July 2006, Vol. XXIII, No. 8
If psychopharmacology sometimes feels like walking a fine line between undertreatment and overtreatment or between letting patients suffer and turning to medications as an "easy fix," then the psychopharmacology of personality disorders can feel like walking a tightrope without a net.
Unfortunately, the published literature is helpful only up to a point. There are not many peer-reviewed randomized controlled trials, given the prevalence and debilitating nature of some of the personality disorders. Studies have tended to be comparatively small and, especially in light of the chronic nature of much Axis II pathology, short in duration.1 Few, if any, trials have compared in size and scope to the large government- or industry-funded studies of the major mood and psychotic disorders.
The studies that have been published often have had high dropout rates, with positive results limited to certain domains and other areas of disturbance largely unimproved.2 There also has been a growing trend, perhaps caused by liability or institutional review board concerns, for treatment studies to include only patients who are comparatively stable and free of active suicidality. Few recent personality disorder drug studies have included, for example, the intermittently unsafe borderline patients often seen in clinical practice. This means that there are virtually no randomized clinical trials of increasingly popular agents like topiramate, lamotrigine, or the atypical antipsychotic in borderline or other personality disordered patients who are or recently have been actively suicidal.
Nevertheless, meaningful strides have been made over the past several years, with several novel agents demonstrating statistically significant results. Most promising are major discoveries relating to neurobiologic abnormalities and neurobiologic changes in response to provocation or treatment that have helped elucidate the constitutional factors that underlie personality disturbance.3,4 Much of this recent outcome and neurobiologic research has focused on aspects or dimensions of the Axis II disorders in addition to or instead of the categoric syndromes defined by the DSM.
Personality disorders are phenotypic expressions of biologic vulnerabilities and arise from a combination of predisposing genotypes and experiential factors.5 These factors may consist of present-day life stressors and/or early abuse and neglect, the effects of all of which may be modulated by damage to the CNS as it develops during childhood and adulthood.6 Although the strong correlation between negative life experiences and personality disorders has become a cornerstone of our understanding of some Axis II pathology, it should be remembered that heredity also appears to play a crucial role in the pathogenesis of personality disorders.
A Norwegian twin study suggested that personality disorders are more strongly determined by genetics than are almost all the Axis I disorders.7 Moreover, the psychological damage caused by abuse and neglect may be modulated by deleterious effects on the "hard-wiring" of the CNS.8 It should not, therefore, be surprising that many patients with personality disorders exhibit structural and functional brain abnormalities, or that certain medications may be beneficial in these disorders.
Borderline personality disorder
By far the most studied personality disorder, borderline personality disorder (BPD), is also the only one to be the subject of an American Psychiatric Association practice guideline.9 Pioneering BPD studies that suggested the utility of neuroleptics and mood stabilizers, as well as the risks of benzodiazepines and tricyclic antidepressants, are by now well known.10,11 Early neurobiologic research implicated disturbances of the serotonin system in the suicidality and impulsive aggression characteristics of this population.12,13 Currently, the best validated medications for BPD may be the selective serotonin reuptake inhibitors (SSRIs)/serotonin-norepinephrine reuptake inhibitors (SNRIs), antipsychotics, and anticonvulsants.
In placebo-controlled trials, fluoxetine has been shown to be helpful with anger,14 irritability, aggression, and overall clinical status,15 although it did not appear to benefit patients receiving concurrent dialectical behavior therapy (DBT)16 and was less efficacious than olanzapine or the olanzapine-fluoxetine combination for depression and impulsivity-aggression.17 In a rare pharmacologic study using decreases in suicide attempts as the principal outcome measure, paroxetine was superior to placebo in a mixed sample of multiple attempters, the majority of whom had BPD.18 Both sertraline and citalopram were found to decrease the frequency of BPD—and of several other personality disorders—in patients entered into a randomized, non–placebo-controlled depression treatment study.19 Most recently, fluvoxamine was better than placebo for rapid mood shifts in BPD but not for impulsivity or aggression.20
Exploration of serotonergic abnormalities has continued to dominate neurobiologic research into BPD and its component dimensions. Trait impulsivity in persons with personality disorders has been associated with blunted hormonal responses to the serotonergic probe ipsapirone.21 Positron emission tomography (PET) scans of patients who have personality disorders with impulsive aggression have shown blunted metabolic responses to fenfluramine in the orbitofrontal, adjacent ventromedial, and cingulate cortices.22 Fluoxetine treatment of BPD in a sample of patients with impulsive aggression was associated not only with clinical improvement but with PET findings of increased metabolic rate in the orbitofrontal cortex.23
During the first wave of pharmacologic research in BPD, neuroleptics were among the most consistently efficacious agents, although follow-up studies to one famous set of trials failed to support haloperidol's overall usefulness.24,25 Two important studies demonstrated a decrease in suicide attempts or self-harm behaviors with depot neuroleptics.26,27 Unsurprisingly, atypical antipsychotics have supplanted neuroleptics in more recent research. Olanzapine has proved better than placebo on scales of anxiety, paranoia, anger/ hostility, and interpersonal sensitivity, although not depression28; overall clinical status29; andwhen added to DBTdepression, anxiety, and impulsive aggression.30 More recently, aripiprazole was found to be superior to placebo for depression, anxiety, anger, and most of the domains measured by the SCL-90-R.31 The (partial) efficacy of antipsychotics in BPD may reflect underlying causative abnormalities of the dopaminergic system.32
Anticonvulsants have proved attractive to clinicians and researchers working with BPD because of the disorder's resemblance, in certain respects, to bipolar disorder, and because BPD patients' "anger attacks" sometimes seem seizure-like in their suddenness and irresistibility. Early encouraging results with carbamazepine10 were followed by a published trial in which the active drug showed no greater efficacy than placebo. 33 In a controlled study of the effect of divalproex on aggression, the intent-to-treat analysis failed to reveal any significant benefit, but subgroups with cluster B personality disorders34 and BPD35 showed improvements on various measures of anger, aggression, and overall clinical status. Divalproex has also been associated with amelioration of interpersonal sensitivity, anger/hostility, and aggression in BPD patients with comorbid bipolar II disorder.36
Newly popular anticonvulsants for BPD include lamotrigine and topiramate, both of which have been associated with decreases in self-report measures of anger in BPD patients.37-39 A recent topiramate study is noteworthy among BPD drug trials for using an usually broad range of outcome measures. Significant treatment-associated improvements were noted in somatization, interpersonal sensitivity, anxiety, hostility, phobic anxiety, and global severity on the SCL-90-R; 4 of 8 scales of the Inventory of Interpersonal Problems; and all 8 scales of the SF-36 Health Survey, an instrument measuring health-related quality of life.40
Innovative, recently studied agents to treat BPD include ethyl-eicosapentaenoic acid, which sometimes is used for bipolar and other mood disorders, and was found to be superior to placebo at reducing aggression and the severity of depressive symptoms in BPD.41 Two different doses of the -adrenergic agonist, clonidine, were given to BPD patients experiencing strong urges to engage in self-injurious behavior; both were associated with significant decreases in dissociative symptoms, suicidal ideation, and wishes to commit self-harm.42
Intravenous naloxone, in the only published placebo-controlled trial of an opiate antagonist known to us, appeared to reduce dissociative symptoms but failed to show greater efficacy than placebo.13 Nevertheless, BPD treatments that mimic or block narcotic activity show some promise. Many BPD patients mutilate themselves or engage in other behaviors like bulimia and substance abuse that also may reflect abnormalities of the endogenous opiate system. BPD patients who self-mutilate have shown diminished pain sensitivity,44,45 which, in turn, has been correlated with dissociation.46 Interestingly, responders to therapy for BPD show normalization of previously impaired nociception.47 It may be that physical self-destructiveness, impulsivity, and dissociation in BPD represent the final pathways of various possible disturbances in the painprocessing system, and that medications that help offset these disturbances will provide substantial benefit.