Selective serotonin reuptake inhibitors (SSRIs) and suicide in adults: meta-analysis of drug company data from placebo controlled, randomised controlled trials submitted to the MHRA's safety review
BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7488.385 (Published 17 February 2005) Cite this as: BMJ 2005;330:385
All rapid responses
When the randomised control trial data submitted by pharmaceutical
companies is used for a meta-analysis to investigate whether SSRIs are
associated with increased risk of suicide a number of limitations are
there. First of all suicide is a very rare event even in patients with
severe depression. As correctly pointed out by the authors even after
pooling the data from hundreds of RCTs the study did not have sufficient
power and results should be interpreted cautiously. The RCTs had only few
cases to confirm or rule out an increased risk of suicide with SSRIs. To
complicate this further there is an obvious chance of underreporting of
suicide related outcomes in these clinical trials because the research
question in most of these RCTs would be focused on the efficacy of the
drug. Since the primary aim is not to check the association of SSRIs and
suicide, the reported adverse event may have been considered to be due to
the mental illness rather than medication by the researchers which would
lead to under-reporting. Reporting of events in subgroups is again
difficult because of the lesser number of patients. Even though there is a
less chance of differential reporting in placebo controlled trials it is
very important to make sure that allocation concealment was adequate,
which may influence the degree of effect. The selective inclusion of drug
company data would result in a location [publication] bias.
According to MHRA data [Incidence of possible suicide-related events
by treatment group and age group: adult active-control trials] data of
Paroxetine and a comparator in the age group above 70 shows an increase
incidence of possible suicide-related events in Paroxetine group.
[Paroxetine n/N (%) - 4/457 (0.9%); Comparator n/N (%) - 1/390 (0.3%) ;
Odds ratio (95% CI) - 3.4 (0.4, 30.8) ; p-value 0.38 ]
In spite of a wide confidence interval an large p –value this is an
interesting finding and we wonder whether it would be due to the
differences in the pharmaco-dynamics in the elderly population. In that
case this would be very significant because of the closeness to the real
life scenario than the research setting as observed by Alexander M
Ponizovsky in rapid responses to this article on 22 February 2005.
In the meta analysis the authors attempted to quantify heterogeneity
and recognized that heterogeneity between trials and individual products
is likely to have been masked. However we find that the paper is subject
to a high level of clinical heterogeneity because, the data is summarised
for all treatment indications, except for Citalopram. Moreover difference
in the length of the follow-up in different trials and, randomization
ratios are not adequately reported which could lead to over or under
estimation of risk.
Finally when the data provided by Prof. Healy is included in the
analysis there is a major shift of the study result in favour of placebo.
The call for urgent research by the authors to identify people at risk is
the most appropriate approach to clarify this problem. ,
Competing interests:
None declared
Competing interests: No competing interests
Dear Editor,
Professor Healy (18 February 2005) raises concerns about two of the
numbers used in our meta-analysis. As we highlighted in our paper, the
data were drawn from the report of the Medicine and Healthcare products
Regulatory Agency’s (MHRA) review of the safety of SSRIs (1). We can
confirm that the Expert Working Group’s report did, in fact, include one
suicide amongst placebo treated patients in placebo-controlled trials of
citalopram for depression (see Table 7.16 on page 84). Likewise data on
paroxetine are available in the report (see section 7.2.1 on page 74) and
were included in our meta-analyses. Three of the four suicides in the
placebo-controlled trials of paroxetine were reported as having occurred
in the post-treatment period and the one paroxetine suicide occurred on-
treatment.
We have communicated with the MHRA concerning Professor Healy’s
concern with the paroxetine suicide data. In further consultation with GSK
(the license holder for paroxetine) the MHRA confirm that the four
suicides in adult placebo-controlled randomised trials of paroxetine were
distributed as described above (personal communication: MHRA).
