Healthy Skepticism Library item: 12526
Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.
Publication type: news
Mansell P.
Lilly hits back on antidepressant trial publication
PharmaTimes 2008 Jan 21
http://www.pharmatimes.com/WorldNews/article.aspx?id=12678&src=EWorldNews
Full text:
Eli Lilly has gone on the offensive over suggestions that it has suppressed negative clinical trial data on its antidepressant Prozac (fluoxetine hydrochloride).
The US company singled out a news report in The New York Times on a review of antidepressant studies registered with the US Food and Drug Administration as “inaccurate” and likely to have “created a strong false impression with readers that Lilly suppresses data”.
The Times report did not mention any specific data on Prozac and the ratio of positive or negative trial results that were either published or unpublished. However, it did twice name Prozac as part of an overall trend identified in The New England Journal of Medicine, whereby 31% of FDA-registered studies on antidepressants were not published in medical journals and, out of 36 trials regarded by the agency as having “negative or questionable” outcomes, 22 were not published and 11 were tailored in publication to suggest a positive outcome.
While the published literature indicated that 94% of the clinical trials conducted with 12 antidepressants were positive, FDA analysis revealed that just 51% were positive, the NEJM article said. It estimated the “increase in effect size” from these disparities – i.e., the amount by which the antidepressants’ perceived effectiveness was inflated by failure to publish negative studies – at 32% overall, with a range from 69% (Serzone, Bristol-Myers Squibb) to 11% (Paxil CR, GlaxoSmithKline). Prozac was close to the bottom of this scale, with an estimated effect size of 14%.
According to Lily, The New York Times was inaccurate in citing Prozac, as “the NEJM article didn’t identify a single Prozac study as unpublished”. Moreover, two trials with another Lily antidepressant, Cymbalta (duloxetine hydrochloride – estimated effect size: 33%) listed in the appendix to the NEJM paper as unpublished had in fact been published in peer-reviewed journals, the company said.
“The results of HMAT-A and HMAQ-B were published twice — first in the Autumn 2002 issue of Psychopharmacology Bulletin, and again in the Primary Care Companion Journal of Clinical Psychiatry in 2003,” Lilly stated. “In addition, these studies were presented at one or more medical congresses that require peer review of abstract submissions and they also have been available to the general public on LillyTrials.com since 2004.”
The point of contention here was that the authors of the NEJM paper “decided not to count studies as “published” if the manuscript included data from two or more studies,” the company noted, adding: “While this methodology might be suitable for an academic discussion, it’s clearly not the appropriate standard for determining whether a company has been transparent in disclosing its data.”
Industry leader
Lilly claims to be “an industry leader” in ensuring transparency of clinical trial data, committed to disclosing its results publicly whether they are favourable or not and “in an accurate, objective and balanced manner in order for our customers to make more informed decisions about our products”. In December 2004, it pointed out, “Lilly was widely recognized as the first pharmaceutical company to voluntarily launch a clinical trials registry, where we post the results of all Lilly-sponsored registration clinical trials for all of our marketed products dating back to 1994, and all clinical trials for marketed products since December 2004.”
The US industry association, Pharmaceutical Research and Manufacturers of America (PhRMA), was quoted in The New York Times as criticising the NEJM study for not mentioning the efforts industry and government had made since 2004 to make clinical trial data more transparent.
