Sarah Houlton

Speaking at the British Association Festival of Science in September, Richard Sullivan, Cancer Research UK's head of clinical programs, urged all biotech and pharma companies to follow the example of GlaxoSmithKline and Eli Lilly, both of which recently began to register and publish every one of their clinical trials.

GSK announced its decision in June, and the first data appeared on its publicly accessible website at the beginning of September. The company has been under particular pressure in recent months following allegations that it hid unfavorable results of clinical trials containing harmful effects of its big-selling antidepressant Paxil/Seroxat (paroxetine).

But the issue is not merely burying bad news. Sullivan says that data from more than a quarter of all cancer trials never see the light of day, making it difficult for doctors to decide on the best possible treatment for their patients. This figure comes from a study of more than 500 Phase III cancer trials presented at ASCO (American Society of Clinical Oncology) meetings over 10 years—26 percent of which were not published within five years of their presentation. Lack of time and money were the most common reasons given for nonpublication. However, the study found that trial results that did not show the investigational drug was better than the standard treatment were much less likely to get published than positive ones.

"If only positive results are published, this can distort medical literature and leave doctors thinking a treatment is more effective than it actually is," Sullivan says. "This in turn can affect the validity and findings of subsequent reviews, treatment decisions, and clinical practice guidelines."

But it's not just the publicly known trials that aren't being published. Sullivan claims that a lack of regulation in the industry means that one-third of all clinical trials are never even registered, much less published. And, suggests Fran Balkwill, Cancer Research UK's head of translational oncology, all scientists have a responsibility to ensure that their data are accessible to the wider community. "If the results from trials never reach the public domain," says Balkwill, "researchers could waste time and unknowingly repeat lines of inquiry that have already been proven unsuccessful."

Cancer Research UK believes that the only answer is a single, comprehensive database that would track the progress of all trials and ensure that results—good or bad—are published once trials are finished. "Registers for cancer trials exist, but they are not compulsory and are far from being complete," says Sullivan. "There needs to be one comprehensive database so trials can be monitored and a wealth of information can be accessed from a central place."

And the editors of 13 of the world's most important medical journals, including Lancet, New England Journal of Medicine, and Journal of the American Medical Association, have called for honest reporting of clinical trials, which would mean revealing the existence and results of all trials, regardless of outcome. Starting next July, their journals will require any trials they publish to have been publicly registered before patient enrollment commenced.

This is certainly a step in the right direction, and it is hoped that the plethora of other medical journals follow suit. If nonregistration precludes publication in the more prestigious journals, it should also encourage other pharma companies to follow the lead of GSK and Lilly. By being open and honest about trial results—good and bad —pharma companies can begin to regain the trust of a public that is increasingly skeptical about their motives.

Sarah Houlton, PhD, is Pharmaceutical Executive's global correspondent. She can be reached at sarah@owlmedia.co.uk