JOURNAL OF CHILD AND ADOLESCENT PSYCHOPHARMACOLOGY Volume 26, Number 1, 2016 ª Mary Ann Liebert, Inc. Pp. 65–73 DOI: 10.1089/cap.2015.0167
Pharmacotherapy of Aggression in Child and Adolescent Psychiatric Disorders Tina Gurnani, MD,1 Iliyan Ivanov, MD,1,2 and Jeffrey H. Newcorn, MD 2
Abstract
Objective: Aggression is a common, yet complex, behavioral complaint, and a frequent indication for referral to child and adolescent psychiatrist treatment. This article reviews the evidence supporting pharmacotherapy of aggression in youth, with a primary focus on impulsive aggression (the primary indication for this intervention). Relevant diagnostic considerations and consensus guidelines are discussed. Methods: Articles examining the role of medications in the treatment of aggression in youth with pathological aggression were identified using PubMed and MEDLINE databases over the past 15 years (2000–2015); selected articles published prior to 2000 and deemed to be of high relevance were searched and also included. Search terms included: Aggression, aggressive, disruptive behavior, conduct, youth, children, and adolescents. Cited references were also searched for relevant articles. Results: There are a number of evidence-based medication treatments for aggression, which are generally best considered in the context of differential diagnosis and ongoing evidence-based psychosocial interventions. Impulsive aggression is generally considered the type of aggression most amenable to medication, but other aggression subtypes may also possibly respond to treatment. Medication classes with positive evidence include the psychostimulants and a-2 agonists (in the presence of attention-deficit/hyperactivity disorder [ADHD] and/or disruptive behavior disorders), mood stabilizing agents, and atypical antipsychotics. Published guidelines recommend systematic and adequate trials of medications in sequential order, to optimize response and minimize polypharmacy. Guidelines for safety monitoring are available for many of the medications used for aggression in youth, and are also discussed. Conclusions: Aggression in children carries a high risk of poor outcomes, and, therefore, a better understanding of treatment options is a high priority. The available literature points to the importance of identifying the underlying disorder, when possible, and using this information to guide treatment selection. Future studies are needed to better inform the treatment of aggression across disorders, and the treatment of different aggression subtypes.
Introduction
persistent concerns regarding efficacy, safety, and polypharmacy (Crystal et al. 2009). Therefore, evidence-based recommendations are needed to guide implementation of appropriate therapeutic interventions. Although aggression is a behavioral construct present in all people to some extent, and can be adaptive for survival, it is highly impairing when it is pathological and persistent. Prosocial or adaptive aggressive behaviors (i.e., self-defense) are distinguished from pathological, maladaptive, or antisocial aggressive behavior. In maladaptive aggression, the intensity, frequency, duration, or severity of the aggressive response is out of proportion to its causes, and may occur in the absence of specific triggers. Pathological aggression can occur in the context of specific psychiatric disorders, or it can be a nonspecific manifestation of anger or frustration. Clinically, the different levels of aggression are commonly differentiated by a person’s motivation and that person’s ability to
P
roblems with aggressive behavior affect 10–25% of youth (Loeber and Farrington 2001), and constitute one of the most common causes for referral to child and adolescent psychiatrists. Youth with aggressive behavior often have profound disruption in a range of functional domains, including academic and occupational attainment, social and family relationships, and psychological development. Consequently, aggression is commonly a target of both psychosocial and medication treatment. However, the urgency of the condition often dictates the use of medication. Given the dearth of child and adolescent psychiatrists, a significant number of prescriptions for aggressive behavior are provided by primary care clinicians; for example, approximately one third of antipsychotic prescriptions for children during 1995–2002 were by non-mental health providers (Cooper et al. 2006). In addition, there have been
1
Department of Psychiatry, Mount Sinai-St. Luke’s Hospital Center, New York, New York. Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, New York.
2
65
66 control the behavior, and the intended goal. Impulsive aggression (IA) is generally reactive, unplanned, and overt, with the aggressor perceiving the outcome in a negative light. Planned (i.e., predatory, instrumental) aggression is often covert, and although it may be associated with a negative emotion, the aggressor anticipates a positive outcome. IA is more characteristic of clinical samples and amenable to pharmacological and psychosocial treatments; in contrast, planned aggression is more often characteristic of delinquency. However, although these descriptors are often conceptualized as dichotomous, impulsive/reactive and planned aggression can also occur in the same individual. In a recent study using an aggression scale measuring reactive/impulsive and proactive/ planned aggression, the two scores were found to be distinct but moderately correlated, with r = 0.42 (Kaat et al. 2015). In youth with planned aggression, psychosocial intervention may be useful, whereas pharmacotherapy has not traditionally been considered to be effective, unless IA is also present. Callous-unemotional traits (which are traditionally thought to be related to planned aggression) have also been studied with respect to conduct behavior problems in children, and seem to be associated with higher rates of aggression (Kahn et al. 2012), especially proactive aggression (Viding et al. 2012; Helseth et al. 2015). Different domains of aggression may be associated with distinct neurobiological mechanisms and treatment indicators. For example, in adults, IA has been shown to be related to low serotonergic function or reactivity in adults. However, data are less consistent in children, and do not offer clear directives regarding the approach to pharmacotherapy (discussed subsequently) (Newcorn et al. 2016). Risk factors for IA occur at an early age, often starting at 4 or 5 years of age (vs. 6–7 years of age for planned aggression) (Dodge et al. 1997; Vitaro et al. 2002); therefore, treatment of young children is sometimes necessary. Nevertheless, transient disruptive behaviors are common during the preschool years, so it is important to distinguish pathological aggression from developmental variations. Young boys are especially prone to aggression, and this gender bias is consistent worldwide. Girls tend to exhibit more covert aggression than boys, and less intense behavioral disturbance. Early aggression is a key predictor of later delinquency (Kotch et al. 2008); approximately half of school-age children who are aggressive continue to exhibit such behavior in adolescence. Key risk factors for aggression can be potential treatment targets. A history of childhood trauma has repeatedly been shown to predict later violent behavior, with cumulative exposure to trauma giving rise to increased risk. Sexual abuse may also be a risk factor (Harford et al. 2014). Parental separation and conflict, poverty, harsh discipline, and parental criminality are also associated features. Other risk factors include friendships with delinquent peers, living in high-crime neighborhoods, witnessing community violence, and peer rejection (especially early in childhood) (Hong and Espelage 2012). The large number of environmental, familial, and social contextual factors associated with risk for aggression underscores the importance of psychosocial treatment. Interestingly, poor peer relationships and a history of physical abuse are often specifically associated with IA, whereas aggressive role models in the family and parental incarceration have been shown to be associated with covert aggression (Newcorn et al. 2016). Low intelligence quotient (IQ) and poor language skills also constitute risk factors for early and persistent disruptive behavior (Department of Health and Human Services 2001), with the presumption that the range of adaptive skills in youth with poorly developed language skills is restricted.
