Doctoring the Evidence: GlaxoSmithKline Pushes Depression Drug


A striking young woman with a talent for painting, Sharise Gatchell was - like many teenagers - painfully self-conscious. At 14, she moved with her family from South Africa to England where at first she found it hard to make friends. By the age of 16 she was acutely embarrassed that her periods hadn't started and felt she was sweating excessively. Just the sort of thing to make adolescence even more painful than usual. Her mother, Stephanie, took her to see a consultant physician at the local hospital. Stephanie Gatchell recalls:

"During that consultation she became a bit emotional because she was explaining to him about the problems she had and how it affected her social life. Then she started crying and I was amazed. He asked her, 'Sharise, do you sometimes feel like ending it all?', and I couldn't believe it when she said 'Yes I do.' And then he started talking about paroxetine and suggesting that she try it."

"Off-label" Prescriptions

Paroxetine - better known as Paxil in the U.S. - is an antidepressant. It is licensed only for adults, but doctors are allowed to prescribe any medicine if they think it will help their patient, a practice called "off-label" prescribing. Sharise was not the only teenager getting paroxetine off-label. Around 7,000 children a year were on the drug in the UK; and many more in the U.S.

Within days of starting on paroxetine, Stephanie noticed a dramatic change in her daughter: she became more confident. But with the confidence came aggression and worse:

"One day in the kitchen her sleeve pulled up slightly and I noticed that there were cut marks on her left arm. I couldn't believe my eyes. She was obviously self-harming while she was on the drug, something she's never ever done before."

Stephanie persuaded her daughter to stop taking paroxetine. But in 2003 Sharise went back on it without telling her mother. When her parents returned from a weekend break, Stephanie immediately realized something was terribly wrong:

"I went upstairs before my husband came in, and the moment I got to the landing at the top and I turned round and looked, she was hanging from the loft hatch. I tried to revive her, but before I even started I realized I was too late."

Next to her daughter's suicide note lay a packet of paroxetine. Stephanie instinctively blamed the drug and now blames herself for letting her daughter take it. But she couldn't have known what the drug's manufacturer had known for years.

GlaxoSmithKline Experiments on Children

UK-based GlaxoSmithKline (GSK), the second largest drug company in the world, which recorded 2006 sales of over $45 billion, had begun a series of clinical trials in the mid-1990s to test whether paroxetine would work in depressed children. Paroxetine had already been hailed as a wonder drug in adults as a treatment for everything from depression and stress to anxiety and even shyness. By the new millennium 100 million paroxetine prescriptions had been written worldwide, bringing in $2 billion a year for GSK and placing the antidepressant a close second to Prozac in popularity.

With the adult market sewn up, the company sought new ways to make money from the drug, or "life cycle management," as the approach is known in the pharmaceutical industry. The U.S. Food and Drug Administration wanted to boost the number of medicines tested for children and had introduced an incentive that would give companies a six-month extension on their patent just for carrying out pediatric trials. For paroxetine alone that would be worth $1 billion. And if GSK could be the first company to prove its antidepressant was safe and effective for children, the rewards would be even higher, as paroxetine could become the market leader.

Hundreds of children were recruited from around the world to take part in three clinical trials. One group was given the drug, the other a placebo. They were randomized controlled trials (RCT) where neither the children nor their doctors knew whether they were taking the active drug or the placebo until the end of the study. This is widely accepted as the best way of working out whether a drug causes a particular effect: the gold standard in terms of evidence.

But the outcome of these the trials was not what GSK had been hoping for. Paroxetine proved no better than placebo. In the biggest trial, Study 329, which was conducted across several sites in the U.S., 11 of the 93 children who took paroxetine developed serious side-effects; seven had to be hospitalized. Significantly more had self-harmed or attempted suicide on the drug than on placebo.

The BBC (British Broadcasting Corporation) obtained confidential case reports from Study 329 which detail what happened to them:

"[13 days after starting paroxetine] the patient became very angry. He punched pictures, broke glass and sustained lacerations that required six sutures... he expressed hopelessness and possible suicide thoughts."

