Primary Care Psychiatry, 2000 6:23–28 © LibraPharm Limited Emergence of induced suicidality

David Healy In the course of a randomised double-blind crossover study comparing the effects of reboxetine and sertraline in a group of healthy volunteers, two volunteers became suicidal on sertraline. This paper describes the charac- North Wales Department of teristics of the reactions experienced by both subjects. These problems Psychological Medicine were associated with a combination of akathisia and disinhibition. Hergest Unit Dysphoric or akathisic responses on their own to either did not lead Bangor LL57 2PW, UK to suicidality in this group of subjects. Primary Care Psychiatry 2000; 6:23–28, Copyright © 2000 by LibraPharm Limited Received 22 January 2000; accepted 14 March 2000

Keywords Reboxetine Sertraline Suicide

Introduction nomenon in general cleared up on Following the release of fluox- discontinuation of treatment and etine a number of other SSRIs In 1990, Teicher and colleagues re-emerged in a number of cases on appeared on the market including reported on the emergence of suici- re-exposure to the original treat- sertraline and paroxetine. It seems dality on fluoxetine in a group of six ment. In cases where it re-emerged clear that in general these are patients {1}. These reports were fol- there were reports that agents, associated with common profiles of lowed-up by reports from King et al. which theoretically might block the both main effects and side effects. All {2}, Creaney et al. {3}, Rothschild and appearance of a 5HT mediated prob- have received licenses for a similar Locke {4} and Wirshing et al. {5}, lem, were able to minimise or block set of nervous conditions. All pro- among others, reporting other cases the emergence of suicidality on re- duced a set of side effects including where suicidality appeared to emerge exposure. extra-pyramidal side effects, which in individuals taking fluoxetine. Nevertheless, despite a num- are not generally seen with other There were a number of ber of indicators for very clear causal {6}. Both sertraline factors associated with these case linkage between the intake of fluoxe- and paroxetine have been associated reports that argued for a strong tine and suicidality, there were also a with reports of akathisia {6,7,8,9,10, causal connection. In general there number of aspects of the reports, 11,12,13}. was consistency across the reports as which argued for some caution in On this basis there would to the time of onset of the problems the interpretation of what was hap- seem therefore to be a possibility following intake of fluoxetine. There pening. One was that some but not that other SSRIs might similarly appeared to be a dose response all of the patients were on other induce suicidality. A meta-analysis of relationship with problems liable to . Another was that some trials involving the SNRI milnacipran emerge on a higher dose. There was but not all had lengthy psychiatric compared with SSRIs showed a sig- some agreement as regards the histories with evidence of personal- nificantly increased rate of suicidality probable mechanism leading to the ity problems. It was, in general, less on treatment with SSRIs {14}. A ran- difficulties. This was termed than clear what the description of domised placebo control trial of akathisia, although whether it was this problem emerging in tertiary paroxetine in recurrent major quite the same phenomeno n as care centres might mean for the depression showed a higher rate of akathisia traditionally associated with wider world of primary care antide- suicide attempts on paroxetine than neuroleptics was less clear. The phe- pressant prescribing. on placebo in this group of patients 23 Primary Care Psychiatry 2000, Vol 6 No 1 Emergence of antidepressant induced suicidality

{15}. These data supplement unpub- dose of the drugs was either 4mgs of When on reboxetine, two sub- lished RCT data from Lilly suggesting reboxetine for the first five days of jects reported becoming depressed. a significantly higher rate of suicide the reboxetine arm with an option to In both cases they likened the expe- attempts in patients taking fluoxe- increase to 4mgs bd if tolerated or rience to post-childbirth baby-blues. tine compared with placebo or other sertraline 50mgs for the first five days Neither of these two subjects or any non-SSRI antidepressants {16,17}. of the sertraline period with an other subject while taking reboxe- Against this background we option to increase to 50mgs bd if tol- tine had suicidal ideation. In con- report the findings from a double- erated. trast, two subjects taking sertraline blind randomised crossover trial of At baseline, subjects com- developed clear suicidal ideation, sertraline and reboxetine in healthy pleted a Karolinska