Play Live Radio
Next Up:
Available On Air Stations

Scientific Case Still Open On 2001 Anthrax Attacks


This is SCIENCE FRIDAY. I'm John Dankosky. Ira Flatow will be back next week. Coming up later this hour, a look at yogurt bacteria and whether they help out your gut. But first, 10 years ago, in September and October of 2001, letters containing anthrax spores showed up in news bureaus and Senate mail rooms.

The spores ended up in postal facilities and mailboxes, killing five people and sickening 17. The investigation of those attacks dragged on for years. Then in 2008, the FBI was prepared to file charges against its number one suspect - Army microbiologist Bruce Ivins - when he took his own life, overdosing on Tylenol.

Last year, the FBI and the Department of Justice closed the case, concluding that the late Dr. Ivins acted alone in executing the anthrax attacks. But how conclusive was the scientific evidence against Dr. Ivins? Several recent investigations, led in part by two of our next guests, have asked that question. And that's what we'll be talking about this hour.

Looking ahead, what have we learned since 2001? Are we better prepared to solve biomedical mysteries like the anthrax case today? You can give us a call. Our number is 1-800-989-8255. That's 1-800-989-TALK. If you're on Twitter, you can tweet us your questions by writing the @ sign followed by scifri. If you want more information on what we'll be talking about this hour, you can go to our website at, where you will find links to our topic.

Now let me introduce our guests. Stephen Engelberg is a managing editor at ProPublica here in New York and a reporter on ProPublica's anthrax series with McClatchy and PBS Frontline. He joins us in our New York studios. Welcome to SCIENCE FRIDAY.


DANKOSKY: Paul Keim is a microbiologist and evolutionary biologist at Northern Arizona University and at the Translational Genomics Research Institute in Flagstaff, Arizona. He joins us from KNAU. Welcome to the show, Dr. Keim.

PAUL KEIM: Thanks, John, I'm glad to be here.

DANKOSKY: And David Relman is professor of medicine and microbiology and immunology at Stanford. He's also chief of infectious diseases at the VA Palo Alto Health Care System in Palo Alto. He joins us today from WYPR in Baltimore. Welcome back to SCIENCE FRIDAY, Dr. Relman.

DAVID RELMAN: Thank you very much.

DANKOSKY: I want to start with you, Paul Keim. Maybe you can set the scene for us. Your lab was called upon to do some scientific work for the anthrax investigation. What was your role?

KEIM: So my lab is an evolutionary genetics laboratory, and we'd been working on how bacillus anthraces, you know, or anthrax as it's commonly called, evolved and spread around the world. It's an extremely biological entity. It grows very slowly in some ways, and it changes in evolutionary time very slowly, and so it was a challenge that we wanted to tackle.

And so that's what we did is we went into the genome, tried to identify rapidly evolving regions that we could focus in on. This was in the mid-'90s and well before we could sequence all genomes. And what we were able to do, then, is reconstruct the evolutionary pathway for anthrax on a global scale.

So what happened then was in October of 2001, we received a call from the FBI and said, you know, there's this unusual case, can you help us out. And so we said sure. And so they threw the first culture on an airplane, a jet out of Atlanta and flew it into Flagstaff. And we worked all night and came back and said this looks like the laboratory strain that we call the Ames strain, which to us was really the first piece of evidence that this is a bioterrorism event and not just a natural case of anthrax because that particular strain is very rare in nature.

So that was really our role is in the strain identification and continuing to screen samples from the crime scene, helping to understand what was a part of the crime scene, and wasn't - there were natural cases of anthrax going on at that same time. And it was real quick, using our DNA technologies, to say this is not the same as what was seen in the letters, and so leave it alone, or turn it over to public health officials.

DANKOSKY: Stephen Engelberg, maybe we can pick it up from there. Your investigation at ProPublica takes a look at the way the anthrax case was handled. Maybe you can talk a bit about some of the findings in your stories and also a little bit of what Paul Keim and others were doing.

ENGELEBERG: Well, sure. I'd like to just remind your listeners that the case kind of proceeded in two halves. It began, and for several years focused, on a government scientist by the name of Dr. Stephen Hatfill, who the FBI released significant portions of it, which were persuaded was the perpetrator.

