German Journal of Psychiatry

ISSN 1433-1055



Helmut Peter, F.-Michael Stark, and Dieter Naber

From the Dept. of Psychiatry and Psychotherapy, University of Hamburg, Germany. Corresponding author: Dr. H. Peter, Department of Psychiatry and Psychotherapy, University of Hamburg-Eppendorf, Martinistr. 52, D-20246 Hamburg, Germany, Tel.: 040 - 47174249, Fax: 040 - 47175025, E-mail: Peter@uke.uni-hamburg.de


The course of illness and treatment of a 27-year-old female patient suffering from chronic depression and a dependent and borderline personality disorder is presented. Various psycho- and pharmacotherapeutic treatment methods proved unsuccessful over a period of several years. Finally, a combined treatment with behaviour therapy (BT) and fluoxetine was successful. In addition to a complete remission of symptoms, the treatment led in turn to an improvement in the underlying personality disorder. Of particular interest are the possible synergistic effects and the temporal course of BT and fluoxetine treatment. Aspects of combined treatment involving psycho- and pharmacotherapy are discussed on the basis of a literature review on the topic of SSRIs in the case of personality disorders and treatment-resistant depression (German J Psychiatry 1998;1:15-30).

Keywords: fluoxetine - behaviour therapy - treatment-resistent depression - personality disorder

Received: April 14, 1998 

Published: April 24, 1998


Treatment-resistant depression occurs in 10-30% of all depressive episodes which lack somatic causes (Phillips & Niernberg, 1994). However, it must be considered that there are no clear and generally valid criteria for treatment failure (Nierenberg, 1990), and that frequently used success criteria (e.g. 50% reduction of symptoms) can lead to an overly optimistic view of therapy success. This is because some patients continue to suffer from substantial depressive symptoms even after the attainment of operational criteria. In the case of inadequate treatment success, it must naturally be guaranteed that this is not due to an insufficient therapy. Psychiatric co-morbidity - especially substance dependence, eating disorders and personality disorders - is, in the case of depression, connected with a resistance in particular to antidepressant treatment (Phillips & Nierenberg, 1994).

While pharmacological studies on numerous substances have been able to demonstrate efficacy in the treatment of depression, systematic efficacy studies of psychotherapeutic procedures have, up to now, only been carried out for behaviour therapy (BT) and interpersonal psychotherapy (IPT). In a review of the efficacy of psychotherapeutic procedures for the treatment of depression, Hand (1994) suggests that it is not possible to make a final judgement on the basis of the existing data. The results of a multi-centre study by the NIMH (Elkin et al., 1989) suggest that both BT as well as IPT represent successful intervention strategies. In a comparison of both procedures, IPT was more successful than BT at the end of therapy; however, this advantage could no longer be maintained by the time of an 18 month follow-up. In fact, BT exhibited the most convincing results 18 months after the conclusion of therapy. All studies before and after the NIMH study pointed to equivalent therapy effects of BT and IPT in the short-term. In the long-term, BT was - according to Hand - clearly superior. For severe, endogenous depression, the data situation is not yet convincing for a psychotherapy indication.

In meta-analytical efficacy comparisons of various psycho- and pharmacotherapeutic procedures, two studies (Dobson, 1989; Hautzinger, 1993) found no different effects for the individual intervention strategies - neither for single nor for combined applications. This picture nevertheless changes as soon as a distinction is made between light or moderately severe depression, and severe disorders. In the case of severe depression, antidepressants, best studied in the case of imipramine, are the therapy of choice with the quickest and most extensive prospects of improvement. Long-term successes are presumably best attainable with - as early as possible - a combination with BT. In the case of severely disturbed depressive patients, the issue is, according to the present state of knowledge, not so much a case of "either-or, but rather the optimal combination of pharmaco- and behaviour therapy" (Hand, 1994).

There are only a few controlled studies on the treatment of therapy-refractory depression. Some studies indicate that a differentiated drug treatment can lead to better therapy results. Thus, there are indications that depression with anxiety and agitation symptoms (Partiot et al., 1997) or an impairment of autoaggressive impulse control (Amin et al., 1984; Montgomery et al., 1981) responds better to selective serotonin-reuptake inhibitors (SSRI) than to tricyclic antidepressants (TCA). Lithium augmentation led to therapy success in 30-65% of therapy resistant depressions (Phillips and Nierenberg, 1994). An antidepressant combination treatment, e.g. with fluoxetine and a TCA, can lead to an increase in the success rate to 87% (Weilburg et al., 1989). That the combination with psychotherapy can result in additional positive effects can only be presumed at the present time. Evidence from controlled studies is still lacking.