Personality one of which reached extremely seri- volunteers. This was aimed at explor- Questionnaire {18}, a Tridimensional ous proportions. Both these individ- ing modes of action of antidepres- Personality Questionnaire {19}, a uals appear to have had elements of sant drugs on levels of wellbeing and Profile of Mood States (POMS) {20}, a akathisia and emotional blunting but in particular the serenic effect that Positive and Negative Affect Scale other subjects had either akathisia or appears associated with the use of (PANAS) {21}, a Social Adaptation Self emotional blunting without becom- SSRIs which may mediate their ther- Evaluation Scales (SASS) {22} and a ing suicidal. W e report on these apeutic effect. BIS-BAS scale {23}. The effects of both cases in more detail, reconstructing drugs in the whole group on mood events from the diaries kept by both scales, personality inventories, side subjects. Methods effect rating scales and other mea- sures will be reported elsewhere Twenty healthy volunteers aged POMS, PANAS and SASS scales Case A between 28 and 52, with a mean age were completed on a daily basis as of 37.8 years, were recruited to a well as a daily diary of impressions of The first case was a 30 year old study comparing reboxetine with the functional and physical effects of woman. She was randomise d to sertraline on a range of personality, each drug. V olunteers were actively reboxetine initially. This made her self-report and quality of life mea- encouraged to consult their partners conspicuously relaxed and slightly sures. The study was aimed at estab- or others as to any changes that these sedated on the first day . She lishing the effects of antidepressants others noticed in them over each described the effect as a chill pill. In on levels of wellbeing in subjects not two-week period. A focus group was subsequent days she found normally currently depressed. There were conducted at the end of the study stressing events less stressful. She 9 males and 11 females recruited aimed at establishing whether there described the effect as keeping her from among the administ rative, were effects characteristic of either normal temperamental self on the medical and nursing members of the drug. All ratings were done blind. straight and narrow. She had poor North W est W ales district general The blind was only broken two weeks sleep but was uncertain if this was hospital psychiatric unit, as well as after the study was completed. owing to childcare or the drug. four others known to members of During the two weeks on what the unit. Ethical permiss ion had turned out to be reboxetine, she been obtained from the North West Results reported dry mouth, sleeplessness, Wales Ethical Committee. W ritten reduced appetite, nausea and consti- consent to inclusion was obtained A preliminary analysis of the results pation, which she attributed to the from each subject. All volunteers indicates that two-thirds of subjects drug. were free of medical conditions. responded to one or other of the After randomisation to sertra- None were on concurrent drug treat- two drugs but not the other with a line, within the first few days she ment. None had a history of previous ‘better than well’ response. appeared to become avoidant. She psychiatric illness. The two volun- Response appears to have been pre- complained of a stiff jaw (an teers whose experiences are dicted by personality factors similar extrapyramidal side-effect , some reported here have given consent for to those outlined by Joyce in a study variant of which was reported in 45% this information to be reported. of depressed subjects randomised to of the group while on sertraline). Subjects were randomised to either relatively selective noradren- She had a migraine, which she did reboxetine, a selective noradrenaline ergic or serotonergi c reuptake not attribute to the drug, along with reuptake inhibitor, or sertraline, a inhibitors {24}. The group also nausea, malaise, restlessness, agita- selective 5HT reuptake inhibitor, in a showed changes on the KSP that tion, anxiety , vivid emotions, racing crossover design so that a propor- have been reported in depressed thoughts and ruminations, which tion received reboxetine for two patients following SSRIs {25}. Finally she did attribute to the drug. weeks followed by a two week drug in the group as a whole a greater From the end of the first week free period and thereafter sertraline number of subjects expressed a she was reported by colleagues to be for two weeks or alternatively sertra- preference for sertraline than for somewhat restless and fidgety. T o line followed by reboxetine. The reboxetine. some she appeared akathisic but not 24 Primary Care Psychiatry 2000, Vol 6 No 1 Emergence of antidepressant induced suicidality so clearly that