And there were some interesting things in his background that made them suspicious, and eventually they even searched his house and his lab. They emptied an entire pond. They had anthrax-sniffing dogs barking when they walked by him. So there was some evidence there.

But ultimately, the case against Hatfield fizzled out. In fact, he ultimately sued the government and won a $5.8 million settlement. They then focused on Dr. Ivins. At this point, you're looking for a needle in a haystack. You're looking for somebody who sent letters. There are literally hundreds of people who have access to the strain of bacteria that Dr. Keim has mentioned, the Ames strain.

And so the question was how to narrow that, and ultimately the FBI focused on him because of a series of circumstantial things. They were never able to find specific forensic evidence. What our investigation did is looked at some of those things and looked at some of the science, and we found things to question.

DANKOSKY: And just to be clear, some of the circumstantial things had to do with his personality, his behaviors, having nothing to do with science at all.

ENGELEBERG: Yes, and, you know, it would be fair to say that when you get to know the personality of Dr. Ivins, which was largely hidden from his colleagues, that he was not the perfect guy you'd want to have handling deadly germs. He had severe psychological problems. He was obsessed with a sorority, a particularly sorority. He took late-night drives. He did a number of things that would make any reasonable person suspicious.

DANKOSKY: Dr. Relman, you were vice-chair of the National Academy's panel that investigated how scientific evidence was handled in the anthrax case. What was that panel's conclusion? And was the scientific evidence against Ivins as thorough or conclusive as has been described by the FBI?

RELMAN: Well, first of all, we were asked to look at the scientific data and the conclusions that were drawn from it. We were not asked to comment upon or assess the probative value of the scientific data.

Our conclusion, our bottom line, was that the evidence linking the material in the letters to the material in the flask that happened to be found in Bruce Ivins' lab was consistent with an association, a relationship between the two, but was not conclusive or definitive.

DANKOSKY: What was so challenging about this, Paul Keim, this investigation?

KEIM: Well, it was very much a changing landscape from a technology standpoint. You know, during this same period of time, the human genome was completed, and our ability to go in and do comprehensive genome analysis was changing.

And so, you know, the fact that we could do that and that we had pretty much an open checkbook to go out and try new technologies, meant that we did it. And so we were trying to keep up to speed the entire time.

You know, the evidence linking the letter material back to this flask, RMR1029, involved a new level of analysis that none of us had ever done before, and that was to go in and look at individual types within a culture. Previously, everyone had just looked at the culture en mass. And in fact when we did that - and this was done with Claire Fraser and Jacques Ravel at TIGR - when we did that we found that the material in all the cases looked the same. And that's because we were looking at the cells en mass.

And so kind of the breakthrough that led back to that flask involved scientists identifying morphological differences within the culture, subculturing that, and then doing the whole genome analysis on those individual, single-cell-derived colonies, and then taking that back and trying to find a pattern of mutants that were in the letters that was similar to the pattern that was in the flask.

And so the breakthrough there, is, in fact, we're no longer trying to do genetic analysis on a large number of cells, but we're trying to look inside those cells to see what the composition or the heterogeneity within those cultures would be. Really a new approach to this type of science, and again, one that Dr. Relman just referred to that the committee looked at very closely.

DANKOSKY: So we talk about the investigation into the anthrax attacks of 2001. You can call us at 1-800-989-8255 if you have questions, 1-800-989-TALK. Paul Keim, do you think that you'd be called upon again in the event of a bioterrorism attack? I mean, would you get a call like this if this same sort of thing happened today?

KEIM: It would be a little different type of call. I think there was a level of desperation in the FBI's voice when they were calling me back in 2001. At that time, the government didn't have the capability to handle dangerous pathogens like this. They didn't have the linkage between handling dangerous pathogens and genomics.

And my laboratory had that, because we'd been studying the evolutionary patterns of this select agent, bacillus anthraces. The government has spent, I assume, billions of dollars building new facilities up at Fort Dietrich. So they now have the capability to not only handle the material, but also do the genomic analysis in-house.

