Tuesday, October 25, 2005


It's tiresome, but I guess I'll have to do it. Again.

Freelance journalist Wendy Orent is at it once more, pooh-poohing the threat of avian influenza and, not incidentally, cashing in. Orent gets paid by the word and specializes in being the messenger of "don't worry, be happy," a highly saleable commodity when everyone else is writing about the dangers of bird flu. Since she is one of the few--the very few--hewing to this line, she has a ready made market in newspapers whose idea of "balance" is a version of "on the one hand" paired with "on the other hand." As one of the regulars on PBS's NewsHour with Jim Lehrer once wryly commented about that show's tendency to do the same thing, it's like "And now for another view of the Holocaust . . . " (Disclaimer: I still like the show).

OK, that was snide. But I am irritated by having to waste bandwidth on this lightweight. What's her argument?
News reports make the threat even more ominous. In resurrecting the 1918 pandemic virus, the deadliest flu strain of all time, researchers recently learned that this strain was far deadlier than any other human virus — it killed mice, while normal human flu won't even ruffle a mouse's fur. They also found out that all of its genes came, directly or indirectly, from birds. Unlike the pandemics of 1957 and 1968, the 1918 version didn't arise from a combination of bird and mammal genes. Instead, the bird genes evolved into a human virus that killed as many as 50 million people.

This means, say breathless news reports, that what happened in 1918 could happen again, this time with H5N1. (LA Times)
When last I addressed the Orent story (March 2005) she was making a big deal about the fact that Taubenberger was saying they weren't bird genes. Never mind. There's more. Like the Professor of Aeronautical Engineering at MIT who patiently explains to his class why anything designed like a bumblebee couldn't possibly fly, she impatiently explains to the stupidos at CDC, WHO and the rest of the scientific establishment why she, Wendy Orent, anthropologist, can demonstrate that bird flu couldn't possibly be a problem. In doing so, she doesn't mind lobbing a few misleading rhetorical asides. For example, right after the last sentence quoted above, she goes on to say:
But Peter Palese doesn't think so. He is lab director at Mount Sinai Hospital in New York, where the technique that re-created the 1918 genes — known as reverse genetic engineering — was developed. He and associate Adolfo Garcia-Sastre contend that what the resurrected virus really shows is how supremely adapted it is — how well its parts fit together, how perfectly it works. The sublime malignance of the 1918 virus doesn't lie in one part but rather in how the genes function together. Evolution shaped this virus to be a sleek, effective killing machine.
It so happens I also have talked to Peter Palese about this. He has publicly expressed reservations that the next pandemic will be from H5N1 but not that there will be a next pandemic. He doesn't know if it will be H5, H7 or H9 or something else. He told me his hunch was against H5N1 because of a paper from 1991 that showed 7% seroprevalence against H5N1 in rural China, although not elsewhere. But he also said it was only a hunch and he didn't really know. I disagree with his assessment, but I respect him. What he does not say is that the concern over H5N1 is just hype and misplaced, which is what Orent implies he says.

Orent's main point, however, is that it isn't likely a bird virus can now evolve to be both efficiently transmitted in humans and highly lethal at the same time. The fact it did this in 1918 was a unique outcome of the crowded conditions of soldiers in World War I (citing the speculations of Paul Ewald). Without those conditions, she claims, the virus could never have attained the degree of lethality it did. Her only evidence to this effect is in essence a "thought experiment." If the virus killed quickly and efficiently, there would be insufficient opportunity to be transmitted to another host. This is a simple-minded view that doesn't take account of the complexity of evolutionary dynamics where one has opposing tendencies within the host and between hosts, multigene interactions at different scales and simultaneous polymorphisms within each host. It is known that such dynamical systems often produce counter-intuitive results, so arguing from the crudest level as does Orent can be misleading. But more important than these theoretical issues are the empirical counter-examples.

We know of many diseases that are highly transmissible and also highly lethal in immunologically virgin populations. Smallpox and measles are two examples that wiped out vast numbers of indigenous North Americans when Europeans arrived, and these diseases continued to do so through the last century. More bizarrely for Orent's argument, we have the pneumonic plague (The Black Death) of the 14th century (hardly a time of overcrowding), a disease about which Orent herself has written a book with the scaremongering title, Plague: The Mysterious Past and Terrifying Future of the World's Most Dangerous Disease. Maybe she resents an upstart virus displacing her own favorite doomsday agent.

More to the point, however, is that we don't need H5N1 to be highly and quickly lethal. We only need it to be highly transmissible, which is advantageous in evolutionary terms. The 1918 virus, for example, had a case-fatality estimated to be "only" two or three percent, but when you have infected a significant portion of the world's population (immunologically naive to H5N1), 2% is quite a lot of dead people. Thus if there is a 40% infection rate (perhaps comparable to 1918), we would have 2.4 billion infections, very few (percentage-wise) fatal. But a 2% case-fatality rate (no one's idea of super lethality) is still 50 million deaths. Fifty million deaths is not exactly a tunafish sandwich, I'd say.

In addition, influenza is infectious before symptoms (and hence incapacitation) occur, unlike SARS or Ebola. Orent likes to say that the reason wild ducks are only mildly affected is that dead ducks don't fly. But infectious people do fly--on airplanes. In addition, diseases that don't make animals sick often make people very sick. Thus viruses can continue to circulate in large animal reservoirs like aquatic waterfowl and still sicken incidental hosts like humans or other animals.

I'd go on to address her other arguments except she doesn't have other arguments. Her claim that the jump to virulence requires a long period of adaptation and is unlikely is just plain false. When you have billions of virions experimenting in each host and hundreds of millions of hosts, random mutations of the right kind are not rare. There are several mechanisms to produce this genetic variation, including reassortment, missense mutations and recombination. The latter doesn't even require co-infection, since we know of instances where the influenza virus has changed from Low Pathogenic to High Pathogenic via non-homologous recombination with its own genome. Moreover we now know that the protein changes needed are often extremely small, almost trivial.

But why go on. When you're selling crap by the word, someone's blog post isn't going to stop you. I'm sure we haven't heard the last from Wendy Orent. Too bad for us.