Chilled Virology
"On reading," by Simon Wain-Hobson, is a weekly discussion of scientific papers and news articles around gain of function research in virology.
Since January 2024, Dr. Wain-Hobson has written weekly essays for Biosafety Now discussing risky research in virology. You can read his entire series here.
This OpEd from US National Institutes of Health top brass was published three weeks after the first meeting of the US NSABB which had been asked to look at manuscripts by Drs. Fouchier and Kawaoka describing H5N1 avian flu engineering.
The opening gambit is the sharp contrast between the deaths from seasonal flu which represents less than 1% of those infected and the lethality of non-transmissible H5N1 flu virus infections – something close to a staggering 50%. If ever H5N1 adapted to efficient airborne transmission the impact could be colossal.
Space requirements no doubt prevented the authors from writing more. Unfortunately, they missed adding two important points making for a misleading story. Up to the end of 2011 H5N1 infection among humans had been clearly documented for 592 cases. This equated to 592 x 50%, or approximately 296 deaths. By contrast, seasonal flu kills anywhere between 290-650,000 people worldwide. Conclusion? A highly transmissible virus that kills less than 1% is way more dangerous than highly pathogenic viral spillovers. Not complicated; it has always been this way.
The second point is that infectious disease people often talk about a tradeoff between pathogenesis and transmissibility. It’s not set in stone, but it happens. For example, compare the dangerous Delta variant of the COVID-19 virus which was supplanted by the milder but much more transmissible Omicron strain. Indeed, Fouchier and Kawaoka talked up such a tradeoff.
A subsequent DURC/GOF paper reported an engineered H7N1 bird flu strain was capable of airborne transmission without loss of 60% lethality in ferrets. For the macabre record, this is less than HIV without anti-virals which is close to 100%, but whopping lethality for any airborne infection - around 30X greater than Spanish flu. Despite this no regulatory or oversight dog barked even though the work was funded by a grant from the NIAID. There was a microdiscussion.
About these novel lab viruses, We cannot predict whether it or something similar will arise naturally, not when or where it might appear. Exactly. Can we stop here?
In defining the mutations required for mammalian transmission, public health officials are provided with genetic signatures that, like fingerprints, could help scientists more readily identify newly emergent, potentially harmful viruses, track their spread and detect threatening outbreaks. Public health officials can’t handle these mutations. Experiments on natural viruses are way more informative (Getting it right). This reflects the belief that the pathogenesis of a virus can be read from its genome, which it can’t (1918 and all that). Just learning of the complexities and subtleties associated with disease of avian flu A viruses in chickens, geese, ducks, mice and guinea pigs makes light of this belief.
And just to ram home the point about genotype to phenotype predictions, several papers have shown that the H5N1 adaptive mutations described by Fouchier didn’t confer the same phenotype in another strain of H5N1. Such a result was always on the cards and yet doesn’t get a mention unless it was covered by We cannot predict whether it or something similar will arise naturally, not when or where it might appear.
We learn that Identifying threatening viruses can also facilitate the early stages of manufacturing vaccines that protect against such a virus in advance of an outbreak. Virologists were rightly scared of H5N1 from when it first raised its ugly head in 1997. Period. They didn’t need the Fouchier and Kawaoka work to get going on making vaccines to H5N1 avian flu. Proof? A PubMed search shows 1384 hits for ‘H5N1, influenza, vaccine’ from 1997 to 2011, the year before publication of the Fouchier and Kawaoka papers. What is this sentence doing here?
Then the authors get to the crunch. The question is whether benefits of such research outweigh risks. The answer is not simple. A highly pathogenic bird flu virus transmissible in humans could arise in ways not predicted by laboratory studies. And it is not clear whether this laboratory virus would behave in humans as it does in ferrets. So even before the publication of these controversial studies it was clear that the answers were anything but simple.
Rather than accepting this, the OpEd continues with a nonetheless, which leaves On reading dubitative. Nonetheless, new data provide valuable insights that can inform influenza preparedness and help delineate the principles of virus transmission between species. After 27 years of keeping tabs on H5N1 bird flu viruses, they are not spilling over to humans any more than before. Indeed, since the Egyptian outbreak of 2015, there have been very few cases.
For the moment, avian H5N1 flu doesn’t present much of a risk (Getting it right). Equally, long entrenched horse and dog flu viruses haven’t seeded human outbreaks or epidemics.
The OpEd closes with talk of Safeguarding against the potential accidental release or deliberate misuse of laboratory pathogens is imperative. Throughout the DURC/GOF controversy, very few have tackled publicly the dual use part, aka how to stop a malicious mind doing something dangerous or stupid with this information. We learn that scientists, journal editors and funding agencies involved are working together to ensure that access to specific information that could be used to create dangerous pathogens is limited to those with an established and legitimate need to know.
So what happened down the line? The papers were published without redaction of any specific information some 5-6 months after these words were written. Need to know ended up meaning the whole world. This includes the H7N1 flu virus experiment where 60% lethality in ferrets was not traded upon acquisition of airborne transmissibility.
Thinking it through, it does makes sense that the whole world needed to know about such risky research.
The OpEd also made an error of judgement. It preempted future discussion by stating that the risk was nevertheless worth taking. It was published while the Fouchier and Kawaoka manuscripts were confidential. Only a handful of people had seen them. Yes, Dr. Fouchier had given a talk in Malta where he had talked up his work, although scientists know that only by reading a paper can you apprehend the fine details. And while a cliché, the devil is often in the scientific details. Virologists couldn’t forge an opinion either way which was unfair and unscientific.
Administrators are enablers not gatekeepers to science. It is for the field to decide the outcome of the science, not those with vested interests having funded the Fouchier and Kawaoka work. And if this takes time, say several years, so be it. As is now understood, the papers didn’t deliver on the promises made by them, that is the prediction of the next pandemic strain and with it the generation of preemptive drugs and vaccines. This has been dealt with.
The corollary of such a public OpEd was the chilling effect on anyone who did not share their views. The NIH is such a huge funder in infectious disease research, many thought twice before speaking out against such work. This is not an opinion. More than one US virologist has shared their fears with On reading or others. This no doubt led the NSABB to change its initial recommendation of redacting the key data concerning the mutations that conferred airborne transmission.
From the beginning, we had an OpEd which highlighted the huge unknowns associated with the Fouchier and Kawaoka work but pushed on. Nonetheless. More flight from reason.
 | A guest post by
|
Great to read these words from a stalwart virologist!
"We cannot predict whether it or something similar will arise naturally, not when or where it might appear".
This seems clear even to someone with minimal knowledge of the field. As the author writes, the discussion should have ended there.
Two things as a virologist who has worked extensively with avian influenza viruses: 1. I believe avian H5N1 is not as deadly as they make it out to be. Remember they only count laboratory confirmed cases but we know as with any infectious agent there are people who may get subclinical infection and/or mild infections. Those people will not seek medical care and even if they do they most likely won't be tested to determine the etiologic agent of their illness. 2. I worked on a project where we showed avian H5N1 viruses had varying pathotypes in ducks and were able to map the lethal phenotype to several amino acids in one of the influenza genes. So it is possible to find markers of pathogenesis. That being said, that doesn't mean it will translate across species as we now there is host range restriction for viruses.