Transition from methylphenidate or amphetamine to atomoxetine in children and adolescents

1168_quintana 6/29/07 9:59 AM Page 1168

Clinical Therapeutics/Volume 29, Number 6, 2007

Transition from Methylphenidate or Amphetamine
to Atomoxetine in Children and Adolescents with
Attention-Deficit/Hyperactivity Disorder—A Preliminary
Tolerability and Efficacy Study
Humberto Quintana, MD1; Edward A. Cherlin, MD2; David A. Duesenberg, MD3;
Mark E. Bangs, MD4; Janet L. Ramsey, MS4; Peter D. Feldman, PhD4;
Albert J. Allen, MD, PhD4; and Douglas K. Kelsey, MD, PhD4
1Department of Psychiatry, Louisiana State University Health Sciences Center, New Orleans, Louisiana;
2Valley Clinical Research, El Centro, California; 3St. John’s Mercy Medical Center, Chesterfield, Missouri;
and 4Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana

ABSTRACT (ADHDRS-IV-Parent:Inv), was assessed from baseline
Background: The primary treatment for attention- to end point.
deficit/hyperactivity disorder (ADHD) has been psy- Results: Of the 62 subjects enrolled in the study, 39
chostimulants. Recently developed nonpsychostimulant (62.9%) were diagnosed as ADHD-combined type.
treatments have allowed certain patients to switch Similar proportions were receiving methylphenidate
from a psychostimulant to a nonpsychostimulant. How- (51.6%) and amphetamine (48.4%). Slightly more
ever, the outcomes of such switches have not been sys- wished to switch due to inadequate response (53.2%)
tematically studied. than intolerability (46.8%). Nine subjects discontin-
Objective: The purpose of this pilot study was to ued at various times during the course of the study
assess treatment tolerance and efficacy during a cross- (patient or parent/caregiver decision [4], AE [2], pro-
taper transition from methylphenidate or ampheta- tocol violation [2], and lack of efficacy [1]). Mean
mine to atomoxetine among children and adolescents (SD) ADHDRS-IV-Parent:Inv total scores (n = 59, last-
with ADHD. observation-carried-forward) improved significantly
Methods: This pilot study was conducted in pa- from baseline (visit 2) to an end point (32.1 [10.5] vs
tients (aged 6–17 years) with incomplete responses 22.6 [14.0]; P < 0.001). Of the 58 subjects answering
(failure to obtain full reduction/elimination of symp- in the atomoxetine monotherapy phase, 38 (65.5%)
toms) or intolerance of adverse events (AEs) during reported a preference for atomoxetine treatment over
psychostimulant treatment. Patients continued ongo- their previous psychostimulant. Tolerability results were
ing psychostimulant treatment during the first week of as follows: 26 (44.1%) of 59 patients reported ≥1 AE,
the study. Transition to atomoxetine began by admin- the most common being somnolence (4 [6.8%]), fa-
istering atomoxetine 0.5 mg/kg ⅐ d plus full-dose psy- tigue (3 [5.1%]), decreased appetite (3 [5.1%]), cough
chostimulant for 1 week, followed in the second week (3 [5.1%]), headache (3 [5.1%]), and contact dermati-
by 1.2 mg/kg ⅐ d atomoxetine plus half-dose psycho- tis (2 [3.4%]). No clinically severe AEs were reported.
stimulant. Patients remained on 1.2 mg/kg ⅐ d atomoxe-
tine monotherapy for the remaining 5 weeks. This
This research was presented in part at the Scientific
stepwise transition was enacted due to the difference
Proceedings of the 52nd Annual Meeting of the American
in pharmacodynamics between the psychostimulants Academy of Child and Adolescent Psychiatry, October 18–23,
and atomoxetine. Applying a stepwise cross-titration 2005, Toronto, Ontario, Canada.
allowed for better control of ADHD symptoms during
Accepted for publication April 6, 2007.
the intervening period. Change in ADHD symptoms, doi:10.1016/j.clinthera.2007.06.017
as measured by the mean change in the Attention- 0149-2918/$32.00
Deficit/Hyperactivity Disorder Rating Scale-IV- Printed in the USA. Reproduction in whole or part is not permitted.
Parent Version: Investigator-administered and -scored Copyright © 2007 Excerpta Medica, Inc.

