Palese: Don’t censor live-saving science

11 January 2012

peter paleseRenowned influenza virologist Peter Palese has penned an opinion column for the science journal Nature in which he uses his experience in reconstructing the 1918 pandemic influenza virus strain to question the censoring of H5N1 results by the National Science Advisory Board for Biosecurity (NSABB):

My colleagues and I were at the centre of a similar controversy in 2005, when we reconstructed the 1918 flu virus, which had killed up to 50 million people worldwide.

As Palese and colleagues readied a manuscript describing these remarkable findings, the NSABB intervened. Palese explained why publishing the work would allow studies on why the virus is dangerous. The NSABB agreed and allowed publication. That was an important stimulus for work on the virus:

After we published our full paper…researchers poured into the field who probably would not otherwise have done, leading to hundreds of papers about the 1918 virus. As a result, we now know that the virus is sensitive to the seasonal flu vaccine, as well as to the common flu drugs amantadine (Symmetrel) and oseltamivir (Tamiflu). Had we not reconstructed the virus and shared our results with the community, we would still be in fear that a nefarious scientist would recreate the Spanish flu and release it on an unprotected world. We now know such a worst-case scenario is no longer possible.

In light of this positive experience, Palese does not understand why the NSABB today does not want to make public the mutations that allow aerial transmission of H5N1 in ferrets. He believes that knowing the mutations will allow a rapid response if they are observed in nature. His conclusion:

The more danger a pathogen poses, the more important it is to study it (under appropriate containment conditions), and to share the results with the scientific community. Slowing down the scientific enterprise will not ‘protect’ the public — it only makes us more vulnerable.

Why is the NSABB taking a different stance on influenza research today compared with 2005? A major factor may be the perception that the fatality rate of H5N1 influenza virus in humans is 60%, compared with 2.5% for the 1918 H1N1 strain – although the latter is not insignificant. As I have pointed out previously, that assumption is incorrect.

Palese hopes that the scientific community will convice the NSABB to change its mind – otherwise who will enter a field in which you cannot publish your most interesting results?

  • Anonymous

    Looking forward to Peter’s talk tomorrow at Yale University.  Very timely.

  • Wolcllark

    Didn’t University of Wisconsin have a disease escape few years back?

  • Brian Hanley

    Agree completely.

  • http://www.mikebarnkob.dk/ Mike Barnkob

    Is it really true, that this science will not be published? I think, by reading the comments and quotes from the NSAAB just here on your website, it is clear, that their recommendation is simply to not publish the whole methodology, but still the rest of the article. Is this false?

    If the authors get a Nature article out of it, I am sure the authors will be happy about it; and would the moral then be opposite from what Palese suggests, namely that the more dangerous the work and the more stir up you can make, the better chance of going into a high impact journal and the more students you attract? I really don’t think either description is right and it removes focus from an important discussion.

    The really question is how we as scientist deal with dual use science. In this case there seems to be an informed and sciencetific body – the NSAAB – that have looked at the article and made a decision. I might have misunderstood something, but that really seems like a good way of handling this type of science to me. If you disagree with this setup, then it falls to you to suggest a better system. I for one do believe we do have to take into consideration the greater good of publishing our work; that should not be overruled by an ideal of total freedom in what we do in the lab or what we publish.

    Best,
    Mike Barnkob

  • gsgs

    http://www.nature.com/news/don-t-censor-life-saving-science-1.9777

    Palese:
     > We then took an existing influenza virus and, one by one, swapped its genes
     > with those from the 1918 virus, eventually recreating a live version

    could that be easily redone by terrorists ?

     > Giving the full details to vetted scientists is neither practical nor sufficient.

    would he be one of those ?

     > Once 20–30 laboratories with postdoctoral fellows and students have such
     > information available, it will be impossible to keep the details secret.

    so do it in military grade labs only ?!

     > “Who will want to enter a field in which you can’t publish your most scientifically
     > interesting results?”

    this would only be a small special (sub)field.
    I could ask: what research are we,the public, willing to pay for ?

     > Knowing which mutations render the virus more dangerous could help on a
     > public-health level — if an outbreak of bird flu occurs in Taiwan, for instance,
     > and researchers sequence the virus and see those mutations, we would know
     > to ramp up the production of appropriate vaccines and antiviral drugs.

    ahh, if an outbreak of bird flu occurs in Taiwan, then we should ramp up the production
    anyway – no matter if those mutations were included or not

     > Incidentally, I believe that the risk of future outbreaks in humans is low: H5N1 has
     > had the opportunity to cause widespread pandemics for many, many decades,
     > yet it has not done so.

    and I remember that’s exactly why Palese concluded in 2006 that H5N1 were just not capable
    of this. Does he still think so after the Fouchier/Kawaoka experiments ?
    Mexflu needed 13 years to take off. Equine H3N8 needed 40 years before it jumped to dogs.
    H5N1 had had not enough opportunities yet to evolve 10 or more generations in humans
    as Fouchier did with the ferrets.

