DocCheck reported already one and a half years ago about the dark secrets kept in the vaults of major pharmaceutical companies. An objective evaluation of many common medications is made difficult due to the lack of data from major studies. This involves data that was indeed collected but never reached the public via journals. For example, in oncology: from 1989 to 2003 about 700 phase III trials were presented at the annual meetings of the American oncologists. Every fifth investigation still lays in the drawer of the sponsor even five years after the presentation. The majority of them would have had a significant impact on treatment. For medications that were approved between 1998 and 2000 in the U.S., five years later more than half of all relevant studies were still absent from the list of publications.
Oseltamivir under criticism
One preparation in particular continually ignited stormy disputes in recent years over allegedly withheld trial data: oseltamivir, developed and marketed by the Roche Group against flu viruses. In 2009, researchers of the Cochrane network had doubts for the first time about its antiviral activity being really as effective as shown in the publications. Recently a new assessment from Germany’s Institute for Quality and Efficiency in Health Care (IQWiG) appeared, which attested to an earlier subsidence of flu symptoms using oseltamivir. Whether it could prevent serious course of flu is not something able to be judged without the relevant study documentation. While Roche insists that it has brought all the key data to the public, Peter Doshi of the Cochrane community and others have requested access to additional documentation.
Missing trial data: Controversial publication rights
By now, most clinical trials get registered centrally. Besides Roche’s flu remedies, there are still a number of other medications for which full documentation is missing. Several initiatives intend to now change that. YODA (Yale Open Data Access) aspires to have companies voluntarily supply incomplete or incorrectly reported data. Harlan Krumholz of Yale University, the person responsible for this initiative, justifies his actions as follows: “With the current rules the owners of the data benefit from its exclusivity. But for products used commercially there arises with this privilege also a responsibility to share all knowledge about risks and benefits.” In at least one case YODA has already had success: The firm Medtronic supplied all data on their controversial recombinant agent for bone regeneration (human bone morphogenetic protein-2).
A second initiative from Peter Doshi of Johns Hopkins University School of Medicine in Maryland wants much stricter dealings with the industry: The sponsors should publish unpublished trial data within one year or forfeit their ownership of it. This data could then be made available in relevant journals by other scientists or institutions to all interested parties. In his call RIAT stands for “restoring invisible and abandoned trials“. Of course, the publication should be of appropriate quality, be guaranteed by independent experts and having stringent requirements placed on the text.
Disclosure of individual patient data?
In his article in the British Medical Journal Doshi writes that he is in possession of more than 150,000 pages of study documents that are waiting to be put in summary form to inform the scientific and medical community. Documents which he has fought for through the Freedom of Information Act. That such action works is something of which Doshi is certain: “I see no legal obstacles to the publication of orphaned studies”, he is quoted in taz Berlin as saying, “but it is difficult to predict whether those who do not like our foray will try to stop this by legal process”.
The pharmaceutical companies are defending themselves however. In the words of the European umbrella organisation EFPIA: “It is however necessary to protect patient data and commercially sensitive information in the process.” Therefore the entire body of raw data should not be uploaded. Even if patient data were to be made anonymous, an experienced puzzle solver could nonetheless draw conclusions on the identity of the participants. Additionally, no doctor or researcher with a publication at hand would, even in the case of an oft used medication, fight through hundreds or even thousands of pages of masses of data in order to learn about risks or side effects of a medication.
Doshi want to introduce the RIAT initiative over the course of the next year at various conferences and hopes to find consensus with the authorities and with the industry as well. Cooperating partners on the part of journals already exist: The British Medical Journal and PLOS Medicine would be willing to publish data even if it were already a little long in the tooth, but perhaps still controversial.
EMA: Staggered information sharing
Regarding the publication of future studies, the EU and its medications agency EMA have recognised that things cannot continue on as if ‘business as usual’. “Publication and Access to Clinical Trial Data” is one draft plan of the EMA according to which the operator of the studies is supposed to pass their data to the public or to interested outsiders. Data that does not identify individual patients and also contains no commercially exploitable trade secrets should then obtain an “open access” label. This procedure should also apply to individual requests for study data files on individual participants once restrictions targeting potential abuse have been passed. Data on investigational, bioanalytical data or in-vitro data which is commercially usable, such as in the development of generics, should be released only in consultation with the sponsor of the EMA.
Increasing pressure on pharmaceutical companies
Initiative AllTrials, supported by the Cochrane Society, the British Medical Journal and the Drug Commission of the German Medical Association, calls for registration of all trials. Their results should be made publicly available no later than one year after completion. This also applies to what are already past studies. That the enforcement of such principles is not easy was shown by a judgment of the European Court of Justice in Luxembourg from April this year, wherein the judges prohibited the EMA from passing on data from two trials for agents against pulmonary fibrosis and rheumatoid arthritis. Nevertheless, the companies want by yielding to general pressure to become more open and to have the applications for issue checked with independent bodies. Retentive behaviour is the outcome of, if for no other reason, of fear of the competition: every third request for data to the EMA comes from the drug companies themselves.
“Study data belongs to the patient”
Courts in 1998 forced several companies in the tobacco industry into publication of internal documents on research and marketing of their products. They finally revealed the attempts made to influence researchers and to play down the effects of smoking. Whether the data on oseltamivir, a medication which governments and private citizens bought and stored away in large volumes during the influenza pandemic in 2009, brings similarly sobering findings is uncertain. However, one group should in any case have the right to know those results which would not have emerged without their involvement: “The study data should belong only to one group of people: the patients”, says Gerd Antes, director of the German Cochrane Centre, clearly. “Because they can be harmfully affected if not all the data is laid out on the table”.