Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Pediatr Drugs 2009; 11 (3): 203-226

ADIS DRUG EVALUATION 1174-5878/09/0003-0203/$49.95/0

ª 2009 Adis Data Information BV. All rights reserved.

Atomoxetine
A Review of its Use in Attention-Deficit Hyperactivity Disorder in
Children and Adolescents
Karly P. Garnock-Jones and Gillian M. Keating
Wolters Kluwer Health | Adis, Auckland, New Zealand, an editorial office of Wolters Kluwer Health, Philadelphia,
Pennsylvania, USA

Various sections of the manuscript reviewed by:


J. Graham, Department of Psychiatry, University of Dundee, Dundee, UK; D.E. Greydanus, Department of Pediatrics and Human
Development, Michigan State University, Kalamazoo, Michigan, USA; F. Levy, School of Psychiatry, University of New South Wales, Sydney,
New South Wales, Australia.

Data Selection
Sources: Medical literature published in any language since 1980 on ‘atomoxetine’, identified using MEDLINE and EMBASE, supplemented by AdisBase (a proprietary database
of Wolters Kluwer Health | Adis). Additional references were identified from the reference lists of published articles. Bibliographical information, including contributory unpublished
data, was also requested from the company developing the drug.
Search strategy: MEDLINE, EMBASE and AdisBase search terms were ‘atomoxetine’ and [(‘attention deficit hyperactivity disorder’ or ‘ADHD’) and (‘infants’ or ‘children’ or
‘adolescents’)]. Searches were last updated 10 February 2009.
Selection: Studies in pediatric patients with attention-deficit hyperactivity disorder who received atomoxetine. Inclusion of studies was based mainly on the methods section of
the trials. When available, large, well controlled trials with appropriate statistical methodology were preferred. Relevant pharmacodynamic and pharmacokinetic data are also
included.
Index terms: Atomoxetine, attention-deficit hyperactivity disorder, ADHD, children, adolescents, pharmacoeconomics, pharmacodynamics, pharmacokinetics, therapeutic
use, tolerability.

Contents

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
2. Pharmacodynamic Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
2.1 Effects on Neurotransmitter Transporters and Receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
2.2 Other Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3. Pharmacokinetic Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
3.1 Absorption and Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
3.2 Metabolism and Elimination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
3.3 Special Populations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
3.4 Drug Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
4. Therapeutic Efficacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
4.1 Comparisons with Placebo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
4.1.1 Short-Term Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
4.1.2 Longer Term Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
4.1.3 In Stimulant-Naive Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
4.2 Comparisons with Stimulants or Standard Current Therapy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
4.3 In Patients with Co-Morbid Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
5. Tolerability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
5.1 Specific Adverse Events . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
6. Pharmacoeconomic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
204 Garnock-Jones & Keating

7. Dosage and Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220


8. Place of Atomoxetine in the Management of Attention-Deficit Hyperactivity Disorder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Summary
Abstract Atomoxetine (Strattera) is a selective norepinephrine (noradrenaline) reuptake inhibitor that is not clas-
sified as a stimulant, and is indicated for use in patients with attention-deficit hyperactivity disorder
(ADHD).
Atomoxetine is effective and generally well tolerated. It is significantly more effective than placebo and
standard current therapy and does not differ significantly from or is noninferior to immediate-release
methylphenidate; however, it is significantly less effective than the extended-release methylphenidate for-
mulation OROS methylphenidate (hereafter referred to as osmotically released methylphenidate) and
extended-release mixed amfetamine salts.
Atomoxetine can be administered either as a single daily dose or split into two evenly divided doses, has a
negligible risk of abuse or misuse, and is not a controlled substance in the US. Atomoxetine is particularly
useful for patients at risk of substance abuse, as well as those who have co-morbid anxiety or tics, or who do
not wish to take a controlled substance. Thus, atomoxetine is a useful option in the treatment of ADHD in
children and adolescents.
Pharmacologic The mechanism of action of atomoxetine is unclear, but is thought to be related to its selective inhibition of
Properties presynaptic norepinephrine reuptake in the prefrontal cortex. Atomoxetine has a high affinity and selectivity
for norepinephrine transporters, but little or no affinity for various neurotransmitter receptors. Atomox-
etine has a demonstrated ability to selectively inhibit norepinephrine uptake in humans and animals, and
studies have shown that it preferentially binds to areas of known high distribution of noradrenergic neurons,
such as the fronto-cortical subsystem.
Atomoxetine was generally associated with statistically, but not clinically, significant increases in both
heart rate and blood pressure in pediatric patients with ADHD. While there was an initial loss in expected
height and weight among atomoxetine recipients, this eventually returned to normal in the longer term. Data
suggest that atomoxetine is unlikely to have any abuse potential. Atomoxetine appeared less likely than
methylphenidate to exacerbate disordered sleep in pediatric patients with ADHD.
Atomoxetine is rapidly absorbed, and demonstrates dose-proportional increases in plasma exposure. It
undergoes extensive biotransformation, which is affected by poor metabolism by cytochrome P450 (CYP)
2D6 in a small percentage of the population; these patients have greater exposure to and slower elimination
of atomoxetine than extensive metabolizers.
Patients with hepatic insufficiency show an increase in atomoxetine exposure. CYP2D6 inhibitors, such as
paroxetine, are associated with changes in atomoxetine pharmacokinetics similar to those observed among
poor CYP2D6 metabolizers.
Therapeutic Efficacy Once- or twice-daily atomoxetine was effective in the short-term treatment of ADHD in children and
adolescents, as observed in several well designed placebo-controlled trials. Atomoxetine also demonstrated
efficacy in the longer term treatment of these patients. A single morning dose was shown to be effective into
the evening, and discontinuation of atomoxetine was not associated with symptom rebound. Atomoxetine
efficacy did not appear to differ between children and adolescents. Stimulant-naive patients also responded
well to atomoxetine treatment.
Atomoxetine did not differ significantly from or was noninferior to immediate-release methylphenidate in
children and adolescents with ADHD with regard to efficacy, and was significantly more effective than
standard current therapy (any combination of medicines [excluding atomoxetine] and/or behavioral
counseling, or no treatment). However, atomoxetine was significantly less effective than osmotically re-
leased methylphenidate and extended-release mixed amfetamine salts.
The efficacy of atomoxetine did not appear to be affected by the presence of co-morbid disorders, and
symptoms of the co-morbid disorders were not affected or were improved by atomoxetine administration.

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 205

Health-related quality of life (HR-QOL) appeared to be positively affected by atomoxetine in both short- and
long-term studies; atomoxetine also improved HR-QOL to a greater extent than standard current therapy.
Tolerability Atomoxetine was generally well tolerated in children and adolescents with ADHD. Common adverse events
included headache, abdominal pain, decreased appetite, vomiting, somnolence, and nausea. The majority of
adverse events were mild or moderate; there was a very low incidence of serious adverse events. Few patients
discontinued atomoxetine treatment because of adverse events. Atomoxetine discontinuation appeared to
be well tolerated, with a low incidence of discontinuation-emergent adverse events. Atomoxetine appeared
better tolerated among extensive CYP2D6 metabolizers than among poor metabolizers.
Slight differences were evident in the adverse event profiles of atomoxetine and stimulants, both im-
mediate- and extended-release. Somnolence appeared more common among atomoxetine recipients and
insomnia appeared more common among stimulant recipients.
A black-box warning for suicidal ideation has been published in the US prescribing information, based on
findings from a meta-analysis showing that atomoxetine is associated with a significantly higher incidence of
suicidal ideation than placebo. Rarely, atomoxetine may also be associated with serious liver injury; post-
marketing data show that three patients have had liver-related adverse events deemed probably related to
atomoxetine treatment.
Pharmacoeconomic Treatment algorithms involving the initial use of atomoxetine appear cost effective versus algorithms in-
Evaluation: volving initial methylphenidate (immediate- or extended-release), dexamfetamine, tricyclic antidepressants,
or no treatment in stimulant-naive, -failed, and -contraindicated children and adolescents with ADHD. The
incremental cost per quality-adjusted life-year is below commonly accepted cost-effectiveness thresholds, as
shown in several Markov model analyses conducted from the perspective of various European countries,
with a time horizon of 1 year.

1. Introduction factors.[2,4,6] Dopamine and norepinephrine (noradrenaline)


abnormalities are believed to be associated with ADHD, as are
Attention-deficit hyperactivity disorder (ADHD) is a neuro- alterations in regional cerebral volumes and reduced metabo-
behavioral disorder, characterized by one or both of its subtypes lism in the prefrontal cortex and striatal regions.[2,4,6]
(inattention and hyperactivity/impulsiveness).[1,2] Symptoms may
It has been demonstrated that ADHD patients use sig-
manifest as early as 3 years of age; however, most diagnoses occur
nificantly more health services than children without ADHD,
when the patient is aged 7–10 years.[3] Approximately half of
both before and after diagnosis.[7] The total financial burden of
childhood ADHD patients show symptomatic features continuing
ADHD to patients and their families in the year 2000 in the US
into adulthood.[1] ADHD has a prevalence of 3–9% in children and
was estimated to be $US31.6 billion.[2]
adolescents in the US,[1] and 4–8% worldwide.[4] European rates
appear lower than US rates; however, this has been described as an Currently, where pharmacologic treatment is deemed ap-
effect of using diagnostic criteria based on the International propriate, recommended ADHD treatments include methyl-
Classification of Diseases (ICD), as opposed to Diagnostic and phenidate, dexamfetamine, and atomoxetine (Strattera).[8]
Statistical Manual of Mental Disorders, 4th Edition (DSM-IV)[5] Lisdexamfetamine is also approved for the treatment of
criteria.[1] Male sex, low socioeconomic status, and young age are ADHD.[9] Of these, methylphenidate, dexamfetamine, and lis-
all associated with a higher prevalence of ADHD.[6] dexamfetamine are all CNS stimulants.[8,9] There are several
Co-morbid psychiatric disorders are common among patients concerns with stimulant use, including the risk of growth re-
with ADHD, including oppositional defiant disorder (ODD), tardation, development of tics, and sudden cardiac death, as
conduct disorders, mood disorders, and anxiety disorders.[2,6] well as a potential for abuse or misuse.[2]
More than two-thirds of children with ADHD have a co-morbid Atomoxetine is an orally administered selective norepinephrine
condition.[2] reuptake inhibitor that is approved for the treatment of ADHD in
The etiology and pathophysiology of ADHD are, thus far, various countries including the US[10] and the UK.[11] It is not
unknown; however, genetic and developmental factors have been classified as a stimulant, and is not a controlled substance in the
implicated, as have, to a smaller degree, environmental and social US. This article reviews the pharmacologic properties and clinical