I think what would happen is they would call me, but it would be more for my scientific expertise and insights rather than actually handling and doing the analysis.

DANKOSKY: Stephen Engelberg, from what you've learned, do you think that things have changed substantially as to how an investigation like this would be undertaken?

ENGELEBERG: I think it has, and I'd like to hear from both of your guests a little bit more about this. But my sense from interviewing them and others is that the techniques available today are far, far, far more acute in terms of detecting the very thing we're trying to detect. So Dr. Keim, for example, today, if you were given a sample of anthrax from a letter, an unknown letter, how would you approach it differently, and how much more sure could you be in the result?

KEIM: Well, we would certainly immediately to whole-genome analysis. Again, this is a technology that's approachable. It's cheap. It's not quite as fast as we'd like, but it's getting there. You know, a good model for this is what happened with the German outbreak of E. coli recently, where they began to analyze with traditional methods, and there was some confusion, in fact, before they got the whole genome out.

They, you know, there was this accusation that the Spanish cucumbers, for example, were the source of the outbreak, when ultimately, it came back to some sprout production. And it was the whole genome and the identity of the particular strain involved that was based upon a whole genome analysis that led to very rapid diagnostic tests and eventually led to the sprouts, where they never did culture a live E. coli from. But the DNA technologies were able to point them to - link the sprouts to the clinical cases.

And this all took place in a timeframe of two to three weeks. You know, it took us, you know, what, eight years to get where we were at in the anthrax investigation. And they did that investigation in three weeks. Amazing.

DANKOSKY: We're talking with Paul Keim, a microbiologist and evolutionary biologist at Northern Arizona University. Stephen Engelberg is also here, managing editor at ProPublica. They've been doing a series on the anthrax investigations with McClatchy and PBS Frontline. David Relman joins us, as well. He's a professor of medicine, microbiology and immunology at Stanford, and he joins us from Baltimore today.

If you want to join our conversation: 1-800-989-8255, 989-TALK. We'll be right back after this short break.




DANKOSKY: I'm John Dankosky, and this is SCIENCE FRIDAY, from NPR. Today, we're talking about the 2001 anthrax attacks and the scientific evidence in that case. My guests today are Stephen Engelberg, managing editor at ProPublica. Paul Keim is here from Northern Arizona University. David Relman is here from Stanford, as well.

We're taking your calls at 1-800-989-8255. And we'll get to those in just a moment. I'm wondering, Dr. Relman: How much of the investigation of Bruce Ivins, the main target of this investigation, how much of this was really based on science, and how much was based on other things?

RELMAN: It's an interesting question. It's one that we really weren't in a position to be able to answer. We were simply asked to look at the science. But, of course, it's a blurry line between what constitutes science done for science's sake in the setting of an investigation, and what is other kinds of investigatory lead and data that bear on the case, as well.

DANKOSKY: Well, how is scientific investigation different than a criminal investigation, for instance?

RELMAN: So there are, of course, similarities, and yet I think one of the differences is that when one goes about science for science's sake, there are certain sort of basic questions that are being asked, hypotheses posed and a plan set out. And as the plan is pursued, new data arise, and new leads become available.

Now, this is similar to science in a criminal case. The difference, I think, is that in the setting of science for science's sake, one can pursue multiple leads at the same time, turn back in time and go off in different directions, whereas in a case there is a certain pressure to get to an answer and bring a case to trial.

So it's a more open-ended procedure in the scientific world, and much more of a directed, deliberate and sort of linear path in the criminal investigation world.

KEIM: And what our investigation found was that as the case was built, they became less and less interested in doing things that might undermine the case. One example, which Dr. Relman's panel pointed to, was at one point, there were environmental samples of anthrax taken from one of the crime scenes, and the prosecutor declined to sequence them.

And, you know, they said, well, it was resources. It was time. But the fact is, there were unlimited resources for this case. My guess - and here it's just a guess - is that they didn't want to ask a question they didn't have the answer to. They were afraid it would show something unexpected. And I think for scientists, that's exactly what you'd want to do, is find something unexpected.

