I am not throwin' away my shot
April 28, 2021 11:17 AM Subscribe
Nice find, storybored!
posted by They sucked his brains out! at 11:53 AM on April 28, 2021
posted by They sucked his brains out! at 11:53 AM on April 28, 2021
Very nice! I had a totally different workflow in my head; one that involved synthesizing the mRNA directly from nucleotides. But then again I really don’t know the engineering aspects of large scale biochemical (or other) manufacturing at all, other than watching How It’s Made a lot. But my brother-in-law is a mechanical engineer who specializes in production and packaging lines. So I sent him this; not just because he would like it, but maybe more information will help push my vaccine-hesitant sister (I haven’t pressed her on why, but I’m sure Fox News is involved) toward getting the shot.
posted by TedW at 12:09 PM on April 28, 2021 [1 favorite]
posted by TedW at 12:09 PM on April 28, 2021 [1 favorite]
"The Chesterfield facility is Pfizer’s only source of plasmids for its Covid-19 vaccine. But finishing the vaccine requires several more steps in two other facilities."
Well that's slightly terrifying. I hope they have off-site backups.
posted by kaibutsu at 12:12 PM on April 28, 2021 [3 favorites]
Well that's slightly terrifying. I hope they have off-site backups.
posted by kaibutsu at 12:12 PM on April 28, 2021 [3 favorites]
Brazil Rejects the Gamaleya Vaccine and Russian Vaccine Behavior, two posts from In The Pipeline about another widely-used vaccine's production issues (and the dissemination thereof).
posted by lalochezia at 12:25 PM on April 28, 2021 [1 favorite]
posted by lalochezia at 12:25 PM on April 28, 2021 [1 favorite]
Also really cool to see the yields involved here: 1.5 million doses from a liter of source material.
It would be cool if we could throw a few GPUs in a tank of nutrient broth and just let 'em multiply...
posted by kaibutsu at 12:26 PM on April 28, 2021 [2 favorites]
It would be cool if we could throw a few GPUs in a tank of nutrient broth and just let 'em multiply...
posted by kaibutsu at 12:26 PM on April 28, 2021 [2 favorites]
Since I work in this field: how many of the lab techs there earn a wage that lets them comfortably rent a 1 bedroom apartment in Kalamazoo, alone?
There's billions of dollars sloshing around and a lot rides on each and every lab tech or GMP 1 operator to complete the tasks within the validated parameters of the process. But removing any financial stressors that may bleed into work performance is not considered here in DC and I can't imagine that its any different in MI.
A lot of the new employees at this multi-billion dollar international mega corp live with their parents until they peel out for a different job.
posted by Slackermagee at 12:27 PM on April 28, 2021 [18 favorites]
There's billions of dollars sloshing around and a lot rides on each and every lab tech or GMP 1 operator to complete the tasks within the validated parameters of the process. But removing any financial stressors that may bleed into work performance is not considered here in DC and I can't imagine that its any different in MI.
A lot of the new employees at this multi-billion dollar international mega corp live with their parents until they peel out for a different job.
posted by Slackermagee at 12:27 PM on April 28, 2021 [18 favorites]
It would be cool if we could throw a few GPUs in a tank of nutrient broth and just let 'em multiply...
As someone who is basically made of nutrient broth, I am opposed to this idea
posted by aubilenon at 12:55 PM on April 28, 2021 [11 favorites]
As someone who is basically made of nutrient broth, I am opposed to this idea
posted by aubilenon at 12:55 PM on April 28, 2021 [11 favorites]
I can't believe they completely glossed over the part where they add the microchips.
posted by vverse23 at 1:12 PM on April 28, 2021 [13 favorites]
posted by vverse23 at 1:12 PM on April 28, 2021 [13 favorites]
> "There’s no weekend breaks. There’s no whitespace in the project plan. You hire people, you put everyone you can on."
