Documentos de Académico
Documentos de Profesional
Documentos de Cultura
doi: 10.1111/j.1528-1167.2012.03688.x
SPECIAL REPORT
*Amedeo Avogadro University, Novara, Italy; yRush Medical College at Rush University, Chicago, Illinois, U.S.A.;
zVivantes Humboldt-Klinikum, Berlin, Germany; and xBeth Israel Deaconess Medical Center,
Harvard University, Boston, Massachusetts, U.S.A.
In 2008, the U.S. Food and Drug Administration (FDA) patients), and ‘‘other conditions’’ (48% patients). In the
issued an alert to health care professionals about an main analysis, almost 28,000 people taking AEDs and
increased risk of suicide ideation and behavior in people >16,000 people taking placebo were considered. There were
treated with antiepileptic drugs (AEDs) (FDA 2008a). Such four completed suicides altogether, all in people taking
a conclusion came from a meta-analysis of multicenter- AEDs and none in those taking placebo. The FDA con-
randomized placebo controlled trials of 11 AEDs. Pharma- cluded that patients receiving AEDs were significantly more
ceutical companies had been previously asked to submit likely to experience suicidal behavior or ideation compared
data from these trials, with at least 30 patients involved, with placebo (odds ratio [OR] 1.8; 95% confidence interval
regardless of indication. Spontaneously reported suicidality [CI] 1.24–2.66). In addition, they observed that the relative
occurring during double-blind trials with an AED (or within risk (RR) versus placebo was higher in patients treated for
1 day of stopping) were sought and categorized. Data were epilepsy (RR 3.6; 95% CI 1.3–12.1) than for selected psy-
provided on the use of carbamazepine, felbamate, gabapen- chiatric illnesses (RR 1.6; 95% CI 1–2.4) or other condi-
tin, lamotrigine, levetiracetam, oxcarbazepine, pregabalin, tions (e.g., migraine and neuropathic pain; RR 2.0; 95% CI
tiagabine, topiramate, valproate, and zonisamide used for 0.8–4.8) (FDA 2008b) where RR were not significant. As a
epilepsy (25% patients), psychiatric indications (27% result of the analysis, the FDA required that all manufactur-
ers of drugs in this class include a warning in their labeling
Accepted August 6, 2012; Early View publication September 20, 2012. and develop a medication guide for patients, informing them
Address correspondence to Marco Mula, Division of Neurology, Uni-
versity Hospital Maggiore della Carit, C.so Mazzini, 18, 28100 Novara,
of the risks of suicidal thoughts or actions. The FDA also
Italy. E-mail: marco.mula@med.unipmn.it suggested that the warning language be carefully worded
Wiley Periodicals, Inc. and that it mention the risk of not treating the underlying
ª 2012 International League Against Epilepsy condition (Busco, 2008).
199
200
M. Mula et al.
The FDA data was received with great skepticism by clini- 3 Most epilepsy trials (92%) included patients taking
cians and the professional societies, including the American adjunctive therapy (compared with 14% of psychiatric
Epilepsy Society, and some investigators questioned the trials and 15% of other medical trials). It is, thus, unclear
validity of its findings, after identifying serious methodologic whether the higher suicidality rates in the epilepsy trials
flaws (Hesdorffer & Kanner, 2009) and suggested that the were due to drug interactions, given the high proportion
FDA’s concern might have been excessive (Hesdorffer & of epilepsy trials designed with polytherapy, or whether
Kanner, 2009; Mula et al., 2010), and that the risk of stop- they potentially were due to the low suicidality risk asso-
ping (or not even starting) AEDs in people with epilepsy ciated with carbamazepine and valproate—both drugs are
would be far greater than this hypothetic small increased risk the most common comparison drugs in these trials and
of suicidality (Mula & Sander, 2010). The International Lea- demonstrated favorable psychotropic properties (Ettin-
gue Against Epilepsy (ILAE) appointed a Task Force under ger, 2006).
the Commission on Neuropsychobiology to formulate a pro- 4 Suicidal behavior was greater in certain geographic
posal for a consensus on treatment issues in patients with epi- regions. For example, the odds ratio of suicidality was
lepsy and psychiatric problems. Members of this task force 1.38 (95% CI 0.9–2.13) in North American studies and
(MM, BS, AMK) have been selected according to the follow- 4.53 (95% CI 1.86–13.18) in studies done elsewhere.
