German researchers claim firm hid trial data showing drug to be “ineffective and potentially harmful.”
A meta-analysis of largely unpublished trials with Pfizer’s EU-marketed antidepressant reboxetine shows the drug is “ineffective and potentially harmful”, German researchers claim. The Institute for Quality and Efficiency in Health Care (iQWiG) team suggests sneaky publication tactics by Pfizer mean that up to three-quarters of reboxetine trial data were never released. The iQWiG suggests that the selective data Pfizer did allow to be published actually overestimated the benefits of reboxetine compared with placebo by 99–115%, and of reboxetine compared with SSRIs by 19–23%. Having had to effectively pry the missing trials out of Pfizer’s hands to carry out its meta-analysis, the iQWiG researchers are now calling for mandatory publication of clinical trial data, including data on older agents, worldwide.
Results from the meta-analysis are published in the British Medical Journal (BMJ), in a paper titled, “Reboxetine for acute treatment of major depression: systematic review and meta-analysis of published and unpublished placebo and selective serotonin reuptake inhibitor controlled trials.”
In a separate paper in the same issue of BMJ, two of the meta-analysis authors, iQWiG’s head of the department of drug assessment Thomas Kaiser, Ph.D., and deputy Beate Wieseler, Ph.D., say that although the U.S. and Canada have never granted reboxetine full approval, the drug has been marketed in several countries in Europe since 1997. Data from the new meta-analysis has now led Germany’s Federal Joint Committee to confirm plans to stop reimbursement of reboxetine.
Doubts about the effectiveness of reboxetine were highlighted last year. In February 2009 a paper published in The Lancet reviewed 117 randomized controlled trials from 1991 up to Nov 30, 2007. It compared a range of individual antidepressants in the acute treatment of unipolar major depression in adults. This concluded that “reboxetine was significantly less efficacious than all the other antidepressants tested”.
In June 2009 iQWiG published an initial report on its review of bupropion, mirtazapine, and reboxetine. This concluded that “in the case of reboxetine, iQWiG abandoned its analysis of the study data that was publicly available at that time, because it was evident that the manufacturer, Pfizer, was concealing almost two-thirds of all data collected in trials to date. An analysis of the available data would have produced a biased picture. Despite several requests, Pfizer had steadfastly refused to provide IQWiG with a list of all published and unpublished data.”
At the time, a literature search suggested that reboxetine had been tested in about 4,600 patients in 16 trials. However, data on only about 1,600 of these patients had ever been published. Pfizer was reluctant to provide iQWiG with the remaining trial data.
The published meta-analysis carried out after Pfizer did eventually cough up the missing studies included data from acute treatment trials evaluating reboxetine compared with placebo in 2,256 patients, and reboxetine versus SSRI comparisons in another 2,641 patients. The results, based on the Hamilton depression rating scale, found no significant difference between reboxetine and placebo in terms of remission and response rates, and suggested Pfizer’s drug performed worse than SSRI’s overall for both these outcomes. Reboxetine therapy in addition led to a higher rate of withdrawal due to adverse events than placebo, and treatment withdrawal rates were also higher in reboxetine patients than in those treated using fluoxetine.
“To our knowledge, this is the first systematic review of a comprehensive evidence base of published and unpublished acute treatment trials of reboxetine versus placebo or SSRIs in adults with major depressive disorder,” notes the iQWiG team, led by project manager Dirk Eyding. “The results of our review largely contradict the findings of previous systematic reviews and analyses of reboxetine versus placebo and reboxetine versus active comparators.”
Drs. Kaiser and Wieseler suggest that the FDA Amendments Act of 2007 has partially solved the problems associated with publication bias in drug research. It requires standardized data on protocol and results to be publically posted on www.clinicaltrials.gov for all trials (except Phase I) of drugs subject to regulation by the agency. The European regulatory authorities is also in the process of implementing similar legislation, which will make disclosure of data from the clinical trials database EudrCT mandatory, the authors add.
However, the authors point out, the FDA Act does have loopholes. It doesn’t require the publication of trials completed before September 27, 2007, or of trials of drugs that have never been approved. “Thus, although comprehensive information will in the future be available on newer drugs, published information on older drugs will remain biased, resulting in an unfair comparison of old and new treatment options,” they point out.
The loopholes also mean negative findings on drugs never approved in the U.S. do not have to be made public, even if the drug is approved in other territories. The current regulations in addition allow for a potentially two-year delay in the publication of trials for approved drugs tested in new indications. They are also insufficient to tackle the problem of reporting selected outcomes. “This requires registration of full study protocols, including any amendments, and of plans for statistical analysis,” the authors stress.
The iQWiG team are recommending new measures to help close these loopholes both in the U.S. and globally. These include:
- Extension of the FDA Amendments Act to include drugs for which approval was declined and worldwide implementation of this type of legislation
- Public access to regulatory databases containing trials of older drugs not covered by current law
- Greater data sharing between regulatory authorities, as well as re-evaluation of a drug if approval is declined elsewhere
- Legal obligation for manufacturers to provide all requested data to health technology assessment bodies without commercial restrictions to publication