How is science made in a pharmaceutical company, when billions of Euros may depend on the outcome of a scientific experiment? A partial, if incomplete, glimpse at the process can be seen in the documents released last year in a court case in the US on the AstraZeneca medicine Seroquel. (Seroquel is a “blockbuster” drug, authorised for the treatment of schizophrenia and other conditions, with annual sales of over $4 billion.)
The documents were widely covered in the US media but my focus was on how the company dealt with the science regarding the weight gain that seemed to be associated with the medicine from the beginning.
The most widely reported document was perhaps the“smoke and mirrors” e-mail from 1997, where the writer praised “Lisa” for having done a great “smoke and mirrors job” on Trial No 15′
A later e-mail from 1999 commented: “Thus far we have buried Trials 15, 31, 56 and are now considering COSTAR” although in this case the writer seemed to question if the policy of “cherry picking” data should or could continue. Of one study he said: “The reporting of the COSTAR will not be easy. We must find a way to diminish the negative findings. But, in my opinion, we cannot hide them.”
Some studies indicated little or no weight gain, including a case study of one patient that was apparently used in a brochure for Seroquel. The study was authored by an important “customer” of the company, although internal company e-mails were strongly critical of research done by the author or his associates – on issues such as informed consent (lack of), and quality (low) of clinical practice.
Overall, the documents show the company wrestling for many years with the issue of weight gain and how best to present it, or at least gain a comparative advantage. The concept of “weight neutral” was mooted as a description of the outcome. This could cover, so far as I could see, the possibility that some patients gained weight and others lost weight. The problem with this of course is that the patient who gains weight would not be consoled by the fact that another patient lost weight.
In June 2000 it was agreed that the phrase limited weight gain would be replaced by the phrase “weight gain” but internal opposition “from the commercial side” appeared to delay the change for two years.
Study 41 (on Seroquel SR) was described as “red”, meaning a “failure of the primary outcome measures of the study…”. One e-mail cited clear instruction from the highest level …not to discuss details surrounding Study 41 with any external customers” – meaning prescribers and researchers.
The disclosed documents also included e-mails relating to requests from third parties for research funding. One 1998 request was approved internally “provided we see the data before publication and can exercise the right to withhold publication if it muddies the waters too much”.
A request in 1997 for funding for pre-clinical research on Seroquel drew the following response: “…the decision to fund any further pre-clinical work will depend on the competitive advantage that the work can demonstrate for Seroquel…… This is primarily due to the fact that R &D is no longer responsible for Seroquel research – it is now the responsibility of Sales and Marketing.
In an e-mail in 2000 a researcher, who was refused funding, was given a sample of Seroquel and advised “Should you be able to differentiate Seroquel from other atypical antipsychotic drugs, I would be more than happy to present your results to the business side for possible further support”.
The company’s US branding team strongy opposed a proposed study that would have included a comparison between Seroquel and a competing product, on the grounds inter alia that “US prescribers believe SEROQUEL has a better safety profile than risperdai . Data showing risperda! has a better safety profile than SEROQUEL may impact US sales.
In response to media reports on these and other disclosures AstraZeneca said, inter alia, that the extracts were taken out of context. They also said that AstraZeneca “shared all relevant and required data with the FDA”. Obviously I cannot contradict that claim.
On the issue of context, the disclosed documents do show that there were pressures militating against good science around a medicine potentially (and actually in this case) worth billions of dollars in sales. I don’t know if these pressures were effective or not. However, if certain studies were not published, and not subject to public review, the best science was not available and patients and doctors may have been deprived of the opportunity to make an informed decision on different therapy options.
Science is a kind of competition between ideas, and it works best when everyone plays.
Even third party researchers of the highest integrity may be unconsciously biased if they are told they will gain from one outcome rather than another. As for researchers whose integrity is not of the very highest…
The documents I’ve quoted come from one company. Are the same pressures at work in other large pharmaceutical companies. I would be glad to hear readers’ views on this. More to the point, what can be done to counteract these pressures? How can we improve the way in which science is made in this context?
Your comments please.
Sequel to Seroquel
I have focused here on the issue of weight gain but the company has faced a number of lawsuits on other issues in recent years. One lawsuit alleging illegal marketing was settled without admission of liability for around half a billion dollars. A similar settlement may be reached with some 20 000 or more patients/plaintiffs in a product liability class action, also without admission of liability and, I fear, without any further disclosure of the company’s internal documents on diabetes.
The weight gain issue has not gone away. Earlier this year, the UK Pharmaceutical Industry self regulatory body ruled as misleading a company claim (in 2004) that Seroquel had a “favourable weight profile..” as compared to its rivals.