"Without trust, medical research is doomed."
-Catherine D. DeAngelis, MD, MPH, editor, Journal of the American Medical Association
Scientists, like judges, are supposed to be impartial. They are expected to weigh evidence and examine data without emotion or prejudice. If a researcher reports that a certain drug is effective or a specific test is useful, that conclusion is assumed to have been based on an objective evaluation of the data, not on the possibility that the scientist owns stock in the company that makes it.
Over the years, however, medical research and business-particularly the pharmaceutical industry-have become thoroughly and deeply intertwined. As the New England Journal of Medicine noted in a May 18, 2000 editorial:
The ties between clinical researchers and industry include not only grant support, but also a host of other financial arrangements. Researchers serve as consultants to companies whose products they are studying, join advisory boards and speakers bureaus, enter into patent and royalty arrangements, agree to be the listed authors of articles ghostwritten by interested companies, promote drugs and devices at company-sponsored symposiums, and allow themselves to be plied with expensive gifts and trips to luxurious settings. Many also have an equity interest in the companies.
In such an environment, some people question whether commercial interests have an untoward influence on research-both on the conduct and reporting of particular trials and on decisions about what gets studied in the first place. In fact, AIDS activists have been critical and skeptical for years.
Opinion on the gravity of the problem varies. In the summer of 2001, editors of many of the world's leading medical journals announced that they were developing new guidelines to ensure the scientific integrity and independence from commercial control of studies they publish. In response, Bert A. Spilker, Senior Vice President for Scientific and Regulatory Affairs of the Pharmaceutical Research and Manufacturers of America (PhRMA), the leading industry trade group, told the Associated Press that the editors' concerns were "patently absurd." Since then, however, PhRMA has retreated from that stance. PhRMA spokesman Jeff Trewhitt recently told BETA, "The position we've taken is that if [the new set of guidelines] is what makes them feel comfortable, we'll do our best to make the transition to the new policy smooth." He emphasized that "the vast majority of studies have scientific integrity."
Most studies may well have scientific integrity, but enough horror stories have been made public through the media to tarnish the image of both science and industry. These incidents also have alarmed the editors of prestigious journals such as The Lancet and the Journal of the American Medical Association (JAMA).
University of California at San Francisco (UCSF) AIDS researcher James O. Kahn, MD, was at the center of one such episode. Dr. Kahn, who serves as Associate Director of the Positive Health Program at San Francisco General Hospital, never expected to become a poster child for scientific integrity. Yet that is precisely what he became when he headed a 2,500-person trial of Immune Response Corporation's HIV-1 Immunogen (brand name Remune), a therapeutic vaccine that the developers hoped would enhance the body's defenses against HIV, boosting the effectiveness of antiretroviral therapy. To date, Dr. Kahn remains slightly uncomfortable in the role.
In May 1999, based on the fact that Remune was seen to produce no discernible effect on either survival or viral load, a Data and Safety Monitoring Board ordered an early end to the trial (see "News Briefs" on page 6 in the Winter 2002 issue of BETA). Tension quickly arose between the company and the investigators. Immune Response- which sponsored and paid for the study-issued a press release giving the situation a hopeful spin, including what Dr. Kahn called "information that was not supported by the data." The press release notably failed to include commentary from Dr. Kahn and colleagues.
Things went downhill from there as Immune Response battled the researchers over how viral load data should be analyzed. The company wanted to include a viral load analysis of a small subgroup of study subjects at particular time points, which appeared to show improvement. Yet the overall data showed no impact on viral load, and the company's analysis, Dr. Kahn explained, "was not specified in the protocol and was done after the protocol was unblinded, so they knew who got assigned what and how." The scientists felt the company was "data dredging"- selectively picking through the results in an attempt to find something positive.
Immune Response, which had responsibility for processing data from the 77 study sites, refused to turn over the final data to Dr. Kahn and colleagues. The company also refused to provide the addresses of all the study site investigators so that Dr. Kahn and his team could circulate their manuscript to them. Frustrated, the scientists submitted their article to JAMA, which published it despite the missing data on November 1, 2000. "Our decision to publish this study is based on the belief that the integrity of the research process must be protected and preserved," editor Catherine D. DeAngelis, MD, MPH, wrote in an accompanying editorial. The issue included several other articles looking at the influence of commercial ties on research.
