Deadly Medicines and Organised Crime

by Peter Gotzsche

  AstraZeneca switches patients to expensive me-again omeprazole

  The power of money in corrupting doctors’ judgements is perhaps best illustrated by the stereoisomers. Usually, only one of the two halves, which are mirror images of each other, is active, but when the patent runs out, the company may patent the active half, a trick called evergreening, or ‘me-again’. Our patent laws are really weird since they allow this, which merely benefits the company for no societal gain.

  The proton-pump inhibitor omeprazole (Losec, Prilosec), used for stomach ulcers and related conditions, was the world’s best-selling drug in the late 1990s. When the patent ran out in 2001, AstraZeneca had extracted the most active half, which has its own chemical name, esomeprazole (Nexium). Generic versions of omeprazole were ready to enter the market at a much lower price than Losec and, in a rational world, all patients would now be treated with a cheap version of omeprazole. This didn’t happen. AstraZeneca used illegal methods to keep competitors away.4 It abused its dominant market position; lied to patent lawyers, patent offices and courts in several countries about the date at which omeprazole had originally been given marketing authorisation; replaced a capsule formulation of the drug with tablets and withdrew the capsule authorisation, which made it impossible for manufacturers of generic drugs to market the capsules.

  AstraZeneca produced flawed trials that purported to demonstrate that Nexium was slightly better than Losec. Instead of comparing equivalent doses, AstraZeneca compared 40 mg Nexium with 20 mg Losec, which is a much higher dose.5 It is ludicrous to ‘prove’ that something is better than itself. If I drink four beers instead of one, my mental capacity will deteriorate more, but this doesn’t mean that a beer is stronger than a beer. AstraZeneca did a meta-analysis of three such trials showing that more patients with reflux oesophagitis were healed on the high dose than on the low dose after 4 weeks.5 The result was shown as a relative risk of 1.14, which isn’t informative. I therefore redid the meta-analysis and found a risk difference of only 0.08. Thus, by treating 13 patients (= 1/0.08) with the high dose, one more patient would get an effect, at a cost that was about 30 times higher.

  Thirty times! It would seem impossible to get any doctor to use such a drug, but doctors are willing to do almost anything, no matter how stupid it is, while they say that the information they get from the drug industry is valuable for them (see Chapter 9). AstraZeneca’s violent attack on common sense worked, aided by a series of shady marketing techniques at extremely high cost; the company used $500 million in the United States for its campaign in just 1 year.6 Five hundred million dollars for selling a drug that was 30 times more expensive than a drug that contained the same active substance. What a waste.

  In Germany, AstraZeneca launched seeding trials, and one-quarter of all general practitioners participated in the hoax and were paid for starting patients on Nexium and making a note of how it went.7

  Seeding trials increased the German drug budget with €1 billion in 2008.8 Companies pay doctors as much as €1000 per patient; the patients don’t give informed consent; and the health insurance companies pay for the drugs. This looks like paying kickbacks, but bribery of doctors is legal in Germany if they work in private practice.9 Self-employed physicians (about one-third of all doctors) that accept up to €10 000 from drug companies in cash – or gifts such as computers, equipment or even holidays – will not face corruption charges. Germany’s Supreme Court ruled in 2012 that drug companies cannot be penalised either when paying German freelance physicians to prescribe their drugs. The case leading to the verdict involved a drug salesperson who paid cash to doctors, amounting to a 5% kickback on each product they prescribed. The company’s official explanation was that the money was remuneration for delivering academic presentations, but these seminars never took place. Even more astonishingly, the head of the German Medical Association, Frank Ulrich Montgomery, shared the court’s view that the rights of doctors to operate in an independent professional capacity should be protected. He added that the media coverage of the case was part of a wider behind-the-scenes agenda to tarnish the reputation of doctors. I doubt the media are better at tarnishing the reputation of doctors than the doctors themselves.

  AstraZeneca was also ‘creative’ in Denmark, selling Losec to the hospitals for only 1% of the price, whereas the patients had to pay the full price when they left the hospital. The company used the same trick with Nexium, which was sold for 2% of the price. Because of such tricks, hospitals are now obliged to use the same drug as would be preferred outside hospital.