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
206 Garnock-Jones & Keating

profile of oral atomoxetine in children and adolescents with dopamine levels in the nucleus accumbens and striatum remained
ADHD. constant. Methylphenidate significantly (p < 0.05) increased extra-
cellular norepinephrine and dopamine levels in the prefrontal cor-
2. Pharmacodynamic Properties tex, and dopamine levels in the nucleus accumbens and striatum,
indicating a potential difference from atomoxetine in mechanism
Atomoxetine is a (-) isomer of an ortho-methylphenoxy ana- of action.[16] It was hypothesized that the inhibition of dopamine
log of nisoxetine, and is a derivative of phenoxypropylamine.[12] uptake in the prefrontal cortex, but not in the dopamine trans-
Its mechanism of action in the treatment of ADHD is unclear, porter-rich nucleus accumbens and striatum, was due to baseline
but is thought to be related to its selective inhibition of pre- nonselective dopamine uptake by norepinephrine transporters in
synaptic norepinephrine reuptake in the prefrontal cortex, re- the prefrontal cortex.[16] The absence of extracellular dopamine
sulting in increased noradrenergic transmission, important for accumulation in the nucleus accumbens and striatum suggests that
attention, learning, memory, and adaptive response.[13,14] atomoxetine is unlikely to produce tics or have abuse potential.[16]
Atomoxetine also selectively inhibits norepinephrine uptake
2.1 Effects on Neurotransmitter Transporters in humans.[20] Four healthy male volunteers received atomox-
and Receptors etine 20 mg twice daily for 1 week, following a week of placebo
administration. When an infusion of norepinephrine was
Atomoxetine has a high affinity and selectivity for norepine-
administered, the mean pressor response for atomoxetine
phrine transporters, as demonstrated in radioligand binding stu-
versus placebo was 12.4 versus 5.2 mmHg per microgram of
dies in rat brain synaptosomes,[12,15] as well as in clonal cell lines
norepinephrine per minute (p = 0.054).[20] The increase with
transfected with human neurotransmitter transporters.[16] The af-
atomoxetine on day 1 was 261% greater than that observed with
finity constant (Ki) values for atomoxetine inhibition of nor-
placebo.[20] There was a significant correlation between the
epinephrine, serotonin, and dopamine transporters (in MDCK
pressor response to norepinephrine and the plasma con-
and HEK 293 cells) were 5, 77, and 1451 nmol/L.[16] Atomox-
centration of atomoxetine (p = 0.002).[20] In contrast, when an
etine was associated with a much lower affinity (Ki >1 mmol/L)
infusion of tyramine was administered, the mean pressor re-
for choline, GABA, and adenosine transporters,[16] as well as
sponse for atomoxetine versus placebo treatment was 4.5 versus
many other neurotransmitter receptors, ion channels, second
7.9 mmHg per milligram of tyramine per minute (p = 0.003),
messengers, and brain/gut peptides.[16]
and the increase with atomoxetine on day 1 was 70% of that
In vitro radioligand binding studies of atomoxetine in the
observed with placebo.[20] Serotonin uptake into platelets was
human brain have demonstrated that it has little or no affinity
not affected by atomoxetine administration in this study.[20]
for various neurotransmitter receptors.[17] Ki values for ato-
A randomized, double-blind, crossover study using tran-
moxetine binding to muscarinic, a-adrenergic, histamine H1,
scranial magnetic stimulation in nine healthy volunteers de-
and serotonergic (5-HT1A and 5-HT2) receptors were
monstrated that both atomoxetine 60 mg and methylphenidate
940–10 900 nmol/L; for atomoxetine binding to dopamine D2
30 mg significantly (p < 0.05) decreased cortical inhibition and
receptors the Ki was >35 000 nmol/L.[17]
increased cortical facilitation to extents that did not sig-
When norepinephrine and serotonin depletion was induced
nificantly differ between treatments, indicating the possibility
in the rat brain in vivo, using the transporter-specific neurotox-
of a shared cortical target for ADHD treatment.[21]
ins p-CA and DSP-4, atomoxetine was shown to inhibit the deple-
tion of norepinephrine, but not serotonin, in a dose-dependent
manner.[16] Methylphenidate did not inhibit either transporter, 2.2 Other Effects
leading to normal depletion of both neurotransmitters.[16]
Localization studies, using quantitative autoradiography in Atomoxetine was associated with modest increases in heart
the rat brain, suggest that atomoxetine preferentially binds to rate and BP in children and adolescents with ADHD, according
areas of known high distribution of noradrenergic neurons, to the results of a pooled analysis of clinical trial data.[22] In the
such as the fronto-subcortical system, which controls attention short term (up to 9 weeks’ therapy), significantly greater increases
and motor behavior.[18,19] in mean heart rate (+7.8 vs +1.5 beats/minute [bpm]; p < 0.001)
In rats, intraperitoneal atomoxetine significantly (p < 0.025) and mean diastolic BP (DBP; +2.1 vs -0.5 mmHg; p = 0.002) were
increased extracellular norepinephrine and dopamine levels in the seen with atomoxetine than with placebo; there was no signifi-
prefrontal cortex by up to 290% and 323% of basal levels; serotonin cant between-group difference in the change in systolic BP (SBP).
levels did not significantly differ from baseline.[16] Extracellular Patients receiving atomoxetine for ‡1 year had increases in mean

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 207

heart rate of <10 bpm and mean increases in BP that were small Sexual development did not appear to be affected by atomoxe-
and not considered to be of clinical significance.[22] The in- tine treatment in an analysis (available as an abstract) of
cidence of abnormally high heart rate or BP in atomoxetine 15 months’ treatment with atomoxetine compared with placebo in
recipients is discussed in section 5.1. Atomoxetine treatment was children and adolescents with ADHD using Tanner staging.[30]
not associated with QT prolongation in these clinical trials.[22] Children with ADHD receiving atomoxetine 1–1.8 mg/kg/day
Heart rate appears to increase to a greater extent among had a smaller increase in sleep-onset latencies than those receiving
atomoxetine recipients compared with stimulant recipients (sec- methylphenidate 0.9–1.8 mg/kg/day (12.06 vs 39.24 minutes;
tion 4.2). In a study comparing the efficacy of atomoxetine with p < 0.001), according to results from a randomized, double-blind,
that of the extended-release methylphenidate formulation OROS crossover study.[31] Other actigraphy measures demonstrated that
methylphenidate (hereafter referred to as osmotically released atomoxetine recipients worsened to a significantly lesser extent
methylphenidate), recipients of atomoxetine showed a significantly than methylphenidate recipients, with regard to total sleep interval
greater change in heart rate (+6.4 vs +3.0 bpm; p < 0.05).[23] (p = 0.004) and assumed sleep time (p = 0.016), although methyl-
Another study demonstrated a greater increase in heart rate among phenidate recipients showed an improved interrupted sleep time
atomoxetine versus immediate-release methylphenidate recipients compared with a worsened interrupted sleep time among atomoxe-
(+8.51 vs +4.76 bpm; p = 0.005).[24] Where reported in other active tine recipients (p = 0.025), as well as the number of sleep interrup-
comparator trials, atomoxetine did not differ significantly from tions worsening to a significantly (p = 0.011) greater extent among
immediate-release methylphenidate[25] or extended-release mixed atomoxetine recipients than among methylphenidate recipients.[31]
amfetamine salts[26] with regard to the increase in heart rate. When monoamine oxidase inhibitors (MAOIs) have been
Atomoxetine is unlikely to lead to abuse.[27,28] A rando- administered concomitantly with other drugs that affect brain
mized, double-blind, crossover trial in 16 healthy volunteers monoamine concentrations, reports of serious, sometimes fatal,
who were nondependent light drug users compared the effects reactions have occurred. Thus, coadministration of atomox-
of placebo, atomoxetine 20, 45, or 90 mg and methylphenidate etine and MAOIs is contraindicated.[10,11] Other drugs that
20 or 40 mg on subjective, physiologic, and psychomotor interact with atomoxetine include pressor agents (e.g. dopa-
measures.[27] Results demonstrated that atomoxetine did not mine), with subsequent effects on blood pressure, and high-
significantly differ from placebo in perceptions of stimulant dose nebulized[11] or systemically administered[10,11] b2-agonists
and euphoric effects, but methylphenidate was associated with (e.g. albuterol [salbutamol]), with possible effects on heart rate
significantly higher perceptions of these effects than placebo and blood pressure.[10,11] Atomoxetine should therefore be
(p < 0.05). Atomoxetine 90 mg was associated with ‘bad’ and coadministered with caution with pressor agents or b2-agonists.
‘sick’ feelings, differing significantly from placebo in these Atomoxetine made no difference to the intoxicating effects of
measures (p < 0.05).[27] Data from a study involving six healthy ethanol or the cardiovascular effects of methylphenidate.[10]
volunteers with a recent history of nontherapeutic stimulant The UK prescribing information states that atomoxetine
abuse who received methylphenidate 5–30 mg, atomoxetine should also be used with caution with drugs that affect nor-
15–90 mg, dexamfetamine 2.5–15 mg, triazolam 0.06–0.375 mg, epinephrine levels (such as antidepressants or certain deconge-
and placebo suggest that, while behavioral effects of atomox- stants), as there is a potential for additive or synergistic effects;
etine overlap somewhat with psychomotor stimulants, it has a caution is also advised with drugs that lower the seizure threshold
low abuse potential.[28] This conclusion is potentially supported (e.g. antidepressants, antipsychotics), as there is a potential risk of
by the lack of extracellular increase of dopamine in the nucleus seizures with atomoxetine.[11] When atomoxetine is administered
accumbens and striatum (section 2.1).[16] with QT-prolonging drugs (e.g. antipsychotics or class IA and III
Atomoxetine recipients demonstrated an initial loss in both antiarrhythmics), drugs that cause an imbalance in electrolytes
expected weight and height, although these shortfalls peaked at (e.g. thiazide diuretics), or drugs that inhibit cytochrome P450
15 and 18 months, respectively, and returned to expected measure- (CYP) 2D6, there is a potential for an increased risk of QT pro-
ments by 36 and 24 months, according to the interim results of a longation, according to the UK prescribing information.[11] These
long-term, open-label extension study.[29] Persistent decreases from precautions are not specified in the US prescribing information.[10]
the expected measurements appeared to occur in patients who
were taller or heavier than average before treatment.[29] The 3. Pharmacokinetic Properties
study involved pediatric patients (n = 1312; 5-year data n = 61)
who were enrolled in one of 13 clinical atomoxetine trials and This section focuses, where possible, on results from an open-
who subsequently received long-term atomoxetine treatment. label pharmacokinetic study conducted in children and adolescents

