- Email: [email protected]
Non Case Study in the French Pharmacovigilance Database
Thérapie 2015 Septembre-Octobre; 70 (5): 425–432 DOI: 10.2515/therapie/2015026
PHARMACOEPIDEMIOLOGY
© 2015 Société Française de Pharmacologie et de Thérapeutique
Drug-induced Depression: a Case/Non Case Study in the French Pharmacovigilance Database Claire Lafay-Chebassier1, François Chavant1, Sylvie Favrelière1, Véronique Pizzoglio2, Marie-Christine Pérault-Pochat1 and the French Association of Regional Pharmacovigilance Centers 1 2
CHU de Poitiers, Service de Pharmacologie Clinique et Vigilances, Poitiers, France Agence Nationale de Sécurité du Médicament et des produits de santé (ANSM), Direction de la Surveillance, Saint-Denis, France
Text received September 30th, 2014 ; accepted January 27th, 2015
Keywords: depression; drug; adverse drug reaction; case/non case study
Abstract – Depression is a complex disorder with heterogeneous clinical anomalies whose neurobiological understanding still remains unclear. Medications have been implicated as potential causes of depression but for many of them, data are controversial. The present study aims to investigate association bet ween drugs and reports of depression. We used the case/non case method in the French pharmacovigilance database (FPVD) to identify drugs associated with depression. Cases were reports of depression in the FPVD between January 2007 and December 2011. Non cases were all other reports during the same period. Data were expressed as reporting odds ratio (ROR) with their 95% confidence interval. Of the 114 692 reports recorded in the FPVD during the studied period, we identified 474 cases of depression. For the majority of the patients, they were considered as “non serious” (56%) and evolution was favorable (64%). Significant RORs were found for antiepileptics (topiramate, levetiracetam), anti-infective and especially anti-retroviral drugs (efavirenz, emtricitabine, tenofovir, etravirine, raltegravir), interferons and other agents including isotretinoin, methylphenidate, sodium oxybate, varenicline, montelukast, flunarizine, adalimumab, anastrozole. Taking into account the limits of the methodology, the present study described associations with mainly expected drugs belonging to various therapeutic classes but it also found a signal with some anti-retrovirals. On the contrary, we did not find some assumed associations like cardiovascular medications, antimalarial. For most of the drugs, one or more mechanisms were found to explain these depressogenic effects on the basis of animal and human literature. Even if such associations need to be confirmed by further prospective studies, cautions are necessary for many drugs to early detect depressive symptoms.
Mots clés : dépression ; médicament ; effet indésirable ; étude cas/non cas
Résumé – Dépression induite par les médicaments : étude cas/non-cas dans la banque nationale de pharmacovigilance. La dépression est une pathologie complexe, aux signes cliniques hétérogènes et dont les mécanismes neurobiologiques restent encore mal connus. Les médicaments peuvent être à l’origine de dépression mais pour beaucoup d’entre eux, les données sont controversées. L’objectif de cette étude est d’évaluer l’association entre les médicaments et les notifications de dépression. Nous avons utilisé la méthode cas/non-cas dans la banque nationale de pharmacovigilance (BNPV) afin d’identifier les médicaments associés à la survenue de dépression. Les cas étaient les notifications de dépression enregistrées dans la BNPV entre janvier 2007 et décembre 2011. Les non-cas correspondaient à toutes les autres notifications enregistrées pendant la même période. Les données étaient exprimées par un reporting odds ratio (ROR) avec son intervalle de confiance à 95 %. Sur les 114 692 notifications enregistrées dans la BNPV au cours de la période d’étude, nous avons identifié 474 cas de dépression. Pour la majorité des patients, ils étaient considérés comme non graves (56 %) et l’évolution était favorable (64 %). Des ROR significatifs étaient retrouvés pour des anti-épileptiques (topiramate, lévétiracétam), des anti-infectieux et notamment des antirétroviraux (éfavirenz, emtricitabine, ténofovir, étravirine, raltégravir), des interférons et d’autres molécules incluant l’isotrétinoïne, le méthylphénidate, l’oxybate de sodium, la varénicline, le montélukast, la flunarizine, l’adalimumab et l’anastrozole. En tenant compte des limites liées à la méthodologie, cette étude a décrit des associations avec principalement des médicaments attendus appartenant à des classes thérapeutiques diverses mais il a également été retrouvé un signal avec certains antirétroviraux. A l’inverse, nous n’avons pas retrouvé les associations précédemment décrites avec des médicaments cardiovasculaires ou antipaludéens. Pour la plupart des médicaments, un ou plusieurs mécanismes ont été retrouvés pour expliquer ces effets dépressogènes. Même si ces associations nécessitent d’être confirmées par des études prospectives, des précautions sont nécessaires pour de nombreux médicaments afin de détecter précocement des symptômes dépressifs.
Abbreviations: see end of article.
Article publié par EDP Sciences
426
Lafay-Chebassier et al.
1. Introduction Mood disorders represent one of the most prevalent forms of mental illness. Depression is a complex heterogeneous disorder with a wide spectrum of anomalies including depressed mood, anhedonia, sleep disorders, fatigue, loss of self-esteem, negative thinking and suicidality. Functional brain imaging and neurobiological techniques have revealed specific functional and structural brain abnormalities but the neurobiological understanding of depression still remains unclear. Genetic and environmental factors such as stress and emotional trauma, viral infections, medical conditions have been implicated in depressive symptoms. Medications may also cause neuropsychiatric symptoms and, in literature, a wide variety of drugs have been implicated as potential causes of new-onset depression or worsening of established depression.[1] Unfortunately, contradictory data have been reported for many agents leading to the lack of conclusive data.[1,2] Indeed, it can be difficult to determine whether the suspected drug is truly causing the depressive symptoms or whether its use was concomitant with an endogenous depression on the basis of case reports. However, identification of drugs associated with depressive symptoms is important because untreated depression can evolve into suicidal ideation and suicide attempts and the lack of adherence to therapy is often closely related to the central adverse effects caused by drugs. Thus the aim of the present study was to evaluate association between drugs exposure and reports of depression using the case/ non case method in the French pharmacovigilance database.
2. Methods The data source used is the French pharmacovigilance database (FPVD) that has been previously described.[3] Briefly, the French pharmacovigilance system was created in 1973 and, since 1985, all spontaneous reports of adverse drug reactions (ADRs) have been recorded into the FPVD. According to French law, every health professional must report “serious” and/or “unexpected” adverse drug reactions to their Regional Center of Pharmacovigilance (31 centers in France). “Serious” ADRs are defined as reactions resulting in death, life-threatening event, hospitalization (or prolongation of existing hospitalization), persistent or significant disability or incapacity, congenital abnormalities/birth defect or other significant medical event. ADRs are defined as “unexpected” if their nature, seriousness or evolution is not consistent with data given by the drug monograph. For each ADR report, information about the patient (age, sex, medical history), drug exposure (to the suspected drug and other associated non-suspected drugs) and ADR characteristics (“serious” or “non-serious”, “expected” or “unexpected”, causality score, outcome) are recorded in the FPVD. ADRs are coded according to the Medical Dictionary for Regulatory Activities (MedDRA, version 15.1).[4]
© 2015 Société Française de Pharmacologie et de Thérapeutique
2.1. Case/non case method The case/non case method measures the disproportionality of combination between a drug and a particular adverse drug report (ADR) in a pharmacovigilance database.[5] This method can be used to generate signals from a pharmacovigilance database. Cases are defined as reports of the ADR of interest and non cases as all other reports of ADR. For each drug of interest, the association with the ADR was assessed by calculating an ADR reporting odds ratio (ROR) with its 95% confidence interval (CI). The ROR is the ratio of reporting of one specific event versus all other events for a given drug compound.[6] 2.2. Selection of cases and non cases In our study, we collected ADRs notified between January 1st, 2007 and December 31st, 2011. Cases were identified using standardised MedDRA queries (SMQ) “Depression (excl. suicide and self injury)”. After a careful analysis of the comments of spontaneous reports, we excluded all cases that were not “depression”. Non cases, using as controls, were all the remaining ADR reports recorded in the database during the same period. Drug exposition was defined by the presence in the report of the drug coded “suspect” according to the World Health Organization (WHO) criteria, whatever the level of causality assessment. We selected drugs for which four or more reports of “depression” had been registered in the FPVD. If available, the following data were also collected from each report: age, sex, seriousness and evolution of the ADR. 2.3. Statistical analysis Collected data were compared between reports defined as depression (cases) and all other reports in the database (non cases). We calculated a ROR to compare risk of exposure to different drugs in cases and non cases. The RORs are given with their 95% CIs. The 95% CIs were calculated using the Woolf’s method.[7] A p value <0.05 was considered statistically significant. Statistical analyses were performed with Epi Info version 3.5.4.