DANKOSKY: Dr. Relman, is it possible that there were key anthrax samples missing from the FBI repository from abroad, maybe from U.S. government labs?

RELMAN: It's possible. I think in any case like this, you're always going to be confronted with the problem of knowing what is the true universe within which one needs to search. We were told of a certain set of samples, and there were a large number that were collected from all over the world. And yet we did become aware, for example, of some environmental samples that were collected overseas which gave inconsistent results in the presence of Ames anthrax.

And these samples, because they didn't yield a cultivated organism, never made their way into the repository, where the more formal, deliberate testing was undertaken. So that's one bit of evidence or one indication that there may have been certain kinds of samples that never really made their way into the final repository.

DANKOSKY: Martin is in Los Angeles. Hi there, Martin. Go ahead. You're on SCIENCE FRIDAY.

MARTIN: Good afternoon. I'm the commander of the homicide bureau of our department, which also encompasses our scientific investigation division. And I can tell you from a law enforcement perspective - I sit on a joint antiterrorist task force here in California, and we are much more concerned with a possible bioterrorism than, let's say, a dirty bomb or something similar to that.

When we started looking into this a little more closely, we were - found that it was pitifully inadequate how these pathogens and agents are controlled just on campuses in our state and people that have access to them on a daily basis. So the probability - simply because of the fact that it's a whole lot easier to get some of these things than it is nuclear components, it put a good chill down our spines.

DANKOSKY: And the security, Stephen Engelberg, you reported on Dr. Ivins' lab - very, very lax.

ENGELEBERG: Yes, now in fairness, it has been significantly tightened since the case of the anthrax letters. But I think your caller's point is well-taken. I did a book 10 years ago with Judith Miller and Bill Broad of the New York Times on this subject, and I was startled, as well, as to how easy it would be to do this.

Now, you know, again, in this case what we're talking about is supposedly is an insider who did it. But you can't rule out other possibilities, because it is that easy.

DANKOSKY: I want to go Mark, who's calling from Oklahoma City. Hi, Mark.

MARK: Yeah, I was kind of following up on that. I saw the "Frontline" episode, and it raised some questions with me, the first one being is my understanding is is there's - when it comes to the highest level of biohazard containment, there's Fort Dietrich and maybe a handful of other places in the U.S., like the CDC. And I was always assuming that there was a high degree of psychological screening required of people allowed to deal with those kinds of pathogens. In which case, how did a guy like Dr. Ivins - who they say is mentally unstable - get access to that kind of facility?

DANKOSKY: Yeah, what can we say about that? What do we know about how you get a security clearance? Dr. Relman?

RELMAN: Well, there is - there has been a fairly extensive procedure in place, and there was one with the U.S. Army and the Department of Defense. It has been reassessed and reevaluated. But I think the bottom line is we can't be absolutely positive that there won't be individuals that harbor instability or other motives that we simply can't detect.

And I think we're going to have to rely upon a sensitized and sort of well-tuned-in, motivated scientific community to speak up when they see things that don't seem right. This is where we need to sort of buy the goodwill of the scientific community and all those who work in this area of science because there is no foolproof method.

DANKOSKY: So we've talked a bit about the circumstantial evidence and some of the science. Dr. Keim, I'll ask you first: Do you think, in the end, that the scientific evidence here was solid enough to go to court to charge Dr. Ivins?

KEIM: Well, I can't tell you whether it was enough to charge Dr. Ivins. I can only speak to the part of the investigation that I worked on, which was, really, the Ames strain. And, you know, for seven years that's all I thought about, was going to court. I have testified in murder trials and other DNA trials before, and so I was aware of the standards that were necessary when you are on the stand.

And for the part of the investigation we were tasked with, we were definitely ready to go to court. We had to change from being an academic laboratory to putting, you know, things in place like evidence tracking, chain of custody. We had highly validated DNA assays. We had - you know, things that you'd do in an academic lab to speed things up just went by the wayside as we started to have double and triple-witnessing of every single procedure.

So I would have gone to court and I would have been able to say quite precisely that this was a strain that came from a laboratory. It was very unlikely that it would have come from nature. You know, and when you move into other types of evidence, the nuance becomes the language that the expert witness uses.