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
208 Garnock-Jones & Keating

aged 7–14 years with a DSM-IV diagnosis of ADHD who adolescents; however, atomoxetine may be administered with or
received either a single 10 mg dose of oral atomoxetine (n = 7) without food.[10] The steady-state volume of distribution
or atomoxetine 20–45 mg twice daily for 11 weeks (n = 16) (2.25 L/kg[32]) indicates that atomoxetine primarily distributes
[dosage was not adjusted for weight].[32] Most doses were not into total body water.[10] Plasma protein binding (mainly to al-
administered within 1 hour of a meal. Although it was not a bumin) of atomoxetine was 98% at therapeutic concentrations.[10]
requirement, all patients were extensive metabolizers of
CYP2D6 substrates (extensive metabolizers). The vast majority 3.2 Metabolism and Elimination
of the general population are extensive metabolizers, with 7% of
Atomoxetine undergoes extensive biotransformation.[10] In
Caucasians and 2% of African-Americans being poor meta-
extensive metabolizers, it is mainly metabolized via the CYP2D6
bolizers of CYP2D6 substrates (poor metabolizers);[10] some
enzymatic pathway, in which atomoxetine is oxidated to form
pharmacokinetic data are available for pediatric poor meta-
4-hydroxyatomoxetine (the major metabolite), which is then
bolizers.[33] Dosage adjustments are recommended in the poor
glucuronidated to form 4-hydroxyatomoxetine-O-glucuronide,
metabolizer population (section 7).
the main excreted metabolite (accounting for >80% and <17% of
Atomoxetine pharmacokinetics have been demonstrated to
the total dose in urine and feces); <3% of the total atomoxetine
be similar for adults and children or adolescents, once weight is
dose is excreted as unchanged drug.[10,34] Poor metabolizers are
adjusted for.[32] Therefore, some data presented in this section
unable to metabolize as efficiently using the CYP2D6 pathway;
are from studies in adult subjects, when no pediatric data are
metabolism in these individuals occurs mainly via the CYP2C19
available. These data were obtained from a review article[34] and
pathway, forming N-desmethylatomoxetine.[10,34] Several other
from the US prescribing information.[10]
CYP isoforms are able to form 4-hydroxyatomoxetine and
While there are few published pediatric data for poor me-
N-desmethylatomoxetine; thus, 4-hydroxyatomoxetine is still the
tabolizers, it has been demonstrated in adult poor metabolizers
most common metabolite, even in poor metabolizers.[10,34]
that values for the area under the plasma concentration-time
4-Hydroxyatomoxetine has similar pharmacologic activity to
curve (AUC) and maximum plasma concentration (Cmax) are
atomoxetine (although, unlike atomoxetine, it does have rela-
10- and 50-fold higher than in extensive metabolizers, and that
tively high affinity for the human serotonin transporter),[34] but
atomoxetine elimination is slower, with a plasma half-life (t1=2 )
circulates in plasma at lower concentrations (e.g. at 1% of ato-
of »24 hours.[10]
moxetine concentrations in extensive metabolizers).[10] Con-
A population pharmacokinetic, one-compartment model
versely, N-desmethylatomoxetine has much less pharmacologic
was constructed, using data from five studies in pediatric pa-
activity than atomoxetine and 4-hydroxyatomoxetine,[34] but also
tients receiving dosages of atomoxetine 10–90 mg/day, ad-
circulates at lower concentrations.[10]
ministered twice daily.[33] This model demonstrated that drug
Atomoxetine has a short half-life (just over 3 hours); this
clearance among pediatric poor metabolizers is 9-fold lower
explains the low amount of atomoxetine accumulation (mean
than that among extensive metabolizers.
9% in pediatric patients) observed at steady state.[32]
Time-invariant pharmacokinetics were indicated by the pe-
3.1 Absorption and Distribution diatric study; t1=2 , apparent clearance, and apparent volume of
distribution were all similar after a single dose (3.12 hours,
Oral atomoxetine is rapidly absorbed; Cmax (144 ng/mL) was
0.455 L/h/kg, and 1.96 L/kg, respectively) and at steady state
reached in 2 hours among children and adolescents with
(3.28 hours, 0.477 L/h/kg, and 2.25 L/kg).[32]
ADHD (all of whom were extensive metabolizers) receiving a


single dose of atomoxetine 10 mg.[32] The AUC from time zero 3.3 Special Populations
to infinity (AUC1) was 645 ng h/mL.
Corresponding data for children and adolescents with ADHD Extensive metabolizers with moderate (Child-Pugh class B)


receiving multiple doses of 40–90 mg/day are 537 ng/mL (Cmax at or severe (Child-Pugh class C) hepatic insufficiency have in-
steady state), 1.73 hours, and 2250 ng h/mL.[32] creased atomoxetine exposure compared with healthy volun-
Adult studies demonstrated an absolute oral bioavailability of teers.[10] Therefore, dosage reduction is recommended in these
»63% in extensive metabolizers and »94% in poor metabolizers.[10] patients (section 7).[10,11] Following a single dose of atomox-
Atomoxetine demonstrated dose-proportional increases in plas- etine 20 mg, AUC1 was significantly higher in adults with


ma exposure.[32] When administered with food, atomoxetine was moderate (n = 6) or severe (n = 4) hepatic impairment than in
absorbed at a slower rate, with a 9% lower Cmax in children and healthy volunteers (1.59 vs 0.85 mg h/mL; p < 0.05).[35]

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 209

Extensive metabolizers with end-stage renal disease showed no duration investigated the efficacy of atomoxetine versus placebo
significant difference from healthy volunteers when exposure was in a stimulant-naive population.
corrected for dosage.[10] Atomoxetine pharmacokinetics were not
influenced by sex or ethnic origin (other than Caucasians having a 4.1.1 Short-Term Treatment
higher likelihood of being poor metabolizers).[10] Patients in the short-term trials were randomized to receive
once-[38,40-43] or twice-daily[37,39] atomoxetine or placebo. One fixed-
3.4 Drug Interactions
dose trial titrated patients to a target dosage of atomoxetine 0.5, 1.2
Atomoxetine was not associated with clinically important in- or 1.8 mg/kg/day.[37] Three other trials titrated patients to a target
hibition or induction of CYP isoenzymes, including CYP1A2, dosage of atomoxetine 1.0[38] or 1.2[42,43] mg/kg/day, although the
CYP3A, CYP2D6, and CYP2C9.[10] No dosage adjustment is dosage could be further increased to 1.5[38] or 1.8[42,43] mg/kg/day if
considered necessary for drugs metabolized by CYP3A or required. The remaining four trials titrated patients according to
CYP2D6.[10] therapeutic response; permitted atomoxetine dosages were
Healthy extensive metabolizers receiving both atomoxetine 0.8–1.8,[41] £1.8,[40] or £2[39] mg/kg/day. Where specified, final mean
20 mg twice daily and paroxetine 20 mg once daily, a potent atomoxetine dosages were 1.3,[38,41-43] 1.4,[40] and 1.5[39] mg/kg/day.
CYP2D6 inhibitor, demonstrated pharmacokinetic parameters Eligible patients aged 6–18 years in the short-term trials had a
for atomoxetine that were similar to those seen among poor meta- DSM-IV diagnosis of ADHD, as confirmed by the Kiddie Sched-
bolizers.[36] Coadministration of paroxetine with atomoxetine ule for Affective Disorders and Schizophrenia for School-Age
was associated with 3.5-, 6.5-, and 2.5-fold increases in atomox- Children-Present and Lifetime version (K-SADS-PL),[37,38,41-43]
etine steady-state Cmax, AUC from time 1 to 12 hours, and t1=2 , the K-SADS-Epidemiologic version (K-SADS-E; Chinese ver-
respectively.[36] Dosage adjustment is recommended for pediatric sion),[40] or an unspecified version of K-SADS.[39] Most patients
patients receiving potent CYP2D6 inhibitors (see section 7).[10] In were required to have a minimum ADHD-RS total score of 25 for
vitro studies suggest that CYP2D6 inhibitors have no effect on boys and 22 for girls (or 12 for the inattentive or hyperactive/
atomoxetine pharmacokinetics among poor metabolizers.[10] impulsive score),[40] or ADHD-RS total or inattentive or hyper-
Atomoxetine had no effect on the binding of warfarin, aspirin active/impulsive subscale scores ‡1[42] or 1.5[37-39,43] SD above age
(acetylsalicylic acid), phenytoin, or diazepam to human albumin, and sex norms. Exclusion criteria included below-average intelli-
or vice versa, according to results from in vitro drug-displacement gence levels,[37,39-42] use of other psychotropic medication,[37-43]
studies.[10] Gastric pH-elevating drugs (e.g. antacids, omeprazole) weight <25[39,41] or <20[40] kg or >60[40] kg, or a history of or cur-
had no effect on the bioavailability of atomoxetine.[10] rent psychosis,[37-41,43] bipolar disorder,[37-41,43] or serious medical
illness.[37,38,40,42,43] Two trials also excluded patients who were
4. Therapeutic Efficacy poor metabolizers.[39]
The primary endpoint for all short-term placebo-controlled
The focus of this section is on data from large (n >100), fully trials was the change in the investigator-administered ADHD-RS
published, randomized, controlled trials investigating the efficacy total score (parent[37-40,43] or teacher[41,42] version) from baseline
of atomoxetine in children and adolescents with ADHD. It to endpoint.
should be noted that the vast majority of atomoxetine clinical Additional endpoints included scores on the inattentive and
trials included dosages that were potentially higher than the hyperactive/impulsive subscales of the ADHD-RS,[37-40,42,43] res-
maximum approved dosage of 1.4 mg/kg/day or 100 mg/day. For ponse rate (see table II for definitions of response),[38,39,41-43] and
definitions of some of the rating scale abbreviations and de- CPRS,[37-42] CTRS,[38,40] CGI-ADHD-S,[39,40,43] and CGI-S[37,38,42]
scriptions of rating scales referred to in this section, see table I. scores; as well as health-related quality of life (HR-QOL), rated on
4.1 Comparisons with Placebo the CHQ psychosocial summary score.[37,41] Two studies[38,43]
specifically investigated morning and evening efficacy using
The efficacy of oral atomoxetine was compared with that of the DPREMB, both original[38] and revised (DPREMB-R)[43]
placebo in the treatment of children and adolescents with ADHD versions.
in eight 6- to 9-week, randomized, double-blind, multicenter, The majority of patients in treatment groups were diagnosed
short-term, fully published trials;[37-43,58] one multinational with the combined subtype (55–81%); inattentive and hyperactive/
study with two randomized, double-blind, placebo-controlled impulsive subtype proportions ranged from 19% to 41% and from
phases has investigated longer term results.[46,47] Two rando- 0% to 4%, respectively.[37-43] The mean patient age ranged from
mized, double-blind, multicenter trials of 10[59] and 12[60] weeks’ 9.1 to 11.5 years.[37-43] The majority of patients were male