1168 Volume 29 Number 6

1168_quintana 6/29/07 9:59 AM Page 1169

H. Quintana et al.

Both mean (SD) diastolic (2.4 [7.8] mm Hg; P = throughout childhood and well into adulthood. The
0.031) and systolic (2.4 [7.9] mm Hg; P = 0.029) mainstay of pharmacotherapy has been psycho-
blood pressures increased significantly from baseline stimulants, particularly dextroamphetamine, methyl-
to end point. Electrocardiography revealed a signifi- phenidate, and preparations containing 1 of these
cant increase in mean (SD) heart rate (9.2 [11.6] bpm; 2 medications as their functional agent. In 2003, ato-
P < 0.001) and a corresponding decrease in mean (SD) moxetine, a norepinephrine reuptake inhibitor and a
RR interval (–77.8 [98.2] ms; P < 0.001). Statistically nonpsychostimulant, was approved by the US Food
significant, but mild, increases in diastolic pressure and Drug Administration, providing an alternative to
and heart rate were observed. the psychostimulants for the treatment of ADHD in
Conclusion: These children and adolescent patients children and adults. As with most medications, not all
were successfully switched from methylphenidate or treatments are effective for all patients, and some pa-
amphetamine to atomoxetine treatment, with result- tients will respond better to one medication than to
ing improvement in ADHD symptom severity from another.9 In fact, ~60% to 90% of patients will re-
baseline in this pilot study. (Clin Ther. 2007;29: spond to treatment with the various psychostimu-
1168–1177) Copyright © 2007 Excerpta Medica, Inc. lants,10–13 and similar rates have been reported for
Key words: amphetamine, atomoxetine, disor- patients receiving atomoxetine.9,14,15 More will be
der, attention-deficit/hyperactivity methylphenidate, known about differential response rates when a study
psychostimulant. funded by the National Institute of Mental Health,
comparing atomoxetine and methylphenidate, is com-
pleted (Newcorn et al., in progress). Nonetheless, over-
INTRODUCTION all response to treatment might be associated with
Attention-deficit/hyperactivity disorder (ADHD) is a the response criteria being applied, the person provid-
common neuropsychiatric condition of childhood, ing the evaluation (patient vs parent or physician),
characterized by developmentally inappropriate levels and the comorbidities that are present. Adverse events
of hyperactivity and impulsivity and persistent diffi- (AEs) have been found to occur during treatment in
culties in maintaining sustained attention.1 The condi- upwards of 40% of patients.16,17 A proportion (25%–
tion is both common and widespread—the prevalence 30%) of patients will remain underresponsive to—
in the United States is frequently cited as being be- or intolerant of (that is, will experience AEs)—psy-
tween 3% and 7% among school-aged children,1 while chostimulant treatment for ADHD, regardless of the
estimates of the prevalence of ADHD or hyperkinetic medication. Therefore, it remains the responsibility of
disorder from studies in other countries have varied the treating physician to determine which medication to
widely between 1% and 19.8%,2,3 due in large part prescribe and the optimal treatment for each patient.
to differences in methodology and diagnostic use of In view of the frequent need (25% –30%) to switch
the International Statistical Classification of Diseases patients from one medication to another, it was of in-
and Related Problems, 10th Revision4 and the Inter- terest to examine the process and determine if any is-
national Classification of Diseases, Ninth Revision, sues regarding efficacy or tolerance might arise in
Clinical Modification.5 ADHD can have a major im- switching between a psychostimulant and a non-
pact on the life of the individual and on the whole psychostimulant during treatment for ADHD. While
family. Compared with their non-ADHD peers, chil- many patients have made the switch from a psycho-
dren and adolescents with ADHD have higher rates of stimulant to atomoxetine, an examination of the liter-
school suspension and are less likely to graduate high ature (Index Medicus, between 2000 and present)
school.6 As young adults, they have 3 times as many suggests that the outcomes of such switches have not
traffic accidents,7 are at far greater risk for sexually been systematically studied. Accordingly, this report
transmitted disease and unwanted pregnancy,6 and are details the findings of a preliminary (pilot) study con-
likely to have more frequent problems with the law.8 ducted to assess a method of transitioning from a psy-
ADHD is a chronic condition, although the severity chostimulant to nonpsychostimulant. In particular, we
of symptoms of impulsivity and hyperactivity tends to wanted to identify any changes in ADHD symptoms
lessen during the transition to adolescence and adult- during a cross-tapering transition from 2 of the lead-
hood. Nonetheless, treatment might need to continue ing psychostimulants, methylphenidate or ampheta-

June 2007 1169

1168_quintana 6/29/07 9:59 AM Page 1170

Clinical Therapeutics

mine, to atomoxetine in patients that had an incomplete Study Design and Treatment
response or had experienced significant AEs during This study consisted of 3 study periods. Following
their current regimen of psychostimulant therapy. enrollment, patients were evaluated on 2 successive
Specific hypotheses to be tested were that switching visits (visits 1 and 2) spaced ~1 week apart, during
from a psychostimulant to atomoxetine was: (1) not which they continued to receive ongoing psychostimu-
associated with a worsening of ADHD symptoms; and lant treatment at their current dose. Patients were
(2) not associated with the development of intolerable screened during this period (Study Period I [SP-I]), the
new AEs. diagnostic interview was administered, baseline as-
sessments were acquired, and written informed con-
PATIENTS AND METHODS sent was obtained. During the succeeding transition
Patient Sample period (SP-II), patients received 0.5 mg/kg ⅐ d atomoxe-
This pilot study (Study B4Z-MC-LYCI) was con- tine plus a full-dose of their current psychostimulant
ducted at 6 sites in the United States between July (methylphenidate or amphetamine) as prescribed by
2004 and January 2005. Patients were male or female their physician (who was not associated with this
outpatients and were ≥6 years of age at study initiation study) for 1 week; followed by 1 week in which they
and <18 years of age at study completion. Patients received 1.2 mg/kg ⅐ d atomoxetine plus a half-dose of
had to meet the Diagnostic and Statistical Manual psychostimulant. This stepwise transition was enacted
of Mental Disorders, Fourth Edition,18 diagnostic cri- due to the difference in pharmacodynamics between
teria for ADHD (any subtype), as determined by a the psychostimulants and atomoxetine; whereas psy-
physician-investigator’s clinical assessment and the chostimulants have a short t1/2 that allows for rapid
Schedule for Affective Disorders and Schizophrenia control of symptoms at treatment initiation and re-
for School-Aged Children–Present and Lifetime sults in a rapid return of ADHD symptoms after treat-
Version19 (K-SADS-PL) semistructured interview, and ment cessation,11,12 atomoxetine has a more gradual
be of normal intelligence. Eligible study patients had onset of action.15,21 Thus, applying a stepwise cross-
to be receiving ongoing treatment with a single psy- titration might allow for greater control of ADHD
chostimulant at a dose considered by their doctors symptoms during the intervening period.
as an effective dose, and patient/parent had to be Assessment of vital signs, weight/height, treatment
amenable to dose reduction. Patients needed to have response, and AEs were made at the end of each week
experienced AEs while on psychostimulant therapy or (visits 3 and 4). In cases where a dose of psychostimu-
to have had an unsatisfactory response to the psycho- lant was not easily reduced by half, the investigator
stimulant, such that the treating physician or parent/ was instructed to “round up” to the next available dose.
patient desired to change the patient’s therapy. Patients unable to tolerate 1.2 mg/kg ⅐ d of atomoxe-
Exclusion criteria included a good response to psy- tine, whether through lack of efficacy, as measured by
chostimulant therapy, because there would be no jus- significant changes of ≥20% on ADHD rating measures
tification for switching therapy. Other exclusion or an unbearable AE (eg, excessive sleepiness during
criteria included a history of bipolar I or II disorder, the day, insomnia, or abdominal [gastric] upset), were
psychosis, or a pervasive developmental disorder, any discontinued from the study. The third study period
seizure disorder, a recent history of drug or alcohol (SP-III) lasted ~5 weeks, during which patients re-
abuse, hypertension, or a determination by the inves- mained on 1.2 mg/kg ⅐ d of atomoxetine monotherapy,
tigator to be at serious suicidal risk. Prior to partici- and all psychostimulant treatment was discontinued.
pation, patients and their parents or guardians were An assessment was made after ~2 weeks (visit 5) and
provided with a complete description of the study and again ~3 weeks later (visit 6). The entire study lasted
a child assent form, and a parent or legal guardian ~8 weeks. Patient compliance was assessed at each
signed an informed consent document approved by visit by direct questioning and pill count. A patient
the study site’s institutional review board. The study was considered significantly noncompliant if he or she
was approved by site-specific local institutional re- failed to take ≥80% of the prescribed doses of study
view boards and/or a central institutional review medication during any visit interval. Noncompliance
board and followed the guidelines for Good Clinical with the recommended protocol or medication treat-
Practices, as set forth by the Declaration of Helsinki.20 ment instructions, such as failing to take medications


