Very little is known publicly about the Yugoslavia paroxetine (Seroxat UK, Paxil US) trial. What is public, makes very interesting reading. Interesting because it covers the subject of withdrawal, something which GSK have denied for years is a problem with their antidepressant.
I'm going to try and break this down so it's easily digestible. The information was taken from a Paxil lawsuit that was filed in California on August 19, 2000. (Nguyen & Farber, plaintiffs vs. SmithKline Beecham Corporation) GSK resolved the suit in January 2002. The results of the resolution, including any settlement by GlaxoSmithKline, were never announced.
The legal brief for the lawsuit is a very interesting document, in as much that it mentions clinical trials carried out in Yugoslavia (as the country was known then)
During this time GlaxoSmithKline were known as SmithKline Beecham (SKB)
According the the legal brief, here's what happened...
In 1988, a clinical trial in Yugoslavia commenced. The trial was sponsored by SKB and could have (should have) been an ideal opportunity for SKB to address the withdrawal issue. The trial was known as the "Yugoslavia trial" or "relapse trial."
At this time, SKB were seeking approval of Paxil and the Yugoslavia trial was to show the FDA (the US drug regulator) how effective Paxil was in treating depression - they would also try to show the FDA how it was important to keep taking Paxil and not to stop... because if you did stop then you would go into relapse, in other words, SKB were trying to prove that stopping Paxil meant the patient's original illness would return.
Here's how they achieved this...
First off, participants in the clinical trial had to have suffered a major depressive disorder and experienced a history of recurring depression.
In "Phase I" of the trial ALL patients were given Paxil. "Phase II," of the trial saw those same patients split into two groups. One group (Group A) remained on Paxil while the second group (Group B) shifted to placebo, a sugar coated pill.
Those in Group B would have, in essence, been going through a cold turkey withdrawal of Paxil - something that GSK have clearly stated that no patient should do when coming off Paxil. For the purposes of this particular clinical trial, however, it appears that they wanted patients to suffer withdrawal symptoms so they could present the FDA with evidence that Group B were suffering a relapse of their original illness - ergo, the illness had returned once they had stopped taking Paxil. In actual fact (my opinion and the opinion of many others) SKB were suggesting that those experiencing withdrawal issues were actually experiencing a return of the original depressive symptoms. Win-Win for SKB.
It was SKB's intention to show the FDA that Paxil patients do not "relapse." So, it appears they masked the results. From a business point of view I can understand this. Understand but not agree with.
Shortly after Phase II of the trial commenced it is alleged that the placebo group (Group B) began experiencing massive Paxil withdrawal symptoms. This occurred in 1989 and 1990. In the Yugoslavia trial, when the placebo patients walked into the doctor's office suffering withdrawal symptoms, the SKB agents saw relapse criteria, not symptoms of withdrawal.
Here's the clever bit...
When clinical trials are set up they must have a protocol, an official procedure or system of rules. The protocol in the Yugoslavia trial was, it has to be said, genius.
Doctors were not looking for a withdrawal problem, they were, according to the protocol, looking for a relapse.
This from the legal brief...
When a patient goes to the doctor's office, scientifically validated medical procedures should be followed by the physician in order to attain an accurate diagnosis. In the case of ongoing clinical trials, however, as with the "Hamilton" suicidality indices, the physicians as well as the patients are on a pre-planned glide path in accordance with the trial protocol. Given this was a relapse study, medical ethics require the responsible clinician to abandon the trial for patient emergencies and safety, and to always follow sound medical procedures. Nevertheless, the high stakes nature of the clinical trials likely prevailed in the PI's mind. This does not necessarily suggest that the PI is intentionally cheating. What is does suggest are that test objectives taint the prism through which the PI observes patient symptoms. In the Yugoslavia trial, when the placebo patients walked into the doctor's office suffering withdrawal symptoms, the SKB agents saw relapse criteria, not symptoms of withdrawal. Alternatively, it is alleged on information and belief that many physicians in Phase II of the Yugoslavia study indeed detected and reported Paxil withdrawal symptoms for what they were. They did this in written patient summaries submitted to SKB headquarters. However, because of SKB's study design and desire to avoid the "withdrawal" issue at all costs, SKB chose to interpret the withdrawal data as "relapse" data and reported it as such to the FDA.
Here's how SKB spun the data.
- At the six month mark in Phase II, SKB reported placebo patients suffered 450% more relapse incidents than did the Paxil patients.
- Eighteen placebo patients suffered "relapse" while only four Paxil patients did. This published result was to show that patients staying on Paxil continued to enjoy a normal, "depression free" life, but that those abandoning the drug would suffer relapse back into a depressive state.