The 50th Annual National Institute of Mental Health (NIMH) New Clinical Drug Evaluation Unit (NCDEU) Meeting,Boca Raton,Florida,June 14–17,2010: Posters Most Relevant to Child and Adolescent Psychopharmacology

Applying Discrete Choice Experiments in Mental Health—An Example on Parents' Preferences in Attention Deficit Hyperactivity Disorder (ADHD) Treatment

Objectives: While assessing and allocating treatment options for children with attention deficit hyperactivity disorder (ADHD), health care providers should take the parents' preferences into account since the parents' acceptance and support is crucial for treatment success. Discrete choice experiments (DCE) are increasingly applied in the health sector to elicit the participants' (e.g., patients', insured persons', caregivers') preferences for health care programs and products (i.e., to weigh the importance these persons attribute to treatment characteristics). The applicability of DCEs in mental health and in particular with primary caregivers in ADHD is investigated in this study.

Methods: Relevant characteristics of an “ideal” medical ADHD treatment from the parents' perspective were collected by a literature review and a qualitative study with focus groups in order to achieve content validity. In the subsequent quantitative study, preferences for an ADHD treatment were investigated by using direct measurement (23 aspects, Likert-scale) and a DCE (eight pairs of treatment options, six dichotomous characteristics). Preferences were analysed with multivariate procedures (factor analysis, probit/logit models).

Results: Questionnaires were completed by 121 primary care givers (101 mothers, 16 fathers, four others) on behalf of their school age child (six to 14 years, 87% male). All characteristics were statistically significant in the DCE. Highest relative importance was attributed to an improvement in the child's social situation and its emotional state (38% and 21%). Another essential characteristic was a long-lasting drug effect (18%), whereas side effects (8%), dosing flexibility (7%) and discretion (8%) were assessed as less important.

Limitations: Preferences reflect cultural attitudes and might not be completely generalizable to other contexts. Participants were clinically referred patients and not a population-based sample.

Conclusions: The presented study demonstrates feasibility of methods applied to caregivers (parents) of children with ADHD in order to systematically assess treatment preferences. This work may eventually lead to improved ADHD treatment and a more efficient resource allocation in mental health care.

Source of Funding: Janssen Cilag.

Psychometric Properties of the Wender-Reimherr Adult Attention Deficit Disorder Scale

Background: The Wender-Reimherr Adult Attention Deficit Disorder Scale (WRAADDS) is a clinician rated scale for assessing adult attention deficit hyperactivity disorder (ADHD) based on the Utah Criteria. ¹ It assesses seven symptom areas: attention difficulties, hyperactivity/restlessness, temper, affective lability, emotional over-reactivity, disorganization and impulsivity. While three of these mirror the DSM-IV criteria, the other four (temper, affective lability, emotional over-reactivity and disorganization) are either associated features or not included. Their inclusion makes the WRAADDS unique from DSM-based scales. While it has proven effective in clinical trials, the lack of published psychometric information has limited its use. This report documents both reliability and validity for the WRAADDS.

Methods: The normative sample consisted of 120 males and females ages 20 through 49 inclusive. Exclusion criteria included known personal or family history of ADHD, personal history of an Axis I disorder in the last three months and any history of hospitalization or treatment for a psychotic disorder. The ADHD subjects were taken from clinical trials of adult ADHD: (1) two identical clinical trials of atomoxetine; (2) a clinical trial of OROS-MPH; and (3) a clinical trial of Methylphenidate Transdermal System.

Results: Test-retest using 51 adult ADHD subjects from two studies was excellent (r = 0.96). Inter-rater reliability using 67 adult ADHD subjects from two studies was excellent (r = 0.75). Internal consistency was acceptable (Cronbach's alpha = 0.78). Removal of any of the seven subscales resulted in only a small decrease in internal consistency. Concurrent validity was assessed using the Conners' Adult ADHD Rating Scales (CAARS). ² The total WRAADDS was significantly correlated with the Total CAARS (r = 0.501, df = 534, p < 0.001). The hyperactivity +impulsivity subscales of the WRAADDS correlated with the hyperactivity/impulsivity subscale of the CAARS (r = 0.601, df = 534. p < 0.001). The attention + disorganization subscales of the WRAADDS correlated with the inattention subscale of the CAARS (r = 0.430, df = 534, p < 0.001). Discriminate validity was evaluated by distinguishing between adults with and without ADHD (matched for age, sex and race). There was a significant difference between the normative and ADHD subjects for the total WRAADDS (F1,228 = 1075.6, p < 0.001) and each of the seven symptom areas (p < 0.001).

Conclusions: The WRAADDS assesses a more comprehensive set of adult ADHD symptoms than DSM-IV based scales. These data indicate that it has adequate reliability and validity for continued use in research and treatment of adult ADHD.

Source of Funding: In part by an unrestricted educational grant from Eli Lilly, and Company.

Cardiovascular Safety of Stimulant Medications for Attention Deficit Hyperactivity Disorder (ADHD): A Review of Data from Placebo-Controlled and Open-Label Extension Trials

Background: Stimulants are a mainstay of attention deficit hyperactivity disorder (ADHD) treatment. Despite their efficacy and long history of use, recent concerns exist regarding their potential adverse cardiovascular (CV) effects. ^1,2 The nature and magnitude of these effects is controversial. This data review evaluates a decade's worth of published reports in the context of stimulant treatment of ADHD and the risk of CV events.

Methods: All placebo-controlled and open-label extension trials published after 2000 reporting heart rate (HR), blood pressure (BP) and/or QT interval data in children, adolescents or adults with ADHD treated with methylphenidate (MPH) or mixed amphetamine salts (MAS) were included. CV data from all treated subjects were compared across studies.

