| Even though the Director of the WHO said yesterday that you need 2-3 months for clinical trials, our Chief Medical Officer, Sir Liam Donaldson, appears adamant that first doses will arrive in early autumn. Although he back-peddled a little by saying "We then have to license it properly and that may take a little time", still he appeared rather confident (and I might add more than a little self-congratulatory), that we'd be one of the first nations to receive vaccines, that all 60 million doses will be available before the end of the calendar year, presumably to be given out asap to protect the British people against this H1N1 virus.
Which got me thinking, what exactly is going on here? I mean, Margaret Chan notwithstanding, Sir Liam appears to be quite certain that we could be doing mass vaccinations pretty d****d quickly, and not several months down the road after clinical trials are all done. So what's the deal here, with the 'fast-tracking in 5 days' thing? Can they really license and roll out a pandemic vaccine for tens of millions of people in 5 days, without clinical trials?
I looked, and the answer is, YES THEY CAN.
What's more, it's perfectly legal - they don't even need to change any laws or invoke such things as Emergency Use Authorization like what the US government would need. At least that is as far as I can determine.
Let's take a look at how pandemic vaccines are licensed in the EU. I must first raise a caveat, that this is a much simplified exploration of a complex subject, by no means comprehensive. Also, for the sake of clarity, I'd prefer that specific issues about individual vaccines be explored separately, some time hopefully in the near future.
First of all, the European Medicines Agency (EMEA) is the EU regulatory body, the equivalent of the FDA. There are, however, important differences. Since the EU is made up of sovereign member states, there are 2 tracks to licensure. Companies can license their product with individual regulatory agencies, like the MHRA in the UK, but since approval from the EMEA allows marketing in all 27 member states, companies of course much prefer the EMEA route. Assessment of applications in the EMEA is under the Committee for Medicinal Products for Human Use or CHMP, made up of one expert from each member state. Since not all countries have similar levels of expertise or excellence in governance, there are IMO consequences from this practice.
OTOH, member states do have the discretion to approve or deny a license to any product approved by the CHMP/EMEA. After all, they are sovereign countries. An example is Fluad, the adjuvanted seasonal flu vaccine that is approved for use in the EU for those aged 65 or older. Even though the approval is there, not every country has licensed or used it.
With regards to influenza vaccine approval, the EMEA has 3 different tracks - interpandemic, pandemic, and pre-pandemic vaccines. - An interpandemic vaccine is basically your seasonal flu vaccine, made with the circulating seasonal strains.
- A pandemic vaccine is a vaccine made with a virus of pandemic potential, and intended to be used only when a pandemic breaks out.
- A pre-pandemic vaccine is made with a virus of pandemic potential, but licensed to be used before a pandemic breaks out.
Since a pre-pandemic vaccine will be used in the absence of a pandemic, the risk/benefits calculations are different from a pandemic vaccine. In essence, the EMEA requires a pre-pandemic vaccine to fulfill the same standards as the interpandemic or seasonal vaccine, for safety and immunogenicity etc.
For pandemic vaccine, however, the EMEA offers a different procedure for fast-tracking, where companies submit a 'mock-up dossier' for approval, and, when a pandemic happens, they file for strain variation. The process is shown in these 2 slides (source).
 
As you can see from the flow-chart, the idea is when a pandemic is declared, the manufacturer does not need to re-submit the whole dossier, only information on the new strain used, to apply for 'pandemic variation'. Here's what the EMEA guideline says:
In the case of an actual pandemic, the vaccination programme would need to be implemented as soon as possible. Provided that the mock-up and final pandemic vaccines are similar other than in strain content and the dose schedule is unchanged, the final pandemic vaccine may be approved for use by means of a variation that addresses only the quality issues and without the provision of clinical data.
In other words, once the vaccines are made, the EMEA can grant approval for marketing the vaccines based solely on product quality assessments, and no further clinical trials are required before the vaccines are properly, legally, approved, to be given to the citizens of these 27 countries.
