What Limits Should Be Placed on Biomedical Research in Response to Security Concerns?

John Rennie addressed the following remarks to the audience gathered for the Hans Bethe Award Ceremony.

Thank you. It's a great pleasure to be on this panel today, grappling with the topic of security-based limitations on biomedical research. In the interest of candor, I should begin by pointing out that none of the three of us [John Rennie, Kumar Patel, Victor McElheny] is actually a biomedical researcher or security specialist. So we might be a bit like nuns commenting on the merits and difficulties of sex education-our heart is in the right place, but you may have to make allowances for some inexperience or naivete in our views.

"Naïve"-that's a word that has come up a lot in discussions of this subject in recent years. For example, George Poste, a former head of SmithKline Beecham and the chair of a Dept. of Defense task force on terrorism, told Nature back in November 2001 that biology must "lose its innocence" about how legitimate scientific work could be perverted to malicious ends.

He made an interesting point. Physicists for more than 60 years have been inescapably aware of how their nuclear research could be coopted for evil ends. It's an awareness that security restrictions enforce and drill home at every turn. The physics community has become well-acquainted with advising the government on potential strike and counterstrike development, and with drafting regulations that try to preserve the needed priorities of both research and security. Biologists have certainly always known that their work could be misused in horrible ways-germ warfare goes back deep into history, and worries about the intentional or unintentional creation of novel disease organisms has been the stuff of laws and studies and even pop culture. But biologists have not always had to worry about the government stepping in to interfere with their research for reasons of security.

I think that several very specific events or actions frame our current view of tensions between biomedical research and security. One, of course, was the set of terrorist attacks on 9/11, about which enough has been said. Another was the mailing of weapons-grade anthrax, which took place on the heels of 9/11 and which has never been attributed to any culprit.

Aside from those acts of terrorism, we also have two acts of research. In 2001 Australian researchers were genetically modifying mousepox virus with the intention of developing a contraceptive vaccine to curtail the rodent population explosion. To their amazement and dismay, they found that inserting an interleukin gene inadvertently converted relatively mild mousepox into a mouse superpathogen. The publication of that work prompted worries that some would-be bioterrorist who couldn't obtain a sample of smallpox might create his own disease instead. Then in 2002 Eckard Wimmer of SUNY Stonybrook published a paper showing that it was possible to synthesize poliovirus from scratch, using the publicly available viral genome.

Finally and most recently, we have an act of policy on scientific communication. In mid-February 2003, the editors of more than 30 journals publishing biology research announced that they would adopt a voluntary policy of self-policing papers for information that might be useful to terrorists. Nobody likes to talk about this as self-censorship, but it is in many ways a matter of semantics.

How serious or significant will this self-policing by the journals be? The great worry is that security screens might prevent valuable discoveries from becoming public, or that information essential to understanding exactly how experiments were performed will be excised, making it difficult for other researchers to replicate or scrutinize published reports. Both those results could be disastrous for science.

But in fact, it all depends on how the editors choose to exercise this policy. It is purely voluntary, and each journal will interpret it in its own way. The editors who have commented on this decision publicly say that they expect only one or two papers a year may run afoul of these restrictions. And even in those cases, the editors would probably still publish the papers' results, leaving out just enough details to mitigate the damage they might do. (I believe the editors pointedly remarked that they would still publish both the mousepox and poliovirus synthesis papers I mentioned earlier.) The editors claim a commitment to making sure that the flow of essential scientific information would not be impeded, and I think they deserve the benefit of the doubt.

	“... it’s reasonable to worry about how new biological research could be of service to terrorists, but before society cracks down too hard on open research and discovery, it should make a stronger case that new research really abets terrorism.”

A natural question to ask is, why did the editors choose to introduce this policy at all? Let's not be so cynical that we don't credit their announced desire to prevent terrorists from doing harm. But it's likely too that the editors felt it was essential for them to act as a way of heading off more ham-fisted outside review by the government.

Starting shortly after 9/11 and the anthrax attacks, rumors started to circulate that the Bush Administration was leaning on biomedical journals to censor delicate information. These rumors were denied on both sides for quite some time, but at this point it seems to be acknowledged by Ronald Atlas, the president of the American Society of Microbiology, and others. Moreover, there were reports that early drafts of the Homeland Security bill in 2002 would have established authority for that agency to limit the publication of sensitive papers; that condition was withdrawn after protests. On January 9 of this year, the National Academy of Science and the Center for Strategic and International Studies convened a meeting to discuss conflicts between open scientific publication and terrorism concerns; during those proceedings, it was suggested that Congress might feel an obligation to legislate some kind of biology research oversight for itself. In short, government limits on biology research and publication were in the air.

My own perspective on the new journal policy straddles the fence, I suppose. In practice, this is not going to interfere much with scientific communication, so even though as a journalist I'm predisposed to be a First Amendment absolutist, I'm not going to wring my hands about how this journal policy will wreck or cripple science. On balance, if the editors were really feeling heat from the government, or feared with good reason that it was about to be turned up, they were probably wise to take this on themselves. There's every good chance that if the editors are dedicated to making good research public, this self-policing will become a non-issue over time.

On the other hand, for roughly the same reasons, I think that it's going to do almost nothing to further our national security.

After all, how much do newly published research results help terrorists, anyway? Remember, the most sophisticated approach to bioterror ever taken was the weaponized anthrax attack through the mail a year and a half ago-and my understanding is that nobody has ever published a paper about how to weaponize anthrax.

A lot of the information that budding bioterrorists might want to use is already out there in the public domain. Intelligence experts all fear that large numbers of underemployed researchers who had experience with making secret bioweapons in the former Soviet Union are potentially available to terrorists with cash. And of course, nature itself is full of nasty pathogens already-Ebola virus, anyone?

My feeling is that it's reasonable to worry about how new biological research could be of service to terrorists, but before society cracks down too hard on open research and discovery, it should make a stronger case that new research really abets terrorism. Moreover, before government's protective iron hand closes too tightly on research, it should do more to sensibly govern the commercial availability of actual materials that could be put to malicious ends. Controlling materials may be easier than controlling information, while doing more direct good. And such restrictions shouldn't fall more heavily on biological materials than on chemical or radiological ones, given their tremendous potential for misuse.

Back in November 2001, we at Scientific American arranged a vivid demonstration of how easy it was to obtain dangerous chemicals. At a time when the whole country was on alert against terrorists, we picked up the phone, called a chemical supply company, and conspicuously ordered all the ingredients needed to make the nerve gas sarin. These chemicals were delivered through the mail to an office in midtown Manhattan, no questions asked. Theoretically, I had enough material in my office to kill thousands of people.

It's important to realize that terrorists in the real world aren't usually going to be like the villains in James Bond movies. They aren't Dr. No, with a secret underground lab inside a volcano. They're far more likely to be fanatics working out of a garage or a cave. They aren't that sophisticated, and they don't need to be.

For all the fearful emphasis on bioweapons, chemical weapons and radiological weapons, it's still most likely that terrorist attacks will continue to employ conventional explosives and the like. Ordinary explosives are cheap and easy to make. The Oklahoma City bomber made a truckload out of fertilizer. Of course, explosives are only one of the mundane weapons potentially available. The terrorists of 9/11 killed thousands of people by hijacking and crashing airplanes. Earlier this year, the D.C. snipers spread fear across several states by killing individuals with rifles, over and over again. If I were al-Qaeda and I had secret operatives in the U.S., I would tell them to join the N.R.A., buy rifles and spread out across the country. A loose, decentralized network of snipers like that could paralyze much of the nation and would be almost impossible to stop.

Never forget that the great goal of terrorists is not to kill, but to spread terror. They want to make our society grind to a halt. They can do this by spreading fear directly, or by making us armor ourselves into a cocoon.

A problem with instituting restrictions on biomedical research, or any other kind of research, in response to this "war on terrorism" is that it's not like a conventional war. There isn't a well-defined enemy; there isn't a well-defined set of targets for us to protect or attack; there isn't a defined timeframe outside of which it's possible to think of resuming normal practices. Essentially, we're opening the door to policing research forever.

Some limits on biomedical research are prudent, appropriate, and don't infringe on liberties essential to progress. But before we adopt more severe limits, let's insist that the case be made that they are truly necessary.

John Rennie is Editor-in-Chief of Scientific American.