Results: Three placebo-controlled trials evaluated therapeutic MAS (three to six weeks treatment; N = 659); one reported significant increases from baseline in mean HR (5.0 bpm, p < 0.05); none noted significant BP or QT interval changes. Four open- label extension trials evaluated MAS (15 wks–2 yrs; N = 3087); all reported significant increases in HR (1.5 to 4.4 bpm, p < 0.001) and systolic BP (0.6 to 3.5 mm Hg, p < 0.05 or p < 0.001), and three noted QT interval increases (2.7–7.2 ms, p < 0.05). Three placebo-controlled trials evaluated MPH treatment (4–6 wks; N = 253); two trials reported significant HR increases (4.5 and 7.0 bpm; p < 0.001), and one reported increases in systolic BP (1.5 mm Hg, p < 0.05), diastolic BP (4.0 mm Hg, p < 0.001), and QT interval (7.0 ms, p < 0.05). One of two MPH open-label extension trials (6 mo–1 yr; N = 602) observed significant increases in HR (3.9 bpm, p < 0.001), systolic BP (3.3 mm Hg, p < 0.001) and diastolic BP (1.5 mm Hg, p < 0.001). No deaths or hospitalizations due to CV adverse events were reported in any of the studies.

Conclusions: Therapeutic use of stimulants for the treatment of ADHD can increase HR, BP and QT interval. Although the magnitude of these effects appear small, we recognize both a potential for acute consequences in children with pre-existing cardiac disease and uncertainties regarding sequelae of long-term exposure to these small effects in all children. Screening for occult disease with a baseline electrocardiogram (ECG) is reasonable to consider for all patients starting stimulants.

Source of Funding: Novartis Pharmaceuticals Corporation.

Effects of Stimulants on the Cerebellar Morphology in Attention Deficit Hyperactivity Disorder (ADHD)

Background: Neuroimaging research of attention deficit hyperactivity disorder (ADHD) has consistently demonstrated brain abnormalities in youth with ADHD with most prominent differences documented in the cerebellum. The fact that the developmental trajectories for the cerebellum remain parallel for both patients and controls suggests that both genetic and environmental influences in ADHD are fixed and independent of medication treatment. More specifically, decreased volume of the superior cerebellar vermis appears to represent an important substrate of these non-progressive anatomical changes that underlie ADHD whereas the cerebellar hemispheres are more plastic, state-specific markers that may potentially constitute targets for biological interventions. Therefore, implementation of neuroimaging techniques that can differentiate between cerebellar substructures may provide useful additional information about the effect of biological treatments on regional cerebellar morphology.

Objective: To evaluate the effect of stimulant treatment on the cerebellar morphology in youth with ADHD with the use of surface morphometry. Based on previous reports It is hypothesized that children with ADHD who received naturalistic stimulant treatment will exhibit larger regional volumes in the cerebellar hemispheres when compared to untreated counterparts.

Methods: We compared 42 children and adolescents with ADHD and 57 control ages eight to 18 years, in a cross-sectional, case-control study design with measures of surface morphology of the cerebellum used as main outcome measures. Of the ADHD group 31 children have received naturalistic treatment with stimulant agents from minimum of three to maximum of 108 months (Mean = 43.3, SD + 29.1).

Results: Surface maps showed significantly decreased regional volumes of the anterior-lateral surfaces of the cerebellar hemispheres bilaterally (p < 0.0001) in individuals with ADHD versus controls. Conversely, youth with ADHD who received stimulant treatment exhibited significantly larger volumes in these same areas compared to untreated counterparts.

Conclusions: These results confirm previous findings demonstrating reduced hemisphere regional volumes of the cerebellum in youth with ADHD compared to controls. They also provide preliminary data indicating that these same areas are relatively enlarged in treated versus untreated patients with ADHD.

Source of Funding: National Institute on Drug Abuse/American Association for Child and Adolescent Psychiatry (K23 PA-00-003); National Institute of Mental Health (MH068318, MH59139).

A Study of the Coadministration of Guanfacine Extended Release and a Psychostimulant for the Treatment of Attention Deficit Hyperactivity Disorder: Design and Rationale

Background: Increasingly, children and adolescents with attention deficit hyperactivity disorder (ADHD) are given combined medication regimens for refractory or comorbid ADHD. Despite this, few controlled studies are available on the concomitant use of two medications for ADHD. One such combination of interest is the coadministration of α₂-adrenoceptor agonists and psychostimulants. Currently, α₂-adrenoceptor agonists are often administered nightly, whereas little data exist on differences in efficacy and tolerability of daytime versus nighttime administration. We describe a controlled study of a recently approved extended-release preparation of guanfacine (GXR), a selective α₂-adrenoceptor agonist, added to a stable psychostimulant regimen for the treatment of ADHD in subjects with suboptimal response to the psychostimulant.

Methods: Pediatric and adolescent subjects aged six to 17 years with a DSM-IV-TR diagnosis of ADHD confirmed by structured interview were eligible to participate. Subjects were required to have an ADHD Rating Scale Version IV (ADHD-RS-IV) score of ≥ 24 at baseline, and be on a stable, once-daily dose of a long-acting psychostimulant (amphetamine or methylphenidate) for ≥ 4 weeks with suboptimal response in the opinion of the investigator. Subjects were randomized (1:1:1) to receive their baseline psychostimulant regimen in addition to placebo, GXR dosed in the morning, or GXR dosed in the evening. Guanfacine was increased in 1-mg weekly increments up to a maximum optimal dose of 4 mg/d. The five-week dose-optimization phase was followed by a three-week dose-maintenance phase. ADHD outcome was assessed weekly using the ADHD-RS-IV (the primary efficacy measure). Secondary efficacy measures included the Clinical Global Impressions–Improvement Scale, the Clinical Global Impressions–Severity of Illness Scale, the Conners' Global Index–Parent (assessing symptoms in the morning and in the evening separately), the Parent's Global Assessment, the oppositional subscale of the Conners' Parent Rating Scale–Revised Long Form, and the Wil-Hammer Before School Functioning Scale. Sleep parameters were assessed with the Post-Sleep Questionnaire. Safety was assessed through vital signs, reports of adverse events, electrocardiograms, physical examination and laboratory monitoring.

Results: Of the 615 subjects who were screened, 461 were randomized. There were 59 enrolling sites. Enrollment lasted approximately 65 weeks.

Conclusions: The current study demonstrates an approach to addressing combination therapy and timing of medication administration in ADHD. Despite the complexity of such multisite studies, the current study has demonstrated the feasibility of combined medication trials for design and recruitment. Given the complexity of ADHD and the need for controlled studies of combined treatments, well-conducted studies are essential to inform the field.