Department of Social Medicine, University of Bristol, Bristol BS8 2PR
David Gunnell
professor of epidemiology
Wolfson Institute of Preventive Medicine, Queen Mary, University of
London; London EC1M 6BQ
Julia Saperia
research assistant
Deborah Ashby
professor of medical statistics
1. Report of the CSM Expert Working Group on the safety of selective
serotonin reuptake inhibitors. 2004.
http://www.mhra.gov.uk/news/2004/SSRIfinal.pdf
Competing interests:
DG and DA were members of the MHRA’s expert working group on the safety of SSRIs. We acted as independent advisors, receiving travel expenses and a small fee for meeting attendance and reading materials in preparation for the meeting. DA has spoken on the methodology of adverse drugs reactions in HIV at a scientific meeting attended by several pharmaceutical companies, and sponsored by GSK. An honorarium was paid to her department. JS has no competing interests.
Competing interests: No competing interests
There has been much recent discussion regarding the apparent increase
in suicidality in patients treated with selective serotonin reuptake
inhibitors, particularly in early stages of treatment. Despite
difficulties in interpreting data from post-mortem tissue, there appears
to be some evidence to suggest that a reduction in dopamine turnover in
the basal ganglia is associated with suicide (Bowden et al., 1997). This
observation appears consistent with evidence that reduced activity of
dopaminergic neurons is associated with feelings of dysphoria (Voruganti
et al., 2001), whilst increases in activity lead to reward and euphoria.
Serotonergic neurons originating in the raphe nucleus modulate the
firing of dopaminergic neurons in the ventral tegmental area (VTA), the
origin of mesocorticolimbic dopaminergic projections to the nucleus
accumbens and frontal cortex. An interaction between serotonergic and
dopaminergic pathways has been proposed as a possible mechanism underlying
the therapeutic action of antidepressants (D`Aquila, et al., 2000), and an
increase in dopaminergic activity is associated with long-term
antidepressant administration (Bonhomme and Esposito, 1998). However,
electrophysiological studies in rodents have shown that acute
administration of fluoxetine and other serotonin selective reuptake
inhibitors (SSRIs), reduces the activity of dopaminergic neurons in the
VTA (Prisco and Esposito, 1995; Di Mascio et al., 1998). The authors of
these studies suggest that this may account for the akathisia sometimes
associated with SSRI therapy. However, it also seems conceivable that a
reduction in the activity of these neurons may result in suicidal thoughts
due to dysphoria and a reduction in motivation during initial treatment
with SSRIs. Following chronic fluoxetine administration (once daily
administration for twenty-one days), the reduction in dopaminergic
activity associated with acute fluoxetine challenge was lost. As neuronal
function returns to normal over a longer period of drug administration,
motivation and feelings of pleasure may return, leading to a reduction in
suicidal thoughts and improvement in affect. This may explain why the risk
of suicide associated with antidepressants is increased during early
stages of treatment, in particular during days one to nine (Jick et al.,
2004), but appears to be attenuated as treatment continues.
References:
Bonhomme, N., Esposito, E., 1998. Involvement of serotonin and
dopamine in the mechanism of action of novel antidepressant drugs: a
review. J. Clin. Psychopharmacol., 18(6): 447-454.
Bowden, C., Cheetham, S.C., Lowther, S., Katona, C.L.E., Crompton,
M.R., Horton, R.W. 1997. Reduced dopamine turnover in the basal ganglia of
depressed suicides. Brain Res., 769(1): 135-140.
D`Aquila, P.S., Collu, M., Gessa, G.L., Serra, G., 2000. The role of
dopamine in the mechanism of action of antidepressant drugs. Eur. J.
Pharmacol., 405: 365-373.
Di Mascio, M., Di Giovanni, G., Di Matteo, V., Prisco, S., Esposito,
E., 1998. Selective serotonin reuptake inhibitors reduce the spontaneous
activity of dopaminergic neurons in the ventral tegmental area. Brain Res.
Bull., 46(6): 547-554.