3. Results Of the 114 692 reports recorded in the FPVD from the January 1st, 2007 to December 31st, 2011, the SMQ allowed to identify 1691 cases and we validated 474 cases of depression, representing 0.41% of all reports (114 218 were non cases). Among these cases, 255 (54%) were observed in women. Mean age was 44.6 years (median 44 years, range 5 – 93 years). The maximal
Thérapie 2015 Septembre-Octobre; 70 (5)
Drugs and Depression
427
Fig. 1. Repartition of the 474 cases of depression according to age.
number of cases occurred between 35-54 years (n = 105) [figure 1]. Most cases (56%) were considered as “non serious”. For the majority of the patients (64% of the cases), evolution was favorable with symptoms disappearing. However evolution was unknown for 44 patients (9% of the cases). In 6 cases, depression led to the death of the patients by suicide. The patients were treated by isotretinoin in 4 cases, an association of ropirinole and L-DOPA/benserazide in 1 report and an association of infliximab, cyamemazine, escitalopram and zolpidem for the last one. Among the 474 cases, 200 drugs were mentioned as “suspect”. In 378 observations, only one drug was defined as “suspect”. In 75 cases, two different drugs were involved with in 23 cases, the association interferon (IFN)+ribavirine. Eleven other case reports involved three suspected drugs and 10 other reports four or more suspected drugs. Significant RORs were found for 22 drugs mentioned in table I. Among them, the main therapeutic classes were antiepileptic medications (20 cases with topiramate and levetiracetam), anti-infective agents (76 cases) and especially anti-retroviral drugs (53 cases with efavirenz, emtricitabine, tenofovir, etravirine, raltegravir), IFNs (46 cases with IFN-alpha2b, PEGIFNalpha2a, PEGIFN-alpha2b, IFN-beta1a), synthetic retinoid medications (57 cases with isotretinoin), psycho-analeptic (4 with methylphenidate), and numerous other agents including sodium oxybate, varenicline, montelukast, flunarizine, adalimumab, anastrozole. It should be noted that an association was also found with 2 other medications, rimonabant and veralipride that have been withdrawn from the market. The repartition of age and sex of patients for each drug is described in table II. For most of them, the mean age was included between 35 and 54 years, range of ages where the maximum number of cases occurred. Nevertheless, the mean age was younger for methylphenidate, isotretinoin, montelukast, topiramate and was older for veralipride, anastrozole. Discrepancies were also observed for the sex repartition. Depression reports were more important in women treated with anastrozole, veralipride, sodium oxybate,
© 2015 Société Française de Pharmacologie et de Thérapeutique
topiramate, adalimumab, flunarizine. Cases in male population implicated anti-virals, isotretinoin, methylphenidate, montelukast.
4. Discussion The present study investigated the putative association between exposure to drugs and reports of depression using the case/non case approach based on individual reports of adverse drug reactions in the FPVD. In our study, depression reports were slightly higher in women (54%). This was in accordance with that is observed in general population in France where occurrence of depression was more elevated in women than in men.[8] The maximal number of cases occurred between 35 and 54 years. In France, in general population, the prevalence of depression is the highest between 20 and 34 years.[8] This difference could be explain that our studied population included only patients with a treatment and so probably older than general population. Concerning drugs associated with depression, our study suggests that agents from various therapeutic classes have been associated with depression: antiepileptic medications, anti-infective agents and especially anti-retroviral drugs, interferons (IFNs), synthetic retinoid medications, psycho-analeptic, and numerous other agents including sodium oxybate, varenicline, montelukast, flunarizine, adalimumab, anastrozole. The strongest association was found with veralipride, isotretinoin, flunarizine, varenicline, rimonabant. Repartitions of age and sex were different according to molecules and these discrepancies could be at least in part explained by the treated pathologies. Topiramate and levetiracetam are anticonvulsants that were significantly associated with depression in the present study. There are several possible causes of psychiatric disorders in patients with epilepsy and most antiepileptic drugs have been linked with the development of depression in a small percentage of patients with
Thérapie 2015 Septembre-Octobre; 70 (5)
428
Lafay-Chebassier et al.
Table I. Risk of exposure to drugs and occurrence of depression in the French pharmacovigilance database (FPVD). Frequency = very common (≥1/10), common (≥1/100 to <1/10), uncommon (≥1/1,000 to <1/100), rare (≥1/10,000 to <1/1,000) and very rare (<1/10,000).
Drugs
Number of Number of cases of depression in ADRs the FPVD in the FPVD
Antiepileptics
20
Levetiracetam
507
13
Topiramate
216
7
Antivirals
76
Mention in French summary of product characteristics
ROR (95% CI)
p
Depression (common)
6.5 (3.7, 11.3)
<0.001
Depression (very common)
8.2 (3.8, 17.5)
<0.001
Efavirenz
317
13
Depression (common)
10,6 (6.0, 18.6)
<0.001
Emtricitabine
660
15
No data
5.8 (3.4, 9.7)
<0.001
Etravirine
90
4
No data
11.3 (4.1, 30.9)
<0.001
6.8 (3.0, 15.3)
<0.001
15.4 (10.0, 23.8)
<0.001
No data
4.8 (2.9, 8.1)
<0.001
Depression (very common), suicidal thoughts (rare), suicide and suicide attempt (very rare)
12.1 (4.9, 29.7)
<0.001
Depression, suicide (undetermined frequency)
7.7 (2.8, 21.0)
<0.001
19.9 (13.3, 29.7)
<0.001
24.4 (12.6, 46.9)
<0.001
Raltegravir
222
6
Ribavirine
399
23
Tenofovir
788
15
Interferons
Suicide attempt, depression, mood disorders, suicidal thoughts, suicidal behaviour (particularly in patients with history of psychiatric disease) [uncommon] Depression (very common), mood disturbance (common), suicidal thoughts (uncommon), suicide (rare)
46
IFN-alpha2b
106
5
IFN-beta1a
130
4
PEGIFN-alpha2a
373
27
PEGIFNalpha2b
111
10
Others
Depression (very common), mood disorder (common), suicidal thoughts (uncommon), suicide (rare) Depression (very common), mood disorder (common), suicide, suicide attempt and suicidal thoughts (uncommon)
181
Adalimumab
629
6
Mood disorders (depression included) [common]
2.3 (1.0, 5.3)
<0.05
Anastrozole
183
4
Mood disorder (undetermined frequency)
5.4 (2.0, 14.7)
<0.001
Isotretinoin
298
57
Depression and aggravated depression (rare), suicidal thoughts, suicide attempt and suicide (very rare)
64.7 (47.7, 87.6)
<0.001
Flunarizine
20
4
Depressive syndrome (common)
60.8 (20.2, 182.4)
<0.001
10.0 (3.7, 27.3)
<0.001
8.9 (3.3, 24.3)
<0.001
33.6 (22.6, 50.1)
<0.001
13.5 (5.5, 33.4)
<0.001
34.8 (25.9, 46.9)
<0.001
Methylphenidate
101
4
Montelukast
111
4
Rimonabant
250
29
Sodium oxybate
95
5
Varenicline
484
55
Veralipride
32
13
Depression (common), suicidal thoughts and mood disturbance (uncommon), suicidal behaviour (suicide included) and transitory depressive state (very rare) Depression (uncommon), suicidal thoughts with suicide attempts (very rare) Depressive disorders and mood disorders (common), suicidal thoughts (uncommon) Depression (common), suicide attempt (uncommon), suicidal thoughts (undetermined frequency) Mood disorder and depression (uncommon), suicidal thoughts (undetermined frequency) Depressive syndrome with or without anxiety (very rare)
169.5 (83.2, 345.2) <0.001
ADRs: adverse drug reports; CI: confidence interval; FPVD: French pharmacovigilance database; ROR: reporting odds ratio.