  A couple of years ago, I discussed ulcer drugs with a chief gastroenterologist at a meeting. He firmly believed that Nexium was a better drug than Losec and therefore used Nexium. I fail to understand this. Are my colleagues dumb or corrupt? I cannot see other possibilities. Roughly half of those in treatment with proton-pump inhibitors have no appropriate indication,10 and expenditure on these drugs was €10 billion globally in 2006. It is difficult for patients to stop, as the use of the drugs disturbs the hormonal homeostasis. This builds up an excessive production of counteracting hormones, which may cause severe gastric symptoms if treatment is stopped abruptly.11

  The rebound phenomenon is a problem with many of our drugs and it is often misinterpreted to mean that the patients need to increase the dose or to take the drug forever, although a much better option would have been to taper off the drug slowly or to take the drug only intermittently, e.g. if you have heartburn. The rebound phenomenon is the reason why we have an epidemic of happy pills (see Chapter 17).

  Pfizer has provided a most bizarre example of me-again. Aricept (donepezil) was the biggest player in the lucrative market for Alzheimer’s disease with over $2 billion in annual sales in the United States alone.12 Four months before the expiry of the patent, the FDA approved a new dose, donepezil 23 mg, which would be patent protected for three more years, whereas the old doses of 5 and 10 mg were not. The advertising was directed towards patients and contained untrue statements, but the scam worked.

  One would have hoped people were clever enough to take either 20 or 25 mg of the drug to save money, but no. And the FDA failed us badly again. Its own medical reviewers and statisticians recommended against approval, as the 23 mg dose didn’t produce a clinically meaningful benefit whereas it caused significantly more adverse events, particularly protracted vomiting. The reviewers added that the adverse events could lead to pneumonia, massive gastrointestinal bleeding, oesophageal rupture and death.13 This didn’t impress the director of the FDA’s neurology division, Russel Katz, who overruled his scientists.

  I must use strong language now. What the hell is going on? We know that big pharma is evil,14 but what about our drug agencies? Why do they side with evil and deceitful drug companies?

  References

  1 Gale EAM. Post-marketing studies of new insulins: sales or science? BMJ. 2012; 344: e3974.

  2 Harris G, Pear R. Drug maker’s efforts to compete in lucrative insulin market are under scrutiny. New York Times. 2006 Jan 28.

  3 Yudkin JS. Post-marketing observational trials and catastrophic health expenditure. BMJ. 2012; 344: e3987.

  4 Hawkes N. AstraZeneca must pay €52.5m fine for anticompetitive tactics, rules European court. BMJ. 2012; 345: e8396.

  5 Edwards SJ, Lind T, Lundell L. Systematic review of proton pump inhibitors for the acute treatment of reflux oesophagitis. Aliment Pharmacol Ther. 2001; 15: 1729–36.

  6 Relman AS, Angell M. America’s other drug problem: how the drug industry distorts medicine and politics. The New Republic. 2002 Dec 16: 27–41.

  7 Grill M. Kranke Geschäfte: wie die Pharmaindustrie uns manipuliert. Hamburg: Rowohlt Verlag 2007.

  8 Tuffs A. Germany sees rise in post-marketing studies. BMJ. 2009; 339: b4199.

  9 Hyde R. German doctors free to take cash from drug firms. Lancet. 2012; 380: 551.

  10 Forgacs I, Loganayagam A. Overprescribing proton pump inhibitors. BMJ. 2008; 336: 2–3.

  11 McKay
AB. Overprescribing PPIs. BMJ. 2008; 336: 109.

  12 Schwartz LM, Woloshin S. How the FDA forgot the evidence: the case of donepezil 23 mg. BMJ. 2012; 344: e1086.

  13 Lenzer J. FDA is criticised for licensing high dose donepezil. BMJ. 2011; 342: d3270.

  14 Goldacre B. Bad Pharma. London: Fourth Estate; 2012.

  16

  Blood glucose was fine but the patients died

  The story of rosiglitazone is one of death, greed, and corruption … The trust between doctor and patient, researcher and participant, or author and editor is undermined when the foundations on which evidence is built are treated with such casual contempt.