1168_quintana 6/29/07 9:59 AM Page 1171

H. Quintana et al.

as instructed or noncompliance with drug counts, was variance (ANCOVA), using a model with terms for
grounds for discontinuation. baseline score, psychostimulant type (methylpheni-
date or dextroamphetamine), visit, primary reason for
Efficacy and Tolerability Assessments switching, and visit-by-reason-for-switching interac-
The primary objective of the study was to assess tion. Baseline-to-end point changes in efficacy mea-
the change in ADHD symptoms, as measured by the sures were also assessed using a last-observation-
mean change in the Attention-Deficit/Hyperactivity carried-forward ANCOVA with terms for baseline
Disorder Rating Scale-IV-Parent Version: Investigator- score, psychostimulant type, and primary reason for
administered and -scored22 (ADHDRS-IV-Parent:Inv) switching. For the ADHDRS-IV-Parent:Inv total
total score from visit 2 (baseline, full-dose psycho- score, if a single item was missing in any subscale, the
stimulant monotherapy) to the SP-III end point (visit 5 mean score for all other items in the subscale was im-
or 6, atomoxetine monotherapy). Secondary objec- puted as the score for the missing item. If ≥1 item was
tives included comparison of the efficacy of atomoxe- missing in either subscale, the total score was also
tine with that of the psychostimulants using the considered missing. To examine individual responses to
ADHDRS-IV-Parent:Inv Inattention and Hyperactivity/ treatment, ADHDRS-IV-Parent:Inv total scores were
Impulsivity subscales and the Clinical Global also expressed as T-scores. A T-score is a transforma-
Impressions—Improvement23 of ADHD scale (CGI- tion of the raw score based on normative data, in
ADHD-I; a single-question, 7-point scale ranging from which the mean for healthy individuals is 50 and each
1 = very much improved to 7 = very much worse). change of 10 points represents 1 SD. Improvement or
Information regarding symptoms was provided by worsening of symptom severity by 1 SD could there-
parents on behalf of child subjects, or by both the par- by be assessed by examining the prevalence of decreas-
ents and subjects if the latter were adolescents. The es or increases, respectively, in T-score by 10 points.
ADHDRS-IV-Parent:Inv was assessed at each visit Changes in vital signs and laboratory measures were
(visits 1–6), and the CGI-ADHD-I at baseline through analyzed using a paired t test, while electrocardio-
visit 6. Tolerability was monitored by collection of graphic changes from baseline were assessed with the
data concerning AEs, as well as assessment of labora- Wilcoxon signed rank test. All statistical tests were per-
tory results (chemistries, hematology, urinalysis, elec- formed using a 2-sided, 0.05 significance level unless
trocardiography, and vital signs). Assessment of AEs otherwise specified.
and vital signs was performed at baseline and at each
postbaseline visit. Electrocardiography was performed RESULTS
at baseline and visits 3 (mid-transition) and 6. All Study Entry and Baseline: Psychostimulant
other laboratory results were acquired at baseline and Monotherapy, Visits 1 and 2
visit 6. Patients who discontinued prematurely were Seventy-one patients were screened for the study,
administered all end point assessments at the time of and 62 (87.3%) were enrolled (Figure). Nine patients
discontinuation in accordance with the principles of were excluded for the following reasons: failure to
intent-to-treat,24 and all measures in this report were meet entry criteria (4), patient or caregiver decision to
defined as a priori in the study protocol. withdraw (3), lost to follow-up (1), and discontinued
due to sponsor decision (1). Patients enrolled in the
Statistical Analyses study had an overall mean (SD) age of 11.5 (2.8) years
Approximately 60 patients were planned for this (range, 6.1–17.5 years) and a mean (SD) weight of
study. As this study was considered a pilot, this sam- 41.1 (15.5) kg (range, 22.0–70.0 kg). Most patients
ple size was estimated to provide 80% power, at a sig- (47 [75.8%]) were male and whites accounted for the
nificance level of 0.05, to detect a difference of majority (46 [74.2%]); the rest being either Hispanic
≥5.5 points in a mean ADHDRS-IV-Parent:Inv total (10 [16.1%]) or black (6 [9.7%]). Most patients (39
score from baseline to the SP-III end point, assuming [62.9%]) were diagnosed with ADHD of the com-
a per-patient SD of ≤13.5 points and assuming that bined type, while 22 (35.5%) were diagnosed as pre-
≥50 patients could provide a visit 5 or 6 observation. dominantly inattentive and 1 (1.6%) as predominantly
Mean efficacy scores and pertinent subscale scores hyperactive/impulsive. Genetic information regarding
were assessed by repeated-measures analysis of co- cytochrome P450 2D6 activity was available for 60 pa-