GURNANI ET AL. Clinical Approach Treatment planning with aggressive youth is based on a comprehensive understanding of the nature, severity, and context of the symptoms; a confluence of psychological, behavioral, and cognitive considerations; and the profile of risk and protective factors that characterize the individual patient and family. In addition, the intensity and developmental course of symptoms must be considered, as well as the potential consequences (e.g., harm to self or others) that could arise from the aggressive behaviors and how likely this is to happen. In emergent situations, especially in which the child or adolescent has already caused injury to others, or has access to means of inflicting harm (such as firearms), inpatient hospitalization and/or acute medication intervention with sedative agents may be necessary. In the less emergent outpatient setting, there are two general approaches to treating aggression in youth: One is the symptombased approach (e.g., empirical treatment of presenting aggressive symptoms), and the other is to treat the underlying disorder (assuming one can be identified, which is often but not always the case). Current consensus guidelines recommend treating aggression in the context of specific disorders as a first step, and to develop a rational plan for pharmacotherapy in this context ( Jensen et al. 2007). The most common underlying conditions are attentiondeficit/hyperactivity disorder (ADHD) and disruptive behavior disorders (DBD) such as conduct disorder (CD) or oppositional defiant disorder (ODD), especially in younger children (Connor et al. 2002). Mood disorders and neurodevelopmental disorders are also common, as are posttraumatic stress disorder (PTSD), autism spectrum disorder (ASD), and intellectual disability (ID). Other potential underlying conditions may include anxiety disorders, tic disorders, specific learning disorders, and even adjustment disorders. Medical workup may be required to rule out exposure to toxins, seizures, infections, head trauma, and substance use. Several guidelines have been developed for the treatment of aggression in youth, based on both clinical research findings and expert consensus. The Treatment Recommendations for the use of Antipsychotics for Aggressive Youth (TRAAY) guidelines offer recommendations outlined in 14 steps. Key steps include: Psychosocial treatments as the cornerstone of first-line intervention, use of medication targeted to the primary underlying disorder as the next step (such as stimulants for ADHD, selective serotonin reuptake inhibitors [SSRIs] for depressive or anxious symptoms), avoiding polypharmacy when possible, offering a trial of an atypical antipsychotic when psychosocial measures and first-line medication have failed (starting at a low dose and titrating gradually), allowing at least 2 weeks of an antipsychotic trial at an adequate dose before considering it ineffective, use of crisis intervention strategies to minimize the need for emergency chemical and physical restraints, assessing for side effects routinely, ensuring an adequate trial before switching or combining medications, switching to a different atypical antipsychotic if the first one was ineffective, considering addition of a mood stabilizer if there was only a partial response to the first antipsychotic trial, tapering medications if an inadequate response to multiple agents is seen, and tapering of antipsychotic medication after ‡6 months of an adequate response (Pappadopulos et al. 2003). Recently, the Center for Education and Research on Mental Health Therapeutics (CERT) developed similar guidelines for primary care clinicians and mental health providers (i.e., Treatment of Maladaptive Aggression in Youth [T-MAY]). Key recommendations are to use evidence-based parent and child skills training as the first line
PHARMACOTHERAPY OF AGGRESSION intervention, treating the underlying disorder thought to give rise to the behavior, careful consideration of the potential risks of and contraindications for the initiation of medications, avoiding polypharmacy, careful monitoring of treatment response, and close medical monitoring of side effects (Scotto Rosato et al. 2012). Treatment of Aggression in the Context of Specific Psychiatric Disorders ADHD and DBD Psychostimulants are considered to be the most effective medication treatment for ADHD, with mean effect sizes for core symptoms generally ranging from 0.8 to 1.0. In addition, stimulants generally produce improvement in aggression, with positive findings seen in virtually all studies that have examined this question. One meta-analysis examined the effect of stimulants on covert and overt aggression in children with ADHD, finding overall weighted mean effect sizes of 0.84 for overt aggression and 0.69 for covert aggression (Connor et al. 2002). There is also good evidence that stimulants are effective in treating ODD/CD symptoms in children with comorbid ADHD (Kaplan et al. 1990; Gadow et al. 2008). Notably, in a placebo-controlled study of 84 youth with CD, two thirds of whom also met criteria for ADHD, Klein et al. (1997) found that methylphenidate (MPH) reduced behaviors specific to CD, independent of severity of ADHD symptoms. However, a substantial proportion of children with ADHD do not achieve sufficient reduction in aggressive behavior with stimulant treatment ( Jensen et al. 2007). Blader et al. (2009) conducted an open trial of stimulants as the first stage of a stimulant augmentation trial in children 6–13 years of age with ADHD plus ODD/CD and aggression thought to be refractory to stimulants. Of note, although all children referred to this study purportedly had inadequate response, retitration to the optimal dose using both stimulant classes produced improvement in approximately half of the children. However, even in the case of positive response to stimulants, stimulant treatment of aggression is limited by the temporal duration of the medication, because stimulants have relatively short duration of action, which is often true even with many of the current long-acting formulations. Because aggression can be present at any time of the day, treatments that last all day are highly desirable. Nonstimulant medications such as atomoxetine (ATX) and the extended release a-2 adrenergic agonists are United States Food and Drug Administration (FDA) approved for ADHD, generally have a longer duration of action than stimulants, and can be used for ADHD plus aggression, although there are no data that directly compare the effects of stimulants and nonstimulants on aggression. It is of note that clinical trials with nonstimulants have mainly focused on improvements in core ADHD symptoms, even though aggression and oppositionality are frequent targets of nonstimulant treatment (particularly with the a-2 agonists) in the clinical setting. Studies evaluating response in youth with ADHD plus ODD with both clonidine and guanfacine have demonstrated improvement (Palumbo et al. 2008; Wilens et al. 2012). Clonidine has been shown to be effective in reducing aggression in adolescents with ADHD (Hazell and Stuart 2003), and guanfacine has been shown to increase frustration tolerance and decrease irritability (Connor et al. 2010), symptoms that are frequently present in youth with IA. A meta-analysis of ATX clinical trials showed a small effect size (ES) (overall mean ES = 0.18 in 4 randomized controlled trials [RCTs]) for pediatric aggression with a dosage of 1.3mg/kg/day (Pappadopulos et al. 2006). Furthermore, ATX was found to only be
67 modestly effective in treating ODD symptoms in youth with ADHD (Biederman et al. 2007); therefore, ATX may not be a first choice recommendation. Several studies (e.g., Findling et al. 2000; Armenteros et al. 2007) have shown that antipsychotics, especially second generation agents, can be effective when used together with stimulants for aggression in ADHD, although they are not specifically approved by the FDA for this purpose and should be used with caution. In a double-blind, RCT, the Treatment of Severe Childhood Aggression (TOSCA) study, children 6–12 years of age with ADHD and ODD or CD and severe physical aggression received stimulant medication for 3 weeks, titrated for optimal effect, while parents received parent training for behavioral management. If subjects either failed to show sufficient clinical response or deteriorated, treatment was augmented with risperidone or placebo. At the end of 9 weeks, the risperidone group showed moderate improvement in aggression ratings (Aman et al. 2014). Improvement was mostly noted in impulsive, unplanned aggression and disruptive behavior, rather than callous, planned aggression. However, reevaluation of moderators of response in the TOSCA study found that higher levels of callous-unemotional traits predicted better response in disruptive symptoms regardless of treatment assignment, although it should be noted that the measurement of callous-unemotional behaviors was via an ad hoc composite variable based on several rating scales, which has not been validated (Farmer et al. 2015). Similarly, Blair et al. (2014) suggest that the use of stimulants and atypical antipsychotics in combination may be helpful for youth with CD with callous-unemotional traits, which is more difficult to treat and generally thought to have a poorer prognosis. Several open-label trials (e.g. Kronenberger et al. 2007) have been conducted