  Editorial, The Lancet1

  The FDA approved rosiglitazone (Avandia) in 1999 although there were more thrombotic heart events with the drug than with placebo or active comparators (relative risk 1.8, 95% confidence interval 0.9 to 3.6).2

  The FDA reviewer had adjusted for time on drug, which brought the relative risk down to 1.1. However, as stated in the package insert, the drug increased LDL cholesterol by 19%, which explains its harmful effect on the heart. The cholesterol-lowering drug ezetimibe was approved in 2002 based on a 15%–18% reduction in LDL cholesterol, which was presumed to confer cardiovascular benefits. Thus, a lowering of LDL cholesterol by 15%–18% without evidence of clinical benefit led to drug approval in one case, whereas an increase by the same amount with clinical evidence of harm didn’t lead the FDA to reject rosiglitazone. This illustrates again the failure at drug agencies in protecting public health.

  In Europe, the EMA was so concerned that it rejected the drug, only to approve it a year later despite there being no new evidence. It isn’t clear why, but Silvio Garattini was on EMA’s committee and has described how the companies bring forward paid opinion leaders who give favourable presentations at committee meetings.3

  A member of the committee told the BMJ that he had been contacted by respected members of the diabetes community who urged him to approve the ‘wonder’ drug. Garattini’s view was that there was no need for the drug, as there already were so many that were more or less the same.3 He explained that long-term trials required after marketing approval are highly beneficial for the companies, which have every reason in the world to be so slow with the trials that the drug was off-patent when the bad results came in. An even better strategy was to ignore the demands, and in fact, only about a third of FDA requests for post-marketing studies are ever carried out.3

  In 1999, the company, then known as SmithKline Beecham, completed a trial that found more cardiac problems with rosiglitazone than with pioglitazone, but according to an internal email, ‘These data should not see the light of day to anyone outside of GSK.’3,4 Instead of publishing the results, the company spent the next 11 years trying to cover them up.4 Mary Anne Rhyne, a GlaxoSmithKline spokeswoman, said that the company had not provided the results of its study because they ‘did not contribute any significant new information’.4 Apparently it did, also for Glaxo, as the results made the company decide against further comparisons!

  In 2004, the WHO sent Glaxo an alert about cardiac events and the company performed a meta-analysis that confirmed this, which it sent to the FDA and the EMA in 2006. However, none of the agencies made the findings public because of the proprietary nature of companies’ trial results.3 This absurd interpretation of ownership of data and results is not only deeply unethical, it is also wrong, as it violates the fundamental principles on which the European Union is founded (see Chapter 11).5 But as long as we allow regulators to believe in their own nonsense and putting profits before the survival of patients, it allows the companies to ‘push the drug aggressively and hope they can make a billion dollars before someone finds out’, as former editor of the New England Journal of Medicine, Jerome Kassirer, expressed it.6 Rosiglitazone was Glaxo’s second-best-selling drug, at about $3 billion a year,3 and Glaxo behaved like drug pushers in the street, as they could have informed the public about the dangers with its drug but didn’t.

  In 2006, Glaxo sent an updated analysis to the FDA with five more trials confirming the harm, but, yet again, the FDA failed to warn the patients and the physicians.2 Perhaps the FDA was duped by an observational study Glaxo had also submitted, performed by a commercial vendor, which showed no increase in risk? However, Glaxo had carefully avoided to report to the FDA what this study had shown when rosiglitazone was compared with pioglitazone. This comparison showed that rosiglitazone led to more admissions to hospital with myocardial infarction than pioglitazone.2 I believe the omission is scientific misconduct, given that Glaxo already knew that pioglitazone is a better drug.