Case Report

The case of a 27-year-old female patient, who suffered almost continuously over a 5½ year period from a severe, therapy refractory depressive disorder with chronic suicidality, is presented here. The disorder had developed against the background of a mixed personality disorder with dependent, anxious-avoiding, and borderline elements. Treatment took place during this time completely under in-patient or day clinic conditions.

Psychiatric History

From as early as the age of nine, the patient felt unhappy and had lasting anxious changes to her basic mood. The first depressive episode occurred during preparations for her school-leaving examination, accompanied by apathy and suicidality. She did not seek professional help at this time, however. At the age of 21, during the leaving examination for her apprenticeship, there was a recurrence of clinically relevant depressive symptoms with severe inner tension and fears of failure. During this time, she attempted suicide for the first time, taking an undetermined quantity of propanolole. With the psychopathological clinical picture of a severe depressive disorder, with feelings of guilt, fear and failure, she was admitted as a psychiatric in-patient following further persistence of suicidality. She was in an anorexic condition (44 kg, with a height of 162 cm).

Family History

The father's brother suffered from a depressive disorder and committed suicide when the patient was 16 years old. The father had an alcohol addiction which may have masked a depressive disorder. Impairment of his aggressive impulse control occurred under the influence of alcohol. No psychic abnormalities are known in the mother's family.


The patient is the youngest of three children (sister +2 years, half-brother +5 years). The mother, widowed in her first marriage, is described as being strict and at the same time anxious. She had the role of the family breadwinner because the father, as an alcoholic, did not pursue a regular occupation. When in an alcoholised state, he would insult and threaten the entire family and was especially violent towards the mother and the half-brother. The mother was frequently chased out of the house with the children and locked out for days on end. As a child, the patient was spared the father's violent outbursts; however, he frequently sought a close, and possibly abusive contact with her when he was drunk. Because of the special relationship with her father, she was entrusted by the remaining family members with the role of mediator at an early stage, whereby the hazardous proximity to the alcoholised father took on a central significance in the maintenance of the family structure. The remaining family members tolerated and sanctioned the risk to the patient for the sake of protection from the father's violence. The patient developed a chronic anxious tension in this constellation, which persisted over many years. She received no emotional support from her mother. Instead, the satisfaction of her natural childhood needs for motherly attachment and attention was denied, because this was forbidden by the father, while not being permitted by the emotionally distanced and strict mother.

The patient grew up in a village community in which she was subject to constant social control. As a member of an alcoholic's family despised in the village, she experienced early stigmatisation and social isolation. A deep-seated impairment of self-esteem and marked social anxieties developed early on, inhibiting normal social integration outside of the family. She attempted to compensate her feelings of inferiority by outstanding academic achievements at school. Despite academic success, she could not fulfil her own achievement expectations, and was tortured by fears of failure during her entire time at school. Following school lessons, she preferred to remain alone in a boiler-room and did her homework there. At the onset of puberty, the few activities with her peers were largely reduced to formal school contacts, so that any form of psychosexual development was lacking.

During this time, the mother filed for divorce and there was a separation from the father. Years later, the latter emigrated to South America without taking any further interest in the family.

Following the school-leaving examination, with the second-best marks of her year, the patient left home and took up training as a child nurse after a voluntary social year. It was her express wish "to concern herself completely with the care of small children." The new domestic situation in a sisters' home contributed to further isolation, because here also she was unsuccessful at integrating into her social surroundings. Anxious-depressive decompensation and a clinical manifestation of the psychiatric disorder ultimately developed under the pressure of the leaving exam.

Psychiatric Treatment

Initial Psychiatric Treatment as an In-patient

After the first suicide attempt, a 1.5-year in-patient treatment followed, with a focus on a psychoanalytic approach. According to the patient, a drug treatment with 75 mg amitriptyline took place over the entire period. These measures failed to have a lasting positive effect on the depressive syndrome or the chronic suicidality. Because documentation from this treatment period is largely lacking, the reasons for the unsuccessful adherence to the drug regime are unknown. During the long period of residence as an in-patient, she lost her few social contacts. The chronic suicidality prevented discharge and return to her own home, so that a therapeutic home was planned in the course of further treatment.