No wonder Sir Liam is so optimistic, that we WILL have a vaccine in early autumn. Exactly what kind of vaccine(s), and how safe they are without clinical trials, is another story.
Now, let's take a closer look at the fast-tracking requirements. First question is, would a vaccine now made for H1N1 be considered 'similar other than in strain content' from the H5 vaccines used to support the mock-up dossier approval? I haven't found a definite answer to that but reading between the lines, especially the recently updated Revised recommendations for pharmacovigilance plans for H1N1 pandemic, which details the post-approval clinical trials that need to be done, I believe the answer is yes.
So, apart from these requirements for post-approval clinical investigations, no clinical data on either safety or effectiveness is needed BEFORE approval by the EMEA. Assuming of course they do not vary the dose, but since any dose variation - requires clinical trials (how else are you going to find out what is the appropriate dose, without testing it?)
- automatically voids the fast-track option (see box above),
- giving an advantage to your competitors, and
- if you don't test, nobody can say your dose is inappropriate,
how likely do you think companies will voluntarily do dose-ranging studies to determine the safe and effective dose?
Bear in mind that this protocol was developed with H5N1 in mind. All 3 vaccines were approved based on the H5N1 virus, which is notoriously non-immunogenic, requiring 90ug to induce an immune response in the absence of an adjuvant, as compared to 15ug per strain in the seasonal vaccine. The Novartis vaccine uses 7.5ug, GSK 3.8ug, both with adjuvants. I have a suspicion that the low immunogenicity of H5 may have been due to the modification of the HA (to remove/change the part that makes it HP), since the Baxter whole virus vaccine made from wild-type (ie unmodified) virus is immunogenic without adjuvant at 7.5ug.
We have NO idea how immunogenic the vaccine seed virus for current pandemic H1N1 will be. Can data generated from H5N1, a very unusual virus that is/was not well adapted to humans AND genetically modified (except for the Baxter vaccine), be extrapolated to H1N1, a pandemic virus with a totally different origin and much more adapted to humans? The issue of dose-ranging is made more acute by the use of adjuvants, since adjuvants lack universality in their actions, ie how an adjuvant behaves with one antigen may be drastically different when the antigen is substituted. If H1N1 is a more immunogenic strain than H5, and we use the same dose of antigen and adjuvant, what is the risk of excessive immune response causing adverse reactions? I don't think anyone has any idea.
In addition, the current pandemic has a CFR a couple of logs lower than H5, so the risk/benefits tolerance from the public will be correspondingly different. Would parents be willing to let their kids be vaccinated, or would HCWs roll up their sleeves to take the shot in the spirit of public service, if they know the following?
From the EMEA guidance (emphasis mine):
Unless the mock up vaccines are also suitable for use as final pandemic vaccines, ALL the safety data on the final pandemic vaccines will have to come from real life use.
In other words, we/they will be the guinea pigs. Or ferrets. Or whatever...
The reason why this is such a concern, is because, to start with, the standard required for the mock-up dossier submission, is already much lower than what is required for normal vaccines, eg for seasonal or pre-pandemic use. Indeed, the EMEA guidance admits that:
However, even if mock-up vaccines are suitable, the post-marketing safety data will be extremely important, since the core pandemic dossier will be too limited in size for full assessment of safety.
Exactly how limited? The EMEA guidance says the database should be sufficient to detect adverse reactions or events at a frequency of approximately 1%
Is that good enough, to support a mass vaccination campaign without clinical trials? If you vaccinate half the UK population, ie 30 million people, a severe adverse event with an incidence of 0.1% would affect 30,000 people, but this would NOT be detectable with the trials. In any case those were done with a different virus at doses that may or may not be appropriate for the current pandemic. And so on and so forth...
Is that good enough, Sir Liam? In case you didn't notice, this is not H5N1. This is not a catastrophic or even a severe pandemic; we don't have to take drastic measures, at least, not just yet, and not without careful deliberation. It's SWINE flu, not BIRD flu, did anyone brief you on that?
Hello, is anyone home?
Can someone please make sure the UK government has a remote control with a PAUSE button that WORKS? |
