Biomarkers of ADHD Treatment Response (BAT)

The goal of this translational biomarker study is to use electroencephalography (EEG) to identify brain signatures that will predict a child's response to two of the most commonly prescribed ADHD medications, methylphenidate and mixed amphetamine salts. The main questions the investigators aim to answer are: Do children with ADHD who show symptom reduction with methylphenidate have different EEG profiles than children who do not respond well to methylphenidate? Do children who respond better to mixed amphetamine salts than to methylphenidate have unique EEG profiles? The investigators will measure brain activity before the participating children have tried any stimulant medications, and then again after a 3-week trial of Concerta (methylphenidate). Participants who do not show significant symptom improvement on Concerta will then complete a 3-week trial of Adderall (mixed amphetamine salts), and the study will measure brain activity while those children are on the best dose of Adderall. The investigators will collect information from the child, caregivers, and teachers each week to measure ADHD symptom improvement and side effects. This study will therefore follow the typical treatment approach used in the Boston Children's Hospital Developmental Medicine Clinic, but the investigators will add measures of brain functioning before and after medication.

Attention deficit hyperactivity disorder (ADHD) is a highly prevalent neurodevelopmental disorder associated with psychiatric, social, academic, occupational, and health impairments across the lifetime. Although pharmacological interventions for pediatric ADHD are safe and effective, there is considerable variability in treatment response at the individual level. As a result, identification of optimal medication class and dose is often not attained in community clinical settings. The current application constitutes a translational biomarkers study aimed at identifying electroencephalography (EEG) and event related potential (ERP) biomarkers of preferential response to two commonly prescribed psychostimulants among children with ADHD. The results of this investigation will improve understanding of individual differences in neurobiological mechanisms of ADHD and provide preliminary data for a large-scale clinical trial aimed at developing a precision medicine care model for pharmacological treatment of ADHD. With the support of the Translational Neuroscience Center Clinical Research Operations services and in collaboration with the Boston Children's Hospital Primary Care Center (CHPCC), the investigators will execute a sequential crossover design study examining pre-treatment EEG and ERP biomarkers of response to methylphenidate (MPH; Concerta) among all children and to mixed amphetamine salts (MAS; Adderall-XR) among children with suboptimal response to MPH. Additional pre- and post-treatment assessments will be integrated with standard clinical care provided by Dr. Chan (co-PI) in the Division of Developmental Medicine. The study will recruit 30 stimulant-treatment-naïve children with ADHD, ages 7-11, from the CHPCC. Additionally, analyses will capitalize on Dr. Arnett's (co-PI) existing EEG/ERP data on 40 typically developing (TD) children in the same age range to maximize power for statistical comparisons. The investigators hypothesize that, consistent with Dr. Arnett's prior work, the EEG and ERP profiles will differentiate children with positive response to MPH versus preferential response to MAS. Specifically, the investigators hypothesize that MPH responders will have reduced P3 ERP amplitude and normal aperiodic spectral slope, while MAS preferential responders will have normal P3 amplitude and flatter aperiodic spectral slope. The investigators expect that slow individual alpha peak will be associated with reduced response to both MPH and MAS, as suggested by prior literature. Additionally, the investigators hypothesize that at optimal dosing, treatment-related change in EEG/ERP biomarkers will be associated with ADHD symptom improvement; this will indicate that individual differences in psychostimulant response reflect individual differences in the neurobiological etiology of ADHD symptoms. The results of this pilot study will support application for federal funding for a large-scale clinical trial. The long-term outcomes of this line of research stand to benefit children and families with ADHD, as well as children with other primary diagnoses commonly associated with ADHD (e.g., autism spectrum disorder; genetic syndromes). Moreover, differences in neurophysiological correlates of differential stimulant response have potential to increase our knowledge of neural mechanisms underlying psychostimulant medication effects.

This is a sequential partial-crossover design. All participants will complete the Baseline visit, the methylphenidate trial, and a follow-up lab visit. Participants with < 30% symptom improvement on methylphenidate will then complete a 1-2 week washout, followed by a trial of mixed amphetamine salts and a second follow-up lab visit.

3-week trial of oral mixed amphetamine salts, extended release, administered only to children who do not show at least 30% improvement during the Concerta/Methylphenidate trial.

Vanderbilt ADHD Ratings from the parent and teacher will be collected at baseline and weekly during each of the 3-week medication trials to guide further dosing and study disposition. Patients with at least 30% ADHD symptom reduction on both parent and teacher rating scales at the end of the 3-week Concerta trial will return to standard clinical care with their primary care provider. Patients with less than 30% ADHD symptom reduction at the end of the 3-week Concerta/Methylphenidate trial will complete a 1-2 week medication washout and then a 3-week Adderall-XR trial. Vanderbilt items are rated on a scale of 0-3 and averaged. Higher scores indicate more symptoms.

