The mathematics of early vaccine approval

OCTOBER IS OVER. And today is the US election day — with no vaccine authorised yet, even for emergency use. Amidst the most optimistic forecasts, the hoped-for COVID-19 shots did not arrive. It turns out that Pfizer's aspirational line — that it might have vaccine data by Halloween — did not happen, either. America's "Warp-speed" program to hasten vaccine trials and availability have been impressive, but even had its own limits.

HIGH STATISTICAL BAR: A vaccine approval must clear some mathematical hurdles, especially in the US. Proving a vaccine both protects and is safe is a high bar, statistically. A candidate shot needs to clock up some good numbers to get the US Food and Drug Administration (FDA)'s emergency-use authorisation. Now, Dr Anthony Fauci, America's top infectious disease expert, suggests a COVID vaccine might not be available on even a limited basis until January.

HIGH INFECTION RATE: Clearing the statistical bar on safety and efficacy needs to pass the gold standard of trials: double-blind, placebo-controlled, randomised. With the recent spike in COVID-19 cases, there's one upside: it may increase the rate of infection to validate on-going vaccine trials. Amidst a high infection rate, this could potentially lift a big hurdle in getting to the required threshold to warrant an "interim analysis". Vaccine researchers and the review board supervising them are gingerly looking for that data. Here's how the COVID-19 vaccine approval process works:

1. NATURAL INFECTION TAKES TIME: Some participants get a shot of the experimental vaccine (experimental group); some get a placebo (control group). Neither recipients nor the vaccinators know (double-blind) whether they're getting a placebo or experimental shot. Researchers measure how many get sick in each group. This takes time. Both groups are expected to get infected — naturally, not deliberately — from community transmission.

2. CHECKING PRELIMINARY DATA: Results are checked at several points in the trial (called "interim analysis") to see if the vaccinated participants have a lower rate of illness. In order to gain FDA authorization for use against the coronavirus, a vaccine needs to show firm evidence that it protects at least half of those inoculated.

3. 50% RATE NOT GOOD ENOUGH IN EARLY DATA: Barely clearing 50% isn't good enough to declare success after just a few months. Statistically, when numbers are low, there's a higher probability that what looks like a protective effect results from chance rather than the vaccine's good work.

4. HIGHER BAR: A higher bar for the first analysis — an efficacy of around 75% or better — is desired. As for an interim analysis: Moderna (testing mRNA-1273 on 30,000 volunteers) says that after 53 cases, it will take its first "peek" at the data.

5. WHAT A GOOD VACCINE MUST HAVE: If interim data show the vaccine is capable of preventing COVID-19 in 75% of patients, it will be seen as a good indicator. If first data analysis is inconclusive, the review board overseeing the 30,000-person trial won't reassess the shot until cases double (to 106). Again, this would depend on natural community-based transmission rate of the virus. And it needs time.

6. DATA PEEK: US pharma giant Pfizer, for its part, plans a data peek after just 32 confirmed symptomatic infections. (As of October 27, 2020, Pfizer researchers have not yet conducted an analysis of the efficacy of the vaccine). Not reaching just 32 cases in its trial to warrant an an interim analysis suggests participants may be catching COVID-19 less frequently than expected.

7. AN IMPORTANT BIT TO REMEMBER: Data may be skewed by "functional unblinding". It simple means this: the presence (or lack) of side effects might let volunteers "guess" if they got the vaccine. This could lead them to take different virus risks — thus potentially slowing and skewing trial data. This particular issue does not come into play until there are enough infections.

8. US APPROVAL PROCESS: The US Food and Drug Administration’s (FDA’s) Center for Biologics Evaluation and Research (CBER) is responsible for regulating vaccines in the United States. The vaccine product sponsor follows a multi-step approval process, which typically includes: (a) An Investigational New Drug application; (b) Pre-licensure vaccine clinical trials; (c) A Biologics License Application (BLA); (d) Inspection of the manufacturing facility; (e) Presentation of findings to FDA’s Vaccines and Related Biological Products Advisory Committeeexternal icon (VRBPAC); (f) Usability testing of product labeling.

9. VACCINE PRODUCTION OVERSIGHT: After a vaccine is approved, FDA continues to oversee its production to ensure continuing safety. Monitoring of the vaccine and of production activities, including periodic facility inspections, must continue as long as the manufacturer holds a license for the vaccine product. FDA can require a manufacturer submit the results of their own tests for potency, safety, and purity for each vaccine lot. FDA can require each manufacturer submit samples of each vaccine lot for testing.

10. CHECKING FOR SIDE EFFECTS: Once a trial vaccine is administered, there's a system to track side effects. The FDA has a Vaccine Adverse Event Reporting System (VAERS), a national vaccine safety surveillance program co-sponsored by the FDA and the CDC. VAERS collects and analyzes information from reports of adverse events (side effects) that occur after the administration of US licensed vaccines. Reports are welcome from all concerned individuals: patients, parents, healthcare providers, pharmacists, and vaccine manufacturers. To submit a report, use VAERS’ reporting pageexternal icon. [Source: FDA]