"Got Toxic Milk?": A Rejoinder

by Milton Leitenberg and George Smith

Recently, the New York Times published an opinion piece that served as a bioterrorism threat assessment. Entitled "Got Toxic Milk" and written by Stanford University professor Lawrence M. Wein, it made the claim that a single terrorist could contaminate the milk supply with a lethal toxin by following the instructions in a jihadi manual available on the Internet.

The opinion described large numbers of casualties, was inflammatory and, we think, flawed in its understanding of terrorist capabilities. It was based upon a paper entitled "Analyzing a Bioterror Attack on the Food Supply: The Case of Botulinum Toxin in Milk" which was accepted for publication in the scientific journal, Proceedings of the National Academy of Sciences (PNAS). The paper was distributed to journalists, including those at CNN, who reported that it had provoked a protest from the US Department of Health and Human Services. HHS claimed that the paper gave terrorists a roadmap for their operations and that it should not be published.

It is not a roadmap, it is a mathematical model built upon a thin supposition regarding what terrorists can do. Wein starts with the assumption that terrorists can make gram quantities of botulinum toxin. For PNAS publication, the reasoning for this was delivered in a little over a paragraph and based on three citations. The first is an article published in the Journal of Bacteriology in 1971. While it deals with the production of Clostridium botulinum toxin, it is more specifically about the preliminary characterization of another protein from the microbe which activates the toxin.

The second citation for terrorist capability was not from another scientific journal at all but, surprisingly, from a short news piece published by the New York Times on April 27, 2003 written by Judith Miller, then embedded with the US Army's Mobile Exploitation Team Alpha in Iraq. It also is not about toxin production capabilities in the hands of terrorists. It is a very brief interview with a scientist who worked in Iraq's bioweapons program and it delivers one brief and contextually meaningless claim about that country's production of botulinum toxin. Neither UNSCOM nor UNMOVIC could verify his statement, nor find verification for it in any internal Iraqi government report.

The third citation is used to bolster argument for possible advanced production methods in the hands of terrorists. It, too, is also not from a scientific publication, but a "paper for discussion" attributed to Richard Danzig, a former US Secretary of the Navy. Of course there are advanced methods for production, the question is whether these are within the capabilities of "terrorists."

Taken together, these three citations are imputed to provide credibility for the capabilities of terrorists to produce grams of botulism toxin. For the New York Times, Wein supposed that a terrorist could also get botulism toxin from a black laboratory located outside the United States. These last claims can be interpreted as the equivalent of a terrorism deus ex machina -- an unconvincing event brought into the plot of a story in order to resolve an involved and complicated situation. In this case, they are furnished to sidestep the obstacles to making large amounts of botulism toxin so that the PNAS paper can proceed with its calculation of casualties.

After reading the original opinion piece and the PNAS paper, we approached the editor of the New York Times with a counter piece and it was turned down. The editor replied that "As a matter of policy, we do not publish rebuttals on the op-ed page."

While it may not be the policy of the New York Times, the paper should be accountable and receptive to a challenging position when it chooses to publish what is effectively "news" or "quasi-news" extracted from a scientific report, particularly one of extraordinarily menacing flavor at the same time as it appears to be grossly inaccurate.

What follows is our critique of "Got Toxic Milk?" that the New York Times wasn't interested in publishing.

"Got Toxic Milk?" Reconsidered

On May 30, the New York Times published a guest editorial that described a potential attack on the nation's milk supply by a terrorist armed with a gallon jug containing a few grams of botulinum toxin. That prognostication and warning, written by Stanford University professor Lawrence M. Wein, was dire. The toll of poisoned was estimated in the hundreds of thousands.

For it to be plausible, one has to accept several assumptions. The first of these is that terrorists can easily brew the amount of toxin cited. But can they?

Lawrence M. Wein's initial claim concerning terrorist knowledge on production of botulinum toxin was an alarming one. He wrote that all that was necessary would be for a single terrorist to have the jihadi manual called "Preparation of Botulism Toxin," secured from the Internet.

We have a copy of the 28-page jihadi manual. It is an oft-stated canard that terrorists, or a single one, can simply download their capabilities for mass death from the world wide web. The assistance that the manual is alleged to confer is greatly exaggerated. While its text certainly appears technical to laymen, its compiler does not explain, except in the most general terms, how to obtain a toxic strain of Clostridium botulinum in the first place. Any strain of the bacterium which produces botulinum toxin won't do, an aspect even noted in the manual. Many strains of Clostridium botulinum in nature produce very little or no toxin. Finding the right one in nature out of literally 600 or 700 strains can take a long time. For example, the task took the pre-1969 US offensive BW program many man-years of work by highly trained and competent professionals.

Wein also posited that botulinum toxin could be bought from an overseas black-market lab. In the real world no "black market" botulinum toxin producer is known to exist.

No real world "terrorist" group - excluding the perpetrator of the US Amerithrax events - has been recognized as having the professional capacity in personnel or equipment to even follow the instructions that the manual does contain. Following the manual's directions requires sophisticated equipment, special reagents, and substantial experience. These do not exist in "jihadi" camps.

In addition, the jihadi manual does not describe processes or professional art that would enable one terrorist to produce gram quantities of botulinum toxin.

Possession of a patched-together electronic manual on botulinum toxin production does not magically make for a shortcut around the experience, education and labor involved in gaining hands-on expertise in the most lethal applications of microbiology. The portrayal of such alleged terrorist documents as true and accurate indications of what terrorists can achieve in this area is not the best assessment.

In the larger work from which Dr. Wein drew his guest editorial, he recognizes that three basic assumptions that lie at the heart of his calculations are open to great uncertainty, as much as three orders of magnitude or more for each of the three. The first is the presumed production capability of the terrorist, only discussed in part above. The second is the degree of lethality of the material presumably produced. The third is the degree of inactivation of the toxin during pasteurization of milk. One of the two main recommendations offered by Dr. Wein was that milk producers should seek to increase the degree of toxin destruction during pasteurization.

The US government and the International Dairy Foods Association have in fact collaborated over the past few years in practical discussions of this problem. An internal document indicates that about 12 months ago, processors of fluid milk quietly raised their pasteurization times and temperature, a step which would significantly reduce by several orders of magnitude the survival of any botulinum toxin that had been added to milk. Dr. Wein's suggestion has by and large already been implemented.

Thinking about the unthinkable has become a way of life in the war on terror. But too often, in our opinion, have the debates on securing the country against the threat of bioterrorism degenerated into worst case scenarios which assume an easy and accomplished technical capability for mass killing already or soon to be in the hands of terrorists. Our assessment is that the possible variability in the three key assumptions means that, taken together, they could result in a difference of nine orders of magnitude from the numbers presented by Dr. Wein, that is produce a result only one-billionth as much.

There is therefore an extraordinary degree of uncertainty associated with Dr. Wein's estimates. The analysis of real and practical intelligence reveals a vastly different, more complicated, and much less frightening picture.