Anthrax expert says NRC report supports FBI

Feb 16, 2011 (CIDRAP News) – The National Research Council's (NRC's) report on the FBI's anthrax investigation amounts to a general endorsement of the agency's scientific approach, even though the NRC found that the purely scientific evidence on the source of the anthrax used in the 2001 attacks was not conclusive, a leading anthrax expert said today.

"I actually have been telling people this is a qualified endorsement of the science in the [FBI] investigation," Paul S. Keim, PhD, a Northern Arizona University microbiologist who helped the FBI investigate the anthrax attacks, told CIDRAP News.

Keim praised the NRC report, saying, "I think the committee did a great job. It's one of the most comprehensive accountings of the investigation anywhere. It makes for a good read. It seems to be factual."

The FBI concluded in 2008 that government scientist Bruce Ivins had sent the anthrax-laced letters to three media offices and two US senators; Ivins committed suicide as the FBI was preparing to charge him. After a number of experts questioned the FBI's scientific methods, the FBI asked the NRC to review its investigation.

Yesterday an NRC committee released a lengthy report saying that the anthrax used in the attacks was genetically similar to anthrax in a flask in Ivins' custody, but that this evidence did not conclusively prove it came from that source. The FBI investigation did not rule out other possible sources for the deadly spores, the committee found.

Keim said the NRC panel is not saying the FBI was wrong, only that the scientific evidence wasn't as strong as the agency suggested.

"The FBI always said the scientific evidence wasn't definitive," he said. He said the scientific investigation yielded information about how and where the anthrax was made and the probable source.

"It was never single source; it was always linked to a set of samples constructed in US military labs," Keim said. "The [NRC] committee came back and said the FBI's conclusion and review of the data was consistent with that. Nothing they found said the FIB was wrong. They said the evidence wasn't as strong as the FBI was saying, and they're probably right with that."

Keim—who described himself as a friend of Ivins' who was surprised when the probe led to him—noted that some media headlines have said the NRC committee doubts the link to Ivins. "The committee isn't saying that. . . . All the major conclusions that the FBI came to, the committee said, 'Yeah, the evidence is consistent with that.'"

The NRC panel reviewed the biological, physical, and chemical methods the FBI used in its investigation, particularly the "microbial forensics," and left aside the traditional forensic methods such as fingerprints and fiber analysis. The committee made no statements about the probability of Ivins' involvement.

Keim said his lab was involved in the investigation of the 2001 anthrax attacks from the beginning, starting with analyzing anthrax DNA from the first victim, Bob Stevens. Because the FBI didn't have a biosafety level 3 lab at the time, Keim's lab became the repository for anthrax samples that the FBI gathered from the around the world for comparison with the letter anthrax.

The FBI analyzed 947 samples and found 8 with the same genetic signature, based on four mutations, as anthrax from flask RMR-1029, which was in Ivins' custody at the US Army Medical Research Institute for Infectious Diseases, the NRC report explains.

However, as many as 100 people, not just Ivins, had access to RMR-1029, Keim and others have observed. "So the science didn't pinpoint the letters to Bruce . . . . It never was what we'd call a definitive identification of Ivins," he said. The FBI used other kinds of evidence in building its case, he added.

One point on which the NRC panel took a slight exception to the FBI had to do with the FBI's suggestion that Ivins tried to hide his role in the attacks by submitting for testing an anthrax sample from a different source than RMR-1029, according to Keim.

Ivins had to submit a second sample after there was a flaw in the way he submitted the first one, Keim explained. When the FBI eventually analyzed that second sample, it was found not to contain the anthrax colony "morphotypes" seen in the letter anthrax. After further testing, the FBI concluded that Ivins had taken the sample from a different source in an effort to hide his role.

The NRC panel, however, concluded that the pattern seen in the sample Ivins submitted could have occurred by chance, Keim said. The panel "went back and said, 'It's not as strong as you said. If Bruce had done this 100 times, this could've happened once. So you can't say it's 100%, you can only say 99%.'"

Keim added, "The difference between 100% and 99% is probably not important in this case. It was only a 1% chance. In the scientific literature we routinely publish things with 99% probability."

On another issue, Keim said he hopes the NRC report will lay to rest the idea that silicon was added to the letter anthrax to facilitate its airborne dispersal, which had implied that the perpetrator had highly specialized skills.

In an early analysis at a military lab, investigators found a higher than expected level of silicon in the anthrax, which made headlines and sparked speculation that someone had added silicon to the preparation, Keim explained. But subsequent analysis under FBI auspices at Sandia National Laboratory showed that silicon wasn't added but was biologically incorporated into the anthrax cells, he said.

"Tests have shown that cells can take silicon out of the glass and store it. So the FBI concluded that the perpetrator didn't do this," he said. The NRC committee agreed with the FBI on this, he added. "So the faulty conclusion about added silicon hopefully has been put to rest by this report."

In other comments, Keim said the NRC report's mention of an FBI investigation of a possible al-Qaida link to the attacks was a complete surprise.

The report says the FBI revealed only last December that it had collected and analyzed samples from an undisclosed foreign site in 2004 on the basis of information that al-Qaida was trying to develop an anthrax program. Some samples from these sites tested positive for Bacillis anthracis by polymerase chain reaction, but cultures were negative. A later and larger set of samples tested entirely negative.

The NRC committee said it was given only fragmentary information about the findings. It concluded that the data were inconclusive as to the presence of anthrax at the foreign site, and added that several scientific and technical issues should be explored in more detail.

Looking to the future, Keim commented that some of the kinds of studies the NRC panel recommended are already under way. He said his lab has a Department of Homeland Security grant to study anthrax culturing patterns as possible strain signatures, and the Pentagon is funding similar research by others.

See also:

Feb 15 CIDRAP News story "NRC: Data insufficient for firm conclusion in anthrax case"

NRC report

This week's top reads