A bad day for science

I look forward to a holiday season explaining this controversy to family. I was discussing it with my father several weeks ago when the story broke and used many of the same points as ya’ll in TWiV. Not being a blogger (or terribly professional as far as social media) I think this is the most effective way to spread rational discussion of this issue. Other ideas?

Has this ever happened in any other fields outside of virology, where ‘independent’ groups have effectively shut down communication of research? And are examples of times when we have been grateful that specific knowledge was shared that could have also been prevented from being released into the public domain?

You have my complete agreement! Below is my letter to the NSABB.
========================================================
Dear NSABB:
    I was not able to locate emails for all of you. The list of missing
members is at the end. My sincere apologies to those not included.

Comment

    I would like to point out a couple of things. 

    First, the part of the study that I have read which is being
considered for redaction is specific mutation point sequence info. This
information is useful to scientists and diagnosticians. It is not
useful to Violent Non-State Actors (VNSA’s) we are concerned with for
reasons I will explain.

    The kind of parties who could make use of that information to
manufacture an organism would require high tech equipment and
laboratory access. This requires money and considerable education as to
how to use it.  It requires a significant ability to creatively solve
problems. In the context of polio, smallpox and the 1917 virus genomes
being public information, the addition of this H5N1 data does not
represent anything significant. If those parties were going to
manufacture an organism using sequence data, they would already be
doing it. Please note that there is plenty of money, and there are
laboratories with such equipment today, in regions known for financial
support of such terrorist groups.

    The VNSA’s of concern are almost entirely of quite limited
technical education with very few exceptions.  Those parties do not
have a scientists and labs at their disposal to make use of such
information.  They have already declared that they are seeking
biological weapons. There have been fatwas approving of the use of CBW
on enemies. A certain gentleman is in prison in England for such a
fatwa. If those VNSA’s had the capacity to manufacture a bioweapon
using sequence data, they would have used that capacity already and
done so. So far they have not produced a bioweapon with such equally
dangerous sequence information.

    Yes, not having deployed a bioweapon yet (presuming Dr. Ivins guilt)
does not mean they never will. Perhaps they are trying to make a
weapon now, we cannot know for sure. But if
they are doing so now, they are using previously published information.
This addition doesn’t matter much. It’s just another lollipop in the
candy store for such parties.

    Second, what is useful information to the VNSA’s we are concerned
with is the general method. Infecting an animal and passaging from one
to the next is not hard. It does not require high tech apparatus for
the most part. It is something that could be done in a cave in the
mountains, or in a compound in a city. It could be done in a basement
anywhere in the developed world. For people unconcerned with human
life, it could be done using people who were either prisoners taken in
battle, kidnapped, tricked, or even volunteers for the cause. It could
be done in a few tents.

    That passaging technique has been published to the general public
twice before. Laurie Garret had it in one of her books, I think it was
“Betrayal of Trust: The Collapse of Global Public Health”. I remember
reading it also in Ken Alibek’s “Biohazard”.  The passaging method has
now been published in several semi-popular articles about the
deliberations of the committee on this matter. Thus, that horse is out
of the barn and well into the next county. Not much point to barring
the barn door now. Best to suck it up and improve readiness.

Conclusion

    I think the greatest concern is that the general method of
passaging microbes hit the mass media. The VNSA’s we are concerned
about are well plugged in to our mass media, not so much into
scientific journals. I am all too aware of the weird goldfish bowl that
US news media create to those outside our borders. While we are barely
aware of them, we are in their face every day. Just turn on the TV or
browse CNN overseas and it’s all right there, from Kim Kardashian’s
tribulations to Congressional testimony.

    Yes, one could ask if the VNSA’s would have problems transporting H5N1. But I imagine that they would solve that using suicide human vectors. Those gentlemen can cut the Gordian knot of biosafety which we worry over so much. For them it would be a kill ratio calculation. Just
like any other commanders, they want positive kill ratios for their expenditure of personnel. A human biodeployment would be highly positive, much more positive than a car or truck bomb. Note that human suicide vectors would be the best way to introduce human transmissible
H5N1. I could go on about what they would likely instruct their suicide vectors to do, but I don’t think that is necessary. Our best bet if they do it is that we catch it early.

    Our primary concern is response capability. That capability is best served by dissemination so that there is no problem with diagnostics or tracking of H5N1 that is close to natural evolutionary crossover. After all, nature is quite the bioterrorist, with no human help. Is nature the worst bioterrorist? So far, yes. Not the worst in theory, but so far it is. We must keep in
mind that medicine is bioterrorism response and diagnostics are surveillance.  We should not interfere with those.