ing criteria: (1) to be boarded in both adult neurology and Such differences strongly suggest serious methodologic
adult psychiatry with certified clinical practice in adult neu- errors in data gathering (Hesdorffer & Kanner, 2009;
rology and adult psychiatry; (2) to work in an epilepsy center Hesdorffer et al., 2010).
as an epileptologist as main clinical practice; (3) to have pub-
lished relevantly in epileptology and neuropsychiatry of epi- Available Evidence on AEDs
lepsy. An additional member (SS) has been included as a
liaison with the ILAE Commission on Therapeutic strategies.
and Suicide
The present article represents an expert consensus statement In the last 3 years, a number of retrospective cohort and
on the issue of suicide during treatment with AEDs. Refer- case–control studies, using observational study methodol-
ences were identified by searches of Medline/PubMed using ogy, investigated whether or not there is an association
the terms ‘‘epilepsy,’’ ‘‘antiepileptic drugs,’’ ‘‘suicide.’’ Only between AEDs and suicidality (Gibbons et al., 2009, 2010;
articles published in English in international peer-reviewed Andersohn et al., 2010; Arana et al., 2010; Olesen et al.,
journals were considered. The reference list of relevant arti- 2010; Patorno et al., 2010; VanCott et al., 2010; Redden
cles was hand-searched for additional publications (e.g., et al., 2011). The majority of studies included data from
book chapters or review papers) if relevant to the discussion. patients with epilepsy, whereas some authors focused on
Abstracts of presentations, even from international con- psychiatric subjects only (Gibbons et al., 2009). In general
gresses, have not been included. Discussions at international terms, results are contradictory among the studies, with
workshops have been considered (Kanner et al., 2012). some studies reporting an increased suicidality risk (Ander-
sohn et al., 2010; Olesen et al., 2010; Patorno et al., 2010;
VanCott et al., 2010) and others observing no increased
Limits of the FDA Meta-Analysis risk (Arana et al., 2010; Gibbons et al., 2010). However,
An analysis of the FDA document demonstrates several all of these studies are being affected by a number of limi-
methodologic flaws, which include: tations (Hesdorffer et al., 2010; Mula & Hesdorffer, 2011)
1 The assessment of suicidality was based on ‘‘spontane- such as the failure to adjust for past suicidal behaviors
ous’’ reports of patients and not gathered in a systematic (Andersohn et al., 2010; Olesen et al., 2010). In one study,
prospective manner in every patient who was randomized the authors adjusted for a history of self-harm that does not
to study drug or placebo (Hesdorffer & Kanner, 2009; imply suicidal behavior, being two different entities from a
Hesdorffer et al., 2010). psychopathologic point of view (Mula & Hesdorffer,
2 The FDA’s warning indicates an increased risk of suicide 2011). Another study took into account past suicidality but
with all AEDs, despite the fact that statistical significance over an unspecified period of time (VanCott et al., 2010),
was found in only 2 (i.e., topiramate and lamotrigine) of whereas other authors (Arana et al., 2010; Patorno et al.,
the 11 AEDs studied. Furthermore, inclusion of three 2010) excluded subjects with a past history, or family his-
additional studies of lamotrigine resulted in the loss of tory (Arana et al., 2010), of suicidality before the study
statistical significance for this AED. Two other AEDs, entry. The issue of a past history of suicide behavior repre-
valproic acid and carbamazepine, actually yielded a sents an important variable because suicide is a highly
‘‘small protective effect.’’ The FDA’s decision to present recurrent phenomenon (between 14% and 17% after 1 year
the risk as involving all AEDs stemmed from a concern and >30% over 10 years) (Owens et al., 2002; Kapur et al.,
that singling out specific AEDs might have changed pre- 2006) and because prescribing practices may be influenced
scribing practices, rather than emphasize the suicide risk, by prior suicidal behaviors (Hesdorffer et al., 2010; Mula
a reasoning that was not based on scientific facts. & Hesdorffer, 2011), considering that some AEDs are
Epilepsia, 54(1):199–203, 2013
doi: 10.1111/j.1528-1167.2012.03688.x
201
Antiepileptic Drugs and Suicidality
thought to have positive effects on mood and others nega- Psychiatric Adverse Events of
tive (Ettinger, 2006). Therefore, in the absence of adjust- AEDs in Patients with Epilepsy
ment for prior suicidal behavior, it is impossible to
determine, from the published studies, whether AEDs are The psychotropic potential of AEDs in patients with epi-
associated with suicidal behavior (Mula & Hesdorffer, lepsy is related to direct and indirect mechanisms (Mula &
2011). Sander, 2007). A number of studies suggest that treatment
with some AEDs is associated with the occurrence of symp-
toms of depression (Table 1), whereas other compounds
The Issue of Suicide in Epilepsy have positive psychotropic properties. As far as first-genera-
In the general population, suicide represents the 11th tion compounds are concerned, authors agree that there is a
cause of death and the second in the group aged 25– link between barbiturates and depression, whereas carba-
34 years (Gelder, 2009). It is more common in men than in mazepine probably has mood stabilizing and antimanic
women, particularly in developed countries, whereas effects (Rodin et al., 1976; Dodrill & Troupin, 1977;
attempted suicide is more represented among females (Sa- Robertson & Trimble, 1983). Among second-generation
dock et al., 2009). In patients with epilepsy, the overall risk AEDs, vigabatrin (Levinson & Devinsky, 1999), tiagabine
of committing suicide is about three times higher than that (Trimble et al., 2000), and topiramate (Mula et al., 2003a)
of the general population (Harris & Barraclough, 1997; have been linked to treatment-emergent depressive
Christensen et al., 2007; Bell et al., 2009). Several studies symptoms, whereas levetiracetam with dysphoria and mood
have attempted to identify reasons for such an increased lability (Mula et al., 2003b).