The course of treatment, including that of the suicidal decompensation, was largely dominated by dependent, anxious-avoiding and borderline personality traits which fulfilled the criteria for corresponding personality disorders (ICD 10 F61.0). An extreme dependent relationship with the attending female therapist developed, to the exclusion of other persons. Imminent or anticipated separation (e.g. discharge) led in every case to a reactivation of the latent suicidal impulse. Shortly before one of the planned transfers, there was a second serious suicide attempt (attempt at strangulation and suffocation with a belt and plastic bag). After 1.5 years of unsuccessful treatment, there was a change of therapist, in view of the irresolvable emotional complications in the patient-therapist interaction, which led to a third suicide attempt. The patient threw herself in front of an underground train and suffered a fracture dislocation with incomplete paraparesis on the level of the lumbar vertebrae 1/2.

Second Psychiatric Treatment as an In-patient

After a six-month neurological rehabilitation treatment as an in-patient in another clinic, exclusively in-patient psychiatric treatment was resumed. The paraparesis and the resulting additional problems for coping with everyday life led to a worsening of the depressive syndrome, so that the psychiatric admission, originally planned as a transitional measure, lasted for a further 1.5 years. An interim attempt at discharge into a therapeutic home failed after only three months.

Therapeutic efforts now focused on a stabilising psychiatric treatment. Behaviour therapy sessions took place once a week, concentrating on coping with daily life problems. Attempted drug treatment played a more significant role. Various psychotropic drugs were employed during this time, partly administered in a combination treatment of three different substances. Antidepressant medication: trimipramine up to a maximum of 75 mg, viloxazine up to a maximum of 600 mg. Benzodiazepines: nitrazepam up to a maximum of 5 mg, lorazepam up to a maximum of 8 mg. Neuroleptics: melperone up to a maximum of 100 mg.

Day Clinic Treatment

After 3.5 years of treatment as an in-patient and several frustrated attempts at transferring the patient to a complementary therapeutic institution, further treatment was planned in a day clinic setting. To this end, the patient was transferred to the psychiatric department of the University Hospital. Curative treatment goals had meanwhile given way to a supportive approach. The therapeutic goal in the day clinic was the gradual severance of the dependence on an in-patient setting, which in an extreme fashion had become a substitute for lacking social contacts and family connections.

The first impression in the admission interview was that of an alert, reflective, in spontaneous rapport fluent, communicative, almost logorrhoeic patient who hardly seemed depressed. In view of the outlined pre-history and course of illness, the patient's self-portrayal seemed oddly parathymous, or at least dissimulating. An agitated-anxious depressive mood with latent to open aggressiveness only became perceptible in later interviews. The patient denied having acute suicidal impulses; however, chronic latent suicidality had to be assumed on the basis of the case history. There were no signs of paranoid symptoms. Her thought processes were, however, dominated by depressive content. Her depressive, self-deprecating cognitions were accompanied by considerable feelings of guilt and insufficiency. Signs of apathy, a psychomotor retardation or an affective vacancy were not determined.

The day clinic treatment, lasting 19 months in total, was marked by various crises. The patient had to be transferred to in-patient treatment seven times, sometimes also under closed ward conditions, because of suicidal or parasuicidal decompensation (on four occasions, there were actual suicide attempts or parasuicidal actions). Conditions of existential anxiety occurred repeatedly, accompanied by menacing fantasies and nightmares, including recollections of physical and sexual threats by her father during early childhood. The patient herself described a deep-seated disturbance of self-esteem associated with a strong yearning for death.

In the day clinic treatment, her relationships with the therapists were characterised by marked ambivalence. Her intense wish for emotional attachment was eclipsed by a deep-seated mistrust in the stability of the relationship and by her deep conviction regarding her own worthlessness in relationships. She often provoked crises in the therapeutic relationship through rejection, sometimes hostile, deprecating and self-deprecating behaviour, which for the patient involved a threatening and anxiously anticipated instability in the relationship. The therapist's absence at weekends, during holidays and due to illness triggered deep despair and reactivated her yearning for death.