Jick, H., Kaye, J.A., Jick, S.S., 2004. Antidepressants and the risk
of suicidal behaviours. J.A.M.A., 292(3): 379-381.
Prisco, S., Esposito, E., 1995. Differential effects of acute and
chronic fluoxetine administration on the spontaneous activity of
dopaminergic neurons in the ventral tegmental area. Br. J. Pharmacol.,
116: 1923-1931.
Voruganti, L., Slomka, P., Zabel, P., Costa, G., So, A., Mattar, A.,
Awad, A.G., 2001. Subjective effects of AMPT-induced dopamine depletion in
schizophrenia: Correlation between dysphoric responses and striatal D2
binding ratios on SPECT imaging. Neuropsychopharmacology, 25(5): 642-650.
Competing interests:
None declared
Competing interests: No competing interests
In an exhaustive Bayesian meta-analysis, Gunnell et al (1) explored
the suicide risk from pharmaceutical data on the different selective
serotonin reuptake inhibitors (SSRIs) currently on the market. Their
results give rise to some comments.
First, it has been observed by clinicians since the beginning of the use
of antidepressants that the first weeks of treatment of a severe
depression are accompanied by a higher risk of suicide, because of a drug-
induced motor disinhibition which is not yet accompanied by a mood
improvement. Probably a phenomenon such as akathisia mentioned by
Fergusson et al (2) adds up to the clinical picture. This element is
briefly mentioned by Gunnell et al (1) in the discussion but not
developed. During the first weeks of antidepressant treatment, patients
need to be closely monitored for this risk.
The global findings of a trend toward a protective effect of SSRI
concerning suicidal thoughts (pooled odds ratio (OR) 0.77) compared to a
trend towards an increased risk of self-harm (pooled OR 1.57) is somewhat
paradoxical even though the results are not statistically different. More
surprising is the heterogeneity of results among the different types of
SSRIs shown on the Figure. Why sertraline would show a protective effect
for suicidal thoughts and simultaneously increase the risk of self-harm?
Also strange is the difference of risks between citalopram and
escitalopram, while escitalopram is simply the active S-enantiomere of
citalopram. A difference of safety profile between the two compounds is
surprising. More generally, there is currently no strong biological
rationale to explain such heterogeneity in a pharmacological class of
drugs with the same mechanism of action.
In the Discussion, the authors mention also that according to a review by
the Medicine and Healthcare products Regulatory Agency (MHRA), there was
little evidence to show difference in the risk profile between SSRI and
other antidepressants. The two accompanying papers (2,3) show indeed that
the suicidal risk appears similar for SSRI and tricyclic antidepressants.
We think that this is an essential point and that the risk associated with
suicide should be assessed for the whole class of antidepressants.
The next stage would be to quantify the risk of suicide attempts according
to time since start of antidepressant. In terms of benefit-risk ratio,
there is probably an initial period where the risks are increased compared
to the benefits. It is only when the full improvement of the mood takes
place that the full benefit of antidepressants will be observed. It would
be interesting and useful to confirm by evidence-based methods the old
clinical observations.
1. Gunnell D, Saperia J, Ashby D. Selective serotonin reuptake
inhibitors (SSRI) and suicide in adults: meta-analysis of drug company
from placebo controlled, randomised controlled trials submitted to the
MHRA’s safety review. BMJ 2005:330;385
2. Fergusson D, Doucette S, Cranley Glass K, Shapiro S, Healy D,
Hebert P, Hutton B. Association between suicide attempts and selective
serotonin reuptake inhibitors: systematic review of randomised controlled
trials BMJ 2005;330:396
3. Martinez C, Rietbrock S, Wise L, Ashby D, Chick J, Moseley J,
Evans S, Gunnell D Antidepressant treatment and the risk of fatal and non-
fatal self harm in first episode depression: nested case-control study.
