long-term tolerability and effectiveness of once-daily mixed amphetamine salts (adderall xr) in...

Long-Term Tolerability and Effectiveness of Once-DailyMixed Amphetamine Salts (Adderall XR) in

Children With ADHD

JAMES J. MCGOUGH, M.D., JOSEPH BIEDERMAN, M.D., SHARON B. WIGAL, PH.D.,

FRANK A. LOPEZ, M.D., JAMES T. MCCRACKEN, M.D., THOMAS SPENCER, M.D.,

YUXIN ZHANG, PH.D., AND SIMON J. TULLOCH, M.D.

ABSTRACT

Objective: To evaluate the long-term tolerability and effectiveness of extended-release mixed amphetamine salts (MAS

XR; Adderall XR�) in children with attention-deficit/hyperactivity disorder (ADHD). Method: This was a 24-month, multi-

center, open-label extension of TWO placebo-controlled studies of MAS XR in children with ADHD aged 6 to 12 years.

Subjects (N = 568) began treatment with MAS XR 10 mg once daily, with 10-mg weekly dose increases to optimal effec-

tiveness (maximum dose, 30 mg/d). Effectiveness was assessed with analysis of quarterly Conners Global Index Scale,

Parent version (CGIS-P) scores. Tolerability was assessed by monitoring adverse events (AEs), vital signs, physical ex-

aminations, and serial laboratory tests.Results:Significant improvements (.30%, p, .001) in CGIS-P scores were main-

tained during long-term treatment. Treatment was well tolerated, and most AEs were mild. The most frequently reported

drug-related AEs included anorexia, insomnia, and headache. The incidence of drug-related AEs increased with increasing

MAS XR dose, suggesting a dose relationship. Changes in laboratory values and vital signs were modest and not clinically

meaningful. Conclusions: In children with ADHD, once-daily 10 mg–30 mg MAS XR was well tolerated and significant

behavioral improvements were consistently maintained during 24 months of treatment. J. Am. Acad. Child Adolesc. Psy-

chiatry, 2005;44(6):530–538. Key Words: attention-deficit/hyperactivity disorder, amphetamine, Adderall XR, long-term

treatment.

Attention-deficit/hyperactivity disorder (ADHD) is themost common neurobehavioral disorder of childhood,affecting an estimated 5% to 7% of school-age children(Wolraich et al., 1996). Core features include symp-

toms of inattention, impulsivity, or hyperactivity thatare developmentally inappropriate. Although ADHDwas once thought to affect only children, follow-upstudies indicate that it is a chronic disorder that fre-quently continues into adulthood (Barkley et al.,2002). Over the lifespan, ADHD is associated withincreased academic difficulties, family and social dys-function, occupational challenges, substance abuse/de-pendence, and antisocial behavior (Wilens et al., 2002).

Stimulant medications have proven useful in ADHDtreatment for more than a half century (Bradley, 1937).There are more published short-term stimulant studiesfor ADHD than for any other child psychiatric disorder(Greenhill et al., 1999). However, few prospective long-term studies have addressed stimulant effectiveness andsafety, and the ultimate impact of sustained treatmenton long-term functioning is unknown.

Although stimulants are effective ADHD treat-ments, the necessity for multiple daily doses of

Accepted December 21, 2004.Drs. McGough andMcCracken are with the Division of Child and Adolescent

Psychiatry, UCLA Neuropsychiatric Institute, Los Angeles; Drs. Biederman andSpencer are with the Department of Psychiatry, Massachusetts General Hospital,Boston; Dr. Wigal is with the Child Development Center, University ofCalifornia– Irvine; Dr. Lopez is in private practice at the Children’s DevelopmentalCenter, Maitland, FL; Drs. Zhang and Tulloch are with Shire PharmaceuticalDevelopment Inc., Wayne, PA.This research was supported by Shire Pharmaceutical Development Inc.,

Wayne, PA. The authors acknowledge the contributions of the SLI381.302 StudyGroup and thank Tracey I. Fine, M.S., and Dana Randall, Pharm.D., for as-sistance with manuscript preparation.Reprint requests to Dr. McGough, 300 UCLA Medical Plaza, Suite 1414,

Los Angeles, CA 90095; e-mail: [email protected]/05/4406–0530�2005 Journal of the American Academy of

Child and Adolescent Psychiatry.

DOI: 10.1097/01.chi.0000157550.94702.a2

530 J . AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 44:6, JUNE 2005

immediate-release compounds led to difficulties withcompliance and social stigma, often resulting in treatmentdiscontinuation (Claxton et al., 2001; Hack and Chow,2001). The availability of once-daily formulations is likelyto increase compliance with stimulant treatment (Swanson,2003). Two previous studies demonstrated a 12-hour du-ration of effect and short-term safety and efficacy of anextended-release formulation of mixed amphetaminesalts (MAS XR; Adderall XR�) (Biederman et al.,2002; McCracken et al., 2003). This study examinedthe long-term tolerability and effectiveness of MAS XR.

