Sown the seeds: An analysis of published trials in clinical oncology.

Abstract

6068 Background: Competition in the pharmaceutical marketplace has resulted in “seeding trials” which appear to serve little scientific purpose other than to gain a financial hold in the marketplace. Hallmarks of “seeding” clinical trials [Kessler, et al N Eng J Med 1994] include trial design that does not support stated research goals, industry sponsorship without novel findings or poor scientific rationale, and drugs introduced into an already crowded therapeutic class, i.e. “me too” drugs. We evaluated clinical trials published in high impact oncology journals to ascertain the prevalence of trials bearing the characteristics of a seeding trial. Methods: We conducted a PubMed search using “clinical trial”, “oncology”, published 2/2005 – 2/2010, in Ann Oncol, J Clin Oncol, Lancet, Lancet Oncol, or the N Engl J Med. All phase I/II, II and III studies were included. Phase I, hematology, radiation oncology, and pediatric studies were excluded. Data collected included disease site, phase, funding source, journal, and whether the trial fit criteria for “seeding” as previously defined. Results: 1781 articles have been evaluated to date. 528 met criteria for analysis. 130 were considered seed studies. Seed studies per journal were Ann Oncol 26 of 128, J Clin Oncol 88 of 340, Lancet 8 of 22, Lancet Oncol 5 of 16, and N Engl J Med 3 of 22. Studies published per disease site and percent seed studies were melanoma 19 and 31.6%, breast 106 and 30.2%, GI 126 and 20.6%, lung 101 and 29.7%, GU 65 and 24.6%, gynecology 40 and 30%, head and neck 24 and 12.5%, CNS 24 and 12.5%, sarcoma 11 and 18.2%, and advanced cancer 12 with zero seeds. Overall phase totals were phase I/II 16, phase II 272, and phase III 240. Seed studies per phase were phase II 28.7%, phase III 21.7% and zero for phase I/II. 210 studies were solely industry sponsored, and of these, 37.1% were seed studies. Funding was not listed for 73 studies. Conclusions: The integrity of clinical trials is challenged in the race to claim major market share. Our analysis of clinical trials published in high impact oncology journals found seeding trials account for 24.6% of publications. 37.1% of industry sponsored trials are seed studies. Seed studies are more common among the major disease sites or sites for which targeted therapies exist.