IMAGES.COM/CORBIS
In 1997, Tyrone Hayes agreed to do a study sponsored by Novartis. But he had concerns. Hayes, an amphibian specialist at the University of California, Berkeley, feared that favourable results would make it seem as though he had been "paid off ".
As it turns out, that issue never came up. Via the environmental consulting firm Ecorisk based in Ferndale, Washington, Novartis (now Syngenta) contracted Hayes to study the affects of atrazine, a widely used pesticide, on the hormonal systems of frogs. To Hayes's surprise, he found that at doses as low as one part per billion (one-third of the three p.p.b. level that the Environmental Protection Agency (EPA) allows in US drinking water), the chemical seemed to affect the growth of the larynx in male frogs.
"My study ended up being quite damning for the sponsor," says Hayes. Then he ran into a different problem. Hayes had signed a contract with Ecorisk that gave the company the right to approve publication of his research.
Hayes says that Ecorisk demanded he repeat the experiment but then didn't release funding. Attempts to reach Ecorisk at the company's headquarters were unsuccessful. Ronald Kendall, a toxicologist at Texas Tech University in Lubbock, who headed the Ecorisk panel, was also unreachable, but he has previously denied Hayes's allegations that the company tried to manipulate data. "Ultimately, I think I was quite naive," Hayes says, noting that at the time not many biologists with his background took on industry contracts. The ensuing fight between Hayes, Ecorisk and Novartis resulted in plenty of media attention and controversy.
Key resource
S. BATILORO
Lisa Bero thinks that researchers don't always fully appreciate what rights they might be signing away.
Industry funding can provide valuable research support for academics, but such arrangements must be handled with care. Cases like Hayes's have demonstrated that understanding the caveats and contracts involved in such work could help researchers to avoid awkward or even career-damaging ramifications.
Funding is often sought from industry — especially when government and academic sources are meagre. A 2005 survey of Norwegian researchers, by Magnus Gulbrandsen at the Norwegian Institute for Studies in Research and Education in Oslo and Jens-Christian Smeby of Oslo University College, found that about 25% of university faculty members in natural science and medicine had some industry funding, and two-thirds of technology faculty members did1. And in a not-yet-published survey of more than 1,800 life-science researchers at 125 universities in the United States, Bradford Barham and Jeremy Foltz, professors of agriculture and applied economics at the University of Wisconsin-Madison, found that 20% of respondents reported receiving industry funding in the past three years. This includes 50% of food researchers and 37% of botanists, entomologists and zoologists, but only 4% of geneticists and 6% of ecologists and evolutionary biologists.
Contracts such as the one Hayes signed are not unusual. Often these agreements make it difficult for study authors to publish what they think the results show, according to Peter Gøtzsche, director of the Nordic Cochrane Centre in Copenhagen, Denmark. In a study of 44 industry-financed pharmaceutical trials approved by Denmark's regional scientific ethical committees in 1994 and 1995 that resulted in publication, Gøtzsche found that the industry sponsor either owned the data or was required to approve a manuscript before publication in 50% of the cases, although resulting papers often failed to note those restrictions. In 36% of the trials reviewed, the sponsor reserved the right to terminate the trial at any time. The study also assessed the first 44 industry-initiated trials that were approved in 2004, and in 61% of cases the sponsor either owned the data or had the right to approve the manuscript2.
J. GROW
BACHRACH
David Ludwig (top) and Dennis Ausiello have different takes on the subject of industry funding for academics.
Researchers do not always fully appreciate what rights they are signing away when they agree to these conditions, says Lisa Bero, a professor in the School of Pharmacy at the University of California, San Francisco. "It is so hard to get funding these days that they may not think through the ramifications," she says. Bero has seen many contract clauses mandating the right to review manuscripts, delay publication or terminate a study at any point in her studies of the tobacco and pharmaceutical industries. She cites Betty Dong, a professor of clinical pharmacy at the University of California, San Francisco, who spent seven years fighting to publish results of a study funded by Flint Laboratories comparing Flint's version of a thyroid medication, Synthroid, to three others. When, in 1990, Dong found that all four performed similarly, the company held up publication. Dong eventually prevailed and the work3 was published in 1997.
