Coleg Meddygaeth Prifysgol Cymru

University of Wales College of Medicine

hergest.bmp (2190 bytes)
Adran Meddygaeth Seicolegol / Department of Psychological Medicine
Adran Cymru y Gogledd / North Wales Department
Pennaeth Adran / Head of Department
Yr Athro /Professor Michael J. Owen
Dr. D. Healy (Cyfarwyddwr)


Dr June Raine, Director of
Post-Licensing Division, MCA
Market Towers, 1 Nine Elms Lane
LONDON SW8 5NQ 19th June 2001

Dear Dr Raine,   


Many thanks for your letter of the 8th June. There are a number of ambiguities as regards the listing of agitation in the Summaries of Product Characteristics of all SSRIs. First, neither the BNF nor the ABPI versions of these list agitation as an adverse effect for Seroxat for instance. I have little doubt that there is such a listing in some Summaries of Product Characteristics but perhaps you can tell me whether there is any evidence as to the likelihood of GPs for example reaching for the individualised SPCs of each of the SSRIs rather than for their BNF or for a copy of the ABPI Compendium.

Second do you know whether this listing is likely to be interpreted either as a warning or a statement of causality rather than something of a dismissal of a link.

This question seems to me to be critical. If your interpretation is that the information being presented to physicians is that agitation can be caused by SSRIs (and the data fully supports this interpretation), then it seems to me almost logically impossible for you not to advise physicians that thoughts of suicide or harm that emerge early in the course of treatment with SSRIs must. in some cases be linked directly to the agitating drug with which they are being treated.

Given that you don't offer this advice at present, I can only assume you consider that the reference to agitation on SSRIs refers to an association of uncertain causal significance - despite the appearance of agitation and suicidality in healthy volunteers. Not knowing of suicidality or agitation in healthy volunteers, I would expect most doctors to interpret your reference to agitation in these terms.

I am pleased to see that you will in July be reviewing some aspects of the Healthy Volunteer work that has been done. In the light of previous correspondence with Dr Jones, it seems that hitherto you've had summaries of this data prepared for you by the companies rather than the raw data. Can you tell me, however, who will be doing the review and what will they review? Can you also tell me what criteria you have, if any, as regards links to interested parties, for an outside expert called in to review such material? More generally, I would appreciate knowing if you have formal criteria governing the links an expert/advisor being consulted by the MCA may have with a company whose product is under review for licensing in the first instance or under review because of some problem subsequently.

An independent review of these studies will I believe show you that agitation occurs in up to 25% of the volunteers exposed to paroxetine - and other SSRIs - in multiple-dose studies. It is very difficult in the light of this to see how you cannot avoid a warning about the occurrence of drug-induced agitation and suicidality in patients being treated with Seroxat or other SSRIs.

In SmithKline's healthy volunteer panel you will also find that there was a suicide. It happened some weeks after treatment had been discontinued. Quite reasonably at that historical point in time the investigators did not link the suicide to paroxetine. However it may be time to re-open the question of what happened to this volunteer.

The first point is that none of the investigators were trained to elicit problematic behavioural or mental effects emerging on these drugs. The volunteer in question will not have been systematically or properly assessed at the time of paroxetine intake. There has to be a considerable chance that the volunteer suffered in much the same way that the healthy volunteers in the study we ran in this department suffered. Without the index of suspicion we had as to the possibility of a problem we would not have unearthed the findings that I subsequently wrote up and sent to you.

The article that was published in Primary Care Psychiatry referred to the fact that our volunteers took several weeks to recover their mental equilibrium following what had happened to them. While the active suicidality that emerged with treatment disappeared once the treatment had stopped, both volunteers were left with severely impaired self-confidence, dysphoria and nervousness for weeks afterwards.

Had either committed suicide during this period I would not have felt able to reassure you that the Sertraline that we had given them which had made them suicidal for the first time in their lives had not contributed to the subsequent suicide. In fact there would have to be a strong index of suspicion that it had contributed, albeit in ways that have not yet been properly mapped out. I think you need to consider SmithKline's death in this light.