    Consequently, I think our biggest risk is creating the problem the committee was formed to prevent because of deliberations generating publicity about the general method.

it depends on the details, which we don’t know.
And on the cost to repeat the experiment.
Well, it took them almost 10 years
How much would publishing the details reduce the cost
to repeat it, to create the strain ?

they can threaten to use it to pressure governments
for money or to withdraw troups or free jailed friends.
We saw it before. If they are desperate enough, they may risk the pandemic.
They may even hope that the epidemic would be contained.
But we won’t know it in advance and the threat would still be
too terrible to even risk it.

Check out the December 21 NPR story on this decision.

We have known about this virus since 1997. The first of many papers that showed a successful protection by a vaccine candidate against HPAI H5N1 in a mammlian system appeared less than a year later due to concerns about the pandemic potential of the virus: Protection against a lethal avian influenza A virus in a mammalian system. Riberdy JM, Flynn KJ, Stech J, Webster RG, Altman JD, Doherty PC. J Virol. 1999 Feb;73(2):1453-9.
The same year saw successful vaccine candidates in ferrets published.
My point is, even if the virus the Fouchier and Kawaoka groups created were a threat we already have many vaccine candidates for it that could be put into production. I wonder if there is a expt. in either paper showing protection by such a vaccine.

“The only thing we have to fear, is fear itself.”

This security council is blinding itself to the reality that these experiments are not so cutting-edge that any gifted molecular virologist would be able to get to the same point within a few years or maybe even months if they really want to.

This restriction is disturbing to a lay person because I agree it’s better to have open discourse in the research community.  In this case the crux seems to center on ferrets being infected and the impression that an infected animal can readily transmit the infection to humans.  After all that’s what happens with rabies, right?  Lots of discussion about why this isn’t so with H5N1 would really be helpful.  All I can argue, and perhaps any politician or lay-official can argue, is that so far very few human cases have resulted this way through contact with birds (ducks only?) and in those cases the contact was extreme, such as sleeping nightly in the same cage as the ducks.  Are there some good analogies for this sort of thing, like one I heard that during WWII 10,000 bullets were fired for each person killed.   Research done on every bacteria and virus-borne disease entails risk and periodically a researcher dies in the process.  Would a presentation of some of these efforts help put the current research in a more reasonable light?  Let’s agree that there’s some risk.  That’s true with everything: look at coal mining with cave-ins and black lung!  Is the risk with this H5N1 work reasonable and in line with other research — on polio, on yellow fever, on anthrax, on smallpox, influenzas….  That’s my 2-bits worth.

As a virologist and a member of NSABB, I would like to make a few comments mainly to clear up some misconceptions that have crept into this discussion.

First and foremost, there is absolutely no doubt that this research had a very legitimate goal, to understand how the avian H5N1 virus might become transmissible among humans, using the well-established ferret model system as a surrogate. This has very important implications for the public health community worldwide. One should remember that until this work, it was unclear whether H5N1 even could be transmitted among mammals. These papers show that it can happen.

Second, this is not the same as the debate over the reconstruction of the 1918 virus in 2005. That virus was much less lethal in man. Moreover, there is some immunity against H1N1 in the population and medical care has improved greatly since 1918. Despite these advances, however, the current H5N1 mortality rate is in the 60% range. Might that rate drop and the virus become less virulent as a whole  if it does indeed begin to exhibit human-to-human spread? It’s possible, but we have no way of knowing that. I, for one, am not anxious to see that human experiment take place.

Third, I object to the use of the term “censorship.” This is not censorship, but rather our responsibility as a scientific community. If H5N1 were to start spreading from human to human, be it by natural or man-made means, we are not prepared to handle it. There is NO pre-existing immunity and a reliable vaccine is not available, nor are the quantities of antiviral drugs that would be required to deal with a pandemic that would take only a few short months to establish itself. In this context, I agree that it is key that people with a legitimate need to know this information must be given access to it. And this is not limited to the influenza virology community. I know that both Nature and Science are insisting that this be the case if they are to agree to witholding some of the work from publication.

Finally, let me comment on the argument that terrorists would not use an agent whose spread cannot be controlled. One needs to look no further than suicide bombers to know that many modern day terrorists do not care about collateral damage or even their own lives. Moreover, it is well known, in public sources, that terrorist groups have expressed an interest in developing biological weapons.

