WASHINGTON — Scouring the anthrax-laced mail that took five lives and terrorized the East Coast in 2001, laboratory scientists discovered a unique contaminant — a tiny scientific fingerprint that they hoped would help unmask the killer.
One senior FBI official wrote in March 2007, in a recently declassified memo, that the potential clue "may be the most resolving signature found in the evidence to date."
Yet once FBI agents concluded that the likely culprit was Bruce Ivins — a mentally troubled, but widely regarded Army microbiologist — they stopped looking for the contaminant, after testing only a few work spaces of the scores of researchers using the anthrax strain found in the letters. They quit searching, despite finding no traces of the substance in hundreds of environmental samples from Ivins' lab, office, car and home.
It's been two and a half years since Ivins committed suicide in the face of prosecutors' threats to charge him with five murders, each carrying a potential death sentence. It's been more than a year since the Justice Department, despite lacking hard proof, formally declared that Ivins "perpetrated the anthrax letter attacks."
But the FBI's decision not to fully test for the distinct bacterial contaminant, pieced together by McClatchy in interviews with scientists, federal law enforcement officials and in a review of recently declassified bureau records, could reignite the debate over whether its agents found the real killer.
The Justice Department closed the eight-year investigation, said to cost as much as $100 million. However, none of the circumstantial evidence it found showed that Ivins prepared the deadly powder, scrawled "Death to America" in a seeming mimic of al Qaida, or twice sneaked away on six and a half hour roundtrip drives to drop them in a Princeton, N.J., mailbox.
If the FBI got the right man, then there is no consequence to its decision to stop hunting for bacillus subtilis, a harmless bacterial contaminant that resembles anthrax. But if Ivins was innocent, then the killer is at large, and the bureau may have missed a big opportunity.
Some scientists and ex-colleagues of Ivins, who spent 27 years studying anthrax at the U.S. Army Medical Research Institute of Infectious Diseases at Fort Detrick, Md., remain convinced of his innocence and believe the FBI erred in limiting the testing.
"This was not an incidental finding," said Martin Hugh-Jones, a retired professor of veterinary medicine at Louisiana State University and one of the world's foremost anthrax experts. "The FBI had what I would call an institutional fingerprint. Whoever had that strain of (bacteria) has to answer to the investigators."
Hugh-Jones, who knew Ivins, believes he lacked the expertise to make the anthrax powder. He contends that the bureau "dismissed" the importance of the contaminant, but concedes that "a bit of housekeeping" could have made it untraceable by the time testing began years later.
One of four federal anthrax investigators, made available to McClatchy on the condition of anonymity, described the contaminant as a clue that "didn't pan out." The official said that the bureau tested "thousands" of samples for the substance, but that included 1,057 anthrax samples submitted by various labs. He wouldn't say how many researchers' work areas were tested.
Some 12,000 pages of bureau records made public to date reflect tests on hundreds of samples gathered in searches surrounding Ivins, but little evidence of tests on other researchers' lab spaces or their stocks of the contaminant.
"They've got thousands of samples?" Hugh-Jones echoed. "But were they thousands of the right samples?"
The mysterious mailing of five anthrax-filled letters to media firms and politicians in New York, Washington and Boca Raton, Fla., sickened 22 people, five fatally, and forced 32,000 others to take antibiotics for weeks. Letters sent to Democratic U.S. Sens. Patrick Leahy of Vermont and Tom Daschle of South Dakota contained a purer, especially deadly anthrax powder, causing lengthy shutdowns of a Senate office building and a major postal facility.
Occurring shortly after al Qaida hijackers seized and crashed four passenger jets on Sept. 11, 2001, the mailings ignited fears that Osama bin Laden or Iraqi President Saddam Hussein had loosed a deadly biological weapon.
Lab tests, however, soon showed that the anthrax in the letters was a strain used solely at 18 U.S., Canadian and European bio-weapons facilities.
Searching for a domestic perpetrator, FBI agents, postal inspectors and lab scientists seemed to spare nothing in their push to narrow a huge suspect list. Initially, they paid up to $1 million for a single genetic lab test, hastening the development of a new field of microbial forensics.
FBI agents locked on Ivins after 2007 tests showed a genetic match between the mailed anthrax and spores in a flask in his lab. He'd shared the contents with others. Testing all samples submitted by labs, the FBI found eight with mutations matching those in Ivins' anthrax, and soon eliminated all suspects but Ivins.
Colleagues and friends knew Ivins as a first-rate scientist who played the organ at church and livened parties with juggling routines, music and limericks. Even after learning of his decades-long obsession with a college sorority and his threat shortly before his suicide to carry out a mass shooting, some of them challenge the FBI's decision, after his death, to elevate him from prime suspect to killer.
