MHRA-LOGO.gif (6825 bytes) Medicines and Healthcare products
Regulatory Agency

Market Towers
1 Nine Elms Lane, London SW8 5NQ


Mr Medawar
Social Audit Ltd
P.O. Box 111
London NW1 8XG

27th March 2007

Telephone 020 7084 2000  Fax 020 7084 2353
E-mail [email protected]


Dear Mr Medawar,    

Thank you for your letter dated 12th March 2007. The majority of the enquiry points raised in your letter do not fall within the remit of FOI, as you are not seeking specific items of information. Where you have asked for specific items of information they have been answered under FOI (see attached).

I will address the other matters raised in your letter herewith.

The interview that Sir Alasdair Breckenridge did with Panorama back in September 2004 was a two and a half hour pre-recorded event. The parts you have selectively quoted from have already been edited from that large amount of original footage and his performance/the content of the final programme should be viewed within that context. As you point out, Sir Alasdair was disappointed with the final outcome but editorial rights is not within the gift of the interviewee in such situations. It was felt inappropriate at the time to make ‘an issue’ about this with the producers, after the event as it would have been of no benefit to the Agency in arrears.

Sir Alasdair stands by the comments he made in the full pre recorded interview. The information he imparted in that interview was that which was available to him at the time (September 2004). The SSRI Expert Working Group that you refer to did not conclude until the 6th December 2004.

I would like to reassure you that our Chairman and other spokespeople for the MHRA always ensure that they give robust and fact based judgements to ensure that benefits to patients and the public justify the risks.

Yours sincerely

Sara Coakley
Media Relations Manager



FOI ANSWERS TO QUESTIONS Mr Medawar letter dated 12 March 2007


4. (i) We cannot confirm this statement as we do not hold the relevant data

    (ii) Yes

                    6. No to both questions

7. The Chairman undertook no specific training in preparation for that Panorama interview. A verbal briefing was provided in advance of the interview from colleagues within the Agency. No written briefing exists.

8. No

9. No

10. Please see below minutes from EWG regarding lessons learnt.

(i) No

(ii) No.





Extract from above report


This review of SSRIs has considered a wide range of data from meta-analyses, randomised controlled clinical trials (RCTs), epidemiological studies and from spontaneous suspected ADR reports, including reports from patients. This chapter summarises the key findings of the review and considers what further research would be desirable to provide better information to patients and prescribers.


11.1 Main findings

The main findings of the review are as follows.


Use of SSRIs in adults


Suicidal behaviour – adults

• There is epidemiological evidence that the risk of self-harm in depressed patients is greatest around the time of presentation to medical services. It is general clinical experience that the risk of suicide may increase in the early stages of treatment for depressive illness.

• Careful and frequent patient monitoring by healthcare professionals and, where appropriate, other carers, is important in the early stages of treatment, particularly if a patient experiences worsening of symptoms or new symptoms after starting treatment.

• Studies indicate that increases in the prescribing of SSRIs have not been associated with an increase in population suicide rates, although interpretation of these findings is difficult as a range of factors influence population trends in suicide.

• From the available clinical trial data, both published and unpublished, a modest increase in the risk of suicidal thoughts and self-harm for SSRIs compared with placebo cannot be ruled out. There is insufficient evidence from clinical trial data to conclude that there is any marked difference between members of the class of SSRIs, or between SSRIs and other antidepressants with respect to their influence on suicidal behaviour.

• Evidence from non-experimental studies based on the General Practice Research Database indicates that there is no increased risk of suicidal behaviour with SSRIs compared with TCAs.

• There is no clear evidence that there is an increased risk of self-harm or suicidal thoughts when SSRIs are discontinued.

• Evidence of a relationship between suicidal behaviour and increasing/decreasing dose is not robust; however, patients should be monitored around the time of dose changes for any new symptoms or worsening of disease.


Withdrawal reactions

• All SSRIs may be associated with withdrawal reactions on stopping or reducing treatment. Paroxetine and venlafaxine seem to be associated with a greater frequency of withdrawal reactions than other SSRIs. A proportion of SSRI withdrawal reactions are severe and disabling to the individual.

• The most commonly experienced withdrawal reactions are dizziness, numbness and tingling, gastrointestinal disturbances (particularly nausea and vomiting), headache, sweating, anxiety and sleep disturbances.

• To minimise withdrawal reactions on stopping SSRIs, the dose should be tapered gradually over a period of several weeks, according to the patient’s need.

• There is no clear evidence that the SSRIs and related antidepressants have a significant dependence liability or show development of a dependence syndrome according to internationally accepted criteria [either DSM-IV or ICD-10].


Dose response

• For the majority of SSRIs in the treatment of depressive illness, clinical trial data do not show an additional benefit from increasing the dose of an SSRI above the recommended daily dose.

• In the absence of evidence of a benefit from increasing the dose, good practice would be to maintain patients on the lowest efficacious dose.

• If a patient is not doing well after starting treatment, the possibility of an adverse reaction to the drug should be considered. Patients should be monitored for signs of restlessness or agitation, particularly at the beginning of treatment. Increasing the dose in these circumstances may be detrimental.


Use of SSRIs in children and adolescents

• The balance of risks and benefits for the treatment of depressive illness in under-18s is judged to be unfavourable for paroxetine (Seroxat), venlafaxine (Efexor), sertraline (Lustral), citalopram (Cipramil), escitalopram (Cipralex) and mirtazapine (Zispin). It is not possible to assess the balance of risks and benefits for fluvoxamine (Faverin) due to the absence of paediatric clinical trial data. Only fluoxetine (Prozac) has been shown in clinical trials to be effective in treating depressive illness in children and adolescents, although it is possible that, in common with the other SSRIs, it is associated with a small increased risk of

self-harm and suicidal thoughts. Overall, the balance of risks and benefits for fluoxetine in the treatment of depressive illness in under-18s is judged to be favourable.