everyone would lessness of her legs. She said that she suddenly decided she should go and notice. She recognised her restless- was not doing much work, that she throw herself in front of a car, that ness and had found that with a couldn’t cope with anyone’ s emo- this was the only answer. It was as if degree of concentration and fiddling tions and she tried to occupy herself there was nothing out there apart she could disguise the effect. She by doing paper work but found it dif- from the car, which she was going to also noted other changes in herself ficult to focus her attention. She throw herself under. She didn’t think that she could not initially clearly described swings of emotion, with of her partner or child. She was walk- describe. misery predominating but she was ing out of the door when the phone In view of the side effects up to not depressed. She was advised to went. This stopped the tunnel of sui- this point, the blind guesses of study stop the drug and agreed to do so. cidal ideation. She later became dis- monitors taking into account both She did not stop. In retrospect, it traught at what she had nearly done drugs were that she was on reboxe- was almost as if she could not stop and guilty that she had not thought tine rather sertraline. She was herself from taking the tablets. A of her family. noticed by the monitors by the end number of other disinhibited things She was taken for a walk and of the first week to have become happened including telling her life appeared to gather herself. Later in needy of time and company. There history to a colleague, whom she did the day she completed a diary entry were moments when she appeared not know well, leaving him con- for the night before to supplement to become preoccupied and emo- cerned. the brief entries of the previous day, tional but on questioni ng would Her diary records impulsive- which included a hope that she mention that she would be well in a ness, , over-sensitivity as would make it through the night. In short while, as her mood had been well as marked suspicion. It also this she described being jumpy, anx- swinging ‘ one minute doom and records a that she had an ‘old ious and suspicious. Her mind was gloom the next sunshine and laugh- me’ and another. The other was like racing and spiralling out of control. ter’. Toward the end of the first week a bit of her childhood self, easily Then it went blank except for the on treatment, her diary records, and moved to emotion, simplistic, aware clear thought that she must kill her- she reported to a study monitor, two of social etiquette but not inclined to self violently by throwing herself incidents involving quite atypical follow it, impatient, selfish and irre- beneath a car or train. This clear behaviour for her. She reported a sponsible. The ‘ old me’ could only thought appeared irresistible and its lack of guilt about something she watch what was happening and was appearance seemed to put an end to was concerned she might be guilty helpless to stop the other bit which the anxiety . It was trance-like and about when she came off the was in control. only broken by a phone-call, which medication and was herself. This as it The following day it turned out came when she was felt she was turned out was the case. she had not stopped the tablets and about to act on the basis of this idea. Over the first weekend she had the medication was discontinued by Contact was maintained over a nightmare about having her throat study monitors. She looked unwell the weekend. By the Monday she slit, so that it gaped open and she but seemed to be making a great declared herself to be back to her- bled to death in the bed. She didn’t effort at trying to appear normal. She self. She looked better but remained get back to sleep. She did not was still akathisic, at a level that vulnerable and was clearly apprehen- increase the dose of sertraline over would be apparent to most sive about talking about what had the weekend. Versions of this night- observers. That night she was seri- happened. Both study monitors and mare recurred on two succeeding ously suicidal, although the tablet the research subject were still blind nights. she was due to take that night had at this point as to the identity of the At the start of week two, she been stopped. medication. The subject herself remained restless, withdrawn and On the Friday she telephoned remains very disturbed at what has preoccupied. The possibility of halt- early in the morning, distressed and happened. ing the drug was raised but she tearful after the previous night. Her During this period daily POMS opted to carry on, despite feeling as conversation was garbled. She and PANAS ratings were undertaken. though she had a combination of described almost going out and On the POMS the words