© 2015 Société Française de Pharmacologie et de Thérapeutique
Thérapie 2015 Septembre-Octobre; 70 (5)
Drugs and Depression
429
Table II. Repartition of age and sex for each drug.
Number of cases of depression in the FPVD
Age mean, years
13
42.2
6 (46)
7
30.6
6 (86)
Efavirenz
13
46.8
2 (15)
Emtricitabine
Drugs
Female sex n(%)
Antiepileptics Levetiracetam Topiramate Antivirals
15
45.3
0 (0)
Etravirine
4
49.3
0 (0)
Raltegravir
6
47.7
3 (50)
Ribavirine
23
45.7
9 (39)
Tenofovir
15
45.3
0 (0)
IFN-alpha2b
5
53.2
2 (40)
IFN-beta1a
4
44.3
2 (50)
PEGIFN-alpha2a
27
43.4
11 (41)
PEGIFN-alpha2b
10
49.5
2 (20)
Adalimumab
6
46.7
5 (83)
Anastrozole
4
62.5
4 (100)
Isotretinoin
57
20.4
21 (39)
Flunarizine
4
49.8
3 (75)
Methylphenidate
4
17.5
1 (25)
Montelukast
4
28.8
0 (0)
Rimonabant
29
51.8
16 (57)
5
39.0
Varenicline
55
43.8
33 (60)
Veralipride
13
57.1
13 (100)
Interferons
Others
Sodium oxybate
5 (100)
FPVD: French pharmacovigilance database
predominately 3 molecules including topiramate.[9] This one induces depressive symptoms in up to 10% of patients, in a dosedependent manner, in case of rapid dose titration and in presence of risk factors while levetiracetam presents an intermediate risk, with prevalence of depression of about 4% or lower.[9] Antiretroviral medications including efavirenz, emtricitabine, tenofovir, etravirine, raltegravir were also associated with depres-
© 2015 Société Française de Pharmacologie et de Thérapeutique
sion in this study. Among them, the non-nucleoside reverse transcriptase inhibitor, efavirenz, is frequently associated with doserelated neuropsychiatric disorders as depressive symptoms. Plasma and intracellular drug concentrations and cytochrome P450 2B6 genotype have been found to be predictors of these early neuropsychological disturbances. The present study also found associations with etravirine, raltegravir, emtricitabine, tenofovir while depression
Thérapie 2015 Septembre-Octobre; 70 (5)
430
is an ADR only mentioned in the summary of product characteristics (SPC) of raltegravir. For the 4 drugs, no data are available in the literature about their potentially depressogenic properties. Among anti-infective agents, ribavirine was also associated with depression. We can postulate that this association is not linked to a depressive effect of ribavirine itself but is the result of the commonly combination of ribavirine with interferon-alpha known to be a depressogenic drug (as described further) in the treatment of chronic hepatitis C. In this way, ribavirine did not exhibit depression like-effect in mice.[10] IFNs were also reported in our study among drugs associated with depression. IFN-alpha is well known to induce a lot of neuropsychiatric side effects including depression in a substantial proportion of patients. However, psychiatric comorbidities are common with the patients using IFN-alpha. Concerning IFN-beta, there has been significant concern that it similarly causes depressive symptoms. Nevertheless, current data are contradictory and the pathology itself significantly increased risk for depression.[1] Among the other drugs with an association, we found isotretinoin and varenicline. Isotretinoin has been associated since its marketing with psychiatric side effects including depression and carries a black box warnings related to these risks. However, controversy exists concerning the risk of depression in treated patients.[1] Moreover, this drug is used during adolescence and puberty, a group of age particularly vulnerable to emotional disturbances. Concerning varenicline, warnings with regard to worsening mood and suicidal ideation were also carried since its introduction. However a connection is well established between smoking and depression and smoking cessation itself can precipitate depressive symptoms in patients with nicotine dependence. Moreover, this seems paradoxical if varenicline also has antidepressant effect in animal and human.[11,12] We also observed possible association with central nervous system (CNS) stimulant or depressant, like methylphenidate or sodium oxybate respectively. Psychiatric side effects such as depression are mentioned in the SPC of methylphenidate as a frequent adverse effect. However this molecule seems to have antidepressant properties linked to its amphetamine-like central nervous system stimulant action and may be prescribed for off-label use in treatment-resistant cases of depression in cancer.[13] Like methylphenidate, depression is an adverse effect mentioned in pharmacological information for sodium oxybate even though this molecule has antidepressant properties.[14] Other drugs like montelukast, flunarizine, adalimumab were also associated with depression, an adverse effect mentioned in their SPC. Concerning montelukast, the Food and Drug Administration (FDA) recently issued a warning raising concerns about the suicidality potential of this antiallergy medication. In France, depression is an adverse effect mentioned in its SPC with low frequency. Nevertheless, strong data also suggest a link between allergic disorders and depression.[15] For flunarizine, a non-selective calcium antagonist used in migraine prophylaxis, it should be noted
© 2015 Société Française de Pharmacologie et de Thérapeutique
Lafay-Chebassier et al.
that depression and migraine headaches frequently co-occur. Mood changes including depression are mentioned in the SPC of the tumor necrosis factor-α (TNF-α) antagonist, adalimumab, prescribed in patients with autoimmune diseases. The prevalence of depression in these patients is known to be higher than in the general population. Inflammatory cytokines certainly interact with virtually every pathophysiologic pathway relevant to depression. However, no other data are available in the literature concerning adalimumab induced-depression and even if the notion regarding the complete independence of the CNS from the peripheral immune system has changed,[16] the depressogenic effect of this agent is not easy to explain because it does not cross the blood–brain barrier. Moreover, in literature only limited and contradictory data exist with other TNF-α inhibitors. On one hand, psychiatric side effects including depression are reported as uncommon with infliximab treatment in the SPC. On the other hand, infliximab has been shown to decrease depression in rat model of chronic mild stress[17] and exerts antidepressant effects in human.[18] Anastrozole was also associated with depression in our study and is described to induce mood changes (without details) in the SPC. No data are available in the literature about this potentially depressogenic property in a disease strongly associated with depressive symptoms again. Finally, an association was also observed with 2 other medications, rimonabant and veralipride. Marketing authorisations of these 2 drugs have been withdrawn throughout Europe because their benefits do not outweigh their risks. Numerous cases of severe depression and suicidal thoughts were reported with rimonabant. In the same way, the use of veralipride was associated with side effects, including depression, anxiety, sleep disorders, tremor and tardive dyskinesia. For most of the drugs, the underlying mechanisms of these depressogenic effects are not well known. On the basis of animal and human literature, one or more neurochemical mechanisms were found to explain such symptoms: 1/ alterations in serotoninergic system for efavirenz,[19] IFN,[20] isotretinoin,[21] methylphenidate,[22] varenicline,[23] 2/ interference with cytokines and stress for efavirenz,[24] IFN,[25] isotretinoin,[26] 3/ inhibition of hippocampal neurogenesis for IFN,[27] isotretinoin,[28] 4/ alterations in cholinergic transmission for varenicline,[29] 5/ GABA modulation for antiepileptics,[30] sodium oxybate…[31] Nevertheless, the pathways by which these effects are generated remain complex. Some drugs, especially, IFN, isotretinoin, certainly interact with virtually more than one pathophysiologic pathway relevant to depression including together neurotransmitter metabolism, neuroendocrine function, synaptic plasticity… The associations found in this study concerned mainly expected drugs ie drugs for which depression was mentioned in the SPC. However, we did not find either some known associations like cardiovascular medications as beta-blockers, antimalarial as mefloquine for example. We also observed discordant data between frequencies in the SPC and calculated ROR. For example, depression
Thérapie 2015 Septembre-Octobre; 70 (5)
Drugs and Depression
is mentioned as rare in SPC of isotretinoin while we showed a strong association (ROR 64.7, 95% CI [47.7, 87.6]). It is also the case for varenicline: uncommon in the SPC and ROR 34.8, 95% CI (25.9, 46.9). These ascertainments can be partially explained by the method used. The case/non case method is a simple, quick tool for assessing and detecting associations between a specific adverse drug effect and exposure to drugs in real conditions of use. It presents several limitations.The first one is linked on the under-reporting of ADRs in general. Only a fraction of the real adverse effects of drugs are reported in pharmacovigilance. Thus, severe and/or unexpected ADRs linked to newly marketed drugs are more likely to be reported than others. Secondly, the context of use of the most of these drugs could also have influenced the results. Indeed, it is difficult to assess the involvement of drugs in the occurrence of depression because the pathology treated itself can precipitate depressive symptoms. In conclusion, taking into account the limits of the methodology, this study confirmed the association of depression with mainly expected drugs belonging to various therapeutic classes and used to treat common conditions. Nevertheless, it remains difficult to ascertain a drug as being a causative agent for depression because of the multifactorial pattern of the pathology and the fact that depression is common in patients with medical illness. This study should alert clinicians and caution and very careful monitoring are necessary for many drugs to early detect depressive symptoms, especially in patients with current or past depression where they should be carefully used. Author disclosures. None. Abbreviations. ADR: adverse drug report; ADRs: adverse drug reactions; CI: confidence interval; CNS: central nervous system; FDA: Food and Drug Administration; FPVD: French pharmacovigilance database; IFN: interferon; MedRA: Medical Dictionary for Regulatory Activities; ROR: reporting odds ratio; SPC: summary of product characteristics; WHO: World Health Organization.