risk. In fact, suicidality and epilepsy have a complex rela- In some cases, treatment-emergent depressive symptoms
tion, based on several variables. In the general population, are associated with a sudden complete control of seizures
about 90% of people who successfully commit suicide have (the forced normalization phenomenon) (Ring et al., 1993),
at least one psychiatric disorder at that time (Barraclough, whereas in others they are unrelated to this. Nevertheless, a
1987). Epilepsy can be complicated by psychiatric comor- rapid titration of the drug (Mula et al., 2009) in patients with
bidities, but it is unlikely that such comorbidity is the only drug-refractory mesial temporal lobe epilepsy and a past
responsible element. A Danish study pointed out that the history of depression (Mula et al., 2007) have been identi-
rate ratio of suicide in people with epilepsy is still doubled fied as major determinants in the majority of cases. Particu-
even after excluding people with psychiatric comorbidity larly, several studies pointed out that the development of
and adjusting for various factors and that it increases by psychiatric adverse events (which could potentially facili-
32-fold in the presence of comorbid mood disorders and by tate the occurrence of suicidal ideation and behavior) is
12-fold in the presence of anxiety disorders and schizo- more likely in patients at risk of developing psychiatric
phrenia. Some have suggested a link with temporal lobe disorders, either because of a past psychiatric history or a
epilepsy (Harris & Barraclough, 1997). However, a recent family psychiatric history (Kanner et al., 2003; Mula et al.,
study, using retrospective and prospective methods, found 2007).
no epilepsy-related factors (Hara et al., 2009). Postictal sui-
cidal ideation is relatively frequent in patients with treat- Conclusions
ment-resistant partial epilepsy, having been identified in
13% of 100 consecutive patients with a median duration of 1 Although some (but not all) AEDs can be associated with
24 h (Kanner et al., 2004). Furthermore, a bidirectional treatment-emergent psychiatric problems that can lead to
relation has been identified between suicidality and suicidal ideation and behavior, the actual suicidal risk is
epilepsy, whereby patients with a history of suicidal behav- yet to be established, but it seems to be very low. The risk
ior have a fivefold higher risk of developing epilepsy in a of stopping AEDs or refusing to start AEDs is signifi-
population-based study conducted in Iceland (Hesdorffer cantly worse and can actually result in serious harm
et al., 2006). This bidirectional relation raises the question
of common pathogenic mechanisms operant in both
conditions such as serotonin dysfunction, a hyperactive
hypothalamic-pituitary-adrenal axis, as well as glutamate Table 1. Treatment-emergent psychiatric adverse
and c-aminobutyric acid (GABA)-ergic disturbances events of AEDs in patients with epilepsy
(Hecimovic et al., 2011). Depression Psychoses Irritability/Emotion lability
In conclusion, all the complexities of the relation between Barbiturates Ethosuximide Felbamate
epilepsy and suicide are still far from being elucidated, but Tiagabine Levetiracetam Lamotrigine
they are clearly multifactorial with biologic, constitutional, Topiramate Phenytoin (toxic levels) Levetiracetam
and psychosocial variables being implicated. Yet, given the Vigabatrin Topiramate
Zonisamide Vigabatrin
increased risk of suicide in patients with epilepsy, screening Zonisamide
for suicide is a relevant issue.