most con- bad PMS and pre-exam nerves. Her killing herself. She was visited at sistently endorsed were lively, active, diary records that she thought she home. She recounted that the night cheerful and vigorous during the might have an that was previously she had felt complete reboxetine block. On sertralin e, aggravating things and a hope that blackness all around her. All she especially in the second week, the when that cleared she would feel could focus on was the pen and the words endorsed were tense, worn better. questionnaire in front of her but she out, unhappy , fatigued, sad, con- By Wednesday , she had couldn’t write anything down. She fused, shaky , discouraged, on edge, become withdrawn, was ruminating felt hopeless and alone. It seemed miserable, bewildered and nervous over impulsive, disinhibited things that all she could do was to follow a were regularly endorsed. On the she had done, was tearful and not thought that had been planted in her PANAS, she had a mean positive herself. She had very obvious rest- brain from some alien force. She word score of 19.7 in week one and 25 Primary Care Psychiatry 2000, Vol 6 No 1 Emergence of antidepressant induced suicidality

20.3 in week 2 on reboxetine, and but most probably was, as she to find her mother extremely fright- 25.3 in week one on sertraline falling endorsed agitation as a side effect of ened about what had happened. to 19.4 in week 2. Her negative word the drug. Close colleagues and her There was a contrast between rebox- score was 10 in week one on reboxe- mother noticed this and other disin- etine and sertraline in this regard in tine, 11.4 in week two, rising to 12.7 hibited events of the kind described that reboxetine had made her feel for week one on sertraline and 22.9 below. All agreed that she was calm but in a manner that left her for week two. adversely affected by the second pill still able to feel fear. In contrast on and she was advised particularly by sertraline, she felt aggressive and her mother never to go on anything fearless. She found herself thinking Case B like that again. on several occasions when faced There was a reluctance to with situations with people that she B was a 28-year old woman at the mention any of this to others and could ‘deck’ them. time of entry to the study. She had only a brief mention in a post-study She increased her dose of ser- no history of psychiatric or medical focus group led to a subsequent dis- traline to 100 mg as per protocol. problems. She smoked and took cussion in which these details Thereafter she felt worse, so that she approximately two units of alcohol became clear. One of the intriguing reduced the dose back to 50 mg her week. Her only significant period features of this woman’ s experi- three days before the end of the of stress was three years previously enced endorsed by Case A was a con- study. Following the increase in when she separated from a partner. cern that elaborating on what was dose, she had two disturbing noctur- At that point in time the thought of happening her would lead to others nal episodes on consecutive nights. suicide had crossed her mind but thinking she was crazy – no-one had She is a woman prone to lucid without any intent, plans or active warned her beforehand that this dreaming and both sleep walking ideation of any sort. could possibly happen on this drug. and sleep talking with extensive She was randomised to rebox- Her diary entries and her rating memory the following day for what etine followed by sertraline 50mgs scales were therefore economical has happened the previous night. for five days increasing up to with the truth. Her partner regularly reports her as 100mgs. On reboxetine at some While on reboxetine on the being awake during these episodes. point during the two weeks, she suf- POMS, this subject endorsed the It is difficult therefore to determine fered from sleeplessness, constipa- words worn out, and weary during what happened but her recollection tion, dry mouth with some reduction the first two weeks, switching to sad, is, while awake or lucidly dreaming, of appetite, sexual dysfunction and annoyed, miserable, unhappy and that she spent a long period lying in impairment of concentration. In gen- angry during week two. On the her bed fantasising about hanging eral, however, she found herself PANAS her positive scores dropped herself from a beam across the bed- more confident, calm and energetic. from 17.6 and 15.4 on reboxetine to room ceiling. She was aware that On going on sertraline she 11.4 and 12.6 on sertraline. Her neg- these thoughts were accompanied noticed the following effects, which ative scores did not change. by an abnormal lack of concern as to she attributed to the drug: nausea, However, there is a group of over 6 whether her partner, mother