Is this definitive? Is it consistent? I'm highly confident, or I suspect that it's true. In science, we like to deal in probabilities, but frequently, when you get into court, those probabilities have to be translated into a language when you're addressing a jury. And that language doesn't always equate perfectly with a P value, or a probability value, that we would prefer to be using in science.

And doubtlessly, part of the evidence that was being developed was going to be less strong than the evidence that we had. But again, what we were working on was ready for court.

DANKOSKY: Dr. Relman, what do you think? Do you think that there was enough evidence to go to court?

RELMAN: Again, it would be hard for me to say. I think there certainly were data that were suggestive and consistent with a link between the letters and the flask. But, you know, I think we have to step back and really ask: What is it that we expect science to be able to do for us?

I don't think we should be expecting that a scientific experiment is going to reveal a result that points to a person. It may point to a possible source or an evolutionary history or a set of relationships. But, you know, we have a tendency to put science on a pedestal sometimes, and in this case, like many others, we would have to hope that there are other kinds of evidence to bring this case, you know, fully around and point it in the right direction.

KEIM: Just to be clear on that, I mean, what we're talking about is the question of whether or not a particular anthrax culture, the person who made the letters actually took this and grew it somewhere else. It's not taken out of the flask and then dried. So they grew it someplace separately.

Did that person take it from this particular flask? If so, there are well over 200 people who had access to this particular flask. But that's not the end of the story, because this material, this very same material, was in other places. So, for example, to say that one of the research laboratories in Ohio did not contain the perpetrator, the FBI needed to look at those people and ask questions like: Did they have enough time to drive to the Princeton, New Jersey mailbox? And science can't answer that question because it's the same stuff. And so there's no scientific test that can prove to you that it came - that one genetically identical thing came from Ohio or from Fort Detrick, Maryland.

DANKOSKY: Paul Keim, knowing what you know about growing anthrax, do you think that Dr. Ivins had the capability to produce these dry spores, the ones that ended up in the letters in such a high quantity?

KEIM: I'm going to have to decline to answer that. I'm not a spore-production expert, and I, you know, really don't know what Bruce had in his laboratory, and whether it would have been able to do that. That's just beyond my area of expertise. I'm sorry, John.

DANKOSKY: No. I apologize. David Relman, you want to weigh in? I mean, do you think that this is even possible to do what Dr. Ivins had at his disposal?

RELMAN: Well, you know, we actually were never presented with what we had at his disposal. And what there might have been at someone's disposal in 2009 when we began our work could certainly have been something very different than what was at someone's disposal in 2001. So there are a lot of unknowns. And we - and to be honest, we were not presented with a specific scenario, a set of resources and equipment and other reagents, you know, that might have been the, you know, the scenario through which these things were made.

ENGELEBERG: Now, that said, if you interviewed Dr. Ivins' colleagues at Fort Detrick, there are one or two who say he could have done it and a larger number who say he couldn't have done it. And frankly, none of the people speaking are really truly expert in the area of spore cultivation. So having spent some time interviewing people on this, I would say that it's still a difficult question to answer.

DANKOSKY: Paul Keim, one recent anthrax investigation you worked on involved contaminated heroin in Scotland. I'm wondering if you can quickly tell us about that investigation.

KEIM: Yeah. There were a lot of parallels between this recent outbreak of anthrax in Scotland and the anthrax letters. And, of course, we're able to take advantage of our ability to do DNA analysis. But in late 2009 into 2010, there were a series of anthrax cases among drug users in Scotland, and what the investigators quickly tied together was that they were all injecting or smoking heroin. And so this epidemiological leak is very suspicious. And they were able to isolate live Bacillus anthracis, the bacterium, out of the victims, although they never actually got it out of the heroin itself.

So they brought us in, and we worked with the British government and the health protection agencies up in Scotland to try to figure out exactly what type of anthrax was being involved. And like much of science, we're able to disprove things more easily than we're able to prove them, but the things we disproved were pretty important. For example, we very quickly were able to say that this outbreak among the drug users was not associated with the anthrax letters. It was not the aim strain.