with quetiapine, suggesting that this medication can be used successfully with MPH in patients with ADHD and aggression, with comorbid ODD or CD. Prescription rates of antipsychotic medication have increased in children and adolescents in recent years (Burcu et al. 2014), mainly because of their use in treating aggression and mood dysregulation. However, there are concerns regarding their longterm use, primarily because of metabolic side effects. Furthermore, concomitant use of stimulants and antipsychotics does not seem to combat the metabolic effects of antipsychotics, contrary to speculation (Linton et al. 2013). There is a long history of using mood stabilizers in youth with aggressive CD, dating back to the seminal study of Campbell et al. (1984), which compared the effects of lithium and haloperidol in 61 treatment-resistant hospitalized children. Both medications were superior to placebo in reducing behavioral symptoms, but haloperidol was more sedating. Subsequent studies of lithium have supported this finding, though ESs have been lower (e.g., Campbell et al. 1995; Malone et al. 2000). Valproic acid has also shown efficacy in ameliorating aggression in youth with ADHD, but studies are limited (Patel and Barzman 2013). A preliminary crossover study by Donovan et al. (2000) assessed 20 children with DBD with explosive temper or mood lability, who received 6 weeks of either divalproex treatment or placebo, and then 6 weeks of the reciprocal treatment. In addition, an RCT conducted with adolescent juvenile offenders with CD treated patients with either highdose (500–1500 mg/day or plasma level of 50–120 lg/mL) or lowdose (£ 250 mg/day) divalproex over 7 weeks (Steiner et al. 2003). In the first study, divalproex had superior response rates over placebo, and in the second study, high-dose divalproex yielded greater improvement in clinician ratings and in self-report of impulse control. More recently, Blader et al. (2009) evaluated the efficacy of divalproex in 27 children 6–13 years of age with ADHD plus
68 ODD/CD plus aggression. Those with persisting aggressive behavior after stimulant treatment was optimized were given either divalproex or placebo for 8 weeks, with subjects receiving behavioral therapy throughout. Children treated with divalproex had a significantly higher rate of remission on aggression ratings than those treated with placebo. It is of note that ratings of irritability and mood dysregulation were predictors of inadequate response to stimulant monotherapy. ID and developmental disorders Aggression is a frequent associated feature in youth with ID and ASD. In years past, medication for these conditions was often prescribed off label, with the most frequent medications used being the atypical antipsychotics (de Bildt et al. 2006). Because of this, and preliminary findings indicating beneficial effects of risperidone in this population (Van Bellinghen and De Troch 2001; Aman et al. 2002; Snyder et al. 2002), more definitive studies were conducted to obtain FDA approval for this practice. In 2002 and 2004, two double-blind, placebo-controlled RCTs with risperidone were conducted in children with autism and DBD, in which significant improvements in irritability were observed on the Aberrant Behavior Checklist (ABC)–Irritability subscale after an 8 week treatment period (McCracken et al. 2002; Shea et al. 2004). Similarly, Nagaraj et al. (2006) found improvements in irritability and aggression in a 6 month double-blind, placebo-controlled RCT of 39 young children (2–9 years of age). A 24 week, multisite, randomized, parallel groups trial comparing risperidone alone (subjects could be switched to aripiprazole if ineffective) to a combination of risperidone and parent training therapy, found that combination therapy was somewhat superior to medication alone, with a 14% lower dose of risperidone used at the end of the study (Aman et al. 2009). Subsequent research with aripiprazole also resulted in FDA approval for this indication. Politte and McDougle (2014) reported improved ratings of aggression, with similar or slightly lower response rates than are typically seen with risperidone. There is insufficient evidence that other antipsychotics, such as olanzapine and quetiapine, are similarly efficacious for aggression in the context of developmental disabilities, although both medications are approved for bipolar disorder in adolescents and presumably might yield improvement in aggression in these populations. Beta blockers are often prescribed off label for youth with ID, as they have shown positive results in several studies; however no RCTs for aggression have been conducted (Ward et al. 2013). Nonstimulant medications for ADHD, such as clonidine and guanfacine, have also been used, particularly when hyperactivity/ impulsivity and sleep problems are present. In addition, moodstabilizing medications have been prescribed off label for aggression in youth with developmental disabilities. Divalproex has shown some efficacy in a small double-blind, placebo-controlled study (Hollander et al. 2010) in children and adolescents 5–17 years of age with ASD; subjects who received divalproex had significantly less irritability/aggression on the ABC–Irritability subscale than those treated with placebo after 12 weeks of treatment. Mood disorders Behavior problems and aggression are frequently encountered in youth with depressive disorders, which to some extent distinguishes depression in youth from depression in adults. In a longitudinal study by Kovacs et al. (1988), children with depression were found to have a 36% risk of developing CD by age 19. PuigAntich (1982) separately published observations that approxima-
GURNANI ET AL. tely one third of boys with major depressive disorder in a clinical trial of imipramine for major depression met criteria for CD; in this subset of participants, response to imipramine was followed by resolution of CD symptoms in the majority of children. There is also evidence that serotonergic and noradrenergic antidepressants may decrease impulsivity and aggression (Bond 2005). For example, Coccaro et al. (2009) conducted a double-blind RCT of fluoxetine in 100 adults with intermittent explosive disorder (IED) and concurrent IA. In this study, treatment was associated with significant reductions in ratings on the Modified Overt Aggression Scale (M-OAS), with no relation to any antidepressant or antianxiety effect. Interestingly, whereas low serotonergic function early in life is associated with high levels of aggression, risk taking, and premature death in nonhuman primates, the role of serotonin levels in childhood aggression is less clear, with some studies finding high serotonergic reactivity and others finding low reactivity (Newcorn et al. 2016). To summarize, in depression in youth, conduct problems and aggression are not infrequently present. In these cases, treatment of depression with SSRIs may lead to improvements in aggression ( Jacobs et al. 2010) provided there is not long-standing difficulty with behavior. However, use of SSRIs or other antidepressants for treatment of aggression in children and adolescents in the absence of depression is not currently recommended. Aggression is frequently reported in pediatric bipolar disorder, and a growing literature has addressed this topic. Several studies have shown divalproex to be effective in reducing aggression in manic bipolar youth (e.g., DelBello et al. 2004) and youth at high risk for bipolar disorder (Saxena et al. 2006b). Less evidence exists for other anticonvulsants. Lithium is approved by the FDA for use in children and adolescents 12–18 years of age with bipolar disorder; however, serum level monitoring requirements may limit its use in these populations. Nevertheless, lithium generally ameliorates aggression and irritability in pediatric bipolar illness (Nevels et al. 2010). Antipsychotics such as risperidone, quetiapine, aripiprazole, olanzapine, and recently asenapine, are approved for youth with bipolar disorder. Some studies have shown particular agents to have efficacy for aggressive behaviors specifically. In a comparator study of divalproex versus quetiapine for impulsivity and reactive aggression in children with bipolar disorder and ODD or CD, both medications were similarly efficacious; however, there was no placebo control group (Barzman et al. 2006). In a case series of youth with aggression refractory to other mood stabilizers, aggressive behaviors showed improvement with addition of risperidone to the medication regimen (Saxena et al. 2006a). Furthermore, in a prospective double-blind, placebo-controlled 6 week outpatient trial, youth with bipolar disorder and comorbid DBD and aggression experienced greater improvement in manic symptoms with risperidone versus divalproex than youth with bipolar disorder without comorbidity (West et al. 2011). Psychotic disorders In youth with psychotic disorders such as schizophrenia, antipsychotics are the first line treatment; this is also true when there is associated aggression. Several studies (e.g. Kumra et al. 2008) have shown efficacy of this class of medications in schizophrenia spectrum disorders. TRAAY guidelines and American Academy of Child and Adolescent Psychiatry Practice Parameters recommend the use of antipsychotic medication for this population, typically using a second generation agent (Pappadopulos et al. 2003; McClellan and Stock 2013).