June 2007 1171

1168_quintana 6/29/07 9:59 AM Page 1172


Screened
(N = 71) Excluded (n = 9)
Study
Failed to meet criteria (4)
Period Visit Week
Patient/caregiver decision (3)
1 0 Lost to follow-up (1)
Enrolled Sponsor decision (1)
(n = 62)
I

2 1
Completed lead-in
(n = 62)
Discontinued (n = 3)
II Patient decision (2)
Adverse event (1)
Completed cross-taper
4 3 (n = 59)
Discontinued (n = 6)
Patient/caregiver decision (2)
III Protocol violation (2)
Completed titration Lack of efficacy (1)
(n = 53) Adverse event (1)
6 8

Figure. Flowchart representing disposition of children and adolescent patients who had experienced incomplete
responses or intolerance of adverse events during psychostimulant treatment.

tients, of whom 56 (93.3%) were characterized as ex- citing inadequate response as their reason for wanting
tensive metabolizers. Baseline concurrent diagnoses to switch than among those citing AEs.
indicated 16 patients with oppositional defiant disor-
der, 2 with conduct disorder, and 1 each with major Transition Period: Psychostimulant Plus
depressive disorder, dysthymia, or generalized anxiety Atomoxetine, Visits 3 and 4
disorder. The K-SADS-PL assessment and diagnosis Fifty-nine (95.2%) patients completed the 2-week
were made by the primary investigator (H.Q.). cross-taper transition period (SP-II), with 3 discontinu-
Patients were grouped with regard to those who ations; 1 due to an AE (fatigue, visit 4) and 2 due to
had been receiving methylphenidate (32 [51.6%]) and “patient decision” (visit 5). No patients discontinued
those receiving amphetamine (30 [48.4%]). Thirty- the study at this point due to lack of efficacy. Treatment
three (53.2%) patients cited inadequate response as compliance was 91.7% at visit 3 and 96.7% at visit 4.
their reason for wanting to switch medications, while Relative to baseline (visit 2), least squares mean
29 (46.8%) cited AEs. Study patients’ symptoms were ADHDRS-IV-Parent:Inv total scores (Table I) decreased
characterized as moderate to severe, with an imputed significantly following the addition of 0.5 mg/kg ⅐ d of
mean (SD) ADHDRS-IV-Parent:Inv total score at atomoxetine to full-dose psychostimulant (visit 3) and
visit 1 of 33.2 (10.5). The mean (SD) ADHDRS-IV- decreased further following the increase of atomoxe-
Parent:Inv total T-score was 74.1 (12.4), indicating tine dose to 1.2 mg/kg ⅐ d and the reduction of the psy-
that patients’ average severity was ~2.4 SDs above the chostimulant dose to one half (visit 4). Least squares
age- and sex-matched norms. The imputed mean (SD) in- mean scores for the Inattention and Hyperactivity/
attention subscale score of the ADHDRS-IV-Parent:Inv Impulsivity subscales also reflected this pattern of im-
(19.6 [5.5]) suggested somewhat greater impairment provement (Table I), as did CGI-ADHD-I scores
than did the mean (SD) imputed hyperactivity/ (mean [SE]: visit 3, 3.2 [0.1], n = 61; P < 0.001; visit 4,
impulsivity subscale score (13.6 [7.3]). Mean baseline 2.8 [0.1], n = 60; P < 0.001).
ADHDRS-IV-Parent:Inv total score and total T-score More than half (32 [51.6%]) of the patients report-
were noted to be significantly higher among patients ed an AE during the course of the study, but with


1168_quintana 6/29/07 9:59 AM Page 1173

H. Quintana et al.

Table I. Attention-Deficit/Hyperactivity Disorder Rating Scale-IV-Parent Version: Investigator-administered and -
scored (ADHDRS-IV-Parent:Inv) total score and subscores by visit (least squares mean [SE]): Relative to
baseline (visit 2) in children and adolescent patients who had experienced incomplete responses or intol-
erance of adverse events during psychostimulant treatment.*

Hyperactivity/
No. of Inattention Impulsivity
Treatment Visit Week Patients Total Score Subscore Subscore

Full-dose psychostimulant only 2 1 61 32.1 (0.9) 18.9 (0.5) 13.2 (0.5)
Full-dose psychostimulant plus
atomoxetine 0.5 mg/kg ⅐ d 3 2 61 24.4 (1.3) 14.2 (0.7) 10.2 (0.7)
Half-dose psychostimulant plus
atomoxetine 1.2 mg/kg ⅐ d 4 3 60 20.8 (1.4) 12.5 (0.8) 8.3 (0.7)
Atomoxetine monotherapy 5 5 59 22.6 (1.7) 13.3 (0.9) 9.2 (0.8)
Atomoxetine monotherapy 6 8 56 22.5 (1.6) 13.5 (0.9) 9.1 (0.8)

*Repeated measures least squares mean scores were significantly reduced for the ADHDRS-IV-Parent:Inv total score and both
subscores at all postbaseline time points (P < 0.001; analysis of covariance with terms for visit 1 score, psychostimulant type,
visit, primary reason for switching, and visit-by-reason-for-switching interaction).

