This post came in response to a discussion about an article in The Atlantic that cast some doubts about how America would react to a vaccine. If you want to read the entire conversation, try dropping in on the Covid thread here.
Hi - I have a PhD in viral immunology and worked in the field for 20 years, the last 10 with a large company that makes vaccines and I currently work in R&D for flu vaccine development. There are a lot of questions on many posts in response to the Atlantic article. Since freshmanjs listed many of them here I'll give responses to the best of my knowledge here
A vaccine is likely to limit serious disease rather than all disease. - Yes, the COVID-19 vaccine as with many vaccine is not likely to provide sterilizing immunity but will hopefully reduce death, hospitalizations and severe cases greatly. If the "death rate" could drop from say 1% to 0.1% and the hospitalization rate from 15% to 1.5% that would be great. SARS-CoV-2 is not going to disappear a few months after a vaccine is available but if we can manage it like the other respiratory viruses that circulate (flu, RSV, parainfluenza, common corona, rhino) would be the goal
A lot of doses will be required and that is logistically complicated - It seems very likely that the first COVID vaccine approaches in the pipeline will require 2 doses with the second one occurring about 3-4 weeks after the first. And then subjects would likely have their peak Ab response about 2 weeks after the second dose. So vaccine Ab immunity will peak about 6 weeks after first dose. Anytime that a vaccine requires 2 dose it is logistically much more difficult but it is not a failing of these first COVID vaccine approaches it is common that vaccine to naive (unencountered) viruses require 2 doses
The 2009 flu vaccine only ended up being used in 20% of the population because demand was low, as the disease had become much less prevalent - Not as much of the 2009 "swine" flu H1N1 vaccine was used as was produced is true, but it wasn't due to that strain of flu becoming less prevalent. The H1N1 pdm/A/California strain that appeared in 2009 was the exact H1N1 strain in the seasonal flu vaccine for 7 years (meaning it was most prevalent circulating strain). The swine flu vaccine process began in April (after it was discovered) in March and large amounts of doses were made by Oct/Nov which is impressive. However, that specific vaccine was not needed as much as was thought because pre-existing T cell immunity from previous (but different H1N1) seemed to lessen the severity of the disease in most people
Flu mist is scary because it's alive. - FluMist is live but it attenuated through cold temperature adaptation. It only replicates at around 33C when the human body is 37C. So while there is risk with any live vaccine, FluMist was quite safe. I do have issue with the NYT article about how live, intranasal vaccines would be preferable for coronavirus. While it is a true statement that intranasal administration of a live vaccine would theoretically be preferred because it is more likely to induce IgA Abs (the type associated with mucosal response in the lungs), the reality is that has no one has effectively proved that in dozens of years in human flu vaccine research. FluMist did show better response in kids (we gave it to our son when he younger) it was largely thought that improvement was due to enhanced T cell responses to FluMist and not to IgA Abs. VaxArt (mentioned in the NYT article) has been trying an oral flu vaccine approach for awhile but AFAIK have not gotten to a phase 3 efficacy trial yet. IgG Abs induced by intramuscular administration seem likely to work well enough for COVID-19 (early results are promising but not yet definitive
Other thoughts on timing and manufacturing supply chain: This is an issue not be given enough attention in the rightfully optimistic articles about Ph 1 success. Moderna is supposed to start their Phase 3 efficacy trial next week. If everything works perfectly it will take 3 months to know if the trial was successful. Not all 30,000 people can get the vaccine on Day 1. Normally it would take 2 months to enroll this many people. Assume they streamline it to 1 month. As I said above it will take about 6 weeks to reach peak Ab response with a 2 dose regimen. So in an efficacy trial that would be the bare minimum time to determine if the vaccine is better at preventing/reducing disease than the placebo. So that's 10 weeks. Need some time to compile the data and present to FDA. This would normally take months but even if it fast tracked that's 2 weeks and about 3 months total.
However, as the Atlantic article pointed out we have to manufacture 100 million doses, which is a huge undertaking. For my company it takes normally 5-6 months of highly entwined processes to make, fill vials, package, and distribute 100 million doses of flu vaccine. The filling, packaging, and distribution requires a lot of expertise but that expertise can be bought. The tricky part is upscaling your manufacturing processes from say thousands of research grade doses to millions of manufacture grade doses. This process can theoretically go quicker with mRNA but can be fraught with lots of pitfalls. Moderna has contracted with Lonza to do this but who knows. In addition to the technical and logistical difficulties, normally you don't start manufacturing a vaccine at risk (before FDA licensure). Not a sound financial strategy. Now Operation Warp Speed is likely to help provide $$$ to start these processes at risk but I'd say we have 1-2 months minimum from licensure to having say 10 million doses available.
Deciding who gets these doses and how to distribute these doses should be underway by the federal government right now. But I have seen no efforts on this front. NIH/FDA have experience in the running trial and regulatory approval but not in distribution. That is generally left to the individual companies. That doesn't seem like a good idea for COVID.
So in summary. I am more optimistic than the tone of the Atlantic article that we will have effective vaccines. I agree with Jesse Goodman (former FDA official quoted in the Atlantic article) that we are probably 5 months from say 10 million doses of a vaccine. This won't be wrapped up by Christmas. Not even close. However, I think that with the second wave of companies (that are a few months behind Moderna/Pfizer) we will have hundreds of millions of vaccine available by late spring early summer 2021. Enough to vaccine all in US that want it. If enough people get the vaccine, I think we could have a situation where COVID-19 is a bit more like a normal respiratory virus public health situation by fall 2021. However, sadly we won't really know if that is the case until we get through the 2021 flu season. So, in my opinion there is a likely path at the end of the tunnel I just don't think it is quite as close as some articles make it seem.