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
210 Garnock-Jones & Keating

Table I. Definition of efficacy rating scale abbreviations and description of rating scales used in clinical studies[23-26,37-57]
Rating scale Range Description
ADHD-Rating Scale (ADHD-RS) 0–54 18-item rating scale, each item corresponding to a symptom contained in the DSM-IV
ADHD diagnosis. Each item is scored from 0 (never or rarely) to 3 (very often). It
includes the subscales ‘inattentive’ and ‘hyperactive/impulsive’. Parent and teacher
versions are both available
Child Health and Illness Profile-Child, NA 76-item parent- or patient-rated quality-of-life rating scale. The total score is the mean
Adolescent or Parental Edition (CHIP-CE, score of the five domains (satisfaction, comfort, resilience, risk avoidance, and
CHIP-AE or CHIP-PRF) achievement), which is then standardized to a t-score (a mean – SD of 50 – 10, based
on norms of a sample of US children). A higher score implies a higher quality of life
Children’s Health Questionnaire (CHQ) 0–100 50-item parent-rated quality-of-life rating scale, measuring 14 physical and
psychosocial concepts. A higher score implies a higher quality of life
Clinical Global Impressions-Improvement 1–7 7-point scale for rating the improvement of mental illness, taking into account the total
scale (CGI-I) clinical experience. Rating is from 1 (very much improved), through to 4 (no change), to
7 (very much worsened)
Clinical Global Impressions-ADHD-Severity 1–7 7-point scale for rating the severity of ADHD, taking into account the total clinical
scale (CGI-ADHD-S) experience. Rating is from 1 (normal) to 7 (extremely ill)
Clinical Global Impressions-Severity scale 1–7 7-point scale for rating the severity of mental illness, taking into account the total clinical
(CGI-S) experience. Rating is from 1 (normal) to 7 (extremely ill)
Conners’ Parent/Teacher Rating Scale 0–3 per item Parent- or teacher-rated ADHD rating scale with direct links to the DSM-IV. Each item is
(CPRS/CTRS) scored from 0 (never) to 3 (very often). Studies varied in the number of items included in
their analyses
Daily Parent Ratings of Evening and 0–52 13-item parent-completed questionnaire, examining behavior in the morning and
Morning Behavior scale (DPREMB) evening. Each item is scored from 0 (not present) to 4 (extremely problematic)
Daily Parent Ratings of Evening and 0–33 11-item parent-completed questionnaire, examining behavior in the morning (3 items)
Morning Behavior-Revised scale and evening (8 items). Revised version of the DPREMB. Each item is scored from
(DPREMB-R) 0 (no difficulty) to 3 (a lot of difficulty)
Pediatric Quality of Life Inventory (PedsQL) 0–100 4-subscale quality-of-life rating scale. A higher score implies a higher quality of life
Swanson, Kotkin, Agler, M-Flynn, and 0–6 per item 13-item rating scale representing classroom behavior (6 deportment items, 7 attention
Pelham behavioral rating scale (SKAMP) items). Each item is scored from 0 (normal) to 6 (maximum impairment). The deportment and
attention subscale scores are acquired by calculating the mean of the items in each scale
Swanson, Nolan, and Pelham Rating 0–3 per item 26-item rating scale (18 items for ADHD symptoms, 9 items for ODD symptoms). Each
Scale-Revised (SNAP-IV) item is scored from 0 (not at all) to 3 (very much). Scores are yielded in three domains:
inattention, hyperactivity/impulsivity, and oppositional
ADHD = attention-deficit hyperactivity disorder; DSM-IV = Diagnostic and Statistical Manual of Mental Disorders, 4th edition; NA = not applicable;
ODD = oppositional defiant disorder.

(70–90%),[37-43] and 48–61% had received previous stimulant placebo than atomoxetine recipients; this was shown to have no
treatment.[38,40-43] effect on the efficacy conclusions.
The most common (‡10% of patients) co-morbid disorders Once- or twice-daily atomoxetine was effective in the short-
included ODD,[37-43] learning disorders,[41,42] elimination dis- term treatment of ADHD in children and adolescents (table II).
orders,[39] and phobias.[39] The mean improvement from baseline in ADHD-RS total score
Assessments were based on the modified intention-to-treat (primary endpoint) was significantly greater among atomox-
(mITT) population[37-43] using last-observation-carried-for- etine than placebo recipients in all eight short-term trials
ward (LOCF) imputation.[37-41,43] Between-group differences (reduction of 10.3–17.3 vs 5.0–9.3; all p < 0.05)[37-43] [table II].
in baseline characteristics within each trial were not significant, In addition, results from the trials reporting inattentive and
except for in a combined analysis of two studies,[39] which hyperactive/impulsive subscale scores[37-40,42,43] demonstrated
showed a significantly (p < 0.05) higher mean Wechsler In- that these scores also improved to a significantly greater extent
telligence Scale Intelligence Quotient (WISC-IQ) score among after atomoxetine versus placebo administration (table II). Of the

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 211

six trials reporting response rates, five[38,39,41,43] reported sig- between patients receiving any of the three dosages of atomox-
nificantly higher response rates among atomoxetine than placebo etine (0.5, 1.2, and 1.8 mg/kg/day) and placebo recipients for the
recipients (59–60% vs 25–40%; all p < 0.05); the remaining trial[42] change from baseline in the CHQ psychosocial summary score
reported no significant difference (69% vs 43%) [table II]. (+4.4, +6.0, and +9.1 vs -0.9; all p < 0.05 vs placebo) [baseline
The two studies investigating HR-QOL, rated on the CHQ scores of 32.9, 35.4, 31.3, and 35.2, respectively].
psychosocial summary score, reported differing results. One Other efficacy measures, including CPRS,[37-42] CTRS,[38,40]
study[41] reported no significant difference between atomoxetine CGI-S,[38,42] and CGI-ADHD-S[39,40,43] scores, were also significantly
and placebo recipients in the change from baseline in the CHQ (p < 0.05) improved among atomoxetine versus placebo recipients.
psychosocial summary score (+7.1 vs +3.7; baseline scores of 32.5 Two studies specifically investigating evening efficacy of a
and 32.1). The other study[37] reported significant differences single morning dose of atomoxetine found that it had a positive

Table II. Efficacy of oral atomoxetine (ATO) in the short-term treatment of children and adolescents with attention-deficit hyperactivity disorder (ADHD).
Results from eight randomized, double-blind, placebo (PL)-controlled, multicenter trials in patients (pts) aged 6–18 years.[37-43] The primary endpoint in all
studies was the ADHD-Rating Scale (ADHD-RS) total score
Study Treatment Age of pts Treatmenta No. of pts Mean ADHD- Mean change from baseline in Responsec rate
duration (y) (mg/kg/day) RSb total score ADHD-RSb score (% pts)
(wk) at baseline
total inattentive hyperactive/
impulsive
Brown et al.[41] 7 8–12 ATO 0.8–1.8d 99 65.6e -10.3***e 66**
PL 51 64.4e -5.0e 36
Gau et al. [40]
6 6–16 ATO £1.8 d
69 36.7 -17.3** -8.7* -8.7***
PL 29 37.1 -9.3 -5.2 -4.1
Kelsey et al.[43] 8 6–12 ATO 1.2d 126 42.1 -16.7* -8.3* -8.5* 63***
PL 60 42.3 -7.0 -4.1 -2.9 33
Michelson et al.[37]f 8 8–18 ATO 1.2d 84 39.2 -13.6* -7.0* -6.6*
ATO 1.8d 82 39.7 -13.5* -6.8* -6.7*
PL 83 38.3 -5.8 -2.5 -3.2
*** ***
Michelson et al. [38]
6 6–16 ATO 1.0d
84 37.6 -12.8 -7.1 -5.7*** 60***
PL 83 36.7 -5.0 -2.9 -2.1 31
*** *** ***
Spencer et al. [39]g
9 7–12 ATO £2.0 d
64 41.2 -15.6 -7.5 -8.0 64***
(Study 1) PL 61 41.4 -5.5 -3.0 -2.5 25
*** *** **
Spencer et al. [39]g
9 7–12 ATO £2.0 d
63 37.8 -14.4 -7.6 -6.9 59*
(Study 2) PL 60 37.6 -5.9 -3.0 -2.9 40
Weiss et al.[42] 7 8–12 ATO 1.2d 100 38.9 -14.5*** -7.5* -7.0*** 69
PL 51 36.7 -7.2 -4.3 -3.0 43
a Treatment was administered either once daily in the morning[38,40-43] or in divided doses twice daily, in the morning and early evening.[37,39]
b Investigator-administered parent[37-40,43] or teacher[41,42] version.
c Defined as ‡25% reduction from baseline in ADHD-RS total score;[38,39,43] an endpoint t-score using chi-square analyses that was no worse than 1 SD below
age and sex norms;[41] or ‡20% reduction from baseline in ADHD-RS total score.[42]
d Of the four fixed-dose trials, one titrated patients to a target dosage of ATO 0.5, 1.2, or 1.8 mg/kg/day,[37] and three titrated patients to a target dosage of
ATO 1.0[38] or 1.2[42,43] mg/kg/day, although the dosage could be further increased to 1.5[38] or 1.8[42,43] mg/kg/day if required. The remaining four trials
titrated patients according to therapeutic response; permitted ATO dosages were 0.8–1.8,[41] £1.8,[40] or £2[39] mg/kg/day.
e The ADHD-RS total score was standardized to a t-score (a mean – SD of 50 – 10, based on norms of a sample of children) in this study.
f An additional ATO treatment group (ATO 0.5 mg/kg/day) was included, but results have not been reported as this group was only present to show any dose-
dependent effect and was not included in the primary analysis.
g Methylphenidate treatment group was included as a positive control, but the data are not reported.
*
p < 0.05, ** p < 0.01, *** p £ 0.001 vs PL.

