Problems: SSRI anti depressants for aspergers, adhd, autism

The rise and fall of the wonder-drugs

New Scientist vol 183 issue 2454 - 03 July 2004, page 36

There is a crisis of confidence in many psychiatric drugs. In the first of a two-part series about mind medicines, James Kingsland looks at the Prozac class of antidepressants. Once thought to make you feel 'better than normal', now there are fears that they raise the risk of suicide, and some even question how well they work

(continued)

It appears that manufacturers have been keeping a wealth of research secret. Last month the state of New York filed a lawsuit against GlaxoSmithKline (GSK), alleging that it suppressed research revealing the suicide risk to under-18s from its SSRI, Seroxat (Paxil in the US). GSK denies the charge, saying it publicised the results in journals or at medical conferences.

So what are we to make of the conflicting claims surrounding these drugs? Should SSRIs still be used at all? Can we trust anything the drug companies say? With campaigners calling for new laws to force firms to publish all their research data, the SSRI saga could have far-reaching implications.

Depending on the result of the New York lawsuit, as well as ongoing inquiries, the pharmaceutical industry could be facing its biggest ever shake-up. Richard Brook of the UK mental health charity Mind, says: "There is a clear case for government changing this situation as a matter of urgency."

The story begins more than 30 years ago. The antidepressants available then - which are still used today - are reasonably effective. Trials show they help, on average, about two-thirds of patients to recover, compared with one-third of those given placebos. But patients often stop taking these drugs because of side effects. A major group called the tricyclics, for example, can cause dry mouth, constipation and irregular heart beat.

These drugs indiscriminately affect a range of chemicals that relay the signals between nerve cells, namely serotonin, dopamine, noradrenaline and acetylcholine. Researchers began to suspect the drugs' antidepressant effects stemmed from their raising levels of serotonin. Suppose a drug could be found that selectively targeted this chemical? It should have the therapeutic effects without the side effects.

US firm Eli Lilly found just such a drug, which works by blocking the reabsorption of serotonin from nerve synapses in the brain - hence "selective serotonin reuptake inhibitor" (SSRI). It was launched in 1988 as Prozac. Trials seemed to show that it was about as effective as the tricyclics but without the side effects. It was also found to be far less deadly if patients took an overdose - an important advantage for any antidepressant.

Soon everyone was talking about it. The word was that Prozac was worth taking even if you were just down in the dumps. More than an antidepressant, it was becoming a lifestyle drug, and was nicknamed "bottled sunshine". In his seminal book, Listening to Prozac, US psychiatrist Peter Kramer described how the medicine could turn pessimists into optimists and conquer shyness.

With successive launches of other SSRIs through the 1990s, firms began marketing them not just for depression but also for conditions that doctors would previously have barely recognised as illnesses, let alone treated with drugs. The broadest licence is held by Seroxat, made by GSK. As well as depression and its common attendant, anxiety, Seroxat can be taken for obsessive-compulsive disorder, post-traumatic stress disorder, panic attacks and "social phobia". Last year, GSK's worldwide sales of Seroxat were £2.7 billion.

But right from the start there were hints that all was not well. In 1990, US doctors reported in a medical journal that six patients had become suicidal when they started taking Prozac. Similar reports followed, the common factor seeming to be that patients became highly agitated soon after starting treatment. But the drug firms were dismissive. Was it so surprising that people with depression became anxious or suicidal? The cause was the disease itself, they said, not the medication.

These worries did nothing to slow the steady expansion of the SSRI market during the 1990s (see Graph). Whenever challenged, the companies argued that the evidence was "anecdotal" - collections of cases rather than systematic trials.

This was a major coup because drug firms routinely keep secret the reams of research data they hold on their products. Firms applying for a licence for one of their drugs are legally obliged to disclose the results of relevant trials to regulators, but there's no guarantee that this always happens. And firms commonly request the data is kept confidential for commercial reasons, so there is no way for outsiders to scrutinise the regulators' decisions.

