Study questions whether drugs such as Prozac are any better than placebos
A widely-reported analysis of clinical trial data for Prozac and related antidepressant drugs has claimed that the medicines work little better than chemically-inactive placebos in all but the most severely depressed patients. The study’s authors, led by Irving Kirsch at the University of Hull, UK, conclude that there is little benefit in prescribing such chemical treatments to depressed patients unless alternative treatments have failed to provide a benefit. Here, Chemistry World answers some of the big questions surrounding the study.
Which drugs were involved in the study?
Four of the six most widely prescribed antidepressants approved by the US Food and Drug Administration (FDA) between 1987 and 1999: Prozac (fluoxetine), Effexor (venlafaxine), Seroxat/Paxil (paroxetine), and Serzone (nefazodone). The drugs are made by some of the world’s largest pharmaceutical companies - US firm Eli Lilly makes Prozac, and the UK’s GSK makes Seroxat.
The two other bestsellers, Zoloft/Lustral (sertraline) and citalopram, were not included in the analysis as the team were unable to obtain all the necessary data.
How do the drugs work?
They are all selective serotonin reuptake inhibitors (SSRIs). Serotonin (also known as 5-hydroxytryptamine) is a neurotransmitter that relays signals from one nerve cell to the next, across a small gap called a synapse, within the brain. SSRIs were developed based on the theory that depression is caused by a lack of neurotransmitter stimulation at the receiving neuron. Blocking the reuptake of serotonin by the body would leave more of the chemical in the brain, boosting its stimulatory effect.
What did the study actually find?
The team carried out a meta-analysis on clinical trial data that had compared patients’ responses to the drug with the effects of a placebo. Meta-analysis is a statistical method that combines the results from all the studies into one grand conclusion. Studying a large data set in this way can reveal overall trends not apparent from individual trials.
Kirsch and colleagues found that there was a statistically significant benefit in the use of SSRIs over placebo - but that the difference was smaller than the standard of ’clinical significance’ set down by the UK’s National Institute for Clinical Excellence (NICE) for all but the most depressed patients. NICE uses the data to form clinical recommendations, which among other inputs takes into account any particular treatment’s value-for-money for the National Health Service.
Interestingly, the team also found that patients’ response to placebo across all the trials was ’exceptionally large’ - an indication of the complexity of the disorder. It was only the fact that the most severely depressed patients showed a much lower response to placebo that made the drug response clinically significant in this group of patients.
So did the study include any new data?
Not exactly. But Kirsch did combine published data with a number of unpublished studies which they obtained from the FDA using the Freedom of Information Act. These clinical trial studies had been sent to the regulator by the manufacturers when the drugs were originally submitted for approval. Kirsch says that by including this data, his meta-analysis avoids the data bias caused by pharmaceutical companies selectively reporting positive results.
While the trial data had all been filed with the FDA, some of the unpublished data had not been made available to NICE when the institute was deciding whether to recommend the drugs for the NHS. NICE says it cannot force companies to submit their unpublished data.
I bet the drug companies are hopping mad?
Absolutely. GSK, who make Seroxat, responded that they strongly believe the scientists’ interpretation is incorrect - and that only a small subset of the total clinical data on antidepressants had been included in their meta-analysis. The company also says that all its clinical trials data on Seroxat are freely available on their website.
Prozac manufacturer Eli Lilly have similarly defended their drug’s performance, stating that the Prozac is one of the world’s most-studied medicines, and that extensive scientific and medical experience has demonstrated it as an effective antidepressant.
The message was echoed by David Nutt, a psychopharmacologist at the University of Bristol. ’There is good evidence of antidepressant efficacy from other types of studies, especially relapse prevention, which show significant benefit over placebo,’ he said.
Are the results a surprise?
Depends if you believe them or not. Previous meta-analyses, including one by Kirsch, have already shown that SSRIs do not provide a large benefit over placebos. Yet there are many other trials that show significant, positive results from these drugs.
NICE already recommends that SSRIs should not be used in cases of mild depression, and that the drugs should only be prescribed for patients with moderate to severe depression as part of a wider treatment regime that includes counselling.
NICE told Chemistry World that a review of the clinical guidelines for the treatment of depression was already underway, which would include the results of Kirsch’s study. The updated guidelines are due to be published in June 2009. Following the publication of Kirsch’s study, NICE clinical and public health director Peter Littlejohns stressed the importance of being given access to unpublished trial data in order to develop effective guidelines.
In the wake of the study, the Royal College of Psychiatrists has urged people taking antidepressants not to discontinue their medication without consulting a doctor.
James Mitchell Crow
Enjoy this story? Spread the word using the ’tools’ menu on the left
I Kirsch et al, PLoS Med, 2008, DOI: 10.1371/journal.pmed.00550045