Multimodal Behaviour Therapy Treatment

The severe clinical picture made systematic therapy planning and treatment impossible. Therapeutic sessions concerning biographical events often led to crisis-like worsening which made further exploration difficult, often impossible. The exploration sessions either were broken off by the patient or immediately led to therapeutic treatment. The stages of therapy below in part overlapped with shifting focus. The chronological sequence implied by the numbering is intended exclusively as a rough orientation:

1. Building Up of the Therapeutic Relationship

The patient's need for attachment, with simultaneous mistrustful rejection, at first impeded content-related therapeutic work, so that the stability of the therapeutic relationship became the centre of attention in the treatment. In the case of every necessary crisis-related transfer to the ward, the patient was reassured of the continuity of both the treatment offered as well as of the female therapist to whom she related most closely. The constant experience of continuity, in spite of repeated separation provoked by the patient (e.g. admission to a closed ward), made possible a gradual building up of trust in the therapeutic relationship which, thanks to the special efforts of the therapist, developed into a sound treatment alliance.

2. Examination of the Traumatising Childhood Experiences

Traumatising childhood experiences were made a central theme in individual sessions, on the one hand to promote further hypothesis development and, on the other hand, because recurring memories of these experiences increasingly revealed themselves to be an intrapsychic condition for the origin and maintenance of the depressive syndrome. This was because they essentially contributed to chronic affective labilisation as well as to the development of the autodestructive self-image, due to culpable processing. The patient's attempts to avoid examination of these memories were unsuccessful, because they intruded intensively into her everyday life, despite her efforts. The dysfunctional avoidance behaviour meant that an adequate processing and emotional distancing failed to occur. In therapy, confrontation with the upsurge of emotions had to take place in cautious steps which were bearable for the patient. Through graduated emotional exposure, an initiation of the necessary grief could be permitted, as well as an open experience of her suppressed anger. The fact that she could cope with the released emotions and was not, as she had feared, overwhelmed and destroyed by them, led to a reduction in her anxiety. The repeated experience that the anticipated condemnation by the therapist did not occur promoted trust in the therapeutic relationship and at the same time led to a correction of the depressive self-image. An examination of the disturbance model presented in the therapy enabled a distanced reflection on her past and a new evaluation of her childhood memories.

3. Treatment of the Depressive Cognitions (in accordance with Hautzinger et al., 1992)

Depression-specific automatic thoughts and basic assumptions were identified during biographical analysis. An indirect and cautious procedure in Socratic dialogue proved to be the most successful in the treatment of these cognitions. Again, the experience that she was not condemned by the therapist, despite the exposure of supposed weaknesses and mistakes, was helpful in this situation. This enabled reflection on the deep-seated attributions of guilt. As therapy progressed, a direct treatment of the depressive attitudes and a corresponding correction was increasingly possible in the dialogue. Far-reaching changes of dysfunctional depressive cognitions could be attained in this phase, and there was a change from a self-destructive behavioural pattern to an active coping style in crisis situations.

4. Treatment of the Social Anxieties and Deficits

In the course of treatment, social anxieties and deficits were integrated into the therapy plan and treated both in an individual as well as in a group setting. After step-wise analysis of problem and alternative behaviour, considerable difficulties arose in the implementation of exercise units. These could only be mastered under certain conditions through role plays. An exceptionally high level of anxiety and the persistence of depressive symptoms, but also motivation problems - especially the activation of fears of separation in the case of steps taken outside of the day clinic - were the main reasons for this. The first exercise steps in a real situation were finally enabled by constant encouragement from the therapist. Initially for the therapist's sake, and later because of an understanding of their necessity, the patient carried them out and thus reduced her anxiety by means of repeated exposure. As a result, it was possible to develop or to improve social skills in diverse social contexts (making contact, holding a conversation, group activities, dealing with public authorities) with therapeutic guidance (behaviour analysis, role plays, exercises). This also involved changes in her sometimes brusque, rejecting interaction behaviour. In this way, she was finally able to achieve the building up of social activities and contacts which she consistently maintained, despite the continued persistence of depressive mood. She regularly participated in activities outside of the day clinic (sports for the disabled, a self-help group for physically disabled women, adult college choir, concerts, weekend seminars), which she nevertheless often found strenuous and, in part, an ordeal. Her understanding of the necessity of these measures, as well as regular motivation by the therapist, helped her to persevere with these behaviour changes despite insufficient improvement of mood.

5. Change of Current Lifestyle and Life Planning

In particular it was the last focal point of therapy, inherently associated with release from dependence on hospital, which led to an activation of her fears of separation, and thus to the occurrence of crises. Nevertheless, an examination of these themes took place, because a change in behaviour pattern and in the cognitive evaluation of such crises had developed during therapy. Suicidal thoughts and actions were substituted by admittedly still inadequate, but not life-threatening retreat behaviour in overload situations (overdoses with the medication, in order to sleep through the following day).