METHOD

Subjects

Participating subjects were previously enrolled in one of twodouble-blind, placebo-controlled MAS XR studies. The first was a6-week, randomized, double-blind, crossover trial of three dosesof once-daily MAS XR (10, 20, or 30 mg), placebo, and an activecontrol (Adderall 10 mg) in 51 children (McCracken et al., 2003).This analog classroom study demonstrated an onset of action within1.5 hours after administration of MAS XR, as well as significantimprovements in behavioral and academic measures assessed up to12 hours. The second was a 4-week, multicenter, randomized, dou-ble-blind, parallel-group, placebo-controlled trial of once-dailyMAS XR (10, 20, or 30mg) or placebo in 584 children with ADHD(Biederman et al., 2002). This study was conducted in a naturalistichome and school setting and assessed diurnal variation in responsesbased onmorning and afternoon assessments. Children were recruitedinto both studies using a combination of advertising and distributionof information about study participation at local outpatient clinics.Eligible subjects in the current long-term study completed their pre-

vious trial without any clinically relevant adverse events (AEs) or hadwithdrawn from either study for reasons other than AEs. Subjects wereboys and girls, 6 to 12 years old, mostly with combined subtype, withvital signs in the normal range, who satisfiedDSM-IV criteria for a pri-mary diagnosis of ADHD (American Psychiatric Association, 1994).All diagnoses were based on psychiatric evaluation and selected mod-ules of the computerized parent version of the National Institute ofMental Health Diagnostic Interview Schedule for Children Version4.0 (Shaffer et al., 2000). All study procedures and informed con-sent/assent forms were approved by institutional review boards at eachparticipating research site. Subjects aged 7 years and older providedwritten assent for study participation, and parents or caregivers pro-vided written consent for their child’s enrollment after receiving bothwritten information and an oral explanation of study procedures.Children were ineligible to participate if they had a coexisting

psychiatric disorder requiring medication treatment, history of sei-zure, hypersensitivity to MAS, hyperthyroidism, or glaucoma. Pro-hibited concomitantmedications includeda2 agonists, anticonvulsantdrugs, and medications that affect blood pressure, heart rate, or cen-tral nervous system performance. There were no restrictions on con-comitant psychotherapy or other psychosocial interventions.

Study Design

This was a multicenter, 24-month, open-label, extension studyconducted at 45 research sites. The study comprised 28 clinic visits,

including the baseline visit, visits at weeks 1 through 4, and monthlyvisits thereafter. Subjects chose either to remain in the study or towithdraw at month 12. Whenever possible, the baseline visit for thelong-term study was the final study visit of the previous 6-week an-alog classroom study or the 4-week natural setting study. Of the 568patients who continued to receive MAS XR in the open-label exten-sion study, 525 participated in the natural setting study and 43 par-ticipated in the analog classroom study. The longest possible time offstudy medication was 5 months between antecedent controlled stud-ies and this extension.

Dosing

After the baseline visit, all subjects were initiated on a once-dailyoral dose of MAS XR 10mg regardless of theMAS XR dose receivedin previous studies. At weekly clinic visits during the first month, thedose could be increased in 10-mg increments to 20 or 30 mg/day toachieve optimal effectiveness as determined by the study physician.A 10-mg decrease in dose was allowed at any clinic visit if tolerabilitywas a concern. Subsequent dose adjustments in 10-mg incrementsbetween the three doses were permitted at monthly clinic visitsthroughout the study. ‘‘Drug holidays’’ limited to weekends andbrief vacations were allowed with prior approval from the investi-gator. Treatment compliance was assessed at each clinic visit by cap-sule count; noncompliance was defined as taking,80% or.120%of study medication.

Assessments

The primary effectiveness measure was the 10-item ConnersGlobal Index Scale, Parent version (CGIS-P) (Conners, 1997).The CGIS-P is a validated, standard short-scale instrument for as-sessing ADHD symptoms and monitoring treatment over time.Item scores range from 0 (‘‘not true at all’’) to 3 (‘‘very much true’’).The scale was administered at each clinic visit and, whenever pos-sible, was completed by the same parent or caregiver.

Safety Assessments

Safety was assessed by analysis of AEs and vital signs (e.g., sittingblood pressure and pulse) recorded at each study visit, height andweight at baseline and months 12 through 24, laboratory tests(hematology, serum chemistry, and urinalysis) conducted at baselineand 6-month intervals, physical examinations performed at baselineand 12-month intervals, and electrocardiograms (ECGs) performedat baseline and months 12, 18, and 24. A urine pregnancy test wasrequired at baseline and 3-month intervals during the study for fe-male subjects who had reached menarche.AE data were collected by spontaneous report and investigator

queries of subject and caregiver at each visit. Individual AEs werecounted separately at each visit that they were described. If a subjectreported an AE, the investigator assessed severity (mild, moderate,severe) and relationship to the study drug (not related, possiblyrelated, probably related), based on consideration of illness orprogression of disease, absence of other reasonable etiologies, thetemporal sequence of effects, and knowledge of the study drug basedon clinical experience. Serious AEs (SAEs) were defined as AEsresulting in death or hospitalization, or were life threatening or med-ically significant.