no placebo control group with which to compare, the caregiver decision (2), protocol violation (2), AE (1),
treatment association could not be determined. The and lack of efficacy (1).
most common events (≥2%) were nausea (5 [8.1%]), ADHDRS-IV-Parent:Inv scores indicated little change
fatigue (3 [4.8%]), and headache (3 [4.8%]). There in symptom severity from the end of the preceding
were no serious AEs reported in this period. Mean transition period to the visit 6 end point (Table I).
(SD) diastolic blood pressures increased significantly Relative to study baseline, however, significant im-
from baseline to visit 4 by 2.8 (8.0) mm Hg (n = 57, provements were seen on the ADHDRS-IV-Parent:Inv
t1 = 2.61; P = 0.012). No other vital signs changed sig- total score and the Inattention and Hyperactivity/
nificantly from baseline. Electrocardiography revealed Impulsivity subscale scores. The mean (SE) ADHDRS-
a significant increase in mean (SD) heart rate (+6.2 IV-Parent:Inv total T-score was reduced from baseline
[10.4] bpm, n = 60; P < 0.001, Wilcoxon signed rank (72.8 [0.9], n = 61) to an end point (62.6 [1.7], n = 56;
test) and a corresponding decrease in mean (SD) RR P < 0.001), indicating an improvement by ~1 SD.
interval (–58.3 [114.1] ms, n = 60; P < 0.001). No Improvement (ADHDRS-IV-Parent:Inv total T-score
clinically significant changes were seen in any other elec- reductions of ≥10 points) was seen in 34 (57.6%) of
trocardiographic measures, including data-corrected QT 59 patients, with a reduction of symptom severity by
interval. as much as 3.2 SDs. Corresponding baseline-to-end
point decreases in the actual ADHDRS-IV-Parent:Inv
Completion: Atomoxetine Monotherapy, total score ranged from –10 to –31. Worsening (T-score
Visits 5 and 6 increases of ≥10 points) was seen in 6 (10.2%) patients,
By the end of the atomoxetine maintenance phase with an increase in symptom severity by as much as
(visit 5 or 6, SP-III), the final mean (SD) dose of ato- 1.6 SDs. Corresponding increases in actual ADHDRS-
moxetine was 1.19 (0.15) mg/kg · d (n = 59; range: IV-Parent:Inv total score ranged from 8 to 16.
0.55–1.38 mg/kg · d). Treatment compliance was 98.3% Nineteen (32.2%) patients showed no substantial im-
at visit 5 and 82.1% at visit 6. Overall treatment com- provement or worsening. The end point mean CGI-
pliance for the entire study was 100%. Fifty-three ADHD-I score also reflected significant improvement
(89.8%) of the 59 patients entering SP-III completed (least squares mean [SE], 2.9 [0.2], n = 56; P < 0.001
the 5-week treatment. Six patients discontinued; parent/ vs baseline, repeated measures ANCOVA), as it did

June 2007 1173

1168_quintana 6/29/07 9:59 AM Page 1174


for all other postbaseline visits (P < 0.001 for visits erate AEs associated with psychostimulant treatment
3–6). Reason for wanting to switch (AEs, inadequate were successfully switched to a nonpsychostimulant
response) was not a significant factor in patients’ change using the proposed medication transition method,
in scores on any measure of efficacy. with no notable increase in treatment intolerance or
Twenty-six (44.1%) patients reported ≥1 AE, the lack of efficacy. Patients who completed this study
most common (≥2%) being somnolence (4 [6.8%]), (>90%) who had moderate-to-severe levels of ADHD
fatigue (3 [5.1%]), decreased appetite (3 [5.1%]), symptoms at baseline showed significant reductions in
cough (3 [5.1%]), headache (3 [5.1%]), and contact der- symptoms during the 7-week transition from amphet-
matitis (2 [3.4%]). Again, no clinically severe AEs were re- amine or methylphenidate to atomoxetine. Although
ported. Both mean (SD) diastolic (2.4 [7.8] mm Hg, n = similar transition or crossover studies have been per-
53, t1 = 2.21; P = 0.031) and systolic (2.4 [7.9] mm Hg, formed between 2 psychostimulants25,26 or between a
n = 53, t1 = 2.24; P = 0.029) blood pressures increased psychostimulant and an off-label nonpsychostimulant
significantly from baseline (visit 2) to visit 6. treatment for ADHD such as bupropion,27 guanfa-
However, no other vital signs changed significant- cine,28 or modafinil,29 the present study is, to our
ly from baseline. As in SP-II, electrocardiography re- knowledge, the first to examine the transition from a
vealed a significant increase in mean (SD) heart rate psychostimulant (methylphenidate or amphetamine)
(+9.2 [11.6] bpm, n = 57; P < 0.001, Wilcoxon signed to a nonpsychostimulant (atomoxetine) that has been
rank test) and a corresponding decrease in mean (SD) approved for treatment of ADHD in children, adoles-
RR interval (–77.8 [98.2] ms, n = 57; P < .001). No cents, and adults.
other electrocardiographic measures changed signifi- A few studies have examined the feasibility of tran-
cantly. Statistically significant changes from visit 1 to sitioning a patient from one ADHD medication to an-
visit 6 were observed in several laboratory measures, other type of medication for the treatment of ADHD.
but none was considered clinically meaningful by the Pelham et al30 conducted a placebo-controlled crossover
study investigators (Table II). evaluation in children and adolescents of the efficacy
of immediate-release methylphenidate (20 mg/d BID;
DISCUSSION N = 21), sustained-release methylphenidate (20 mg/d QD),
The present pilot study suggests that these patients sustained-release dextroamphetamine (10 mg/d QD),
who were inadequately responsive to or could not tol- and pemoline (56.25 mg/d QD) concluding that the

Table II. Laboratory measures in children and adolescent patients who had experienced incom-
plete responses or intolerance of adverse events during psychostimulant treatment and
were transitioned to atomoxetine 1.2 mg/kg ⅐ d over the course of 8 weeks (N = 58). All
changes were deemed clinically nonsignificant.

