U.S. Scientists Faulted on Biotech Consulting

See allHide authors and affiliations

Science  28 May 2004:
Vol. 304, Issue 5675, pp. 1222
DOI: 10.1126/science.304.5675.1222

Until last week, two federal scientists with a hot new idea for using proteins in blood to detect cancer were on a roll. They had teamed up with a bioinformatics company in Bethesda, Maryland, called Correlogic Systems, published a high-profile proof of principle in The Lancet, and launched a clinical research program that attracted hundreds of outsiders. They worked out an unusual collaboration between their agencies, the National Cancer Institute (NCI) and the Food and Drug Administration (FDA). And they had signed up as consultants with a second firm, Biospect Inc. in South San Francisco.

Then the roll came to a halt. The co-inventors—NCI's Lance Liotta and FDA's Emanuel Petricoin—got caught up in an ongoing investigation of federal scientists' interactions with industry. A Capitol Hill committee summoned them on 18 May to respond to allegations of improper conduct. Their supposed misdeed: They had become paid consultants for Biospect (now called Predicant Biosciences) without informing Correlogic, which claims that the deal risked helping a competitor.

House Oversight and Investigations Subcommittee chair James Greenwood (R-PA) and other panel members said this “secret” consulting may have slowed a public-private partnership on a lifesaving cancer test. “This is an outrage,” Greenwood said, calling it an example of poor oversight by the National Institutes of Health (NIH) (Science, 21 May, p. 1091). Liotta and Petricoin respond that they fully complied with their agencies' ethics rules, and their consulting agreements were officially approved.

The issue dates back to April 2002, when NCI and FDA signed a Cooperative Research and Development Agreement (CRADA) with Correlogic. According to a summary, the partners were to “utilize Correlogic's proprietary software technology” to identify protein patterns indicating disease states. Correlogic also signed an exclusive licensing agreement with NCI to commercialize the technology.

A few months later, Liotta and Petricoin were approached by Biospect, a new company with close ties to NCI: Former NCI Director Richard Klausner was a board member, and two former NCI tech-transfer officials were employees. Liotta and Petricoin received approval from their NCI and FDA ethics officers to consult at a rate eventually set at $3250 for 1 day a month. In testimony last week, Petricoin said they were paid to “survey the public domain for applications” for Biospect's technology, which Liotta described as instrumentation for separating and analyzing biological fluids. Explicitly excluded from their agreements was “pattern analysis,” Correlogic's specialty.

Tangled commitments?

Legislators question NCI's Lance Liotta (left) and FDA's Emanuel Petricoin about officially sanctioned consulting deals.


A year ago, Correlogic CEO Peter Levine told Science, he heard from “people in the industry” about Liotta and Petricoin's Biospect activities. He also bumped into them in a Bethesda building where the two companies both rented office space. “I was shocked,” Levine says. “These are the same two gentlemen we were sharing all our thinking and planning with.”

Levine complained in July 2003 to NCI, which rereviewed Liotta's consulting agreement. NCI officials Anna Barker and Carl Barrett testified that they found no conflict with Liotta's official duties. But by that time, Levine says, “we became very guarded in what we would say” to Liotta and Petricoin.

Until a few weeks ago, NIH appears to have continued to support Liotta's Biospect deal, from which he has earned $49,375. A 4 May printout from NCI's ethics database deemed the arrangement “recommended,” even though the companies “do business in the same area,” because Barrett concluded that the work differed from Liotta's official duties on the CRADA with Correlogic. But an NCI ethics official added the note that “this is a very technical distinction” and flagged the case for higher review. The debate soon became moot: Liotta ended the agreement after learning that Biospect had downloaded NCI's publicly available protein- spectrum data. He could not be “absolutely sure” Biospect's research would not “overlap with my government work,” he testified. Petricoin also ended his consultancy after FDA concluded that the agency regulates Biospect's activities. FDA is now launching a review of all staff consulting agreements.

Today, an NIH official says that the agency would not approve Liotta's request because of the “appearance” of a conflict. Levine argues that the scientists could not have kept the two companies' work separate: “If you're being paid $3250 a day by a client, you're going to forget where you learned something.”

What would have been the right thing to do? Greenwood suggested that NIH should have told Correlogic of the proposed Biospect deal and asked for prior approval. Petricoin questioned that, however. Taken to an extreme, “they could claim that the field of technology is their domain” and “there would be only one CRADA: the very first one,” he said. Some tech-transfer officials agree that more than one CRADA for a given technology can be desirable. The Correlogic situation “is a tradeoff,” says John Raubitschek, patent counsel for the Department of Commerce.

The only good news for NIH was that the committee praised NIH Director Elias Zerhouni for taking steps to disclose more information. Last week, for example, he ordered NIH staff to report past consulting agreements. “We are starting to achieve positive changes,” said Commerce Committee chair Joe Barton (R-TX). Greenwood expects to hold at least one more hearing on conflicts at NIH.

Navigate This Article