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
212 Garnock-Jones & Keating

effect compared with placebo. One study[43] reported significant 4.1.2 Longer Term Treatment
improvement on an overall evening parent-rated scale (p < 0.05 vs A multicenter study has investigated longer term atomoxetine
placebo), as well as on five of the eight evening subscale scores administration in this population. This study was in two phases:
(all p < 0.05 vs placebo). The other study[38] reported significant the first[47] investigated 9-month relapse prevention in responders
improvement on two of the nine evening items (both p < 0.05 vs (response defined as a decrease of ‡25% in ADHD-RS total score
placebo), but did not report an overall evening result. and a CGI-S score of 1 or 2) to an initial 12-week period of open-
Of the six trials[37,38,40-43] that did not exclude poor meta- label atomoxetine treatment; the second[46] was a re-randomized
bolizers, only one reported efficacy results in this popula- 6-month extension phase in recipients of atomoxetine in the first
tion.[37] However, a pooled analysis of four randomized, phase. Patients were aged 6–15 years and had a DSM-IV diagnosis
double-blind studies[37-39] reported that significantly (p = 0.002) of ADHD confirmed by K-SADS-PL, and a symptom severity of
greater reductions from baseline in mean ADHD-RS-IV total ‡1.5 SD above US age and sex norms. Exclusion criteria included
scores were observed among poor metabolizers (-20.9; n = 30) bipolar or psychotic disorders, unstable medical illness, and con-
than among extensive metabolizers (-14.1; n = 559).[61] Baseline comitant psychotropic medication (other than atomoxetine).
mean ADHD-RS-IV total scores were 38.9 and 40.3 in this The primary endpoint for both phases was time to relapse
pooled analysis. Response rates (percentage of patients with a (defined as an increase in ADHD-RS total score to 90% of the
‡25% decrease from baseline in ADHD-RS-IV total score) baseline score plus an increase in CGI-S score by ‡2
were also significantly higher among poor metabolizers than points).[46,47] Additional endpoints included relapse rate[46,47]
among extensive metabolizers (80% vs 59%; p = 0.033). and ADHD-RS total,[46,47] CGI-S,[47] CHQ,[46,47] CPRS,[46,47]
Atomoxetine efficacy does not appear to differ between and CTRS[46,47] scores.
children and adolescents, but it does appear more effective Of the 604 patients who entered the initial 12-week atomox-
among older children than younger children. A meta-analysis etine treatment period (target dosage 1.2 mg/kg/day [as two equal
of six trials, involving children aged 6–11 years (n = 510 ato- doses in the morning and evening], maximum 1.8 mg/kg/day),
moxetine, n = 341 placebo) and adolescents aged 12–17 years 416 responded to treatment and were randomized in a double-
(n = 107 atomoxetine, n = 69 placebo) receiving atomoxetine or blind manner to 9 months of atomoxetine (at the same dosage;
placebo for 6–8 weeks, revealed no significant differences in the n = 292) or placebo (n = 124) treatment. Of the atomoxetine re-
effects on ADHD symptoms (rated on the ADHD-RS, the cipients, 163 were re-randomized to a further 6 months of ato-
CGI-S, and the CPRS), response rates, or time to response.[44] moxetine (n = 81) or placebo (n = 82) treatment. The mean final
Another meta-analysis of six 6- to 9-week trials compared atomoxetine dosages were 1.56 mg/kg/day for the 9-month period
atomoxetine efficacy versus placebo among young children and 1.55 mg/kg/day for the subsequent 6-month period.[46,47]
(6–7 years; n = 184 atomoxetine, n = 96 placebo) versus older The only common (‡10% of patients) co-morbid disorder in
children (8–12 years; n = 544 atomoxetine, n = 316 placebo).[62] this study was ODD. The mean patient age ranged from 10.1 to
It found that, while both age groups showed a significantly 11.0 years, and 89–90% of patients were male.[46,47] Combined
greater improvement in ADHD-RS scores and response rates and inattentive subtype proportions in treatment groups ran-
among atomoxetine versus placebo recipients (p < 0.05), older ged from 73% to 74% and from 21% to 23%; 5% of patients were
children had significantly (p < 0.05) greater improvements in hyperactive/impulsive.[46,47] A total of 50–54% of patients in
ADHD-RS scores than younger children, regardless of whether one phase had previously received stimulant therapy.[47]
they were receiving atomoxetine or placebo.[62] Assessments were based on the mITT population using
Discontinuation of atomoxetine does not appear to be as- LOCF imputation. Between-group differences in baseline
sociated with symptom rebound. A prospective pooled analysis characteristics within each phase were not significant.
of two »9-week randomized, double-blind, placebo-controlled Atomoxetine was effective in the longer term treatment of
trials involving a total of 194 (102 atomoxetine £2 mg/kg/day ADHD in children and adolescents. Among responders, ato-
and 92 placebo recipients) children aged 7–12 years with moxetine was associated with a significantly longer mean time to
ADHD was carried out.[45] The trial demonstrated that among relapse than placebo in both the 9-month (217.7 vs 146.1 days;
patients originally administered atomoxetine, ADHD-RS total p < 0.001)[47] and 6-month (160.5 vs 130.8 days; p = 0.008)[46]
scores, while worsening on treatment discontinuation (p < 0.001 periods. Relapse rates for atomoxetine and placebo recipients
vs original placebo recipients), did not return to pretreatment were 22% and 38% (p < 0.01)[47] in the 9-month, and 3% and 12%
levels after a 1-week discontinuation phase, during which all (relative risk ratio for relapse with placebo vs atomoxetine 5.6;
patients received placebo in a single-blind manner. 95% CI 1.2, 25.6)[46] in the 6-month periods.

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 213

Mean total ADHD scores increased to a significantly smaller psychoactive medication;[59,60] a need for immediate pharma-
extent among atomoxetine compared with placebo recipients in cotherapy;[59] and psychotherapy.[59,60]
both periods (9-month: +6.8 vs +12.3, p < 0.001; 6-month: +1.7 vs Primary endpoints were the change from baseline in total
+7.8, p < 0.001); mean inattentive and hyperactive/impulsive ADHD-RS score[60] and HR-QOL, rated on the CHIP-CE
symptom scores also increased to a significantly smaller extent achievement domain (primary endpoint data not available from
among atomoxetine versus placebo recipients (all p < 0.01).[46,47] this study).[59] Other endpoints included ADHD-RS total,[59]
Other efficacy endpoints revealed differing results in the two inattention,[59,60] and hyperactivity/impulsivity[59,60] scores,
phases. In the first,[47] both CGI-S and CPRS scores increased to a CGI-S[59] or CGI-ADHD-S[60] scores, CGI-I scores,[59] response
significantly (p < 0.05) lesser extent among atomoxetine versus rates (the proportions of patients with a ‡25% or ‡40% im-
placebo recipients; however, the change in CTRS scores did not provement from baseline in ADHD-RS total score),[59] and
significantly differ between treatment groups. In the second phase, CHIP-CE, -AE and -PRF.[60]
however, CTRS score improved to a significantly greater extent Most patients in both studies were male (81%[59] and 80%[60]),
among atomoxetine than placebo recipients, and CPRS score with a mean age of 12[59] or 10[60] years. A total of 78%[59] and
changes did not significantly differ between treatment groups.[46] 63%[60] had the combined ADHD subtype; 4%[59,60] had the hy-
Atomoxetine recipients demonstrated a significantly smaller peractive subtype, and 18%[59] and 33%[60] had the inattentive sub-
decline in HR-QOL compared with placebo recipients, as type. The most common co-morbid disorder was ODD (20%[59]
assessed by mean CHQ psychosocial summary scores, in the and 26%[60]); other common (‡10% of patients) co-morbidities
9-month (-5.6 vs -9.5; p = 0.016)[47] but not the 6-month (-0.9 included tics (14%[59] and 17%[60]) and anxiety disorders (13%[60]).
vs -2.9)[46] period. Assessments were based on the mITT population,[59,60] using
Atomoxetine appears to maintain efficacy for at least 2 years LOCF imputation.[59] No differences were reported between
with no evidence of drug tolerance. Two meta-analyses of 13 groups in baseline characteristics.[59,60]
trials each, one in 219 adolescents (aged 12–18 years)[49] and one Atomoxetine was effective in stimulant-naive pediatric pa-
in 97 young children (aged 6–7 years)[48] with ADHD who were tients with ADHD. ADHD-RS total scores decreased to a
treated with atomoxetine for a minimum of 2 years, showed significantly (p < 0.001)[59,63] greater extent with atomoxetine
that atomoxetine retained significant (p < 0.001) improvement than with placebo (-19.0 vs -6.3[59] and -12.8 vs -4.7[60]);
at endpoint versus baseline in ADHD-RS scores.[48,49] Dosage baseline scores in atomoxetine and placebo recipients were 38.9
escalation was not required.[49] and 39.5[59] and 39.1 and 39.5.[60]
Atomoxetine was associated with an increased efficacy re-
4.1.3 In Stimulant-Naive Patients lated to a longer treatment duration. After 12 weeks, the
Two randomized, double-blind, placebo-controlled, multi- ADHD-RS total score was significantly improved compared
center trials have investigated the efficacy of atomoxetine in with after 6 weeks of atomoxetine treatment (p = 0.0132).[63]
stimulant-naive patients aged 6–15[60] or 7–15[59] years with a Where reported,[59] other endpoints support the efficacy of
K-SADS-PL-confirmed DSM-IV diagnosis of ADHD.[59,60] atomoxetine. ADHD-RS inattention and hyperactivity/
Some additional efficacy data for the 12-week study[60] were taken impulsivity subscale scores, as well as CGI-S and -I scores, were all
from an abstract.[63] Patients were randomized to treatment with significantly (p < 0.001) improved among atomoxetine versus pla-
atomoxetine (n = 49[59] and 99[60]) or placebo (n = 50[59,60] ) for cebo recipients.[59] Significantly (p < 0.001) more atomoxetine than
10[59] or 12[60] weeks. The atomoxetine dosage, taken in the morn- placebo recipients had a ‡25% (71% vs 29%) or ‡40% (63% vs
ing, was 0.5 mg/kg/day (40 mg/kg in patients weighing >70 kg[59]) 14%) improvement in ADHD-RS total scores.[59]
for the first 1[59] or 2[60] weeks and increased to a target[60] dosage HR-QOL was significantly improved with regard to risk avoid-
of 1.2 mg/kg/day (80 mg/day in patients weighing >70 kg[59]) for ance and achievement among atomoxetine versus placebo recipi-
the rest of the treatment period.[59,60] ents; no significant difference was noted between treatment groups
Eligible patients had an ADHD-RS total score of ‡1.5 stan- in the other three domains (satisfaction, comfort, and resilience).[60]
dard deviations above the US[59] age[59,60] and sex[59] norms for When rated on the CHIP-PRF, risk avoidance and achievement
their diagnostic subtype, were stimulant-naive,[59,60] and were improved to a greater extent with atomoxetine than with placebo
newly diagnosed.[60] Exclusion criteria included impaired in- (+7.89 vs -0.64 [p < 0.001] and +4.94 vs +1.55 [p = 0.042]); base-
tellect;[59,60] serious medical illness;[59] a history of psychosis,[59,60] line scores were 31.7 and 34.1 for risk avoidance and 33.2 and
bipolar disorder[59,60] or pervasive developmental disorder;[60] 33.1 for achievement.[60] When rated on the CHIP-CE/AE
alcohol or drug abuse[59,60] within the past 3 months;[59] use of (combined), risk avoidance again improved to a greater extent