Importantly, it was not the British biological weapon strain. The British had actually tested anthrax as a biological weapon in the 1940s and '50s in Scotland, and so the locals were, of course, already pointing fingers at the government and saying this is residual from that biological weapons program. But it wasn't true. It was a very different strain than that. We had a lot of information about anthrax in Afghanistan, as you can imagine, with the terrorist threat coming out of that country.

And that's where heroin in Europe is produced. Ninety percent of all heroin that ends up in the U.K. comes from Afghanistan. And again, we're able to quickly rule that out. Now, we have these really large genetic databases of anthrax. It's kind of like the law enforcement equivalent of Dakota's database that they use for human forensics. So we've developed this over the last 10 years. And when we queried that database with the genetic type that we found in the drug users, it matched up with a smuggling route that snakes its way through the Middle East.

And so we're able to actually tie it to several independent isolates from the Middle East. So the law enforcement officials now believe what happened was that that the heroin was contaminated with anthrax spores as it was being smuggled, possibly, you know, in animal hides or maybe adulterated with some type of animal product.

DANKOSKY: Hang on for one second, doctor. I'm John Dankosky. This is SCIENCE FRIDAY from NPR.

KEIM: So, you know, from a public relations, political and even public health standpoint, this is very important to try to identify a source or, even better, rule out alternate sources. And so, again, this is all based upon research that had been developed during the anthrax letters and methodologies.

DANKOSKY: I want to get to...

RELMAN: You know, it's...

DANKOSKY: Go ahead, Dr. Relman.

RELMAN: I was going to say - this is Dave Relman. Paul Keim has done some incredible work, and this just, again, points to the potential value of having a lot of expertise and understanding of certain organisms. But as we look forward - I think that's an important thing at this point - the next event is not going to look like the last event, and the next event may not involve anthrax. And it may involve an organism that we know less about or that can be cultured less easily or has been engineered by somebody, by the perpetrator.

And we may not have attack material the way we did this past time. And for all of those reasons, I think we have to now sit down and really think about what are the tools and kinds of understandings that we need to develop to be able to address the next time.

DANKOSKY: Well, what do you think we do need, Dr. Relman?

RELMAN: Well, I think the kind of thing that Paul has developed for many other kinds of organism, both the specifics of the population structures of other kinds of threat agents, but also just the tools and the conceptual understanding of how to understand populations and evolution, the kinds of experiments that can teach us more, but also just the technical tools of how one now goes about sampling the environment, making use of very small amounts of material, which is likely what we're going to have in front of us the next time.

DANKOSKY: Stephen Engelberg, quickly, what have you learned about our ability to protect against the next bioterror attack?

ENGELEBERG: Well, I've always felt that there's an overlap between public health and biodefense. And, you know, in this case, I think we learn a lot of things about how to deal quickly with the public health threat that we saw. You know, the movie "Contagion" is actually pretty accurate. One of my favorite lines in it is the question about is it weaponized, some government official asked. And the CDC guy says, no, the birds have weaponized it, and they're doing a heck of a job. And I think that's true in these biological events that you really got to look at your - the fundamentals of your public health system first. And I think it's better today than it was 10 years ago.

DANKOSKY: Stephen Engelberg is the managing editor at ProPublica here in New York, a reporter on ProPublica's anthrax series, done in conjunction with McClatchy and PBS "Frontline." Thank you so much for being here.

ENGELEBERG: Thank you.

DANKOSKY: Thanks also to Paul Keim, a microbiologist and evolutionary biologist at Northern Arizona University and at the Translational Genomics Research Institute in Flagstaff. Paul, thank you for being here.

KEIM: Thanks a lot, John.

DANKOSKY: And Dr. David Relman is a professor of medicine, microbiology and immunology at Stanford. He is also chief of infectious diseases at the VA Palo Alto Health Care System in Palo Alto. Thank you all of you for joining us today. Now, after the break, we turn to some friendly microbes, the bugs that live in yogurt. Do they help out your gut? Stay with us. Transcript provided by NPR, Copyright NPR.