Borderline Personality Disorder - Pharmacotherapy

During the past two decades, the use of medication in the care of borderline patients has changed from being occasional to being expected. In this section, we begin by discussing some general principles and then discuss specific approaches based on existing empirical evidence. The borderline patient should be encouraged to participate in decisions about the use of pharmacological agents.

An N-of-1 trial, in which a particular medication regimen is compared with the patient’s prior treatment, is a useful way to engage a patient in therapy. This approach has particular merit for borderline patients because it encourages their collaboration in assessing the medication, diminishes their fears of being controlled, prevents a long-term commitment to interventions that do more harm than good, and makes the expectations for outcome of treatment clearer.

Another general principle of pharmacotherapy is that borderline patients require special attention to monitor effectiveness and judge side effects. Because research has shown a lack of agreement between self-rated and observer-rated measures of outcome, it is useful to monitor both or to choose specific behavioral outcomes to evaluate the patient’s response. Before prescribing the medication, the clinician should assess the patient for symptoms (which are consistent with the profile of side effects of the medication to be used). This evaluation provides a baseline for monitoring the appearance of new symptoms during the trial of new medications. Completing self-ratings and observer ratings of the symptoms targeted for treatment will assist the clinician in evaluating the response to medication.

Soloff suggested a symptom-specific approach to pharmacotherapy. For BPD patients, four areas of dyscontrol can be targeted for pharmacotherapy: 1) cognition, 2) affect, 3) impulse, and 4) anxiety. In the discussion of pharmacotherapies that follows, we examine these four putative areas of dyscontrol and familiarize clinicians with the evidence for the role of the various classes of medications (

Table 83-2).

Cognitive Dyscontrol

Problems of cognitive dyscontrol that are common in borderline patients include depersonalization/derealization: illusions, ideas of reference, and brief paranoid states. Several randomized controlled trials of traditional antipsychotic medications in borderline patients have been completed. These studies have generally supported the effectiveness of low-dose antipsychotics over placebo. The response to antipsychotics has not been restricted to improvement in psychotic-like symptoms. Anxiety, obsessive-compulsive symptoms, affective symptoms, and suicidal behaviors all significantly improved in patients taking antipsychotics compared with those taking placebo.

Several case reports and series of the effective use of novel antipsychotics, including risperidone, clozapine, and olanzapine, in borderline patients have been published. Clozapine treatment generally has been restricted to intractable patients with Axis I psychotic disorders and BPD. Despite frequent side effects, clozapine reportedly has reduced troubling symptoms, including severe self-mutilation, in these patients.

Low doses of antipsychotics, including novel agents, have been shown to be effective as nonspecific tranquilizers that reduce, in the short term, the severity of a broad range of symptoms found in acutely distressed borderline patients. Novel agents may be effective with a lower risk of side effects, although randomized controlled trials are needed to be definitive about their role in BPD.

Affective Dyscontrol

Affective dyscontrol problems of borderline patients include depression and other indications of dysregulation of mood, such as volatility and intensity of feelings. We review the efficacy of medication for depression and mood lability separately.

The efficacy of antidepressants for depression in patients with BPD has been examined in several randomized controlled trials. Tricyclic antidepressants have not shown significant effectiveness for depression in borderline patients; Soloff et al. reported a worsening of symptoms in some patients who were given amitriptyline.

Monoamine oxidase inhibitors have shown some promise in the treatment of BPD, particularly for patients with atypical depression. Soloff et al.  failed to demonstrate much superiority of phenelzine over haloperidol and placebo with regard to observer-rated depressive symptoms. Phenelzine did have significant effects on anger and hostility compared with placebo.

Selective serotonin reuptake inhibitors (SSRIs) have been examined in patients with BPD in two published randomized controlled trials. In general, various symptoms of the disorder significantly improved; however, anger and impulsivity may be particularly targeted with these medications. Studies with dual-action agents, such as venlafaxine, are beginning that have reported effectiveness in an open trial with BPD patients.

