Case study: mpox vaccine
In a sudden public health emergency caused by mpox in 2022-23 there was not enough supply of the only mpox vaccine (Jynneos/Imvanex, made by Bavarian Nordic) to immunise the at-risk population (men who have sex with men). Unlike other case studies presented, this affected all countries, e.g., in the US over 3mln doses were needed to vaccinate high-risk population and the estimated shortage was 50%. UK had a stockpile of 50,000 and moved to order 150,000 (source). It’s safe to assume that in other countries the shortages would be larger.
Jynneos/Imvanex (live smallpox and mpox vaccine) was initially approved for SC injection in two doses. The fractional dosing approach is to give two 1/5 doses (0.1mL instead of 0.5mL) intradermally. This is a “regular” vaccination, just done at 5-15 degree angle and using a smaller needle and syringe.
What was done in 2022
In August 2022 FDA gave emergency use authorization (EUA) for this FD based on the only RCT from 2015 (n=300, comparing 0.1mL ID arm to 0.5mL IM); immunogenicity was deemed “nearly identical,” as was safety (source). A few things to note:
However, there are no defined correlates of protection against mpox, i.e. we don’t actually know how well mpox virus-neturalising antibodies protect against disease or transmission, so this was definitely a decision made with considerable uncertainty.
Note that this was an actual EUA for this specific mode of application. This is in contrast to other recommendations which are made not by drug regulators and constitute off-label vaccine use guidelines.
CEO of Bavarian Nordic wrote to the FDA comissioner to raise some concerns about using small sample to make decisions
EMA’s Emergency Task Force advised use of FD based on the same data. UKHSA piloted use of FD and UK’s JCVI recommended use of intradermal doses.
UK’s pilot was also because they wanted to resolve programmatic issues
In parallel there was a debate on First Doses First. In the US FDA advised states against it, while some experts suggested that it may work.
What we learned since 2022
Following the public health emergency we learned that:
Based on an immunogenicity study (N=225), the antibody titres were similar between SC full dose and ID 1/5 dose two weeks after second dose; 1/10th ID dose was inferior (source).
2 fractional doses are superior to one SC standard dose and no different to 2 SC standard doses per this NEJM study (N=145).
A large mpox vaccination effectiveness study (case-control design, N = 2,193 cases, 8,319 controls) found lower effectiveness of one dose compared to two doses (VE = 36% with one, 66% with two doses) but no comparison of effectiveness of ID vs SC administration route as the study was not powered to detect them.
Studies also note that a very important determinant of immune response is previous smallpox immunisation. There has been some discussion about general level of antibodies following two doses being low and therefore need for a third dose of vaccine in the future.
Data from US roll-out suggests that both options are well-tolerated, although ID has more local reactions. People complete their 2 doses at same rates regardless of ID vs SC.
It seems that generally the conclusion is that this worked well, but there isn’t yet enough evidence about ID to make strong conclusions.
Modelling
In retrospect fractionation was very likely the right decision, but what about public health benefits assessment at the time?
Only modeling in the public domain that I am aware of has been done by Dimitrov et al. This is adaptation of a compartmental epidemiological model (a very typical approach) applied to a population in Seattle, by experienced modellers; I discuss it in modeling health benefits of optimal dosing. I was told by one of the authors that this did not drive any policy.
However, there has been a lot of modelling of mpox outbreaks in the US done by a group at CDC led by Ian Spicknall, e.g. here. I am uncertain what role the modelling has played in the decisions and it would be good to understand that.