The ‘best’ drugs may simply be those with the most shamelessly biased data. Bias is often introduced already in the design of the trial, but independent physicians who challenge the design can be fired and may acquire a negative reputation among other drug companies as well for not being ‘cooperative’.7
One of the best safeguards we have against biased results is to establish a central adjudication committee, blinded for the drugs, which decides whether an adverse event occurred or not. However, if such a committee is fed biased and selected information from the sponsor, it will end up putting its quality stamp on a deceitful trial. This seems to have occurred for three major cardiovascular trials that were all published in the drug industry’s preferred journal, the New England Journal of Medicine.8,9,10 Independent investigators compared the number of heart attacks as reported by a central adjudication committee in the publications with those reported to the FDA for the same trials.11 It turned out that what was published was seriously misleading and favoured the sponsor’s drug over the control drug in all three cases.
The names of the drugs, trials and sponsors were prasugrel (TRITON, Daiichi Sankyo and Eli Lilly),8 rosiglitazone (RECORD, GlaxoSmithKline),9 and ticagrelor (PLATO, AstraZeneca).10 Compared to the FDA records for the individual study sites, the committee more than doubled the difference between the sponsor’s drug and the comparator in the TRITON and PLATO trials, from 72 to 145 and from 44 to 89 heart attacks, respectively, whereas in the RECORD trial, heart attacks went down from 24 to 8, which was also beneficial for the sponsor.11
These differences are really remarkable. The probability that the larger difference in the PLATO trial had occurred by chance is so low that it will happen in only one of five trillion trials,11 or about once in 20 billion years, which is longer than the universe has existed. In the TRITON trial, the definition of heart attack was changed to a very liberal one towards the end of the trial, raising the heart attack rate to an unprecedented 10% on the control drug, which is also highly suspicious. Finally, an FDA scientist showed that the adjudication of the events in the RECORD trial was also seriously flawed (see Chapter 16).
In the not so distant past, the situation was better. Academic, independent clinical investigators were key players in design, patient recruitment, and data interpretation in clinical trials.12 Twenty-five years ago, I led the Nordic Coordination Office for AIDS trials, and after we had conducted a trial sponsored by the Nordic Medical Research Councils,13 we negotiated with a drug company about performing a trial with the company’s product, sponsored by the company. During a meeting with company representatives and academic investigators from all over the world, I suggested a change to the trial protocol, which was in the patients’ interest, as it addressed the – undoubtedly negative – impact of the drugs on the patients’ quality of life. To my big surprise, an Australian professor remarked that my proposal was not in the company’s best interest. I was so baffled to discover that an academic investigator who was going to enrol patients behaved like this that I still recall his name: David Cooper. In the coffee break, I discussed the event with some of my colleagues who were equally appalled as I was that Cooper seemed to put profits before patients, and one guessed on the amount of money he received by ‘consulting’ for the company.
In the end, we decided to perform another large AIDS trial in the Nordic countries alone, funded by Bristol-Myers Squibb that respected that our academic freedom was not for sale. We did everything ourselves. We wrote the protocol, monitored the trial, analysed it and wrote the report for publication, after which I visited the company’s headquarters in Connecticut and told them about our results.14 The company never interfered with anything we did. It was a rare example of what I consider the ideal way of collaborating with a drug company.
Today, academic investigators have little or no input into trial design, no access to the raw data and limited participation in data interpretation.12 A saying commonly attributed to Josef Stalin is that those who cast the votes decide nothing whereas those who count the votes decide everything. The drug industry has hijacked clinical trials for marketing purposes, thereby making a mockery of clinical investigation, misusing a powerful tool, and betraying the trust and altruism patients exhibit when they volunteer to participate in trials.12
We have investigated the lack of academic freedom and honest scientific inquiry. In 1994–95, the research ethics committees in Copenhagen approved 44 industry-sponsored trials, which were subsequently carried out and published. It was stated explicitly in 22 of the 44 trial protocols that the sponsor owned the data or needed to approve the manuscript or both.15 Not a single one of the 44 trial reports mentioned anything about that the clinical investigators had participated in the trials with tied hands and had effectively accepted that if the results or their interpretation didn’t please the sponsor, they might never be published.
When we submitted our results to JAMA, we were met with the usual industry excuse that these were old trials and it’s much better now. In agreement with the editor, we therefore sampled a new set of protocols, from 2004, for studies that were ongoing. The industry’s practice had not improved; it had become worse. There were 27 protocols out of 44 that stated ownership to data or control over the publication, similar to 1994–95, but it seemed that the industry now tried to hide what it was doing. Thirteen of the new protocols mentioned separate publication agreements with the investigators, in contrast to none of the protocols from 1994–95, and none of these secret agreements were available in any documents filed with the research ethics committees.