PHARMACOTHERAPY OF AGGRESSION Symptom-Based Approach for Treatment of Aggression Limited research supports use of medication treatments when an underlying primary disorder is not identified, which may be at least partially the result of the heterogeneity in aggressive symptoms and difficulty in reliably assessing aggression. Atypical antipsychotics are increasingly prescribed in this situation, including by primary care providers. IA is the subtype of aggression thought to be most amenable to pharmacotherapy; however, there has been some recent research indicating positive response to pharmacotherapy in other aggression subtypes; therefore, further research is required. For example, little is known about the response of youth with CD and callous, unemotional traits to pharmacotherapy. Although the presumption is that medication might not be useful for this subgroup of youth, Blader et al. (2013) recently reported similar response to stimulants in aggressive youth 6–12 years of age with ADHD with and without callous, unemotional traits. Although this finding seems counterintuitive, it is also instructive. As many children with callous and unemotional traits also show features of IA, treatments that target IA symptoms may have a role in this population. Assessment of Aggression in Youth Assessment of aggressive behavior requires a multistep, multiinformant approach. The clinical interview of the patient and family is often the starting point of clinical assessment, but obtaining collateral information from teachers and other adults is also essential. Further, the mental status examination is useful for identifying cognitive-emotional capacities, including verbal skills, affect regulation, and capacity for empathy in relating to others; the latter capacities can be useful in anticipating the potential for treatment response. A variety of psychometric instruments are also informative for clinical evaluation and monitoring treatment. Conners’ Parent and Teacher Rating Scales (Conners et al. 1998a,b), the Child Behavior Checklist (CBCL), ABC, and Nisonger Child Behavior Rating Form (NCBRF) are broadband rating scales that elicit information about the child or adolescent’s general behavior and functioning in relation to age and gender norms. Although these scales are not specifically devised to assess aggression, they have aggression subscales. Narrowband scales such as the Children’s Aggression Scale (CAS), M-OAS, and Buss-Perry Aggression Questionnaire (Buss and Perry 1992) provide more specific data regarding the context, severity, and precipitants of aggression; the latter scales are most often used in research, but can also be helpful in the clinical setting, especially when evaluating covert aggressive behaviors. The Young Mania Rating Scale is specific to mania but can be used to monitor aggression in the context of bipolar illness. In cases in which aggression is a prominent symptom but the primary diagnosis might be difficult to identify, it may be useful to distinguish between impulsive and premeditated aggression (because IA is generally considered to be a more specific target for pharmacotherapy), and to select behaviors that occur with sufficient frequency to detect change with treatment. The Impulsive/ Premeditated Aggression Scale has been validated in adolescents and young adults, and has been used to differentiate aggressive behavior among adolescents with CD (Mathias et al. 2007). Although no studies have been conducted to evaluate the utility of this instrument as a tool for monitoring treatment response, the scale can be helpful to establish baseline characteristics in aggressive youth and guide decisions regarding treatments that affect impulsivity regardless of diagnosis. Finally, the Antisocial
69 Behavior Scale (ABS) (Brown et al. 1996) is a measure that distinguishes between reactive/impulsive aggression and proactive aggression. The ABS has been validated in clinical samples and was recently used in the TOSCA study (Table 1) (Aman et al. 1985; Achenbach et al. 1991; Sorgi et al. 1991; Aman et al. 1996; Tasse´ et al. 1996; Achenbach and Rescorla 2001; Halperin et al 2002, 2003; Stanford et al. 2003). Other objective tools, such as neuropsychological testing to assess response inhibition, consequence sensitivity, startle response, and psychophysiological measures of autonomic sensitivity, can be useful for characterizing domains of aggression and/or impulsivity, but are not, in and of themselves, sufficient to diagnose IA. The first two (i.e., neuropsychological testing and consequence sensitivity) can provide important information about underlying cognitive processes related to risk as well as the capacity to successfully utilize behavioral treatment. The others are best reserved for the research setting. Safety Considerations and Monitoring of Pharmacotherapy Medications for aggression should be used judiciously and with close patient monitoring, given potential safety concerns regarding the various agents mentioned in this review. With regard to psychostimulants, the potential for slightly delayed weight and growth attainment, as well as cardiovascular risk, has been extensively debated. However, growth trajectory is not a major issue in the majority of youth and there is no evidence that there is elevated risk for sudden death or other severe potentially severe cardiac outcomes (Cooper et al. 2011; Habel et al. 2011). Although obtaining a baseline electrocardiogram (ECG) is not essential prior to treatment, it is suggested in patients with arrhythmias, hypertension, structural cardiac defects, or a family history of untoward cardiac events. Alpha-2 adrenergic agonists are associated with changes in heart rate, blood pressure, and QTc interval. Sedation and vital sign changes usually resolve over the course of treatment, and QTc changes are small and do not lead to significant adverse events. ATX carries a rare risk for liver toxicity. Routine liver function testing before initiating treatment is not recommended; however, a thorough workup is indicated in patients at risk or with related symptoms. ATX and the antidepressants carry an FDA black box warning for suicidal thinking in individuals
Pharmacotherapy of Aggression in Child and Adolescent Psychiatric Disorders Tina Gurnani, MD,1 Iliyan Ivanov, MD,1,2 and Jeffrey H. Newcorn, MD 2
Abstract
Objective: Aggression is a common, yet complex, behavioral complaint, and a frequent indication for referral to child and adolescent psychiatrist treatment. This article reviews the evidence supporting pharmacotherapy of aggression in youth, with a primary focus on impulsive aggression (the primary indication for this intervention). Relevant diagnostic considerations and consensus guidelines are discussed. Methods: Articles examining the role of medications in the treatment of aggression in youth with pathological aggression were identified using PubMed and MEDLINE databases over the past 15 years (2000–2015); selected articles published prior to 2000 and deemed to be of high relevance were searched and also included. Search terms included: Aggression, aggressive, disruptive behavior, conduct, youth, children, and adolescents. Cited references were also searched for relevant articles. Results: There are a number of evidence-based medication treatments for aggression, which are generally best considered in the context of differential diagnosis and ongoing evidence-based psychosocial interventions. Impulsive aggression is generally considered the type of aggression most amenable to medication, but other aggression subtypes may also possibly respond to treatment. Medication classes with positive evidence include the psychostimulants and a-2 agonists (in the presence of attention-deficit/hyperactivity disorder [ADHD] and/or disruptive behavior disorders), mood stabilizing agents, and atypical antipsychotics. Published guidelines recommend systematic and adequate trials of medications in sequential order, to optimize response and minimize polypharmacy. Guidelines for safety monitoring are available for many of the medications used for aggression in youth, and are also discussed. Conclusions: Aggression in children carries a high risk of poor outcomes, and, therefore, a better understanding of treatment options is a high priority. The available literature points to the importance of identifying the underlying disorder, when possible, and using this information to guide treatment selection. Future studies are needed to better inform the treatment of aggression across disorders, and the treatment of different aggression subtypes.