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
214 Garnock-Jones & Keating

Table III. Efficacy of atomoxetine (ATO) vs other attention-deficit hyperactivity disorder (ADHD) medication in the treatment of children and adolescents with
ADHD. Results from six randomized, open-label[25,50,51] or double-blind,[23,24,26] multicenter trials comparing the efficacy of ATO with immediate-release
methylphenidate (MPH),[24,25] osmotically released MPH (OR MPH),[23,51] extended-release mixed amfetamine salts (MAS)[26] or standard current therapy
(SCT)[50] in patients (pts) aged 6–16 years
Study Treatment Age of pts Treatmenta No. Mean ADHD-RSb Mean CHIP-CE Mean SKAMP Response
duration (y) of total score total t-score deportment score ratec
(wk) pts (% pts)
baseline change baseline endpoint baseline change
Kemner et al. [51]
3 6–12 ATO d
850 38.6 -16.0 e
69
(FOCUS) OR MPHd 473 39.9 -20.2**e 80**
Kratochvil et al. [25]
10 7–15 (boys) ATO £2 mg/kg/day f
178 39.4 -19.4 e

7–9 (girls) MPH 5–60 mg/day 40 37.6 -17.8e


Newcorn et al. [23]
6 6–16 ATO 0.8–1.8 mg/kg/day 222 40.9 -14.4 45ze
OR MPH 18–54 mg/day 220 40.0 -16.9* 56*zzg,e
PL 74 41.7 -7.3 24e
Prasad et al.[50] 10 7–15 ATO 0.5–1.8 mg/kg/day 104 45.5 23.5-h 23.2 38.4-e 79-/65-i
j h e
(SUNBEAM) SCT 97 45.6 33.7 23.9 30.8 48/35i
Wang et al.[24] 8 6–16 ATO 0.8–1.8 mg/kg/day 162 38.6 -21.1 77e,k
MPH 0.2–0.6 mg/kg/day 164 37.4 -21.6 82e
Wigal et al.[26] 3 6–12 ATO 0.5–1.4 mg/kg/day 101 1.63 -0.13e 38/28l
**e
(StART) MAS 10–30 mg/day 102 1.44 -0.56 70**/68**l
a ATO was administered either once daily,[24,26,51] in divided doses twice daily,[23,25] or one or the other.[50] MPH was administered either twice daily[24] or one
of once, twice or three times daily.[25] OR MPH and MAS were both administered once daily.[23,26,51]
b Studies used an investigator-administered version of the parent[24,25] or an unspecified[23,50,51] version of ADHD-RS.
c Defined as a ‡25%[50,51] and/or ‡40%[23,24,50] reduction from baseline in ADHD-RS total score; or a ‡25% improvement on the SKAMP deportment or
attention scales.[26]
d Dosage was individually tailored to simulate clinical practice, on the basis of clinical judgment and the US FDA-approved prescribing information.
e Primary endpoint.
f Cytochrome P450 2D6 poor metabolizers received ATO 0.2–1.0 mg/kg/day.
g ATO did not demonstrate noninferiority to OR MPH.
h Presented as value at endpoint, not change from baseline.
i Data presented as ‡25%/‡40% response rate.
j SCT included any combination of medicines (excluding atomoxetine) and/or behavioral counseling, or no treatment.
k ATO was noninferior to MPH.
l SKAMP deportment/attention scale response rate.
ADHD-RS = ADHD-Rating Scale; CHIP-CE = Child Health and Illness Profile-Child Edition; FOCUS = Formal Observation of Concerta versUs Strattera;
PL = placebo; SKAMP = Swanson, Nolan, and Pelham Rating Scale; StART = Strattera/Adderall Randomized Trial; SUNBEAM = Study into the broader
efficacy of atomoxetine; * p < 0.05, ** p < 0.001 vs ATO; - p < 0.001 vs SCT; z p = 0.003, zz p £ 0.001 vs placebo.

with atomoxetine than with placebo (+3.60 vs +0.03; p = 0.006) 6–16 years with ADHD has been evaluated in six randomized,
from baseline scores of 47.6 and 49.1; however, no significant open-label[25,50,51] or double-blind,[23,24,26] multicenter, fully pub-
treatment difference was noted for achievement on this scale.[60] lished trials lasting 3–10 weeks (table III).[23-26,50,51] Active com-
parators included immediate-release methylphenidate,[24,25]
4.2 Comparisons with Stimulants or Standard osmotically released methylphenidate,[23,51] extended-release
Current Therapy mixed amfetamine salts (this study was conducted in a laboratory
school setting),[26] and standard current therapy (any combination
The efficacy of oral atomoxetine compared with other ADHD of medicines [excluding atomoxetine] and/or behavioral counsel-
medication in the treatment of children and adolescents aged ing, or no treatment).[50] It should be noted that two of these trials

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 215

(comparing atomoxetine with extended-release mixed amfetamine studies excluded patients who had not responded to previous
salts[26] and osmotically released methylphenidate[51]) were of only ADHD treatment.[23,25,51]
3 weeks’ duration; full benefits of atomoxetine often take several Primary endpoints were ADHD-RS investigator-adminis-
weeks (potentially up to 8 weeks)[8] to occur.[2,64] Additionally, one tered total score (parent[25] or unspecified[51] version), CHIP-CE
of the studies comparing atomoxetine with immediate-release me- total t-score,[50] SKAMP deportment score,[26] and response
thylphenidate reported preliminary results only, from a study in- (defined as a ‡40% reduction from baseline in ADHD-RS total
vestigating relapse prevention, and was not powered for score) rate.[23,24] Additional endpoints included response rates
comparisons between the two drugs;[25] despite this, statistical (see table III for definitions),[26,50,51] and ADHD-RS,[23,24,50,51]
comparisons were still reported and are included in this section. One CPRS,[23-25] CGI-ADHD-S,[23-25] CGI-S,[50] and SKAMP at-
trial titrated atomoxetine recipients to a target dosage of atomox- tention[26] scores. Two studies[23,26] also investigated HR-QOL,
etine 1.2 mg/kg/day (maximum permitted dosage was 1.4 mg/kg/ using the PedsQL[26] or CHQ.[23]
day); the extended-release mixed amfetamine salt dosage was in- The proportions of patients in treatment groups with ADHD of
creased in 10 mg increments at 1-week intervals to a final dosage of the combined, inattentive and hyperactive/impulsive subtypes
30 mg/day.[26] One trial administered atomoxetine or osmotically ranged from 57% to 100%, from 0% to 39% and from 0% to 13%,
released methylphenidate with an individually tailored dosage, to respectively.[23-26,50,51] The most common co-morbid disorders
simulate the clinical setting.[51] The remaining trials titrated patients (‡10% of patients) were ODD[23-25,50] and elimination disorders.[25]
according to therapeutic response, with permitted atomoxetine Mean patient age ranged from 8.6 to 11.1 years,[23-26,50,51] and
dosages of £2,[25] 0.8–1.8,[23,24] 0.5–1.8[50] mg/kg/day, immediate- 69–100% of patients were male.[23-26,50,51] A total of 23–67% of
release methylphenidate dosages of 5–60 mg/day[25] or 0.2–0.6 mg/ patients had received prior ADHD treatment.[23,24,51]
kg/day,[24] and osmotically released methylphenidate dosages of Where stated, assessments were based on the mITT popu-
18–54 mg/day.[23] Where stated, the mean final dosages were lation[23-26,50] using LOCF imputation.[23-25,50] Between-group
1.08,[51] 1.40 (among extensive metabolizers only),[25] 1.45,[23] 1.5,[50] differences in baseline characteristics within each trial were ei-
and 1.37[24] mg/kg/day for atomoxetine, and were 31.3 mg/day[25] ther not significant[23,26] or accounted for in the analyses,[50,51]
and 0.52 mg/kg/day[24] for immediate-release methylphenidate and except for one significant (p < 0.05) difference in sex propor-
1.01[51] and 1.16[23] mg/kg/day for osmotically released methylphe- tions[25] and another (p < 0.05) in patient age.[24] Two trials were
nidate. Certain limitations of these trials are discussed in section 8. noninferiority studies; noninferiority of atomoxetine to im-
In the trial comparing atomoxetine with standard current mediate-release methylphenidate[24] or osmotically released
therapy, at baseline 75.3% of patients in the SCT arm received methylphenidate[23] was established if the lower limit of the 95%
only pharmacotherapy, 3.1% received simple behavioural confidence interval (CI) for the difference between groups in
counselling, and 10.3% received both pharmacotherapy and response rate was greater than -18%[24] or -15%.[23]
simple behavioural counselling; 11.3% received no treat- Atomoxetine did not differ significantly from[25] or was non-
ment.[50] Pharmacotherapy included immediate-release me- inferior to[24] immediate-release methylphenidate, with regard to
thylphenidate (37.3%), extended-release methylphenidate the primary endpoints (change from baseline in ADHD-RS total
(47.0%), clonidine (3.6%), or combinations thereof (12%). score[25] and response rate[24]) in children and adolescents with
Eligible patients met DSM-IV criteria for ADHD (any sub- ADHD (table III). In terms of response rate, the lower limit of the
type[23-25,50,51] or either combined or hyperactive/impulsive sub- 95% CI for the difference between atomoxetine and immediate-
types[26]), confirmed by K-SADS-PL[23,24,50] or an unspecified release methylphenidate was -11.7%.[24] Recipients of atomox-
K-SADS version,[25] and had to have an investigator-administered etine did not differ significantly from those receiving immediate-
ADHD-RS total score of ‡24,[51] ‡25 for boys and ‡22 for girls,[24] release methylphenidate in the mean change from baseline on
‡1.5 SD above age and/or sex norms,[23,25,50] or >12 for a specific ADHD inattention (-11.3 vs -12.0[24] and -9.9 vs -9.3[25]) or
subtype.[24] One trial required a CGI-S score of ‡4,[51] another hyperactivity/impulsivity (-9.7 vs -9.5[24] and -9.5 vs -8.5[25])
a CGI-ADHD-S score of ‡4.[24] Exclusion criteria included subscale scores, although both treatment groups showed scores
other psychiatric disorders (except ODD;[51] e.g. depression,[26] significantly lower than baseline (both p < 0.001).[24,25] Other
bipolar disorders,[23-26,50] anxiety disorders,[23,24,26] psychotic dis- additional endpoints (CPRS and CGI-ADHD-S scores) did not
orders[23-26,50]); a history of seizure,[23,26,50,51] tic disorder,[23-26,51] differ significantly between groups.[24,25]
mental retardation[51] or developmental disorder;[23,24,26,50,51] Atomoxetine was significantly less effective than osmotically
Tourette’s syndrome;[24-26,51] use of concomitant psychotropic released methylphenidate[51] and extended-release mixed amfe-
medication;[24,26,50,51] and serious medical illness.[25,50] Three tamine salts,[26] both controlled-release formulations of stimu-