Immune Response promptly filed an arbitration claim against Dr. Kahn and UCSF, asking for up to $10 million in damages allegedly caused by publication of the study. The university filed a counterclaim demanding release of the withheld data. The dispute was settled on September 11, 2001, with the company dropping its claim for damages and agreeing to release the data.
Immune Response declined to comment, referring BETA to a publicist who said only, "The most I'm authorized to say is that there has been a mutually agreed-upon settlement." Scientists are not so reluctant to speak about the situation. "I thought what Immune Response did was absolutely disgraceful," commented longtime AIDS researcher John P. Moore, PhD, of Cornell University Medical College, echoing the sentiments of many. "It's one of the worst examples I've ever heard."
The Immune Response case was the second such public battle for UCSF. In 1990 researcher Betty Dong, PharmD, completed a study showing that generic thyroid hormone preparations worked as well as the name-brand version, called Synthroid. The name-brand manufacturer, which had financed the study expecting its product to be shown superior, suppressed the findings for seven years, after requiring Dr. Dong to sign a contract (against university policy) giving the company control over publication of the data. Amid much controversy, the study eventually was published in the April 16, 1997 issue of JAMA, accompanied by a pointed editorial.
Such horror stories represent the worst-case scenario-a company overtly trying to manipulate or suppress research for commercial advantage. Only a few such instances have come to light, but it is unclear if these are aberrations or hints at the tip of a larger iceberg. And what about commercial influences that are less powerful, but perhaps equally problematic?
"I don't know that I can give you a number or a percentage" for how often such conflicts occur, commented JAMA executive deputy editor Phil B. Fontanarosa, MD. "I suspect that these types of issues come up all the time, and that many times they're negotiated and a happy medium reached."
"My guess is it's very common," Dr. Kahn said, adding that industry influence is not necessarily limited to company-funded studies. "Even if you have the umbrella of some group like the National Institutes of Health [NIH], if the NIH is conducting a study of a medication that's owned by a company, that company is part of it. In my experience, everybody has an angle."
A significant amount of industry money goes into biomedical research; by some estimates, the pharmaceutical industry spends nearly double the amount spent by the U.S. government and private foundations combined. And that, Dr. Fontanarosa noted, produces great public benefit: "If not for drug companies pouring large amounts of money into research and development, we wouldn't have good drugs for HIV, cancer, heart disease, and other illnesses. There's a reliance on industry research and development. That's a good thing because a lot of patients benefit. The key is insuring that the money is well spent for the public good."
In theory, companies as well as the public are better off when research is conducted rigorously and reported accurately. "It's really a very stupid idea to play fast and loose with the data," noted PhRMA spokesman Trewhitt. "We think that if you look at the whole record over the years, you see a very large amount of progress. We have more and better medications for a whole host of diseases, and it starts with the clinical testing. That underscores that the vast majority of these studies do have scientific integrity."
However, a large volume of data also documents the downside of industry involvement. Citing an armful of studies, JAMA editor Dr. DeAngelis wrote in her November 1, 2000 editorial, "When an investigator has a financial interest in or funding by a company with activities related to his or her research, the research is lower in quality, more likely to favor the sponsor's product, less likely to be published, and more likely to have a delayed publication."
In 1986 Richard Davidson, MD, MPH, of the University of Florida analyzed 107 drug studies published during 1984 in several major journals; the analysis was published in the May/June 1986 issue of the Journal of General Internal Medicine. Dr. Davidson found a statistically significant relationship between funding source and outcome, with company-funded trials much more likely to favor the new, experimental treatment over traditional therapies. He wrote that "in no case was a therapeutic agent manufactured by the sponsoring company found to be inferior to an alternative product manufactured by another company."
This does not necessarily mean that the studies were biased or that negative results were suppressed, he noted. For example, if a trial is clearly showing no benefit, a company may decide to cut its losses by stopping it early and moving on to more productive efforts. Still, the pattern in this and other research (reflected also in a review of studies analyzing the cost-effectiveness of cancer drugs, which was published in the October 20, 1999 issue of JAMA) is clear: published studies supported by pharmaceutical companies are much more likely to favor those companies' products than studies with independent funding.