Despite the therapy successes described above, the persistent depressive mood, with marked feelings of hopelessness and fears of failure which could not adequately be influenced cognitively, remained a decisive barrier to further positive development.

Pharmacological Treatment

At the commencement of day clinic treatment, the patient took viloxazine 600 mg, lorazepam 2.5 mg, and chloral hydrate 500 mg. The long-term goal of the pharmacological treatment was the adjustment to an effective antidepressant medication and, given the high dependence potential of the patient, the gradual reduction of the lorazepam dosage. However, changes in medication during the first half of treatment were in particular influenced by the crisis-like worsening of the symptoms on a number of occasions. During this time, psychotropic drugs such as perazine and buspirone were employed because of the repeated agitated-anxious and depressive-suicidal decompensations. The psychopathological picture and the hitherto unsuccessful treatment with a TCA, as well as with two atypical antidepressants, suggested the employment of an SSRI. A systematic change of the antidepressant medication was not attempted, because even the smallest changes in well-being, e.g. caused by changes in medication, provoked acute suicidality and thus endangered the treatment setting in the day clinic. A renewed, longer-term admission as an in-patient was inconsistent with the original therapeutic goal, namely to achieve independence from such a setting. This would have reactivated all the problems of separation, and was furthermore rejected by the patient. There were further reservations against using SSRIs, since suicidality is still listed as one of their relative contra-indications.

Only once the acute risk of suicide had abated was an SSRI therapy carried out under day clinic conditions. With 20 mg fluoxetine, there was a clear improvement in mood after around 14 days, accompanied by further behaviour changes already planned before the change in medication, but not implemented. In the following four months, the patient was able to achieve further changes in lifestyle, e.g. taking up various social contacts and activities. These were also increasingly accompanied by a sense of pleasure and joie de vivre. Her body weight returned to normal during this time. It was now possible in psychotherapeutic work to concentrate on the central theme of relationship continuity and to plan and carry out a staged release from dependence on a therapeutic setting, so that a discharge from hospital was possible six months after the commencement of fluoxetine therapy.

Some months later, the patient took a further step towards autonomy. She left the therapeutic flat-sharing community, and shared an independent flat with a female friend whom she had met in the course of her social activities. Here, it was possible for her to build up a sound bond and, for the first time since her early childhood, to overcome her social isolation and loneliness. Furthermore, for the first time in her life she was able to develop her first sexual relationship with a man. She took on a course of study and was able, accompanied by many kinds of uncertainty, but without the familiar fears of failure and achievement, to cope with the initial demands. These changes in the patient's lifestyle can, in view of the intrapsychic problems reaching far back into childhood, be considered far more than a mere abatement of a depressive syndrome. Rather, changes were exhibited in the premorbid personality traits that predisposed towards the depressive disorder. The patient has now been in a stable mental state for a total of two years under fluoxetine. For the past year, she has been treated exclusively as an out-patient.


The case report presented above describes the course of illness of a 27-year-old female patient who suffered for 5½ years from a severe depressive disorder with chronic suicidality against the background of dependent, anxious-avoiding, and emotionally unstable personality traits. The severe, therapy-refractory suffering of the patient made in-patient (or day clinic) and pharmacological treatment necessary over the entire period. Given the unsuccessful therapy over many years, the surprising and rapid remission following commencement of SSRI therapy poses a series of questions and aspects worthy of discussion, in particular related to the sequentially interacting affects of psycho- and pharmacotherapy.

From a pathogenic point of view - despite the peculiarities in the course of illness related to intensity and chronicity of the symptoms - a very typical constellation existed, suggesting a multi-factor, biopsychosocial origin of the disorder. The family history points to a significant biological-genetic component, with an accumulation of psychiatric illnesses in the father's family. By contrast, given the biographical and anamnestic data, the disorder can be understood both for the psychoanalytically orientated as well as for the behaviour therapy orientated therapist as a psychogenic disorder. This constellation contains the danger of a one-sided dichotomisation of the presumed causes of illness and of the resulting treatment concepts. In the case of the patient described above, the focus of treatment during the first 1.5 years was psychoanalytically orientated in-patient therapy which, given the ominous biography and the self-image of the patient, appeared to be justified. This attempt at treatment failed tragically. A marked dependence on the attending female therapist developed. At the same time, there was a complete loss of the rudimentary social network with a consecutive chronicity of the problems. Pharmacological measures were employed during this treatment phase primarily in the sense of accompanying medication. Antidepressant medication was administered over the entire period which, in spite of insufficient effects on the target symptoms, did not entail a change in dosage (too low) or substance.