THEN HIRE MORE PEOPLE. At the scale and importance of this project there's no lack of financial resources, you don't have to skimp out in this one key area.
posted by JHarris at 1:20 PM on April 28, 2021 [9 favorites]
THEN HIRE MORE PEOPLE. At the scale and importance of this project there's no lack of financial resources, you don't have to skimp out in this one key area.
posted by JHarris at 1:20 PM on April 28, 2021 [9 favorites]
Ctrl+F orphans
Phrase not found
Huh. TIL.
posted by The Tensor at 1:27 PM on April 28, 2021 [1 favorite]
Phrase not found
Huh. TIL.
posted by The Tensor at 1:27 PM on April 28, 2021 [1 favorite]
THEN HIRE MORE PEOPLE
Hiring more people is one of the things they say they're doing in your quote from the article! But it's not clear to me though that "no weekend breaks" means everyone's working huge inhumane hours, or just that if they don't have weekend shifts their equipment is operating at 5/7 capacity.
posted by aubilenon at 1:28 PM on April 28, 2021 [7 favorites]
Hiring more people is one of the things they say they're doing in your quote from the article! But it's not clear to me though that "no weekend breaks" means everyone's working huge inhumane hours, or just that if they don't have weekend shifts their equipment is operating at 5/7 capacity.
posted by aubilenon at 1:28 PM on April 28, 2021 [7 favorites]
Evidently they're not hiring enough? Over the years I have come, with some cause, to look at these kinds of corporate statements in the worst possible light, and once in a while it is unjustified. If "no weekend breaks" doesn't mean they're prevailing upon employees to heroically make up for lacks produced by their own hiring practices then I will happily emit a mea culpa, but that phrase is a sign.
Anyway. lalochezia's links are very interesting, indicating that Trump's (and presumably Putin's) anti-fact "fake news" messaging has infected the Russian vaccine producer's messaging. Derek Lowe's comment on it is refreshingly direct to hear from a learned source:
"These claims are bullshit. Posting them is a disgrace."
posted by JHarris at 1:47 PM on April 28, 2021 [2 favorites]
Anyway. lalochezia's links are very interesting, indicating that Trump's (and presumably Putin's) anti-fact "fake news" messaging has infected the Russian vaccine producer's messaging. Derek Lowe's comment on it is refreshingly direct to hear from a learned source:
"These claims are bullshit. Posting them is a disgrace."
posted by JHarris at 1:47 PM on April 28, 2021 [2 favorites]
Well that's slightly terrifying. I hope they have off-site backups.
What matters is the factory equipment, not anything they could back up off site. You could make more plasmids over the weekend in a million labs around the country; what you couldn't do is manufacture at scale.
one that involved synthesizing the mRNA directly from nucleotides
Polymer synthesis is tough for biomolecules. The spike protein is about 3800 base pairs long. If you tried to build it via linear chemical synthesis, and got 99% yield at each step, you'd need 40,000 tons of starting material to get a microgram of RNA. (That's not quite how you'd do it, but there's a reason why you get the bugs involved.)
THEN HIRE MORE PEOPLE. At the scale and importance of this project there's no lack of financial resources, you don't have to skimp out in this one key area.
I understand the sentiment is routed in valuing the humanity of workers, but ironically the underlying assumption needed for this to be the solution is that people are interchangeable raw materials who can be used as inputs in a process as needed.
That's not how that works in this area. Hiring people and training them takes time. There was *no* bulk scale manufacturing of mRNA a year ago. None, anywhere in the world. The vaccines haven't even been approved for six months. And somehow you've gotten enough, from a standing start, to basically cover the entire US population.
You're trying to build capacity from the ground up. You have a finite number of people who can hire, supervise, train, review and test the new employees, and probably you are asking them to do both things. New procedures are being written, QC tests are uncovering new problems, deviations filed, root causes analyzed, and so on. You're running through cycles of these things and if each cycle takes a few weeks longer and the newly trained people are working fewer hours, you can easily imagine vaccination progressing six months or a year behind the current schedule.