Introduction
persistent concerns regarding efficacy, safety, and polypharmacy (Crystal et al. 2009). Therefore, evidence-based recommendations are needed to guide implementation of appropriate therapeutic interventions. Although aggression is a behavioral construct present in all people to some extent, and can be adaptive for survival, it is highly impairing when it is pathological and persistent. Prosocial or adaptive aggressive behaviors (i.e., self-defense) are distinguished from pathological, maladaptive, or antisocial aggressive behavior. In maladaptive aggression, the intensity, frequency, duration, or severity of the aggressive response is out of proportion to its causes, and may occur in the absence of specific triggers. Pathological aggression can occur in the context of specific psychiatric disorders, or it can be a nonspecific manifestation of anger or frustration. Clinically, the different levels of aggression are commonly differentiated by a person’s motivation and that person’s ability to
P
roblems with aggressive behavior affect 10–25% of youth (Loeber and Farrington 2001), and constitute one of the most common causes for referral to child and adolescent psychiatrists. Youth with aggressive behavior often have profound disruption in a range of functional domains, including academic and occupational attainment, social and family relationships, and psychological development. Consequently, aggression is commonly a target of both psychosocial and medication treatment. However, the urgency of the condition often dictates the use of medication. Given the dearth of child and adolescent psychiatrists, a significant number of prescriptions for aggressive behavior are provided by primary care clinicians; for example, approximately one third of antipsychotic prescriptions for children during 1995–2002 were by non-mental health providers (Cooper et al. 2006). In addition, there have been
1
Department of Psychiatry, Mount Sinai-St. Luke’s Hospital Center, New York, New York. Department of Psychiatry, Icahn School of Medicine at Mount Sinai, New York, New York.
2
65
66 control the behavior, and the intended goal. Impulsive aggression (IA) is generally reactive, unplanned, and overt, with the aggressor perceiving the outcome in a negative light. Planned (i.e., predatory, instrumental) aggression is often covert, and although it may be associated with a negative emotion, the aggressor anticipates a positive outcome. IA is more characteristic of clinical samples and amenable to pharmacological and psychosocial treatments; in contrast, planned aggression is more often characteristic of delinquency. However, although these descriptors are often conceptualized as dichotomous, impulsive/reactive and planned aggression can also occur in the same individual. In a recent study using an aggression scale measuring reactive/impulsive and proactive/ planned aggression, the two scores were found to be distinct but moderately correlated, with r = 0.42 (Kaat et al. 2015). In youth with planned aggression, psychosocial intervention may be useful, whereas pharmacotherapy has not traditionally been considered to be effective, unless IA is also present. Callous-unemotional traits (which are traditionally thought to be related to planned aggression) have also been studied with respect to conduct behavior problems in children, and seem to be associated with higher rates of aggression (Kahn et al. 2012), especially proactive aggression (Viding et al. 2012; Helseth et al. 2015). Different domains of aggression may be associated with distinct neurobiological mechanisms and treatment indicators. For example, in adults, IA has been shown to be related to low serotonergic function or reactivity in adults. However, data are less consistent in children, and do not offer clear directives regarding the approach to pharmacotherapy (discussed subsequently) (Newcorn et al. 2016). Risk factors for IA occur at an early age, often starting at 4 or 5 years of age (vs. 6–7 years of age for planned aggression) (Dodge et al. 1997; Vitaro et al. 2002); therefore, treatment of young children is sometimes necessary. Nevertheless, transient disruptive behaviors are common during the preschool years, so it is important to distinguish pathological aggression from developmental variations. Young boys are especially prone to aggression, and this gender bias is consistent worldwide. Girls tend to exhibit more covert aggression than boys, and less intense behavioral disturbance. Early aggression is a key predictor of later delinquency (Kotch et al. 2008); approximately half of school-age children who are aggressive continue to exhibit such behavior in adolescence. Key risk factors for aggression can be potential treatment targets. A history of childhood trauma has repeatedly been shown to predict later violent behavior, with cumulative exposure to trauma giving rise to increased risk. Sexual abuse may also be a risk factor (Harford et al. 2014). Parental separation and conflict, poverty, harsh discipline, and parental criminality are also associated features. Other risk factors include friendships with delinquent peers, living in high-crime neighborhoods, witnessing community violence, and peer rejection (especially early in childhood) (Hong and Espelage 2012). The large number of environmental, familial, and social contextual factors associated with risk for aggression underscores the importance of psychosocial treatment. Interestingly, poor peer relationships and a history of physical abuse are often specifically associated with IA, whereas aggressive role models in the family and parental incarceration have been shown to be associated with covert aggression (Newcorn et al. 2016). Low intelligence quotient (IQ) and poor language skills also constitute risk factors for early and persistent disruptive behavior (Department of Health and Human Services 2001), with the presumption that the range of adaptive skills in youth with poorly developed language skills is restricted.