In contrast to the above, the procedure during the second in-patient treatment phase was characterised by the process of progressive chronicity, and was lastingly influenced by a severe suicide attempt with resulting paraparesis. The psychotherapeutic curative treatment approach largely gave way to a general psychiatric procedure with stabilising and structuring interview contacts and the minimal goal of re-establishing the patient's independence of hospital psychiatry with the aid of complementary treatment facilities. During this in-patient treatment phase, which lasted barely two years, various antidepressants were administered, sometimes at the maximum dosage. At the same time, tranquillisers and neuroleptics were administered, similarly without much success.

Transfer to the university's psychiatric department was followed by a third treatment period in a day clinic. The focus of the first treatment phase was the overcoming of problems in patient-therapist interaction. These had developed against the background of a personality-related relationship disorder on the patient's part. Building on this, content-related therapeutic goals were striven for, in particular related to the depressive cognition as well as the social deficits and the deep-seated fears of failure against the background of the traumatising childhood experiences. This represented the basis for changes in attitude and behaviour with respect to current lifestyle and life planning. There was a change of pre-existing behavioural patterns in crisis situations as well as in passive, avoiding, and autodestructive conflict behaviour. Nevertheless, the psychotherapeutic measures did not have sufficient impact on the depressive symptoms. The commencement of fluoxetine treatment at this point led to a rapid abatement of the depressive mood and, as a result, to a far-reaching psychosocial adaptation with positive changes extending into the personality.

The surprising turning-point in the chronic course of the illness, which far exceeded an abatement of the depressive syndrome, is a relatively rare phenomenon in everyday clinical terms. Given the drastic changes immediately after use of the SSRI, the question must be raised as to whether the observed effect can be attributed exclusively to the medication, or rather to a synergistic effect between psycho- and pharmacotherapy. In order to review the first possibility, it is necessary to clarify the matter of how fluoxetine differs positively in its effect on the symptoms of the patient compared with that of other antidepressants and, if need be, whether its use at an earlier stage, despite existing suicidality, would have been possible or should even have been carried out. The second possibility, namely the assumption of a synergistic effect between psycho- and pharmacotherapy, would appear to be justified in the light of the course of illness and therapy. With respect to the potential clinical implications, it represents an exciting alternative.

In studies on SSRIs, there are indications that in the case of symptom constellations such as in the case presented - with chronic suicidality, disturbed aggressive impulse control, and agitated-anxious mood - a serotonin deficiency can be of decisive pathogenic significance:

A serotonin disorder first became associated with depression at the end of the 1960s. 5-hydroxyindole acetic acid (5-HIAA), a breakdown-product of serotonin metabolism, is reduced in the cerebrospinal fluid of some, but not all depressive patients (Åsberg et al., 1976). An explanation for this non-uniform finding was first found in later studies; in particular, it was suicidal patients who were affected. Subsequent studies confirmed these results and showed that this involved an extremely stable and culture-independent context. The suicidality hypothesis was further specified in the course of time. A metabolic disorder of the serotonin balance is associated less with depression or some other diagnostic category; rather, it is associated with an impairment of aggressive behaviour (Montgomery, 1991).

Therapeutic studies indicate that fluoxetine can be particularly effective in the case of agitated-anxious depressions (Partiot et al., 1997) and borderline personalities (Coccaro et al., 1990; Cornelius et al., 1990; Coccato and Kavoussi, 1997; Norden 1989). A positive effect on the depressive mood and impulsive aggressiveness could be demonstrated, in the case of borderline patients, with a dosage of 5-80 mg fluoxetine. In a comprehensive meta-analysis (Montgomery, 1989), fluoxetine was found to be superior in comparison with reference antidepressants (TCA) and a placebo in patients with agitated depression. A positive effect on the agitated-anxious syndrome could also be demonstrated for other SSRIs. In a comparison between paroxetine, placebo and active control (mostly TCA) on about 3,000 patients, Sheehan et al. (1992 ) found that paroxetine significantly reduced anxiety symptoms, while the active control condition showed no effect on anxieties. In two comprehensive meta-analyses, a superiority of fluvoxamine in the treatment of anxiety symptoms in the case of depression could be demonstrated compared to imipramine (Montgomery, 1991). These results are in accordance with the known positive effects of serotonin reuptake inhibitors in the treatment of anxiety disorders. Paroxetine also seems to have a favourable effect on personality-related characteristics of affectivity. In an initial double-blind, placebo-controlled study (Knutson et al., 1996), a significant reduction of the "negative affects" and hostility in the case of healthy test persons could be attained one and four weeks after commencement of treatment, in comparison with a placebo group.

