Transparent clinical trials data

A report on the decision of drug manufacturers and international pharmaceutical associations to reveal the results of clinical trials, whether or not they were successful

Clinical trials are an essential part of the drug discovery process, with huge amounts of money being spent by the pharma companies on trying to establish that their potential new products work - and work better than the alternatives already on the market. But one of the many criticisms levelled at the industry is that it publishes only the 'good' results, while a full analysis of trials whose results are not what might have been hoped never seeing the light of day.

Now, in response to major criticism by the editors of some of the most important journals in which the companies publish their results, it looks as though the industry has, in principle, agreed to a system by which the results of all trials on drugs that reach the market will be made available in a publicly accessible database, regardless of the outcome of the trial.

Two companies are already making all their trials data available to anyone who is interested. GlaxoSmithKline led the way, in response to a number of allegations that it had buried unfavourable results on its blockbuster antidepressant Seroxat (paroxetine). It first said that it planned to set up a trials data website in June 2004, and the data first started being deposited on the website last September. According to the company's chairman of r&d, Tadataka Yamada, the register will be a 'major advance in providing on-line access to information to support patient care, facilitating access to study summaries by putting them on a single internet site'.¹

clear answers

Lilly made a similar announcement at the beginning of August. Its public clinical trials registry² contains data on all trials from Phase I through to Phase IV on its marketed products, plus announcements of the commencement of all Phase III and IV studies. 'Lilly understands that patients, customers and critics are looking for transparent answers that provide value to the healthcare decision making process,' says Lilly's chairman, president and ceo Sidney Taurel. '[The registry] represents a comprehensive effort to publicly disclose Lilly's clinical trial information. These actions should prove to be invaluable for patients and the medical community as they seek to make informed decisions about Lilly medicines.'

The proposed new system has been put together by four international pharma associations - the European Federation of Pharmaceutical Industries and Associations (EFPIA), the International Federation of Pharmaceutical Manufacturers and Associations (IFPMA), the Japanese Pharmaceutical Manufacturers Association (JPMA) and the Pharmaceutical Research and Manufacturers of America (PhRMA). At the beginning of January, they published the Joint Position on the Disclosure of Clinical Trials Registries and Databases, under which summary results of industry-sponsored clinical trials on medicines that are approved for marketing will be disclosed on a cost-free, publicly accessible database, regardless of the trial's outcome.

In addition, details of all ongoing clinical trials will be publicly registered at the outset of the trial. The industry has undertaken to implement both parts of the announcement in 2005, and has also committed to make information available on all clinical trials, other than very early exploratory trials. And even the results of these early trials will be published if they have 'significant medical importance'.

trial details

The registry of ongoing trials should contain sufficient basic information about the trials so that interested potential participants and their doctors can find out how they might enrol in the trial. At the very least, this should include, in addition to the title, a description of the trial in lay terminology; trial phase; type of trial, such as interventional; the trial's status; its purpose, for example treatment, diagnosis or prevention; and intervention type - whether it is a drug, a vaccine and so on; and what condition or disease it is aimed to treat. In addition it should contain the trial's key eligibility criteria, such as gender and age; its location, and contact information. The trials should also be given a unique identifier so that in future it will be easy to track down the results and marry them up with the original protocol.

According to EFPIA, the results of trials should normally be published within one year of the medicine's approval, and for post-approval trials, within a year of their completion.

public awareness

'The industry recognises that there are important public health benefits associated with making clinical trial information more widely available to healthcare practitioners, patients and others,' says EFPIA's director general, Brian Ager. 'By making public not just the results of trials that have taken place - whether positive or negative - but also information those that are just starting, the industry has made a major step towards achieving greater transparency.'

The move has also been welcomed by the Association of the British Pharmaceutical Industry, which was involved in the discussions as a member of EFPIA. Director general, Richard Barker says: 'The UK-based industry took a world lead in providing the public with information about clinical trials work more than 18 months ago when the ABPI established its own website for companies to register information about trials. We are very pleased that the industry globally is moving to build on this by providing information about trials wherever they may have been conducted, and the initiative has the full endorsement of the industry in the UK.'

unsuccessful trials

The information will be published as a standard non-promotional summary. This will contain details of the trial's design and methodology, the results of primary and secondary outcome measures, as well as safety results. The database will also provide a link to any publication of the trial results in a peer reviewed medical journal.

The obvious problem with information about unsuccessful trials being undisclosed is the insinuation that the company running the trial has something to hide. However, it could all too easily be a lot more damaging than that.

As Richard Sullivan, Cancer Research UK's head of clinical programmes, told the British Association Festival of Science in Exeter last year, if a trial isn't published, it can be nigh on impossible for a doctor to make a properly informed decision about treatment - a particular problem in the cancer field. Indeed, at least a quarter of all cancer trials are never published. Of more than 500 Phase III trials that had been given preliminary announcements at American Society of Clinical Oncologists meetings over a decade, 26% were not published within five years of the presentation.³

The most common reasons given for not publishing the results were lack of time and shortage of money. Worryingly, however, if the trial showed that the investigational drug was no better than the standard treatment, then its results were much less likely to be published than trials that had a positive outcome.

This is not necessarily a sinister manipulation of the data, and it is perhaps understandable: the researchers lost interest because the trial didn't work and so they spent their time and effort elsewhere. And, anyway, the important journals are only really interested in publishing positive trial outcomes.

It is, however, not a good situation. '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 said. 'This in turn can affect the validity and findings of subsequent reviews, treatment decisions and clinical practice guidelines.'

Furthermore, he claims that the current lack of regulation of trials means that around a third of all clinical trials are never even registered.

And, claims 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, researchers could waste time and repeat lines of inquiry that have already been proven unsuccessful.'

addressing criticism

The recent announcement covers only industry sponsored trials, however, and ABPI's Barker has called for other bodies that sponsor trials to follow suit.

But if the proposals are fully implemented, it does mean that the vast majority of trials will have their results placed in the public domain, regardless of whether a full peer reviewed paper is produced. This will greatly increase transparency, and the appearance of trial results where a new treatment is shown to be no better - or even worse - than existing therapies will certainly go some way towards addressing the criticism that the pharma industry hides the bad news.

It should also mean that less valuable time is wasted by clinical researchers repeating trials that had already proved unsuccessful elsewhere.