Switching strains in mRNA vaccines

Posted in Health, Random thoughts, Technology trends at 12:13 am by ducky

There is a lot of nervousness right now about new strains of the SARS-CoV-2 virus. There is now very good evidence that the B.1.1.7 (“UK”) strain is more transmissible than the original strain (which I will call the “Wuhan” strain because I haven’t seen anyone give it a name). There are also strains out of South Africa and Brazil which have similar mutations and are similarly worrying.

So I have been wondering how fast Moderna and Pfizer/BioNTech could tweak their strains to produce vaccines targeting the UK strain instead of the Wuhan strain. That has not been entirely straightforward to figure out, but I think I have finally pieced it together from reading, mostly from this and this.

The things that need to be done include:

  • Get the DNA sequence of the new strain’s spike protein. That’s something which comes from outside the company. (Zero additional time.)
  • Modify the sequence slightly in completely predictable ways: add a start block on, add an end block, convert all the thymine with 1-methyl-3’-pseudouridylyl, translate some bases into equivalent sequences to get a higher proportion of cytosine and guanine, etc. (See Reverse Engineering the source code of the BioNTech/Pfizer SARS-CoV-2 Vaccine for more on these algorithmic modifications.) (Educated guess: probably fifteen minutes of additional time, not even worth counting.)
  • Modify the sequence slightly to improve manufacturability. This will be a proprietary step that the company will do. Given how small the physical changes are between the Wuhan strain and the UK strain, I suspect that this step will not take long, and possibly could be skipped. (Guess: 0-1 days.)
  • Convert the text sequence into physical DNA, the “template DNA”. This is a completely straightforward, common process, but I don’t know how long it takes to get the amount of DNA needed. (Guess: 1-4 days)
  • Verify that you got the right DNA, I assume using a PCR test. (6 hours, 0.25 days)

After this point, all the rest of the production steps are exactly the same as are already being done for the Wuhan strain. If I were Pfizer or Moderna, I would already have done the previous steps for all of the scary variant strains and have a bunch of DNA on ice, ready to ship out if the green light was given. So it might in fact be zero additional days. (2-6 weeks, 6 weeks)

I have not heard of any official floating the idea of modifying the vaccine, probably because it looks like the Wuhan vaccine will be good enough against the UK strain. However, if there started to be rumbling about how the vaccine should be changed, if I were Pfizer or Moderna, I would actually go through all the steps to make enough doses of the UK vaccine for testing if need be.

At this point, the new vaccine would need some testing. According to the FDA officials, this would not need to be a full randomized clinical trial:

It might be the first time we do it, we’ll check an immunogenicity study. But it’s not going to have to be another 30,000 patient clinical trial. Those immunogenicity studies are usually 400 patients, just to make sure that we have the right check of what’s coming out. And even that may not be necessary after we check at the first one or two times. So I think we’ll have a way of evolving here with these.

If I understand correctly, immunogenicity studies are those which check to make sure that the vaccine causes the desired immune response.

This would mean that the the company would vaccinate volunteers — apparently about 400 — give them enough time for their immune system to react

  • Recruit test participants. Hopefully they would have volunteers ready to go. (Zero additional days.)
  • Inject test participants with their first dose. One vaccinator plus support staff in a high-throughput clinic can vaccinate 30 people per hour, so with appropriate application of resources, it should be no problem to get them all vaccinated in one day. (1 day.)
  • Wait for the test participants’ immune systems to react to the vaccines. (14 days.)
  • At this point, the test participants’ blood could be tested to see if it mounts the desired immune response. One might think they could stop there, but regulatory agencies probably would want to check after the second dose. Moderna ran their first trials with a 21 day delay; Pfizer did a 28 day delay. (7 or 14 more days)
  • Give all the volunteers their second dose. (1 day)
  • Wait for the volunteers’ immune systems to react to the vaccines. (14 days.)
  • Take blood samples from the test participants. (1 day)
  • Test blood to see if it mounts the desired immune response. (Guess: 2 days)
  • Write up report. I think this wouldn’t take long because they could copy and paste a lot from the Wuhan vaccine’s report. (Guess: 2-7 days.)
  • At this point, it would go to the regulatory bodies.