Source of Funding: Shire Development Inc.

Predictors of Risperidone Serum Concentration in Youths

Background: When given alone, risperidone may be more rapidly metabolized in youths compared to adults. However, little is known about its metabolism in a clinical sample, where polypharmacy is common. Such data are important since common risperidone side effects, like hyperprolactinemia, are dose-dependent. ¹

Methods: Seven to 17 year-old patients treated with risperidone for at least six months were enrolled. ¹ Treatment with other antipsychotics, but not other psychotropics, led to exclusion as did the presence of serious medical or neurological disorders, the use of contraception, or pregnancy. Demographic and clinical variables were collected upon enrollment and the medication history was extracted from the medical record. In order to measure risperidone and 9-hydroxy-risperidone concentrations, a fasting (in 93%) serum sample was obtained, before the morning dose (in 93%).

Results: One hundred-seven participants (92% males, median age: 11.3 years, interquartile range [IQR]: 4.3) were recruited. Attention deficit hyperactivity disorder (86%) and disruptive behavior disorders (65%) were the most common diagnoses, though comorbidity was the rule. At enrollment, the median daily dose of risperidone was 0.03 mg/kg (IQR: 0.3) and the median treatment duration was 2.2 years (IQR: 2.4). The most commonly co-prescribed psychotropics included: psychostimulants (71%), selective serotonin reuptake inhibitors (50%), and α₂- agonists (32%). Cytochrome CYP2D6 inhibitors were divided in three non- overlapping categories: prominent (fluoxetine, paroxetine, fluvoxamine or bupropion, 30%), intermediate (sertraline, 9%), and weak inhibitors (citalopram or escitalopram, 13%). Multiple linear regression analysis revealed that the serum levels of both risperidone and its metabolite were strongly associated with the weight-adjusted dose of risperidone and time since the last dose (all p values < 0.001). In addition, risperidone concentration, but not that of its metabolite, increased with Tanner stage (p < 0.04) while male sex (p < 0.03) and body mass index z score (p < 0.002) predicted a higher level of only 9-hydroxy-risperidone. Finally, the use of CYP2D6 inhibitors was more strongly correlated with risperidone concentration (p < 0.0001) than with its metabolites (p < 0.05). The duration of risperidone treatment was not associated with either concentration. The models accounted for 55% of the variance in risperidone level, 64% of its metabolites, and 70% of the variance in their combined concentration.

Conclusions: In chronically-treated youths, the metabolism of risperidone depends on the stage of sexual development while that of 9-hydroxy-risperidone varies with sex and body fat. Moreover, CYP2D6 inhibitors more strongly affect risperidone metabolism than that of its metabolite.

Source of Funding: NARSAD, the National Center for Research Resources (RR00059) and the National Institute of Mental Health (R21MH080968 and K23MH085005).

Persisting Racial Disparities Despite Increased Psychotropic Use across a Decade for Medicaid-Insured Youth

Background: Volunteer survey data ¹ and some claims data ^2,3 suggest reduced racial disparities in psychotropic drug use, specifically related to increased use among minority youth. This study was designed to learn whether earlier findings are true for more recent years; if they relate to increased Hispanic enrollment and since state-Children's Health Insurance Programs (s-CHIP) were implemented.

Methods: Three one-year data extracts (1997; 2001; 2006) from a mid-Atlantic Medicaid program were organized from outpatient dispensed medication and enrollment files; three study groups comprised zero to 19 year olds with ten to 12 months of continuous enrollment/year. Psychotropic users were defined as youth with one or more dispensed prescriptions for antipsychotics, antidepressants, alphaagonists, stimulants, atomoxetine, anticonvulsant-mood stabilizers, lithium, anxiolytic/hypnotics and antiparkinsonians. Descriptive data included annual psycho-tropic medication use prevalence (per 100 enrollees) for the three yearly groups, total and according to age group, gender, race/ethnicity, eligibility and region. The major questions for logistic regression analysis were: (1) did the prevalence of use change across the decade adjusting for covariates? and (2) what is the impact of the decade on racial disparities in psychotropic drug use?

Results: (1) The study population rose from 208,950 in 1997 to 371,819 in 2006 with large increases in Hispanic youth and in s-CHIP eligible youth. (2) Total unadjusted psychotropic prevalence rose from 6.9% in 1997 to 11.2% in 2006. When adjusted for age group, gender, race/ethnicity, region and eligibility, the odds of psychotropic use was two-fold greater at the end of the decade (Odds Ratio [O.R.] = 2.10, 95% Confidence Interval [C.I., 2.05, 2.14]). (3) In separate models for 1997 and 2006, the racial disparity showed no significant change for African-American/White psychotropic use (in 1997, O.R. = 0.42, [CI = 0.41, 0.44]; in 2006, O.R. = 0.43, [CI = 0.42, 0.44]). The change for Hispanic/White youth, however, showed a significant increase in the racial disparity from 1997, O.R. = 0.55 (CI = 0.48, 0.62) to 0.40, (CI = 0.38, 0.43). Prominent differences according to age group, gender, eligibility group and region reflect known sociodemographic, social and clinical vulnerability as well as regional variations.

Conclusions: Despite increased access to psychotropic treatment and overall growth in use across the decade (1997–2006), Medicaid-insured minorities are less than half as likely to have psychotropic treatments.

Source of Funding: Pharmaceutical Health Services Department.

Metabolic Effects of Olanzapine in Children with Autistic Disorder

Background: Antipsychotics are the best-studied drugs for reducing symptoms in children with autistic disorder. ¹ A safety concern with antipsychotics is use-associated increases in metabolic parameters including weight, body mass index (BMI), fasting glucose and fasting lipids (triglycerides, total cholesterol, high-density lipoprotein [HDL], low-density lipoprotein [LDL]). These changes are associated with increased risk of cardiovascular disease and diabetes. This study examines the effect of olanzapine on metabolic parameters in a population of children with autism who participated in a clinical trial of olanzapine.