"It's irresponsible," said Gerry Andrews, who was Ivins' boss at the time of the mailings.
"I'd rather have a fallible, but more honest FBI, where they say he's our number one suspect, but we really don't know."
Andrews insisted, however, that Ivins and his colleagues "didn't have anything to do with it."
Retired Army Lt. Col. Jeffrey Adamowicz, who supervised Ivins in 2003 and 2004, expressed dismay that the search for the contaminant was cut short.
Adamowicz said that anyone with access to spores from Ivins' flask — or to anthrax he shipped to other labs — needed only "a teeny tiny microscopic drop of that culture to grow their own."
Despite the FBI's cutting-edge work, controversy has followed the "Amerithrax" inquiry.
One person close to the investigation, who requested anonymity to avoid harming relationships, suggested that FBI officials felt "trapped" by Ivins' suicide.
"If they ever had any doubts, once he committed suicide, they had to unite," this person said. "Otherwise, you've driven an innocent man to suicide. And that's a terrible thing."
The law enforcement officials bristled at such assertions, saying that they were seeking Justice Department approval to indict Ivins in the days before he died, but also had risked exposing witnesses to alert his lawyer that he might be a danger to himself or others.
In February, a National Academy of Sciences panel challenged the bureau's finding that a genetic match meant that the wet anthrax in Ivins' flask was the "parent" of the dry powder in the envelopes. The panel said that link wasn't definitive.
Meantime, at the request of skeptics in Congress, the Government Accountability Office recently began an extensive review of the FBI's handling of the inquiry, in which a former Army microbiologist, Steven Hatfill, collected a $5.8 million court settlement after he was mistakenly targeted and publicly identified.
The law enforcement officials stressed that they agreed Ivins was the mailer based on "the totality of the evidence" gathered in gumshoe investigating, not just lab tests.
In a 91-page summary of the inquiry last year, the Justice Department alleged that Ivins: feared that Congress might discontinue an anthrax vaccine program to which he'd devoted his career; misled FBI agents in 2002 by providing anthrax samples that weren't from his flask; had the ability to use the lab's equipment to dry anthrax into fine powder, and was a night owl in his lab in the weeks before the letters were mailed.
The FBI got some corroboration last month when an expert panel concluded that Ivins was the killer, after conducting an unusual, posthumous, court-approved review of his psychiatric records.
Lab scientists didn't identify the genetically unique strain of B. subtilis until December 2005. It was in letters sent to NBC News anchor Tom Brokaw and the New York Post, but wasn't in the Senate letters.
B. subtilis is harmless, but looks and behaves so much like anthrax that researchers have used it to simulate how anthrax spores would act if made into an airborne spray.
Its presence in the letters, LSU's Hugh-Jones said, suggests that somebody grew anthrax using equipment contaminated during earlier B. subtilis experiments.
In March 2007, an FBI advisory panel of six scientists recommended expansive testing for both the mutations in the anthrax and the B. subtilis strain, describing the latter as perhaps the most promising clue to date.
One of the unnamed law enforcement officials said that the FBI arranged for extensive studies of B. subtilis. It also tested for, but didn't find the contaminant in a lab at the Army's Dugway Proving Ground in Utah, which years earlier grew anthrax that went into Ivins' flask and which also received anthrax from him, the official said. Tests also were conducted in work areas of unidentified parties who were under "under investigation," but weren't anthrax researchers, he said.
But once the four mutations in the mailed anthrax were linked to Ivins' flask, there seemed little value to testing the equipment, countertops and B. subtilis stocks in the labs of researchers whose anthrax didn't match Ivins' spores, another of the law enforcement officials said.
Jacques Ravel, a lab scientist who aided the FBI while with the Institute for Genomic Research in Rockville, Md., shrugged off the B. subtilis lead as "a long shot," saying that the contaminant is found "everywhere" in the air and soil and wasn't used much at the time by bio-weapons labs.
However, a 2004 paper in a science journal described a study of B. subtilis by researchers at Dugway, the Battelle Memorial Institute's operations at Dugway, and the Army's Aberdeen Proving Ground in Maryland. Unlike Ivins, researchers at Dugway and Battelle both worked with dry anthrax powder.
Nonetheless, the National Academy of Sciences' panel accepted the FBI's finding that the incomplete testing for B. subtilis lead "did not provide useful forensic information." But, the panel said deep in its report, such clues "should be investigated to their fullest" in the future.
MORE FROM MCCLATCHY