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
216 Garnock-Jones & Keating

lants, in the change from baseline in ADHD-RS total score[51] in patients with ADHD and co-morbid ODD has also been
and SKAMP deportment score[26] (table III). In terms of response conducted.[54]
rate, atomoxetine was not noninferior to osmotically released Eligible patients were children or adolescents (aged 5–18 years;
methylphenidate (lower limit of the 95% CI for between-group mean 9.3–14.6 years)[52-57] with a DSM-IV diagnosis of ADHD[52-57]
difference of -21%); the response rate was subsequently shown to and a total or subscale ADHD-RS score of ‡1.5 SD above age
be significantly higher in osmotically released methylphenidate and sex norms,[52,55,57] and/or a Children’s Depression Rating
recipients than in atomoxetine recipients (table III).[23] In addi- Scale-revised total score >36[53] or ‡40.[55] Patients were also re-
tion, response rates were significantly higher with osmotically quired to have a DSM-IV diagnosis of an anxiety disorder,[52,53]
released methylphenidate[51] or extended-release mixed amfeta- depressive disorder,[53] major depression,[55] Tourette’s syn-
mine salts[26] than with atomoxetine in two other trials (table III). drome,[57] tic disorders,[57] autism spectrum disorder,[56]
Mean SKAMP attention score improvement was also significant- or ODD.[54] Diagnoses were confirmed using the K-SADS-
ly lower among atomoxetine than extended-release mixed PL.[52-55,57] One study[52] focused on patients who did not
amfetamine salt recipients (-0.08 vs -0.49; p < 0.001),[26] and ato- respond during the placebo lead-in period, based on a £25%
moxetine recipients demonstrated a significantly (p < 0.01) lower reduction in Pediatric Anxiety Rating Scale (PARS) score.
improvement on the CPRS and CGI-ADHD-S scales compared Patient numbers ranged from 16[56] to 226,[54] with most studies
with recipients of osmotically released methylphenidate.[23] involving between 100 and 200 patients.[52,53,55,57]
In a subanalysis, the response rates among stimulant-ex- Not all primary endpoints related to efficacy with regard to
posed patients (n = 301) receiving atomoxetine versus osmoti- ADHD (most focused on both ADHD and the co-morbid
cally released methylphenidate were significantly different condition,[52,54-57] one focused on tolerability[53]); the main
(37% vs 51%; p = 0.03) and only osmotically released methyl- ADHD endpoints of each trial (ADHD-RS total,[52,53,55,57]
phenidate differed significantly from placebo (23%; p = 0.002); DSM-IV ADHD symptoms,[56] and SNAP-IV ADHD sub-
however, among stimulant-naive patients (n = 191), the two scale[54] scores) are focused on in this section.
active treatment groups did not differ significantly from each Patients did not receive concomitant psychotropic treatment
other (57% vs 64%) and both differed significantly from pla- for their co-morbid conditions in most studies;[52,54-57] the one
cebo (25%; p = 0.004 and p £ 0.001, respectively).[23] exception[53] involved all patients receiving atomoxetine and
When compared with standard current therapy, atomoxetine being randomized to fluoxetine or placebo.
had a significantly greater positive impact on HR-QOL, as The efficacy of atomoxetine does not appear to be affected by
assessed by the mean CHIP-CE total t-score at endpoint[50] the presence of co-morbid disorders, and symptoms of the
(table III). Atomoxetine was also significantly more effective than co-morbid disorders were either not adversely affected or im-
standard current therapy in terms of ADHD-RS total score and proved in these studies.[52-57] Atomoxetine was more effective
response rate at endpoint (table III), as well as on the ADHD-RS than placebo in the treatment of ADHD in patients with co-
subscale scores (inattention 12.1 vs 17.3; p < 0.001; hyperactivity/ morbid anxiety disorders (mean change in ADHD-RS total score
impulsivity 11.3 vs 16.3; p < 0.001).[50] Atomoxetine was asso- of -10.5 vs -1.4; p < 0.001),[52] major depression (mean change in
ciated with a significant improvement versus standard current ADHD-RS total score of -13.3 vs -5.1; p < 0.001),[55] tic disorders
therapy with regard to CGI-S score (p < 0.001).[50] (including Tourette’s syndrome) [mean change in ADHD-RS
total score of -10.9 vs -4.9; p < 0.01],[57] autism spectrum dis-
4.3 In Patients with Co-Morbid Conditions orders (mean DSM-IV ADHD hyperactive/impulsive symptom
score at endpoint of 10.4 vs 14.5; p < 0.01; no significant between-
While most of the major efficacy trials investigating atomox- group difference in inattentive symptom score),[56] and ODD
etine have allowed the inclusion of patients with co-morbid (mean change in SNAP-IV ADHD hyperactive/impulsive sub-
ODD, other disorders are also common in patients with ADHD, scale score of -4.6 vs -2.2; p < 0.01; change in inattentive subscale
and were either not permitted in these trials or patient numbers score of -5.0 vs -2.2; p < 0.001).[54]
with these co-morbid disorders were low. Therefore, randomized, In general, symptoms of co-morbid conditions were not wors-
double-blind, placebo-controlled trials,[52-57] mainly of multi- ened by the use of atomoxetine;[54,55,57] for example, in children
center design,[52-55,57] have been conducted in patients with co- with co-morbid tic disorder Yale Global Tic Severity Scale total
morbid anxiety disorders,[52,53] depressive disorders,[53,55] tic dis- scores in atomoxetine recipients did not differ significantly from
orders (including Tourette’s syndrome),[57] and autism spectrum those in placebo recipients (-5.5 vs -3.0) and Tic Symptom Self-
disorders (including Asperger’s syndrome).[56] An additional trial Report total scores were -4.7 and -2.9, while Clinical Global

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 217

Impressions tic/neurologic severity scale scores were significantly In individual placebo-controlled trials, significantly (p < 0.05)
improved with atomoxetine (-0.7 vs -0.1; p = 0.002).[57] more atomoxetine than placebo recipients reported decreased
In addition, in some trials[52,56] the co-morbid conditions appetite (18–36% vs 4–17%),[38-40,42,43] somnolence (15–17% vs
were improved by treatment with atomoxetine; for example, in 2–4%),[42,43] vomiting (15% vs 1%),[38] nausea (12–17% vs
patients with anxiety disorders Pediatric Anxiety Rating Scale 0–2%),[38,40] asthenia (11% vs 1%),[38] fatigue (10% vs 2%),[43] and
scores in atomoxetine recipients were significantly improved dyspepsia (9% vs 0%).[38]
with respect to placebo (-5.5 vs -3.2; p < 0.012).[52] Where reported, the severity of adverse events among ato-
When administered alone, atomoxetine did not significantly moxetine recipients was generally classified as mild[51] or mild
differ from atomoxetine plus fluoxetine with regard to ADHD to moderate.[24,26,50,55] Most studies either did not report the in-
(mean change in ADHD-RS total score of -20.5 vs -24.0) or cidence of serious adverse events[23,25,37,38,40,41,43,46,47,52-54] or stated
anxiety symptoms (mean change in Multidimensional Anxiety that there were none among atomoxetine recipients.[26,39,42,50,55,57]
Scale for Children score of -11.3 vs -13.4).[53] Depression Where reported, serious adverse events included aggressive beha-
measures were inconclusive in this study: there was a significant vior,[56] partial seizure,[24] and prolonged crying (fear of death),[51]
difference between groups in favor of combination treatment in and had a low incidence.
the change in mean Children’s Depression Inventory score A total of 3% of atomoxetine recipients (48 of 1613) versus
(-5.4 vs -8.8; p = 0.43), but no significant difference was found 1% of placebo recipients (13 of 945) discontinued treatment
in the change in mean Children’s Depression Rating Scale- because of adverse events in short-term studies, with adverse
Revised score (-17.6 vs -20.4). events leading to discontinuation including irritability, som-
nolence, aggression, nausea, vomiting, abdominal pain, con-
stipation, fatigue, feeling abnormal, and headache.[10]
5. Tolerability Atomoxetine appears better tolerated among extensive me-
tabolizers than poor metabolizers. The US prescribing in-
This section focuses primarily on data from the US prescrib- formation reported that adverse events occurring in ‡5% of
ing information,[10] supplemented by data from two meta- poor metabolizers and either twice as frequently or significantly
analyses[65,66] and the trials reported in section 4.[23-26,37-43,46,47,50-57] (p-value not stated) more frequently among poor metabolizers
Oral atomoxetine was generally well tolerated in children and than extensive metabolizers included tremor (5% vs 1%), de-
adolescents with ADHD. Common (‡5% of atomoxetine recipi- pression (7% vs 4%), constipation (7% vs 4%), decreased weight
ents and reported by numerically more atomoxetine than placebo (7% vs 4%), and insomnia (15% vs 10%).[10] A pooled analysis
recipients) adverse events from placebo-controlled trials of £18 of 14 studies reported significant differences between poor
weeks’ duration are shown in figure 1, and included headache, (n = 237) and extensive (n = 3017) metabolizers receiving ato-
abdominal pain, decreased appetite, vomiting, somnolence, and moxetine in the incidence of decreased appetite (24% vs 17%;
nausea.[10] p = 0.008), insomnia (11% vs 7%; p = 0.035), abrasion (5% vs
2%; p = 0.012), and tremor (5% vs 1%; p < 0.001).[61]
ATO
Dizziness
PL Atomoxetine appears generally well tolerated in the long-
Irritability
term treatment of children and adolescents with ADHD. Dis-
Fatigue
continuations due to adverse events in long-term placebo-
Nausea
controlled trials among atomoxetine versus placebo recipients
Somnolence
occurred in 3%[47] and 1%[46] versus 1%[47] and 1%.[46] A total of
Vomiting
66% atomoxetine versus 54% placebo recipients reported at
Decreased appetite
least one new or worsened adverse event.[47] Common adverse
Abdominal pain
events included headache (10% atomoxetine vs 9% placebo
Headache
recipients) and nasopharyngitis (8% vs 9%).[46]
0 5 10 15 20 Atomoxetine appears associated with weight loss to varying
Incidence (% patients)
degrees in comparison with other ADHD medication. Overall, in
Fig. 1. Tolerability of oral atomoxetine (ATO) in pediatric patients with placebo-controlled trials the incidence of weight loss was 3% of
attention-deficit hyperactivity disorder. Incidence of common (‡5% of ATO
atomoxetine vs 0% of placebo recipients.[10] Long-term placebo-
recipients [n = 1597] and reported by numerically more ATO recipients than
placebo [PL] recipients [n = 934]) treatment-emergent adverse events asso- controlled trials demonstrated that a switch to placebo was as-
ciated with ATO and PL. Data from trials of £18 weeks’ duration.[10] sociated with a greater increase in weight than atomoxetine

