Michael Osterholm, his face a pink-cheeked scowl, looked out across the table, beyond the packed room at the New York Academy of Sciences, and out through the windows. The New York Academy of Sciences is housed on the fortieth floor of 7 World Trade Center, and their endless bank of windows affords a staggering view of Manhattan, Brooklyn, and New Jersey. One reason that its view is so magnificent is that there’s a huge gap in the skyline–and a huge gouge in the ground–where the Twin Towers once stood.
Osterholm had come here from Minnesota, where he runs a research center for infections diseases and terrorism, to talk Thursday night about the threat of a new kind of flu sitting in labs in the Netherlands and Wisconsin. In nature, it’s a flu that spreads easily between birds but doesn’t travel well from human to human. The Dutch and Wisconsin scientists had found ways to get this bird flu, known as H5N1, to move between ferrets. For Osterholm, ferrets were uncomfortably close to humans on the evolutionary tree. And so he, along with other members of an advisory board, issued a recommendation in December that key information in the papers about the research should be left out.
Osterholm looked out at the empty space beyond the windows. “Who would have imagined that you could use box cutters to take down the World Trade Center?” Osterholm asked. The risk from the new bird flu might seem equally unlikely, he warned, but it could end up being far more devastating. “We can’t afford to be wrong.”
The bird flu controversy first started to bubble up in September, when Ron Fouchier of the Erasmus Medical Center in Rotterdam described some of his unpublished results at a scientific meeting in Malta. It kicked into high gear when the National Science Advisory Board on Biosecurity issued their ruling, which Fouchier and Yoshihiro Kawaoka have agreed to. In January, the researchers agreed to stop doing any H5N1 research for two months, during which time the scientific community would try to come up with a plan about how to deal with such controversial research.
Viruses very often spark controversies, but often the controversy is between the scientists who study them and groups of people beyond the academy. Think of HIV denialism, of the non-existent link between vaccines and autism, of the purported connection between the XMRV virus and chronic fatigue syndrome. The new bird flu controversy is different. It’s split the scientific community wide open. I’ve written about this controversy in recent weeks over at Slate, as well as here at the Loom. Like most reporters covering the story, I’ve sampled the sharply opposing viewpoints of scientists over the phone or via emails. But on Thursday night, we got to see this debate in person. The New York Academy of Sciences brought together a group of experts to talk about new virus, and whether self-censorship is a prudent protection or a dangerous precedent. I wasn’t sure what to expect; I was a bit worried it might have turned out to be a fairly dry discussion of how to inspect the hood equipment in virus labs. Instead, we witnessed explosive confrontation between scientists who think we may be facing a world-destroying catastrophe, and others who think our fear of non-existent threats is going to destroy science’s power to help us out of clear and present dangers.
The panel included two members of the National Science Advisory Board on Biosecurity: Michael Osterholm and Arturo Casadevall of Albert Einstein College of Medicine. They both made it clear that they were speaking at the meeting as individuals, rather than as official spokesmen for the board. But they presented a fairly united front. The board has been around for eight years, and it has only considered issuing a recommendation twice. The first time was in 2005, when scientists unearthed the bodies of victims of the 1918 flu epidemic, which killed an estimated 50 million people. The researchers isolated the 1918 virus and sequenced its genes. The board decided they had no objections about publishing the research. But six years later, they decided that, as bad as the 1918 flu might have been, the risk of an H5N1 outbreak was worse.
One big factor in their recent decision was the mortality rate when H5N1 gets into people. The World Health Organization’s official estimate is 60%. The 1918 flu, by contrast, had a death rate of about two percent. If H5N1 could gain the ability to spread among humans–either naturally, or through a lab experiment–it could bring that fearsome death rate to the entire world. “It’s the lion king of infectious diseases,” Osterholm said, no doubt dismaying Disney lawyers across the country.
Sitting a few seats down the panel from Osterholm was Peter Palese, one of the world’s leading experts on flu, who works at Mount Sinai Medical School. Palese disputed Osterholm’s apocalyptic warnings. Where Osterholm burned hot, Palese kept cool, but he did not hide his utter rejection of the board’s decision. Just because a flu virus can be transmitted by another mammal species, he argued, doesn’t automatically mean it can spread among humans. In fact, ferrets are rather delicate in the face of a flu infections, easily suffering from brain damage. Our closer relatives among the primates, by contrast, don’t get sick from flu at all. (Jon Cohen explores the ferret question in depth in a news article for Science.)
Palese also questioned whether H5N1 is all that dangerous. He argued that the World Health Organization based its mortality rate only on the people who came into hospitals and tested positive for H5N1. But this particular strain of bird flu mostly strikes people in poor countries, especially in southeast Asia, where medical services are scarce. The people who make it to a hospital could well be a small fraction of all the people who come down with H5N1.
“The asymptomatic people are not being counted,” Palese said. If those extra people only got sick for a few days and then got on with their lives, the true mortality rate might be far less than 60% “It’s really much lower,” he said, pointing to surveys in Thailand and other countries that revealed evidence that a fair number of people had been exposed to H5N1 at some point in the past. (Palese recently published this same argument in the Proceedings of the National Academy of Sciences.)
This argument positively enraged Osterholm. He had clearly read Palese’s recent PNAS commentary and had prepared a rebuttal. “What you’re saying is just propaganda,” he told Palese. The trouble with Palese’s numbers were that they came from lousy studies, Osterholm argued. There are many ways to overestimate how many people have been exposed to a particular virus. A common test involves fishing for antibodies in blood samples. If your test isn’t precise enough, you may end up dredging up antibodies to other viruses. Osterholm had gone through surveys of H5N1 exposure, setting aside the lousy studies and tallying up the results from the best of the bunch. He came up with an estimate of .6% or less. If very few people have been exposed, the recorded deaths from H5N1 represent a frighteningly high rate.