Although the need to treat a coexisting major depression in patients with BPD is common, clinicians should approach treatment carefully. The depression should be documented to persist for at least 2-3 weeks, and clinicians should recognize that these symptoms often resolve because of the benefits of being admitted to a hospital. Major depression in the context of BPD may represent a different syndrome than major depression alone, and this may explain why the coexistence of depression is not a strong predictor of a patient’s response to antidepressant medications. In treating major depression plus BPD, a less complete response to antidepressant medication can be expected and the effects of the medication may take longer to appear. Similarly, depressed patients with BPD will probably respond to electroconvulsive therapy, but the response may be short-lived .

After these complications have been considered, a trial of antidepressant medication appears to be warranted for those borderline patients with persistent depressive symptoms. Evidence suggests that the first line of treatment should be serotonergic reuptake inhibitors. These agents are relatively safe and have few side effects. Because of this, Soloff recommends that a second SSRI should be tried if the first SSRI is ineffective. The next choice would be a monoamine oxidase inhibitor. Some evidence indicates that these agents are effective in these patients, but patients must comply with a diet regimen.

The role of mood stabilizers in BPD is still undetermined. Randomized controlled trials of lithium, carbamazepine, and divalproex have shown effectiveness, particularly against impulsivity and suicidality, but less so for mood stabilization. Other open-label studies of valproate and lamotrigine and a randomized trial of carbamazepine have suggested that effectiveness may be limited to a subset of patients.


Impulsivity, expressed in actions such as recklessness, bingeing, promiscuity, and impulsive suicide attempts, is a significant problem in patients with BPD that sometimes is amenable to pharmacological treatment. Links et al. offered evidence that lithium may be useful against anger and suicidal symptoms in patients with BPD. Soloff et al. found that phenelzine was effective in reducing anger and hostility. Gardner and Cowdry showed that carbamazepine reduced symptoms of behavioral dyscontrol, whereas alprazolam significantly increased it.

A growing body of literature confirms the relation between low levels of 5-hydroxyindoleacetic acid in cerebrospinal fluid and a history of suicide attempts and aggression. Coccaro et al. reported diminished central serotonin responsiveness in borderline patients to fenfluramine challenge and suggested that increasing serotonin neurotransmission might decrease impulsive suicide attempts and aggression. Montgomery showed flupenthixol to be effective in reducing episodes of suicidal behavior over a 6-month follow-up period. As noted, evidence indicates that fluoxetine may reduce anger and impulsivity in borderline patients, and Coccaro and Kavoussi demonstrated effectiveness of fluoxetine in patients with personality disorders characterized by impulsive aggressiveness.

Several agents are effective in reducing impulsivity; therefore, agents should be chosen based on safety in overdose and side-effect profile. Approaches that require further study are the use of naltrexone in repetitive self-harm behavior and psychostimulants in impulsive borderline patients with residual adult symptoms of attention-deficit/hyperactivity disorder.


The severe anxiety observed in many borderline patients can be classified into two types: somatic and psychic. These subtypes may help guide pharmacotherapies.

Somatic anxiety is experienced in the body and is associated with impulsivity and a stimulus-seeking pattern of behavior. It is most often associated with antisocial behavior and a histrionic cognitive style. This type of anxiety is best treated with a monoamine oxidase inhibitor or by catecholaminergic drugs.

Psychic anxiety is most often observed in patients who have a low tolerance to stimulation and a high anticipation of harm. These patients use obsessional cognitive styles. This type of anxiety may be best treated with a long-acting benzodiazepine such as clonazepam. Alprazolam, however, and perhaps other short-acting benzodiazepines can disinhibit impulsive and violent actions and lead to dependency.

Comorbid anxiety disorders may be more common in BPD patients than previously recognized, and treatment of these comorbid conditions should be a priority. Recent case reports have suggested that naltrexone may reduce flashbacks and self-mutilative behaviors by borderline patients .

Next article Borderline Personality Disorder - Residential Therapies » »


Provided by ArmMed Media
Revision date: June 11, 2011
Last revised: by Dave R. Roger, M.D.