References 1.
Celano CM, Freudenreich O, Fernandez-Robles C, et al. Depressogenic effects of medications: a review. Dialogues Clin Neurosci 2011; 13(1): 10925
431
in the French pharmacovigilance system database. Br J Clin Pharmacol 1997 Nov; 44(5): 513-8 7.
Woolf B. On estimating the relation between blood group and disease. Ann Hum Genet 1955 Jun; 19(4): 251-3
8.
Chan Chee C, Gourier-Frery C, Guignard R, et al. The current state of mental health surveillance in France. Sante Publique 2011 Nov-Dec; 23 Suppl 6: S13-29
9.
Mula M, Sander JW. Negative effects of antiepileptic drugs on mood in patients with epilepsy. Drug Saf 2007; 30(7): 555-67
10. Abdel-Salam OM. Antinociceptive and behavioral effects of ribavirin in mice. Pharmacol Biochem Behav 2006 Feb; 83(2): 230-8 11. Philip NS, Carpenter LL, Tyrka AR, et al. Varenicline augmentation in depressed smokers: an 8-week, open-label study. J Clin Psychiatry 2009 Jul; 70(7): 1026-31 12. Rollema H, Guanowsky V, Mineur YS, et al. Varenicline has antidepressantlike activity in the forced swim test and augments sertraline's effect. Eur J Pharmacol 2009 Mar 1; 605(1-3): 114-6 13. Ng CG, Boks MP, Roes KC, et al. Rapid response to methylphenidate as an add-on therapy to mirtazapine in the treatment of major depressive disorder in terminally ill cancer patients: a four-week, randomized, double-blinded, placebo-controlled study. Eur Neuropsychopharmacol 2014 Apr; 24(4): 491-8 14. Bosch OG, Quednow BB, Seifritz E, et al. Reconsidering GHB: orphan drug or new model antidepressant? J Psychopharmacol 2012 May; 26(5): 618-28 15. Goodwin RD, Kroenke K, Hoven CW, et al. Major depression, physical illness, and suicidal ideation in primary care. Psychosom Med 2003 JulAug; 65(4): 501-5 16. Banks WA, Erickson MA. The blood-brain barrier and immune function and dysfunction. Neurobiol Dis 2010 Jan; 37(1): 26-32 17. Karson A, Demirtas T, Bayramgurler D, et al. Chronic administration of infliximab (TNF-alpha inhibitor) decreases depression and anxiety-like behaviour in rat model of chronic mild stress. Basic Clin Pharmacol Toxicol 2013 May; 112(5): 335-40 18. Ertenli I, Ozer S, Kiraz S, et al. Infliximab, a TNF-alpha antagonist treatment in patients with ankylosing spondylitis: the impact on depression, anxiety and quality of life level. Rheumatol Int 2012 Feb; 32(2): 323-30 19. Zheve GT. Neuroprotective Mechanisms of Nevirapine and Efavirenz in Model of Neurodegeneration. Rhodes: Rhodes University; 2008 20. Schaefer M, Winterer J, Sarkar R, et al. Three cases of successful tryptophan add-on or monotherapy of hepatitis C and IFNalpha-associated mood disorders. Psychosomatics 2008 Sep-Oct; 49(5): 442-6 21. O'Reilly K, Bailey SJ, Lane MA. Retinoid-mediated regulation of mood: possible cellular mechanisms. Exp Biol Med (Maywood) 2008 Mar; 233(3): 251-8 22. Attar-Herzberg D, Apel A, Gang N, et al. The serotonin syndrome: initial misdiagnosis. Isr Med Assoc J 2009 Jun; 11(6): 367-70
2.
Kotlyar M, Dysken M, Adson DE. Update on drug-induced depression in the elderly. Am J Geriatr Pharmacother 2005 Dec; 3(4): 288-300
23. Betry C, Etievant A, Oosterhof C, et al. Role of 5-HT3 receptors in the antidepressant response. Pharmaceuticals 2011; 4(4): 603-29
3.
Moore N, Biour M, Paux G, et al. Adverse drug reaction monitoring: doing it the French way. Lancet 1985 Nov 9; 2(8463): 1056-8
4.
Brown EG, Wood L, Wood S. The medical dictionary for regulatory activities (MedDRA). Drug Saf 1999 Feb; 20(2): 109-17
24. O'Mahony SM, Myint AM, Steinbusch H, et al. Efavirenz induces depressive-like behaviour, increased stress response and changes in the immune response in rats. Neuroimmunomodulation 2005; 12(5): 293-8
5.
Favreliere S, Lafay-Chebassier C, Alkhidir F, et al. Drug-induced dementia: a case/non-case study in the French Pharmacovigilance database. Therapie 2007 Nov-Dec; 62(6): 507-11
6.
Moore N, Kreft-Jais C, Haramburu F, et al. Reports of hypoglycaemia associated with the use of ACE inhibitors and other drugs: a case/non-case study
© 2015 Société Française de Pharmacologie et de Thérapeutique
25. Capuron L, Neurauter G, Musselman DL, et al. Interferon-alpha-induced changes in tryptophan metabolism: relationship to depression and paroxetine treatment. Biol Psychiatry 2003 Nov 1; 54(9): 906-14 26. Chen XN, Meng QY, Bao AM, et al. The involvement of retinoic acid receptor-alpha in corticotropin-releasing hormone gene expression and affective disorders. Biol Psychiatry 2009 Nov 1; 66(9): 832-9
Thérapie 2015 Septembre-Octobre; 70 (5)
432
27. Kenis G, Prickaerts J, van Os J, et al. Depressive symptoms following interferon-alpha therapy: mediated by immune-induced reductions in brain-derived neurotrophic factor? Int J Neuropsychopharmacol 2011 Mar; 14(2): 247-53 28. Crandall J, Sakai Y, Zhang J, et al. 13-cis-retinoic acid suppresses hippocampal cell division and hippocampal-dependent learning in mice. Proc Natl Acad Sci U S A 2004 Apr 6; 101(14): 5111-6 29. Risch SC, Cohen RM, Janowsky DS, et al. Physostigmine induction of depressive symptomatology in normal human subjects. Psychiatry Res 1981 Feb; 4(1): 89-94
© 2015 Société Française de Pharmacologie et de Thérapeutique
Lafay-Chebassier et al.