Bias can be subtle. A lengthy 1996 article by UCSF researchers Lisa A. Bero, PhD, and Drummond Rennie, MD, published in the International Journal of Technology Assessment in Health Care described many types of bias detected in specific industry-sponsored studies, including use of unrepresentative patient populations, skewed dosing that tilts the study in favor of the new therapy, and misleading data analyses or conclusions not supported by the data. "Our review," Drs. Rennie and Bero wrote, "has indicated that the biases introduced have almost always favored the new drug product over the old."
A major concern is that the net of links between companies and researchers can be extended to affect even research that is not company-sponsored. In an analysis of financial disclosure records for UCSF researchers, who are covered by particularly strict state disclosure requirements, Dr. Bero and colleague Elizabeth A. Boyd, PhD, found a small but steadily growing percentage of researchers who reported a variety of connections with private firms. Connections ranged from small speaking fees, to paid consulting arrangements and membership on scientific advisory boards, to memberships on boards of directors and stock ownership.
Industry's reach goes beyond individual researchers. In 1980 Congress passed the Bayh-Dole Act, aimed at fostering technology transfer-moving research into practical applications. The measure encouraged universities with federal grants to patent discoveries, license them to industry, and share the proceeds with faculty researchers. By most accounts the law has been highly successful, but it has also encouraged development of the web of relationships between companies and academic researchers that now raises questions.
"One thing we found is that not only is the number of investigators with relationships [to industry] increasing, but the complexity of those relationships is increasing," Dr. Bero commented. Gauging the influence of such relationships is "very hard to do," she added. "You see the worst cases, but the ones that aren't so egregious often don't get publicly disclosed. It's very hard to tell."
Industry influence on AIDS research has been an ongoing concern of the activist community since the late 1980s. In the HIV/AIDS arena, it is not unusual at international AIDS conferences to see a researcher present data at a daytime conference session and then discuss the same data-presumably for an honorarium-at a company-sponsored promotional reception the same evening. Other investigators serve as advisors or consultants. "At the top end there are people who probably double their annual income as company consultants," said Dr. Moore of Cornell, politely declining to name names. "Some people are consultants to every company."
Dr. Moore, who himself was a paid consultant to Schering-Plough for a time, is not opposed to all such arrangements. "I think I benefited from it [being a paid consultant], and I think they benefited from me," he said. "I think it's fine in small doses, but you can get buried in it."
Periodically the activist community accuses the industry of wielding undue influence on the NIH's AIDS Clinical Trials Group (ACTG), the largest government-funded HIV/AIDS clinical trials network. Mark Harrington of New York's Treatment Action Group (TAG) has been particularly scathing in describing the ACTG's interactions with Hoffmann-La Roche over ACTG 229, a key trial of the protease inhibitor (PI) saquinavir that was begun in the early 1990s while he was a community representative on what was then called the Primary Infection Committee.
"Roche, which had done the Phase I study, had not tested the drug up to the maximum tolerated dose," Harrington wrote. "Most of the drug did not, it appeared, get into the bloodstream. This was of concern because it might weaken the drug's effect, or even lead to drug resistance.… We were sitting around the table discussing the [trial] proposal, and I kept bringing up uncomfortable questions about the dose and the design." After "an embarrassed silence," the committee chair and others "left the room with the Roche representative and huddled in the hallway.… Then they came back into the room and told us that everything was okay, the dose proposed for ACTG 229 looked just fine."
That dose-600 mg three times daily-became the approved dose of the first saquinavir formulation, Invirase, which proved far weaker than later PIs. Suspicions that the dose was suboptimal were confirmed, and Roche later introduced a new formulation, Fortovase, which provides much higher blood levels of the drug. Harrington's assessment: "Researchers and regulators were complicit in allowing Roche to impose a shoddy dose and a shoddy design on the ACTG."