• The safety profiles of the different products in clinical trials in children and adolescents differ across studies. However, an increased rate of a number of events, including insomnia, agitation, weight loss, headache, tremor, loss of appetite, self harm and suicidal thoughts, were seen in those treated with some of the SSRIs compared with placebo.


Young adults

• There is no clear evidence of an increased risk of self-harm and suicidal thoughts in young adults of 18 years or over. However, given that individuals mature at different rates and that young adults are at a higher background risk of suicidal behaviour than older adults, as a precautionary measure young adults treated with SSRIs should be closely monitored.


11.2 Research requirements

The review has identified several areas for further research on the use of antidepressants.

• The effectiveness of SSRIs in mild depression has not been clearly demonstrated in RCTs. Without clinical trials in mild depression it is not possible to assess the balance between benefit and harm for these drugs in this indication.

• Many TCAs have not been subjected to the same level of assessment in clinical trials as have modern antidepressants (eg SSRIs). A systematic review of published and unpublished placebo-controlled RCTs of TCAs is required to assess whether there is any evidence of an increased risk of suicidal thoughts, self-harm and suicide in association with these drugs.

• Primary care research is required to investigate the patterns of use for the most commonly prescribed antidepressants, and factors governing GPs reasons for prescribing them.

• Questionnaire-based measures should be developed and evaluated for use in primary care to help GPs identify those patients who would most benefit from antidepressant treatment.


Psychiatric adverse effects of SSRIs: pharmacological considerations

• Further research is necessary into the effect of specific enzymes such as CYP2D6 on the rate of metabolism of SSRIs, and the impact this has on adverse effects and withdrawal reactions. Likewise, there is emerging evidence that the risk of

adverse drug reactions may be related to specific genotypes. There should be further research into the pharmacogenetic determinants of efficacy and toxicity associated with SSRIs.

• Research is needed into the effect of SSRIs on other chemical receptors and whether this has a bearing on SSRI efficacy or toxicity.

• Clinical trials should be conducted to study pharmacological and other approaches to offset the worsening of symptoms described by some patients around the time of commencing treatment.


Safety and efficacy in children and adolescents

• There should be further research on the epidemiology of depressive illness in children and young people.

• Research is required to better understand why antidepressants generally appear to be ineffective in children and adolescents.

• There should be clinical trials of antidepressants or non-drug interventions in children and adolescents, which include regular monitoring for the occurrence of self-harm, suicidal thoughts and other adverse effects.

• Large randomised controlled trials should be conducted comparing psychological interventions in depressed children and young people to placebo medication and antidepressants.

• There should be studies to estimate the rate and nature of withdrawal effects in children and adolescents.

• Cochrane systematic reviews on drug and non-drug treatments of depressive illness in children and young people should be conducted, incorporating published and unpublished data.


Withdrawal reactions and potential for dependence

• There should be trials in adults, adolescents and children to determine the best way to withdraw from antidepressants; such research might include an assessment of the substitution of short-acting SSRIs with long-acting SSRIs during withdrawal.

• Further research is needed to determine the relationship of withdrawal reactions with dose and duration of treatment.

• Studies are required to quantify the abuse liability of SSRIs and related compounds.


Suicidal behaviour

• There is anecdotal evidence of an increased risk of self-harm early in antidepressant therapy; further investigation and quantification of this phenomenon will help inform the monitoring requirements for patients initiating antidepressant therapy.

11.3 Recommendations for the conduct of future clinical trials of antidepressants

• To enable comparison between substances, consideration should be given to developing standard clinical trial protocols for assessment of dose, withdrawal reactions, and duration of therapy for antidepressants.

• Safety and efficacy should be assessed separately for age groups 18-30 years and >30 years.

• Study results should be presented using ‘time to event’ analyses to assess whether there is an excess risk early in treatment.

• All clinical trials should be carefully monitored for an excess risk of self-harm.

• Suicidal behaviour is rare in clinical trials; therefore many trials are too small to reliably detect any difference in the risk of self-harm and suicide between treatment groups. Standardisation of clinical trial protocols should enable results from future studies to be combined, and should enable the detection of any increased risk within particular sub-groups or at any specific time-points.

• Information on patients leaving a study for ‘lack of efficacy’ should be closely examined for suicidal events prior to unblinding.

• Information on patients experiencing suicidal events during the ‘run-in’ period should be clearly identified.

• Clinical trials of the same substance in a range of indications should be designed in such a way that any safety concerns relating to specific indications should not be confounded by the study design.

• Trials should be of long enough duration to reflect likely use in practice.

• Whilst placebo-controlled trials provide important information on efficacy, studies against an active comparator should be a requirement for all new antidepressants. The clinical trial protocol must have an appropriate and relevant comparator for the indication prescribed at an appropriate dose; it would be

preferable if clinical experts could design a set of standard protocols which could be used in clinical trial design.

• These studies should be designed to provide clear evidence of any efficacy of increased doses. The most appropriate trial is the randomised non-responder trial where patients are all treated with the same starting dose. Those that fail to respond are then randomly allocated to either have their dose increased or to remain on the same dose. The trial is blinded, meaning that patients and investigators do not know which group they were assigned to and so their assumptions cannot affect the responses. At the end of the trial the dose increase group can be compared with the group that did not have their dose increased, and the benefit of the dose increase can be assessed. The studies should be designed to be large enough to provide accurate information for each proposed dose level. The minimum effective dose should also be ascertained.