For confidentiality reasons, we were only allowed to see those pages in the new protocols that addressed publication rights. For the old protocols, we had access to everything and it was clear that the sponsors had tight control over their trials. It was stated in 16 protocols that the sponsor had access to accumulating data, e.g. through interim analyses and participation in data and safety-monitoring committees. Such access was disclosed in only one corresponding trial article. An additional 16 protocols noted that the sponsor had the right to stop the trial at any time, for any reason; this was not noted in any of the trial publications. The sponsor therefore had potential control over a trial in progress in 32 (73%) of these studies. When the sponsor can peep repeatedly at the data as they accumulate, there is a risk that the trial will be stopped when it is favourable to the sponsor. Trials reported as having stopped early for benefit exaggerated the effect by 39% compared to trials of the same intervention that had not stopped early.16
None of the protocols or trial publications stated that the investigators had access to all of the data generated from the trial or had final responsibility for the decision to submit for publication without requiring approval from the sponsor.
These findings are deeply worrying. Among the protocols we examined, a sponsor had the potential to prevent publication in half of the trials and had recourse to practical or legal obstacles in most of the others. Surveys of US medical schools17,18 have shown that they frequently engage in industry-sponsored research that fails to adhere to editorial guidelines regarding trial design, access to data, and publication rights.19
A survey from 2005 was particularly shocking. It showed that 80% of the medical schools would allow a multicentre trial agreement that granted data ownership to the sponsor, and 50% would allow the sponsor to write up the results for publication and let the investigators review the manuscript and suggest revisions.18 Ownership of the data was a tough issue; 25% replied that the negotiations were very difficult.
Even after the contract had been signed, 82% of the medical schools had experienced difficulties in a 5-year period, and in one case, the sponsor refused to send the final payment because they didn’t like the results!
The researchers could not study the trial agreements directly because sponsors generally require that institutions keep them confidential. It is likely that the extent of the problems was underestimated, as it is uncomfortable to admit in an au
diotaped telephone interview that your institution accepts highly dubious practices. Nonetheless, 69% of the administrators said that competition for research funds created pressures on them to compromise the conditions in the contract.
This study shows that academic drug research in the United States has been almost totally corrupted by industry. The companies shop between the various academic centres and choose those who are least willing to raise uncomfortable questions. The Association of American Medical Colleges held talks with drug company officials to explore the development of standardised contract terms, but the discussions fell apart when drug company executives baulked.19
Here is an example of the consequences of the corruption. In 2003, the FDA was reviewing unpublished data from studies in its possession on the use of selective serotonin reuptake inhibitors (SSRIs) in children and adolescents to see if the drugs increased the suicide risk. The academics at the medical schools who had published positive results of these drugs were worried and issued a report in January 2004 defending the effectiveness of the drugs and disputing evidence that their use increased suicidal behaviour. Subsequently, however, the FDA determined that such a risk existed (see Chapter 18). The academic researchers had contacted the companies to get access to the data they had themselves generated, but some drug companies refused to turn over the data. This decision could not be disputed because the medical schools, in agreeing to run the trials, had signed agreements with the drug makers that kept the data confidential.19
According to the voluntary principles of the Pharmaceutical Research and Manufacturers of America, sponsors own the study database and
have discretion to determine who will have access to the database … Sponsors will make a summary of the study results available to the investigators. In addition, any investigators who participated in the conduct of a multi-site clinical trial will be able to review relevant statistical tables, figures and reports for the entire study at the sponsor’s facilities, or other mutually agreeable location.20
Don’t you find it scary that the only people in the world who have seen the entire dataset in industry trials are company employees? I do.
If, despite all the precautions, disaster strikes and the results show that the competitor’s drug is best, the easiest way out is to bury the trial. An industry insider once told me that in such a predicament the investigators were informed that, most unfortunately, the company had screwed up the randomisation so it wasn’t possible to say which patients had received which of the two cancer drugs being compared. That ended any discussion about publication before it even started.
The situation has deteriorated substantially. In 1980, 32% of biomedical research in the United States was financed by the industry, and in 2000, it was 62%.21 Currently, most trials are industry sponsored, both in the EU and in the United States.18,22 However, the proportion of the industry’s projects that go to academic medical centres has decreased dramatically, from 63% in 1994 to 26% in 2004.20 It is now mainly private companies, so-called contract research organisations (CROs), that run trials and some of them also work with marketing and advertising; yet another sign that industry trials are marketing ploys.
In order to compete with the CROs, academic medical centres have set up clinical trials offices and openly court the industry, offering the services of their clinical faculties and easy access to patients.23 Thus, instead of fighting the corruption of academic integrity, the academics participate in a race to the ethical bottom, making it less and less likely that any outsiders will ever get to see the data.
Doctors have accepted that they are no longer partners in the clinical research enterprise, but merely provide patients for the trials, in return for publications and various benefits, above all financial support that can be used for other research at the clinic or as a supply to the doctor’s private economy. Specialists may receive as much as $42 000 for enrolment of one patient in a trial, which the US Department of Health and Human Services described in a report with the telling title, Recruiting Human Subjects: pressures in industry-sponsored clinical research.24 With such copious amounts of money at stake, it is difficult to believe that patients are never coerced into participating.
