Kill it before it grows: IRBs and seeding trials

Hard to detect, but watch for red flags

IRBs striving to ensure they're not inadvertently approving a seeding trial have a tough job ahead of them.

Because of the nature of the trials, and the fact that they are likely to have plausible-sounding scientific rationales, they may be hard to ferret out, says Kevin Hill, MD, MHS, a clinical fellow in psychiatry at McLean Hospital in Belmont, MA. Hill and his colleagues published an article in the August issue of the Annals of Internal Medicine about a Vioxx study they identified as a seeding trial.1 Merck & Co. Inc., the maker of Vioxx, has denied that the study was a seeding trial.

"Seeding trials have really evolved over time," Hill says. "Earlier papers about seeding trials talked about open-label trials with lots of sites. If that were the way seeding trials currently were configured, it would be a bit easier to identify them. But over time, the trials have evolved into fairly scientifically sound trials that you wouldn't be able to otherwise identify."

But there are signs that can make an IRB suspicious, if members are willing to ask questions and exercise some extra diligence.

Question the science

Hill says that simply asking the investigator, "Is this a seeding study?" is a good start, since it puts the burden on the investigator to be truthful. But it's possible that the investigator doesn't know the true origin of the trial.

One way IRBs can attack the problem is to concentrate on the scientific merit of the study under consideration, say Hill and Mark Schreiner, MD, chairman of the Committee for the Protection of Human Subjects at The Children's Hospital of Philadelphia, PA.

"The seeding trial often does not have a real scientific hypothesis," Schreiner says. "So the question is, does this study have a true scientific objective that can only be answered by a clinical trial?

"And then is the study adequately powered to address that objective? Are the comparators the right comparators? Is it necessary to use a placebo or should it be an active comparator? Is the dose that's being selected the correct dose? Is there an adequate search of the literature?"

Schreiner notes that most of the studies that might fall into the category of a seeding trial are Phase IV studies conducted after a drug has been approved. In those instances, IRBs can ask whether the trial is initiated by an investigator, or by the drug's sponsor.

An editorial in the Annals that accompanied the Hill article suggested other clues an IRB could consider suspicious: An open-label design, no control group or a short-term study of a chronic disease. "None of these clues is highly specific, but institutional review boards should start asking questions when a study has several of them," the editors wrote.2

Patient/physician ratios

A red flag can go up, Hill says, if the study has an unusually large number of individual physicians participating, or if the number of patients per/physician is unusually small. That might indicate that the purpose of the trial isn't to answer a question, but to put the drug in the hands of lots of doctors.

For that reason, Schreiner says, central IRBs may have more likelihood of being presented with a potential seeding trial.

David Forster, JD, MA, CIP, vice president in the Office of Compliance for Western IRB in Olympia, WA, says the question of whether a proposed study was a seeding trial has been asked occasionally over the years.

"You never had any true knowledge of whether it was or not," Forster says. "We often disapproved studies for lack of scientific validity or an inappropriate risk-benefit ratio, but I do not remember when we've ever said we're disapproving this study because it's a seeding trial."

The criteria that IRBs follow in reviewing studies don't directly address this question, he says. "If you go through it and the risks are minimized, the risk-to-benefit ratio is acceptable, the knowledge is supposed to be scientific knowledge, not marketing knowledge. We have these checklists we use and there's really nothing in the regulations that addresses this issue."


  1. Hill KP, Ross JS, Egilman DS et al. The ADVANTAGE seeding trial: a review of internal documents. Ann Intern Med. 2008 Aug 19;149(4):251-8.
  2. Sox HC, Rennie D. Seeding trials: Just say "no." Ann Intern Med. 2008 Aug 19;149(4):279-80.