30. Petty F. GABA and mood disorders: a brief review and hypothesis. J Affect Disord 1995 Aug 18; 34(4): 275-81 31. Ghose S, Winter MK, McCarson KE, et al. The GABAB receptor as a target for antidepressant drug action. Br J Pharmacol 2010 Jan; 162(1): 1-17 Correspondence and offprints: Claire Lafay-Chebassier, Service de Pharmacologie Clinique, Centre Régional de Pharmacovigilance et de Renseignement sur les Médicaments, CHU de Poitiers, 2 rue de la Milétrie, 86021 Poitiers, France. E-mail: [email protected]
Thérapie 2015 Septembre-Octobre; 70 (5)
PHARMACOEPIDEMIOLOGY
© 2015 Société Française de Pharmacologie et de Thérapeutique
Drug-induced Depression: a Case/Non Case Study in the French Pharmacovigilance Database Claire Lafay-Chebassier1, François Chavant1, Sylvie Favrelière1, Véronique Pizzoglio2, Marie-Christine Pérault-Pochat1 and the French Association of Regional Pharmacovigilance Centers 1 2
CHU de Poitiers, Service de Pharmacologie Clinique et Vigilances, Poitiers, France Agence Nationale de Sécurité du Médicament et des produits de santé (ANSM), Direction de la Surveillance, Saint-Denis, France
Text received September 30th, 2014 ; accepted January 27th, 2015
Keywords: depression; drug; adverse drug reaction; case/non case study
Abstract – Depression is a complex disorder with heterogeneous clinical anomalies whose neurobiological understanding still remains unclear. Medications have been implicated as potential causes of depression but for many of them, data are controversial. The present study aims to investigate association bet ween drugs and reports of depression. We used the case/non case method in the French pharmacovigilance database (FPVD) to identify drugs associated with depression. Cases were reports of depression in the FPVD between January 2007 and December 2011. Non cases were all other reports during the same period. Data were expressed as reporting odds ratio (ROR) with their 95% confidence interval. Of the 114 692 reports recorded in the FPVD during the studied period, we identified 474 cases of depression. For the majority of the patients, they were considered as “non serious” (56%) and evolution was favorable (64%). Significant RORs were found for antiepileptics (topiramate, levetiracetam), anti-infective and especially anti-retroviral drugs (efavirenz, emtricitabine, tenofovir, etravirine, raltegravir), interferons and other agents including isotretinoin, methylphenidate, sodium oxybate, varenicline, montelukast, flunarizine, adalimumab, anastrozole. Taking into account the limits of the methodology, the present study described associations with mainly expected drugs belonging to various therapeutic classes but it also found a signal with some anti-retrovirals. On the contrary, we did not find some assumed associations like cardiovascular medications, antimalarial. For most of the drugs, one or more mechanisms were found to explain these depressogenic effects on the basis of animal and human literature. Even if such associations need to be confirmed by further prospective studies, cautions are necessary for many drugs to early detect depressive symptoms.
Mots clés : dépression ; médicament ; effet indésirable ; étude cas/non cas
Résumé – Dépression induite par les médicaments : étude cas/non-cas dans la banque nationale de pharmacovigilance. La dépression est une pathologie complexe, aux signes cliniques hétérogènes et dont les mécanismes neurobiologiques restent encore mal connus. Les médicaments peuvent être à l’origine de dépression mais pour beaucoup d’entre eux, les données sont controversées. L’objectif de cette étude est d’évaluer l’association entre les médicaments et les notifications de dépression. Nous avons utilisé la méthode cas/non-cas dans la banque nationale de pharmacovigilance (BNPV) afin d’identifier les médicaments associés à la survenue de dépression. Les cas étaient les notifications de dépression enregistrées dans la BNPV entre janvier 2007 et décembre 2011. Les non-cas correspondaient à toutes les autres notifications enregistrées pendant la même période. Les données étaient exprimées par un reporting odds ratio (ROR) avec son intervalle de confiance à 95 %. Sur les 114 692 notifications enregistrées dans la BNPV au cours de la période d’étude, nous avons identifié 474 cas de dépression. Pour la majorité des patients, ils étaient considérés comme non graves (56 %) et l’évolution était favorable (64 %). Des ROR significatifs étaient retrouvés pour des anti-épileptiques (topiramate, lévétiracétam), des anti-infectieux et notamment des antirétroviraux (éfavirenz, emtricitabine, ténofovir, étravirine, raltégravir), des interférons et d’autres molécules incluant l’isotrétinoïne, le méthylphénidate, l’oxybate de sodium, la varénicline, le montélukast, la flunarizine, l’adalimumab et l’anastrozole. En tenant compte des limites liées à la méthodologie, cette étude a décrit des associations avec principalement des médicaments attendus appartenant à des classes thérapeutiques diverses mais il a également été retrouvé un signal avec certains antirétroviraux. A l’inverse, nous n’avons pas retrouvé les associations précédemment décrites avec des médicaments cardiovasculaires ou antipaludéens. Pour la plupart des médicaments, un ou plusieurs mécanismes ont été retrouvés pour expliquer ces effets dépressogènes. Même si ces associations nécessitent d’être confirmées par des études prospectives, des précautions sont nécessaires pour de nombreux médicaments afin de détecter précocement des symptômes dépressifs.
Abbreviations: see end of article.
Article publié par EDP Sciences
426
Lafay-Chebassier et al.
1. Introduction Mood disorders represent one of the most prevalent forms of mental illness. Depression is a complex heterogeneous disorder with a wide spectrum of anomalies including depressed mood, anhedonia, sleep disorders, fatigue, loss of self-esteem, negative thinking and suicidality. Functional brain imaging and neurobiological techniques have revealed specific functional and structural brain abnormalities but the neurobiological understanding of depression still remains unclear. Genetic and environmental factors such as stress and emotional trauma, viral infections, medical conditions have been implicated in depressive symptoms. Medications may also cause neuropsychiatric symptoms and, in literature, a wide variety of drugs have been implicated as potential causes of new-onset depression or worsening of established depression.[1] Unfortunately, contradictory data have been reported for many agents leading to the lack of conclusive data.[1,2] Indeed, it can be difficult to determine whether the suspected drug is truly causing the depressive symptoms or whether its use was concomitant with an endogenous depression on the basis of case reports. However, identification of drugs associated with depressive symptoms is important because untreated depression can evolve into suicidal ideation and suicide attempts and the lack of adherence to therapy is often closely related to the central adverse effects caused by drugs. Thus the aim of the present study was to evaluate association between drugs exposure and reports of depression using the case/ non case method in the French pharmacovigilance database.
2. Methods The data source used is the French pharmacovigilance database (FPVD) that has been previously described.[3] Briefly, the French pharmacovigilance system was created in 1973 and, since 1985, all spontaneous reports of adverse drug reactions (ADRs) have been recorded into the FPVD. According to French law, every health professional must report “serious” and/or “unexpected” adverse drug reactions to their Regional Center of Pharmacovigilance (31 centers in France). “Serious” ADRs are defined as reactions resulting in death, life-threatening event, hospitalization (or prolongation of existing hospitalization), persistent or significant disability or incapacity, congenital abnormalities/birth defect or other significant medical event. ADRs are defined as “unexpected” if their nature, seriousness or evolution is not consistent with data given by the drug monograph. For each ADR report, information about the patient (age, sex, medical history), drug exposure (to the suspected drug and other associated non-suspected drugs) and ADR characteristics (“serious” or “non-serious”, “expected” or “unexpected”, causality score, outcome) are recorded in the FPVD. ADRs are coded according to the Medical Dictionary for Regulatory Activities (MedDRA, version 15.1).[4]
© 2015 Société Française de Pharmacologie et de Thérapeutique
2.1. Case/non case method The case/non case method measures the disproportionality of combination between a drug and a particular adverse drug report (ADR) in a pharmacovigilance database.[5] This method can be used to generate signals from a pharmacovigilance database. Cases are defined as reports of the ADR of interest and non cases as all other reports of ADR. For each drug of interest, the association with the ADR was assessed by calculating an ADR reporting odds ratio (ROR) with its 95% confidence interval (CI). The ROR is the ratio of reporting of one specific event versus all other events for a given drug compound.[6] 2.2. Selection of cases and non cases In our study, we collected ADRs notified between January 1st, 2007 and December 31st, 2011. Cases were identified using standardised MedDRA queries (SMQ) “Depression (excl. suicide and self injury)”. After a careful analysis of the comments of spontaneous reports, we excluded all cases that were not “depression”. Non cases, using as controls, were all the remaining ADR reports recorded in the database during the same period. Drug exposition was defined by the presence in the report of the drug coded “suspect” according to the World Health Organization (WHO) criteria, whatever the level of causality assessment. We selected drugs for which four or more reports of “depression” had been registered in the FPVD. If available, the following data were also collected from each report: age, sex, seriousness and evolution of the ADR. 2.3. Statistical analysis Collected data were compared between reports defined as depression (cases) and all other reports in the database (non cases). We calculated a ROR to compare risk of exposure to different drugs in cases and non cases. The RORs are given with their 95% CIs. The 95% CIs were calculated using the Woolf’s method.[7] A p value <0.05 was considered statistically significant. Statistical analyses were performed with Epi Info version 3.5.4.