When I started to work in the drug industry in 1975, there was still a good deal of respect for doctors among industry employees and there were limits as to what one could get away with. There was a reasonable degree of academic freedom for industry investigators and it was more prestigious to work in a clinical trials department than in a marketing department.
In the 1980s, this changed quickly. Marketing people became louder and more aggressive, both internally and towards doctors, and clinical trials became integrated in marketing. Modern business managers or salespersons with little or no sense for science or medicine – sometimes with a background of selling refrigerators or cars, or a lower rank in the military – replaced research heads and took over control not only of clinical research but also basic research, with disastrous consequences for innovation. An industry insider has explained how highly useful drugs like acyclovir for herpes, zidovudine for AIDS, and cimetidine for stomach ulcers hardly made it to the market because the managers couldn’t see the need for them.25 The merger mania created stiff and bureaucratic corporate cultures with milestones, flowcharts and decision trees – which is not how scientists work – and the blockbuster mania changed the focus from innovation to me-too drugs.
In his autobiography, the grand old man in Swedish medicine, cardiologist Lars Werkö, tells a similar story. Werkö spent many successful years at Astra and became the head of its pharmaceutical division, but the company deteriorated when a salesman took over as CEO and started to focus on cough medicines and other useless bazaar products, instead of continuing saving people’s lives with drugs against heart attacks and stroke.26 Werkö was thrown out of the board of directors after he had pointed out on several occasions that the proposals about research put forward by the CEO – who knew virtually nothing about medical research – were based on erroneous assumptions. Werkö explains that to argue methodologically with scientific facts is difficult and takes too much time; what mattered was to sell an idea and have the right supporters. In the academic world you could discuss, demonstrate your preferences, and try to argue your case even if it involved criticism of other participants’ views – which it often did – whereas anything of this kind was unthinkable in Astra’s board of directors where the decisions had been made before the meetings. Objections were not welcomed even when the facts as well as the decisions were obviously erroneous; saving face was more important.
I knew Werkö, who accepted to be on the advisory board for The Nordic Cochrane Centre when I founded it in 1993, and it is very disheartening to read his book about these events. In the past, several drug companies were founded by visionary and idealistic scientists who genuinely wanted to help the patients, e.g. George Merck said in a speech in 1950 that Merck tried never to forget that medicine is for the people and not for the profits.
The science shaded into marketing and the professors ended up as promoters while some industry scientists were sickened by the process they had become involved in,27 but there was nothing they could do. Good manners were gone forever and greed became the norm that trumped everything else. The profit per unit sold has always been much higher in the drug industry than in other industries, e.g. 11% in 1960 compared to 6% in all the Fortune 500 companies, big pharma included.28 But in the 1980s, when the marketers took over, the drug industry’s profits skyrocketed and was 19% in 2011 (see Figure 5.1). In 2002, the combined profits for the 10 drug companies in Fortune 500 exceeded the profits for all the other 490 businesses put together.29
Figure 5.1 Average profit rate (in percent) for Fortune 500 companies (including drug companies) and for drug companies alone
Marketing drugs is so prosperous that the US sales force doubled in just 5 years, from 1996 to 2001, and a paper with the telling title ‘The drug pushers’ described that the average retu
rn for each dollar spent on detailing was 10 dollars!30
Randomised trials were introduced in order to protect us from the many useless treatments on the market, but oddly enough, they have given the ultimate power of knowledge production to big pharma that now use them for getting approval for treatments of little or no value and which are often too harmful.
Marcia Angell, former editor of the New England Journal of Medicine, said in 2010 that ‘It is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgement of trusted physicians or authoritative medical guidelines. I take no pleasure in this conclusion, which I reached slowly and reluctantly over my two decades as an editor.’31
Curt Furberg, a seasoned clinical trialist, lamented the lack of academic freedom in partnerships with industry in this way: ‘Companies can play hardball, and many investigators can’t play hardball back. You send the paper to the company for comments, and that’s the danger. Can you handle the changes the company wants? Will you give in a little, a little more, then capitulate? It’s tricky for those who need money for more studies.’32
The most eloquent description I have found of a system that has broken its social contract with society and with patients – who volunteer for trials to advance science and not to increase commercial profits for a particular company – was given by deputy editor Drummond Rennie at JAMA:33
WHAT IS A TRIAL? The approval process starts with evidence gleaned from clinical trials. It might be instructive to compare the sort of trials with which clinical researchers are familiar with those that go on in the courts. It seems to me fundamental that the legal trial carries credibility and retains force and respect with the public because the various parties, judge, jury, opposing counsels, witnesses and police, are independent one from another.
Deadly Medicines and Organised Crime Page 11