Epilepsia, 54(1):199–203, 2013
doi: 10.1111/j.1528-1167.2012.03688.x
202
M. Mula et al.
Mula M, Hesdorffer DC. (2011) Suicidal behavior and antiepileptic drugs Posner K, Brown GK, Stanley B, Brent DA, Yershova KV, Oquendo MA,
in epilepsy: analysis of the emerging evidence. Drug Healthc Patient Currier GW, Melvin GA, Greenhill L, Shen S, Mann JJ. (2011) The
Saf 3:15–20. Columbia-Suicide Severity Rating Scale: initial validity and internal
Mula M, Sander JW. (2007) Negative effects of antiepileptic drugs on consistency findings from three multisite studies with adolescents and
mood in patients with epilepsy. Drug Saf 30:555–567. adults. Am J Psychiatry 168:1266–1277.
Mula M, Sander JW. (2010) Antiepileptic drugs and suicide risk: could Redden L, Pritchett Y, Robieson W, Kovacs X, Garofalo M, Tracy K,
stopping medications pose a greater hazard? Expert Rev Neurother Saltarelli M. (2011) Suicidality and divalproex sodium: analysis of con-
10:1775–1776. trolled studies in multiple indications. Ann Gen Psychiatry 10:1.
Mula M, Trimble MR, Lhatoo SD, Sander JW. (2003a) Topiramate and Ring HA, Crellin R, Kirker S, Reynolds EH. (1993) Vigabatrin and depres-
psychiatric adverse events in patients with epilepsy. Epilepsia 44: sion. J Neurol Neurosurg Psychiatry 56:925–928.
659–663. Robertson MM, Trimble MR. (1983) Depressive illness in patients with
Mula M, Trimble MR, Yuen A, Liu RS, Sander JW. (2003b) Psychiatric epilepsy: a review. Epilepsia 24(Suppl. 2):S109–S116.
adverse events during levetiracetam therapy. Neurology 61:704–706. Rodin EA, Rim CS, Kitano H, Lewis R, Rennick PM. (1976) A comparison
Mula M, Trimble MR, Sander JW. (2007) Are psychiatric adverse events of of the effectiveness of primidone versus carbamazepine in epileptic out-
antiepileptic drugs a unique entity? A study on topiramate and leveti- patients. J Nerv Ment Dis 163:41–46.
racetam. Epilepsia 48:2322–2326. Ryvlin P, Cucherat M, Rheims S. (2011) Risk of sudden unexpected death
Mula M, Hesdorffer DC, Trimble M, Sander JW. (2009) The role of titra- in epilepsy in patients given adjunctive antiepileptic treatment for
tion schedule of topiramate for the development of depression in refractory seizures: a meta-analysis of placebo-controlled randomised
patients with epilepsy. Epilepsia 50:1072–1076. trials. Lancet Neurol 10:961–968.
Mula M, Bell GS, Sander JW. (2010) Suicidality in epilepsy and possible Sadock BJ, Sadock VA, Ruiz P. (2009) Kaplan and Sadock’s comprehen-
effects of antiepileptic drugs. Curr Neurol Neurosci Rep 10:327–332. sive textbook of psychiatry. Wolters Kluwer Health/Lippincott
Olesen JB, Hansen PR, Erdal J, Abildstrom SZ, Weeke P, Fosbol EL, Poul- Williams & Wilkins, Philadelphia, PA, London.
sen HE, Gislason GH. (2010) Antiepileptic drugs and risk of suicide: a Trimble MR, Rusch N, Betts T, Crawford PM. (2000) Psychiatric symp-
nationwide study. Pharmacoepidemiol Drug Saf 19:518–524. toms after therapy with new antiepileptic drugs: psychopathological
Owens D, Horrocks J, House A. (2002) Fatal and non-fatal repetition of and seizure related variables. Seizure 9:249–254.
self-harm. Systematic review. Br J Psychiatry 181:193–199. VanCott AC, Cramer JA, Copeland LA, Zeber JE, Steinman MA, Dersh JJ,
Patorno E, Bohn RL, Wahl PM, Avorn J, Patrick AR, Liu J, Schneeweiss S. Glickman ME, Mortensen EM, Amuan ME, Pugh MJ. (2010) Suicide-
(2010) Anticonvulsant medications and the risk of suicide, attempted related behaviors in older patients with new anti-epileptic drug use: data
suicide, or violent death. JAMA 303:1401–1409. from the VA hospital system. BMC Med 8:4.