     > Although we know the virus is transmissible between ferrets,
     > little is known about how it will behave in other animals, including humans.

    so how can Palese conclude :
    “Incidentally, I believe that the risk of future outbreaks in humans is low:”
    “little is known” is no negative evidence

  • Goldensail

    A few points:

    H5N1 has had ample time and opportunity to evolve in humans.  The first well documented H5N1 outbreak in a poultry farm occurred in Scotland in 1959 – and surveillance was not happening to any appreciable degree before this.   Think about the practicalities involved in “documenting” an outbreak and then think about the likelihood that such investigations would happen in the early 1900s and before in the areas that are most likely to have H5 outbreaks – rural farms in developing countries. 

    Ferrets are a very sensitive system for studying influenza viruses – they host many influenza viruses from other species (swine, equine, avian) that have never been shown to cause disease in humans.  They are a fine system, but certainly no surrogate for humans.

    No, generation of a virus from sequence information could most certainly NOT be done easily by a terrorist.  Virus rescue is a sophisticated method that requires a enormous resources and technical know-how/training.  Also, it would be a totally impractical weapon – not able to be targeted to a specific population and would probably infect the people who made it first.

  • gsgs

    1959 was a different virus, it reassorted a lot since then.
    Reassortment often is the key for flu-pandemics. One double infection
    with H5N1 and a suitable swine or human virus … very few opportunities
    for that so far.

    So what might terrorists do ? Passaging it repeatedly through ferrets or swine
    or even humans without those starter mutations.

    A freezer and patience could be a good terrorist weapon, as we saw in 1977
    (and almost 2005 with H2N2)

  • http://rybicki.wordpress.com/ Ed Rybicki

    “A freezer and patience could be a good terrorist weapon, as we saw in 1977 (and almost 2005 with H2N2)”…I assume you mean the 1977 H1N1 that MAY have come out of a Russian biowar lab freezer – and the H2N2 that got accidentally mailed around, to reputable labs as it happens?

    Both those incidents relied on infrastructures that could both isolate, culture, and reliably store what is not a particularly stable virus – something the average mid-East cave- or Montana enclave-dweller might struggle with a bit.

    Not only do I think the Fouchier etc work should have been done, I think some H5 reassortant work should too: what is the animal pandemic H5N1 compatible with, and what happens when it reassorts?  As an example, the H5N3 “tern” virus characterised from South Africa in 1961 was linked anecdotally with the 1959 Scottish chicken outbreak – and was also an HPAI.  These things do get around…!

  • Goldensail

    “1959 was a different virus, it reassorted a lot since then.”

    I think that is an important point and, really, THE point when we are talking about the likelihood that H5 viruses can be serious human pathogens…  H5 viruses have existed in close contact with humans and other mammals for a long time and have reassorted and mutated over time as any good flu subtype does.  Still, none of the hundreds of thousands of H5 variants that have emerged have become sustainable human pathogens. 
    No one can say with certainty that it cant/wont happen, but it just seems very unlikely given the mutation rate of flu viruses and the ample opportunity for recombination in mammals that H5 viruses have had.

  • gsgs

    Goldensail wrote, in response to gsgs :

    >> 1959 was a different virus, it reassorted a lot since then.

    > I think that is an important point and, really, THE point when we are talking about the
    > likelihood that H5 viruses can be serious human pathogens…  

    this is not disputed, H5N1 is very pathogenic in humans.
    You mean the likelyhood that it goes pandemic ?
    That’s very much disputed. And that’s what disturbs me about profvrr, he doesn’t
    argue so much that likelyhood but argues the expected CFR instead !

    > H5 viruses have existed in close contact with humans and other mammals
    > for a long time and have reassorted and mutated over time as any good flu subtype does.

    … in birds.

    > Still, none of the hundreds of  thousands of H5 variants that have emerged have
    > become sustainable human pathogens.  

    so did the millions od variants of H9,H7,H4,H6 pathogens

    > No one can say with certainty that it cant/wont happen,

    nothing is certain in the real world outside mathematics. So we need to estimate the probability.

    > but it just seems very unlikely given the mutation rate of flu viruses and the ample opportunity
    > for recombination in mammals that H5 viruses have had.

    recomibination we call in influenza the rearrangement of RNA chunks in one segment during
    coinfection of several strains. This is unlikely in flu (disputed).
    You may mean reassortment, which is the re-shuffling of the 8 segments during coinfection,
    and which often is the source od new dangerous strains.
    This is also rare, even rarer in humans than in birds. And most reassortments are bad viruses
    and eliminated by evolution. But those few that survive pose a great danger.

    You can’t really argue that H5N1 had had ample opportunities to reassort in humans already.
    Nor these 10 or more passages that Fouchier did (without reassortment).
    I doubt that we had had more than 5 passages human to human with H5N1 in nature.

  • Pingback: Moratorium on influenza H5N1 transmission research()