A proposed anti-maladaptive aggression agent classification: improving our approach to treating impulsive aggression

ABSTRACT

Proper drug categorization enables clinicians to readily identify the agents most appropriate for patients in need. Currently, patients with maladaptive aggression do not all always fall into a single existing diagnostic or treatment category. Such is the case for those with impulsive aggression (IA). IA is an associated feature of numerous neuropsychiatric disorders, and can be described as eruptive, aggressive behavior or a ‘short fuse’. Although agents from a broad spectrum of drug classes have been used to treat maladaptive aggression, few have been tested distinctly in patients with IA, and there is no drug specifically indicated by the US Food and Drug Administration (US FDA) for IA. Further, current treatments often fail to sufficiently treat IA symptomatology. These issues create an unclear and inadequate treatment path for patients. Here we will propose the establishment of a class of anti-maladaptive aggression agents to begin addressing this clinical issue. The development of such a class would unify the various drugs currently used to treat maladaptive aggression and streamline the treatment approach towards IA. As an important case example of the range of candidate drugs that could fit into a new anti-maladaptive aggression agent category, we will review an investigational IA pharmacotherapy. SPN-810 (extended-release molindone) is currently being investigated as a novel treatment for children with IA and ADHD. Based on these studies we will review how SPN-810 may be well suited for a new, anti-maladaptive aggression drug class and more precisely, a proposed subgroup of IA modulators. The goal of this review is to begin improving the identification of and therapeutic approach for maladaptive aggression as well as IA through more precise anti-maladaptive aggression agent categorization.

KEYWORDS:

• Impulsive aggression
• maladaptive aggression
• impulsive aggression modulators
• anti-maladaptive aggression agents
• drug classification
• ADHD

1. Introduction

The proper categorization of drugs is one key step in identifying the therapeutic agents best suited to address a patient’s symptomatology. Achieving this goal streamlines treatment by allowing the right group of drugs to be readily identifiable by the clinician for the right group of patients [1]. An issue in properly following this clinical practice arises when treating aggression. Aggression constitutes one of the most common reasons children are referred to psychiatric clinics [2–4]. While aggression can be adaptive to one’s environment and even have a positive impact on the aggressor when it occurs in the appropriate social context, maladaptive aggression is harmful to the acting individual and has negative consequences [5]. Various subtypes of maladaptive aggression can be found in the clinical setting and should be characterized prior to treatment [6,7]. Distinguishing the form of maladaptive aggression prior to treatment is paramount to effectively treating aggressive symptoms, as subtypes of maladaptive aggression vary in their therapeutic responsiveness and therefore may require different management [8]. For example, predatory maladaptive aggression (a subtype of proactive aggression) is diagnostically and therapeutically different from unplanned, reactive aggression [8]. Predatory aggression, for example, is less responsive to pharmacotherapy [8] but may benefit from behavioral intervention [7,8]. While proactive and reactive aggression can be distinguished by available assessment tools, thus allowing a patient’s symptoms to be attributed to an appropriate therapeutic category, the lines between aggressive subtypes that fall within these two broader categories can be blurred by overlapping features [7,9,10]. For example, both overt aggression and impulsive aggression (IA) occur in response to a stimulus, and both hostile aggression and IA are characterized by ‘hot’ or ‘angry’ behavior, yet each of these forms of aggression has unique attributes making them three distinctive subtypes [5,11,12]. The difficulty in distinguishing between overlapping aggressive subtypes can, therefore, pose a significant challenge for clinicians, as ill-defined diagnoses can lead to non-specific treatment recommendations. This interplay between diagnostic ambiguity and the lack of distinct treatment categories can obfuscate a clear path forward for the treatment of maladaptive aggression subtypes and cause frustration for clinicians, parents, and patients [5,7].