ADDERALL XR: LONG-TERM RESULTS

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 44:6, JUNE 2005 531

Statistical Analyses

The CGIS-P total scores were averaged across 3-month periodsand analyzed based on the intent-to-treat (ITT) population usinga two-way analysis of variance (ANOVA) for postbaseline compar-isons. The ITT population included all subjects who received at leastone dose of MAS XR and had at least one valid postbaseline CGIS-Pscore. The ANOVA model included the period (i.e., four quartersper year) and site effect, as well as the subject-within-site effect. Thesite effect was tested against the subject-within-site effect. TotalCGIS-P scores between adjacent periods were compared using a pro-file analysis based on the results from the ANOVAmodel. Subgroupanalyses of mean quarterly CGIS-P scores were stratified by a prioriconditions of treatment at study entry. The three subgroups for thesubgroup analysis were defined as (1) subjects who receivedMAS XRtreatment during transition from the previous study to the presentstudy, (2) subjects who did not receive active treatment during thetransition, and (3) subjects who received placebo during the previousstudy. For both the ANOVA and profile analysis, the type I errorsignificance level was .05.Safety data, including spontaneously reported AEs, were analyzed

using descriptive statistics for all subjects who received at least onedose of MAS XR. For laboratory values, vital signs, and ECG in-terval measurements, differences between baseline and postbaselinevalues were compared using paired t tests. Early withdrawals fromthe study were analyzed by subgroup and study period using theFisher exact test.

RESULTS

Clinical Characteristics

Five hundred sixty-eight subjects initially enrolledand 273 (48%) completed 24 months of treatment.Of the subjects who discontinued treatment before24 months, 87 (15%) withdrew consent, 84 (15%) dis-continued study medication due to AEs, 74 (13%) werelost to follow-up, 26 (5%) terminated for ‘‘other’’ rea-sons (including suboptimal response at 30 mg/day fortwo subjects), 16 (3%) had protocol violations, and 8(1%) lacked efficacy. The following factors did not in-fluence the duration of study participation: sex (p = .3),age (p = .9), race (p = .8), or ADHD subtype (p = .5).Subject demographics and baseline characteristics are

presented in Table 1. Most subjects were white boys(mean age, 8.7 years) with ADHD combined type(91%) and no comorbid psychiatric disorders. Forthe 168 subjects (30%) diagnosed with a psychiatric co-morbidity, diagnoses included oppositional defiant dis-order and/or conduct disorder, anxiety, and depression.Most subjects entered after participation in the naturalsetting study, in which psychiatric comorbidity includedoppositional defiant disorder (136/584 subjects, 23%),conduct disorder (49/584, 8%), separation anxiety dis-order (17/584, 3%), generalized anxiety disorder

(13/584, 2%), and depression (7/584, 1%). Approxi-mately 8% of children who enrolled in this long-termstudy had received an amphetamine formulation for thetreatment of ADHD, 14% had received methylpheni-date, and 41% had received unspecified stimulant med-ications before participation in blinded trials of MASXR.

Overall, 273 subjects receivedMAS XR treatment for.2 years, 284 for .18 months, 330 for at least 1 year,and 238 for ,1 year. Overall medication compliancerate was 94%. The initial MAS XR dose for all subjectswas 10 mg/day and mean doses ranged from 20 mg/dayat 3 months to 22 mg/day at 24 months. The majorityof subjects received 20 mg/day for the duration of thestudy. At 1month, themean (SD) dose was 16 (4.6) mg;at 6 months, 20.2 (7.6); at 12 months, 21.5 (7.5);at 18 months, 21.2 (7.5); and at 24 months, 22.4 (6.9).

Effectiveness

The two-way ANOVA of CGIS-P scores detected astatistically significant time effect by quarter (F7,2346 =4.01; p = .0002) for the ITT population and no siteeffect (F47,512 = 0.92; p = .6). The profile analysis

TABLE 1Demographic and Baseline Characteristics

Subject Populations

Characteristic Safetya (N = 568) ITTb (N = 560)

Age (yr), mean ± SD 8.7 ± 1.8 8.7 ± 1.86–8 yr, no. (%) 264 (46) 263 (47)

9–12+ yr, no. (%) 304 (54) 297 (53)Gender, n (%)Boys 442 (78) 436 (78)Girls 126 (22) 124 (22)

Race, n (%)White 412 (73) 407 (73)Black 67 (12) 65 (12)

Hispanic 53 (9) 53 (9)Asian/Pacific Islander 7 (1) 7 (1)Other 17 (3) 16 (3)

Weight (kg)Mean ± SD 33.5 ± 11.5 33.5 ± 11.6Range 16.4–93.6 16.4–93.6

Height (cm)Mean ± SD 134.4 ± 11.4 134.4 ± 11.2Range 104.1–170.2 104.1–170.2

Note: Numbers may not add up to 100% due to missing data.

ITT = intent to treat.a At least one dose of study medication.b At least one efficacy analysis post-baseline.

MCGOUGH ET AL.


was statistically significant for differences between quar-ter 1 versus quarter 2 scores (F1 = 6.53; p = .01), dem-onstrating that subjects who had been off treatmentbetween studies or who had received placebo in the ini-tial study, as well as those subjects who had receivedtreatment continuously, were all showing continued im-provement with long-term treatment. An analysis ofdose effect on the CGIS-P total score averages wasnot performed.Improvement was maintained in subjects who expe-

rienced uninterrupted treatment from the previousstudies. For subjects whose active treatment was discon-tinued during the transition from the previous studyand for subjects who received placebo in the previousstudy, mean CGIS-P total scores improved by .30%within the first 3 months of treatment. This improve-ment was maintained throughout the remainder of the24-month study and was statistically significant (p ,.001) at each quarter compared with baseline (Fig.1). Drug effectiveness according to ADHD subtypewas not evaluated because of the small number of sub-jects with the hyperactive-impulsive (n = 26, 5% of theITT population) or inattentive subtypes (n = 10, 2%).

Safety and Tolerability

Of the 568 subjects, 525 (92%) reported at least oneAE at some time throughout the study; most AEs werejudged to be mild (63%) or moderate (34%). The mostfrequently reported AEs included headache (15% of

AEs reported during 24 months), anorexia (15%), andinsomnia (11%).Of the subjects who reported AEs, 440 (84%) expe-

rienced at least one AE that was deemed by the inves-tigator to be ‘‘possibly’’ or ‘‘probably’’ related toMASXR.AEs suspected to be treatment related that occurred in‡5% of subjects at any time during the 24 months arepresented in Table 2.The incidence of treatment-related AEs increased

with increasing MAS XR dose: 25% in the 10-mggroup, 51% in the 20-mg group, and 65% in the30-mg group. Of 2,006 total drug-related (or possiblyrelated) AEs, 1,165 (58%) were reported during the firstquarter of the study, and the incidence of AEs dimin-ished over time (Table 2).Eighty-four subjects (15%) withdrew from the study

because of AEs. At time of withdrawal, 31 (37%) sub-jects were receivingMASXR 10mg/day, 40 (48%) werereceiving MAS XR 20 mg/day, and 13 (15%) were re-ceiving MAS XR 30 mg/day. The most frequently re-ported AEs associated with treatment withdrawalincluded weight loss (n = 27), anorexia/decreased appe-tite (n = 22), insomnia (n = 11), depression (n = 7), andemotional lability (n = 4). Of the subjects who withdrewfrom the study due to AEs, 38 (7%) subjects withdrewduring quarter 1, 27 (6%) withdrew during quarter 2, 6(2%) withdrew during quarter 3, 6 (2%) withdrew dur-ing quarter 4, 1 (,1%) withdrew during quarter 5, and2 (,1%) each withdrew during quarters 6, 7, and 8,with percentages based on total subjects remaining intreatment at the beginning of each quarter.Twenty-one SAEs were reported in 18 subjects (3%),

as summarized in Table 3. Only two SAEs, both con-vulsions, were judged to be possibly related to MAS XR.Both were rated mild in clinical severity. Subsequentevaluation indicated that each of these SAEs was an un-recognized preexisting condition. Both subjects werediscontinued from the study, but neither required sub-sequent treatment with anticonvulsant medication.Twelve SAEs were severe, but none of these were con-sidered possibly related to MAS XR.Changes in laboratory values throughout the 24

months were not clinically meaningful, and no trendswere apparent. Mean systolic blood pressure increasedby 3.5 mmHg, diastolic blood pressure by 2.6 mmHg,and pulse by 3.4 beats per minute. There was no rela-tionship between MAS XR dose and pulse or bloodpressure changes. Results from cardiovascular monitoring,

Fig. 1 Mean quarterly Conners Global Index Scale, Parent version (CGIS-P)

scores, stratified by treatment before enrollment in long-term study. Baseline

CGIS-P assessment was conducted at the enrollment visit, before study med-

ication was dispensed.



including vital sings, ECG changes, and cardiovascular-related AEs, will be presented in a subsequent report.

DISCUSSION

Long-term tolerability and effectiveness of MAS XRwere demonstrated in children with ADHD in this2-year, open-label extension of two previous placebo-controlled studies (Biederman et al., 2002; McCrackenet al., 2003). Of the 568 children enrolled in the study,273 (48%) completed 24 months of treatment. Thedose of MAS XR was initiated at 10 mg/day for thefirst week and was increased to optimal effectivenessin 10-mg increments to amaximum daily dose of 30mg.With regard to long-term ADHD treatment, effecti-

veness—the ability to productively participate in activ-ities of daily life in the real world—is arguably moreimportant than efficacy established in highly controlledclinical settings (Wolraich, 1999). In this study, effec-tiveness, as determined by quarterly CGIS-P scores, wasmaintained in the three analysis groups throughout thetreatment period. In subjects without treatment inter-ruption before entering this study, the significant reduc-tion in ADHD symptoms (.30%) achieved during theprevious analog classroom and natural setting studieswas maintained throughout the present study. For sub-jects with treatment interruption between studies andfor subjects who received placebo during the previousstudies, significant symptom improvement was achieved

by the end of the first quarter (.30% reduction inCGIS-P scores) and maintained throughout the study.Behavioral improvements likely occurred soon afterMAS XR treatment was initiated, but the CGIS-Pwas measured monthly and the statistical analysis wasbased on mean quarterly values.

In the absence of large, prospective, long-term stud-ies, the benefit-to-risk ratio of long-term stimulanttreatment for ADHD is of concern. The current studydemonstrated that long-term MAS XR treatment wasboth well tolerated and provided persistent, robust clin-ical benefits. Most AEs were mild. Evidence of a possibledose relationship for AEs was detected whether consid-ering all AEs or AEs that were judged to be related toMAS XR. A corresponding increase in the number ofsubjects withdrawing due AEs as a result of an increaseddose of MAS XR was not evident. The most commonlyreported drug-related AEs (e.g., anorexia, insomnia,headache, weight loss, nervousness, and abdominal pain)were reported more frequently with increasing dose.However, the incidence of AEs diminished quickly overtime, with 58% of all treatment-emergent AEs reportedduring first quarter dose titration. Because only 15% ofsubjects (84/568) withdrew because of AEs, these dataare consistent with previous literature reporting atten-uation of side effects over time with continued stimu-lant treatment (Gillberg et al., 1997). Most subjectswho were intolerant of study medication withdrew fromthe study during the first two quarters; 38 subjects

TABLE 2Treatment-Emergent Adverse Events Reported by ‡5% of Subjects Across All MAS XR Dosesa

Months 1–6(N = 568)

Months 7–12(n = 407)

Months 13–18(n = 330)

Months 18–24+(n = 284)

COSTART TermbSubjects

Reporting (%)No. ofReports

SubjectsReporting (%)

No. ofReports


No. ofReports


No. ofReports

Anorexia 37.1 262 7.6 32 3 11 3.5 10Headache 27.1 256 17.2 113 17.6 82 17.6 80Insomnia 25.7 181 3.9 18 5.8 21 4.2 12

Abdominal pain 18.1 155 4.9 27 7.6 28 7.0 22Nervousness 16.7 122 5.2 27 4.2 16 2.8 8Weight loss 16.9 101 4.2 18 2.4 9 1.4 6Emotional lability 14.1 101 3.4 14 2.7 9 1.8 5

Depression 4.9 33 3.2 14 0.9 3 1.8 5Twitching 4.6 40 2.2 12 1.8 6 0.7 2Thinking abnormal 4.4 31 3.4 17 0 0 0.4 1

Note: COSTART = Coding Symbols for Thesaurus of Adverse Reaction Terms; anorexia = decrease or loss of appetite.a Number of reports indicate each visit in which an adverse event was reported for a particular patient.b The terms listed in this table are those that were considered drug related or possibly related by the investigators for ‡5% of subjects.

MCGOUGH ET AL.


(constituting 45% of subjects who withdrew due toAEs) withdrew during the first quarter, and 27(32%) withdrew during the second quarter.The impact of stimulant medications on growth in

children with ADHD has been an active area of researchand clinical debate. Spencer et al. (1996) detected smallbut statistically significant differences in height in youthswith ADHD compared with similar subjects withoutthe disorder and suggested that height deficits observedin children with ADHDmight represent temporary de-velopmental deviations associated with the disorder it-self. Biederman et al. (2003a) reported no differences inage-adjusted height or weight in girls with and withoutADHD and observed no detrimental impact of psy-chostimulant medications on growth in a medicated co-hort. Biederman et al. (2003b) presented a preliminary

analysis of growth data collected during this long-termstudy using the normative growth charts provided bythe Centers for Disease Control and Prevention (CDC)(2000). Reassuringly, the decrease in expected weight gainwas largest for the heaviest subjects (–7.8 kg for subjectswho were above the 75th percentile on the CDC weightcharts at baseline) and smallest for the lightest subjects(–2.1 kg for subjects who were below the 25th percentileat baseline). Concordant results were also evident for ap-parent treatment-related effects on height, as the heightdeficit was negligible for the shortest patients at end point(a loss of one percentile on the CDC height charts, or areduction in growth rate equivalent to –0.8 cm).For all growth measures, including weight, height,

and body mass index (kg/m2), the reduction in expectedgrowth was not significant in the second year of thestudy, suggesting the effects of MAS XR on growth werenot persistent. In addition, the deficits in growth rateswere smallest for children previously treated with stim-ulant therapy. This finding further supports the hypoth-esis that stimulant-associated growth rate differencesattenuate over time, consistent with findings in otherstudies. A similar z score analysis of growth data froma long-term study of OROS methylphenidate in 237children (aged 6–13) also showed little observed growthdifference during the second year of stimulant treatment(Spencer, 2003). Poulton and Cowell (2003) reportednormalization of height velocity during long-term dex-troamphetamine treatment. A detailed analysis of growthdata collected during long-term treatment withMAS XR,including stratification by dose and time on medication,will be reported separately.The current study represents the largest number of

school-age children receiving an extended-release am-phetamine formulation for a prolonged period of time.This study is a significant addition to the extant litera-ture of longitudinal stimulant studies in children withADHD in general and amphetamine studies in partic-ular. Results of this study are consistent with the resultsof an 18-month, double-blind, placebo-controlled studythat demonstrated long-term amphetamine treatmentto be significantly more effective than placebo in reduc-ing ADHD-related behaviors and improving learning inchildren with severe ADHD (Gillberg et al., 1997). Inthat study, the significant improvements of amphet-amine treatment were maintained throughout the 15-month, double-blind treatment period, and AEs weremild and infrequent.

TABLE 3Serious Adverse Events Reported During Long-Term

MAS XR Treatment

SubjectMAS XRDose (mg)

COSTARTPreferred Term

Relatedto MAS XR Severity

1 10 Convulsion Possibly Mild2 Hostility No Mild3 Hostility No Severe

4 20 Convulsion Possibly Mild5 Accidental injury No Mild6 Intestinal perforation No Mild

7 Asthma No Moderate7 Gastrointestinal

disorderNo Severe

8 Small intestineperforation

No Severe

9 Acute myeloblastic

leukemia

No Severe

10 Gastrointestinaldisorder

No Severe

11 30 Hostility No Moderate

11 Depression No Moderate12 Depression No Severe13 Personality disorder No Severe

14 Accidental injury No Severe14 Gastrointestinal

disorderNo Severe

15 Lymphadenopathy No Severe16 Cellulitis No Severe17 Gastrointestinal

disorderNo Severe

18 Gastrointestinaldisorder

No Severe

Note: MAS XR = extended-release mixed amphetamine salts; CO-

START =Coding Symbols for Thesaurus of Adverse Reaction Terms.



Increased recognition of the genetic basis ofADHD (Faraone and Doyle, 2000; Smalley, 1997)and follow-up studies demonstrating a diagnostic con-tinuity of ADHD from childhood to adulthood supportthe need for long-term treatment (Barkley et al., 1990,2002; Biederman et al., 1998; Weiss et al., 1985). Lon-gitudinal studies of ADHD treatment with other stim-ulant medications, such as methylphenidate, also haveshown persistent therapeutic effects and low incidencesof adverse effects (MTA Cooperative Group, 1999;Schachar et al., 1997; Wilens et al., 2003). A 12-monthinterim report of an extension study of a once-daily for-mulation of methylphenidate (OROS methylpheni-date) noted a similar outcome of persistent clinicalbenefit. The MTA study demonstrated sustained treat-ment response using medication management alone forup to 14 months (multiple daily doses of methylpheni-date) and highlighted the need for ongoing manage-ment of stimulant dose to optimize treatmentoutcomes (Vitiello et al., 2001).

Limitations

This study was limited by its open-label and uncon-trolled design. Ethical considerations preclude inclusionof a placebo control arm in a long-term study because itmight create unnecessary risks for participants from lackof treatment. In the study by Gillberg et al. (1997),which was an 18-month, double-blind, placebo-con-trolled study of amphetamine treatment in childrenwith ADHD, only 27% of the subjects randomizedto receive placebo completed the 15-month period ofdouble-blind treatment. Without controlling for a pla-cebo response in the current study, it is possible thatresults were subject to observer bias, in this case, the par-ent or caregiver completing the CGIS-P. However,mean CGIS-P scores appear to be stable for the 273 sub-jects who completed 24months of treatment.Without aplacebo comparison group, this study also failed to con-sider the reduction in symptom expression over timethat is associated with the natural course of the disorder.Furthermore, we did not assess any potential benefitsthat might have accrued due to concomitant psychoso-cial interventions. Finally, effectiveness measures werebased on parental report only and did not directly assessimprovement in school and other nonhome settings.Eliciting AEs was based on spontaneous subject

report and not structured interviews or rating scales.This approach might underestimate the incidence of

medication side effects. Sponsors of clinical trials shouldbe encouraged to employ structured assessments of AEsin future research. Furthermore, although only 15% ofsubjects dropped out because of side effects, it is unclearhow many of the 37% who reportedly discontinued forother reasonsmight have terminated due to problems tol-erating medication. Although attenuation of AEs withprolonged stimulant exposure is consistent with previousreports, a proportion of this decrease might be due to thedrop out by subjects who did not tolerate the medication.

Given the length of this study, loss of data is unavoid-able because children or caregivers may choose to with-draw from the study for various reasons. A 12-monthinterim report of a long-term study of an extended-release methylphenidate formulation reported that71% of subjects (289/409) completed 12 months oftreatment (Wilens et al., 2003). The number of subjectscompleting at least 12 months of treatment with MASXR (330/568, 58%) and 24 months of treatment(273/568, 48%) is impressive, given the requirementfor 28 study visits and monthly assessments.

Several factors must be considered before generaliz-ing the results of this study to other ADHD popula-tions. Although minority participation in this studyexceeded that reported for many previous ADHD stim-ulant studies, most subjects were white boys withADHD combined type. Most subjects receiving activemedication in the antecedent studies were responsive toMAS XR and did not experience intolerable side effects.Notably, eligibility criteria excluded subjects with co-morbid medical and severe or unstable psychiatric dis-orders. In addition, because 62% of subjects in thepresent study received stimulant treatment (either am-phetamine ormethylphenidate) before entry into the pre-ceding studies, the incidence of AEs reported here mightnot necessarily reflect expected rates in treatment-naivepatients. Finally, data on duration and frequency of drugholidays were not quantified, precluding our ability toassess time off medication as a response moderator.

Clinical Implications

Once-daily doses of 10–30 mg of MAS XR were per-sistently effective in reducing ADHD symptoms and werewell tolerated over a 2-year treatment period in childrenwithADHD. Increasing evidenceof sustained effectivenesswith ADHD medications raises the potential for measur-able improvements in patients’ long-term outcome andadaptive functioning. The long-term effectiveness and

MCGOUGH ET AL.


tolerability of MAS XR demonstrated in this study meritsfurther prospective study of MAS XR in other ADHDpopulations, including adults, adolescents, and patientswith disorders that are commonly comorbid with ADHD,such as oppositional defiant disorder.

Disclosure: Dr. McGough has received support from Eli Lilly (grant/re-search support, speaker’s bureau), Novartis (grant/research support, con-sulting fees, speaker’s bureau), Shire (grant/research support, consultingfees, speaker’s bureau), Pfizer (grant/research support, consulting fees),and New River Pharmaceuticals (grant/research report). Dr. Biedermanhas received support from Shire (grant/research support, consulting fees,speaker’s bureau), Eli Lilly (grant/research support, consulting fees,speaker’s bureau), Pfizer (grant/research support, consulting fees), Ceph-alon (grant/research support, consulting fees, speaker’s bureau), McNeil(consulting fees, speaker’s bureau), Janssen (grant/research support, con-sulting fees), Neurosearch Pharmaceuticals (grant/research support), Stan-leyMedical Institute (grant/research support), Prechter Foundation (grant/research support), Novartis (consulting fees, speaker’s bureau), Johnson &Johnson (consulting fees), New River Pharmaceuticals (grant/research sup-port, consulting fees), and the Lilly Foundation (grant/research support).Dr. Wigal has received support from Eli Lilly (grant/research support),Novartis (grant/research support, speaker’s bureau), Shire (grant/researchsupport, consulting fees, speaker’s bureau), Celgene (grant/research support,speaker’s bureau) Celltech (grant/research support, consulting fees, speaker’sbureau), McNeil (grant/research support, consulting fees, speaker’s bu-reau), Cephalon (grant/research support), and Janssen (grant/research sup-port). Dr. Lopez has received support from Eli Lilly (grant/researchsupport), Novartis (grant/research support, consulting fees, speaker’s bu-reau), Shire (grant/research support, consulting fees, speaker’s bureau),and Novartis (grant/research support, consulting fees, speaker’s bureau),Celltech (consulting fees), and Cephalon (grant/research support). Dr.McCracken has received support from Pfizer (grant/research support, con-sulting fees), Janssen (grant/research support, consulting fees), Solvay(grant/research support), Eli Lilly (grant/research support, speaker’s bu-reau), Shire (grant/research support, consulting fees, speaker’s bureau),Bristol-Myers Squibb (grant/research support), BMS (consulting fees), No-ven (consulting fees). Dr. Spencer has received support from Abbott Labs(grant/research support), Shire (grant/research support, consulting fees,speaker’s bureau), Bristol-Myers Squibb (grant/research support) Cepha-lon (grant/research support), Eli Lilly (grant/research support, consultingfees, speaker’s bureau), GlaxoSmithKline (grant/research support, consul-ting fees, speaker’s bureau), McNeil (grant/research support, consultingfees, speaker’s bureau), Janssen (grant/research support, consulting fees,speaker’s bureau), Celltech (grant/research support, speaker’s bureau)Wyeth-Ayerst (grant/research support, speaker’s bureau), Novartis(grant/research support, consulting fees, speaker’s bureau), and Pfizer(grant/research support, consulting fees). Drs. Zhang and Tulloch arefull-time employees of Shire Pharmaceutical Development, Inc.

REFERENCES

American Psychiatric Association (1994), Diagnostic and Statistical Manualof Mental Disorders, 4th edition (DSM-IV). Washington, DC: AmericanPsychiatric Association

Barkley RA, Fischer M, Edelbrock CS, Smallish L (1990), The adolescentoutcome of hyperactive children diagnosed by research criteria: I. An8-year prospective follow-up study. J Am Acad Child Adolesc Psychiatry29:546–557

Barkley RA, Fischer M, Smallish L, Fletcher K (2002), The persistence ofattention-deficit/hyperactivity disorder into young adulthood as a func-tion of reporting source and definition of disorder. J Abnorm Psychol 111:279–289

Biederman J, Faraone SV, Taylor A, SiennaM,Williamson S, Fine C (1998),Diagnostic continuity between child and adolescent ADHD: findingsfrom a longitudinal clinical sample. J Am Acad Child Adolesc Psychiatry37:305–313

Biederman J, Lopez FA, Boellner SW, Chandler MC (2002), A randomized,double-blind, placebo-controlled, parallel-group study of SLI381 (Add-erall XR) in children with attention-deficit/hyperactivity disorder. Pedi-atrics 110:258–266

Biederman J, Faraone SV, Monuteaux MC, Plunkett EA, Gifford J, SpencerT (2003a), Growth deficits and attention-deficit/hyperactivity disorderrevisited: impact of gender, development, and treatment. Pediatrics 111:1010–1016

Biederman J, Faraone SV, Spencer T (2003b), Extended release mixed am-phetamine salts in ADHD: growth parameter analysis. Poster NR279presented at American Psychiatric Association 156th Annual Meeting;San Francisco, May 19, 2003

Bradley C (1937), The behavior of children receiving benzedrine. Am J Psy-chiatry 94:577–585

Centers for Disease Control and Prevention (2000), Growth charts: UnitedStates; http://www.cdc.gov/growthcharts/ (accessed July 15, 2003)

Claxton AJ, Cramer J, Pierce C (2001), A systematic review of the associ-ations between dose regimens and medication compliance. Clin Ther23:1296–1310

Conners CK (1997), Conners Rating Scale–Revised: Technical Manual. NorthTonawanda, NY: MultiHealth Systems

Faraone SV, Doyle AE (2000), Genetic influences on attention deficit hy-peractivity disorder. Curr Psychiatry Rep 2:143–146

Gillberg C, Melander H, von Knorring A et al. (1997), Long-term stimulanttreatment in children with attention-deficit hyperactivity disorder symp-toms: a randomized, double-blind, placebo-controlled trial. Arch GenPsychiatry 54:857–864

Greenhill LL, Haplerin JM, Abikoff H (1999), Stimulant medications. J AmAcad Child Adolesc Psychiatry 38:503–512

Hack S, ChowB (2001), Pediatric psychotropicmedication compliance: a lit-erature review and research-based suggestions for improving treatmentcompliance. J Child Adolesc Psychopharmacol 11:59–67

McCracken JT, Biederman J, Greenhill LL et al. (2003), Analog classroomassessment of a once-daily mixed amphetamine formulation, SLI381(Adderall XR), in children with ADHD. J Am Acad Child Adolesc Psy-chiatry 42:673–683

MTA Cooperative Group (1999), A 14-month randomized clinical trial oftreatment strategies for attention-deficit/hyperactivity disorder. Arch GenPsychiatry 56:1073–1086

Poulton A, Cowell CT (2003), Slowing of growth in height and weight onstimulants: a characteristic pattern. J Paediatr Child Health 39:180–185

Schachar RJ, Tannock R, Cunningham C, Corkum PV (1997), Behavioral,situational, and temporal effects of treatment of ADHDwith methylphe-nidate. J Am Acad Child Adolesc Psychiatry 36:754–763

Shaffer D, Fisher P, Lucas CP, Dulcan M, Schwab-Stone ME (2000),NIMHDiagnostic Interview Schedule for Children, Version IV (NIMHDISC-IV): descriptive differences from previous versions and reliabilityof some common diagnoses. J Am Acad Child Adolesc Psychiatry 39:28–38

Smalley SL (1997), Genetic influences in childhood-onset psychiatric disor-ders: autism and attention-deficit/hyperactivity disorder. Am J HumGenet 60:1276–1282

Spencer TJ (2003), Long-term once-daily OROS� methylphenidate treat-ment for ADHD: evaluating effect on growth. Poster NR715 presentedat: American Psychiatric Association 156th Annual Meeting; San Fran-cisco, CA: May 21, 2003

Spencer TJ, Biederman J, Harding M, O’Donnell D, Faraone SV, Wilens T(1996), Growth deficits in ADHD children revisited. Evidence for dis-order-associated growth delays? J Am Acad Child Adolesc Psychiatry35:1460–1469



Swanson J (2003), Compliance with stimulants for attention-deficit/hyper-activity disorder. Issues and approaches for improvement. CNS Drugs17:117–131

Vitiello B, Severe JB, Greenhill LL et al. (2001), Methylphenidate dosage forchildren with ADHD over time under controlled conditions: lessonsfrom the MTA. J Am Acad Child Adolesc Psychiatry 40:188–196

Weiss G, Hechtman L, Milroy T et al. (1985), Psychiatric status of hyper-actives as adults: a controlled prospective 15-year follow-up of 63 hyper-active children.24:211–220

Wilens TE, Biederman J, Spencer TJ (2002), Attention deficit-hyperactivitydisorder across the lifespan. Annu Rev Med 53:113–131

Wilens T, Pelham W, Stein M et al. (2003), ADHD treatment with once-daily OROS methylphenidate: interim 12-month results from a long-term, open-label study. J Am Acad Child Adolesc Psychiatry 42:424–433

Wolraich ML, Hannah JN, Pinnock TY, Baumgaertel A, Brown J (1996),Comparison of diagnostic criteria for attention-deficit hyperactivity dis-order in a county-wide sample. J Am Acad Child Adolesc Psychiatry35:319–324

Wolraich ML (1999), The difference between efficacy and effectiveness re-search in studying attention-deficit/hyperactivity disorder. Arch PediatrAdolesc Med 153:1220–1221

MCGOUGH ET AL.


long-term tolerability and effectiveness of once-daily mixed amphetamine salts (adderall xr) in...

Documents