Evaluation of Conflicts of Interest in Basic Science Studies Evaluating Erythropoietin and Erythropoietin Effects of Solid Cancer Cell Lines

Studies have found that when investigators have financial relationships with pharmaceutical manufacturers, they are less likely to criticize the safety, efficacy, or cost-effectiveness of agents supplied by the manufacturers. No study has evaluated the possibility of an association between basic science findings and pharmaceutical funding. This is important for erythropoietic stimulating agents (ESAs), where statements from the Centers for Medicaid and Medicaid Services, FDA, the National Comprehensive Cancer Network, the manufacturers, and the European Medicines Agency highlight tumor progression risks when ESAs are administered to anemic cancer patients. Safety signals originated in 2003 when two trials (BEST (breast cancer) and ENHANCE (head and neck cancer)) investigating the off-label use of ESAs in oncology identified increased mortality and tumor progression concerns. We sought to determine the presence or absence of whether there is an association between pharmaceutical industry support and findings related to erythropoietin receptors (EpoR) in solid cancer cell lines.

A MEDLINE database search was conducted (Keywords: erythropoietin receptor, EpoR, cancer, tumor). Studies investigating EpoR presence, downstream effects of ESA administration (JAK2, STAT5/bcl, MAPK/ERK, PI3K, AKT, and/or NFkB), and/or stimulatory effects of ESA administration (cytoprotective, mitogenic, and/or invasive) in solid tumors were evaluated. Conflicts of interest, affiliations, and funding sources were abstracted. Of the 73 studies included, 5 studies were conducted by investigators who were employed by ESA manufacturers and 4 studies were conducted with research grants from ESA manufacturers. We defined these as “manufacturer-affiliated” studies. Sixty-four studies were conducted with no disclosures of independent of manufacturer affiliations or funding.

Manufacturer-affiliated and non manufacturer-affiliated studies both identified EpoRs in solid cancer cell lines (67% vs 94%), but conflicted studies were less likely to identify downstream effects (JAK2, STAT5/Bclx, PI3, NFkB, AKT, MAPK, ERK) (0% vs 94%), or tumor stimulatory effects (cytoprotection, mitogenic, invasive) (0% vs 73%).

Pharmaceutical company sponsorship of EpoR studies is associated with reduced likelihood of identifying EpoRs on tumor cells, downstream effects of Epo on solid cancer cell lines, or Epo stimulatory effects on tumor cell lines. These findings may have implications for other basic science research studies. Only further analyses can confirm or deny these preliminary findings, but the implications are significant and additional studies in this area are warranted.