3. Results Of the 114 692 reports recorded in the FPVD from the January 1st, 2007 to December 31st, 2011, the SMQ allowed to identify 1691 cases and we validated 474 cases of depression, representing 0.41% of all reports (114 218 were non cases). Among these cases, 255 (54%) were observed in women. Mean age was 44.6 years (median 44 years, range 5 – 93 years). The maximal
Thérapie 2015 Septembre-Octobre; 70 (5)
Drugs and Depression
427
Fig. 1. Repartition of the 474 cases of depression according to age.
number of cases occurred between 35-54 years (n = 105) [figure 1]. Most cases (56%) were considered as “non serious”. For the majority of the patients (64% of the cases), evolution was favorable with symptoms disappearing. However evolution was unknown for 44 patients (9% of the cases). In 6 cases, depression led to the death of the patients by suicide. The patients were treated by isotretinoin in 4 cases, an association of ropirinole and L-DOPA/benserazide in 1 report and an association of infliximab, cyamemazine, escitalopram and zolpidem for the last one. Among the 474 cases, 200 drugs were mentioned as “suspect”. In 378 observations, only one drug was defined as “suspect”. In 75 cases, two different drugs were involved with in 23 cases, the association interferon (IFN)+ribavirine. Eleven other case reports involved three suspected drugs and 10 other reports four or more suspected drugs. Significant RORs were found for 22 drugs mentioned in table I. Among them, the main therapeutic classes were antiepileptic medications (20 cases with topiramate and levetiracetam), anti-infective agents (76 cases) and especially anti-retroviral drugs (53 cases with efavirenz, emtricitabine, tenofovir, etravirine, raltegravir), IFNs (46 cases with IFN-alpha2b, PEGIFNalpha2a, PEGIFN-alpha2b, IFN-beta1a), synthetic retinoid medications (57 cases with isotretinoin), psycho-analeptic (4 with methylphenidate), and numerous other agents including sodium oxybate, varenicline, montelukast, flunarizine, adalimumab, anastrozole. It should be noted that an association was also found with 2 other medications, rimonabant and veralipride that have been withdrawn from the market. The repartition of age and sex of patients for each drug is described in table II. For most of them, the mean age was included between 35 and 54 years, range of ages where the maximum number of cases occurred. Nevertheless, the mean age was younger for methylphenidate, isotretinoin, montelukast, topiramate and was older for veralipride, anastrozole. Discrepancies were also observed for the sex repartition. Depression reports were more important in women treated with anastrozole, veralipride, sodium oxybate,
© 2015 Société Française de Pharmacologie et de Thérapeutique
topiramate, adalimumab, flunarizine. Cases in male population implicated anti-virals, isotretinoin, methylphenidate, montelukast.
4. Discussion The present study investigated the putative association between exposure to drugs and reports of depression using the case/non case approach based on individual reports of adverse drug reactions in the FPVD. In our study, depression reports were slightly higher in women (54%). This was in accordance with that is observed in general population in France where occurrence of depression was more elevated in women than in men.[8] The maximal number of cases occurred between 35 and 54 years. In France, in general population, the prevalence of depression is the highest between 20 and 34 years.[8] This difference could be explain that our studied population included only patients with a treatment and so probably older than general population. Concerning drugs associated with depression, our study suggests that agents from various therapeutic classes have been associated with depression: antiepileptic medications, anti-infective agents and especially anti-retroviral drugs, interferons (IFNs), synthetic retinoid medications, psycho-analeptic, and numerous other agents including sodium oxybate, varenicline, montelukast, flunarizine, adalimumab, anastrozole. The strongest association was found with veralipride, isotretinoin, flunarizine, varenicline, rimonabant. Repartitions of age and sex were different according to molecules and these discrepancies could be at least in part explained by the treated pathologies. Topiramate and levetiracetam are anticonvulsants that were significantly associated with depression in the present study. There are several possible causes of psychiatric disorders in patients with epilepsy and most antiepileptic drugs have been linked with the development of depression in a small percentage of patients with
Thérapie 2015 Septembre-Octobre; 70 (5)
428
Lafay-Chebassier et al.
Table I. Risk of exposure to drugs and occurrence of depression in the French pharmacovigilance database (FPVD). Frequency = very common (≥1/10), common (≥1/100 to <1/10), uncommon (≥1/1,000 to <1/100), rare (≥1/10,000 to <1/1,000) and very rare (<1/10,000).
Drugs
Number of Number of cases of depression in ADRs the FPVD in the FPVD
Antiepileptics
20
Levetiracetam
507
13
Topiramate
216
7
Antivirals
76
Mention in French summary of product characteristics
ROR (95% CI)
p
Depression (common)
6.5 (3.7, 11.3)
<0.001
Depression (very common)
8.2 (3.8, 17.5)
<0.001
Efavirenz
317
13
Depression (common)
10,6 (6.0, 18.6)
<0.001
Emtricitabine
660
15
No data
5.8 (3.4, 9.7)
<0.001
Etravirine
90
4
No data
11.3 (4.1, 30.9)
<0.001
6.8 (3.0, 15.3)
<0.001
15.4 (10.0, 23.8)
<0.001
No data
4.8 (2.9, 8.1)
<0.001
Depression (very common), suicidal thoughts (rare), suicide and suicide attempt (very rare)
12.1 (4.9, 29.7)
<0.001
Depression, suicide (undetermined frequency)
7.7 (2.8, 21.0)
<0.001
19.9 (13.3, 29.7)
<0.001
24.4 (12.6, 46.9)
<0.001
Raltegravir
222
6
Ribavirine
399
23
Tenofovir
788
15
Interferons
Suicide attempt, depression, mood disorders, suicidal thoughts, suicidal behaviour (particularly in patients with history of psychiatric disease) [uncommon] Depression (very common), mood disturbance (common), suicidal thoughts (uncommon), suicide (rare)
46
IFN-alpha2b
106
5
IFN-beta1a
130
4
PEGIFN-alpha2a
373
27
PEGIFNalpha2b
111
10
Others
Depression (very common), mood disorder (common), suicidal thoughts (uncommon), suicide (rare) Depression (very common), mood disorder (common), suicide, suicide attempt and suicidal thoughts (uncommon)
181
Adalimumab
629
6
Mood disorders (depression included) [common]
2.3 (1.0, 5.3)
<0.05
Anastrozole
183
4
Mood disorder (undetermined frequency)
5.4 (2.0, 14.7)
<0.001
Isotretinoin
298
57
Depression and aggravated depression (rare), suicidal thoughts, suicide attempt and suicide (very rare)
64.7 (47.7, 87.6)
<0.001
Flunarizine
20
4
Depressive syndrome (common)
60.8 (20.2, 182.4)
<0.001
10.0 (3.7, 27.3)
<0.001
8.9 (3.3, 24.3)
<0.001
33.6 (22.6, 50.1)
<0.001
13.5 (5.5, 33.4)
<0.001
34.8 (25.9, 46.9)
<0.001
Methylphenidate
101
4
Montelukast
111
4
Rimonabant
250
29
Sodium oxybate
95
5
Varenicline
484
55
Veralipride
32
13
Depression (common), suicidal thoughts and mood disturbance (uncommon), suicidal behaviour (suicide included) and transitory depressive state (very rare) Depression (uncommon), suicidal thoughts with suicide attempts (very rare) Depressive disorders and mood disorders (common), suicidal thoughts (uncommon) Depression (common), suicide attempt (uncommon), suicidal thoughts (undetermined frequency) Mood disorder and depression (uncommon), suicidal thoughts (undetermined frequency) Depressive syndrome with or without anxiety (very rare)
169.5 (83.2, 345.2) <0.001
ADRs: adverse drug reports; CI: confidence interval; FPVD: French pharmacovigilance database; ROR: reporting odds ratio.