I would have bet heavily against them coming anywhere close to this level of production, and the foul ups at J&J (with much older tech) show how easy it would be to have made serious mistakes.
posted by mark k at 3:00 PM on April 28, 2021 [24 favorites]
What matters is the factory equipment, not anything they could back up off site. You could make more plasmids over the weekend in a million labs around the country; what you couldn't do is manufacture at scale.
one that involved synthesizing the mRNA directly from nucleotides
Polymer synthesis is tough for biomolecules. The spike protein is about 3800 base pairs long. If you tried to build it via linear chemical synthesis, and got 99% yield at each step, you'd need 40,000 tons of starting material to get a microgram of RNA. (That's not quite how you'd do it, but there's a reason why you get the bugs involved.)
THEN HIRE MORE PEOPLE. At the scale and importance of this project there's no lack of financial resources, you don't have to skimp out in this one key area.
I understand the sentiment is routed in valuing the humanity of workers, but ironically the underlying assumption needed for this to be the solution is that people are interchangeable raw materials who can be used as inputs in a process as needed.
That's not how that works in this area. Hiring people and training them takes time. There was *no* bulk scale manufacturing of mRNA a year ago. None, anywhere in the world. The vaccines haven't even been approved for six months. And somehow you've gotten enough, from a standing start, to basically cover the entire US population.
You're trying to build capacity from the ground up. You have a finite number of people who can hire, supervise, train, review and test the new employees, and probably you are asking them to do both things. New procedures are being written, QC tests are uncovering new problems, deviations filed, root causes analyzed, and so on. You're running through cycles of these things and if each cycle takes a few weeks longer and the newly trained people are working fewer hours, you can easily imagine vaccination progressing six months or a year behind the current schedule.
I would have bet heavily against them coming anywhere close to this level of production, and the foul ups at J&J (with much older tech) show how easy it would be to have made serious mistakes.
posted by mark k at 3:00 PM on April 28, 2021 [24 favorites]
Since I work in this field: how many of the lab techs there earn a wage that lets them comfortably rent a 1 bedroom apartment in Kalamazoo, alone?
i pay 675 a month for a 2 bedroom apt here in kalamazoo - i make over 18 an hour and am a single parent of a disabled adult child
i'm sure they do alright
posted by pyramid termite at 3:07 PM on April 28, 2021 [6 favorites]
i pay 675 a month for a 2 bedroom apt here in kalamazoo - i make over 18 an hour and am a single parent of a disabled adult child
i'm sure they do alright
posted by pyramid termite at 3:07 PM on April 28, 2021 [6 favorites]
Hunh I also had assumed the RNA was synthetically made but in retrospect, I can see that accomplishing amplification at the DNA step using bacteria is probably safer. from the pics, one might assume that they are checking sequence fidelity solely via electrophoretic migration but I would hope they are also sequencing.
And those Ecoli are the prettiest and most anthropomorphic Ecoli I’ve ever come across. They’re going to be in the next Minions movie.
posted by Tandem Affinity at 6:25 PM on April 28, 2021 [1 favorite]
And those Ecoli are the prettiest and most anthropomorphic Ecoli I’ve ever come across. They’re going to be in the next Minions movie.
posted by Tandem Affinity at 6:25 PM on April 28, 2021 [1 favorite]
Just to say--those are good points mark k.
posted by JHarris at 7:13 PM on April 28, 2021 [2 favorites]
posted by JHarris at 7:13 PM on April 28, 2021 [2 favorites]
Derek Lowe's comment on it is refreshingly direct to hear from a learned source
"Refreshingly direct" is pretty much Derek Lowe's shtick.
posted by flabdablet at 8:49 PM on April 28, 2021
"Refreshingly direct" is pretty much Derek Lowe's shtick.
posted by flabdablet at 8:49 PM on April 28, 2021
It can be very easy to look at the production problems that AstraZeneca is having and think, "wow, they must be really incompetent to mess up like this". It's much more sensible to look at the whole spectrum of what's going in vaccine research and production.