BMJ 2005;330:389
Competing interests:
None declared
Competing interests: No competing interests
Dear Sirs
David Gunnell et al’s study conclusion of a trend towards increased
risk of suicide and self harm caused by SSRIs should be accepted very
seriously, given that it based on data from placebo-controlled trials.
Their conclusions could become substantially stronger, if they have
discussed a methodological problem of clinical trials. This problem is the
discrepancy between the efficacy and safety that an experimental drug
demonstrates in clinical trials and its beneficial and adverse effects in
daily clinical practice. One possible explanation for this discrepancy is
that the service context of clinical trials is essentially different from
routine psychiatric care. The main difference is the selection of patients
characterized by predominance of specific types of symptoms, and the
application of multiple exclusion criteria in clinical trials (in this
case, any manifestation of suicidality). In addition, only the
experimental compound in a restrictively defined dosing regimen is
administered in a trial, whereas polypharmacy and a broad dosage range are
prevalent in everyday practice. As a result, clinical trials recruit
patients who tend to be in a minority, while findings and conclusions have
been generalized to entire populations of patients with similar diagnoses.
These considerations always must be taken into account when we interpret
data obtained from clinical trials.
Gunnell D, Saperia J, Ashby D. Selective serotonin reuptake
inhibitors (SSRIs) and suicide in adults: meta-analysis of drug company
data from placebo controlled, randomized controlled trials submitted to
the MHRA's safety review. BMJ. 2005 Feb 19;330(7488):385.
Competing interests:
None declared
Competing interests: No competing interests
Dear Sirs
The figures for suicides on SSRIs from placebo controlled trials
produced by David Gunnell and colleagues in this article may not be all
they seem. In the Expert Working Group report on SSRIs (EWG)(1), the
table for citalopram trials suggests there was no suicide in the placebo
group. Data on paroxetine are not available from the EWG report, but my
reading of prior submissions on paroxetine suggests there were 4 suicides
on paroxetine in placebo controlled trials (2). The authors note 3
suicides on placebo in the withdrawal phase of these trials. The Safety
Review of paroxetine conducted by M Brecher in 1991 does not suggest these
suicides happened in the withdrawal phase of placebo-controlled trials
(3). Unless the authors can confirm from their own inspection of the raw
data that there were in fact 3 suicides in the withdrawal phase, as
opposed to relying on a company submission, these figures must be in some
doubt.
Based on the above, there may therefore have been 12 suicides in
23,804 SSRI patients versus 6 suicides in 17,022 placebo patients, an odds
ratio of 1.43, or possibly 12 SSRI suicides versus 3 placebo suicides, an
odds ratio of 2.86. Given the confusion about paroxetine, if it is left
out, the figures become 8 suicides in 15,323 SSRI patients versus 3
suicides in 11,214 placebo patients, an odds ratio of 1.96. Adding in
suicides on venlafaxine and mirtazapine from the EWG gives 16 suicides in
23,885 antidepressant patients versus 3 in 14,564 placebo patients, an
odds ratio of 3.1.
Given that antidepressants may reduce the risk of suicide in some
patients, even an odds ratio of 1.0 points to a risk. As things stand at
present, therefore, we have a very clear signal of risk from both clinical
accounts of the problem and clinical trials. Against this background
Gunnell and colleagues call for more research is compelling. But their
suggestion as regards the numbers needed to recruit to an RCT to establish
a drugs ability to reduce the risk of suicide is likely to deter further
research. In contrast, an RCT with a challenge-dechallenge design and a
rating scale sensitive to suicidal ideation, might need less than a
hundred patients to establish conclusively whether there are patients at
risk of induced suicidality. Eli Lilly designed such an RCT in
conjunction with FDA in 1990. The details are available for posting on
the BMJ website.
1 Report on the CSM Expert Working Group on the Safety of Selective
Serotonin Reuptake Inhibitor Antidepressants.
www.mhra.gov.uk/news/2004/SSRIfinal.pdf
2 CSM Expert Working Group on the Safety of SSRIs. Risk:Benefit
Evaluation of paroxetine. Williams J, Wise L, Dunne J, Day S, Davies C,
October 2003.