© 2015 Société Française de Pharmacologie et de Thérapeutique
Thérapie 2015 Septembre-Octobre; 70 (5)
Drugs and Depression
429
Table II. Repartition of age and sex for each drug.
Number of cases of depression in the FPVD
Age mean, years
13
42.2
6 (46)
7
30.6
6 (86)
Efavirenz
13
46.8
2 (15)
Emtricitabine
Drugs
Female sex n(%)
Antiepileptics Levetiracetam Topiramate Antivirals
15
45.3
0 (0)
Etravirine
4
49.3
0 (0)
Raltegravir
6
47.7
3 (50)
Ribavirine
23
45.7
9 (39)
Tenofovir
15
45.3
0 (0)
IFN-alpha2b
5
53.2
2 (40)
IFN-beta1a
4
44.3
2 (50)
PEGIFN-alpha2a
27
43.4
11 (41)
PEGIFN-alpha2b
10
49.5
2 (20)
Adalimumab
6
46.7
5 (83)
Anastrozole
4
62.5
4 (100)
Isotretinoin
57
20.4
21 (39)
Flunarizine
4
49.8
3 (75)
Methylphenidate
4
17.5
1 (25)
Montelukast
4
28.8
0 (0)
Rimonabant
29
51.8
16 (57)
5
39.0
Varenicline
55
43.8
33 (60)
Veralipride
13
57.1
13 (100)
Interferons
Others
Sodium oxybate
5 (100)
FPVD: French pharmacovigilance database
predominately 3 molecules including topiramate.[9] This one induces depressive symptoms in up to 10% of patients, in a dosedependent manner, in case of rapid dose titration and in presence of risk factors while levetiracetam presents an intermediate risk, with prevalence of depression of about 4% or lower.[9] Antiretroviral medications including efavirenz, emtricitabine, tenofovir, etravirine, raltegravir were also associated with depres-
© 2015 Société Française de Pharmacologie et de Thérapeutique
sion in this study. Among them, the non-nucleoside reverse transcriptase inhibitor, efavirenz, is frequently associated with doserelated neuropsychiatric disorders as depressive symptoms. Plasma and intracellular drug concentrations and cytochrome P450 2B6 genotype have been found to be predictors of these early neuropsychological disturbances. The present study also found associations with etravirine, raltegravir, emtricitabine, tenofovir while depression
Thérapie 2015 Septembre-Octobre; 70 (5)
430
is an ADR only mentioned in the summary of product characteristics (SPC) of raltegravir. For the 4 drugs, no data are available in the literature about their potentially depressogenic properties. Among anti-infective agents, ribavirine was also associated with depression. We can postulate that this association is not linked to a depressive effect of ribavirine itself but is the result of the commonly combination of ribavirine with interferon-alpha known to be a depressogenic drug (as described further) in the treatment of chronic hepatitis C. In this way, ribavirine did not exhibit depression like-effect in mice.[10] IFNs were also reported in our study among drugs associated with depression. IFN-alpha is well known to induce a lot of neuropsychiatric side effects including depression in a substantial proportion of patients. However, psychiatric comorbidities are common with the patients using IFN-alpha. Concerning IFN-beta, there has been significant concern that it similarly causes depressive symptoms. Nevertheless, current data are contradictory and the pathology itself significantly increased risk for depression.[1] Among the other drugs with an association, we found isotretinoin and varenicline. Isotretinoin has been associated since its marketing with psychiatric side effects including depression and carries a black box warnings related to these risks. However, controversy exists concerning the risk of depression in treated patients.[1] Moreover, this drug is used during adolescence and puberty, a group of age particularly vulnerable to emotional disturbances. Concerning varenicline, warnings with regard to worsening mood and suicidal ideation were also carried since its introduction. However a connection is well established between smoking and depression and smoking cessation itself can precipitate depressive symptoms in patients with nicotine dependence. Moreover, this seems paradoxical if varenicline also has antidepressant effect in animal and human.[11,12] We also observed possible association with central nervous system (CNS) stimulant or depressant, like methylphenidate or sodium oxybate respectively. Psychiatric side effects such as depression are mentioned in the SPC of methylphenidate as a frequent adverse effect. However this molecule seems to have antidepressant properties linked to its amphetamine-like central nervous system stimulant action and may be prescribed for off-label use in treatment-resistant cases of depression in cancer.[13] Like methylphenidate, depression is an adverse effect mentioned in pharmacological information for sodium oxybate even though this molecule has antidepressant properties.[14] Other drugs like montelukast, flunarizine, adalimumab were also associated with depression, an adverse effect mentioned in their SPC. Concerning montelukast, the Food and Drug Administration (FDA) recently issued a warning raising concerns about the suicidality potential of this antiallergy medication. In France, depression is an adverse effect mentioned in its SPC with low frequency. Nevertheless, strong data also suggest a link between allergic disorders and depression.[15] For flunarizine, a non-selective calcium antagonist used in migraine prophylaxis, it should be noted
© 2015 Société Française de Pharmacologie et de Thérapeutique
Lafay-Chebassier et al.
that depression and migraine headaches frequently co-occur. Mood changes including depression are mentioned in the SPC of the tumor necrosis factor-α (TNF-α) antagonist, adalimumab, prescribed in patients with autoimmune diseases. The prevalence of depression in these patients is known to be higher than in the general population. Inflammatory cytokines certainly interact with virtually every pathophysiologic pathway relevant to depression. However, no other data are available in the literature concerning adalimumab induced-depression and even if the notion regarding the complete independence of the CNS from the peripheral immune system has changed,[16] the depressogenic effect of this agent is not easy to explain because it does not cross the blood–brain barrier. Moreover, in literature only limited and contradictory data exist with other TNF-α inhibitors. On one hand, psychiatric side effects including depression are reported as uncommon with infliximab treatment in the SPC. On the other hand, infliximab has been shown to decrease depression in rat model of chronic mild stress[17] and exerts antidepressant effects in human.[18] Anastrozole was also associated with depression in our study and is described to induce mood changes (without details) in the SPC. No data are available in the literature about this potentially depressogenic property in a disease strongly associated with depressive symptoms again. Finally, an association was also observed with 2 other medications, rimonabant and veralipride. Marketing authorisations of these 2 drugs have been withdrawn throughout Europe because their benefits do not outweigh their risks. Numerous cases of severe depression and suicidal thoughts were reported with rimonabant. In the same way, the use of veralipride was associated with side effects, including depression, anxiety, sleep disorders, tremor and tardive dyskinesia. For most of the drugs, the underlying mechanisms of these depressogenic effects are not well known. On the basis of animal and human literature, one or more neurochemical mechanisms were found to explain such symptoms: 1/ alterations in serotoninergic system for efavirenz,[19] IFN,[20] isotretinoin,[21] methylphenidate,[22] varenicline,[23] 2/ interference with cytokines and stress for efavirenz,[24] IFN,[25] isotretinoin,[26] 3/ inhibition of hippocampal neurogenesis for IFN,[27] isotretinoin,[28] 4/ alterations in cholinergic transmission for varenicline,[29] 5/ GABA modulation for antiepileptics,[30] sodium oxybate…[31] Nevertheless, the pathways by which these effects are generated remain complex. Some drugs, especially, IFN, isotretinoin, certainly interact with virtually more than one pathophysiologic pathway relevant to depression including together neurotransmitter metabolism, neuroendocrine function, synaptic plasticity… The associations found in this study concerned mainly expected drugs ie drugs for which depression was mentioned in the SPC. However, we did not find either some known associations like cardiovascular medications as beta-blockers, antimalarial as mefloquine for example. We also observed discordant data between frequencies in the SPC and calculated ROR. For example, depression
Thérapie 2015 Septembre-Octobre; 70 (5)
Drugs and Depression
is mentioned as rare in SPC of isotretinoin while we showed a strong association (ROR 64.7, 95% CI [47.7, 87.6]). It is also the case for varenicline: uncommon in the SPC and ROR 34.8, 95% CI (25.9, 46.9). These ascertainments can be partially explained by the method used. The case/non case method is a simple, quick tool for assessing and detecting associations between a specific adverse drug effect and exposure to drugs in real conditions of use. It presents several limitations.The first one is linked on the under-reporting of ADRs in general. Only a fraction of the real adverse effects of drugs are reported in pharmacovigilance. Thus, severe and/or unexpected ADRs linked to newly marketed drugs are more likely to be reported than others. Secondly, the context of use of the most of these drugs could also have influenced the results. Indeed, it is difficult to assess the involvement of drugs in the occurrence of depression because the pathology treated itself can precipitate depressive symptoms. In conclusion, taking into account the limits of the methodology, this study confirmed the association of depression with mainly expected drugs belonging to various therapeutic classes and used to treat common conditions. Nevertheless, it remains difficult to ascertain a drug as being a causative agent for depression because of the multifactorial pattern of the pathology and the fact that depression is common in patients with medical illness. This study should alert clinicians and caution and very careful monitoring are necessary for many drugs to early detect depressive symptoms, especially in patients with current or past depression where they should be carefully used. Author disclosures. None. Abbreviations. ADR: adverse drug report; ADRs: adverse drug reactions; CI: confidence interval; CNS: central nervous system; FDA: Food and Drug Administration; FPVD: French pharmacovigilance database; IFN: interferon; MedRA: Medical Dictionary for Regulatory Activities; ROR: reporting odds ratio; SPC: summary of product characteristics; WHO: World Health Organization.