GURNANI ET AL. Clinical Approach Treatment planning with aggressive youth is based on a comprehensive understanding of the nature, severity, and context of the symptoms; a confluence of psychological, behavioral, and cognitive considerations; and the profile of risk and protective factors that characterize the individual patient and family. In addition, the intensity and developmental course of symptoms must be considered, as well as the potential consequences (e.g., harm to self or others) that could arise from the aggressive behaviors and how likely this is to happen. In emergent situations, especially in which the child or adolescent has already caused injury to others, or has access to means of inflicting harm (such as firearms), inpatient hospitalization and/or acute medication intervention with sedative agents may be necessary. In the less emergent outpatient setting, there are two general approaches to treating aggression in youth: One is the symptombased approach (e.g., empirical treatment of presenting aggressive symptoms), and the other is to treat the underlying disorder (assuming one can be identified, which is often but not always the case). Current consensus guidelines recommend treating aggression in the context of specific disorders as a first step, and to develop a rational plan for pharmacotherapy in this context ( Jensen et al. 2007). The most common underlying conditions are attentiondeficit/hyperactivity disorder (ADHD) and disruptive behavior disorders (DBD) such as conduct disorder (CD) or oppositional defiant disorder (ODD), especially in younger children (Connor et al. 2002). Mood disorders and neurodevelopmental disorders are also common, as are posttraumatic stress disorder (PTSD), autism spectrum disorder (ASD), and intellectual disability (ID). Other potential underlying conditions may include anxiety disorders, tic disorders, specific learning disorders, and even adjustment disorders. Medical workup may be required to rule out exposure to toxins, seizures, infections, head trauma, and substance use. Several guidelines have been developed for the treatment of aggression in youth, based on both clinical research findings and expert consensus. The Treatment Recommendations for the use of Antipsychotics for Aggressive Youth (TRAAY) guidelines offer recommendations outlined in 14 steps. Key steps include: Psychosocial treatments as the cornerstone of first-line intervention, use of medication targeted to the primary underlying disorder as the next step (such as stimulants for ADHD, selective serotonin reuptake inhibitors [SSRIs] for depressive or anxious symptoms), avoiding polypharmacy when possible, offering a trial of an atypical antipsychotic when psychosocial measures and first-line medication have failed (starting at a low dose and titrating gradually), allowing at least 2 weeks of an antipsychotic trial at an adequate dose before considering it ineffective, use of crisis intervention strategies to minimize the need for emergency chemical and physical restraints, assessing for side effects routinely, ensuring an adequate trial before switching or combining medications, switching to a different atypical antipsychotic if the first one was ineffective, considering addition of a mood stabilizer if there was only a partial response to the first antipsychotic trial, tapering medications if an inadequate response to multiple agents is seen, and tapering of antipsychotic medication after ‡6 months of an adequate response (Pappadopulos et al. 2003). Recently, the Center for Education and Research on Mental Health Therapeutics (CERT) developed similar guidelines for primary care clinicians and mental health providers (i.e., Treatment of Maladaptive Aggression in Youth [T-MAY]). Key recommendations are to use evidence-based parent and child skills training as the first line
PHARMACOTHERAPY OF AGGRESSION intervention, treating the underlying disorder thought to give rise to the behavior, careful consideration of the potential risks of and contraindications for the initiation of medications, avoiding polypharmacy, careful monitoring of treatment response, and close medical monitoring of side effects (Scotto Rosato et al. 2012). Treatment of Aggression in the Context of Specific Psychiatric Disorders ADHD and DBD Psychostimulants are considered to be the most effective medication treatment for ADHD, with mean effect sizes for core symptoms generally ranging from 0.8 to 1.0. In addition, stimulants generally produce improvement in aggression, with positive findings seen in virtually all studies that have examined this question. One meta-analysis examined the effect of stimulants on covert and overt aggression in children with ADHD, finding overall weighted mean effect sizes of 0.84 for overt aggression and 0.69 for covert aggression (Connor et al. 2002). There is also good evidence that stimulants are effective in treating ODD/CD symptoms in children with comorbid ADHD (Kaplan et al. 1990; Gadow et al. 2008). Notably, in a placebo-controlled study of 84 youth with CD, two thirds of whom also met criteria for ADHD, Klein et al. (1997) found that methylphenidate (MPH) reduced behaviors specific to CD, independent of severity of ADHD symptoms. However, a substantial proportion of children with ADHD do not achieve sufficient reduction in aggressive behavior with stimulant treatment ( Jensen et al. 2007). Blader et al. (2009) conducted an open trial of stimulants as the first stage of a stimulant augmentation trial in children 6–13 years of age with ADHD plus ODD/CD and aggression thought to be refractory to stimulants. Of note, although all children referred to this study purportedly had inadequate response, retitration to the optimal dose using both stimulant classes produced improvement in approximately half of the children. However, even in the case of positive response to stimulants, stimulant treatment of aggression is limited by the temporal duration of the medication, because stimulants have relatively short duration of action, which is often true even with many of the current long-acting formulations. Because aggression can be present at any time of the day, treatments that last all day are highly desirable. Nonstimulant medications such as atomoxetine (ATX) and the extended release a-2 adrenergic agonists are United States Food and Drug Administration (FDA) approved for ADHD, generally have a longer duration of action than stimulants, and can be used for ADHD plus aggression, although there are no data that directly compare the effects of stimulants and nonstimulants on aggression. It is of note that clinical trials with nonstimulants have mainly focused on improvements in core ADHD symptoms, even though aggression and oppositionality are frequent targets of nonstimulant treatment (particularly with the a-2 agonists) in the clinical setting. Studies evaluating response in youth with ADHD plus ODD with both clonidine and guanfacine have demonstrated improvement (Palumbo et al. 2008; Wilens et al. 2012). Clonidine has been shown to be effective in reducing aggression in adolescents with ADHD (Hazell and Stuart 2003), and guanfacine has been shown to increase frustration tolerance and decrease irritability (Connor et al. 2010), symptoms that are frequently present in youth with IA. A meta-analysis of ATX clinical trials showed a small effect size (ES) (overall mean ES = 0.18 in 4 randomized controlled trials [RCTs]) for pediatric aggression with a dosage of 1.3mg/kg/day (Pappadopulos et al. 2006). Furthermore, ATX was found to only be
67 modestly effective in treating ODD symptoms in youth with ADHD (Biederman et al. 2007); therefore, ATX may not be a first choice recommendation. Several studies (e.g., Findling et al. 2000; Armenteros et al. 2007) have shown that antipsychotics, especially second generation agents, can be effective when used together with stimulants for aggression in ADHD, although they are not specifically approved by the FDA for this purpose and should be used with caution. In a double-blind, RCT, the Treatment of Severe Childhood Aggression (TOSCA) study, children 6–12 years of age with ADHD and ODD or CD and severe physical aggression received stimulant medication for 3 weeks, titrated for optimal effect, while parents received parent training for behavioral management. If subjects either failed to show sufficient clinical response or deteriorated, treatment was augmented with risperidone or placebo. At the end of 9 weeks, the risperidone group showed moderate improvement in aggression ratings (Aman et al. 2014). Improvement was mostly noted in impulsive, unplanned aggression and disruptive behavior, rather than callous, planned aggression. However, reevaluation of moderators of response in the TOSCA study found that higher levels of callous-unemotional traits predicted better response in disruptive symptoms regardless of treatment assignment, although it should be noted that the measurement of callous-unemotional behaviors was via an ad hoc composite variable based on several rating scales, which has not been validated (Farmer et al. 2015). Similarly, Blair et al. (2014) suggest that the use of stimulants and atypical antipsychotics in combination may be helpful for youth with CD with callous-unemotional traits, which is more difficult to treat and generally thought to have a poorer prognosis. Several open-label trials (e.g. Kronenberger et al. 2007) have been conducted