Baseline, Change at End Point,
Analysis Mean (SD) Mean (SD) P

Albumin, g/L 46.0 (2.3) –1.7 (3.1) <0.001
ALT, U/L 17.8 (7.8) –2.1 (6.3) 0.008
Calcium, mmol/L 2.5 (0.1) –0.0 (0.1) 0.004
Total cholesterol, mmol/L 4.2 (0.7) –0.3 (0.5) <0.001
GGT, U/L 12.7 (5.3) –1.1 (2.3) <0.001
Total protein, g/L 75.3 (4.0) –1.9 (3.6) <0.001
Erythrocyte count, ×1012 cells/L 4.9 (0.3) –0.1 (0.3) 0.027
Hematocrit 0.4 (0.0) –0.0 (0.0) <0.001
Cell volume, fL 84.1 (4.4) –1.8 (3.1) <0.001

ALT = alanine aminotransferase; GGT = γ-glutamyltransferase.


1168_quintana 6/29/07 9:59 AM Page 1175

H. Quintana et al.

4 treatments were equally effective. Efron et al25 drew istics that might give the clinician an indication of
similar conclusions regarding immediate-release methyl- how best to tailor a patient’s pharmacotherapy, exami-
phenidate and dextroamphetamine in a large cross- nation of such factors as the patient’s age, sex, racial
over study in children (N = 125). They found that a origin, prior psychostimulant use, ADHD subtype,
clinical response to both medications could be demon- and the presence of specific comorbidities has thus far
strated in 42% to 58% of their study sample, depend- proven largely fruitless, although prior use of psycho-
ing on the efficacy measure being applied, while stimulants among patients enrolled in a subsequent
~15% to 25% had a preferential response to one par- atomoxetine clinical trial might predict a lower likeli-
ticular medication. These figures do not differ greatly hood of response.36 Otherwise, the aforementioned
from those obtained by Newcorn et al9 in their com- demographic factors appear not to be associated with
parison of rates of response to immediate-release differential responsiveness to one medication over
methylphenidate versus atomoxetine in 2 clinical another.9,25 This has led some to conclude that
trials involving children with ADHD. In the results no single attribute or feature can predict a patient’s
of one trial (Study B4Z-US-LYAV), Newcorn et al response to a medication, but that it might be better
found that 11% of children responded preferential- instead to perform limited-duration “test runs” of
ly to methylphenidate (up to 1.8 mg/kg ⅐ d TID), medications to determine which is best for a particu-
24% responded preferentially to atomoxetine (up to lar patient25 and switch treatment when needed. In-
1.8 mg/kg ⅐ d BID), 43% responded to both, and 23% creasing knowledge and future availability of pharma-
failed to respond to either. Similarly, in Study B4Z- cogenomics may assist the clinician in choosing
MC-LYBI, 15% responded preferentially to sustained- medication less likely to be associated with AEs and to
release methylphenidate (up to 54 mg/d QD), 18% re- be effective in patients.37
sponded preferentially to atomoxetine (up to 1.8 mg/kg ⅐ d Given that this was a pilot study, a number of
BID), 46% responded to both, and 21% failed to re- drawbacks are evident. Due to the limited sample size
spond to either. However, the analysis in this second and absence of a placebo comparator, it is unclear to
study might not be a fair comparison, as a typical what extent these results can be generalized to the
clinical dose of atomoxetine is 1.2 mg/kg ⅐ d, whereas clinical setting. The study was not “powered” for ex-
the dose used in Study B4Z-MC-LYBI (maximum, tensive analysis of subgroups. It might have been in-
1.8 mg/kg ⅐ d) was above the US Food and Drug structive, for example, to determine if any differences
Administration (FDA)-approved maximum treatment could be seen in the tolerability outcomes of patients
dose. By contrast, the upper range for the doses of starting on methylphenidate therapy versus those
methylphenidate that were used (1–2 mg/kg ⅐ d) just starting with amphetamine, or if patients who chose
met the FDA-approved maximum. to switch from psychostimulants due to inadequate
According to the available literature, it appears response were less likely to respond to atomoxetine
that ~60% to 80% of patients respond adequately therapy than those switching due to AEs. This trial
to any particular ADHD medication, regardless of was also not a true crossover study. Patients’ treat-
whether the treatment in question is methylpheni- ment regimens included a cross-tapering from psy-
date,9,25,31 an amphetamine,10,11,13,25 a tricyclic anti- chostimulants to atomoxetine, but not the reverse, so
depressant,32–34 or atomoxetine.9 Clinically, in practice, it is not obvious whether a switch from atomoxetine
and in at least initial reports,14,15 a real response rate to a psychostimulant would have identified any toler-
of 65% has been reported for atomoxetine. Thus, ability or efficacy issues. The study was conducted
while many patients will respond well to ≥1 type of under an open label and with no placebo arm. Thus,
medication,9,25 25% to 30% of patients remain who it is unclear how much of the improvement in ADHD
might nonetheless respond preferentially to one medi- symptoms was due to nonspecific factors such as pa-
cation over another. This point, combined with the tient or physician-rater expectation, or to what extent
finding that 70% to 80% of ADHD patients follow the tolerability findings, particularly AEs, were associ-
the recommended medical treatment regimen for an ated with the active medication. Finally, because the
average of <2 months after initiation of any treat- study sample consisted of patients who had previous-
ment,35 underscores the importance of finding alter- ly experienced AEs (a level of intolerance) or had been
native treatments. In the search for patient character- inadequately responsive to psychostimulants, it was