Merck dropped their vaccine trial after disappointing phase I results.
Sanofi and GSK are redoing their trial after making a dosing mistake due to a problem with an assay. The same mistake btw that led to a small sub-group in the AZ trial receiving a half dose. The difference is that AZ still worked at half dose and this vaccine had disappointing results.
CureVac is still not approved.
NovaVax is being considered for approval now and they are having massive scaling and supply chain issues.
AZ is seeing much lower yields in the viral growth phase of the process which is not that surprising given that an order of magnitude difference sometimes occurs between batches in biological processes and tuning these is half magic half science.
J&J had an issue where a contract manufacturer cross-contaminated 15m doses worth in what appears to have been a grossly inadequate facility (not their facility as was directly contracted by US government).
Moderna has managed to hit their US delivery deadlines but has had to make big cuts to European deliveries. The only reason there isn't more coverage of that is that the EU didn't buy very much of it so it doesn't matter.
Apparently the Brazilian pharma regulator found replication competent Adenovirus (!) in the Sputnik vaccine.
Pfizer has more or less hit their delivery schedules but also had to delay and change things as they scale production.
I'm beginning to think that developing and producing vaccines in a year might be hard.
posted by atrazine at 6:29 AM on April 29, 2021 [17 favorites]
Merck dropped their vaccine trial after disappointing phase I results.
Sanofi and GSK are redoing their trial after making a dosing mistake due to a problem with an assay. The same mistake btw that led to a small sub-group in the AZ trial receiving a half dose. The difference is that AZ still worked at half dose and this vaccine had disappointing results.
CureVac is still not approved.
NovaVax is being considered for approval now and they are having massive scaling and supply chain issues.
AZ is seeing much lower yields in the viral growth phase of the process which is not that surprising given that an order of magnitude difference sometimes occurs between batches in biological processes and tuning these is half magic half science.
J&J had an issue where a contract manufacturer cross-contaminated 15m doses worth in what appears to have been a grossly inadequate facility (not their facility as was directly contracted by US government).
Moderna has managed to hit their US delivery deadlines but has had to make big cuts to European deliveries. The only reason there isn't more coverage of that is that the EU didn't buy very much of it so it doesn't matter.
Apparently the Brazilian pharma regulator found replication competent Adenovirus (!) in the Sputnik vaccine.
Pfizer has more or less hit their delivery schedules but also had to delay and change things as they scale production.
I'm beginning to think that developing and producing vaccines in a year might be hard.
posted by atrazine at 6:29 AM on April 29, 2021 [17 favorites]
Apparently the Brazilian pharma regulator found replication competent Adenovirus (!) in the Sputnik vaccine.--atrazine
Well this is interesting news. I found an article that describes what this means further.
Apparently you can design an adenovirus vaccine with the adenovirus being able to reproduce and there can be reasons to do this. But the Sputnik vaccine was not designed or tested this way, so if this was found in the vaccine, then its manufacturing isn't being well controlled, and the side effects of this are not known (because it hasn't been tested this way).
posted by eye of newt at 10:18 AM on April 29, 2021 [1 favorite]
Well this is interesting news. I found an article that describes what this means further.
Apparently you can design an adenovirus vaccine with the adenovirus being able to reproduce and there can be reasons to do this. But the Sputnik vaccine was not designed or tested this way, so if this was found in the vaccine, then its manufacturing isn't being well controlled, and the side effects of this are not known (because it hasn't been tested this way).
posted by eye of newt at 10:18 AM on April 29, 2021 [1 favorite]
Yeah, sorry I should have linked to that which is also where I read it.