More recently, the Gay Men's Health Crisis (GMHC) newsletter Treatment Issues suggested that GlaxoSmithKline (formerly Glaxo Wellcome), maker of several antiretroviral drugs, may exert too much influence on the Adult AIDS Clinical Trials Group (AACTG). In a May 2001 article and accompanying editorial, editor Bob Huff noted that of the six open antiretroviral trials in which treatments were specified, five involved one or both of the Glaxo drugs amprenavir (Agenerase) and abacavir (Ziagen). Citing the work of AACTG Executive Committee chair Robert (Chip) Schooley, MD, on earlier Glaxo trials, Huff wrote, "At the very least, AACTG leaders have demonstrated a poor sense of propriety regarding apparent conflicts of interest between government and industry."
One non-AACTG researcher, who asked not to be named, also complained, "The clique that governs the AACTG, a sort of self-perpetuating oligarchy, is up to their eyeballs in corporate contracts and has been for years." Dr. Schooley, the scientist noted, sits on the scientific advisory board of ViroLogic, a company that makes HIV resistance tests, even as the AACTG is doing numerous studies of such tests.
Without naming individuals, Project Inform founder Martin Delaney commented, "When you get a principal investigator of a study who is a major scientific advisor to the company whose studies they are conducting, that's very common in AIDS research. I think that's wrong.… You have a sort of 'group-think' that goes on. [The researchers] are all buddies. They dominate the ACTG. They dominate the scientific advisory boards of the various companies. There's this constant flow of drug company money affecting the principal leaders."
A related concern is that extensive ties with corporate biomedicine subtly shift the research agenda. Pharmaceutical companies exist to develop new chemicals that they can patent and sell exclusively, and inevitably they are much less interested in nonpatentable approaches that, even if useful, may produce little profit. "No one can corner the market," longtime alternative treatment activist George Carter noted. "If one company can make a particular herb, anyone can."
Carter and others do not expect industry to invest heavily in research areas with little profit potential, but claim that the ACTG has not adequately followed up numerous indications from observational studies suggesting that certain nutrients, for example, might slow HIV progression or mitigate drug side effects. "You can just look at what they study. They study almost exclusively pharmaceutical drugs," Carter said. Indeed, a review of the 95 AACTG studies that were open or in development as of September 2001 reveals only three trials looking at nutritional supplements-one of which involved a proprietary, time-release formulation of niacin.
Delaney agrees that commercial ties can influence the direction of even noncommercial research. "It's not evil people plotting behind closed doors, it's more insidious," he said. "Researchers hang out with the company people and it becomes an insular little group…. They start to see things the same way."
The idea of such indirect influences barely seems to register in the minds of some experts. When BETA posed the issue to Dr. Fontanarosa of JAMA, his first response was a long silence. "That's a good question," he finally replied. "I think we can all think of areas where we would hope there would be additional research. Whether the reason we don't have these areas studied extensively is related to the difficulty of the pathophysiology of these disorders or commercial interests is a matter for debate."
Though Dr. Schooley disagrees with some of the criticisms, he said that excessive corporate influence is "something that a lot of us have been concerned with for years," adding, "The complexity of it is sometimes difficult to present to the public."
Those complexities include benefits of scientific-corporate interactions that are sometimes forgotten, he said, including the opportunity to feed useful information to the people charged with developing drugs. "I see patients and have people I take care of who come in all the time. I'm much more in touch with issues I need to think about in relation to drugs than someone sitting up in Research Triangle Park…. The physician/investigator can say to a company, 'This is not a study that makes sense.' Those kinds of discussions are crucial." [Ed. note: Research Triangle Park in North Carolina is the largest research park in the U.S.; several pharmaceutical and biotechnology companies have facilities there.]
Dr. Schooley said he was not involved in the saquinavir decisions that Harrington criticized, but did strongly take issue with Huff's assessment of the AACTG and its interactions with Glaxo. "When you look at the grid put together in [Huff's GMHC article], it didn't look at what other drugs were being studied," he argued, thus producing "an incomplete list of seleced protocols…a distortion and a misrepresentation."
Dr. Schooley also disputed claims that ACTG research decisions are dominated by an industry-connected clique. "Anyone can propose a study," he said, and each proposal receives several levels of review. "The agenda is generated peripherally, not centrally. It's not as if there are a small number of people proposing trials. It's really a very broadly based group-not what was presented [in Huff's article], a very small group."