"The patient began exhibiting symptoms of disinhibition, grandiosity and expansive mood around week four [after eight weeks] the patient became agitated and said she would kill herself."

"[11 days after starting paroxetine] the patient was hospitalized for psychosis with auditory hallucinations and superficial cuts."

Yet this was a drug that was supposed to make children happier and reduce the risk of suicide.

The company knew it had a problem. In an internal memo to senior executives in 1998 the product director for paroxetine admitted:

"The results of the studies were disappointing. The possibility of obtaining a safety statement from this data was considered but rejected. Consultation of the marketing teams confirmed that this would be unacceptable commercially."

In other words, publishing the data in full would undermine the drug's safety profile and put profits at risk. It took five years for this information to reach the public domain and only then as a result of the BBC investigation. In the intervening years doctors continued to prescribe a drug to tens of thousands of vulnerable children, ignorant of the fact that it had not been shown to work and was harmful to some.

Lawsuits Reveal Questionable Paper Trail

This has been described by British psychiatrist Dr. David Healy as "the biggest medical scandal since thalidomide." Dr. Healy, rather than the regulatory authorities charged with monitoring the safety of medicines, was instrumental in uncovering the evidence at the heart of the scandal. It was a series of U.S. legal actions that led Dr. Healy, as an expert witness, and the legal team with whom he was working, deep into GSK's secret archives, where they found a series of damning internal documents.

What this paper trail reveals is how the company, with no hope of obtaining a children's license, devised a back-door method of marketing paroxetine for childhood depression. In came the spin-doctors to bury the inconvenient facts and spread the good news to doctors around the world that paroxetine was safe and effective for children.

Drug companies discovered long ago that doctors are much more likely to be persuaded by their peers than by the hard sell of sales reps. So GSK built a business relationship with several leading academics to write articles and speak at conferences, effectively selling paroxetine for the company. They are known in the trade as "Key Opinion Leaders."

Internal company documents reveal that Professor Martin Keller, Chair of Psychiatry at Brown University in the U.S., was one of the company's leading Key Opinion Leaders. He was named as the author of Study 329. But the documents cast doubt on whether he actually analyzed the data himself and wrote the final paper. In a memo, Professor Keller thanks a ghost writer for the initial preparation of the study manuscript - a ghost writer who works for a medical public relations company hired by GSK:

"You did a superb job with this. Thank you very much. It is excellent. Enclosed are some rather minor changes from me."

It is clear from the internal correspondence that the ghost writer was making some important decisions about how to present the data for publication. At one point a senior GSK executive comments that: "It seems incongruous that we state that paroxetine is safe yet report so many Serious Adverse Events."

GSK suggests to the public relations person that she make clear paroxetine may have caused all 11 of the serious side-effects. But the final article says: "Of the 11 patients, only headache (1 patient) was considered to be related to the treatment."

The paper never states clearly how many children became suicidal, nor does it explore whether the drug was to blame. Instead, it concludes that paroxetine is "generally well-tolerated and effective".

The next step in GSK's marketing plan was to get the study published. Doctors rely on medical journals to give them advice they can trust. The first journal GSK approached spotted flaws in the study and rejected it. So the company sent it to the premier children's mental health magazine, the Journal of the American Academy of Child and Adolescent Psychiatry. Its peer reviewers also spotted the flaws, as is clear from their confidential comments:

"Overall. Results do not clearly indicate efficacy, authors need to clearly note this."

"The relatively high rate of serious adverse effects was not addressed in the discussion."

"Given the high placebo response rate, are [these drugs] an acceptable first-line therapy for depressed teenagers?"

Nevertheless, the journal published the study. Who knows how many child and adolescent psychiatrists around the world were influenced by the article's misleading conclusion?

At the same time, another of GSK's PR people was saying in an internal memo:

"Originally we had planned to do extensive media relations surrounding this study until we actually viewed the results. Essentially the study did not really show paroxetine was effective in treating adolescent depression, which is not something we want to publicize."