with quetiapine, suggesting that this medication can be used successfully with MPH in patients with ADHD and aggression, with comorbid ODD or CD. Prescription rates of antipsychotic medication have increased in children and adolescents in recent years (Burcu et al. 2014), mainly because of their use in treating aggression and mood dysregulation. However, there are concerns regarding their longterm use, primarily because of metabolic side effects. Furthermore, concomitant use of stimulants and antipsychotics does not seem to combat the metabolic effects of antipsychotics, contrary to speculation (Linton et al. 2013). There is a long history of using mood stabilizers in youth with aggressive CD, dating back to the seminal study of Campbell et al. (1984), which compared the effects of lithium and haloperidol in 61 treatment-resistant hospitalized children. Both medications were superior to placebo in reducing behavioral symptoms, but haloperidol was more sedating. Subsequent studies of lithium have supported this finding, though ESs have been lower (e.g., Campbell et al. 1995; Malone et al. 2000). Valproic acid has also shown efficacy in ameliorating aggression in youth with ADHD, but studies are limited (Patel and Barzman 2013). A preliminary crossover study by Donovan et al. (2000) assessed 20 children with DBD with explosive temper or mood lability, who received 6 weeks of either divalproex treatment or placebo, and then 6 weeks of the reciprocal treatment. In addition, an RCT conducted with adolescent juvenile offenders with CD treated patients with either highdose (500–1500 mg/day or plasma level of 50–120 lg/mL) or lowdose (£ 250 mg/day) divalproex over 7 weeks (Steiner et al. 2003). In the first study, divalproex had superior response rates over placebo, and in the second study, high-dose divalproex yielded greater improvement in clinician ratings and in self-report of impulse control. More recently, Blader et al. (2009) evaluated the efficacy of divalproex in 27 children 6–13 years of age with ADHD plus
68 ODD/CD plus aggression. Those with persisting aggressive behavior after stimulant treatment was optimized were given either divalproex or placebo for 8 weeks, with subjects receiving behavioral therapy throughout. Children treated with divalproex had a significantly higher rate of remission on aggression ratings than those treated with placebo. It is of note that ratings of irritability and mood dysregulation were predictors of inadequate response to stimulant monotherapy. ID and developmental disorders Aggression is a frequent associated feature in youth with ID and ASD. In years past, medication for these conditions was often prescribed off label, with the most frequent medications used being the atypical antipsychotics (de Bildt et al. 2006). Because of this, and preliminary findings indicating beneficial effects of risperidone in this population (Van Bellinghen and De Troch 2001; Aman et al. 2002; Snyder et al. 2002), more definitive studies were conducted to obtain FDA approval for this practice. In 2002 and 2004, two double-blind, placebo-controlled RCTs with risperidone were conducted in children with autism and DBD, in which significant improvements in irritability were observed on the Aberrant Behavior Checklist (ABC)–Irritability subscale after an 8 week treatment period (McCracken et al. 2002; Shea et al. 2004). Similarly, Nagaraj et al. (2006) found improvements in irritability and aggression in a 6 month double-blind, placebo-controlled RCT of 39 young children (2–9 years of age). A 24 week, multisite, randomized, parallel groups trial comparing risperidone alone (subjects could be switched to aripiprazole if ineffective) to a combination of risperidone and parent training therapy, found that combination therapy was somewhat superior to medication alone, with a 14% lower dose of risperidone used at the end of the study (Aman et al. 2009). Subsequent research with aripiprazole also resulted in FDA approval for this indication. Politte and McDougle (2014) reported improved ratings of aggression, with similar or slightly lower response rates than are typically seen with risperidone. There is insufficient evidence that other antipsychotics, such as olanzapine and quetiapine, are similarly efficacious for aggression in the context of developmental disabilities, although both medications are approved for bipolar disorder in adolescents and presumably might yield improvement in aggression in these populations. Beta blockers are often prescribed off label for youth with ID, as they have shown positive results in several studies; however no RCTs for aggression have been conducted (Ward et al. 2013). Nonstimulant medications for ADHD, such as clonidine and guanfacine, have also been used, particularly when hyperactivity/ impulsivity and sleep problems are present. In addition, moodstabilizing medications have been prescribed off label for aggression in youth with developmental disabilities. Divalproex has shown some efficacy in a small double-blind, placebo-controlled study (Hollander et al. 2010) in children and adolescents 5–17 years of age with ASD; subjects who received divalproex had significantly less irritability/aggression on the ABC–Irritability subscale than those treated with placebo after 12 weeks of treatment. Mood disorders Behavior problems and aggression are frequently encountered in youth with depressive disorders, which to some extent distinguishes depression in youth from depression in adults. In a longitudinal study by Kovacs et al. (1988), children with depression were found to have a 36% risk of developing CD by age 19. PuigAntich (1982) separately published observations that approxima-
GURNANI ET AL. tely one third of boys with major depressive disorder in a clinical trial of imipramine for major depression met criteria for CD; in this subset of participants, response to imipramine was followed by resolution of CD symptoms in the majority of children. There is also evidence that serotonergic and noradrenergic antidepressants may decrease impulsivity and aggression (Bond 2005). For example, Coccaro et al. (2009) conducted a double-blind RCT of fluoxetine in 100 adults with intermittent explosive disorder (IED) and concurrent IA. In this study, treatment was associated with significant reductions in ratings on the Modified Overt Aggression Scale (M-OAS), with no relation to any antidepressant or antianxiety effect. Interestingly, whereas low serotonergic function early in life is associated with high levels of aggression, risk taking, and premature death in nonhuman primates, the role of serotonin levels in childhood aggression is less clear, with some studies finding high serotonergic reactivity and others finding low reactivity (Newcorn et al. 2016). To summarize, in depression in youth, conduct problems and aggression are not infrequently present. In these cases, treatment of depression with SSRIs may lead to improvements in aggression ( Jacobs et al. 2010) provided there is not long-standing difficulty with behavior. However, use of SSRIs or other antidepressants for treatment of aggression in children and adolescents in the absence of depression is not currently recommended. Aggression is frequently reported in pediatric bipolar disorder, and a growing literature has addressed this topic. Several studies have shown divalproex to be effective in reducing aggression in manic bipolar youth (e.g., DelBello et al. 2004) and youth at high risk for bipolar disorder (Saxena et al. 2006b). Less evidence exists for other anticonvulsants. Lithium is approved by the FDA for use in children and adolescents 12–18 years of age with bipolar disorder; however, serum level monitoring requirements may limit its use in these populations. Nevertheless, lithium generally ameliorates aggression and irritability in pediatric bipolar illness (Nevels et al. 2010). Antipsychotics such as risperidone, quetiapine, aripiprazole, olanzapine, and recently asenapine, are approved for youth with bipolar disorder. Some studies have shown particular agents to have efficacy for aggressive behaviors specifically. In a comparator study of divalproex versus quetiapine for impulsivity and reactive aggression in children with bipolar disorder and ODD or CD, both medications were similarly efficacious; however, there was no placebo control group (Barzman et al. 2006). In a case series of youth with aggression refractory to other mood stabilizers, aggressive behaviors showed improvement with addition of risperidone to the medication regimen (Saxena et al. 2006a). Furthermore, in a prospective double-blind, placebo-controlled 6 week outpatient trial, youth with bipolar disorder and comorbid DBD and aggression experienced greater improvement in manic symptoms with risperidone versus divalproex than youth with bipolar disorder without comorbidity (West et al. 2011). Psychotic disorders In youth with psychotic disorders such as schizophrenia, antipsychotics are the first line treatment; this is also true when there is associated aggression. Several studies (e.g. Kumra et al. 2008) have shown efficacy of this class of medications in schizophrenia spectrum disorders. TRAAY guidelines and American Academy of Child and Adolescent Psychiatry Practice Parameters recommend the use of antipsychotic medication for this population, typically using a second generation agent (Pappadopulos et al. 2003; McClellan and Stock 2013).