Add that all up, and it comes out to about 7-8 weeks from finishing the vaccine production to the data delivered to regulators.

The regulators would need to evaluate the data. Different regulators take different lengths of times, but the Pfizer vaccine took about three weeks from the application to approval by the US Food and Drug administration, and about the same amount of time for Health Canada to approve it. A new strain would have far less data to pore through, so I would guess it would be only a week or so. (Guess: 1 week)

In summary, my best guess is 4-6 weeks to manufacture the new vaccine, 7-8 weeks to do clinical trials, and 1 week for approval, or a total of 12-15 weeks.

Addendum 2021-02-06: According to a Washington Post article on 2021-01-25, Moderna has started working on a vaccine for the P.1 variant.

The scientific and pharmaceutical race to keep coronavirus vaccines ahead of new virus variants escalated Monday, even as a highly transmissible variant first detected in people who had recently traveled to Brazil was discovered in Minnesota.

Moderna, the maker of one of the two authorized coronavirus vaccines in the United States, announced it would develop and test a new vaccine tailored to block a similar mutation-riddled virus variant in case an updated shot becomes necessary.


Addendum 2021-02-07: I forgot that the delay between the first and second dose is not 14 days, it’s 21 (for Moderna) and 28 (for Pfizer). I have adjusted accordingly.

Addendum 2021-02-19: Today, an article in the Washington Post said

A prime inventor of the technology behind mRNA vaccines, Drew Weissman, of the University of Pennsylvania, said he has been told by the leader of BioNTech that it could take as little as six weeks to formulate a new mRNA payload and manufacture it to target a variant. Pfizer chief executive Albert Bourla told investors earlier this month that he anticipates that a variant-specific vaccine could be approved in 100 days, including clinical testing and regulatory reviews.

Addendum 2021-02-23: Pfizer says that it has gotten its manufacturing time down to 60 days (from 100).

Addendum 2021-02-25: Moderna is starting a Phase 1 trial of a vaccine specifically against the B.1.351 (South African) strain. Pfizer announced that they are going to do a test also.

Addendum 2021-02-25 from an article dated 2021-02-22: the US FDA has released guidelines on what testing vaccines against variant strains will need to do. They need to vaccinate volunteers with the variant vaccine, and look at the blood to see how big of a response to the virus variants the volunteers’ blood makes. If the response is comparable to the one from the Classic vaccines, it’s a go.

1 Comment »

  1. Jeff Powell said,

    January 24, 2021 at 10:43 am

    This is a nice summary for mRNA vaccines.

    But it begs the question about other kinds of vaccines and how long it will take for them to update their formulation for new variants.

    It seems likely the J&J & Astrazenica vaccines are going to be approved for use soon, and they are simpler to store & ship, and one of them is even a single dose only vaccine, we’re told. It is possible they will become the workhorse vaccines in some places as a result of those differences.

    And I know we manage to create a new flu vaccine every year in response to the mutations the flu has seen, so clearly adapting non-mRNA vaccines is possible, but I have no sense of how long that takes or what it involves, except that it can be accomplished in something less than a 12 month time frame, and involves n awful production process using jillions of actual chicken eggs. Ideally these non-mRNA vaccines are easier to produce than that, but I don’t know.

    In any case, as the variants of Covid keep on coming – which is normal for a new virus entering a population – it will be interesting to see how vaccine makers adapt. mRNA is a great technology, but until they make it easier to ship and store so it can be delivered to remote populations more easily, other kinds of vaccines will need to be available. How rapidly those vaccines can be adapted is an interesting question for much of the world.

Leave a Comment

Comment moderation is enabled. Your comment may take some time to appear.