Cancer Researchers with Industry Ties Report “Rosier” Results

By Stephanie Portes-Antoine, BS and Brandel France de Bravo, MPH

July 2009

With all of the cancer studies being performed today, how can consumers be sure of their accuracy?  A study by Dr. Reshma Jagsi at the University of Michigan and her colleagues indicates that cancer studies are more likely to have positive results when the researchers have ties to the company that makes the product being studied.1

This study focused on cancer research articles published in eight medical journals, including Cancer, the Journal of the National Cancer Institute, the New England Journal of Medicine, the Journal of the American Medical Association, and the Lancet. An author was categorized as having a conflict of interest if he or she reported one or was employed, at the time of publication, by the company that makes the medical product being studied, or if a drug or medical device manufacturer was a source of funding for the study. Dr. Jagsi and colleagues relied exclusively on disclosures and information given in the articles themselves. They did no further investigation into authors’ financial ties to industry such as checking board memberships or employee listings. For this reason, the authors believe that their findings may underestimate the extent and effects of conflicts of interest in cancer research.

The authors reviewed 1,534 articles on cancer studies published in 2006. Twelve percent of the articles included at least one author employed by industry-a company that makes medical products-and 17% of articles declared industry funding. The articles most likely to have conflicts of interest included those with authors from medical oncology departments; articles from North America; and articles where the first or last author was a man (these are most likely to be the senior author or principle investigator).  Studies having to do with diagnostic radiology were least likely to have conflicts, whereas those involving prostate, lung, skin, and hematologic (blood-related) cancer were most likely to have conflicts.

Nearly one-quarter of the articles disclosed a conflict of interest. By looking at funding sources and author affiliations, Dr. Jagsi and colleagues concluded that 29% of articles had an apparent conflict of interest, meaning that disclosures do not tell the whole story. This discrepancy can be due to many factors: some journals may have chosen to omit certain information in the disclosures; different journals have different guidelines and policies on disclosure; and cultural norms regarding disclosure vary from one region of the world to another.

Cancer studies with conflict
Industry funding of study 17%
Industry employee 12%
Consulting 12%
Industry funding of other research by author 11%
Honoraria 9%
Stock 8%
Lecture fees 3%
Corporate board 3%
Industry-supplied drugs or technology 2%
Testimony/legal team 1%
Patent 1%

Source: “Frequency, Nature, Effects, and Correlates of Conflicts of Interest in Published Clinical Cancer Research,” Cancer. 2009; 115:2783-2791. Table in this form reproduced from: http://www.ama-assn.org/amednews/2009/05/25/prsa0525.htm

Industry-funded studies were far more likely to focus on treatment than non-industry funded studies (62% vs. 36%) and much less likely to look at epidemiology, risk factors, and effective means of disease prevention and diagnosis (20% vs. 47%).

Randomized clinical trials were more likely to find that a treatment or intervention improved patient survival if a conflict of interest was present. Dr. Jagsi and his co-authors suggest that several factors may account for this. Industry-funded research may tend to design studies that are likely to show their products are effective. The example they give is trials where a drug is tested against a placebo rather than against a drug already in use.  Also, journals may be more interested in publishing positive results, which would inadvertently favor studies whose authors have conflicts of interest.

Since randomized clinical trials are the gold standard for the adoption of new therapies and technologies, these conflicts of interest or “funder effects” have serious implications for cancer treatment and public health. Even if industry funding does not lead researchers to exaggerate a treatment or product’s benefits, studies have indicated that they may tend to not publish negative findings. For example, an analysis of 44 studies on the cost-effectiveness of new oncology drugs found that those sponsored by industry were less likely to conclude that a drug was not cost-effective than studies funded by agencies or institutions without a profit incentive, such as government or university-funded studies.2 In fact, the industry-funded studies were eight times less likely to assess a drug unfavorably than non-industry studies.

Studies dating back as far as 1986 have shown that clinical trials funded by industry are far more likely to favor new therapies over traditional ones.3 New therapies usually cost more, and usually less is known about their potential risks.

Medical researchers often rely on industry funding, and this is true for cancer researchers. Studies indicate that industry funding affects the type of research carried out as well as the reporting of results. This is why it is important that medical research-which is for the benefit of all, not just those who manufacture drugs and other therapies-receive funding from diverse sources, including those without a profit incentive. Since the new research indicates that not all conflicts of interest are acknowledged in medical publications, journals should implement stricter policies regarding accurate disclosure of potential conflicts of interest. In addition, peer reviewers and editors should scrutinize the study design and data analyses carefully and ask tough questions of authors to ensure the accuracy of the findings and conclusions. These efforts are especially crucial when the studies involve treatment for potentially fatal diseases such as cancer.

References

1. Jagsi R, Sheets N, Jankovic A, Motomura AR, Amarnath S, Ubel PA. Frequency, nature, effects, and correlates of conflict of interest in published clinical cancer research.  Cancer. 2009;115: 2783-2791.

2. Friedberg M, Saffran B, Stinson TJ, Nelson W, Bennett CL.. Evaluation of conflict of interest in economic analyses of new drugs used in oncology. Journal of the American Medical Association. 1999; 282: 1453-1457.

3. Davidson RA. Source of funding and outcome of clinical trials. Journal of General Internal Medicine. 1986; 1:155-158.