References 1.
Celano CM, Freudenreich O, Fernandez-Robles C, et al. Depressogenic effects of medications: a review. Dialogues Clin Neurosci 2011; 13(1): 10925
431
in the French pharmacovigilance system database. Br J Clin Pharmacol 1997 Nov; 44(5): 513-8 7.
Woolf B. On estimating the relation between blood group and disease. Ann Hum Genet 1955 Jun; 19(4): 251-3
8.
Chan Chee C, Gourier-Frery C, Guignard R, et al. The current state of mental health surveillance in France. Sante Publique 2011 Nov-Dec; 23 Suppl 6: S13-29
9.
Mula M, Sander JW. Negative effects of antiepileptic drugs on mood in patients with epilepsy. Drug Saf 2007; 30(7): 555-67
10. Abdel-Salam OM. Antinociceptive and behavioral effects of ribavirin in mice. Pharmacol Biochem Behav 2006 Feb; 83(2): 230-8 11. Philip NS, Carpenter LL, Tyrka AR, et al. Varenicline augmentation in depressed smokers: an 8-week, open-label study. J Clin Psychiatry 2009 Jul; 70(7): 1026-31 12. Rollema H, Guanowsky V, Mineur YS, et al. Varenicline has antidepressantlike activity in the forced swim test and augments sertraline's effect. Eur J Pharmacol 2009 Mar 1; 605(1-3): 114-6 13. Ng CG, Boks MP, Roes KC, et al. Rapid response to methylphenidate as an add-on therapy to mirtazapine in the treatment of major depressive disorder in terminally ill cancer patients: a four-week, randomized, double-blinded, placebo-controlled study. Eur Neuropsychopharmacol 2014 Apr; 24(4): 491-8 14. Bosch OG, Quednow BB, Seifritz E, et al. Reconsidering GHB: orphan drug or new model antidepressant? J Psychopharmacol 2012 May; 26(5): 618-28 15. Goodwin RD, Kroenke K, Hoven CW, et al. Major depression, physical illness, and suicidal ideation in primary care. Psychosom Med 2003 JulAug; 65(4): 501-5 16. Banks WA, Erickson MA. The blood-brain barrier and immune function and dysfunction. Neurobiol Dis 2010 Jan; 37(1): 26-32 17. Karson A, Demirtas T, Bayramgurler D, et al. Chronic administration of infliximab (TNF-alpha inhibitor) decreases depression and anxiety-like behaviour in rat model of chronic mild stress. Basic Clin Pharmacol Toxicol 2013 May; 112(5): 335-40 18. Ertenli I, Ozer S, Kiraz S, et al. Infliximab, a TNF-alpha antagonist treatment in patients with ankylosing spondylitis: the impact on depression, anxiety and quality of life level. Rheumatol Int 2012 Feb; 32(2): 323-30 19. Zheve GT. Neuroprotective Mechanisms of Nevirapine and Efavirenz in Model of Neurodegeneration. Rhodes: Rhodes University; 2008 20. Schaefer M, Winterer J, Sarkar R, et al. Three cases of successful tryptophan add-on or monotherapy of hepatitis C and IFNalpha-associated mood disorders. Psychosomatics 2008 Sep-Oct; 49(5): 442-6 21. O'Reilly K, Bailey SJ, Lane MA. Retinoid-mediated regulation of mood: possible cellular mechanisms. Exp Biol Med (Maywood) 2008 Mar; 233(3): 251-8 22. Attar-Herzberg D, Apel A, Gang N, et al. The serotonin syndrome: initial misdiagnosis. Isr Med Assoc J 2009 Jun; 11(6): 367-70
2.
Kotlyar M, Dysken M, Adson DE. Update on drug-induced depression in the elderly. Am J Geriatr Pharmacother 2005 Dec; 3(4): 288-300
23. Betry C, Etievant A, Oosterhof C, et al. Role of 5-HT3 receptors in the antidepressant response. Pharmaceuticals 2011; 4(4): 603-29
3.
Moore N, Biour M, Paux G, et al. Adverse drug reaction monitoring: doing it the French way. Lancet 1985 Nov 9; 2(8463): 1056-8
4.
Brown EG, Wood L, Wood S. The medical dictionary for regulatory activities (MedDRA). Drug Saf 1999 Feb; 20(2): 109-17
24. O'Mahony SM, Myint AM, Steinbusch H, et al. Efavirenz induces depressive-like behaviour, increased stress response and changes in the immune response in rats. Neuroimmunomodulation 2005; 12(5): 293-8
5.
Favreliere S, Lafay-Chebassier C, Alkhidir F, et al. Drug-induced dementia: a case/non-case study in the French Pharmacovigilance database. Therapie 2007 Nov-Dec; 62(6): 507-11
6.
Moore N, Kreft-Jais C, Haramburu F, et al. Reports of hypoglycaemia associated with the use of ACE inhibitors and other drugs: a case/non-case study
© 2015 Société Française de Pharmacologie et de Thérapeutique
25. Capuron L, Neurauter G, Musselman DL, et al. Interferon-alpha-induced changes in tryptophan metabolism: relationship to depression and paroxetine treatment. Biol Psychiatry 2003 Nov 1; 54(9): 906-14 26. Chen XN, Meng QY, Bao AM, et al. The involvement of retinoic acid receptor-alpha in corticotropin-releasing hormone gene expression and affective disorders. Biol Psychiatry 2009 Nov 1; 66(9): 832-9
Thérapie 2015 Septembre-Octobre; 70 (5)
432
27. Kenis G, Prickaerts J, van Os J, et al. Depressive symptoms following interferon-alpha therapy: mediated by immune-induced reductions in brain-derived neurotrophic factor? Int J Neuropsychopharmacol 2011 Mar; 14(2): 247-53 28. Crandall J, Sakai Y, Zhang J, et al. 13-cis-retinoic acid suppresses hippocampal cell division and hippocampal-dependent learning in mice. Proc Natl Acad Sci U S A 2004 Apr 6; 101(14): 5111-6 29. Risch SC, Cohen RM, Janowsky DS, et al. Physostigmine induction of depressive symptomatology in normal human subjects. Psychiatry Res 1981 Feb; 4(1): 89-94
© 2015 Société Française de Pharmacologie et de Thérapeutique
Lafay-Chebassier et al.
30. Petty F. GABA and mood disorders: a brief review and hypothesis. J Affect Disord 1995 Aug 18; 34(4): 275-81 31. Ghose S, Winter MK, McCarson KE, et al. The GABAB receptor as a target for antidepressant drug action. Br J Pharmacol 2010 Jan; 162(1): 1-17 Correspondence and offprints: Claire Lafay-Chebassier, Service de Pharmacologie Clinique, Centre Régional de Pharmacovigilance et de Renseignement sur les Médicaments, CHU de Poitiers, 2 rue de la Milétrie, 86021 Poitiers, France. E-mail: [email protected]
Thérapie 2015 Septembre-Octobre; 70 (5)