Casadevall granted that perhaps H5N1 wasn’t 60% fatal. But it could be half that and still be a planetary nightmare. Even if it was ten times lower, it would still be far worse than the 1918 flu. “The numbers of unbelievable, any way you look at it,” he said.
Palese was unmoved. The new H5N1 viruses might pose a risk–a small one, in Palese’s mind–but scientists could handle it. All the research that had triggered the controversy wasn’t conducted in someone’s backyard. It was carried out in well-protected labs. Palese noted that the board doesn’t seem to have any objections to the work that’s done these days on smallpox, a virus that killed millions of people every year until it was eradicated in the 1970s. If scientists can in fact safely experiment with dangerous viruses, there is no need to paralyze the scientific community over bird flu. “You can always assume the worst,” Palese said. “But where do we stop being afraid?”
Osterholm glowered at Palese. “You do not represent the mainstream of influenzologists when it comes to this issue on influenza,” he said. I glanced at some of the other journalist in the audience, wondering if Osterholm could see us scribbling notes.
Osterholm stressed that he was not against research on bird flu in general. He just wanted the scientific community to balance the potential costs and benefits. He didn’t see very much significance in the new bird flu work. It wouldn’t help public health workers monitoring H5N1 viruses for lineages that might be evolving into a human pathogen. Nor did he see any benefit for developing vaccines or antivirals. On the other hand, he saw a risk–a small one, possibly–of tremendous devastation.
But when it comes to viruses can we really calculate such ratios of costs to benefits? Vincent Racaniello, a Columbia University virologist who was also on the panel, doesn’t think so. We’re bad at estimating risks. In 1981, for example, Racaniello and his colleagues pioneered a method for making polio viruses: they stuck the virus’s genes on a ring of DNA called a plasmid, which they then inserted into E. coli bacteria. The engineered E. coli spewed out polio genes, which Racaniello could insert into human culture cells, which then made full-blown polio viruses. People worried that Racaniello’s bacteria would get into people’s guts and start a polio epidemic. (It didn’t.)
We’re also bad at determining the benefits of research. Racaniello recalled how microbiologists in the 1950s discovered that E. coli defend themselves against invading viruses by chopping up their genes. Nobody thought much of that discovery for over a decade. But then in the late 1960s, a few researchers realized that they could use E. coli’s enzymes to cut up DNA and then paste them into new combinations. The entire biotechnology industry was born from that late eureka.
“You could have never predicted that,” said Racaniello. “You never know who will do the right experiment. So that’s why you need to give the information to everyone.”
The way things stand right now, everyone will not be getting that information. I tried to follow the reasoning for holding back key parts of the studies, but, honestly, I can’t recount it in a way that makes sense. As far as I could tell, the thinking was somebody just fooling around out of curiosity would be able to use the full information to create a deadly flu. But the fact is that the scientists who produced the new bird flu used standard methods that have been published many times over. I was also confused by how Nature and Science, the two journals where the redacted papers are to be published, will handle distributing the information to those who need to know about it. An editor from Nature talked about how hard it would be to set up a system. I had been expecting them to have a system to unveil for us.
“None of us ever wants to see a redaction again,” said Casadevall. The most sensible way to avoid that would be to figure out a way to make decisions about risks and benefits much earlier in the life cycle of an experiment. If the mission of an experiment is to create a deadly virus, just to see if it can be done, the panelists agreed that that is probably not a study to run. But what kind of system can stop not just these experiments, but other experiments that might present unexpected dangers? Casadevall worries that every graduate student may have to fill out 100-page forms for even the most harmless of experiments. “You’ll kill science,” he said.
Casadevall was expressing a concern that all the scientists on the panel shared: they worry that this affair will keep them from doing research. For now, they’re trying to work out a fairly self-regulating system to handle this sort of controversial research, perhaps in the hopes that the government won’t come sweeping in. But there was one non-scientist on the panel who did her best to make the scientists aware of the world outside their community.
Laurie Garrett, an award-winning health reporter who now works at the Council on Foreign Relations, pointed out that the flu is not just something that American scientists study in their labs. It’s a global problem. There’s a huge amount of resentment in poor countries where bird flu is the biggest threat, not just to humans, but to the poultry industry. “Poor people are killing their chickens for you,” Garrett said. “They’re going bankrupt.”
Making matters worse, as Garrett has recently written, is the distrust that has developed in the developing world towards Western medical research and the pharmaceutical industry. Indonesia, where many of the H5N1 deaths have occurred, has been reluctant to share bird flu samples with Western scientists, for fear that they would make huge profits from vaccines developed from them. The World Health Organization has set up an international agreement for the exchange of wild bird flu strains between different countries, but it’s in fragile shape.
So for all the sparks that flew in New York Thursday night, the real fireworks over the flu are yet to come.
[Update 2/3 9 am: Corrected description of Racaniello's experiment. Thanks to Matt Frieman. 2:50 pm Fixed Fouchier's institution name and month of his talk. Thanks to Jon Cohen. 8 pm: Expanded Osterholm's "mainstream of influenzologists" quote after seeing his objection to a similarly truncated version in Christine Gorman's story for Scientific American and reviewing my own recording. It's a valid clarification .]