At the end if the day, the possible consequences of this virus getting loose have to be balanced against the benefits of publication the papers “as is.” Given the current state of the science, public health preparedness, and the world, my colleagues at NSABB and I believed that the prudent approach was to use caution. We (the scientific community) owe it to the public, who both fund our work and benefit from it, to demonstrate that we are acting responsibly. I would hasten to add that, if these papers are published with no editing and there is a terrorist event with this virus, one can only imagine what the response of the U.S. Congress and other international legislatures might be in terms of putting Draconian restrictions on research. 

Now we know that the ferret-passaged influenza H5N1
virus can be transmitted to people and cause a deadly disease. The detailed of
how to do it has been removed from the publication, but the idea has been seed
and any one can redo the experiment and select for a very virulent strain of
H5N1. So, the discussion should be, what should we do to protect the population
from a H5N1 man-made or nature-made?

Very well put Brian. I especially agree with the notion that nature is the ultimate bioterrorist, and that we should not interfere with medicine and diagnostics.

Many thanks to Vincent and Mike for this blog.

There is a very practical problem with the NSABB advice: how to share the details of the studies with the scientists and public health specialists with a “legitimate need to know”? The NSABB advice is (perhaps somewhat) understandable, but is at the same time impractical;  The problem is that the critical information can not be shared with dozens of countries (hence 100s of scientists) while keeping the details “classified”. As soon a you share confidential info with more than 10 people, the info will be on the street. If NSABB agrees that the details of the info need to be shared with people with a “legitimate need to know”, the only way to do that responsibly is via the usual channels of scientific communication; publication in respected journals.

In Dual Use research, one needs to way “benefits” and “risks” of research. For the H5N1 studies, the benefits are clear; trying to prevent a severe pandemic or reduce the impact if prevention fails. The risks? Negligible. Bioterrorists and rogue countries have easier opportunities than to recreate these H5N1 viruses; Mother Nature is the biggest bioterrorist, and the weapons are in the field for grabs for everyone.

PS. At the time of the NSABB advice on 1918 Spanish flu, the NSABB had no knowledge of herd immunity against H1N1. That information came from the 2009 pandemic. In 2005, the NSABB had to assume that 1918 Spanish flu virus was a real killer also.

Ron, thanks for weighing in here with very important points. I don’t know how the US government, in cooperation with the rest of the world, is going to deal with the very important practical issue that your raise. Similarly, your point about what an actor (state or non-state) might choose as a bioweapon is spot on and has been the topic of much discussion by NSABB (and others) over the years. No matter what we do, we are not going to be able to stop a determined actor: the best we can do is try to make it as difficult as possible for them to act until we have better countermeasures.

Thanks also for pointing out the H1N1 immunity issue: you are correct about the state of our knowledge at that time.

Mike, I apologize if I misinterpreted you. But please re-read your comment above – which was the basis for my response – and try to understand how your statements sound to someone outside your group.

So 2 questions has need to be discussed as well. The first has already been brought up with no answers yet. How do you share the full data with scientists who want to work on this problem and what are the criteria for them working in it? But the next step would be to publish a new result based on that classified data by another group. New antibodies that target novel mutations, or novel interferon antagonists in the passages flu. How do you publish on this in the future and still keep it secret?

The second question is whether these 2 viruses have the same set of mutations that make them transmissible in ferrets or are they different mutation sets that get to the same result? Without knowing the mutations then there is no way to know between the 2 groups whether they are the same orations types or not. In SARS CoV for example there are different but overlapping mutation sets between 3 different mouse adapted viruses that cause the same lethality and pathology. It would be interesting to know how this compares to Influenza. And I don’t think we will get that information from these publications.

Matt

 restrict/concentrate funding for this special issue to certain groups
(Fouchier should like that ?!)

Antivirals that target special mutations ? Are there those (I haven’t seen
any), is it possible,likely,important ?

Apparently we cannot overcome the resistance mutations.

The 2nd question could be answered without increasing terrorism risks
(much)

Well, if they were different, that should make it even easier…

I accept as a scientist and am relieved as a citizen that the putative impact of scientific novelties is controlled by society as a whole (this is the case in democratically elected governments), and not only by scientists themselves. If this means that now and then censorship is imposed on the description of scientific novelties that might be of great harm, so be it.

And, by any scale, here the balance of danger X freedom of publishing scientific results is clearly defined: among all man-made provoked catastrophes, perhaps only a nuclear war is in the same level of global threat as this pandemic H5N1 kind of virus. Even if this engineered strain were a remote threat (and it is not),  it indeed “constitutes a legitimate interest” for society to take precautionary measures.

Therefore I am glad that NSABB is taking this matter seriously (and among one of the measures that seems to me reasonable is suspending the spread of more information about this newly discovered technical possibility to prevent it to fall in wrong hands too quickly).