Social Audit Ltd
P.O. Box 111 London NW1 8XE
Telephone/Fax: 020 7586 7771

[email protected]


Dr Julie Williams, Post-Licensing Division
Medicines and Healthcare products Regulatory Agency
Market Towers, 1 Nine Elms Lane
London SW8 5NQ

16 May 2006


Dear Dr Williams,

I’m writing to you because I’m concerned that the MHRA/CHM/CSM analysis of the suicidality risks with paroxetine make no distinction between the two different formulations of paroxetine – treating the original formulation of immediate release (IR) paroxetine as if identical to the more recently introduced (in the USA), controlled release (CR) version of the same drug. As the CR version of paroxetine is not licensed for use in the UK, I am writing to ask what evidence the MHRA/CHM is able to offer that these two different formulations have identical effects. Since neither the SSRI expert working group, nor the Gunnell & Ashby analysis, referred to any possible differences between these two formulations, I assume that the MHRA will be readily able to provide the evidence on which its determinations were based.

This FOI request was partly prompted by GSK’s recently published new data on the risk of paroxetine-induced suicidal behaviour in depressed patients – and specifically refers to the ‘Paroxetine Adult Suicidality Analysis’, Table 2.04 recording ‘Number and percent of Subjects with Definitive Suicidal Behavior by Indication, Treatment groups and trial indication = MDD’. The crude figures in this Table suggest a six-fold greater risk for depressed patients treated with paroxetine rather than placebo – but the risk appears higher if the results for Paroxetine CR are excluded from the analysis (which would seem appropriate in the UK, since the IR product is not licensed/used here).

I have not been able to identify from the GSK clinical trials database the basic data for four of the 19 trials listed (numbers 002, 003, 009 and 274), but trial numbers 448, 449 and 487 include both paroxetine IR and CR and trial numbers 785, 810 and 874 involved paroxetine CR only. The following tables compare the results presented by GSK (including both paroxetine IR and CR) with the results for controlled- and immediate- release paroxetine on their own:

Suicidality analysis for paroxetine IR and CR combined, versus placebo:

Paroxetine (IR & CR)


11/3455 (0.32%)

1/1978 (0.05%)


Suicidality analysis for paroxetine IR (excluding CR), versus placebo:

Paroxetine IR only


10/2285 (0.44%)

1/1545 (0.06%)


Suicidality analysis for paroxetine CR (excluding IR), versus placebo:

Paroxetine CR only


1/1170 (0.08%)

0/755 (0.00%)

Yes, the number of events is small and, no, these figures aren’t conclusive. I have referred to them simply to underline the significance of this and one further information request: Did GSK (including SKB) ever apply for a Product License for paroxetine CR and, if so, was that application either withdrawn or refused, and when? If there was a license application for controlled-release paroxetine, and it is pretty much identical to the immediate release formulation, one must wonder why it isn’t on the market now.

Thank you for your attention. I look forward to hearing from you.

Yours sincerely,
Charles Medawar