In the case of the treatment of our patient, an SSRI medication would thus not only be indicated, but would even come into question as the therapy of first choice. In contrast to this, there are case reports which indicate that an increase in suicidal actions can occur with SSRIs, so that suicidality is still listed today as a relative contraindication in the directions for use for fluoxetine treatment. Although more recent studies (Beasley et al., 1991; Judge, 1997) show that a reinforcement of suicidal actions is improbable on fluoxetine and fluvoxamine (Wagner et al., 1993), suicidality is still seen as a relative contraindication for SSRI treatment in a German standard textbook on psychiatric pharmacotherapy (Benkert et al., 1996) and has restrictions such as a temporary combination with benzodiazepines. There is hence, despite current study results, uncertainty over the possibilities and limits of treatment of suicidal patients with SSRIs, which can lead to therapeutic reservations especially under out-patient or day clinic conditions.

The far-reaching psychopathological changes following the use of fluoxetine are in many details similar to Kramer's descriptions of personality changes with fluoxetine (1993). In the case presented, partial therapeutic successes had, however, already occurred before SSRI therapy, whereby the alternative assumption of a synergistic interaction of psycho- and pharmacotherapy becomes more plausible. The sequential therapy successes alone, with changes in various psychic functional areas as a result of different therapeutic measures, supports this hypothesis. During psychotherapy, there were far-reaching changes in the central cognitions and behaviour patterns for the depressive disorder prior to fluoxetine treatment. According to an understanding of behaviour therapy, the symptom complex, consisting of self-deprecating cognition, social isolation, lack of perspective in lifestyle and destructive coping strategies, represents a fundamental maintaining condition for the depression. Successful treatment of depression thus requires positive changes in these areas, so that a reduction of the depressive mood with retention of the pathological attitudes and behaviour patterns would only have been of limited stability and without far-reaching consequences. The therapeutic successes prior to the pharmacologically induced improvement of mood are, according to this theory, a fundamental requirement for the exceptional efficacy of the SSRI medication.

Despite the stable cognitive and behavioural changes which were achieved in the BT, the desired reduction in depressive mood did not occur. The insufficient impact on mood became a decisive barrier to further therapeutic success, and could only be overcome by pharmacological influence on the serotonin metabolism. According to BT concepts, it must be assumed that genetic or acquired biological conditions, e.g. a pathological alteration of the central serotonin balance, became a decisive maintaining condition of the disorder at this point in the therapy process, and required a specific biological therapy. The biologically induced improvement of mood, the emotional support in the therapeutic relationship and the cognitive and behavioural changes through the therapeutic work could have resulted in a constellation which was hitherto lacking in the biography of the patient, enabling the subsequent changes in the premorbid personality traits.

The presumption, that interactions between psycho- and pharmacotherapy can occur which have a greater effect than the sum of the individual effects, can only be speculative at the present time. A systematic study of this assumed phenomenon has yet to be conducted. Because of the possible therapeutic implications, however, such a study would be of great clinical relevance. For the behaviour therapist, a series of questions is posed in this context: Can the prospects for success of behaviour therapy be improved by prior pharmacotherapy and, if so, is it possible to develop specific indication criteria for these? Can behaviour therapy successes lead to a potentiation of the effect of subsequent pharmacotherapy and, if so, under what circumstances can such effects arise? Possible consequences must also be considered with respect to pharmacotherapy. Thus, in addition to psychopathological aspects in the choice of medication, the question of the optimal time of commencement of treatment can play a role in therapeutic decisions. Given the considerations presented here, it would in individual cases be possible that a medication employed unsuccessfully could later in the treatment process - after removal of the symptom-maintaining conditions - demonstrate a hitherto unseen therapeutic effect.


Amin MM, Ananth JV, Coleman BS, Darcourt G, Furkas T, Goldstein B, Lapiere YD, Paykel E, Wakelin JS: Fluvoxamine antidepressant effects confirmed in a placebo-controlled international study. Clin Neuropharmacol 1984;7 (Suppl 1):511:581.