The way that all the Adenovirus vectored vaccines work is that they modify an Adenovirus (Ad26 for J&J, a Chimpanzee Adenovirus for AZ, and an Ad26 + Ad5 for Sputnik) in such a way that it cannot replicate, it just enters a cell and then forces the cell to produce the spike protein from Covid-19 but has had a gene removed which means it can't make the other parts of the virus and therefore can't replicate.
The removed gene is called E1 in this case.
The reason that these viruses are selected is that they are not very immunogenic, which is good because or else the immune system would mop them up before they could get into cells either after the first dose or when boosted. These are selected because they are rare (most immune systems have never seen a Chimp adenovirus before). This is probably why AZ works better with a longer interval between the doses, it lets the immune response to the vector abate. Sputnik uses two different viruses (called heterologous prime-boost) for this same reason.
The problem is, having removed the E1 gene from your virus, how do you grow it? All these viruses are grown on a cell line of human embryonic kidney cells which have been modified to insert the E1 gene. So when the virus enters those cells, it finds the proteins that it needs to assemble itself already being made by the host cell.
There are vaccine designs with replicating vectors (in a way, an attenuated virus vaccine is just that) but none of these vaccines are meant to. Anvisa, the Brazilian pharma regulator, found that in every examined batch of one of the two Sputnik vectors, there were replication competent (i.e. they had intact E1 genes and could replicate) viruses.
To be clear, that doesn't mean those vaccines don't work. They might actually work better by promoting a more robust immune response since replicating viruses lead to all kinds of immune signalling. It doesn't mean necessarily that they're not safe either. It does raise some pretty spicy questions though about what else exactly is going on and they certainly weren't intended to do this.
I'm genuinely surprised by this. One of the only high-tech sectors where Russia genuinely does manage to turn its excellent science into internationally competitive products is vaccine manufacture and the USSR was really very good at vaccine development. On the other hand, there's no reason for Anvisa to invent these findings.
Possibilities:
1) Anvisa has made an error with their assays. There are in fact no replication competent vectors.
2) There are but no-one at Gamaleya noticed. Unlikely given the quality of their scientists.
3) There are, they knew there were traces numbers of replication competent viruses in there but decided that based on the trial results and their view on the basic science it didn't seem to be doing any harm and it would be better for public trust if they kept this quiet.
posted by atrazine at 12:10 PM on April 29, 2021 [7 favorites]
The way that all the Adenovirus vectored vaccines work is that they modify an Adenovirus (Ad26 for J&J, a Chimpanzee Adenovirus for AZ, and an Ad26 + Ad5 for Sputnik) in such a way that it cannot replicate, it just enters a cell and then forces the cell to produce the spike protein from Covid-19 but has had a gene removed which means it can't make the other parts of the virus and therefore can't replicate.
The removed gene is called E1 in this case.
The reason that these viruses are selected is that they are not very immunogenic, which is good because or else the immune system would mop them up before they could get into cells either after the first dose or when boosted. These are selected because they are rare (most immune systems have never seen a Chimp adenovirus before). This is probably why AZ works better with a longer interval between the doses, it lets the immune response to the vector abate. Sputnik uses two different viruses (called heterologous prime-boost) for this same reason.
The problem is, having removed the E1 gene from your virus, how do you grow it? All these viruses are grown on a cell line of human embryonic kidney cells which have been modified to insert the E1 gene. So when the virus enters those cells, it finds the proteins that it needs to assemble itself already being made by the host cell.
There are vaccine designs with replicating vectors (in a way, an attenuated virus vaccine is just that) but none of these vaccines are meant to. Anvisa, the Brazilian pharma regulator, found that in every examined batch of one of the two Sputnik vectors, there were replication competent (i.e. they had intact E1 genes and could replicate) viruses.
To be clear, that doesn't mean those vaccines don't work. They might actually work better by promoting a more robust immune response since replicating viruses lead to all kinds of immune signalling. It doesn't mean necessarily that they're not safe either. It does raise some pretty spicy questions though about what else exactly is going on and they certainly weren't intended to do this.