or oth- lethargy, malaise, panic and pain in co-habitees, work colleagues and ers might find her. She is not aware her jaw . Within two days she also study monitors who had become of ever having comparable thoughts noted in her diary that she had aware of the changes listed above before. The reason she did nothing become snappy and much more and below. she explained afterwards was likely to say things that came to mind B, briefly in her diary and at because she was a coward and had a and that this had been noticed by greater length subsequen tly, vestigial concern about being found her colleagues. By the third day, she described finding herself in a state by her son. This episode repeated records herself as more assertive. where she didn’ t think through the itself the following night. She also reported her mood as liable consequences of what she did or There was a strong feeling that to drop, that generally her emotions said. She didn’t appear to feel afraid. while on the drug in some way she were liable to swing, that she was For example, on one occasion while was being controlled and that suicide sluggish, that she had become irrita- driving home with her mother in the might happen. She rationalised that ble at the slightest little thing and car a group of 18 year old boys by the she had only a few days left on the was liable to take comments person- side of the road made obscene ges- drug and was probably therefore safe ally. Her mood does not appeared to tures and shouted at them, while the to continue, particularly with the have been depressed but colleagues car was moving slowly at a tricky dose reduction. The feelings cleared noted her swings from cheerfulness juncture in the road. She stopped to some extent on the lower dose to withdrawal. the car in the middle of moving traf- and cleared completely after the She also found herself restless fic, went over to them and grabbed drug had been discontinued. In her and reported that she didn’t know if one of them, telling him if he did opinion had the drug been contin- she was ‘coming or going’. This did anything like that again she would ued for longer in a situation where not appear to be an obvious akathisia ‘deck’ him. She came back to the car she was a patient seeking help rather 26 Primary Care Psychiatry 2000, Vol 6 No 1 Emergence of antidepressant induced suicidality than a volunteer who could discon- The cases described in this 5. Wirshing WC, Van Putten T, Rosenberg J, tinue the drug, that there might have paper appear to have become suici- Marder S, Ames D, Hicks-Gray T . Fluoxetine, akathisia and suicidality: is been an up to 50/50 chance of a self dal on sertraline with no easy means there a causal connection? Arch Gen harm episode happening. of explaining what happened other Psychiatry 1992; 49:580–581. A final point of note about this than by invoking an SSRI-induced 6. Lane RM. SSRI-induced extrapyramidal case is that the subject was aware suicidality. The mechanism through side effects and akathisia: implications for treatment. J Psychopharmacol 1998; afterwards of perceptions that the which this was mediated appears to 12:192–214. ‘instability’ she had shown on the have been a combination of akathisia 7. Settle C. Akathisia and sertraline. J Clin drug would reflect on her. She would and emotional blunting, as well as Psychiatry 1993; 54:321. be thought to be unstable rather other features suggestive of an 8. Klee B, Kronig MH. Case report of probable sertraline induced akathisia. Am than to have had a drug-induced automatism. The ‘ disinhibition’ J Psychiatry 1993;150; 986–987. instability. Both subjects felt this. reported appears comparable to that 9. Kutz DL . Drug induced akathisia: They are almost certainly accurate to reported by Hoehn-Saric et al. {28} subjective experience and objective some extent. findings. Milit Med 1994;159:A10. on fluvoxamine and fluoxetine. To 10. LaPorta LD. Sertraline induced akathisia. date, the phenomenology of what J Clin Psychopharmacol can happen has not been explored in 1994;13:219–220. Discussion any detail. It is notable that neither 11. Leo R J. Disorders associated with the selective reuptake inhibitors. drug-induced depression nor drug- JClinic Psychiatry 1996; 57:449–454. These cases, which emerged in the induced akathisia experienced alone 12. Altshuler LL, Pierre JM, Wirshing WC, course of a blind controlled study, necessarily led to suicidality in this Ames D.Sertraline and akathisia . J Clin shed new light on the issue of drug study. Indeed in the case of the two Psychopharmacol 1994;14:278–279. induced suicidal ideation. Hitherto 13. Bonnet-Brilhault F, Thibaut F, Leprieur A, volunteers who became suicidal, Petit M. A Case of Paroxetine induced one of the arguments of the sceptics they do not seem to have been