3. Brecher M (1991). Review and evaluation of clinical data original
NDA 20-031. Paroxetine safety review. Department Heath & Human
Services, Washington, Center for Drug Evaluation and Research, Freedom of
Information File F99-22360
Competing interests:
Extensive links to all the major pharmaceutical companies making antidepressants, and expert witness in antidepressant linked legal cases
Competing interests: No competing interests
Suicidal behaviour and SSRIs - updated meta-analysis
Dear Editor,
Last year you
published our article on the risk of suicide from selective serotonin reuptake
inhibitors (SSRIs).1 In the light of recently released data for
paroxetine by its manufacturer GlaxoSmithKline2 we wish to update our results.
In our original
analysis, we were unable to distinguish between occurrences of non-fatal self
harm and suicidal thoughts for patients in paroxetine trials. Our main analysis
of these events therefore excluded the paroxetine data; in a sensitivity
analysis we divided the events equally between self harm and suicidal thoughts.
The new data released by GlaxoSmithKline come from placebo controlled trials of
paroxetine and combine data on completed suicides, attempted suicide and
preparatory acts towards imminent suicidal behaviour
into a single category of “definitive suicidal behaviour”.
For consistency with our original article, we have used data on all
indications, although GlaxoSmithKline have provided a full breakdown. In 57
trials, there were 50/8958 events in the paroxetine arm and 40/5953 in the
placebo arm. The newly released data suggest the figures in relation to
suicidal thoughts were 33/8958 for paroxetine and 25/5953 for placebo. There is
no new information on completed suicides.
We have used the
newly available data to update our meta-analysis. Our findings are similar to
those in our published paper.
Using the same
Bayesian random effects meta-analysis as before, the odds ratio for non-fatal
self harm in patients taking an SSRI compared to placebo is 1.21 (95% credible
interval 0.87 to 1.83). For suicidal thoughts, we have an odds ratio of 0.80
(95% credible interval 0.49 to 1.30). The previous results were an odds ratio
of 1.57 (95% credible interval 0.99 to 2.55) for self harm and an odds ratio of 0.77 (95% credible interval 0.37 to 1.55) for suicidal thoughts. The results suggest
that the overall effect on non-fatal self harm is reduced compared to our
previous estimate, and slightly increased for suicidal thoughts. As before,
this analysis is limited by the length of the trials and inconsistent
collection of safety endpoints. More evidence is needed to reliably assess
specific adverse effects of SSRIs in relation to
their use in particular disorders such as the increased risk of self harm
recently reported for paroxetine in major depressive disorder.3 We feel that it is worth bringing this to the
attention of your readers. Updated tables and figures are available on request.
1. Gunnell D, Saperia J, Ashby D. Selective
serotonin reuptake inhibitors (SSR1s) and suicide in adults: meta-analysis of
drug company data from placebo controlled, randomised controlled trials
submitted to the MHRA's safety review. BMJ. 2005;330:385-388A.
2.
GlaxoSmithKline. Paroxetine and Adult Patients - 2006
analysis.http://www.gsk.com/media/par_current_analysis.htm
(accessed 10 May 2006).
3. Duff
G. Letter to clinical colleagues in
relation to "Paroxetine (Seroxat)
– risk of suicidal behaviour in adults". http://www.mhra.gov.uk/home/idcplg?IdcService=GET_FILE&dID=20961&noSaveAs=1&Rendition=WEB (accessed 23 May
2006).
Competing interests:
DG and DA were members of the MHRA's expert working group on the safety of SSRIs. They acted as independent advisers, receiving travel expenses and a small fee for meeting attendance and reading materials in preparation for the meeting. DA has spoken on the methods of adverse drugs reactions in HIV at a scientific meeting attended by several pharmaceutical companies, and sponsored by GlaxoSmithKline. An honorarium was paid to her department.
Competing interests: No competing interests