ADVANTAGE study on Vioxx raises questions about “seeding studies” & scientific value
Posted by Andrew Smith on August 22, 2008
Earlier this week, a paper was published in the Annals of Internal Medicine that reviewed internal paperwork around MSD’s ADVANTAGE study on Vioxx, conducted shortly before the drug’s marketing approval by the FDA in May 1999. The paper is written by four physicians who took part in the litigation against MSD over Vioxx on behalf of patients, and draws on documents that were obtained to inform these legal proceedings. Some of these internal documents use the word “seeding”, and the authors contend that other aspects of the study also correspond with “seeding studies”, arguing that the FDA, ethics reviewers, investigators and patients were misled when they were not informed of market development and raising product awareness with potential prescribers as motivations behind the study.
This conclusion is the story that has been taken up by news agencies around the world, and also widely around the “blogosphere”… particularly by bloggers specialising in injury litigation. I haven’t yet seen much in the way of more rounded discussion of the issues raised by this paper.
Of course, true seeding studies, which investigate questions of spurious scientific value, using unnecessarily large numbers of investigators to give them information, access and experience with a drug at the optimum time to drive future sales, are “a bad thing”. Whether the ADVANTAGE study fits this description is a question on which I won’t offer an opinion: the authors of the original ADVANTAGE papers strongly dispute this, as demonstrated by a response to this week’s paper by Jonathan Edelman, Executive Director of Merck’s Global Center for Scientific Affairs.
The more important issue, though, is how we can stop seeding studies from being conducted. In an excellent commentary to the current paper, Harold Sox (Editor of AIM) and Drummond Rennie (Deputy Editor of JAMA) remind us that there are many levels on which we can “just say no” to them: regulators and ethics reviewers can withold approval, and each individual investigator can simply decide not to take part. This revolves around an assessment of the scientific value of the study, in the context of the current knowledge and understanding of the drug. This can be a fairly subjective judgement, but Sox and Rennie issue a “call to arms” to consider these issues more deeply every time an investigator considers a study.
My own take on this issue is that it’s all about the scientific value. If a proposed study is appropriately designed to answer a valid question, does it really matter if there are additional motivations behind the scenes? These could be the future sales of a newly-approved drug, but could equally be a portfolio of publications in the run-up to a review of academic funding. In the light of this, maybe they should be renamed “vanity studies”…
The current paper offers an understandable view of a specific study, of detailed interest only to one pharma company and the many lawyers and ex-patients involved in seeking recompense for a drug that didn’t work as well as anyone had hoped. However, the broader issues it raises are of great importance to clincial research professionals everywhere, and a strong reminder that we should always consider validity, appropriate design and scientific value with regard to every study we work on.
crmidwest said
The insights from the article in AIM were illuminating to the serious issue of conflict of interest that is afoot in clinical research. Unfortunately, the article stated that one of the limitations of this study was, “…The use of litigation documents from 1 company provides a useful perspective on the practices of the pharmaceutical industry but may not be generalizable to other companies. We do not know how common these practices are.” It begs the questions, “Can you confirm that ’seeding studies’ are common? Does the data on their incidence and prevalence exist, and can it be obtained apart from having class action lawsuits due to adverse reactions?”
Andrew Smith said
A very interesting point. I know of no data on the prevalence of seeding studies, and even the class action route depends on a drug encountering high-profile problems follwing a successful registration and launch. The classification of a study as as seeding study depends on a judgement of the scientific value of the study, which is likely to be time-sensitive and subjective, based on individuals’ differing knowledge of the current state of (published, unpublished and ongoing) literature. I expect virtually all studies would be considered valid and valuable by the professionals conducting them…
majikthyse said
I suppose seeding studies can’t be all bad. To impress the prescriber they should be able to show that the drug works
. But this raises wider issues, relating to the involvement of sales and marketing in R & D. This blew up in the early 80s (yes I do remember!), and the ABPI erected a Chinese wall between the two. This has been progressively eroded and in the current ABPI code it’s impossible to find any clear policy on it. Do you think it’s time this was made clear again? My view is that it’s quite hard enough ridding clinical trials of bias, without allowing commercial pressure to make it even worse.