June 2007 1175

1168_quintana 6/29/07 9:59 AM Page 1176


not possible to ascertain the relative efficacy and tol- REFERENCES
erability of the 3 active medications ascertained. 1. American Psychiatric Association. Diagnostic and Statistical
Further studies are also needed to guide physicians Manual of Mental Disorders, 4th Edition, Text Revision: DSM-IV-
on how better to choose or predict what might be the TR. Washington, DC: American Psychiatric Association;
best or most appropriate medication treatment for 2000.
2. Gadow KD, Nolan EE, Litcher L, et al. Comparison of
ADHD and other psychiatric disorders among the
attention-deficit/hyperactivity disorder symptom sub-
many options currently available. The correct choice
types in Ukrainian schoolchildren. J Am Acad Child Adolesc
should not only be effective but should also produce
Psychiatry. 2000;39:1520–1527.
minimal AEs in a specific patient. 3. Lord J, Paisley S. The clinical effectiveness and cost-
effectiveness of methylphenidate for hyperactivity in child-
CONCLUSIONS hood. London, UK: NICE, Version 2; August 2000.
The results of this pilot study suggest that these chil- 4. World Health Organization (WHO). International Statis-
dren and adolescent patients with moderate to se- tical Classification of Diseases and Related Health Problems,
vere ADHD symptoms who were unresponsive to or Tenth Revision. 2nd ed. Geneva, Switzerland: WHO;
had experienced AEs during treatment with psycho- 2004.
stimulants and who wished to try a different agent 5. US Dept of Health and Human Services, Centers for
were successfully switched (90% completion rate) Disease Control and Prevention, National Center for
to a nonpsychostimulant treatment using the pro- Health Statistics (NCHS). International Classification of
Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). 6th
posed transition method. Significant reduction of
ed. Hyattsville, Md: NCHS; 2007.
their ADHD symptom severity was observed, as
6. Barkley RA. Attention-Deficit Hyperactivity Disorder: A Handbook
measured by clinical and objective measures. These for Diagnosis and Treatment. New York, NY: Guilford Press;
results remain to be replicated under blinded, con- 1998.
trolled conditions. 7. Barkley RA, Guevremont DC, Anastopoulos AD, et al.
Driving-related risks and outcomes of attention deficit hy-
ACKNOWLEDGMENTS peractivity disorder in adolescents and young adults: A 3-
This study was supported by Eli Lilly and Company, to 5-year follow-up survey. Pediatrics. 1993;92:212–218.
Indianapolis, Indiana. 8. Hechtman L, Weiss G, Parlman T. Hyperactives as young
Dr. Quintana has received research support from adults: Past and current substance abuse and antisocial
Abbott Laboratories, Abbott Park, Illinois; Astra behavior. Am J Orthopsychiatry. 1984;54:415–425.
Zeneca, London, United Kingdom; Janssen Pharma- 9. Newcorn JH, Zhang S, Sutton V, et al. Differential response
to stimulants and atomoxetine. Presented at: The 51st
ceutica Inc., Titusville, New Jersey; Eli Lilly and
Annual Meeting of the American Academy of Child and
Company; and Pfizer Laboratories, Groton, Con-
Adolescent Psychiatry; October 19–24, 2004; Washington,
necticut. He has served on advisory boards for The DC. Abstract.
Osler Institute, Terre Haute, Indiana, and The Lexicor 10. Ahmann PA, Theye FW, Berg R, et al. Placebo-controlled
Medical Technology Corporation, Augusta, Georgia. evaluation of amphetamine mixture–dextroamphetamine
Dr. Cherlin has received research support from salts and amphetamine salts (Adderall): Efficacy rate and
AstraZeneca; Cephalon Inc., Frazer, Pennsylvania; side effects. Pediatrics. 2001;107:E10.
GlaxoSmithKline, Research Triangle Park, North 11. Efron D, Jarman F, Barker M. Side effects of methyl-
Carolina; Eli Lilly and Company; Merck & Co. Inc., phenidate and dexamphetamine in children with atten-
Rahway, New Jersey; New River Pharmaceuticals, tion deficit hyperactivity disorder: A double-blind, cross-
Radford, Virginia; and Shire, Hampshire, United over trial. Pediatrics. 1997;100:662–666.
12. Greenhill LL, Halperin JM, Abikoff H. Stimulant medica-
Kingdom. He is on the advisory board for Eli Lilly
tions. J Am Acad Child Adolesc Psychiatry. 1999;38:503–512.
and Company. Dr. Duesenberg has received research
13. Spencer TJ, Biederman J, Wilens TE, et al. Pharmaco-
support from Abbott Laboratories; Bristol-Myers therapy of attention-deficit hyperactivity disorder across
Squibb Company, New York, New York; Johnson & the life cycle. J Am Acad Child Adolesc Psychiatry. 1996;35:
Johnson, New Brunswick, New Jersey; Eli Lilly and 409–432.
Company; Organon, Roseland, New Jersey; and 14. Buitelaar JK, Danckaerts M, Gillberg C, et al, for the
Shire. He has served on speakers’ bureaus and is on Stomoxetine International Study Group. A prospective,
advisory boards for Eli Lilly and Company and Shire. multi-center, open-label assessment of atomoxetine in