Quite apart from the impact on our volunteers' self confidence of being made suicidal by a drug, there is a further issue in the case of Seroxat. SmithKline's own studies show that healthy volunteers had withdrawal problems including abnormal dreams and thoughts and agitation. What none of us know at this stage is how long such withdrawal is likely to go on. You may wish to note and perhaps even consult with Professor Merton Sandler formerly of the Pathology Department in Queen Charlotte's Hospital, London who took a single dose of reserpine in the early 1960s as part of his research and was significantly dysphoric for a month afterwards.

I have certainly clinically seen patients with mild conditions treated for short periods of time suffer for three months after discontinuing Seroxat. SmithKline's healthy volunteer work amply bears out these possibilities. It would be very difficult for either them or you or me to rule out therefore a contribution from this source to the death of their healthy volunteer.

This issue raises the role of SmithKline Beecham's awareness of withdrawal in their healthy volunteers when they then made an application to the MCA for a licence to prevent relapse in depression claiming that patients re-randomised to placebo who became unwell were suffering from a relapse of their depression. This seems to me to be close to frankly deceitful in the light of their recognition of drug dependence from a period 10 years earlier.

You must know that there are thousands of patients around the country who find, when they attempt to reduce their medication, that they feel very unwell. They are told by general practitioners and pharmaceutical companies that this is a return of the depressive or nervous condition for which they were being treated. It is no such thing. Any minimal awareness of the nature of physical dependence syndromes, the point at which they emerge following dose reductions and their liability to clear up following the reinstitution of treatment, as opposed to the general failure of new depressive episodes to clear up following the reinstitution of treatment, makes it almost certain that for the vast majority of people affected in this country what is involved is a withdrawal syndrome rather than a new depressive episode.

As things stand at present you are compounding the injury of physical dependence with a message that given inappropriately will in its own right have long-term consequences.

You asked for the article involving the further data from our healthy volunteer studies. I enclose a version that has been submitted for peer review.

As regards data from the recent trial in the United States I'm sure you could get the trial transcript in its entirety from SmithKline Beecham overnight if you requested it. What you should also request are details of the Montgomery study in intermittent depressive disorders conducted around 1992/1993. This was a placebo controlled study of paroxetine that yielded a projected annual rate of 45 suicide attempts in the paroxetine treated group versus 12 suicide attempts in the placebo group. Mr Charles Preuss, the lawyer for SmithKline, suggested these results were not statistically significant. I suspect however that the jury found that the early termination of the study and non-publication of the results significant.

You may also care to know that the most serious suicide attempt occurred in the paroxetine treated group and involved what SmithKline refer to as spinal injuries. This case led to an action against St Mary's hospital.

Finally in order to avoid a repeat of the embarrassment of describing a set of studies the MCA sent me as epidemiological studies, it may be time for you to consider some of the following possibilities. How much of the literature being presented to you by companies supporting their position is ghost written, published in non-peer reviewed supplements to journals, or published in journals of which one of the authors is an editor? I think a survey of the material that companies have brought to legal trials in this area, for instance the studies cited by SmithKline in the Tobin trial, will show that a good proportion of that material fails into one or other of the above categories. How much independent science can you identify?

You may not feel that it is your area of responsibility to police issues like this. My problem is that I'm fairly certain that every other arm of Government that I could turn to will respond similarly. This is not a reassuring situation. In the current parlance, it doesn't smack of joined up government.                 

Yours sincerely

David Healy MD FRCPsych
Director, North Wales Department of Psychological Medicine
Uned Hergest, Ysbyty Gwynedd, Bangor, Gwynedd LL27 2PW
Ffn: (01248) 384452 Ffacs: (01248) 371397
Hergest Unit, Gwynedd Hospital, Bangor, Gwynedd LL27 2PW
Tel: (01248) 384452 Fax: (01248) 311397

Or HERE for full index of this correspondence

Contents page
What's New?