Healy discovered that SmithKline's unpublished results suggested that Seroxat causes agitation in about 1 in 4 people. In one trial, a volunteer committed suicide. There is no way of knowing if Seroxat had been responsible in that case, however.

The balance of evidence convinced the jury that the drug had been a major factor influencing Donald Schell's actions. His family were awarded $6.4 million in damages.

While these revelations concerned the use of SSRIs by adults, it was children and adolescents that next became the focus of controversy. In the UK, no SSRI has ever been licensed for treating depression in under-18s, and in the US only Prozac has this licence. So in this age group most SSRIs were being used "off-label" - in other words, outside the terms of their licence (a common occurrence, in fact, for many medicines given to under-18s). Because firms had not applied for a licence, they had never had to prove their drugs were safe and effective in this age-group.

In April last year, however, the UK Medicines and Healthcare products Regulatory Agency (MHRA), set up an expert panel to review the safety and efficacy of SSRIs, starting with their use in under-18s. It asked firms to send in all the product data they had. By December, it had banned all the SSRIs except Prozac for this age group, citing poor evidence of efficacy and a raised risk of self-harm and suicidal thoughts.

The problems for SSRI manufacturers are unlikely to stop here; research into their use in adults is now being scrutinised. For a start, most early trials were not designed to show if the drugs increased suicidal tendencies. Such effects were often disguised under the general heading of "emotional lability", or instability. "It was dressed up in other terms," says Tim Kendall of the Royal College of Psychiatrists' Research Unit in London.

The suicide risk could have been downplayed in another way, Healy says. In the trials it was common practice to label as "placebo suicides" ones that occurred in the so-called "washout" period - when patients stopped taking any medicines in the week before being assigned to either placebo or drug. This would artificially raise the placebo suicide rate, so even if the SSRI increased the number of suicides, it would look no worse than the placebo rate. Healy re-analysed the raw data from trials involving around 17,000 patients in total taking a placebo or an SSRI, discounting washout suicides. SSRIs seemed linked with over twice the risk of suicidal acts as placebo (Journal of Psychiatry & Neuroscience, vol 28, p 331).

Not everyone shares Healy's views of course. His article appeared alongside one by Yvon Lapierre, psychiatrist-in-chief at the Royal Ottawa Hospital, disagreeing with Healy's arguments (p 340). "The case for the SSRIs adding significantly to the risk of suicide is not convincing," Lapierre told New Scientist. He believes that unlike in depression, an increased risk of suicide has not been seen in other conditions for which the SSRIs are licensed - suggesting that in depression, it could be the patient's original illness causing the problem.

But it is not just the drugs' safety that is now being questioned - their effectiveness is too. One criticism of the early Prozac trials in adults is that patients who showed symptoms of agitation were given tranquillisers. Apart from masking potential side effects, any improvement in the patients' condition could have been down to these, rather than Prozac.

The UK's influential National Institute for Clinical Excellence (NICE) is now drawing up guidelines for treating depression in adults, having analysed about 1000 published trials. The draft version concludes that in mild depression, SSRIs perform on average no better than a placebo.

So where do these revelations leave people suffering from depression? Most doctors believe SSRIs still have a role to play, because some patients do seem to be helped by them. Lapierre, for example, says: "The SSRIs have an important role and I do not feel [it] should be diminished by the recent flow of ink on their alleged causative role in the risk of suicide." Even Healy admits that SSRIs have their place. "I have had lots of patients do very well on these drugs," he says. Unfortunately there is not yet any way to predict which patients the SSRIs will help, and which the drugs will harm.

But in future doctors may well become more cautious about dishing out SSRIs. If the final NICE guidelines, due out in October, reflect the current draft, UK doctors may start recommending that patients with mild depression try non-drug treatments first, such as psychotherapy or exercise programmes.