  Rosiglitazone was now the most sold diabetes drug in the world, but in 2007, all hell broke loose for Glaxo. As part of a legal settlement in relation to the company’s fraud with paroxetine (see Chapter 18),3,7 Glaxo was required to post the results of its clinical trials on a website. This enabled independent researchers Steven Nissen and Kathy Wolski to have a closer look at rosiglitazone. Their 2007 meta-analysis of 42 trials, 27 of which were unpublished, showed that the drug causes myocardial infarction and cardiovascular death.5,8,9

  Diabetes drugs are supposed to lower cardiovascular mortality, not increase it, but, as just noted, the shocking news was not news for Glaxo.10 The company had known about this for 8 years but failed to warn the regulatory authorities and the public. Three years later, the US Senate Finance Committee released a 334-page investigation of rosiglitazone and Glaxo, which mentioned internal company emails and documents that give us a rare insight into the conduct of a major drug company.9

  Nissen and Wolski submitted their meta-analysis to the New England Journal of Medicine on 1 May 2007. The manuscript was sent for peer review and only 2 days after submission, an academic peer reviewer broke the rules and faxed the manuscript to Glaxo.9 Despite its confidential nature, Glaxo circulated the manuscript to more than 40 scientists and executives at the highest levels in the company.11 On 8 May, Glaxo’s head of research admitted internally that the FDA and Glaxo itself had come to similar conclusions about the increased risk with rosiglitazone as the submitted meta-analysis did.11 Yet the next day Glaxo had its key lies ready, which they called ‘key messages’, and which were that the meta-analysis was based on incomplete evidence and that the company strongly disagreed with its conclusions.

  Already on 10 May, four Glaxo scientists and executives met with Steven Nissen after having asked for a meeting.9 As Glaxo had previously threatened John Buse (see Chapter 19), Nissen secretly taped the meeting. Because of Nissen’s meta-analysis, Glaxo had decided to unblind the collected data on its ongoing RECORD trial, which the EMA had required the company to carry out because of cardiovascular safety concerns when it approved the drug in 2000.3 An internal email suggested that if the independent academic steering committee for the trial wouldn’t agree to publish interim results, the company would pursue the line that ‘a decision has been made – live with it’.11 Glaxo convinced the steering committee that an interim analysis should be published, but the committee didn’t know that Glaxo had already unblinded the results 2 weeks earlier. The committee apparently believed it was their decision to unblind the study and publish.

  At the meeting with Nissen, an executive said, ‘Let’s suppose RECORD was done tomorrow and the hazard ratio was 1.12.’9 This comment was made 4 days before the company claimed it unblinded the trial and 14 days before the steering committee was asked to approve unblinding. The hazard ratio that was published was about the same, 1.11.

  Funded by Glaxo, Philip Home et al. published what they called ‘an unplanned interim analysis’ electronically in the New England Journal of Medicine only 2 weeks after Nissen and Wolski published their meta-analysis in the same journal on 14 June. Glaxo succeeded to publish a large trial reporting on 4447 patients followed for 4 years, only 7 weeks after they heard about a meta-analysis that threatened the survival of their product. In contrast, it can take companies 5 or 10 years to publish results they don’t like,
if they publish them at all. Companies are surely able to act fast in the case of a drug emergency.

  What made the New England Journal of Medicine decide to publish an unplanned interim analysis of an ongoing trial, to publish it so quickly and to accept it despite its poor design (e.g. the trial drugs weren’t even blinded)? An FDA scientist, Thomas Marciniak, said that the FDA would have found the trial’s design unacceptable.3 My take on this is that the journal has far lower standards for industry trials than for other types of research and that it has allowed its integrity to be corrupted by big pharma for financial gains (see also Chapters 5, 6, 13 and 14).

  There were eight authors. One was from Glaxo and the other seven were ‘consultants’ on company payroll.11 They talked about ‘exceptional circumstances’ (but didn’t specify that these were that one of their comrades had stolen Nissen’s manuscript) motivating them to report unplanned interim findings and they regarded their findings as ‘inconclusive’.12 It’s unbelievable and scandalous that the New England Journal of Medicine let them get away with this. Nowhere is the reader told what the ‘exceptional circumstances’ were and the editors didn’t ensure the authors explained it in the paper. When the final results were published in the Lancet 2 years later,13 they appeared to be false.9 The event rate for heart attacks was less than one-third of that observed in a similar trial with pioglitazone, and the paper claimed that rosiglitazone was administered during 88% of the follow-up, which was mathematically implausible, given other information about the trial.9