lot of motivated engineering minds right now.
      - acqq 7y ago
        I'm not a professional in that area, just based on what I've read, so in case of doubt, check better sources: If you use human donors, you need trained doctors spending time and equipment to select the donors, handle the donors, take care of them during the donation which takes hours (the needle is stuck in their vein during that time), trained professionals using expensive equipment to process the donation, and then trained doctors to perform the infusion on the patient, an hour or two, but you still want keep the patient in the hospital overnight in case of complications, which do happen often enough that it was practice to do so: as in, once in thousand. At the end all that work, it is relatively sure that the described work could result in protection of exactly one patient more than before you started. The percentage of known complications and the potential for failure in every step is what makes it non scalable. With non human donors you can avoid some work, but the second part remains: it's not "just a shot" like when you do a vaccination which is scalable. And while vaccination is scalable, if you're going to vaccinate billions, you also want to be very sure you aren't harming them, that's why speeding up vaccine production is not scalable. Even producing the vaccines is not easy at all to scale. It's not that nobody has attempted hard to achieve this up to now.
echelon 7y ago
You need to create a cell line. Many medicines today use murine (mouse) monoclonal antibodies.
mtsr 7y ago
A monoclonal antibody (-mab) already existed in a lab in the Netherlands, because it had been developed for SARS and turned out to work for SARS-CoV-2. But it will take weeks before this can be industrialized, let alone approved.
dnautics 7y ago
Not really in time. In principle, yes, but antibodies are more complicated than normal molecules (glycosylation, protein folding) so you can really only industrially prepare them in certain types of cells (pretty much Chinese Hamster Ovary, last I checked). Another problem is retrieving the correct antibody gene sequence. All immunoactive antibodies have been subjected to a process of internal evolution and artificial selection. Their DNA is not the same as your germline DNA, through shuffling and active mutation events, and indeed every immune cell that produces antibodies has different DNA than the other ones (that haven't been told to clone themselves) so it's really difficult to retrieve the correct gene sequence. In practice I wouldn't estimate it to be possible to do this in the 4-6 month timeframe. Maybe in the 6-12 month timeframe under emergency duress situations.

jliptzin 7y ago

Definitely not a stupid question

felipemnoa 7y ago

Following your train of thought, could the synthesis be done in the same way that insulin is currently created? i.e. Have some germs synthesize the antibodies.

chispamed 7y ago
Yes, that’s a possibility and currently many monoclonal antibodies are produced that way (basically all drugs ending in “-mab”). However the process is costly and current capacities are not sufficient to supply a whole population regularly with antibodies which is why we still use some older drugs for certain diseases (think MTX for rheumatoid arthritis) and reserve monoclonal antibodies for treatment-resistant cases. I hope that it’ll be possible to protect high risk HCW and ICU patients through antibodies and thereby bring down Covid’s mortality soonish but we still need to buy some time until then and work on a vaccine to protect the population long-term. Right now what we need to do is stay at home and invest in medical care and drug development but with emphasis on the staying at home part because it’s the necessary condition for everything else to work.
- felipemnoa 7y ago
  Thanks for your reply. Very informative.

ethbro 7y ago

Not stupid. NOTE: Not my field, so speaking off basic knowledge here. Happy to be corrected. Tl;dr: because biology at scale is complicated. Longer-form, because antibodies and immunological response are somewhat complicated. Knowing what an antibody looks like, how to produce it, and how to scale that production in a safe and reliable way are all very different things. Older article, but general concepts: http://neobiolab.com/research/in-vitro-and-in-vivo-production-of-antibodies http://neobiolab.com/research/in-vitro-and-in-vivo-productio... ... And because the in vivo methods we know of are ethically dicey. To rephrase it in rough terms -- creating custom cancers that produce the thing we want and giving them to mice, then harvesting concentrated antibodies from the tumors. See: https://en.wikipedia.org/wiki/Hybridoma_technology https://en.wikipedia.org/wiki/Hybridoma_technology

snowmaker 7y ago

Yes, and there are companies working on exactly this. Abcellera is furthest along: https://www.fiercebiotech.com/biotech/lilly-taps-abcellera-to-get-coronavirus-drug-into-clinic-4-months https://www.fiercebiotech.com/biotech/lilly-taps-abcellera-t...