1168_quintana 6/29/07 9:59 AM Page 1177

H. Quintana et al.

non-North American children and Checklists, Norms, and Clinical Inter- dextroamphetamine, and pemoline.
adolescents with ADHD. Eur Child pretations. New York, NY: The Guilford Pediatrics. 1990;86:226–237.
Adolesc Psychiatry. 2004;13:249–257. Press; 1998. 31. Spencer T, Wilens T, Biederman J, et
15. Kelsey DK, Sumner CR, Casat CD, 23. Guy W. ECDEU Assessment Manual for al. A double-blind, crossover com-
et al. Once-daily atomoxetine treat- Psychopharmacology, Revised. Publica- parison of methylphenidate and
ment for children with attention- tion ADM 76-338. Bethesda, Md: placebo in adults with childhood-
deficit/hyperactivity disorder, includ- United States Department of Health, onset attention-deficit hyperactivity
ing an assessment of evening and Education, and Welfare; 1976. disorder. Arch Gen Psychiatry. 1995;
morning behavior: A double-blind, 24. Gillings D, Koch G. The application 52:434–443.
placebo-controlled trial. Pediatrics. of the principle of intention-to-treat 32. Spencer T, Biederman J, Wilens T, et
2004;114:e1–e8. to the analysis of clinical trials. Drug al. Nortriptyline treatment of chil-
16. Ghuman JK, Ginsburg GS, Subra- Info J. 1991;25:411–424. dren with attention-deficit hyperac-
maniam G, et al. Psychostimulants 25. Efron D, Jarman F, Barker M. Methyl- tivity disorder and tic disorder or
in preschool children with attention- phenidate versus dextroampheta- Tourette’s syndrome. J Am Acad Child
deficit/hyperactivity disorder: Clini- mine in children with attention defi- Adolesc Psychiatry. 1993;32:205–210.
cal evidence from a developmental cit hyperactivity disorder: A double- 33. Spencer T, Biederman J, Kerman K,
disorders institution. J Am Acad Child blind, crossover trial. Pediatrics. 1997; et al. Desipramine treatment of chil-
Adolesc Psychiatry. 2001;40:516–524. 100:E6. dren with attention-deficit hyper-
17. Schachter HM, Pham B, King J, et al. 26. Pelham WE, Aronoff HR, Midlam activity disorder and tic disorder or
How efficacious and safe is short- JK, et al. A comparison of Ritalin Tourette’s syndrome. J Am Acad Child
acting methylphenidate for the and Adderall: Efficacy and time- Adolesc Psychiatry. 1993;32:354–360.
treatment of attention-deficit disor- course in children with attention- 34. Wilens TE, Biederman J, Geist DE, et
der in children and adolescents? A deficit/hyperactivity disorder. Pedi- al. Nortriptyline in the treatment of
meta-analysis. CMAJ. 2001;165: atrics. 1999;103:E43. ADHD: A chart review of 58 cases.
1475–1488. 27. Barrickman LL, Perry PJ, Allen AJ, et J Am Acad Child Adolesc Psychiatry.
18. American Psychiatric Association. al. Bupropion versus methylpheni- 1993;32:343–349.
Diagnostic and Statistical Manual of date in the treatment of attention- 35. Perwien A, Hall J, Swensen A,
Mental Disorders: DSM-IV. 4th Edition. deficit hyperactivity disorder. J Am Swindle R. Stimulant treatment pat-
Washington, DC: American Psy- Acad Child Adolesc Psychiatry. 1995;34: terns and compliance in children and
chiatric Association; 1994. 649–657. adults with newly treated attention-
19. Kaufman J, Birmaher B, Brent D, et 28. Taylor FB, Russo J. Comparing deficit/hyperactivity disorder. J Manag
al. Schedule for Affective Disorders guanfacine and dextroamphetamine Care Pharm. 2004;10:122–129.
and Schizophrenia for School-Age for the treatment of adult attention- 36. Newcorn JH, Allen AJ, Sutton V, et
Children–Present and Lifetime ver- deficit/hyperactivity disorder. J Clin al. Characteristics of placebo re-
sion (K-SADS-PL): Initial reliability Psychopharmacol. 2001;21:223–228. sponders and remitters in pediatric
and validity data. J Am Acad Child 29. Taylor FB, Russo J. Efficacy of ADHD clinical trials. Presented at:
Adolesc Psychiatry. 1997;36:980–988. modafinil compared with dextroam- The 50th Annual Meeting of the
20. World Medical Association. Dec- phetamine for the treatment of at- American Academy of Child and
laration of Helsinki: Recommenda- tention deficit/hyperactivity disorder Adolescent Psychiatry; October 14–
tions Guiding Medical Doctors in in adults. J Child Adolesc Psychophar- 19, 2003; Miami Beach, Fla. Abstract.
Biomedical Research Involving macol. 2000;10:311–320. 37. Mrazek DA, Black J, O’Kane D.
Human Subjects. As adopted by the 30. Pelham WE Jr, Greenslade KE, Vodde- Pharmacogenomic screening for de-
18th General Assembly, Helsinki, Hamilton M, et al. Relative efficacy of pressed children and adolescents.
Finland, June 1964, and amended long-acting stimulants on children Presented at: The 50th Annual
by the 52nd General Assembly, with attention deficit-hyperactivity Meeting of the American Academy
Edinburgh, Scotland, October 2000. disorder: A comparison of standard of Child and Adolescent Psychiatry;
21. Farid NA, Bergstrom RF, Ziege EA, methylphenidate, sustained-release October 14–19, 2003; Miami Beach,
et al. Single-dose and steady-state methylphenidate, sustained-release Fla. Abstract.
pharmacokinetics of tomoxetine in
normal subjects. J Clin Pharmacol.
1985;25:296–301. Address correspondence to: Humberto Quintana, MD, Department of
22. DuPaul GJ, Power TJ, Anastopoulos Psychiatry, Louisiana State University Health Sciences Center, 1542 Tulane
AD, et al. ADHD Rating Scale-IV: Avenue, Room 317, New Orleans, LA 70112. E-mail: hquint@lsuhsc.edu

June 2007 1177

Transition from methylphenidate or amphetamine to atomoxetine in children and adolescents

Recommended

Recommended

More Related Content

What's hot

What's hot (19)

Viewers also liked

Viewers also liked (6)

Similar to Transition from methylphenidate or amphetamine to atomoxetine in children and adolescents

Similar to Transition from methylphenidate or amphetamine to atomoxetine in children and adolescents (20)

More from hospital higueras

More from hospital higueras (20)

Transition from methylphenidate or amphetamine to atomoxetine in children and adolescents