Vanderbilt ADHD Ratings from the parent and teacher will be collected at baseline and weekly during each of the 3-week medication trials to guide further dosing and study disposition. Vanderbilt items are rated on a scale of 0-3 and averaged. Higher scores indicate more symptoms.

Change in functioning and ADHD-related impairment will be measured with the Clinical Global Impressions-Improvement scale (CGI-I) after 3 weeks of the Concerta trial. Ratings are made by the clinician on a scale of 1-7, with lower scores indicating fewer symptoms after treatment. Consistent with previous research, scores of 2 "much improved" or 1 "very much improved" will be considered clinically significant improvement and will be used to categorize patients as treatment responders for analyses.

Change in functioning and ADHD-related impairment will be measured with the Clinical Global Impressions-Improvement scale (CGI-I) after 3 weeks of the Adderall-XR trial. Ratings are made by the clinician on a scale of 1-7, with lower scores indicating fewer symptoms after treatment. Consistent with previous research, scores of 2 "much improved" or 1 "very much improved" will be considered clinically significant improvement and will be used to categorize patients as treatment responders for analyses.

Change in functioning and ADHD-related impairment from Baseline to post-treatment will be measured with the Clinical Global Impressions-Improvement scale (CGI-I) 3 months after the final medication trial. Ratings are made by the clinician on a scale of 1-7, with lower scores indicating fewer symptoms after treatment. Consistent with previous research, scores of 2 "much improved" or 1 "very much improved" will be considered clinically significant improvement and will be used to categorize patients as treatment responders for analyses.

Change in functioning and ADHD-related impairment from Baseline to post-treatment, and maintenance of improvement will be measured with the Clinical Global Impressions-Improvement scale (CGI-I) 6 months after the final medication trial. Ratings are made by the clinician on a scale of 1-7, with lower scores indicating fewer symptoms after treatment. Consistent with previous research, scores of 2 "much improved" or 1 "very much improved" will be considered clinically significant improvement and will be used to categorize patients as treatment responders for analyses.

EEG resting aperiodic spectral slope will be derived from the lights-off resting paradigm at baseline, with the hypothesis that unmedicated individual aperiodic slope will predict treatment response to Concerta and/or Adderall-XR. Aperiodic slope will be quantified as the aperiodic exponent from 1-55 hz using the FOOOF toolbox in MATLAB.

EEG resting alpha peak frequency will be derived from the lights-off resting paradigm at baseline, with the hypothesis that unmedicated individual alpha peak frequency will predict treatment response to Concerta and/or Adderall-XR. Alpha peak frequency will be quantified as the highest peak between 7-13 hz for which there is lower power .2 hz below and .2 hz above, indicating a true peak rather than a local maximum.

Average P3 amplitude derived from novel stimuli in a visual oddball task will be measured as the mean amplitude in a window determined via visual inspection and literature review. The hypothesis is that unmedicated individual novelty P3 amplitude will predict treatment response to Concerta and/or Adderall-XR.

Average P3 amplitude derived from cue stimuli in a cued go-no-go task will be measured as the mean amplitude in a window determined via visual inspection and literature review. The hypothesis is that unmedicated individual cued P3 amplitude will predict treatment response to Concerta and/or Adderall-XR.

EEG resting aperiodic spectral slope will be derived from the lights-off resting paradigm at Baseline (unmedicated) and at the end of the Concerta Trial while the child is on their optimal dose of Concerta.

EEG resting alpha peak frequency will be derived from the lights-off resting paradigm at Baseline (unmedicated) and at the end of the Concerta Trial while the child is on their optimal dose of Concerta.

Average P3 amplitude derived from novel stimulus trials in a visual oddball task will be measured as the mean amplitude in a window determined via visual inspection and literature review. The hypothesis is that change in novelty P3 amplitude from Baseline to optimal dose of Concerta will be associated with change in ADHD symptom severity.

Average P3 amplitude derived from cue stimuli in a cued go-no-go task will be measured as the mean amplitude in a window determined via visual inspection and literature review. The hypothesis is that change in cue P3 amplitude from Baseline to optimal dose of Concerta will be associated with change in ADHD symptom severity.

EEG resting aperiodic spectral slope will be derived from the lights-off resting paradigm at Baseline (unmedicated) and at the end of the Adderall-XR Trial while the child is on their optimal dose of Adderall-XR.

EEG resting alpha peak frequency will be derived from the lights-off resting paradigm at Baseline (unmedicated) and at the end of the Adderall-XR Trial while the child is on their optimal dose of Adderall-XR.

Average P3 amplitude derived from novel stimulus trials in a visual oddball task will be measured as the mean amplitude in a window determined via visual inspection and literature review. The hypothesis is that change in novelty P3 amplitude from Baseline to optimal dose of Adderall-XR will be associated with change in ADHD symptom severity.

Average P3 amplitude derived from cue stimuli in a cued go-no-go task will be measured as the mean amplitude in a window determined via visual inspection and literature review. The hypothesis is that change in cue P3 amplitude from Baseline to optimal dose of Adderall-XR will be associated with change in ADHD symptom severity.

In addition to symptom reduction, treatment side effects and tolerance will be examined as a secondary endpoint. Severity of side effects will be rated by the parent and teacher on the ADHD Vanderbilt Rating scales at baseline and weekly during the 3-week Concerta trial. At the end of the 3-week trial, each child will be assigned a summary tolerance rating by the research team: 1 = no side effects, 2=mild side effects, or 3=significant side effects requiring treatment discontinuation.

In addition to symptom reduction, treatment side effects and tolerance will be examined as a secondary endpoint. Severity of side effects will be rated by the parent and teacher on the ADHD Vanderbilt Rating scales at baseline and weekly during the 3-week Adderall-XR trial. At the end of the 3-week trial, each child will be assigned a summary tolerance rating by the research team: 1 = no side effects, 2=mild side effects, or 3=significant side effects requiring treatment discontinuation.

Commission errors on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Concerta trial when the patient is on their optimal dose of Concerta as a measure of treatment effects on inhibitory control. TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Commission errors on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Adderall-XR trial when the patient is on their optimal dose of Adderall-XR, as a measure of treatment effects on inhibitory control. TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Response time on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Concerta trial when the patient is on their optimal dose of Concerta, as a measure of treatment effects on response time (speed). TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Response time on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Adderall-XR trial when the patient is on their optimal dose of Adderall-XR, as a measure of treatment effects on response time (speed). TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Omission errors on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Concerta trial when the patient is on their optimal dose of Concerta, as a measure of treatment effects on attention maintenance. TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Omission errors on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Adderall-XR trial when the patient is on their optimal dose of Adderall-XR, as a measure of treatment effects on attention maintenance. TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Variability of response time on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Concerta trial when the patient is on their optimal dose of Concerta, as a measure of treatment effects on response time variability (consistency). TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Variability of response time on the Tests of Variables of Attention (TOVA) will be examined at baseline and at the end of the 3-week Adderall-XR trial when the patient is on their optimal dose of Adderall-XR, as a measure of treatment effects on response time variability (consistency). TOVA scores are reported as standard scores with a mean of 100 and standard deviation of 15 (range: <40 - >120). Higher scores indicate better performance.

Longitudinal associations between ADHD symptom reduction and internalizing symptoms (depression, anxiety) will be measured using the Behavior Assessment System for Children, 3rd Edition (BASC-3) Internalizing Scale at Baseline and at 3 months after study completion. Longitudinal bivariate associations will be modeled using an autoregressive cross-lagged path model. BASC-3 scores are reported as T-scores with a mean of 50 and standard deviation of 10 (range: 20-120). Higher scores indicate more pathology.

Longitudinal associations between ADHD symptom reduction and internalizing symptoms (depression, anxiety) will be measured using the Behavior Assessment System for Children, 3rd Edition (BASC-3) Internalizing Scale at Baseline and at 6 months after study completion. Longitudinal bivariate associations will be modeled using an autoregressive cross-lagged path model. BASC-3 scores are reported as T-scores with a mean of 50 and standard deviation of 10 (range: 20-120). Higher scores indicate more pathology.

Longitudinal associations between ADHD symptom reduction and externalizing symptoms (oppositional defiance, aggression, conduct problems) will be measured using the Behavior Assessment System for Children, 3rd Edition (BASC-3) Externalizing Scale at Baseline and at 3 months after trial completion. Bivariate associations will be modeled across all 3 time points using an autoregressive cross-lagged path model. BASC-3 scores are reported as T-scores with a mean of 50 and standard deviation of 10 (range: 20-120). Higher scores indicate more pathology.

Longitudinal associations between ADHD symptom reduction and externalizing symptoms (oppositional defiance, aggression, conduct problems) will be measured using the Behavior Assessment System for Children, 3rd Edition (BASC-3) Externalizing Scale at Baseline and at 6 months after trial completion. Bivariate associations will be modeled across all 3 time points using an autoregressive cross-lagged path model. BASC-3 scores are reported as T-scores with a mean of 50 and standard deviation of 10 (range: 20-120). Higher scores indicate more pathology.

Inclusion Criteria: Pediatric patients ages 7-11 seen at the Children's Hospital Primary Care Center Have a diagnosis of ADHD or referred for an ADHD evaluation Have not previously trialed stimulant medication Exclusion Criteria: Diagnoses of intellectual disability, autism, prior suicide attempt, current psychotropic medication use, known genetic syndrome, color-blindness History of nonfebrile seizures Gestational age < 32 weeks Prenatal alcohol or substance exposure Medical conditions that contraindicate psychostimulant use (e.g., cardiac concerns).