Meanwhile, the MHRA expert panel on SSRIs is due to report on their use in adults in the next few months. The equivalent body in the US, the Food and Drug Administration, has already made manufacturers include stronger warnings about suicidal behaviour on labels and is carrying out a major review of the drugs in children, also due to report shortly. The Australian regulatory body has advised doctors to monitor children taking SSRIs, to watch for suicidal tendencies.

But whatever the outcome for SSRIs, some campaigners argue that this whole sorry tale is a sign that there must be fundamental changes in the way pharmaceutical firms operate.

A popular move would be to force firms to publish the results of all trials they carry out. At the moment, there is nothing to stop them carrying out numerous trials in "off-label" conditions, publishing the ones that give positive results and burying negative ones. In April, a paper appeared in The Lancet that laid bare just how this can distort the clinical picture of a drug's risk-benefit ratio (vol 363, p 1341). It was a systematic comparison of the published trials of SSRIs in under-18s with unpublished ones that had come to light from the MHRA's inquiry. While both sets of trials supported Prozac, it was a different story for Seroxat and another SSRI called sertraline (made by Pfizer). The published trials suggest a small benefit from using these drugs, but the unpublished trials suggest the increased risk of suicide and self-harm outweigh the benefits.

"It's outrageous that patients should not have access to information about the effects of the drugs they're expected to take," says Iain Chalmers, a member of the R&D advisory committee of NICE. The companies' defence has long been that journals are not interested in publishing negative or equivocal findings. Richard Ley, a spokesman for the Association of the British Pharmaceutical Industry told New Scientist: "Industry doesn't have control as to what gets printed. If you have a trial that doesn't find anything exciting, who's going to print that?"

"There is no evidence to justify that defence," Chalmers says. He points out that open-access, electronic publishers like BioMed Central are committed to publishing all well-conducted research. "The companies don't even have to go through journals - they could publish on their own websites." There are now calls for a central, worldwide register of all ongoing trials with predicted publication dates, which would prevent firms from burying results they don't like.

Healy is concerned that even this would not go far enough. He wants not just the final, processed results, but all raw data from trials to be made public - including confidential data held by regulators. But surely we can trust official government regulators to correctly interpret the data? Perhaps not - they didn't pick up on the washout suicides, points out Healy, despite having the data since the 1990s.

The regulatory bodies face other criticisms. In March, Richard Brook from Mind resigned from the MHRA's expert panel on SSRIs, complaining that the agency had sat on a potential safety problem for more than a decade. The MHRA had just warned doctors that raising the dose of Seroxat above the recommended level had no clinical benefit, and led to more side effects. Brook was outraged that it had taken them so long to act, and that their warning letter to doctors failed to acknowledge this oversight. US campaigners say the FDA, too, has known of the suicide risk from SSRIs for years without doing anything.

Forcing manufacturers to make all their raw data available might also allay concerns that even published papers can be misleading. In May, an analysis of published studies of SSRIs in children revealed a catalogue of shortcomings in trial design and the presentation of results (British Medical Journal, vol 328, p 879). The effect was to downplay problems and exaggerate benefits, the authors concluded.

Strict new laws may be unpopular with drug firms, of course. Pfizer, for example, told New Scientist: "We do not need legislation to do this, nor is it a realistic suggestion. We already share all appropriate data on safety, quality and efficacy with expert regulators who are best placed to evaluate it." Eli Lilly did not comment. GSK, however, has pledged to put results of all its trials online, although these will only be summaries, not the raw data. "We feel that summaries would be more useful than complete studies," a spokesman says.

But whether they like it or not, the scale of the unfolding crisis for SSRI manufacturers may make at least some legal changes inevitable, Healy believes. The upcoming New York lawsuit against GSK may take years to come to court, but it is being seen as a high noon for the drugs industry (New Scientist, 12 June, p 4). Suits against other firms could follow. For the industry, says Healy, the lawsuit is "just the kind of thing that could make this a very big crisis indeed".

DHD LD