Cases such as those involving Hayes and Dong came to light because the researchers spoke out, but it is difficult to gauge how often sponsors push scientists to change results or not to publish. In another study, Bero compared industry-sponsored and non-industry-sponsored clinical trials from 1999 to 2005 that compared one statin with another. In industry-sponsored studies, she found that the results were about 20 times more likely to support the statin produced by the funder of the trial than in the other trials4.
David Ludwig, director of the obesity programme at Children's Hospital Boston in Massachusetts, found similar trends in the field of nutrition. Ludwig analysed 206 studies of milk, fruit juice and soft drinks, and found that when a company sponsored studies of its own or a competitor's products, the results were four to eight times more likely to be favourable to the company's financial interests than studies funded independently5. Others have found correlations with industry sponsorship in radio-frequency radiation from cell phones and effects on cognitive function.
But undue industry influence can be minimized. Most top US universities prevent sources of aggravation such as delay of publication, says Dennis Ausiello, chief of medicine at Massachusetts General Hospital in Boston. At this hospital and Harvard, for example, pharmaceutical companies cannot prevent scientists from getting access to and using that data for publication, notes Ausiello, who sits on the scientific advisory board for Pfizer.
In Denmark, all trials must be approved by independent scientific ethical committees. This process has provided a wealth of data for researchers such as Gøtzsche to study possible influence from industry. The US National Institutes of Health has a similar registry, but as it only became mandatory in 2007, conducting analyses on industry influence is difficult.
It is so hard to get funding these days that researchers may not think through the ramifications.
Lisa Bero
But no such measure precludes the need for scientists, especially fledgling scientists, to vigilantly maintain the integrity of their work. Hayes decided he had to find a way to get his results out. In 2000, he quit the Ecorisk panel without publishing his results. He then repeated the studies in his own lab using National Science Foundation funding and grants from foundations such as environmental group the WWF. In 2002, he published those results in the Proceedings of the National Academy of Sciences6 and Nature7. (Members of the Ecorisk panel also published their results, which were more favourable toward atrazine than Hayes's. The EPA reapproved atrazine in 2003; the pesticide is banned in Europe.) Hayes continues to study atrazine as a tenured faculty member at the University of California, Berkeley.
Of course, not all researchers can count on government funding to repeat industry-sponsored studies. "I think the major thing is that you don't want a restrictive contract," says Dong. Hayes advises primary investigators to consider their students' careers and not fund dissertation research with grant money that comes with publication restrictions.
Taking precautions
Even though some institutions have automatic contract-review support, scientists should read everything they sign. "What researchers should worry about are the agreements in the back of the protocol," says Gøtzsche, referring to contract addenda including rules about ownership of data and the company's manuscript review rights. Bero urges young investigators to ask someone at their university's technology-transfer office to review the contract language and ensure it conforms to university policies. She also suggests that they don't rely exclusively on funding from industry. "To advance your career it is good to have a diversified pool of funding," she says. "You don't want to be known as the Merck guy or the Glaxo guy."
Ludwig urges investigators to discuss the prospect of getting industry funding. "You could put together a quick focus group," he advises. "Ask three senior colleagues what they think about taking the funding, and what other funding might exist." He emphasizes that he is not arguing that scientists should decline all industry funding. "We have to balance the opportunity with the cost," Ludwig says. Gøtzsche doesn't shy away from a more idealistic stand. "I think research should be driven by important questions and not by earning money. This is one of the big problems we have today.
When it all works out, industry sponsorship can help to answer these questions. For example, development of the drug Gleevec, a kinase inhibitor that made national headlines in 2001 as an anticancer wonder drug, started with academic discoveries and was developed with industry partnerships. "If what you want to pursue is whether kinase inhibitors of the next generation will affect multiple cancers, and you have an opportunity to pursue that with a pharmaceutical company, I think there is nothing wrong with that," says Ausiello.
Even Hayes still works for industry sometimes. He says he recently took on an environmental study for a water company, looking for contaminants. Hayes believes the company's owner to be a responsible funder. "His view is, 'I don't care what the answer is. I'm responsible, I'm liable, I have to take it out of there'," says Hayes.
And if he could travel back in time to when he was first presented with Novartis-sponsored atrazine study, would he still do it? "Yes," he says. "But I would design the contract in a very different way."