Sometimes, he added, decisions about what to study are affected by companies' willingness to donate medicines for a trial. While the ACTG can purchase drugs if necessary, "we must balance the couple million dollars it would cost to buy drugs against our limited budget."
Dr. Schooley added that the ACTG has adequate mechanisms for dealing with potential conflicts of interest. Not only is the network governed by NIH rules (discussed below) that mandate disclosure of potential conflicts, but researchers often take themselves out of decisions in which such conflicts might be present. "We had a proposal for a study involving the ViroLogic assay at a committee that I'm on," he recalled. "I disclosed my involvement on their scientific advisory board and I recused myself from the vote."
At least one person who might object to Dr. Schooley's link with ViroLogic gave the ACTG a vote of confidence: Dean Winslow, MD, Vice President of Regulatory Affairs and Clinical Research for Visible Genetics, which makes a competing HIV resistance test. Dr. Winslow said, "I feel we have [been treated fairly]. It's my impression-and I may sound like a Pollyanna-that most of the ACTG investigators really try to be objective. Overall, the ACTG has been right on target with trying to look at a variety of [resistance testing] systems."
The NIH's ethics standards depend on disclosure, requiring researchers to reveal any ownership, royalties, consulting fees, or other interests of $10,000 or more that "would reasonably appear to be affected by the research" being funded. Where conflicts exist, the university or institution at which the scientist works is expected to act to eliminate or "manage" the conflict-with possible actions ranging from simple disclosure to divestiture of financial interests or disqualification of the researcher from participating in affected studies. Decisions about what remedy to apply are left to the institution.
The Synthroid and Remune controversies have led respected medical journals to tighten their requirements regarding research independence. The International Committee of Medical Journal Editors (ICMJE) standards, published in all participating journals in September 2001, state, "We will not review or publish articles based on studies that are conducted under provisions that allow the sponsor to have sole control of the data or to withhold publication." JAMA, like many leading publications, is now asking principal authors to sign a statement certifying that he or she "had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analyses."
Beyond that, journals increasingly are requiring detailed disclosure of any financial relationships that might have a bearing on a particular study-often regardless of how small the amount of money. The ICMJE specifically urges that such conflicts also be disclosed to study volunteers, and asks researchers to state that they have done so.
But is it enough simply to tell readers that researcher X is a paid consultant to the manufacturer of the drug he studied? "I think disclosure is a key step," Dr. Fontanarosa said. "The important part, from the journals' standpoint, is giving that information to the reader so the person reading can interpret what they're reading based on those relationships."
Others disagree. Asked if disclosure of relationships to industry is enough, Dr. Bero of UCSF said flatly, "No, I don't think disclosure is enough. The financial conflict is still there, just more people know about it." She added that there is "no evidence" that disclosure changes the influence potentially exerted by such relationships.
Delaney agreed that disclosure is inadequate. "First, who's going to read it? I think when they say the issue is disclosure, that's a sort of 'keep-it-in-the-family' type of response." Delaney called for new leadership at the ACTG and, at the very least, "much stricter policies on conflict of interest. All these guys should be banned from playing any substantial role in the drug companies."
Bruce Mirken is a freelance writer based in Washington, DC.
Angel, M. Is academic medicine for sale? N Engl J Med 2000 May 18;342(20):1516-8.
Bero, L. and Rennie, D. "Influences on the quality of published drug studies", Int J Technol Assess Health Care 1996 Spring;12(2):209-37.
Davidoff, F. and others. "Sponsorship, authorship, and accountability", JAMA 2001 Sep 12;286(10):1232-4.
Davidson, R. "Source of funding and outcome of clinical trials", J Gen Intern Med 1986 May-Jun;1(3):155-8.
DeAngelis, C.D. "Conflict of interest and the public trust", JAMA 2000 Nov 1;284(17):2237-8.
DeAngelis, C.D. and others. "Reporting financial conflicts of interest and relationships between investigators and research sponsors" JAMA 2001 Jul 4;286(1):89-91.
Gay Men's Health Crisis. "It's time for new leadership at the AACTG (editorial)". Treatment Issues 2001 May;15(5):2.