Åsberg M, Träskman L, Thoren P: 5-HIAA in the cerebrospinal fluid: a biochemical suicide predictor? Arch Gen Psychiatry 1976;33:1193-1197.

Beasley CM, Jr., Dornseif BE, Bosomworth JC, Sayler ME, Rampey AH, Jr., Heiligenstein JH, Thompson VL, Murphy DJ, Masica DN: Fluoxetine and suicide: a meta-analysis of controlled trials of treatment for depression. Br Med J 1991;303:685-692.

Benkert O, Hippius H, Wetzel H, Gründer G: Psychiatrische Pharmakotherapie. Berlin, Springer, 1996.

Coccaro EF, Astill JL, Herbert JL, Wernicke JF: Fluoxetine treatment of impulsive aggression in DSM-III-R personality disorder patients. J Clin Psychopharmacol 1990;10:373-375.

Cornelius JR, Scloff PH, Perel JM, Ulrich MS: Fluoxetine trial in borderline personality disorder. Psychopharmacol Bull 1990;26:151-153.

Dobson K: A meta-analysis of the efficacy of cognitive therapy for depression. J Consult Clin Psychol 1989;57:414-419.

Elkin I, Shea MT, Watkins JT, Imber SD, Sotsky SM, Collins JF, Glass DR, Pilkonis PA, Leber WR, Docherty JP, et al.: National Institute of Mental Health Treatment of Depression Collaborative Research Program. General effectiveness of treatments (see comments). Arch Gen Psychiatry 1989;46:971-82.

Hand I: Verhaltenstherapie und/oder Pharmakotherapie bei Depressionen? Fortschr Neurol Psychiat 1994;62:44-52.

Hautzinger M: Kognitive Verhaltenstherapie und Pharmakotherapie bei Depressionen - Überblick und Vergleich. Verhaltenstherapie 1993;3:26-34.

Hautzinger M, Stark W, Treiber R: Kognitive Verhaltenstherapie bei Depressionen. Weinheim, Psychologie Verlags Union, 1992.

Judge R: Dangers of fluoxetine. Lancet 1997;349:809.

Knutson B, Wolkowitz OM, Reus VI, Chan T, Cole SA, Moore E, Manfredi F, Terpstra JI, Johnson R: Personality effects of paroxetine in normal humans, American Psychiatric Association Annual Meeting, 1996, New Research Program & Abstracts. New York, pp 94.

Kramer PD: Glück auf Rezept. München, Kösel-Verlag, 1993.

Montgomery S: The efficacy of fluoxetine as an antidepressant in the short and long term. Int Clin Psychopharmacol 1989;4 (Suppl 1):113-119.

Montgomery S: Serotonin und Depression; in Heinrich K, Hippius H, Pöldinger W (Hrsg.): Serotonin. Ein funktioneller Ansatz für die psychiatrische Diagnose und Therapie? Berlin, Springer, 1991, pp 95-109.

Montgomery S, McAuley R, Rani S, Roy P, Montgomery D: A double-blind comparison of zimelidine and amitriptyline in endogenous depression. Acta Psychiatr Suppl 1981;63 (Suppl 290):314-327.

Nierenberg A: Methodologic problems in treatment of resistant depression. Psychopharmacol Bull 1990;26:461-464.

Norden M: Fluoxetine in borderline personality disorder. Prog Neuropsychopharmacol Biol Psychiatry 1989;13:885-893.

Partiot A, Jouvent R, Pierson A, Baruch P, Beuzen J, Ammar S, Wildlöcher D: Are there differential effects of fluoxetine in retarded/blunted affect versus agitated/anxious depressives? A clinical study. Eur Psychiatry 1997;12:21-27.

Phillips K, Nierenberg A: The assessment and treatment of refractory depression. J Clin Psychiatry 1994;52 (Suppl 2):20-25.

Sheehan D, Dunbar G, Fuell D: The effect of paroxetine on anxiety and agitation associated with depression. Psychopharmacol Bull 1992;28:139-143.

Wagner W, Plekkenpol B, Gray TE, Vlaskamp H, Essers H: Safety database on fluvoxamine: analysis and report. Pharmacopsychiat 1993;26 (Suppl.):10-16.

Weilburg J, Rosenbaum J, Biedermann J: Fluoxetine added to non-MAOI antidepressants converts nonresponders to responders: A preliminary report. J Clin Psychiatry 1989;50:447-449.