Review finds conflicts of interest in many cancer studies

12 may 2009--A new analysis finds that a considerable number of clinical cancer studies published in respected medical journals have financial connections to pharmaceutical companies. Published in the June 15, 2009 issue of CANCER, a peer-reviewed journal of the American Cancer Society, the study indicates that conflicts of interest may cause some researchers to report biased results that are favorable to companies.

Ties between clinical researchers and companies that make medical devices and drugs have become increasingly complex and controversial, particularly as more researchers compete for scarce federal research funds. In addition to using industry money to support their research, some investigators receive consulting fees, own stock and hold positions within companies that profit from selling the very products they are investigating. These conflicts of interest have raised concerns that studies with ties to industry are biased and are not designed to provide a true test of medical therapies. Many medical journals now require researchers to disclose potential conflicts of interest in the articles they submit for publication.

To get sense of the frequency and impact of conflicts of interest in clinical cancer research, Dr. Reshma Jagsi of the University of Michigan and colleagues reviewed cancer studies appearing in eight highly regarded journals in 2006. These journals included the New England Journal of Medicine; JAMA; the Lancet; the Journal of Clinical Oncology; the Journal of the National Cancer Institute; Lancet Oncology; Clinical Cancer Research; and CANCER.

Of the 1,534 cancer studies identified in these journals, 29 percent had conflicts of interest that were apparent from review of published author declarations and authorship lists (including industry funding, consulting fees to authors, co-authorship by industry employees, etc.), and 17 percent declared industry funding. Conflicts of interest were most often found in articles with primary authors from departments in medical oncology (45 percent), those from North America (33 percent), and those with male first and senior authors (37 percent).

According to the authors, randomized clinical trials that assessed patient survival were more likely to report a survival advantage associated with the intervention when a conflict of interest was present. These trials are the foundation by which drugs, technologies, diagnostic tests, etc. get approved for use in the clinic and therefore shape the way oncologists practice medicine.

The findings also show that studies with industry funding were more likely to focus on treatment than studies without industry funding (62 percent vs. 36 percent). They were less likely than studies not declaring industry funding to focus on epidemiology, prevention, risk factors, screening or diagnostic methods (20 percent vs. 47 percent).

This analysis revealed that conflicts of interest exist in a considerable number of clinical cancer research articles published in important journals. The authors noted that "attempts to disentangle the cancer research effort from industry merit further attention, and journals should embrace both rigorous standards of disclosure and heightened scrutiny when conflicts exist."

###

Article: "Frequency, nature, effects, and correlates of conflicts of interest in published clinical cancer research." Reshma Jagsi, Nathan Sheets, Aleksandra Jankovic, Amy R. Motomura, Sudha Amarnath, and Peter A. Ubel. CANCER; Published Online: May 11, 2009 (DOI 10.1002/cncr.24306) Print Issue Date: June 15, 2009.