[AR] Re: ... Coronavirus
- From: Norman Yarvin <yarvin@xxxxxxxxxxxx>
- To: arocket@xxxxxxxxxxxxx
- Date: Sat, 21 Mar 2020 22:30:12 -0400
On Sat, Mar 21, 2020 at 04:26:49PM -0400, Henry Spencer wrote:
On Sat, 21 Mar 2020, Norman Yarvin wrote:
There's nothing set in stone about the length of time to field a
vaccine. The often-quoted 18 months (or "at least a year") assumes the
usual approval process: Phase 1 trial, then Phase 2 trial, then Phase 3
trial -- all hurried up, but still basically business as usual. Really
we could put a lot of candidate vaccines out there as soon as they could
be fielded (starting basically now; at least one is already in a Phase 1
trial), and collect statistics on their safety and efficacy as we went.
And if your prototype vaccine turned out to kill more people than the
virus? That is unfortunately *not* an unrealistic possibility.
If that looks like it's starting to happen, you stop giving it to
people. It'd be hard to kill more people with a candidate vaccine
than the hundreds per day that are currently dying of the virus in
Italy. (Not just figuratively hard, but literally hard: you might be
shot if you tried to persist with that obvious a failure.)
The guys doing vaccine development are not following an abbreviated
version of the standard process because they think it's business as usual,
or because they just don't know any better.
Oh, they're doing it because that's what they're supposed to do. The
standard process is enshrined in law and in regulations. Not only
would they be going against habit to do anything else, they might be
put in jail for it, especially if anything went wrong. That's where
the need for political support comes in: rules would have to be
waived, and perhaps laws rewritten.
There are real reasons for
doing some preliminary testing, especially for safety, before unleashing a
new vaccine on the general public. Yes, even in an emergency.
We don't have to do that preliminary testing via the usual procedures,
though. Those procedures were never designed for an explosive
pandemic; they evolved via years of experience, almost entirely with
drugs of much less urgency. That wasn't just innocent, pure
scientific experience, either: plenty of political witch hunting,
lobbying by big pharma, and bureaucratic empire building have gone
into those rules.
The ramp-up should be done somewhat cautiously, but that wouldn't mean
taking anything like a year. If, say, a new factory came online that
could ship a million doses a week...
You really wouldn't want to invest that level of effort in a prototype
vaccine that might not work, or not very well (another quite realistic
possibility -- see the sad history of attempts at AIDS vaccines, and the
hit-or-miss record of flu vaccines).
Yes, I figured it went without saying that nobody would try to start
up a factory capable of millions of doses until thousands of people
had already been dosed from laboratory production and the results from
that were looking good. But even then, you'd want to be cautious in
using the output of the factory: even if its production procedures
were supposed to be identical to those of the laboratory production,
something might have been lost in the scale-up.
By the way, relative to the amount that lockdowns and quarantines are
costing, building a factory (or even ten of them) is not a large
expense. The shortage is more of the skilled people that'd be
required.
As for the flu vaccine, I think that too could benefit from some
serious acceleration in testing. The usual reason one hears for it
being ineffective is that this year's strains are not the ones that
were predicted. It would be nice to be able to redo it quickly to
match the strains that actually are circulating, instead of having to
live with the guesstimates of six months ago.
As for AIDS, that's in the category of diseases that people don't
naturally develop protective immunity to, which always makes a vaccine
difficult. Tuberculosis and malaria are also in that category; they
first tried making a vaccine for TB soon after Koch's discovery of the
causative germ in 1882, and they still don't have a good one. But
although there are rumors of people getting Covid-19 twice, they seem
to be baseless.
Drug design is often quite complicated, but vaccines can be simple:
take a piece of the virus, mix it with an adjuvant, and inject...
Unfortunately, this is one of those areas where a simple concept doesn't
necessarily translate to a simple real implementation. One particular
complication for this virus is the strong possibility that the bad cases
are caused, not by the virus itself being nasty, but by the body's
defences overreacting in not-quite-appropriate ways. (That's thought to
be what happened with the 1918 Spanish Flu.) You *really* don't want a
vaccine that triggers that...
Yes, that's what happened with the old RSV vaccine: it actually made
the disease worse. They used formaldehyde-killed virus; the
formaldehyde likely damaged the viral proteins enough that the immune
system produced dysfunctional antibodies. We probably wouldn't use
formaldehyde-killed virus in the first place these days, but disease
worsening still would be a danger to watch out for. It'd be one of
many reasons to focus the first testing on the people the most at risk
from the virus.
You also don't want to immunize people against a part of their own
body. They try to study multiple sclerosis, for instance, by
immunizing mice against their own myelin. It quite reliably gives the
mice severe nervous system problems. (It's a very poor model for MS,
but that's another topic.) In humans, a problem from vaccines has
been Guillian-Barré syndrome, also an attack on nerves but with very
different symptoms and disease course: it often kills people, but if
it doesn't they usually make a full recovery. The 1976 swine flu
vaccination produced many cases of it, and flu vaccines today still
produce the occasional case. But it manifests itself quickly (usually
within two weeks), so is not some hidden danger that might take a long
trial to discover. Indeed, in general, being immunized against
yourself is not a subtle thing.
We would need politicians willing to take scientific risks, though.
The decisions on this mostly aren't being made by the politicians. Thank
heavens for small mercies. The more ignorant politicians would be happy
to promise a vaccine next month (and then blame somebody else when it
didn't materialize); it's the public-health people, who understand the
problems, saying "18 months, maybe a year if we're lucky".
The same public health people who have beclowned themselves by telling
people that masks are useless?
For better or worse, as this pandemic progresses we can expect
politicians to start to take the attitude that Clemenceau expressed
during the First World War, that "war is too important to be left to
the generals". Indeed, in a lot of wars, generals have had to be
fired and replaced. The factors that lead to promotion in peacetime
can be entirely different from the ones that give success in war. The
same goes for public health officials and extraordinary pandemics:
they may talk a good game prior to the event, but that's different
from playing a good game.
Other related posts: