`Amphetamine Formulations in ADHD
`
`REGINA S. JAMES, M.D., WENDY S. SHARP, M.S.W., THERESA M. BASTAIN, A.B., PATTI P. LEE, M.A.,
`JAMES M. WALTER, M.A., MARK CZARNOLEWSKI, PH.D., AND F. XAVIER CASfELLANOS, M.D.
`
`ABSTRACT
`Objective: To compare the efficacy and time course of single morning doses of Adderall~, extended-release, and imme(cid:173)
`diate-release dextroamphetamine sulfate. Method: Thirty-five children with attention-deficit/hyperactivity disorder, com(cid:173)
`bined type, were given Adderall, immediate-release dextroamphetamine, dextroamphetamine SpansulesG, and placebo
`In a randomized, double-blind, crossover study. Behavior ratings, locomotor activity measurements, and academic mea(cid:173)
`sures were obtained over a period of 8 weeks. Results: All three drugs exhibited robust efficacy versus placebo on nearly
`all measures. The effects of dextroamphetamine Spansules were less robust in the morning, particularly compared with
`Adderall, but they lasted 3 to 6 hours longer, depending on the measure. Although parent behavior ratings and locomotor
`activity showed improvements up to 12 hours after single doses of all three drugs, the number of math problems
`attempted and completed correctly 4 hours after dosing were only robustly increased by Spansules. Concluslons: Both
`immediate-release amphetamines demonstrated earlier onset of effects, but dextroamphetamine Spansules showed
`more sustained effects that were present on a wider range of measures. J. Am. Acad. Child Ado/esc. Psychiatry, 2001,
`40( 11 ): 1268-1276. Key Words: attention-deficit/hyperactivity disorder, dextroamphetamine, Adderall®, psychostimulants.
`
`Stimulants are the drugs of choice for the pharmacological
`treatment of attention-deficit/hyperactivity disorder
`(ADHD), and the most frequently prescribed agent
`remains methylphenidate (MPH), with a positive response
`rate exceeding 70% (Spencer er al., 1996). Several con(cid:173)
`trolled crossover studies suggest chat nearly all children
`with combined-type ADHD who are nonresponders to
`MPH respond favorably to dextroamphetamine sulfate
`and vice versa (Arnold, 1996; Arnold et al., 1978; Elia
`et al., 1991; Sharp et al., 1999). In the past few years, a
`mixture of 75% dextroampheramine and 25% levoam(cid:173)
`phetamine (Popper, 1994) has been aggressively marketed
`under the trade name Adderall®, attaining an estimated
`29% of market share in 2000 (Goodman and Nachman,
`2000). Initial marketing efforts emphasized the existence
`
`Accrptrdjunr 5, 2QOJ.
`From 1hr Child Psychiatry Bra11ch, Natio11al /nstittttr of Mmta/ Hralth,
`Bethesda, MD. Prrsmtrd at 1l1r 47th a11n1111/ mrrting of thr AACAP, Nrw York,
`Ortobrr 2000.
`Thr authors apprniatr thr assistance of NIH Day Program trachrrs Anna
`Davidso11, M.Ed., and Swan job, M.Ed.: rrcrration thrmpist Dtbbir C Marrus.
`C T.R.S.; NIH Clinical Cmtrr staff mrmbrrs Edwina Smith. R.N., and con(cid:173)
`sulting psychologist Barbara B. Kr/In; Ph.D.
`Rtprint rrqursts to Dr. Castrllanos, NYU Child Study Crntrr. 577 First
`Avrnur, Nnu York, NY 10016.
`0890-8567/0I/4011-1268©2001 by the American Academy of Child and
`Adolc:scenr Psychiatry.
`
`of four distinct sales and suggested that they provided dif(cid:173)
`ferential rates of absorption and thus longer efficacy
`(Richwood Pharmaceutical Company, unpublished pre(cid:173)
`scribing information on Adderall tablets, 1997). Recent
`manufacturer-sponsored trials have compared Adderall
`and immediate-release MPH (Pelham et al., 1999; Pliszka
`et al., 2000; Swanson er al., 1998), but there have been no
`comparisons between Adderall and either immediate(cid:173)
`release or extended-release formulations of dextroamphet(cid:173)
`amine sulfate. On the basis of similar terminal half-lives
`for dextro- and levoamphetamine (Hutchaleelaha et al.,
`1994), we hypothesized that Adderall would be compara(cid:173)
`ble in efficacy and rime course with immediate-release dex(cid:173)
`troamphetamine, that Adderall and immediate-release
`dextroampheramine would be more effective than the
`same dose of extended-release dextroampheramine in the
`morning, but that the effects of extended-release dex(cid:173)
`troamphetamine would last longer than those of Adderall
`and immediate-release dextroampheramine.
`
`METHOD
`
`Subjects
`We examined 21 boys and 14 girls (mean age 9.1 ± 1.5, range
`6.9-12.2 years) with a history of severe hyperaaivicy, impulsivicy, and
`
`1268
`
`J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 40:11, NOVEMBER 2001
`
`KVK-TECH EXHIBIT 1053
`
`
`
`inattention who met DSM-IV criteria for combined-type ADHD. In
`addition to the diagnosis of ADHD, 10 children also met criteria for
`oppositional defiant disorder, 12 for an anxiety disorder, 3 for enuresis,
`2 for dysthymic disorder, and 6 for a learning disorder. Children were
`recruited from local schools. Exclusion criteria included Full Scale IQ
`less than 80 on the WISC-III (Wechsler, 1991) or the presence of a
`chronic medical or neurological disease including Toureue's disorder,
`chronic tic disorder, pervasive devdopmental disorders, and mood or
`anxiety disorders requiring current treatment. The sample consisted of
`18 whites, 9 African Americans, 7 Latinos, and 1 Asian American.
`Fifteen subjects were naive to stimulant treatment prior to participa(cid:173)
`tion in the study (fable 1).
`
`Assessment
`Systematic telephone screenin~ were conducted by a research social
`worker to determine that ADHD symptoms were present in at least
`two settings. Conners Parent and Teacher Rating Scales were then
`obtained (CoMers, 1997). Additional tools used in the screening pro(cid:173)
`cess included the Child Behavior Checklist and Teacher's Report Form
`(Achenbach and Ruffie, 2000). Subjects who met smdy diagnostic
`criteria and who had a Conners Teacher Rating Scale Hyperacrivity
`faaor score of 70 or greater were sent copies of consent and assent by
`mail and invited to visit the program. Written informed consent and
`assent were obtained during this initial face-to-face interview. All psy-
`
`TABLE 1
`Demographic Characteristics
`
`Age (years)
`WISC-III
`Verbal standard score
`Performance standard score
`Full Scale standard score
`CBCL Attention Problems T score
`TRF Attention Problems T score
`
`Sex
`Male
`Female
`DSM-IV diagnoses
`ADHD, combined type
`ODD
`Anxiety disorder
`Enuresis
`Dysthymia
`Learning disorder
`Prior stimulant treatment
`Mechylphenidate
`Methylphenidate SR
`Adderall
`Immediate-release dextroamphecamine
`Dextroamphetamine Spansules
`
`Mean
`
`9.1
`
`102.5
`96.6
`99.8
`72.5
`72.3
`
`n
`
`21
`14
`
`35
`10
`12
`3
`2
`6
`20
`11
`3"
`3
`2
`2
`
`SD
`
`1.5
`
`13.6
`14.5
`13.0
`10.2
`10.8
`
`o/o
`
`60
`40
`
`100
`29
`34
`9
`6
`17
`57
`31
`9
`9
`6
`6
`
`Nott: CBCL = Child Behavior Checklist; TRF = Teacher's Report
`Form; ADHD = attention-deficit/hyperactivity disorder; ODD =
`oppositional defiant disorder; SR = sustained release.
`" Includes one subject taking both methylphenidate and methyl(cid:173)
`phenidate SR.
`
`ADDERALL AND DEXTROAMPHETAMINES IN ADHD
`
`chotropic medications were discontinued prior to beginning the study,
`which was approved by our institutional review board.
`
`Procedure
`Children participated in a research school 5 days per week consisting
`of formal academic instruction from 9 A.M. to 12:30 P.M. and therapeu(cid:173)
`tic ~reation (sports, art therapy, structured social skills sessions) from 1
`P.M. to 3 P.M. Behavior management techniques were used extensively
`within the program, although parent training was not provided.
`·
`During a 3-week medication-free observation period, a physical
`and neurological examination was performed and routine laboratory
`studies were obtained (complete blood cell count, electrolytes, uri(cid:173)
`nalysis, liver function test, thyroid studies). A psychoeducational
`evaluation, consisting of the WISC-Ill and the Woodcock-Johnson
`Acl1ievemem Battery-Revised (Woodcock and Johnson, 1989), was
`performed by a clinical psychologist. A child and adolescent psychia(cid:173)
`trist confirmed the final DSM-IV diagnoses by combining infor(cid:173)
`mation from the structured psychiatric interview (Diagnostic
`Interview for Children and Adolescents-Child and Parent versions,
`revised) (Reich, 2000), teacher and recreation therapist racin~ of
`hyperactive behavior, parent mtings, and staff observations.
`Double-blind medications were administered for 8 weeks, followed
`by 2 weeks of open treatment optimization. Each child received 2
`weeks each of Adderall, immediate-release dexrroampheramine, dex(cid:173)
`troamphetamine Spansules, and placebo in random order. Active drugs
`were given in two doses, one per week. Doses, sdected before random(cid:173)
`ization, were based on age, weight, prior medication experience, and
`symptom severity. The overall mean low dose was 7.8 mg (range 5-25
`mg, 0.24 mg/kg) and the mean high dose was 12.8 mg (range 10-30
`mg, 0.39 mg/kg). The dose order was randomized across subjects, but
`the same order, eid1er increasing (n = 18) or decreasing (n = 17), was
`used for a given subject. Given the absence of comparative data, we
`administered equal doses of all three drugs to the first 24 subjects. The
`last 11 subjects received equal doses of both immediate-release formu(cid:173)
`lations, but we increased the dextroamphetamine Spansules doses by 5
`mg to more closely approximate clinical use patterns (fable 2).
`Medications and placebo were contained in identical capsules pack(cid:173)
`aged in dated, coded blister packs by a research pharmacy service and
`were administered by parents at home at a mean time of 7:16 A.M.
`(SD= 21 minutes), on weekdays and at breakfust on weekends. Subjects
`who had not previously received stimulants were first treated with
`immediate-release dextroamphetamine (2.5-5 mg daily for 3 days)
`
`TABLE2
`Dose Schedule
`
`Dosage Level (mg)
`
`First 24 subjects
`Adderall
`Dextroamphetamine
`Dextroamphetamine Spansules
`Last 11 subjects
`Adderall
`Dextroamphetamine
`Dextroamphetamine Spansules
`
`30
`30
`30
`
`5
`5
`5
`
`5
`5
`10
`
`10
`10
`10
`
`15
`15
`15
`
`10
`15
`15
`10
`15 20
`
`20
`20
`20
`
`20
`20
`25
`
`25
`25
`25
`
`25
`25
`
`Note: Two doses were selected for each patient from the listed
`options. The order of the doses (increasing or decreasing) was
`randomly selected and applied to all active drugs for each subjea.
`Drug sequence, including placebo, was assigned randomly.
`
`J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 40:11, NOVEMBER 2001
`
`1269
`
`
`
`JAMES ET AL.
`
`prior to beginning double-blind trials to rule out idiosyncrnric responses
`and minimize adverse effects from starting on a relatively high dose.
`
`Dependent Measures
`Teachers provided weekly ratings of classroom behavior (between 9
`A.M. and noon) using the Hyperactive/Impulsive factor of the
`Conners Teacher Rating Scale (Conners, 1997). The recreation thera(cid:173)
`pist provided weekly scores on the Hyperactivity factor of the
`Children's Psychiatric Racing Scale (Fish, 1985). Academic measures
`were also obtained each day at 11 A.M., about 4 hours after medication
`administration. Students performed a 5-minute timed math task,
`consisting of arithmetic problems with difficulty levels selected indi(cid:173)
`vidually for each child. The number of correct and total responses was
`recorded daily and summed weekly. The nurse coordinaror recorded
`weekly weights and vital signs and assessed medication side effects
`using the Stimulant Side Effect Rating Scale (Barkley et al., 1990).
`Parents recorded time of administration of coded medication and
`provided a weekly rating of adverse effects (Barkley Side Effect Rating
`Scale). Parents of the 28 most recently enrolled subjects were asked to
`provide ratings of hyperactive/impulsive behavior (Conners Parent
`Behavior Racing Scale) for the hours 4 P.M. to 7 P.M. Weekly telephone
`contact and medication slips signed daily by parents confirmed
`compliance. Moror activity was assessed with an Actometer worn on
`the nondominant wrist (Acriwatch, Sun River, OR).
`Adderall and immediate-release dextroamphetamine were obtained
`commercially. When the study began, Smith-Kline Beecham was in the
`process of modifying the manufacturing process for extended-release
`dexcroamphetamine capsules. Pending Food and Drug Administration
`approval (obtained in February 1999) for the new aqueous process,
`Smith-Kline Beecham provided dextroamphetamine Spansules at no
`cost and without restrictions. No funding was sought or received from
`any commercial entity.
`
`Statistical Analysis
`Paramerric analyses used SPSS 10.0 for Windows (SPSS, 1999). A
`mixed repeated-measures analysis of variance (ANOVA) was used to
`initially examine the two between facrors (dose order (increasing versus
`decreasing), dosage equivalence [same dosage versus 5 mg higher dose
`for Spansules]), and the within factors of drug (4 levels) and dose (2
`levels), and interactions. Nonsignificant between-subject facrors were
`removed from subsequent analyses. Significant ANOVA results were
`explored with a priori defined painvise contrast analyses, corrected for
`multiple comparisons. All tests were two-tailed, with nominal overall
`significance set at p < .05 for each measure.
`Daytime Actometer data were divided into 12 hourly periods from 9
`A.M. (beginning of school) to 9 r.M. (earliesr bedtime), averaged over
`weekdays (Monday-Friday) and log-transformed (In [x + l]). Becmse
`of substantial missing data with this measure, primarily from Actometer
`breakage, complete data are available for only 22 subjects. The 22 sub(cid:173)
`jects with complete data did not differ from the remaining subjects on
`age or any severity measures (p > .40). Post hoc drug-drug and drug(cid:173)
`placebo hourly contrasts were corrected for 72 multiple comparisons
`with significance set at p < .0007.
`The Aaiwatch Sleepwatch software (Mini Miner Company, 1999)
`calculates total presumed sleep duration (Sadeh et al., 1994, 1995),
`which was averaged for the week nights of each study week.
`Missing weekly weights (6%) were calculated by interpolation.
`Children with other missing data for a measure were excluded from
`analyses for that measure. Complete parent data were available for 20
`or fewer subjects, depending on the measure. There were no demo(cid:173)
`graphic or severity differences between the subjects with and without
`completed parent reports.
`
`RESULTS
`
`We enrolled 38 children, 3 of whom were excluded
`prior to randomization because of a history of chronic
`motor and vocal tics, IQ< 80, and abnormal EEG find(cid:173)
`ings, respectively. The remaining 35 subjects completed
`the double-blind trial. All subjects met DSM-JV criteria for
`ADHD, combined type. See Table 1 for additional subject
`characteristics. Medication compliance, as documented by
`parent records noting date and time of medication admin(cid:173)
`istration, exceeded 93% (SD === 11.8%). Documented
`compliance was low for two subjects (46% and 57%).
`Although their data are included in all analyses, we verified
`that they did not represent statistical outliers (Table 3).
`
`Between-Group Factors
`
`Dose order (increasing versus decreasing) was randomly
`determined for each subject, and the same sequence was
`used for all three active medications. Significant three-way
`interactions (dose order by drug by dose) were found for
`weight and recreation therapist ratings of hyperactivity/
`impulsivity, as noted below. Dose order was not a signifi(cid:173)
`cant factor with any other main effects or interactions for
`any other measure.
`Similarly, we compared the first 24 subjects, all of
`whom received the same dosages for all three drugs, with
`the last 11 subjects, who received Spansule doses that were
`5 mg higher than the immediate-release formulations.
`Only parent-reported adverse effects showed a significant
`difference between groups (Fi.17 :::: 9.15, p :::: .008), with a
`greater number of adverse effects reported in the subgroup
`with complete data that received a higher Spansule dose
`(n === 4). No other significant main effects or interactions
`were observed for any measures. Accordingly, analyses for
`all other measures were collapsed across the two groups
`(maximum n = 35).
`
`Hyperactivity Ratings
`
`There was a significant effect of medication on the
`Conners Teacher Hyperactivity factor score (~32 :::: 15.70,
`p < .001), obtained in the classroom between 9 A.M. and
`12:30 P.M. Contrast analysis revealed that immediate(cid:173)
`release dextroamphecamine, which did not differ signifi(cid:173)
`cantly from Adderall, decreased teacher-rated hyperactivity
`significantly more than dextroamphetamine Spansules (p ===
`.025). Higher doses were significantly more effective than
`lower doses for all three medications (F134 === 5.38, p = .03).
`In afternoon recreation therapy ( 1 P.M. to 3 P.M.),
`ANOVA revealed a significant main effect of drug on
`
`1270
`
`J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 40:11..NOVEMBER 2001
`
`
`
`ADDERALL AND DEXTROAMPHETAMINES IN ADHD
`
`Measure
`
`Adderall
`
`TABLE3
`Means, Standard Deviations, and Analyses of Variance
`Mean (SD)
`Dex-Span
`1-R Dex
`
`Placebo
`
`CTS
`CPRS
`CPS
`Weight (kg)
`Sleep (hr)
`TAP
`TCP
`SERS-N#
`SERS-N sev
`SERS-P#
`SERS-P sev
`
`50.6
`2.8
`58.3
`32.6
`7.6
`171.6
`164.6
`3.3
`2.7
`6.3
`3.2
`
`(5.6)
`(1.0)
`(13.1)
`(8.0)
`(0.7)
`(56.4)
`(55.9)
`(2.0)
`(1.5)
`(2.7)
`(1.2)
`
`53.7
`2.3
`60.0
`32.5
`7.2
`187.0
`177.6
`2.9
`3.1
`6.7
`3.7
`
`(9.1)
`(1.0)
`( 15.6)
`(7.9)
`(0.5)
`(60.9)
`(61.1)
`(1.8)
`(2.0)
`(2.9)
`(I .5)
`
`50.5
`2.5
`60.5
`32.7
`7.4
`177.4
`167.6
`2.6
`2.7
`6.4
`3.2
`
`(5.4)
`(1.1)
`(14.7)
`(8.0)
`(0.7)
`(42.9)
`(41.2)
`(1.8)
`(1.7)
`(3.5)
`(1.6)
`
`63.1
`3.8
`68.0
`33.3
`7.8
`147.7
`140.2
`2.0
`1.8
`5.9
`2.8
`
`(12.6)
`(1.1)
`(14.5)
`(8.3)
`(0.6)
`(50.7)
`(51.3)
`(1.9)
`(1.2)
`(3.2)
`(1.5)
`
`ANOVA
`df
`
`3,32
`3,29
`3,12
`3,32
`3,19
`3,32
`3,32
`3,23
`3,23
`3,15
`3,15
`
`F
`
`15.7
`35.0
`5.8
`13.4
`9.9
`6.3
`5.6
`3.9
`3.6
`0.3
`2.2
`
`p
`
`.000
`.000
`.01
`.000
`.000
`.002
`.003
`.02
`.03
`.82
`.13
`
`Note: ANOVA = analysis of variance; Dex-Span "'dextroamphetamine Spansules; 1-R Dex= immediate-release dextroam(cid:173)
`phetamine; CTS "' Conners Teacher Rating Scale Hyperaaivity T score obtained from 9 A.M. to 12:30 P.M.; CPRS = Children's
`Psychiacric Rating Scale Hyperactivity factor score obtained between 1 P.M. and 3 P.M.; CPS = Conners Parent Raring Scale
`Hyperaaivity T score obtained between 4 P.M. and 7 P.M.; TAP= total attempted math problems; TCP= total correct math
`problems, obtained at 11 A.M.; SERS-N and SERS-P, Barkley's Side Effect Rating Scale, nurse and parent forms, respecrivdy. For
`SERS-N and SERS-P, # indicates number of adverse effeas reported, and "sev" indicates mean severity of reported adverse effects.
`
`the Hyperactivity factor of the Children's Psychiatric
`Rating Scale (F3,29 = 34.96, p < .001). A higher dose was
`significantly more effective than a lower dose (F1,31 =
`8.65, p = .006). Across doses, all three active drugs were
`significantly more effective than placebo (Bonferroni
`corrected pairwise comparisons, p < .001). In addition,
`dextroamphetamine Spansules decreased hyperactive
`behavior significantly more than Adderall (p = .04).
`Spansules and immediate-release dextroamphetamine
`did not differ significantly. The significant interaction
`between drug, dose, and order was due mainly to weaker
`effects of high-dose Adderall when it was given before
`low-dose Adderall and a lack of dose-related improve(cid:173)
`ment for Adderall in the increasing-dose group (F3,26 =
`4.81, p = .009). By contrast, both dextroamphetamine
`formulations produced dose-related improvements,
`regardless of the dosing schedule.
`Because of parent work schedules and variable after(cid:173)
`school child-care arrangements, complete late afternoon (4
`P.M. to 7 P.M.) parent ratings of hyperactive behavior were
`available for only 15 subjects. Analyses revealed a signifi(cid:173)
`cant drug effect (F3.12 = 5.84, p = .01 ); Bonferroni adjusted
`pairwise comparisons revealed significant improvements
`versus placebo for dextroamphetamine Spansules (p =
`.007), and Adderall (p = .03), with a trend for immediate(cid:173)
`release dexrroamphetamine (p = .053). A higher dose was
`significantly more effective than a lower dose (F1.14 = 8.04,
`p = .01).
`
`Academic Measures
`Timed academic casks were performed each weekday at
`11 A.M. Stimulants significantly increased the number of
`math problems attempted and number of problems done
`correctly (F3.32 = 6.25, p = .002 and F3,32 = 5.58, p = .003,
`respectively). Immediate-release dextroamphetamine and
`dextroamphetamine Spansules both significantly increased
`the number of problems attempted relative to placebo (p =
`.01 and p = .003, respectively). Improvements on Adderall
`did not reach significance; drug-drug differences were not
`statistically significant.
`Accuracy also improved; the number of problems done
`correctly significantly increased over placebo with imme(cid:173)
`diate-release dextroamphetamine (p = .02) and dexcroam(cid:173)
`phetamine Spansules (p = .003). Adderall did not differ
`significantly from placebo, and no significant between(cid:173)
`drug differences were found. A higher dose did not signif(cid:173)
`icantly improve the number of problems attempted or
`problems done correctly.
`
`Locomotor Activity
`As shown in Figure 1, stimulants significantly decreased
`locomotor activity (F3,54 = 5.50, p = .002). Besides the
`highly significant although trivial effect of time of day, the
`interaction of time and drug was also significant (F33,594 =
`2. 15, p = .004). Pairwise contrasts revealed significant
`decreases in activity for all three drugs relative to placebo
`from 9 A.M. to 7 P.M. Dexrroampheramine Spansules and
`
`J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 40:11, NOVEMBER 2001
`
`1271
`
`
`
`JAMES ET AL.
`
`6.8
`
`6.6
`
`6.4 ·
`
`6.2
`
`6
`
`f
`= (,) <
`
`C 5.8
`..J
`
`5.6
`
`5.4
`
`5.2 ·
`
`-+-Adderall
`~Dex Spansules
`-II-Immediate-release Dex
`..._Placebo
`
`*
`
`*
`
`*
`
`*
`
`*
`
`*
`
`a
`C
`
`5 .... -----.---...-----,---""""!"'---0:------...,....--------,---...-----,-----t
`18
`13
`19
`14
`15
`16
`17
`10
`20
`11
`12
`9
`
`Time
`Fig. 1 Hourly log-transfurmc:d Actomc:ter means and standard errors for 22 children with complete: data: values averaged across weekdays and doses. Hourly
`intervals begin at indicated times. Bonfcrroni corrected significant pairwise: comparisons (p < .0007) arc: indicated as follows: "all stimulants bc:cter than placebo;
`+dc:xtroamphc:tamine Spansulcs and Adderall better than placebo; l\immc:diate-rc:lc:asc: dcxcroamphc:taminc: worse: than placebo: "Addc:rall bc:ttc:r than immc:diatc:(cid:173)
`rc:Jcasc: dexcroamphc:taminc:: "dc:xcroamphc:tamine Spansulc:s better than immc:diare-rc:lc:ase dc:xtroamphetaminc:: cAddc:rall better than dc:xttoamphetaminc:
`Spansulc:s; dimmc:diatc:-rc:lc:asc: dc:xcroamphetaminc: better than Addc:rall; •dc:xtroamphetaminc: Spansulc:s better than Addc:rall.
`
`Adderall significantly reduced activity below placebo levels
`until 8 P.M. (p < .0007). Immediate-release dextroamphe(cid:173)
`tamine significantly increased activity relative to placebo
`from 8 P.M. to 9 P.M. (.Fio,203 = 4.85,p < .0001).
`Adderall significandy reduced activity relative co
`immediate-release dextroamphecamine from 9 A.M. to 10
`A.M., and relative to dextroampheramine Spansules from
`9 A.M. to 11 A.M. (p < .0001). Adderall was comparable
`with both dextroamphetamine preparations between 11
`A.M. and 2 P.M. but was significantly more effective than
`immediate-release dexcroamphetamine between 2 P.M.
`and 3 P.M. (p < .0001). lmmediace-release dextroamphet(cid:173)
`amine was significandy more effective than Adderall
`between 3 P.M. and 5 P.M. (p < .0006). However, the
`reverse was true from 7 P.M. to 8 P.M. (p < .0001).
`Dextroamphetamine Spansules were significandy
`more effective than both immediate-release formulations
`between 4 P.M. and 7 P.M. (p < .0001). Spansules were
`
`also more effective than immediate-release dextroamphet(cid:173)
`amine between noon and 2 P.M. (p < .0001).
`
`Sleep Measures
`
`Overall, stimulants significandy decreased presumed
`sleep duration (F3,19 = 9.92, p < .001). While both
`dextroamphecamine Spansules and immediate-release
`dexcroamphetamine significantly decreased sleep dura(cid:173)
`tion compared with placebo (p < .001, and p = .02,
`respectively), sleep duration on Adderall did not differ
`significantly from placebo (p = .47).
`
`Adverse Effects
`
`Parents and nursing staff recorded the magnitude of
`adverse effects on the 17-item Barkley Side Effect Rating
`Scale (O = absent; 9 = serious), although nurses did not
`rate sleep and nightmares. Nurse ratings (n = 29) revealed
`a significantly increased number of adverse effects (F3,23 =
`
`1272
`
`J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 40:11, NOVEMBER 2001
`
`
`
`3.94, p = .02) and greater mean severity of reported
`adverse effects (F3,23 = 3.56, p = .03). Adderall had an aver(cid:173)
`age of 3.3 adverse effects (SD 2.0), compared with 2.0
`{SD 1.9) on placebo (p = .04). However, dextroamphe(cid:173)
`tamine Spansules had greater severity of reported adverse
`effec~ (mean 3.1, SD 2.0) compared with placebo (mean
`1.8, SD 1.2, p = .02). There were no other significant
`drug-placebo or drug-drug differences. None of the indi(cid:173)
`vidual items reached uncorrected significance levels except
`for nurse-rated appetite decrease, which was comparable
`for all three drugs (Table 4).
`
`Parent ratings (n = 20) did not reveal a stimulant
`main effect, although there was a significant dose by
`dose-order interaction (F1,17 = 7.73,p = .01), which indi(cid:173)
`cated decreased ratings of mean adverse event severity by
`the second week of each phase, regardless of dose, med(cid:173)
`ication, or order of administration. Since parents were
`aware that medications changed every 2 weeks, greater
`alertness to side effects in the first week of each drug
`phase might explain decreased ratings of side effect
`severity by the second week
`
`All three stimulants significantly decreased body
`weight (F3.32 = 13.42, p < .001). The significant dose by
`drug interaction (F3•29 = 4.89, p = .007) resulted from sig(cid:173)
`nificantly lower weight on the higher doses of Adderall
`and dextroamphetamine Spansules (p < .001), but not
`on the higher dose of immediate-release dextroamphec(cid:173)
`amine. The significant drug by dose by dose-order inter(cid:173)
`action (F3,29 = 8.82,p < .001) was associated with greater
`recovery of weight by the second week for subjects receiv(cid:173)
`ing Adderall and dextroamphetamine Spansules in the
`increasing-dose group than in the decreasing-dose group.
`
`Best Medication
`
`At the conclusion of each child's participation, before
`determining the identity and sequence of medications,
`each staff member and parents selected the "best week."
`Daily notes and weekly behavior racings were reviewed by
`staff prior to choosing the best week. Parents selected the
`best week in consultation with the social worker, who had
`access to parent-recorded weekly behavior racings. The
`child and adolescent psychiatrist (RX.C. or RS.J.) would
`then facilitate a consensus, while still blind, about which
`stimulant would be instituted for open treatment optimi(cid:173)
`zation. In the absence of a clear consensus, the stimulant
`with the highest Global Assessment Scale score and best
`teacher ratings was chosen as a starting point. High-dose
`(for a given child) immediate-release dextroamphetamine
`
`.s9 a.]
`~ 5l
`c.. a
`.E~
`"'0 g
`.. e
`u·-
`c2 1l
`~~
`~ Q.
`tlJ §
`~ ..
`u e
`u ~
`~Cl
`<
`0 -g
`·2
`l
`1
`
`l.o,,
`
`~
`~ '! ..c
`~ e
`~ rJ a.a
`u~
`~~
`J]
`I U
`~ ~
`C'II
`..
`"'8 u E C e ::>
`·- 0
`-1
`Cl ~ < C
`=; -~
`C'II
`ti:>
`-oC....
`"'0 0 < rJ
`.. "'
`~...::-
`-2
`~<
`II E
`41::t
`::.: u
`w-.!::
`-' u ..
`m ~:::3'
`t!z~
`l:t: 0
`f! -i3
`V) ~
`-gE
`c,s "'0 _c
`0
`C'II
`M,...:
`II V)
`
`~
`
`- - -Qr-...C---Nr-...C---M r-....1/'\-<r
`~~~~~~~-=~~~~9~~
`
`Li.,
`
`..J c..
`
`~I
`z Cl
`
`::I V)
`
`....: C'II
`
`-
`
`~
`
`~
`
`l'l'\O o--c,-1,,0 ON .... r--.o N 00C'\N
`....:...:...:saoO....:N...:o...:..:No...:
`
`................... .,.....,_._, ...... ...-.. ........................................................ ......
`l.l'IC'\-q<IJ'\t-,..-Nl/'\l/'\t-,..C'\N00-00
`6ss666ss6ssssss
`
`\ON'l-\000'<r-Nl/'\N"\l/'\-N'\ON
`000000000000000
`
`~G~°'~~V\V\~00N'00:=:-N'°'
`sssci66ss6s6ssss
`MM -NO 00NNll"I-C,,\O M -
`-N'l
`ooo~..:0000000000
`
`~GN"R~°'G~00°'00~0~:::-
`888ci68SSSS88886
`
`-r-1 .... C'\C'\ll"IN-'q''q' II'\-< OON"\
`OOONOOOoooddooo
`
`Go~oGolr\M°'~Goo:::-N"oo
`s6sci66sss66ssss
`
`NN'\N-C,,Oll"l--t--.11'\C'\C'OO-'<r
`ooom..:0000000000
`
`~
`
`1 d~ci~ci~~ddd~ddd~dci
`
`l/'\~ll'\C'\ll'\\OOt--.C'\-000t-,..C'00\01,,0
`
`Lz..ld~~~~~~d~~~d~~~d~
`
`i,-----------------
`
`M'GRG:::-000°'00~N"~~~~G~
`Nd~~~N~~~~~N~~~ON
`
`ll"ll'l'\00t--.\Oll'\ll'\OOC'\~\O~C'\t--.C'OC'\
`NOOoo..:M..:..:00..:..:0000
`
`c
`u IV)
`cf Cl
`
`°'~°'M'~:::-~RG00R00M°lr\Mlr\lr\
`ciss66cici66666ci66sci
`
`C'OC'Oll'\C'\00_1/"\_t-,.._l/"\00-00\01'1'\'<r
`moooomm..:o..:..:..:Mooo..:
`
`000:::-M'RGGG~:::-oooM°'MR°'
`01 M..:..:MoMM..:..:..:..:MM..:..:o..:
`0\0000'\l/"\\OMC'\00\0--C"Ot--.ll"ll'l'\N
`~OOOOMmcioo..:MMOOO..:
`
`~ ~~~~~~~~~~~~~~~~~
`
`~1-----------------
`
`:::-~lr\o:::-~°'RGlr\RoRGR~~
`N~~N~N~~~~~N~~oo~
`
`ll"l\000~00-C,,C'\OC'\r-...~\Ol/'\Ol/"\M00
`moo..:omM..:00..:..:..:..:000
`
`wu
`~ u E D o ~ ~..!! -6 -6 .s ~ 8 ~ <.2 ·.: O
`.,, _g .5
`~ g .:.n ~~ ·2 a ·c § ~ e -e § ·>< ~ o..~ ~
`UJ
`~ ~zo~~~~~xo~~~~~o~
`
`~ 5
`u ..c
`..C
`.,.
`0 ~
`~
`-s ·- u
`·-
`·- "'8
`..,
`&
`u ~ "'e ~ .,·:u
`,;;
`JS ..i: ~ - c cu ~ C'II .g i n u o
`
`~
`i.l
`C
`·a. bJ) ~
`~ .s
`·;
`j>-. ~ C
`"'u ~CO i.l t
`
`..C
`U::::I
`
`CC,\OV~
`
`~ " '
`
`ij
`..C
`V
`C'II
`
`
`
`JAMES ET AL.
`
`was chosen most frequently (32%), followed by low-dose
`dextroamphecamine Spansules (26%), high-dose Adderall
`(18%), high-dose dexrroamphetamine Spansules (12%),
`and low-dose Adderall and low-dose immediate-release
`dexrroamphecamine, both (6%).
`Parents were given the option of a brief open trial of
`MPH after the conclusion of the controlled trial. Seven(cid:173)
`teen children were given MPH hydrochloride, including
`13 who had nor previously taken it. Open treatment was
`generally adjusted weekly.
`Twelve subjects (34%) were discharged on immediate(cid:173)
`release dexrroamphetamine, 8 (23%) on MPH, 6 (17%)
`on Adderall, 5 (14%) on dextroamphetamine Spansules,
`and 4 (11 %) on a combination of immediate-release dex(cid:173)
`rroamphetamine and dexrroamphetamine Spansules.
`
`DISCUSSION
`
`This is the fuse controlled comparison of Adderall and
`both formulations of dexrroamphecamine sulfate. In addi(cid:173)
`tion, this is the fuse controlled study of dextroamphecam(cid:173)
`ine Spansules manufactured by a new aqueous process and
`it contributes to the limited literature on sustained-release
`amphetamines (Brown et al., 1980; Pelham et al., 1990).
`This study was performed without funding from pharma(cid:173)
`ceutical companies, with the overall goal of informing clin(cid:173)
`ical practice. The strengths of the study included a
`crossover design and a research laboratory school that
`allowed the same trained raters to study all subjects.
`Besides confirming the robust and surprisingly pro(cid:173)
`longed efficacy of a single morning dose of all three stimu(cid:173)
`lants versus placebo on a wide range of objective and
`subjective measures, we also detected some subtler drug(cid:173)
`drug differences. Consistent with its extended-release
`design, dextroamphetamine Spansules were significantly
`less effective on morning teacher ratings than immediate(cid:173)
`release dextroamphetamine, which did not differ signifi(cid:173)
`cantly from Adderall. Objective wrist-mounted Actometers
`confirmed that Adderall was more effective than immedi(cid:173)
`ate-release dextroamphecamine for the first hour and than
`Sp~ules for the first 2 hours of morning classroom time.
`There were no drug-drug differences on our measure
`of academic productivity, but Adderall did not signifi(cid:173)
`cantly increase the number of math problems attempted
`or completed relative to placebo, whereas Spansules had a
`robust effect on both measures (p = .003). However, we
`may have missed the peak efficacy of these compounds
`by obtaining our academic measures nearly 4 hours after
`drug administration. .
`
`During afternoon recreation therapy, all three drugs
`retained efficacy relative to placebo, but dextroamphet(cid:173)
`amine Spansules were still significantly more effective than
`Add