Improvement of nonsuicidal self-injury following treatment with antipsychotics possessing strong D1 antagonistic activity: evidence from a report of three cases

Introduction

Nonsuicidal self-injury (NSSI) is defined as self-injurious behavior without the intent to die [Nock and Favazza, 2009] and occurs in different mental disorders, especially in posttraumatic stress disorder (PTSD) [Dixon-Gordon et al. 2014], borderline personality disorder (BPD), and major depression (MD) [Hawton et al. 2013]. While in DSM-IV NSSI was considered to occur almost exclusively in patients suffering from BPD, in DSM-5 [American Psychiatric Association, 2013] this symptom is understood as a distinct condition since it has been found to be associated with numerous psychiatric symptoms including suicidality, depression and anxiety [In-Albon et al. 2013]. NSSI can take the form of inter alia cutting, burning, scratching, biting, intoxicating, and head-banging and is ‘the most frequent reason for psychiatric visits to medical emergency departments’ [Stanley et al. 2010]. Despite the availability of elaborate psychotherapeutic NSSI interventions [Hawton et al. 1999], this burdensome symptom, which is associated with an increased risk for suicide attempts, is often treatment refractory [Stanley et al. 2010]. Inter alia, antidepressants, especially serotonin reuptake inhibitors, modulators of the endogenous opioid system and various other drugs such as oxcarbazepine were discussed to be potentially effective in NSSI therapy [Stanley et al. 2010; Hawton et al. 1999; Cordás et al. 2006]. In addition, interestingly, there are several reports on the efficacy of atypical antipsychotics (AAs) in NSSI treatment. For instance, we found that clozapine [Chengappa et al.1999; Hammock et al. 2001], olanzapine (OZP) [Hough, 2001], quetiapine (QTP) [Good, 2006], ziprasidone [Libal et al. 2005], and risperidone [Cohen et al. 1998], were reported to be effective in the treatment of NSSI occurring in the context of different psychiatric disorders. However, all these studies did not go beyond the level of case reports. To the best of the authors’ knowledge, there is so far only one randomized controlled trial (RCT) that assessed the efficacy of medication for reduction of NSSI [Turner et al. 2014]. This RCT found aripiprazole to be effective in reducing NSSI behavior [Nickel et al. 2006]. On the other hand, Ruedrich and colleagues stated that AAs improve ‘aggression, but not self-injurious behavior, in adults with intellectual disabilities’ [Ruedrich et al. 2007].

Taken together, the synopsis of the few existing reports on the effects of antidepressant treatment on NSSI has not yielded promising results and the studies assessing opioid receptor-targeting drugs and AAs for NSSI are yet too few to draw any conclusion. This evident lack of drugs for NSSI treatment and the still unclear pathogenesis of NSSI motivated us to systematically screen medical records for cases that demonstrate a successful drug treatment of NSSI behavior.

Methods

Study design

Systematical screening of medical records from inpatients treated on the closed ward of the Max Planck Institute of Psychiatry (MPI-P) in Munich, Germany, between 2009 and 2010 (about 290 adult patients with various psychiatric diagnoses per year, about 5% of them with NSSI) for cases demonstrating a successful drug treatment of NSSI revealed the three cases presented here (Figure 1). All cases with inconsistent documentation of NSSI behavior and/or with parallel application of psychotherapeutic interventions for NSSI were excluded. All three patients presented in this retrospective case series are adult Caucasian females living in Upper Bavaria. All of them did not receive an NSSI-specific psychotherapy and had no relevant psychiatric or somatic comorbidities.

OPEN IN VIEWER

Figure 1.

Persistent remission of nonsuicidal self-injury was associated with the administration of neuroleptics possessing strong D1 receptor antagonistic activity. Shown are the courses of pharmacotherapy in three adult female patients (a–c) suffering from NSSI associated with major depression. The x-axis indicates the day of treatment. Above the x-axis, the frequency of NSSI (times per day, black bars) is shown. The y-axis indicates the doses of the different drugs administered (in mg/day). Graphs were created using the IGOR Pro 6® software (WaveMetrics, Portland, OR, USA). Flupentixol is highlighted in red. Chlorprothixene is highlighted in blue. Zuclopenthixol is highlighted in green. AMI, amitriptyline; CPT, chlorprothixene; d, day; FPX, flupentixol; GPN, gabapentin; LTG, lamotrigine; LZP, lorazepam; OZP, olanzapine; QTP, quetiapine; TPM, topiramate; TRI, trimipramine; VLX, venlafaxine; ZPT, zuclopenthixol.