akathisia and a review of SSRI induced has been that depression leads to sui- depressed. akathisia. Europ Psychiatry cidal ideation and against this back- One of the messages of the 1998;13:109–111. ground it is difficult to make a causal 14. Kasper S. The place of milnacipran in the reserpine story appears to have been treatment of depression . Hum connection between antidepressant that those who did not suffer from a Psychopharmacol 1997;12:S135–141. intake and suicidal ideation {25}. psychiatric illness may have been at 15. Baldwin D.The treatment of recurrent The example of reserpine is of more risk from this drug than others. brief depression. European College of Neuropsychopharmacology Meeting interest here. Reserpine, a psy- th This healthy volunteer study sug- London, Sept 24 1994 chotropic drug with RCT evidence of gests that individuals who are not 16. Lilly Memo re suicides and suicide antidepressant efficacy , was associ- frankly depressed may also be at attempts October 1986. Forsyth vs Eli ated with an emergence of suicidal increased risk of drug-induced Lilly, Plaintiffs’ exhibit 73. 17. Healy D. Guest Editorial: A Failure to ideation and a number of completed problems compared to others. Warn. International J Risk Safety Med (in suicides during the 1950s when it Concomitant drug prescription may press). was widely used. At the time, it was in fact in many cases minimise the 18. Ekselius L, von Knorring L (1999). argued that it must be triggering a risks. Given that antidepressants are Changes in personality traits during treatment with sertraline or citalopram. depressive disorder. A recent commonly prescribed for stress reac- British J Psychiatry 174:444–448. reanalysis of these early studies sug- tions, particularly when given alone, 19. Cloninger CR (1987). A systematic gests that in fact it precipitat ed this point would seem to be of some method for clinical description and akathisia and this many have medi- importance. classification of personality variants: a proposal. Arch General Psychiatry ated the suicidality {26}. Of particular 44:573–588. interest, is the fact that the problems 20. Mc Nair DM, Lorr M & Droppleman and suicides occurred in non-psychi- References LF(1988). Manual for the Profile of Mood atric patients being treated with States: San Diego Calif. Educational and Industrial Testing Services. 1. Teicher MH, Glod C, Cole JO . Emergence reserpine for hypertension, permit- 21. Watson D, Clark LA & Tellegen A (1988). of intense suicidal preoccupation during ting a strong causal claim to be The development and validation of brief fluoxetine treatment. Am J Psychiatry measures of positive and negative affect. made. 1990; 147:207–210. In a recent healthy volunteer The PANAS Scale. J Personality & Social 2. King RA, Riddle MA, Chappell PB, Hardin Psychology 54:1063–1070. study involving randomisation to MT, Anderson GM, Lombroso P. 22. Bosc M, Dubini A, Polin V (1997). droperidol, lorazepam and placebo, Emergence of self-destructive Development and validation of a social we also witnessed an emergence of phenomena in children and adolescents functioning scale, the Social Adaptation during fluoxetine treatment. J Am Acad suicidality that appeared to be linked Self-Evaluation Scale. European Child Adol Psychiatry 1991; 30:171–176. Neuropsychopharmacology 7 , suppl 1, to agitation or akathisia {27}. This 3. Creaney W, Murray I, Healy D. S57–S70. antidepressant study provides Antidepressant induced suicidal ideation. 23. Carver CS, White TL (1994). Behavioral support for the argument that Hum Psychopharmacol 1991; 6:329–332. inhibition, behavioral activation and psychotropic drugs in certain circum- 4. Rothschild AJ, Locke CA. Re-exposure to affective responses to impending reward fluoxetine after serious suicide attempts and punishment: the BIS/BAS scales. stances may induce a suicidality that by 3 patients: the role of akathisia. J Clin J Personality and Social Psychology would otherwise not have happened. Psychiatry 1991; 52:491–493. 67:319–333. 27 Primary Care Psychiatry 2000, Vol 6 No 1 Emergence of antidepressant induced suicidality

24. Joyce PR, Mulder RT, Cloninger CR Exhumed. British J Psychiatry (1994). Temperament predicts 172:376–378. clomipramine and desipramine response 27. Healy D, Farquhar G (1998). The in major depression. J Affective Disorders immediate effects of droperidol. Human 30:35–46. Psychopharmacology 13:113–120. 25. Healy D, Langmaack C, Savage M (1999). 28. Hoehn-Saric R, Lipsey JR, McLeod DR Suicide in the course of the treatment of (1990). Apathy and indifference in depression. J Psychopharmacology patients on fluvoxamine and fluoxetine. 13:106–111. J Clinical Psychopharmacology 26. Healy D, Savage M (1998). Reserpine 10:343–345.

28 Primary Care Psychiatry 2000, Vol 6 No 1