I'm genuinely surprised by this. One of the only high-tech sectors where Russia genuinely does manage to turn its excellent science into internationally competitive products is vaccine manufacture and the USSR was really very good at vaccine development. On the other hand, there's no reason for Anvisa to invent these findings.
Possibilities:
1) Anvisa has made an error with their assays. There are in fact no replication competent vectors.
2) There are but no-one at Gamaleya noticed. Unlikely given the quality of their scientists.
3) There are, they knew there were traces numbers of replication competent viruses in there but decided that based on the trial results and their view on the basic science it didn't seem to be doing any harm and it would be better for public trust if they kept this quiet.
posted by atrazine at 12:10 PM on April 29, 2021 [7 favorites]
I think you're ignoring the most obvious hypothesis: This was detected but the people who detected it were ignored and the results buried. Just based on Lowe's description of their Twitter feed, there's a clear "see no evil" message being sent from the top down.
posted by mark k at 1:39 PM on April 29, 2021 [1 favorite]
posted by mark k at 1:39 PM on April 29, 2021 [1 favorite]
Since I work in this field: how many of the lab techs there earn a wage that lets them comfortably rent a 1 bedroom apartment in Kalamazoo, alone?
Good point! This would be a very effective moment for a strike action.
posted by latkes at 2:03 PM on April 29, 2021
Good point! This would be a very effective moment for a strike action.
posted by latkes at 2:03 PM on April 29, 2021
Yes but even those people need something to tell themselves to justify it. "it's safe anyway, if unintended" might well be that thing.
posted by atrazine at 2:18 PM on April 29, 2021
posted by atrazine at 2:18 PM on April 29, 2021
Yes but even those people need something to tell themselves to justify it. "it's safe anyway, if unintended" might well be that thing.
I guess I'm thinking the core justification is more "I'm just a junior guy, someone above my pay grade needs to take this on." That stays the justification for not blowing up your career until it hits the point that the managers are saying "Well, if the people doing hands on stuff were really worried they would be making a bigger stink about this."
That's often how it turns out these things worked when you dig into it. There's plausible deniability at various steps, but it's not like anyone is actually looking at the evidence and saying it's OK.
posted by mark k at 3:31 PM on April 29, 2021
I guess I'm thinking the core justification is more "I'm just a junior guy, someone above my pay grade needs to take this on." That stays the justification for not blowing up your career until it hits the point that the managers are saying "Well, if the people doing hands on stuff were really worried they would be making a bigger stink about this."
That's often how it turns out these things worked when you dig into it. There's plausible deniability at various steps, but it's not like anyone is actually looking at the evidence and saying it's OK.
posted by mark k at 3:31 PM on April 29, 2021
I expected from the title that this would be an homage to Fred Rogers’s “How crayons are made” segment, with meditative narration and a jazz soundtrack.
Alas, it’s the other kind of awesome.
posted by fantabulous timewaster at 6:55 AM on April 30, 2021
Alas, it’s the other kind of awesome.
posted by fantabulous timewaster at 6:55 AM on April 30, 2021
Glassdoor says the annual salary of lab techs in Kalamazoo is around $39k. If you use a cost of living calculator that translates to $82k in DC. The working conditions in science can be rough unless you make it to the top of the pyramid, but if these numbers are at all accurate I’d feel a lot better employing a tech in Kalamazoo than a postdoc in San Francisco.
posted by en forme de poire at 7:29 AM on April 30, 2021 [2 favorites]
posted by en forme de poire at 7:29 AM on April 30, 2021 [2 favorites]
« Older Why You Should Consider Second Dog | The Great Moose Migration Newer »
This thread has been archived and is closed to new comments
That's a great article, thanks for posting. It's good to see the vaccine manufacturing tech for reals.
posted by storybored at 11:23 AM on April 28, 2021 [4 favorites]