Methods: Patients were 33 children (25 males, eight females), aged three to 11.9 years (mean, 6.58 ± 2), diagnosed with autistic disorder. The study design included a six-week randomized, double-blind, placebo-controlled phase, followed by a six-week open treatment phase. Responders entered a long-term open continuation phase for an additional 20 weeks (32 weeks total study drug exposure). Safety measures included the body mass index (BMI) and laboratory studies obtained at baseline and Week 12. Paired t-tests were performed comparing Week 12 and baseline values to identify any significant olanzapine associated increases in BMI z- score, fasting glucose, and lipids. The Center for Disease Control and Prevention BMI criteria were used to categorize children as obese ( = 95th percentile for BMI), overweight (85th = x < 95th percentile), healthy (5th = x < 85th) and underweight (<5th percentile). ²

Results: At baseline, 70.3% of children were at a healthy weight, 21.6% were overweight, 2.7% were obese, and 5.4% were underweight. By Week 12, 42.4% were healthy weight, 21.2% were overweight, and 36.4% were obese. Paired t-tests comparing the BMI z-score from baseline and Week 12 showed a mean increase of 0.85 ± 0.5 which was significant (t = 9.8, df = 32, p = 0.0000). However, there was no clinically significant change in fasting glucose or in any of the fasting lipid parameters. Overall, the mean increases for the lipids were: fasting triglycerides, 1.8 mg/dL; cholesterol, 3.8 mg/dL; HDL, 0.2 mg/dL; and LDL, 4.5 mg/dL.

Conclusions: The results of this study suggest olanzapine significantly increases BMI when administered to children with autism. Other clinically significant effects on metabolic parameters were not found in this twelve week study.

Source of Funding: Food and Drug Administration (FD-R-002190 [P.I.: Malone]), National Institute of Mental Health (MH073524 [P.I.: Malone]). Placebo and drug were provided by Eli Lilly.

The Effects of Olanzapine on QTc in Children with Autistic Disorder

Background: Antipsychotics are the best-studied drugs for reducing disruptive symptoms such as irritability, hyperactivity and mood lability in children with autistic disorder. A safety concern with antipsychotics is that administration can be associated with prolongation of the QTc, an electrocardiographic measure of ventricular depolarization and re-polarization. QTc prolongation greater than 60 msec from baseline or an absolute QTc of greater than 500 msec have been associated with ventricular arrhythmias and sudden death. Olanzapine has not been shown to prolong QTc in adults. ^1,2 However, the effect of olanzapine on this important index is not well studied in children. To examine the effect of olanzapine on QTc in a population of children with autism who participated in a clinical trial of olanzapine monotherapy.

Methods: Patients were 35 children (26 males), aged three to 11.9 years (mean, 6.3 ± 2), diagnosed with autistic disorder. The study included a six-week randomized, double-blind, placebo-controlled phase, followed by a six-week open treatment phase. Responders at the end of the open treatment phase continued to receive open olanzapine for an additional 20 weeks. Electrocardiograms (ECG) were obtained at baseline, Week 12, and Week 32. They were obtained in a fasting state between 9 and 10 a.m., when drug was at its trough level. To investigate whether there were olanzapine-associated QTc increases, paired t-tests were performed.

Results: The mean QTc at baseline was 409.3 ± 19.1 msec, at Week 12 was 407.3 ± 25.2 msec, and at Week 32 was 411.3 ±40.1 msec. The change in QTc from baseline to Week 12 was −2.1 ± 29 msec (t = 0.42, df = 34, p = 0.68), and from baseline to Week 32 was 2.28 ± 42.7 msec (t = -0.226, df = 17, p = 0.824). No child had a QTc greater than 500 msec during this study. One child with a baseline QTc of 388 msec had an increase of = 60 msec at Weeks 12 and 32.

Conclusions: These results suggest that olanzapine monotherapy does not significantly affect QTc when administered to children with autism.

Source of Funding: Food and Drug Administration (FD-R-002190m P.I.: Malone) and National Institute of Mental Health (MH073524, P.I.: Malone); placebo and drug were provided by Eli Lilly.

Possible Immunomodulatory Effects of Omega-3 Fatty Acids and Olive Oil in a 20-Week Clinical Trial of Children and Adolescents with Tourette's Disorder: Relationships to Treatment Response

Background: The possible role of immune system dysregulation in Tourette's disorder (TD) has been increasingly recognized. Studies implicate cytokine imbalances in TD, including tumor necrosis factor (TNF)-α, interleukin (IL)-2, and IL-12. Omega-3 Fatty Acids (omega-3FA) are widely used as an alternative treatment for TD and are known to have anti-inflammatory properties. The present study is the first double-blind, placebo-controlled trial examining the therapeutic efficacy and cytokine-specific effects of omega-3FA and olive oil (placebo) in children and adolescents with TD. We hypothesized that: (1) improvement of tics and obsessive compulsive disorder (OCD) symptoms would be positively associated with cytokine plasma concentrations in both treatment groups; (2) these associations would be more pronounced in the omega-3FA group.

Methods: Thirty-three children and adolescents with TD (28 males), ages six to 18, were enrolled in a 20-week, double-blind, placebo-controlled trial; 17 were randomly assigned to omega-3FA and 16 to olive oil treatment. Clinician-rated measures included the Yale Global Tic Severity Scale (YGTSS) and the Children's Yale-Brown Obsessive Compulsive Scale (CY-BOCS). Blood samples were collected at 8–9 am after a 12-hour fast at baseline, midpoint, and endpoint weeks and analyzed using enzyme-linked immunosorbent assays. Generalized estimating equations (GEE) based on a binary logistic regression model was used to assess cytokine predictability of OCD and Tic symptom improvement (i.e., CY-BOCS and YGTSS tic score), accounting for age and gender. Spearman correlations characterized associations between cytokines and severity scores.

Results: Table 1 and Table 2 show GEE predictabilities and Spearman correlations.

Conclusions: Findings suggest that IL-6 and TNF-α may play a role in TD treatment response, possibly affected by omega-3FA.

Source of Funding: National Institutes of Health (AT002395, MH077072), Tourette Syndrome Association, and the New York University School of Medicine (M01- RR00096).

Double-Blind, Placebo-Controlled Efficacy and Safety Study of Lisdexamfetamine Dimesylate in Adolescents with Attention Deficit Hyperactivity Disorder (ADHD)

Background: Lisdexamfetamine dimesylate (LDX, Vyvanse^®, Shire U.S. Inc.) is a long-acting prodrug stimulant indicated for attention deficit hyperactivity disorder (ADHD) in children (six to 12 years) and adults, but not adolescents (13–17 years). LDX is endogenously converted to d-amphetamine. This study examines the efficacy and safety of LDX versus placebo in adolescents with ADHD.

Methods: Eligible subjects (13–17 years) with at least moderately symptomatic ADHD (ADHD Rating Scale IV: Clinician Version [ADHD-RS-IV] score ≥ 28) were randomized to placebo or LDX (30, 50 or 70 mg/d) with forced-dose titration in a four-week, double-blind study. Primary and secondary efficacy measures were the ADHD-RS-IV and Clinical Global Impression—Improvement (CGI-I) scale. Safety assessments included adverse events (AEs), vital signs, laboratory findings, physical exam and electrocardiogram (ECG).

Results: Overall, 314 subjects were randomized, 309 included in efficacy analyses, and 49 withdrew (11 due to AEs). At endpoint, changes in ADHD-RS-IV were significantly greater for each LDX dose versus placebo; least squares mean (SE) change was −12.8 (1.25), −18.3 (1.25), −21.1 (1.28) and −20.7 (1.25) for placebo, 30, 50 and 70 mg/d LDX (p < 0.006 for all), respectively. Significant differences in ADHD-RS-IV scores relative to placebo were observed in each LDX group beginning at Week 1 and at each week throughout the study. The percentage of subjects rated very much or much improved at endpoint as measured by CGI-I was significantly greater for LDX (all doses) than for placebo (69.1% versus 39.5% [p < 0.0001], respectively). The most frequently reported LDX treatment-emergent AEs (>5%) were decreased appetite, headache, insomnia, weight decrease and irritability. There were small mean increases in pulse and systolic and diastolic blood pressure with LDX. There were no clinically meaningful trends in ECG.

Conclusions: LDX was effective compared with placebo in decreasing ADHD symptoms from the first week of treatment in adolescents with ADHD. LDX demonstrated a safety profile consistent with previous LDX studies in children or adults.

Source of Funding: Shire Development, Inc.

Monitoring Social Behavior Using the Child Conflict Index in Children with Attention Deficit Hyperactivity Disorder (ADHD) Treated with OROS Methylphenidate

Background: This study evaluated social behavior of children treated with OROS methylphenidate (OROSMPH) for attention deficit hyperactivity disorder (ADHD).

Methods: We analyzed combined data from two open-label dose-optimization studies evaluating OROS MPH in children aged nine to 12 years with ADHD (ClinicalTrials.gov records NCT00799409, NCT00799487). OROS MPH (18 mg/d) treatment was initiated with incremental dosage increases of 18 mg/d every three to seven days to an optimal individualized dosage (maximum 54 mg/d). Subjects continued their optimized dosage for up to six weeks. The Child Conflict Index (CCI), which captures conflict/attention-seeking and negativity/withdrawal behaviors, was administered by telephone to parents/caregivers just before baseline, at each dosage-adjustment visit, and at the final study visit. Changes from baseline total CCI scores were evaluated using paired Student t-tests. CCI was a secondary measure and no adjustments were made for multiple testing.

Results: Of 167 subjects in the safety analysis set, 115 were boys and 52 were girls. Mean optimized OROS MPH dosage was 41.1 mg/d (SD = 12.9). Mean baseline total CCI score was 8.5 (SD = 4.22) for boys and 7.0 (SD = 2.92) for girls. Statistically significant (p < 0.05) improvements in total CCI score occurred at the first dosage adjustment visit (change of −1.9 [SD = 4.24] for boys, −1.0 [SD = 2.61] for girls) and at most subsequent dosage-adjustment visits for each gender group and dosage level (18 mg/d, 36 mg/d and 54 mg/d). At the final study visit, both boys and girls had statistically significant improvements from baseline CCI scores (change of −6.2 [SD = 3.99] for boys, −4.9 [SD = 3.75] for girls; p < 0.0001 for both). Two subjects discontinued because of adverse events (AEs) during the dosage adjustment period. AEs reported by = 10% of subjects were headache, abdominal pain upper, decreased appetite, irritability and initial insomnia. No serious AEs or deaths were reported.

Conclusions: Statistically significant improvement on the CCI occurred during the first week of OROS MPH treatment and was sustained at the final study visit for both girls and boys. No unexpected or severe AEs were reported.

Source of Funding: Ortho-McNeil Janssen Scientific Affairs, LLC.

Late-Day Attention Deficit Hyperactivity Disorder (ADHD) Symptom Control Improvement with Extended-Release Dexmethylphenidate in Children with ADHD of All Ethnicities: A Sub-Analysis

Background: Dexmethylphenidate (d-MPH) allows similar efficacy to be achieved at half the dose of MPH. ¹ This study evaluated (post hoc) the efficacy of dexmethylphenidate extended-release (d-MPH-ER) 30 mg compared with d-MPH-ER 20 mg across children of different ethnicities in a laboratory-classroom setting. ²

Methods: In a double-blind, crossover study, children six to 12 years old with attention deficit hyperactivity disorder (ADHD) stabilized on MPH (40–60 mg/d) or d-MPH (20–30 mg/d) were randomized to receive d-MPH-ER 20 mg/d, 30 mg/d, and placebo for seven days each. The final dose of each treatment was administered in a laboratory-classroom setting. Efficacy was measured using the pre- to postdose change of the average Swanson, Kotkin, Agler, M-Flynn and Pelham (SKAMP) combined (attention and deportment) score at 10, 11 and 12 hours postdose (avg 10–12) for all ethnicities.

Results: A total of 165 children were randomized (94 boys; mean age 9.6 ± 1.8 years); 63 Caucasian, 52 African-American, 37 Hispanic and 13 Other (numbers too small to compare). Avg 10–12 changes from predose in SKAMP-Combined scores were numerically greater for d-MPH-ER 30 mg compared with d-MPH-ER 20 mg in all ethnic subgroups: African-American, −5.30 versus −3.07, respectively, p = 0.090; Hispanic, −6.57 versus −4.73, respectively, p = 0.359; and statistically superior in the Caucasian group, −3.53 versus −0.35, respectively, p = 0.005. All Avg 10–12 changes from predose in SKAMP-Combined scores by ethnicity were numerically and statistically superior to placebo in the d-MPH-ER 30-mg and 20 mg analyses.

Conclusions: ADHD symptoms significantly improved with d-MPH-ER 30 mg compared with d-MPH-ER 20 mg at hours 10–12 in all ethnic parameters, with a statistically significant difference in the Caucasian subgroup. Thus, d-MPH-ER 30 mg may provide further benefit to patients of all ethnic backgrounds who do not obtain optimal late-day symptom control with d-MPH-ER 20 mg.

Source of Funding: Novartis Pharmaceuticals Corporation.

Extended-Release Dexmethylphenidate 30 mg Improves Late-Day Attention Deficit Hyperactivity Disorder (ADHD) Symptom Control in Children with ADHD: A Randomized, Double-Blind Crossover Study

Background: Because no racemization occurs with the pharmacologically active enantiomer of methylphenidate, dex-methylphenidate (d-MPH), similar efficacy can be achieved at half the dose of MPH; a dose of 20 mg d-MPH extended-release (ER) is comparable to 40 mg MPH. The maximum approved MPH dose is 60 mg. Two recent studies suggested a differential response between 20 mg and 30 mg d-MPH-ER during the later part of the day. ^1,2 The current study evaluated the efficacy and safety of 20 mg- and 30 mg- d-MPH-ER in children with attention deficit hyperactivity disorder (ADHD) over a 12-hour laboratory classroom day.

Methods: In a randomized, double-blind, three periods x three treatments, crossover study, six to 12-year-olds with ADHD stabilized on MPH 40–60 mg/d or d-MPH 20–30 mg/d were randomized to receive d-MPH-ER 20 mg/d, 30 mg/d and placebo for seven days each. The final dose of each treatment was administered in a laboratory-classroom. Primary efficacy outcome measured changes in the average Swanson, Kotkin, Agler, M-Flynn and Pelham (SKAMP)-combined score from predose to 10, 11 and 12 hours postdose (avg 10–12). Adverse events (AEs) and vital signs were noted.

Results: A total of 165 children (94 boys; age 9.6 ± 1.8 years) were randomized, and 162 were included in the intent-to-treat analysis. Mean avg 10–12 change from predose in all efficacy outcome measures were significantly greater for d-MPH-ER 30 mg compared with d-MPH-ER 20 mg: SKAMP-Combined score, −4.47 versus −2.02, respectively, p = 0.002; SKAMP-Deportment score, −1.49 versus −0.39, respectively, p = 0.019; SKAMP-Attention score, −2.62 versus −1.33, respectively, p < 0.001; Math Test-Attempted score, 28.03 versus 18.76, respectively, p = 0.002; Math Test-Correct score, 28.02 versus 18.45, respectively, p = 0.002. Most common AEs ( = 3%) were decreased appetite, headache, abdominal pain and tachycardia.

Conclusions: ADHD symptoms improved significantly with d-MPH-ER 30 mg versus d-MPH-ER 20 mg at hours 10–12. d-MPH-ER 30 mg may provide further benefit to patients who do not obtain optimal later-day symptom control with d-MPH-ER 20 mg.

Source of Funding: Novartis Pharmaceuticals Corporation.

Inter-Rater Reliability in the Assessment of Pediatric Schizophrenia Using the Positive and Negative Syndrome Scale (PANSS): Training Results from a Russian Cohort

Background: The onset of schizophrenia is often in late adolescence or early adulthood though in some cases can be earlier with paediatric schizophrenia affecting approximately two in one million in the general population. ¹ The assessment of schizophrenia in childhood or early adolescence demands a somewhat different style of interviewing technique but the core psychopathology is essentially similar to what may be seen in older patients: delusions, hallucinations and a range of other symptoms including social withdrawal and cognitive problems. The Positive and Negative Syndrome Scale (PANSS) is a primary measure in research used to assess these symptoms. In this study we looked at training data to determine if raters in this group were more or less likely to achieve similar reliability to that expected in adult samples. Because this was a Russian cohort and the materials were originally produced in English, we were also interested in potential linguistic or cultural effects as this has emerged as an important issue in the global standardization of research practices.

Methods: Two early adolescent-aged patients with DSM-IV diagnoses of schizophrenia were rated by video using the PANSS in a training exercise with 30 Russian psychiatrists. All materials were translated into Russian and validated and scores were obtained by rating videos that were subtitled in Russian with transcripts available. Inter-rater reliability was obtained by using the intra-class correlation coefficient (ICC) statistic for total scale and for the positive, negative and general subscales.

Results: Raters achieved comparable inter-reliability to that reported in the literature for the rating of videos. There appeared to be little difference in ICCs for total scale and positive, negative and general subscales. The negative subscale was slightly less reliable in this group with 0.976 and 0.979 for pre- and post-training respectively, but this is still in the excellent range. Pre- and post-training ICCs for the positive subscale ICCs were 0.987 and 0.989, general 0.986 and 0.970 and total scale 0.987 and 0.976.

Conclusions: A number of previous studies utilizing this methodology (rating from video) have reported ICCs in this range for the PANSS rating adults. ^2,3,4 Our data suggests that there does not appear to be a significant impact of rating child versus adult patients using the PANSS instrument with very high ICCs obtained in this small cohort of Russian psychiatrists. Linguistic and cultural considerations did not appear to impact overall reliability but may have had a limited effect on individual item reliability. Caution must be exercised in the interpretation of these results due to the limited sample size and because ICCs can be inflated by the reduction of information variance inherent in rating from video.

Source of Funding: ProPhase LLC.

Conversion of Lisdexamfetamine Dimesylate to Active d-Amphetamine

Background: Lisdexamfetamine dimesylate (LDX, Vyvanse^®), a long-acting prodrug stimulant, requires conversion to d-amphetamine (d-amp) for therapeutic activity. LDX is absorbed throughout the intestines, likely by transporter-mediated uptake. However, a candidate transporter, human peptide transporter-1 (hPEPT1), is not highly expressed in the large intestines. Conversion of LDX to d-amp occurs primarily in the blood, specifically red blood cells (RBCs). These in vitro studies examine potential conversion by contents of the large intestines and potential conversion in blood-containing pathologically deformed RBCs.

Methods: In separate studies, fresh cecum contents from three male rats and colostomy specimens from two human donors were diluted with 0.1 M potassium phosphate buffer (pH = 7.4). Incubations up to four hours with LDX (1 μg/mL) at 37°C under N2 stream were performed. Fresh blood samples from two human male donors with sickle cell disease and two healthy control donors were incubated for up to four hours with LDX (1 μg/mL) at 37°C. LDX and d-amp were measured by a validated LC/MS/MS method.

Results: LDX concentrations declined over time with rat cecum contents; after four hours, 7.2% of initial LDX remained. Mean d-amp levels rose concurrently (62.1 to 2735.1 ng/mL at zero and four hours, respectively). Control incubations containing LDX and buffer showed no decline in LDX and no increase in d-amp. No decline in LDX and increase in d-amp were observed with human colostomy specimens. In incubations of blood from the two donors with sickle cell disease, LDX concentrations declined over time with 14.1% and 15.3% of initial LDX remaining after four hours. Similarly, in incubations of blood from two healthy donors, LDX concentrations declined over time with 13.1% and 10.5% of initial LDX remaining. Half-lives of LDX disappearance were 1.30 and 1.36 hours for the donors with sickle cell disease and 1.15 and 1.13 hours for the healthy donors. In blood samples from donors with sickle cell disease and healthy controls, d-amp concentrations rose similarly.

Conclusions: Conversion of LDX occurred with rat cecum contents but not with human colostomy specimens. Any unabsorbed LDX reaching the large intestines may therefore remain intact and be available for absorption by a different transporter to hPEPT1. Delivery of active d-amp from LDX is similar in subjects with or without sickle cell disease.

Source of Funding: Shire Development, Inc.

Feasibility Trial for Information Technology Enabled Treatment of Adolescent Depression

Background: The need for treatment of adolescent depression arises from the cumulative burden of major depressive disorder (MDD) and the costs that untreated adolescent depression have on individual development and society as a whole. Cumulative prevalence rates of MDD among adolescents are reported to be 12–22% by the age of 21 years. Effective treatments have been developed for adolescent MDD. Unfortunately, multiple barriers interfere with the implementation of “gold standard” care particularly due to the scarcity of trained child psychiatrists and other mental health clinicians. This trial focuses testing of an Information Technology Enabled Management System for Treatment of Adolescent Depression (ITEMS-TAD) that guides nurses in the delivery of evidence-based treatment for adolescent MDD.

Methods: We developed a prototype of ITEMS-TAD that includes four main components: a computer assisted telephone interview, a multi-source integrated treatment database, a set of treatment adjuster algorithms, and a multimedia training program. We conducted a Rapid Iterative Testing and Evaluation Study (RITE) with eight nurses and adolescents to refine the content and assess the usability of the technology in its delivery; then a Pilot Efficacy Study (PES) with four nurses and 30 adolescents to evaluate preliminary indicators of efficacy of the ITEMS-TAD approach, monitor therapeutic alliances, and assess patient and provider satisfaction and acceptance. Both the RITE and the PES consisted of three phone-based CBT sessions focusing on teaching basic cognitive behavioral skills to study adolescents. Sessions were delivered weekly and lasted approximately an hour.

Results: To date, eight subjects have completed the RITE and 11 the PES. Preliminary analyses suggest that nurses adhered to the principles of therapeutic alliance. Eight of the adolescents reported high levels of satisfaction. Two of them expressed some concerns with the length and materials used. So far, nearly all subjects (10) found the project useful and would consider a regular course of treatment.

Conclusions: We aim to demonstrate the ability of nurses to establish a positive working relationship with adolescents and families and to teach basic cognitive behavioral skills. If successful, the approach used in this trial will be refined, expanded, and scaled-up to develop a fully explicated version of ITEMS-TAD in Phase II.

Source of Funding: National Institute of Mental Health (RMH085350A).

Genetic Findings of the Vasopressin/Oxytocin System and Childhood Aggression

Background: Childhood onset aggression is a complex behaviour with diverse theories about its etiology and little agreement on how to define the phenotype. There has been a growing recognition that heritable genetic factors may increase the susceptibility to aggressive behaviour and need to be included in explanatory models of how aggressive behaviour develops. For example, twin studies have shown that the callous-unemotional trait, as a subtype of aggression, has a high heritability, making it a good candidate for molecular genetic studies. While neurotransmitter systems have been implicated in the, development and expression of aggression, recent studies in both animals and humans have demonstrated that the vasopressin/oxcytocin system is strongly linked to aggressive behaviour; however, there are no genetic linkage or association studies exploring the association.

Methods: This sample has been described previously and consisted of 165 children, with a mean age of 11 years. Children were recruited via psychiatric referral and all had at least a two-year history of aggressive behaviour. Inclusion criteria consisted of scores above the 90th percentile on the aggressive subscale of the Child Behaviour Checklist (CBC) and Teacher's Report Form (TRF). The Callous-Unemotional subscale of the Antisocial Process Screening Device (APSD) was used to assess levels of the callous-unemotional trait. One single nucleotide polymorphism (SNP) in the AVPR1b gene and eight SNPs in the AVP/OXT gene region were genotyped in these children. Analyses compared genotype frequencies of the aggressive children with healthy comparison adult subjects. Analyses also compared scores on the callous-unemotional subscale of the APSD with genotype frequencies in the aggressive children.

Results: In the AVPR1b receptor, one SNP was significantly associated with aggression in the analysis of cases versus controls. In the AVP/OXT genes, three SNPs were significantly associated with the children's scores on the APSD, indicating an association with the callous-unemotional trait.

Conclusions: The OXT/AVP gene system is associated with aggressive behaviour and with the callous-unemotional trait. This is the first study to demonstrate an association between the vasopressin/oxcytocin gene system and childhood aggression. These findings identify potential new targets for the development of therapeutic pharmaceuticals.

Source of Funding: Centre for Addiction and Mental Health Foundation.

Comparative Efficacy of Daily versus Split Dosing of Atomoxetine on School and Home Functioning in Children with Attention Deficit Hyperactivity Disorder (ADHD)

Rationale: Comparative studies of once versus twice daily dosing of atomoxetine have demonstrated the efficacy of both strategies for treating attention deficit hyperactive disorder (ADHD) in adults ¹ without evidence of differential effects by dosing pattern. However, all prior work relied on subjective symptom reports, and no such study has been completed in children.

Objective: To compare the effect of single versus split dosing of atomoxetine on school and home functioning of children with ADHD.

Methods: Post-hoc analysis compared efficacy of QAM versus BID Atomoxetine dosing during an eight-week randomized open label trial designed to compare the efficacy of behavior therapy (BT) alone versus BT plus atomoxetine. ² All subjects were started on QAM dosing. Switching to BID dosing occurred by Week 4 at the physicians' discretion to address tolerability and efficacy concerns. Thirty-four subjects were maintained on QAM dosing while 22 went to BID dosing. There were no significant differences in demographics, baseline symptom severity or rates of BT usage between dosing groups. The mean dose was higher in the BID group (1.56 mg/kg/day versus 1.35 mg/kg/day, p < 0.05). The primary efficacy measure was change in classroom rule violations scored by direct observation. Teachers and parents also completed ratings of disruptive symptoms, academic functioning and side effects.

Results: While both groups improved over time, there was a significantly greater decrease in classroom rule violations with QAM versus BID dosing (f = 4.25, p < 0.05, effect size (ES) = 0.71). Teacher-rated oppositional defiant disorder symptoms (f = 4.19, p < 0.05, ES = 0.65), peer relationships (f = 5.17, p < 0.05; ES = 0.33), parent-child relationships (f = 5.46; p < .05; ES = 0.50), percentage of daily classroom goals achieved (f = 2.55, p < 0.10, ES = 0.28) and parent-rated academic functioning (F = 4.06, p < 0.05, ES = 0.36) also favored the QAM group. There were no appreciable group differences in side effects.

Conclusions: Compared to split dosing, once daily atomoxetine dosing is associated with greater improvements in school and home functioning in children with ADHD with no worsening of tolerability. Results are limited by the secondary nature of the analyses and small N but merit further investigation.

Source of Funding: Eli Lilly, U.S.A.

Improving Tic-Related Response Inhibition: Comparing the Effects of Dexmethylphenidate to No Medication in Children and Adolescents with Attention Deficit Hyperactivity Disorder (ADHD) and Chronic Tic Disorders

Objective: This is a pilot study using a Tic Suppression Paradigm (TSP) to test whether dexmethylphenidate (dMPH) can facilitate tic suppression in children and adolescents with attention deficit hyperactivity disorder (ADHD) and Tourette's disorder (TD) or chronic tic disorder. The primary hypothesis is that dMPH will improve reinforced tic suppression.

Methods: Children with TD and ADHD were given dMPH on one day and no medication on another in randomized order. On both days, following a baseline period, subjects were reinforced for suppressing tics using a standard TSP.

Results: Data were analyzed on ten subjects completed to date. Relative to the no medication condition, tics were reduced when children were on dMPH. Reinforcing tic suppression produced lower rates of tics compared to baseline, but dMPH did not improve this ability.

Conclusions: Preliminary results of this study indicate replication of prior studies of behavioral tic suppression in youth with TD without ADHD; in addition, our results indicated tic reduction (and not tic exacerbation) with acute dMPH challenge in children and adolescents with ADHD and TD or chronic tic disorder.

Source of Funding: American Association of Child and Adolescent Psychiatry, Elaine Schlosser Lewis Fund, Tourette Syndrome Association.

Anatomical Evidence for Reticular Activating System and Cerebellar Involvement in Adults with Attention Deficit Hyperactivity Disorder: A Retrospective Analysis

Background: There is a growing need to integrate data from functional brain imaging and neuropsychological assessment to elucidate the role of executive functioning in adults with attention deficit hyperactivity disorder (ADHD). ¹ The current study examined functional differences (as measured by regional cerebral blood flow (rCBF)) between adults with ADHD and normal controls during a rest and concentration (CPT-I administration) scan.

Methods: Participants included 82 individuals that were recruited from a clinic in Southern California that provides neurological services including brain imaging. After screening for inclusion/exclusion criteria, there were a total of 41 participants per group. Single Photon Emission Computerized Tomography (SPECT) was used to examine Cerebral Blood Flow (rCBF). Rest and concentration brain images were processed using the Odyssey software and normalized to a standard anatomical space using Statistical Parametric Mapping Version 2 software (SPM2) with gender as a covariate.

Results: Results from voxel-wise comparisons using Family-Wise Error (FWE) corrections demonstrated low rCBF in the left corpus callosum, t(72) = 5.95, p < 0.001, left pons t(72) = 6.11, p < 0.001, left medulla, t(72) = 5.21, p < 0.02, right superior frontal gyrus, t(72) = 5.52, p < 0.01 and right medulla t(72) = 6.38, p < 0.001 in the ADHD group when compared to the normal controls during the rest scan. During the concentration scan, low rCBF in the left corpus callosum, t(80) = 7.17, p < 0.001, left uncus, t(80) = 5.55, p < 0.004, left cerebellar tonsil, t(80) = 6.04, p < 0.001, right superior frontal gyrus, t(80) = 5.34, p < 0.01, right cerebellar tonsil, t(80) = 6.58, p < 0.001, pons, t(80) = 6.79, p < 0.001 and right posterior cingulate t(80) = 5.90, p < 0.001 in the ADHD group when compared to the normal controls. Results from the CPT-I, CCPT Index dichotomous variable, indicated that the ADHD group performed worse when compared to the non-psychiatric normal group, [chi] (1, N = 123) = 15.63, p < 0.001, during the concentration scan.

Conclusions: This may be the first neuro-imaging study that captures the brainstem as a possible anatomical structure in adult ADHD. Results from this study suggest that adult ADHD may involve deficits of the reticular activating system and the cerebellum. Pharmacological agents with a high affinity for the brainstem may innervate the catecholaminergic system to modulate arousal throughout the cortex in adults with ADHD and improve deficits of attention. ²

Source of Funding: None.