PHARMACOTHERAPY OF AGGRESSION Symptom-Based Approach for Treatment of Aggression Limited research supports use of medication treatments when an underlying primary disorder is not identified, which may be at least partially the result of the heterogeneity in aggressive symptoms and difficulty in reliably assessing aggression. Atypical antipsychotics are increasingly prescribed in this situation, including by primary care providers. IA is the subtype of aggression thought to be most amenable to pharmacotherapy; however, there has been some recent research indicating positive response to pharmacotherapy in other aggression subtypes; therefore, further research is required. For example, little is known about the response of youth with CD and callous, unemotional traits to pharmacotherapy. Although the presumption is that medication might not be useful for this subgroup of youth, Blader et al. (2013) recently reported similar response to stimulants in aggressive youth 6–12 years of age with ADHD with and without callous, unemotional traits. Although this finding seems counterintuitive, it is also instructive. As many children with callous and unemotional traits also show features of IA, treatments that target IA symptoms may have a role in this population. Assessment of Aggression in Youth Assessment of aggressive behavior requires a multistep, multiinformant approach. The clinical interview of the patient and family is often the starting point of clinical assessment, but obtaining collateral information from teachers and other adults is also essential. Further, the mental status examination is useful for identifying cognitive-emotional capacities, including verbal skills, affect regulation, and capacity for empathy in relating to others; the latter capacities can be useful in anticipating the potential for treatment response. A variety of psychometric instruments are also informative for clinical evaluation and monitoring treatment. Conners’ Parent and Teacher Rating Scales (Conners et al. 1998a,b), the Child Behavior Checklist (CBCL), ABC, and Nisonger Child Behavior Rating Form (NCBRF) are broadband rating scales that elicit information about the child or adolescent’s general behavior and functioning in relation to age and gender norms. Although these scales are not specifically devised to assess aggression, they have aggression subscales. Narrowband scales such as the Children’s Aggression Scale (CAS), M-OAS, and Buss-Perry Aggression Questionnaire (Buss and Perry 1992) provide more specific data regarding the context, severity, and precipitants of aggression; the latter scales are most often used in research, but can also be helpful in the clinical setting, especially when evaluating covert aggressive behaviors. The Young Mania Rating Scale is specific to mania but can be used to monitor aggression in the context of bipolar illness. In cases in which aggression is a prominent symptom but the primary diagnosis might be difficult to identify, it may be useful to distinguish between impulsive and premeditated aggression (because IA is generally considered to be a more specific target for pharmacotherapy), and to select behaviors that occur with sufficient frequency to detect change with treatment. The Impulsive/ Premeditated Aggression Scale has been validated in adolescents and young adults, and has been used to differentiate aggressive behavior among adolescents with CD (Mathias et al. 2007). Although no studies have been conducted to evaluate the utility of this instrument as a tool for monitoring treatment response, the scale can be helpful to establish baseline characteristics in aggressive youth and guide decisions regarding treatments that affect impulsivity regardless of diagnosis. Finally, the Antisocial
69 Behavior Scale (ABS) (Brown et al. 1996) is a measure that distinguishes between reactive/impulsive aggression and proactive aggression. The ABS has been validated in clinical samples and was recently used in the TOSCA study (Table 1) (Aman et al. 1985; Achenbach et al. 1991; Sorgi et al. 1991; Aman et al. 1996; Tasse´ et al. 1996; Achenbach and Rescorla 2001; Halperin et al 2002, 2003; Stanford et al. 2003). Other objective tools, such as neuropsychological testing to assess response inhibition, consequence sensitivity, startle response, and psychophysiological measures of autonomic sensitivity, can be useful for characterizing domains of aggression and/or impulsivity, but are not, in and of themselves, sufficient to diagnose IA. The first two (i.e., neuropsychological testing and consequence sensitivity) can provide important information about underlying cognitive processes related to risk as well as the capacity to successfully utilize behavioral treatment. The others are best reserved for the research setting. Safety Considerations and Monitoring of Pharmacotherapy Medications for aggression should be used judiciously and with close patient monitoring, given potential safety concerns regarding the various agents mentioned in this review. With regard to psychostimulants, the potential for slightly delayed weight and growth attainment, as well as cardiovascular risk, has been extensively debated. However, growth trajectory is not a major issue in the majority of youth and there is no evidence that there is elevated risk for sudden death or other severe potentially severe cardiac outcomes (Cooper et al. 2011; Habel et al. 2011). Although obtaining a baseline electrocardiogram (ECG) is not essential prior to treatment, it is suggested in patients with arrhythmias, hypertension, structural cardiac defects, or a family history of untoward cardiac events. Alpha-2 adrenergic agonists are associated with changes in heart rate, blood pressure, and QTc interval. Sedation and vital sign changes usually resolve over the course of treatment, and QTc changes are small and do not lead to significant adverse events. ATX carries a rare risk for liver toxicity. Routine liver function testing before initiating treatment is not recommended; however, a thorough workup is indicated in patients at risk or with related symptoms. ATX and the antidepressants carry an FDA black box warning for suicidal thinking in individuals