Defining the form of disruptive behavior exhibited by patients is central to connecting a condition with evidence-based pharmacotherapy. This issue surrounding diagnostics and treatment is perhaps most challenging in the case of IA, which accounts for ~80% of the aggression cases seen in pediatric clinical practices and can lead to life consequences continuing through adulthood [5]. An ongoing obstacle in the identification of effective treatments for IA is the lack of an associated drug classification – there is no specific drug indicated for IA symptomatology [5]. For the purposes of this review, we are defining IA as an eruptive form of maladaptive aggression that is reactive, overt, has been described as a ‘short fuse,’ and occurs as an exaggerated response outside the social context [4,5,13–15]. IA is not yet considered by all clinicians to be a distinct, identifiable condition – and what constitutes a ‘condition’ versus a ‘symptom’ can be a challenge when considering behaviors such as IA and other forms of aggression. IA can be conceptualized as an associated feature observed in several neuropsychiatric disorders across age groups, akin to how pain or fever is a symptom of many illnesses [4,14,16,17]. Much like a fever, once recognized, it is the task of the clinician to diagnose the underlying cause and identify the best treatment path forward [14]. As IA is not yet well-defined in the clinical setting, it is often mistaken for irritability and is frequently diagnosed in the context of other disorders such as intermittent explosive disorder (IED), conduct disorder (CD), disruptive mood dysregulation disorder (DMDD), and oppositional defiant disorder (ODD) [5]. These limitations of the current nosology have been acknowledged for neuropsychiatric disorders in the broader context as well, and the field is undergoing a shift in our approach to addressing these issues. This shift is highlighted by initiatives such as the National Institute for Mental Health (NIMH)’s Research Domain Criteria or RDoC, which does not seek to establish a new diagnostic guide, but provides a unified framework for investigating the key dimensions comprising neuropsychiatric behavior and disorders in a symptom-oriented manner [18,19].

The IA symptom cluster, for example, is difficult to distinguish from the aforementioned behavioral disorders using current diagnostic methods and nosology. Historically, this diagnostic issue is further complicated by the existence of distinct subtypes of behavioral disorders. The two most recent editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-Text Revision and DSM-5) distinguish between the age of onset of CD, for example, wherein CD can be subdivided into childhood-onset CD or adolescent-onset CD [20,21]. These subtypes of CD as well as ODD can each describe repetitive, persistent disruptive and aggressive behavior, yet also encompass planned and/or predatory aggression, which are not components of IA [20]. Similarly, children with IA can be irritable – irritability is defined as having a low threshold for responding angrily to frustration, and in turn, the DSM-5-defined disorder DMDD is characterized by chronic irritability along with temper outbursts [22,23]. Children with IA, therefore, could be described as having temper tantrums much like children with DMDD, but do not fit this diagnosis perfectly either as they may also express their aggression physically and in an intermittent, rather than chronic, manner [20,22]. The intermittent and non-predatory nature of IA make this disorder very similar to IED, yet IED is not broken into subcategories based on age of onset, as has been necessary for CD, and currently is defined under a blanket age of onset for anyone 6 years of age or older (and does not cover those younger than 6) [20]. In the case of CD, field trials have demonstrated that the developmental age of onset can affect the symptoms, level of severity, and functional trajectory of the patient – an important distinction that can apply to children and adolescents with IA as well [4,24–26]. Therefore, while children and adolescents with IA have facets of DMDD or facets of IED, they do not fall perfectly into any of these categories, leaving these patients in a diagnostic gray area difficult to define and treat. This is a common problem inherent in many mood disorders – indeed, it has been reported that DMDD is not reliably distinguishable from ODD and CD, highlighting the deficiencies in our current approach to defining and treating behavioral and mood disorders [23]. This chronic lack of diagnostic definition for children and adolescents negatively affects psychosocial functioning and developmental outcomes for this population [4].