Objective Medical Opinion?

When I first started looking at Study 329 for the BBC in 2002, I was confused about its findings. The published article's conclusions seemed at odds with the details of the study. I went to Philadelphia to interview one of the authors and came across another of GSK's Key Opinion Leaders - Dr. Neil Ryan, an eminent child psychiatrist at the University of Pittsburgh. He was speaking at an American Psychiatric Association symposium sponsored by GSK but I had no reason at that time to think he was anything other than independent.

Dr. Ryan told the audience that paroxetine was effective for children with depression. He didn't mention anything about the apparent high rate of psychiatric side-effects of paroxetine in Study 329. But the more I studied the data, the more convinced I became that paroxetine was causing serious problems for some kids who took it. When I returned to Britain, I pursued him by telephone but he refused to talk to me about it.

Fast forward four years, when the internal GSK documents revealed what had actually been going on behind the scenes. Dr. Ryan had been asking the company that sponsored his research how he should deal with my awkward questions, emailing all my correspondence to them. In one email he said:

"I'll call you again later today and you can advise on how might be best to handle this."

GSK denies that it promoted paroxetine for off-label use through Key Opinion Leaders or any other route, laying the responsibility firmly at the door of individual clinicians.

A statement issued by the company reads: "Any decision to prescribe a medicine outside its authorized indications, in the EU or the U.S., is made by a doctor on the basis of his/her clinical judgement and the interests of their patient."

But where do doctors get the information from that gives them the confidence to prescribe off-label? From medical journals and conferences.

It was the documents that in the end exposed the Key Opinion Leaders and the drug company. Without the U.S. legal action we would have been none the wiser. The drug is now effectively banned in the UK for use in under-18s and in the U.S. it carries a "black box" warning about the risk of suicidal behaviour.

This drug was GSK's golden egg for many years and a lucrative business was built around its promotion. So it is perhaps not surprising the company dismissed my investigation as media scaremongering. What is surprising, though, is how many doctors on both sides of the Atlantic who claim to practice evidence-based medicine still refuse to acknowledge the damage the drugs can do - even though the evidence comes from GSK's own RCTs and the company's internal correspondence clearly shows how they tried to cover it up.

Sales Affected by Warning Labels

Dr. Mike Shooter, former President of the Royal College of Psychiatrists and a child psychiatrist who used to prescribe paroxetine, says:

"I personally felt cheated and heaven only knows what the children, adolescents and their parents and their doctors on the other end of that felt. Very much the same. This has huge implications, right through medicine."

Professor Keller and Dr. Ryan seem unabashed by their role in the scandal. A few months ago Professor Keller co-authored another paper, on the impact of media reporting on prescribing. Rather than exploring how those conducting the trials overlooked the damage to children caused by the drug, he and his co-authors expressed concern that prescriptions of antidepressants to children have declined sharply in the U.S. since the addition of the black box warning - as if the increased risk of suicidal behaviour on the drugs wasn't something that should affect prescribing.

The FDA is now considering whether to remove the black box warning on paroxetine and other similar antidepressants because it is scaring doctors. In the UK, the regulatory authorities are investigating whether GSK acted fraudulently in its conduct over Study 329. Theoretically, criminal charges could be brought, but the medicine's regulator is fully-funded by the drugs industry, so don't hold your breath.

Meanwhile, Stephanie Gatchell and her husband have moved away from the home that holds so many awful memories, to start a new life in Ireland. They can't forgive GlaxoSmithKline for concealing what it knew about paroxetine:

"I can't understand how they could possibly be so devious, and all just because of profit. I think the decision-makers in that company should be brought to justice. They have a lot to explain."

*Shelley Jofre is a reporter with BBC TV's flagship current affairs program: Panorama. To learn more about Sharise Gatchell, see

GlaxoSmithKline issued a statement in response to Shelley Jofre's investigation.

AMP Section Name:Pharmaceuticals
  • 122 Pharmaceuticals
* indicates required