As this was a retrospective study, ethics committee approval was not required.

Results

We identified only five cases allowing the conclusion of a successful drug treatment of NSSI, two of which had to be excluded due to inconsistent documentation.

Discussion

The three cases presented here suggest that antipsychotics with D1 antagonistic activity reduce NSSI behavior: NSSI remitted after treatment with ZPT, the strongest D1 receptor antagonist currently approved for clinical use [Hyttel et al. 1989], both in case 1 (Figure 1a) and in case 3 (Figure 1c). In both of these cases worsening of NSSI occurred in parallel to treatment with VLX. This is in line with the here-suggested dopamine hypothesis of NSSI, as, in higher dosages (300 mg/day in case 1 and 450 mg/day in case 3), VLX is known to inhibit dopamine reuptake [Bourin, 1999; Lemke, 2007]. However, in case 3, we cannot fully exclude that TRI contributed to remission of NSSI, as TRI and ZPT were started in parallel (Figure 1c). In addition, in case 2 (Figure 1b), dosage reduction of CPT (Figure 1b, days 19–23), the third strongest D1 receptor antagonist currently approved for clinical use [Hyttel et al. 1989], together with the application of an insufficient dose of the second strongest D1 antagonist [Hyttel et al. 1989], FPX, as well as complete discontinuation of CPT prior to elevation of FPX up to 18 mg/day (Figure 1b, days 32–33), were associated with an increase in NSSI frequency. Most important, in the same case, these fluctuations in NSSI behavior occurred independently of the constant increase in the dosage of GPN suggesting that sedating GPN is not involved in NSSI reduction. Moreover, the fact that in case 1 (Figure 1a) sedative and antipsychotic treatment with LZP, QTP, and OZP had no effect on NSSI, also supports the conclusion that medication-elicited reduction of NSSI is independent from sedative effects, especially as FPX has no sedative properties.

We found one case series, published 36 years ago, that also reported significantly reduced rates of parasuicidal acts upon treatment with FPX [Montgomery et al. 1979]. 15 years ago, Evans stated that ‘so far no one has attempted to replicate’ this finding of Montgomery and colleagues [Evans, 2000]. Thus, to the best of the authors’ knowledge, the case series presented here is the second clinical paper suggesting a role for strong D1 antagonists in NSSI drug therapy and for the D1 receptor in the pathogenesis of NSSI in humans. Interestingly, and in accordance with our here-presented findings, ZPT has been described to effectively reduce auto-aggressive behavior in mentally retarded patients [Häßler and Reis, 2010], but however, has, to the best of our knowledge, hitherto not been shown to reduce auto-aggression in patients without mental retardation. The fact that our results are also in agreement with findings in rodents - for example, the synthetic D1 receptor antagonist SCH23390 effectively alleviated self-injury in 6-hydroxydopamine lesioned rats [Sivam, 1989] - further supports the role of D1 receptors in the pathogenesis of NSSI.

Interestingly, some of the AAs that have been described to effectively reduce NSSI (see introduction), are also known for exerting D1 antagonistic properties, even though, in comparison with ZPT, FPX, and CPT, only to a small extent (for example, clozapine [Gerlach et al. 1996] and OZP [Zhang and Bymaster, 1999]). However, in turn, the fact that other classes of drugs with no or no significant D1 antagonistic activity were described to also reduce significantly NSSI [Stanley et al. 2010; Hawton et al. 1999; Cordás et al. 2006], strongly suggests that not only D1 receptors but other neurotransmitter receptors/systems are involved in the assumingly complex pathobiology of NSSI.

Since therapeutic recommendations must not be derived from case series but are usually based on meta-analyses of RCTs, there is a strong need for performing RCTs to identify effective medications for NSSI treatment. Our case series wants to stimulate corresponding RCTs and, moreover, to promote studies that evaluate the effectiveness and molecular mechanisms of D1 receptor antagonists in animal models of self-injury. The identification of an effective ‘anti-auto-aggressant’ would undoubtedly be a significant relief for those suffering from NSSI. However, the therapeutic benefit of any treatment should always be weighed against its potential side effects - in the case of antipsychotics, mainly against the risk of developing acute extrapyramidal side effects and tardive dyskinesia [Haddad et al. 2012; Horácek, 2000].