How then can our diagnostic and therapeutic approach to IA be improved? The study of IA may benefit from further investigation and differentiation much like was the case for bipolar disorder and irritability; in patients with bipolar disorder, efforts were made to distinguish chronic irritability from episodic irritability and bipolar disorder, and resulted in a systemic methodology for parsing out the overlapping symptoms of bipolar disorder versus irritability, as well as improved treatment recommendations for each condition [22,27]. Here we are proposing that the same process may benefit patients with IA through improved identification of IA as a condition, followed by improved classification of therapeutics to better help patients. As patients with IA do not fit perfectly into any diagnostic criteria currently defined in the DSM-5, they also do not fall into an associated drug classification, and there is no specific drug indicated for IA symptomatology [5]. Misattributing IA behaviors as manifestations of attention-deficit/hyperactivity disorder (ADHD), IED, CD, DMDD, or ODD can cloud the distinguishable, aggregate set of behaviors comprising IA, and may preclude precise treatment [28,29]. The current practice for the treatment of IA symptoms begins with treating the primary disorder; however, this does not result in an adequate treatment response for all patients [30]. In the context of ADHD, IA remains persistent in some patients even after treatment with ADHD medication and nonpharmacologic therapy [28,29]. Identification of IA within a primary disorder may direct clinicians to choose a treatment regimen that differs from one that might be followed if they were treating the primary disorder alone. Further, the fact that a subset of patients with IA remains symptomatic following treatment with stimulant or nonstimulant therapy suggests that the IA symptom cluster may represent a unique therapeutic target. Many reviews and clinical studies targeting aggressive behaviors have discussed and focused on patients with maladaptive aggression, but the lack of a clear definition of IA based on established diagnostic tools has led to very few of these studies specifically investigating patients with the IA subtype as an inclusion criterion [4]. Exacerbating this, those studies investigating aggression typically exclude patients with a history of abuse, trauma, post-traumatic stress disorder, or other family histories of mental illness, limiting our understanding of IA risk factors and epidemiology. In turn, IA’s obscure definition has subsequently stunted the development and identification of treatments best suited for patients with IA.

As there currently is no therapeutic category to fully accommodate patients with IA, a step toward improving the treatment of these patients may be to create an anti-maladaptive aggression agent category to streamline treatment for patients struggling from childhood through adulthood [5]. Given these challenges in both diagnosing and treating maladaptive aggression such as IA, in this review, we will describe how the establishment of an anti-maladaptive aggression class of drugs may unify the diverse group of agents currently used to treat aggression. Such changes to the classification of therapeutic agents targeting aggression could help stimulate improvements in how we currently diagnose and treat this group of patients. Second, as few studies have been performed within a target population suffering from IA specifically, we will discuss the investigational therapy SPN-810 as one example of a potential IA modulator that may best reach patients in need if it is grouped within the larger anti-maladaptive aggression agent class proposed. As agents previously investigated for the treatment of maladaptive aggression have shown benefit in treating IA, we will discuss how other existing therapeutics may also fit into a new drug classification. Lastly, while issues surrounding IA in all age groups will be discussed, we will focus primarily on pediatric IA. It is our hope that these taxonomical changes to the therapeutic drug class will improve the treatment of IA in the clinical setting.

2. Agents used to treat aggression and the need for an anti-maladaptive aggression agent classification

Currently, there is no drug category specifically for maladaptive aggression agents, including those for IA [31]. Review of the literature indicates that agents belonging to a wide variety of drug classifications are utilized in the treatment of maladaptive aggression, including typical and atypical antipsychotics, as well as non-antipsychotic agents [5,32]. While these drugs are listed together here in this review, they are not readily available as a collective group to the clinician – preventing consistent, reliable identification of the most appropriate agents available to treat aggressive patients. IA has been identified by the US FDA as a treatable disorder [33], yet as IA symptoms can remain refractory even after treatment of the primary disorder [28], IA is not currently being addressed adequately (discussed further below). This clinical gap for those patients ‘resistant’ to current treatments suggests IA may be in part therapeutically distinct from other forms of maladaptive aggression and suggests there may be a need for the development of a drug classification specifically designed to aggregate anti-maladaptive aggression agents and strengthen the treatment approach for IA. Furthermore, because IA behavior encompasses a complex, multifaceted symptom cluster, a multimodal approach to pharmacotherapy has been reviewed as the best strategy for approaching the treatment of aggressive patients. A brief overview of the agents utilized to manage aggressive behavior is provided below; an extensive review of psychopharmacology used for aggression was published previously [34].

2.1. Typical antipsychotics

As mentioned above, one class of drugs utilized to treat aggressive symptoms is the typical antipsychotics (neuroleptics). Initially, both high- and low-potency neuroleptics (i.e. relating to the dosage required to achieve a similar therapeutic effect) were evaluated for the treatment of maladaptive aggression, with some success [32]. Low-potency neuroleptics had high rates of somnolence and other adverse events, however, and fell out of use in clinical practice [30,32,35]. Typical antipsychotics such as haloperidol showed promise but were associated with detrimental side effects such as extrapyramidal symptoms (EPS), including dystonia [36]. Further, while there are data supporting the use of haloperidol and it is indeed utilized to treat aggression, there is a clinical skepticism about whether typical antipsychotics are effective in specifically reducing maladaptive aggression [36,37]. Haloperidol, though still a commonly used antipsychotic for aggression, was recently reviewed as lacking high-quality evidence supporting its use for aggression in the context of psychosis [37]. Additionally, molindone, currently classified as a typical antipsychotic, was used in an early study in children aged 6–11 years who were hospitalized for undersocialized conduct disorder [38]. As will be discussed in more detail below, treatment with low doses of immediate-release molindone (0.5–2 mg/kg/day) significantly improved behavior associated with aggression [38]. While second-generation antipsychotics are believed to have a lower propensity for causing EPS such as tardive dyskinesia, the risk of developing such symptoms with low doses of molindone is thought to be low, except in susceptible individuals [39]. In adults with schizophrenia, for example, molindone was associated with a lower risk for tardive dyskinesia than haloperidol [40]. Further, although 23.5% of the child CD patients reported dystonic side effects with molindone, these did not include events of akathisia (which is common in adults being treated with molindone), and the low dosage administered was well tolerated [38].

2.2. Atypical antipsychotics

Amidst the growing disuse of typical antipsychotics due to their side-effect profiles, atypical antipsychotics were developed to have a reduced propensity for treatment-related EPS when compared with typical antipsychotics [30,32]. Additionally, while the primary point of action of typical antipsychotics centers on the dopaminergic system, the atypical antipsychotics generally exhibit activity toward additional targets implicated in aggression as well, including serotonin, glutamate, and histamine – making this class of agents a potentially more suitable option for multifunctional pharmacotherapy [34]. In children and adolescents, the Treatment of Maladaptive Aggression in Youth: CERT Guidelines, as well as the Treatment Recommendations for the Use of Antipsychotics for Aggressive Youth (TRAAY), provide guidance for physicians on the treatment of maladaptive aggression, including the use of atypical antipsychotics [7,30,32,41]. Consistent with treatment guidelines, atypical antipsychotics such as risperidone, aripiprazole, and quetiapine have been used to treat disruptive aggression, as has clozapine for adults with aggression in schizophrenia [42–46]. Clozapine, for example, was shown to be effective in controlling aggression in adults with schizophrenia or schizoaffective disorder in a randomized, double-blind, parallel-group study compared to olanzapine or haloperidol, yet it has not been studied in children or adolescents with aggression in the context of ADHD [46]. In children with ADHD and severe aggression being treated with parent training and optimized stimulant therapy, the addition of risperidone provided some improvements to aggressive behavior, although the therapeutic effects were moderate and varied between patients [29]. These findings were supported by modest benefits of risperidone in an extended 12-week blinded study [47]. Further, in a 12-month follow-up of the Treatment of Severe Childhood Aggression (TOSCA) study, 45% of the parents rated their children as being impaired by aggressive or noncompliant behavior, and there was no advantage detected after one year of treatment with risperidone [48].

2.3. Non-antipsychotic agents

While evidence supports the efficacy of antipsychotics toward aggressive symptoms, recent meta-analyses indicate relatively small effect sizes across the broad antipsychotic class, despite the potential for serious adverse side effects [49]. Additional treatment options are therefore important for the safe and effective treatment of IA, and success in ameliorating aggression has also been achieved with various non-antipsychotic agents (Table 1).

Table 1. Successful non-antipsychotic anti-maladaptive aggression agents.

Classification	Primary Disorder (Target Behavior)
Guanfacine	ADHD (oppositional behavior),^a ADHD (disruptive, aggressive behaviors)^b
Dexmedetomidine	Substance abuse (agitation)^c
Clonidine	ADHD (oppositional defiant disorder, conduct disorder),^d autism (aggression)^e
Amantadine	Traumatic brain injury (irritability, aggression)^f
Memantine	Alzheimer’s disease (agitation, aggression)^g
Clonazepam	Schizophrenia associated with seizure disorder (aggression)^h
Bupropion	ADHD (aggression, conduct problems)^i
Allopurinol	Dementia (aggression),^j intellectual disability (aggression),^k astrocytoma (self-injurious behavior)^k
Fluoxetine (SSRI)	Borderline personality disorder (IA)^l
Citalopram (SSRI)	ADHD, ODD, CD, major depression, social phobia, generalized anxiety disorder,^m schizophrenia (aggression)^n
Lithium (antimanic)	IA°
Divalproex (anticonvulsant)	ADHD (ODD, CD, aggression),^p CD (impulsivity, aggression)^q

ADHD = attention-deficit/hyperactivity disorder; CD = conduct disorder; GABA = gamma-aminobutyric acid; IA = impulsive aggression; NMDA = N-methyl-D-aspartate; ODD = oppositional defiant disorder; SSRI = selective serotonin reuptake inhibitor.

^aRef [50]., ^bRef [51]. ^cRef [52]., ^dRefs [53,64]., ^eRef [54]., ^fRef [55,56]., ^gRef [57]., ^hRef [58]., ⁱRef [59]., ^jRef [60]., ^kRef [61]., ^lRefs [75,79]., ^mRef [80]., ⁿRef [62]., °Ref [63]., ^pRef [67]., ^qRef [68].

The adrenergic receptor agonist clonidine has shown efficacy against conduct issues in children with ADHD [64]. The development of clonidine for treating oppositional symptoms in ADHD highlights the need for an improved understanding of our treatment approach for IA. Clonidine is efficacious towards ADHD symptoms, and it is possible that clonidine’s effect on the primary disorder could, in turn, be responsible for ameliorating aggressive behavior in children with ADHD [64]. While treating the primary disorder sometimes results in reduction of aggression, this is not the case for all children with IA [65]. For example, post hoc analysis of the Multimodal Treatment Study of Children with ADHD (the MTA study) demonstrated that 26% of the children whose symptoms are managed by ADHD medication exhibited persistent IA [4]. Such examples illustrate the issues surrounding the use of pseudo-specific treatments for IA [66]. This highlights the need for a category of anti-maladaptive aggression agents available for children not being helped by medications that fall under the indication for their primary diagnosis.

In addition to targeting IA through the management of the primary disorder, the anticonvulsant divalproex is effective in treating irritability and aggression associated with personality disorders and autism spectrum disorders [67–71]. Though few studies have targeted IA specifically, double-blind studies using the anticonvulsants oxcarbazepine and phenytoin have shown efficacy in adult patients with IA [72,73]. In the case of oxcarbazepine, the drug was originally tested for an effect on IA, as it was known to have a mood-stabilizing effect [72]. In a randomized, parallel-group, double-blind, placebo-controlled, variable-dose study, outpatients with IA associated with IED experienced clinically significant improvement in IA behavior as indicated by covariance analysis of the Overt Aggression Scale-Modified (OAS-M) scores [72]. Phenytoin has also historically shown specificity towards IA behavior in adults: in multiple double-blind, placebo-controlled studies investigating males in the general population as well as incarcerated inmates with IA behavior, phenytoin resulted in a reduction of both the frequency and intensity of aggressive episodes, but did not show improvement in premeditated aggressive acts [73,74]. Further, psychophysiological data from the double-blind, placebo-controlled, crossover study in adult males from the general population with IA suggested that the hypothesized mechanism of action on IA behavior may involve an improved ability of the patient to evaluate and react to a stimulus [73]. Importantly, while these findings support that IA neuropathology may be a distinct therapeutic target, phenytoin has not been specifically tested in youths with IA and is not typically prescribed in the pediatric clinical setting. In addition to these studies in patients with IA, lithium, antidepressants, and stimulants have also been used to treat a broader scope of aggressive behavior [75–77]. Even high-dose beta-blockers have been used to reduce problem behaviors in patients with intellectual disabilities [32,78]. These cases illustrate the potential successful application of drugs outside of the antipsychotic class in the treatment of maladaptive aggression [75,79,80].

This seemingly diverse group of therapeutics shares one unifying thread – the ability to impact aggression. We believe these agents showing anti-maladaptive aggressive effects should be labeled accordingly – as anti-maladaptive aggression agents – to improve our approach to connecting evidence-based treatments with patients in need. As such, we propose that a new, specific classification of anti-maladaptive aggression agents be created to help restructure the therapeutic approach toward IA. This approach has proven successful for symptom management in other disorders. For example, in patients with autism spectrum disorder, irritability and problem behavior were identified as distinct clinical issues for these patients [81]. This directed the classification and testing of agents indicated for children with irritability and autism spectrum disorder and has led to treatments that have improved such behaviors within this population [81]. Further, the current approach to first treat aggression with drugs targeting the primary disorder has been called into question for other disorders [66]. In the case of cognitive impairment associated with depression, the initial treatment approach assumed cognitive impairment was a result of the depression, and therefore, drugs targeting depression had been used to combat symptoms of cognitive dysfunction [66]. Similarly, even following management of positive symptoms in schizophrenia, cognitive impairment is a symptom that often persists – indicating that cognitive impairment may be a distinct therapeutic target [66]. Such could be the case for IA, where a subset of patients remain symptomatic despite treatment of the primary disorder [4,65]. Recognition of this issue and the potential limits of treatments with a pseudo-specific effect may help minimize some of the regulatory roadblocks inhibiting the development of treatments specific to the symptoms experienced by patients, as it has in this example of cognitive impairment in schizophrenia [66].

2.4. Investigational IA modulator therapies

As mentioned above, there is currently no health authority (e.g. US FDA, European Medicines Agency [EMA], International Society for Pharmacoeconomics and Outcomes Research [ISPOR])-approved medication for the treatment of IA [82,83]. Given that there are no treatment guidelines specifically for IA, therapeutics targeted toward generalized aggression are utilized in the treatment of IA. SPN-810 (extended-release molindone) is currently in development as a novel treatment for IA in patients being treated for ADHD [33]. Although treating aggressive symptoms is not novel, the specific testing of a therapeutic agent in patients with IA has rarely been performed. Previous studies have tested phenytoin and oxcarbazepine (described above) in adult patients with IA behavior defined using the DSM-based criteria for IED [72,73]. Most clinical trials assessing aggressive behavior include patients with generalized aggression; however, the development program for SPN-810, specifically the Phase 3 studies, is unique in that it is evaluating a large targeted population of patients with IA (trials NCT02618408, NCT02618434, and NCT02691182). As IA is more treatment-responsive than other forms of maladaptive aggression [8], future trials such as these targeting individuals with IA, as opposed to primarily predatory aggression, will be important in identifying a new category of anti-maladaptive aggression agents in which IA modulators such as SPN-810 fit. While the pathophysiology of IA is not fully understood, it is thought that the neurological etiology of IA behavior is similar to that of reactive aggression [4,15]. One hypothesis suggests that IA and reactive aggression may result from an impairment in decision-making associated with reduced frontal-cortical control of top-down processing, which, in turn, contributes to disinhibition of the threat-response circuit [15,84]. This contrasts with affective and predatory aggression, which, based on animal models, are thought to be associated with hypothalamic control [85]. Additionally, the RDoC approach has been applied to describe the developmental basis of aggression, with evidence suggesting that aggressive behavior is a result of complex interactions between dysfunctional domains (such as the negative valence domain encompassing fear or sustained threats, compared to the positive valence domain encompassing reward responsiveness), which culminate in a lack of behavioral adaptability within interpersonal interactions [19]. Hence, distinctions between anti-maladaptive aggression agents and IA modulators may help further improve the identification of drugs best targeted toward the various subtypes of aggression, as well.

2.4.1. A case example of an IA modulator

Immediate-release molindone was initially approved for the treatment of schizophrenia under the name Moban® at doses of 100–225 mg/day but has been used at doses as high as 600 mg/day [86–88]. As mentioned above, immediate-release molindone at low doses (0.5–2 mg/kg/day) demonstrated some improvements to aggressive behavior in children with undersocialized conduct disorder, setting the stage for future studies investigating the action of molindone towards aggressive behavior [38]. A proof-of-concept study demonstrated that treatment with low-dose (5–40 mg/day) immediate-release molindone produced dose-related improvements in aggressive behaviors when added to stimulant monotherapy in children aged 6–12 years with ADHD and persistent serious conduct problems [39,89].

Improvements in IA behaviors were seen with low-dose SPN-810 treatment in a Phase 2b, randomized, double-blind, placebo-controlled, parallel-group, dose-ranging study and its open-label extension in children aged 6–12 years diagnosed with persistent IA and ADHD [83,89]. In this study, SPN-810 was well tolerated, with headache, sedation, and increased appetite being the most common adverse events occurring in approximately 10% of the patients [83,89]. The safety and efficacy of SPN-810 were further sustained in a 6-month follow-up study and has supported ongoing Phase 3 studies (trials NCT02618408, NCT02618434, and NCT02691182) [83]. Importantly, as few trials have investigated therapeutics specifically in IA patients, future trials would be informative in establishing the most efficacious agents and most appropriate sequence of treatment for IA behavior.

While the exact mechanism of SPN-810 is unknown, emerging data suggest it functions as an antagonist of the dopamine receptors D_2S and D_2L and the serotonin receptor 5-HT_2B [90]. Although it remains to be confirmed, given these activities and the purported neuropathology of IA, it is suggested that SPN-810 may function by modulating the dopamine limbic pathway and its associated neural networks [15,90,91]. Further, the reduced incidence of EPS at therapeutic doses used for IA suggests that the potency of SPN-810 achieved at low doses is different from that of doses used for schizophrenia, which are highly associated with EPS [92]. This suggests that while immediate-release molindone is classified as an antipsychotic, it may not be accurate to place low-dose (5–40 mg) SPN-810 in this category if the drug could be more precisely described as an anti-maladaptive aggression agent that improves IA behavior. Future studies such as neuroimaging analysis of children and adolescents with IA may help shed further light on the neurobiological etiology of IA, and how specific neurological pathways leading to these behaviors may be best targeted. In IED, for example, white matter abnormalities between the frontal and temporoparietal brain regions have been identified, improving the diagnostic distinction of this disorder [93]. Such studies might not only improve our therapeutic approach to IA but also may help provide further diagnostic criteria for IA etiology.

Studies such as these raise questions about our current approach to classifying drugs used for aggression in general. Importantly, if, when used at low doses, drugs currently designated to the antipsychotic classification do not achieve a therapeutic antipsychotic effect, but could (e.g. in the case of SPN-810) provide therapeutic benefit for aggressive behavior, are we best serving patients by retaining the antipsychotic classification? Classification of a drug as an antipsychotic provides important information about the appropriate patient population for the drug and the safety profile associated with the drug; however, if low doses of this same drug are best suited for a different patient population, a different condition, and are associated with a different safety profile, perhaps such a drug could be further and more precisely classified in an additional category outside the antipsychotic class. Findings such as those described above support the classification of SPN-810 as an IA modulator when used at low doses for the treatment of IA. Furthermore, SPN-810 for the treatment of IA would then also be appropriate for the larger anti-maladaptive aggression agent class proposed. Further still, it remains possible that other antipsychotic agents used at low doses may have a similar anti-aggressive effect and could also perhaps be better grouped within an additional drug classification (i.e. anti-maladaptive aggression agents) at specified doses. Such changes may improve how we approach the treatment of in-need patients suffering from aggressive disorders such as IA.

3. Conclusions

There currently is a population of patients with persistent IA whose symptoms do not fall into any existing therapeutic or specific diagnostic category (e.g. aggression/irritability in autism or aggression/irritability in ADHD). One of the reasons for this clinical gap may be the absence of a drug category specifically identifying anti-maladaptive aggression agents. The success of numerous agents currently grouped within various drug categories builds the case that it may be sensible for these drugs targeting maladaptive aggression to be classified as a separate, unique category. This could help refine the clinician’s treatment approach towards patients in need and may improve patient outcomes. As just one example of an ongoing investigational therapy, at low therapeutic doses (5–40 mg, thus far investigated in pediatric populations), SPN-810 may be described as an IA modulator rather than an antipsychotic, and may, therefore, be appropriate for classification as an anti-maladaptive aggression agent. It may also be important to investigate a more precise classification for other antipsychotics in the future, as it is currently unknown if other antipsychotics possess an anti-aggressive effect with an improved safety profile at low doses. Such investigations may increase the number of agents available to IA patients in need. In the case of SPN-810, this drug was also specifically tested and shown to be efficacious in patients with aggression that was impulsive in nature, and would, therefore, be ideally classified as an IA modulator, given the known difference in treatment responsiveness between subtypes of maladaptive aggression [8,83,89]. For example, it is unknown if SPN-810 is effective in treating predatory aggression, which is less responsive to pharmacotherapy [8] and may not be an appropriate target for SPN-810 or other anti-maladaptive aggression agents best classified within an ‘impulsive aggression modulator’ category [8]. Such improvements could begin to shift our treatment approach toward better targeting the underlying neurological basis of IA behavior as distinct from the primary disorder. Such a classification may also apply then for other drugs that have been tested specifically in adult patients with IA, such as oxcarbazepine [72]. Future work towards defining the nosological criteria of IA as a distinct entity should facilitate both the study of as well as the generalizability of research results when agents with putative effectiveness, such as SPN-810, are studied in the treatment of maladaptive aggression. Distinct drug classifications for anti-maladaptive aggression agents such as these may prevent improper treatment and improve the application of available therapeutics to match the right therapeutics more readily with the appropriate patients. Therefore, the addition of a specific class of anti-maladaptive aggression agents may improve the identification and application of therapeutics targeted toward maladaptive aggression, and represent an important step in improving the treatment of IA.

Declaration of interest

A. Robb discloses the following: Actavis/Forest Laboratories (consultant, research support, travel support); Aevi Genomic Medicine Inc. (data safety monitoring board); AACAP (honorarium, travel support); AAP (honorarium, travel support); Bracket (consultant); Case Western Reserve (honorarium, travel support); Child and Adolescent Psychiatric Society of Greater Washington (honorarium); College of Neurologic and Psychiatric Pharmacists (honorarium, travel support); Eli Lilly (royalties); GlaxoSmithKline (royalties); Guilford Press (royalties); Johnson & Johnson (royalties); Lundbeck/Takeda (advisor, research support, travel support); Neuronetics (data safety monitoring board); Neuroscience Education Institute (honorarium, travel support); Nevada Psychiatric Association (honorarium, travel support); National Center for Advancing Translational Sciences (research support); NICHD (advisor); NIMH (data safety monitoring board); NINDS (research support); NACCME (honorarium, travel support); Pfizer Inc. (research support, stock/equity, travel support); Sunovion Pharmaceuticals Inc. (advisor, travel support); Supernus Pharmaceuticals, Inc. (research support); SyneuRx (research support); University of Cambridge (advisor). S. Schwabe, G. Ceresoli-Borroni, A. Nasser, C. Yu, R. Marcus and S. A. Candler are employees of Supernus Pharmaceuticals, Inc. R. L. Findling discloses the following: Aevi (research support, consultant); Akili (research support, consultant); Alcobra Pharma (research support, consultant); AACAP (speaker’s bureau); Amarex Clinical Research (consultant); American Psychiatric Press (royalties); Bracket (consultant); ePharmaSolutions (consultant); Forest Laboratories, Inc. (research support); Genentech (consultant); Guilford Press (consultant); Ironshore Pharmaceuticals & Development, Inc. (consultant); Johns Hopkins University Press (royalties); KemPharm, Inc. (consultant); Lundbeck Inc. (research support, consultant); Merck & Co., Inc. (consultant); National Institutes of Health (research support, consultant); Neurim Pharmaceuticals, Inc. (research support, consultant); Nuvelution (consultant); Otsuka America Pharmaceutical, Inc. (consultant); PCORI (research support); Pfizer Inc. (research support); Physicians Postgraduate Press (consultant); Purdue Pharma L.P. (research support); Roche (research support); Sage Therapeutics (royalties); Shire (research support); Sunovion Pharmaceuticals Inc. (research support, consultant); Supernus Pharmaceuticals, Inc. (research support, consultant); SyneuRx (research support); Teva Pharmaceutical Industries Ltd. (consultant); TouchPoint (consultant); Tris Pharma, Inc. (consultant); Validus Pharmaceuticals LLC (research support, consultant); WebMD (consultant).

Editorial support was provided by IMPRINT Science, New York, NY, USA, and was funded by Supernus Pharmaceuticals, Inc. The authors have no other 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.

Additional information

Funding

This study was funded by Supernus Pharmaceuticals, Inc.