Regular readers of the Fiddaman blog may have been keeping a close eye on an old study from the early 90's (08 Apr 1993 – 25 Oct 1994) that I've recently found via GSK's clinical trial registry website.
The study, known as GSK Study 356, reported 7 suicide attempts (2 of which were completed suicides). However, these attempts at suicide are not mentioned within the documents GSK has made public. Moreover, the actual number of suicide attempts in Study 356 was 9. (2 of which were completed suicides)
How do we know this?
Documents obtained from the Australian drug regulator, the TGA, show us this but strangely do not tell us in which group the attempts at suicide occurred...until now.
For those who are unfamiliar with the study, it was, to my knowledge, an attempt at finding which of the two antidepressants, namely Prozac and Paxil, were better at treating major depression. The study was sponsored by GSK.
After posting two blogs on the subject, GSK Study ID - 29060/356 - The Missing Suicide Attempts and No Action to be Taken Against GSK for Hiding Suicide Data, the TGA have written to me in efforts to clarify what was revealed by me last month.
However, rather than trying to add clarity to my previous two blogs, the TGA has opened a pretty big can of worms.
In asking them why GSK failed to report the missing suicide attempts they told me the following. It makes for very interesting reading.
Good Afternoon Mr Fiddaman
I have followed up your enquiry about the presentation of information about the Clinical Trial - GSK Study ID - 29060/356 on the GSK study website.
The nine events you have referred to were discussed with investigators at a meeting on November 26th 1993, whilst the study was still ongoing and blind. Seven of those same events were also included in the Dear Investigator Letter which went out to investigators in August 1993. Of the nine events that were discussed with the investigators five events were subsequently coded to the WHO preferred term of drug abuse, through the methodology applied at that time. On the trial summary these are presented as cases of “Drug Abuse (Overdose)”. There were two of these recorded for paroxetine (patients 139 and 147) and three of them recorded for fluoxetine (patients 18, 76 and 144)
The remaining four cases were coded to the WHO preferred term of “Suicide Attempt”. On the trial summary two of these events are listed as “Suicides” (in the fatal SAE section) with one for paroxetine (patient 92) and one for fluoxetine (patient 122). The two other cases are presented as “Suicide attempt”, both for fluoxetine (patients 87 and 142).
The information as presented in the clinical trial summary accords with the safety information as provided in the full study report.
I trust that the information provided is of assistance.
Regards
Bernadette Barton
Assistant Director
Adverse Event and Medicine Defect
Pharmacovigilance and Special Access Branch
Therapeutic Goods Administration
Department of Health
PO Box 100
Woden ACT 2606 Australia
--
So, there you have it, folks. The missing suicide attempts were coded as "drug abuse." They did this, according to the TGA, because this was the preferred term, at the time, used by the World Health Organisation (WHO)
So, breaking down the 9 who attempted suicide we have 6 from the Prozac group and 3 from the Paxil group. (1 patient from each of the groups completed suicide)
It's a head-scratcher for me - why would WHO use such a system?
The medical definition of "suicide attempt" and "drug abuse" differ somewhat.
Suicide Attempt
A non-fatal, self-directed, potentially injurious behavior with an intent to die as a result of the behavior; might not result in injury
Prescription Drug Abuse
Taking medication in a manner or dose other than prescribed
This begs the question: How many other GSK sponsored studies buried suicide attempts in the drug abuse category?
Study 356 was carried out in the early 90's. By the late 90's there was growing concerns regarding Prozac, the concerns surrounding Paxil came later.
Here's a Guardian article from 1999. I'll just post the sub-heading and the link:
It was too good to be true. Prozac, the wonderdrug hailed as the answer to the war against depression and taken by some 37 million people worldwide, is not as harmless as we've been led to believe. Disturbing evidence has now emerged, showing that, after the initial relief and euphoria of the first dose, Prozac can push some patients into so agitated a state of mind that they are a danger not only to themselves, but to others, too. (Full Article)
One has to remember here that Study 356 was sponsored by GSK, the manufacturers of Paxil who, at the time, wanted Paxil to be the blockbuster drug that Prozac already was.
Would doctors have prescribed Prozac knowing that there was a high ratio of suicides in a clinical trial - personally, I think they would have. The marketing for these two drugs was heavy and included incentives for doctors to prescribe them. Eli Lilly reps dining and dashing', whilst GSK reps would persuade doctors at strip bars, amongst other places.
Now we know. It's taken the best part of 12 months to get to the bottom of this study, a study that is almost 25 years old!, and I couldn't have done it without the help of Kathy, who is the moderator of the Facebook page, Australian Antidepressants Class Action & Awareness and an administrator for the Australians For Safe Medicines Facebook page.
Special thanks to the TGA too. They seem a little bit more transparent than their counterparts, the FDA and MHRA.
The British, American, and, indeed, Australian media are disinterested.
Bob Fiddaman
No comments: