Is pharmacotherapy useful for treating personality disorders?

KEYWORDS:

• Antidepressants
• antipsychotics
• antisocial personality disorder
• borderline personality disorder
• drug treatment
• mood stabilizers
• pharmacotherapy
• personality disorders
• psychopharmacotherapy
• schizotypal personality disorder

1. Introduction

Personality disorders are characterized by enduring maladaptive patterns of behavior, cognition, and inner experience exhibited across many contexts and deviating from those accepted by the individual’s culture [1]. In 1995, DeBattista and Glick [2] concluded in a review of the then-available studies (mostly non-randomized) that it was ‘unlikely that pharmacotherapy will ever “cure” personality disorders.’ Now, a quarter of a century of research later, more than 90% of treatment-seeking individuals with any personality disorder (PD) diagnosis are nevertheless prescribed at least one psychotropic medication, according to a cross-sectional survey in the UK [3]. Did the evidence accumulated since 1995 change our understanding of the effectiveness of drug treatment substantially? Is pharmacotherapy useful for treating PDs after all?

A look at published trials of drug treatment for PD since the introduction of the 10 distinct personality disorders in the DSM-III in 1980 clearly shows that only some PDs have been subject to pharmacotherapy research (see Figure 1). More than three quarters (77.5%) of all such trials focus on Cluster B (‘emotional-dramatic’) PDs, with borderline personality disorder (BPD) specifically accounting for approximately two thirds (65.7%) of all studies.

Figure 1. Number of clinical trials of drug treatments for personality disorders by cluster, published 1980 to 2019 (MEDLINE search via PubMed on 10 November 2020)

2. Drug treatment research is restricted to a few PDs

2.1. Cluster B (dramatic, emotional, and erratic PDs): most intensively studied

Though BPD is not the most prevalent PD in the general population, it is highly prevalent in clinical samples [4] and urges action from clinicians, especially in times of crisis that may repeatedly occur in this disorder. Therefore, the search for treatment options has attracted much attention. As BPD includes symptoms of four different clusters (impulsive-aggressive, affective-dysregulative, interpersonal, and cognitive-perceptual), all of which might respond to different classes of pharmacological agents, many different drugs have been studied, randomized controlled trials (RCTs) focusing on antipsychotics and mood stabilizers [5]. The current evidence indicates limited effects of second-generation antipsychotics (esp. olanzapine) for affective-dysregulative and cognitive-perceptual symptoms but also a possible increase of self-harm [5]. For quetiapine, an anti-aggressive effect was observed in a single RCT, but no impact on impulsivity measures [5]. At the same time, hopes that mood stabilizers (MS) provided an alternative treatment option were disappointed by a recent study that outweighs the previous MS evidence by methodological strength and statistical power and observed no effects after one-year treatment with lamotrigine in a large naturalistic sample [6]. While in practice antipsychotics and mood stabilizers are still recognized as treatments for impulsivity-related symptoms, the current evidence does not firmly support this belief.

Antidepressants, though widely used (mostly selective serotonin reuptake inhibitors (SSRIs)) [7], lack supporting evidence for the treatment of BPD [5].

A recent Cochrane review of drug treatments for antisocial personality disorder (AsPD) identified eleven RCTs of mixed samples [8]. Only four of these studies, published between 1994 and 1997, reported treatment outcomes of individuals with AsPD separately, testing two tricyclic antidepressants (desipramine, nortriptyline), a mood stabilizer (phenytoin), and two dopamine agonists (bromocriptine, amantadine) against placebo. Low-certainty evidence for some effects was only found for Phenytoin. However, the evidence is confined to single, unreplicated studies and inconclusive. The widely recognized British treatment guidelines, commissioned by the National Institute for Health and Care Excellence (NICE) recommend that (much as for the case of BPD) pharmacological interventions should not routinely be used in AsPD [9].

The two remaining Cluster B PDs, i.e. histrionic and narcissistic PD, have not been considered in drug treatment research. One reason for this might be that their pronounced interpersonal pathology is regarded as more suitably approached via psychotherapeutic treatment. In addition, individuals affected are more likely to seek treatment because of comorbid disorders, especially depression, or during crisis.

2.2. Cluster A (odd, eccentric PDs): schizotypal PD as a candidate for drug treatment

In contrast to paranoid and schizoid PD, schizotypal PD has been subject to some drug treatment research. It is regarded as a schizophrenia spectrum disorder due to an elevated risk of developing schizophrenia and shared phenomenologic, genetic, and neurobiological features. Furthermore, as in schizophrenia, preliminary research suggests a dysregulation of the dopaminergic system and responsiveness to antipsychotic treatment. To date, there are three placebo-controlled RCTs indicating therapeutic effects of the second-generation antipsychotic risperidone on self-monitoring, executive functions, and positive as well as negative schizophrenia symptoms [10]. Additional small, placebo-controlled experimental studies report positive effects on working memory after administration of dopamine receptor agonists (pergolide, dihydrexidine), an α₂-agonist (guanfacine) and a cholinesterase inhibitor (physostigmine) [11].

The reason why research on paranoid and schizoid PD, in general, is rare might be that these patients tend not to seek treatment because of disorder-inherent symptoms like suspiciousness (paranoid PD) or decreased social motivation (schizoid PD). Nevertheless, the burden due to comorbid conditions such as depression or anxiety-related disorders is often high and may be subject to respective drug treatment.

2.3. Cluster C (anxious, fearfulPDs): no attention to drug treatment yet

Cluster C PDs, including avoidant, dependent, and obsessive-compulsive PD, have not been subject to substantial research activities on the effects of drug treatment so far. Due to shared phenomenology and underlying biological features, social phobia has been used as a model for avoidant PD treatment, calling for cognitive-behavioral psychotherapy as first-line treatment. The remaining Cluster C PDs have also not been regarded as candidates for drug treatment possibly due to perceived interpersonal problems that are primarily subject to psychotherapy. However, individuals affected might seek treatment primarily for comorbid mood disorders.

3. Conclusion

In conclusion, sadly we have to come to the same conclusion as deBattista and Glick 25 years ago [2]: pharmacotherapy cannot be regarded as a ‘cure’ for PDs. Nevertheless, it has a role in the treatment of severe comorbid disorders, especially mood and anxiety disorders that are highly prevalent in individuals with PDs. However, routine use of medication in individuals with PD (still 82% of individuals with a sole diagnosis of PD are prescribed at least one psychotropic medication [3]) is not supported by the evidence and might reflect the urge felt by clinicians to respond immediately to desperate situations, while the access to better-suited specialist psychotherapy is limited [12]. Better availability of specialist would very likely decrease the overuse of medication [13]. As pronounced placebo effects have been observed in BPD samples [14], the role of paying consistent and reliable attention to individuals as a beneficial treatment component, which is sadly oftentimes not a given, deserves further attention [13].

4. Perspective

Currently, the PD field is facing a shift from categorical to dimensional classifications. Besides the 10 PDs, DSM-5 includes a hybrid diagnostic system, the so-called ‘alternative system’ in Section III of the manual [1]. Likewise, the impending ICD-11 classification will waive the concept of distinct PDs and introduce five personality traits (negative affectivity, detachment, dissociality, disinhibition, and anankastia) that can individually be pronounced to a lesser or greater extent (mild to severe) [15]. Only BPD will be kept as a distinct PD pattern, while its diagnostic criteria parallel the corresponding DSM-5 criteria.

It remains to be seen if, how, where, and when this quite radical transformation of the diagnostic system will be adopted in clinical settings, and if the research will embrace these new concepts or continue to be guided by the conventional PD diagnoses of DSM-III to DSM-5. In any case, we observe ‘research fatigue’ in the field of PD drug treatment research. Even for the so far most studied antisocial and borderline PD, it seems to stagnate, while the psychotherapy evidence is rapidly accumulating. As effective specialist psychotherapies are now available [16], the role of pharmacotherapy has diminished. However, the introduction of dimensional diagnostic models may stimulate pharmacological research again using a transdiagnostic perspective of targeting symptom clusters within and across disorders.

5. Expert opinion

The revision of diagnostic systems in the field is timely as out of the conventional 10 different PDs, only few have shown relevance in routine clinical settings as well as academically, not least because of the overlap of symptoms across PDs and hence co-morbidity even among PD diagnoses. At the same time, an atmosphere of confusion and reluctance to refrain from familiar diagnostic entities have evolved from the publication of two different diagnostic models in DSM-5 (though only the conventional categorical model is in effect). ICD-11 will either exacerbate the debate between proponents of categorical and dimensional models or simply coexist but be ignored by international research, as it was the case for ICD-10 emotionally unstable disorder (F60.3) that was in practice used synonymously with the DSM diagnosis of BPD, despite not being exactly the same.

Nevertheless, the dimensional models present an opportunity to guide research from a transdiagnostic perspective and finally make progress in the search for effective drug treatments at least for specific PD domains. New substances are now subject to experimental RCTs [5], specifically, drugs targeting glutamatergic, cholinergic, or glucocorticoid systems, but also ketamine for which rapidly antidepressive but time-restricted effects have been observed in treatment-resistant depression. A substantial body of research has accumulated regarding the effects of oxytocin in BPD, but longer-term, randomized intervention studies are not available yet.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This manuscript has not been funded.