boticus_prime 7y ago

https://www.fiercebiotech.com/research/fast-moving-regeneron-eyes-summer-clinical-trial-for-covid-19-antibody-cocktail-therapy https://www.fiercebiotech.com/research/fast-moving-regeneron... >Regeneron scientists have isolated hundreds of neutralizing antibodies against the SARS-CoV-2 virus from a humanized mouse model as well as from humans who have recovered from COVID-19, the company said Tuesday.

whoisjuan 7y ago

Yes. Synthetic antibodies are a thing, but they are very expensive and very hard to produce (from what I inderstand). Using animals when possible (for example horses, for the production of snake venom antidotes) seems more efficient in cost, volume and time. Take for instance this crisis. They just need the plasma of immune individuals to give temporary immunity to sick or at risk individuals. It doesn’t scale but it works. Some good basic reads: Antigens - https://en.m.wikipedia.org/wiki/Antigen https://en.m.wikipedia.org/wiki/Antigen Humoral Inmunity -https://en.m.wikipedia.org/wiki/Humoral_immunity https://en.m.wikipedia.org/wiki/Humoral_immunity Synthetic antibodies -https://en.m.wikipedia.org/wiki/Synthetic_antibody https://en.m.wikipedia.org/wiki/Synthetic_antibody Antibodies memetic - https://en.m.wikipedia.org/wiki/Antibody_mimetic https://en.m.wikipedia.org/wiki/Antibody_mimetic

andreareina 7y ago

Protein synthesis is complicated. Antivenom is still produced in the same way: inject an animal with small doses of venom, collect the antibodies from the plasma. [1] https://en.wikipedia.org/wiki/Antivenom https://en.wikipedia.org/wiki/Antivenom

ars 7y ago

That's basically what a https://en.wikipedia.org/wiki/Monoclonal_antibody https://en.wikipedia.org/wiki/Monoclonal_antibody is - and those are by far the most expensive drugs there are, with costs in the $100,000/person range.

freepor 7y ago
What would the costs look like if the market scaled up by 1000 or 10000x?
- acqq 7y ago
  The cost comes from the nonscalability of both the production and the use.
  - jlokier 7y ago
    What are the main obstacles to scalability of production and of use? Could the former be amenable to treatment as an engineering problem? There are a lot of motivated engineering minds right now.
    - acqq 7y ago
      I'm not a professional in that area, just based on what I've read, so in case of doubt, check better sources: If you use human donors, you need trained doctors spending time and equipment to select the donors, handle the donors, take care of them during the donation which takes hours (the needle is stuck in their vein during that time), trained professionals using expensive equipment to process the donation, and then trained doctors to perform the infusion on the patient, an hour or two, but you still want keep the patient in the hospital overnight in case of complications, which do happen often enough that it was practice to do so: as in, once in thousand. At the end all that work, it is relatively sure that the described work could result in protection of exactly one patient more than before you started. The percentage of known complications and the potential for failure in every step is what makes it non scalable. With non human donors you can avoid some work, but the second part remains: it's not "just a shot" like when you do a vaccination which is scalable. And while vaccination is scalable, if you're going to vaccinate billions, you also want to be very sure you aren't harming them, that's why speeding up vaccine production is not scalable. Even producing the vaccines is not easy at all to scale. It's not that nobody has attempted hard to achieve this up to now.

echelon 7y ago

You need to create a cell line. Many medicines today use murine (mouse) monoclonal antibodies.

mtsr 7y ago

A monoclonal antibody (-mab) already existed in a lab in the Netherlands, because it had been developed for SARS and turned out to work for SARS-CoV-2. But it will take weeks before this can be industrialized, let alone approved.

dnautics 7y ago

Not really in time. In principle, yes, but antibodies are more complicated than normal molecules (glycosylation, protein folding) so you can really only industrially prepare them in certain types of cells (pretty much Chinese Hamster Ovary, last I checked). Another problem is retrieving the correct antibody gene sequence. All immunoactive antibodies have been subjected to a process of internal evolution and artificial selection. Their DNA is not the same as your germline DNA, through shuffling and active mutation events, and indeed every immune cell that produces antibodies has different DNA than the other ones (that haven't been told to clone themselves) so it's really difficult to retrieve the correct gene sequence. In practice I wouldn't estimate it to be possible to do this in the 4-6 month timeframe. Maybe in the 6-12 month timeframe under emergency duress situations.