Thread by @rvenkayya

THREAD
In light of the Pfizer #vaccine news, a natural question is whether it's feasible to develop “better” #COVID19 vaccines after the first ones are approved?

The answer is yes. It can be complicated but there are ways to do it. 1/

https://twitter.com/rvenkayya/status/1325850822677303297?s=20

First let’s break this down into three questions:
▪️ Why might we want better vaccines?
▪️ Why would it be hard to study new vaccines?
▪️ What are the options for doing this? 2/

WHY MIGHT WE WANT BETTER VACCINES?

After we’ve seen full Phase 3 datasets on the first vaccines, we may desire better efficacy in certain populations, longer protection, greater impact on transmission, improved dosing schedule, or an improved safety profile. 3/

It’s common for companies to improve upon their vaccines after licensure, but this can take years for regulatory or scientific reasons.

With COVID, we have the unique opportunity to choose from multiple vaccine programs to improve the parameters above. 4/

WHY WOULD IT BE HARD TO TEST NEW VACCINES?

Once a vaccine is approved (EUA or licensure) and recommended, it may be considered unethical to do a placebo-controlled efficacy trial in some groups because placebo recipients would be at greater risk than those outside the trial. 5/

An alternative is to compare efficacy of the “new” vaccine to one already approved (“active control”), but it takes a very large trial to show equivalence or greater benefit of the new vs existing vaccine due to potentially low incidence in both arms. 6/

Even if a placebo-controlled trial is justified (see below), there are practical limitations. People may be reluctant to enter a trial where they might receive placebo or a less-desirable investigational vaccine when an approved vaccine is available, or… 7/

…the incidence of disease might be too low to conduct a trial after the rollout of the first vaccines due to population immunity, public health measures, or changes in epidemiology. 8/

WHAT ARE THE OPTIONS AFTER THE FIRST VACCINES ARE APPROVED?

There are several options: (a) the trial can be conducted in groups for whom the vaccine isn’t yet recommended or available, such as healthy adults when initial doses of the vaccine are going to at-risk groups. 9/

This carries the risk that people leave the trial to take an approved vaccine when it becomes available to their group;

(b) the trial can be conducted in countries where vaccines haven't been approved or aren’t widely available. Data generated supports licensure elsewhere. 10/

(c) the trial can be conducted in places where existing vaccines are not broadly used because they don’t support local needs due to efficacy, cost, logistics, or other reasons. This has been examined by an @WHO expert panel. 11/
sciencedirect.com/science/articl…

It is possible to license new vaccines without conducting efficacy studies, if a “correlate of protection” has been identified. A correlate is a biomarker (typically an antibody) that allows predictions of the efficacy in the absence of an efficacy trial. 12/

An example is the hemagglutination inhibition titer (HAI) of antibodies used for licensure of influenza vaccines. A correlate can be used in either a placebo-controlled trial or a non-inferiority trial that compares a new vaccine to a currently-approved vaccine. 13/

Correlate-based trials are much smaller than standard Phase 3 efficacy trials, and they aren’t dependent on epidemiology of the virus. This substantially reduces time, cost and risk of the trial. 14/

A few notes. First, in the setting of multiple vaccines in development, it may be hard to obtain long-term safety and efficacy data because of the difficulty of, and potential ethical barriers to, keeping people in a placebo arm for extended follow-up. 15/

https://twitter.com/HelenBranswell/status/1319627487819800576?s=20

Second, human challenge trials would not replace Phase 3 trials for licensure of future vaccines, but they could de-risk vaccines before Phase 3, and can help to understand the impact of the vaccine on viral replication and shedding, and other aspects of infection. 16/

There have never been so many vaccines in development against the same pathogen. Even if the first vaccines are safe & effective, they’re unlikely to meet everyone’s needs. We’ll certainly find ethical and cost-effective ways to bring better vaccines to the world. 17/

Thanks to @mlipsitch for discussions on these points, and to two members of my team for their input:

▪️ Dr. Gary Dubin, an ID physician and head of our vaccine global medical office.

▪️ Dr. Hansi Dean, an immunologist and virologist, and head of our vaccine research team. 18/

Some good refs. This excellent article from @nataliexdean and @mlipsitch includes a detailed discussion of subgroup-specific analyses, including the value of maintaining blinding after EUA, and methods of assessing indirect protection. 19/

https://twitter.com/ScienceMagazine/status/1320606108927643648?s=20

And a great primer on assessment of vaccine efficacy from @nataliexdean 20/

https://twitter.com/nataliexdean/status/1310613702476017666?s=20

An excellent discussion of the implications of Emergency Use Authorization (EUA) for people in ongoing clinical trials, from @sciencecohen in conversation with @GaviSeth. 21/
sciencemag.org/news/2020/10/e…

If you were motivated enough to read this far, you’ve probably already seen @florian_krammer reasons for optimism about the Pfizer data. 22/

https://twitter.com/florian_krammer/status/1325887332428505090?s=20

Here’s his epic tweetorial on SARS-CoV-2 vaccines taken from his paper in @nature. 23/

https://twitter.com/florian_krammer/status/1310372301314101250?s=20

Finally, here’s a roundup of some reactions to the Pfizer news from @kakape, including from @cepivaccines @drrichardhatchett @wellcometrust @jeremyfarrar @dereklowe @helenbranswell. 24/

https://twitter.com/kakape/status/1325771586067197954?s=20