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
218 Garnock-Jones & Keating

continuation (weight +3.3 vs +1.2 kg; p < 0.001;[47] weight per- recipients did not differ significantly in overall incidence of ad-
centile +9.9 vs +0.72; p < 0.001[46]). Of the active comparator verse events, the incidence of treatment-related adverse events, or
studies, one[24] found significantly (p < 0.001) greater weight loss the incidence of discontinuations due to adverse events.
among atomoxetine versus immediate-release methylphenidate In the trial comparing atomoxetine with extended-release
recipients (-1.2 vs -0.4 kg), and one[23] found significantly lower mixed amfetamine salts, the most commonly occurring treat-
weight loss among atomoxetine versus osmotically released me- ment-related adverse events among atomoxetine recipients were
thylphenidate recipients (-0.6 vs -0.9 kg; p < 0.05). Where re- somnolence (19% of patients), decreased appetite (18%), upper
ported in the other active comparator trials, weight loss did not abdominal pain (15%), and headache, and among extended-
differ significantly between treatment groups.[25,50] Longer term release mixed amfetamine recipients were insomnia (28% of pa-
weight data (£5 years) are presented in section 2.2, and imply that tients), decreased appetite (28%), upper abdominal pain (19%),
this weight loss is not permanent. and anorexia (17%).[26] In this study, a total of 73% of atomox-
Where reported, no clinically significant changes in labora- etine versus 85% of extended-release mixed amfetamine salt re-
tory test values occurred in any of the short-[23-25,40,42,43] or cipients reported treatment-emergent adverse events, 65% versus
long-[46,47] term trials. 74% reported a study medication-related adverse event, and 4%
Atomoxetine was associated with a significantly (p < 0.05) and 7% discontinued treatment because of adverse events.[26]
greater incidence of somnolence (11–26% vs 0–4%),[24,25] nausea A pooled analysis using data from two open-label atomox-
(20% vs 10%),[24] vomiting (12% vs 0–4%),[24,25] anorexia (37% vs etine studies demonstrated that the initiation of therapy using a
25%),[24] and dizziness (15% vs 7%)[24] compared with immediate- twice-daily divided dose regimen compared with a once-daily
release methylphenidate. Immediate-release methylphenidate regimen could potentially decrease the risk of adverse events
was associated with a significantly (p < 0.05) greater incidence of within the first few weeks of treatment.[67] The incidence of
abnormal thinking than atomoxetine (5% vs 0%).[25] One study decreased appetite in the first 2 weeks was significantly higher
found that the incidence of discontinuation due to adverse events among once-daily than among twice-daily atomoxetine re-
did not differ significantly between groups;[25] another showed cipients (14% vs 8%; p = 0.036), as was somnolence (14% vs 4%;
significantly (p < 0.05) more atomoxetine than immediate-release p < 0.001). The incidence of headache was, however, lower
methylphenidate recipients discontinued because of treatment- among once-daily recipients (7% vs 17%; p = 0.003).
emergent adverse events (11% vs 4%), and had an overall sig- Discontinuation of atomoxetine appears to be well tolerated.
nificantly (p < 0.001) greater incidence of treatment-emergent A prospective pooled analysis of two 9- to 10-week trials in
adverse events (87% vs 68%).[24] children aged 7–12 years with ADHD demonstrated that after a
In one study comparing atomoxetine with osmotically released 1-week discontinuation phase, during which all patients received
methylphenidate and placebo,[23] recipients of atomoxetine had a placebo in a single-blind manner, there was a low incidence of
significantly (p < 0.05) greater incidence of somnolence compared discontinuation-emergent adverse events, and no significant dif-
with osmotically released methylphenidate but not placebo (6% vs ferences were observed between patients previously receiving
2% and 4%), whereas osmotically released methylphenidate was atomoxetine and those receiving placebo throughout.[45]
associated with a significantly (p < 0.05) greater incidence of in-
somnia compared with atomoxetine and placebo (13% vs 7% 5.1 Specific Adverse Events
and 1%).[23] Both atomoxetine and osmotically released methyl-
phenidate were associated with a significantly higher incidence of The US prescribing information carries a black-box warning
decreased appetite compared with placebo (14% and 17% vs 3%), regarding suicidal ideation in children and adolescents; it appears
but did not differ significantly from each other. Atomoxetine was to be more common among atomoxetine than placebo recipi-
associated with a significantly higher rate of any treatment-emer- ents.[10] A meta-analysis including 14 pediatric clinical trials
gent adverse events than placebo (67% vs 54%); osmotically re- (12 placebo-controlled and 5 including a methylphenidate treat-
leased methylphenidate did not differ from either other treatment ment arm) was conducted investigating the suicidality of patients
group in this measure (67%).[23] In the other study comparing ato- aged 6–18 years receiving atomoxetine.[66] No suicides occurred in
moxetine to osmotically released methylphenidate, adverse events the trials included in the meta-analysis. The frequency of suicidal
included somnolence (4% of atomoxetine vs 1% of osmotically ideation was greater among atomoxetine (n = 1357) than placebo
released methylphenidate recipients), nausea (5% vs 1%), fatigue (n = 851) recipients (0.37% vs 0%; incidence difference 0.46; 95%
(3% vs 0%), insomnia (2% vs 6%), and decreased appetite (3% vs CI 0.09, 0.83; p = 0.016), but did not differ significantly between
6%).[51] Atomoxetine and osmotically released methylphenidate atomoxetine (n = 558) and methylphenidate (n = 464) recipients

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 219

(0.18% vs 0.22%; incidence difference -0.12; 95% CI -0.62, +0.38). changes,[10] cardiovascular disease,[11] or cerebrovascular dis-
All patients with suicidal ideation in placebo-controlled trials ease,[11] as well as in patients with hypotension.[10,11]
were male and aged between 7 and 12 years. With atomoxetine, A study conducted by FDA officials indicates that atomox-
the number needed to harm for an additional suicide-related etine use (as well as other ADHD medication use) is potentially
event was 227; the number needed to treat for achievement of associated with hallucinations and other psychotic symptoms.[68]
remission of ADHD symptoms was 5.[66] Patients starting atomox- In a pooled analysis from 49 randomized, placebo-controlled
etine therapy should be closely monitored for suicidal thinking clinical trials of extended-release mixed amfetamine salts (four
and behavior, clinical worsening, or unusual changes in behavior.[10] trials), dexmethylphenidate (seven trials), extended-release
Rarely, atomoxetine may be associated with severe liver in- methylphenidate (tablets [four trials] and capsules [four trials]),
jury.[10] A retrospective study investigated the incidence of liver- long-acting extended-release methylphenidate (three trials), me-
related adverse events in pediatric and adult patients treated with thylphenidate transdermal system (eight trials), modafinil (five
atomoxetine in clinical trials, as well as among spontaneous post- trials), and atomoxetine (14 trials), the rate of psychosis/mania
marketing reports of adverse events.[65] Among recipients of events per 100 person-years was 1.48 in the drug group compared
atomoxetine in clinical trials, 41 of 7961 (0.5%) had hepatobiliary with 0 in the placebo group. Of the eleven psychosis/mania events
events that were considered possibly related to atomoxetine; most occurring in the pooled drug group, four occurred among ato-
of these were mild increases in ALT or AST levels.[65] Four years moxetine recipients.[68]
after the market launch of atomoxetine, a total of 351 liver- There have been post-marketing spontaneous reports of
related adverse events had been reported (among a total of seizures.[10,11] No fatal overdoses occurred in clinical trials, and
4 328 000 recipients of atomoxetine; <0.01%). Of these, 69 were there have been no post-marketing reports of death by overdose
explainable by factors unrelated to atomoxetine, 146 had too little with atomoxetine alone.[10]
information to assess fully, 133 had possible confounding factors
and were deemed possibly atomoxetine-related, and 3 found 6. Pharmacoeconomic Considerations
atomoxetine to be probably related. These three patients all re-
covered after atomoxetine discontinuation.[65] Patients who have This section provides a brief overview of recent pharmaco-
jaundice or laboratory evidence of liver injury should discontinue economic analyses of atomoxetine in the treatment of chil-
atomoxetine treatment; treatment should not be re-initiated.[10] dren[69-72] and adolescents[70] with ADHD. The cost effectiveness
There is the potential for cardiovascular effects with ato- of atomoxetine compared with other treatments (immediate- or
moxetine administration. Both the US[10] and the UK[11] prescrib- extended-release methylphenidate,[69-72] dexamfetamine,[69,71] tri-
ing information include precautions regarding cardiovascular cyclic antidepressants,[71] or no medication[69,70,72]), as a cost-
effects. Caution should be used when prescribing atomoxetine to utility analysis, has been investigated in one fully published
children and adolescents with serious structural cardiac ab- paper[69] and three abstracts with attached posters[70-72] (table IV).
normalities[10,11] (a cardiac specialist should be consulted in these All four analyses used a Markov model, incorporating 10,[72]
patients)[11] or other serious heart problems,[10] as sudden death 14,[70] 18,[69] or 22[71] health states, to estimate the costs and ben-
has been reported in these patients;[10,11] caution should also be efits of the treatment strategies, and estimated the incremental
used in patients with congenital long QT interval, acquired long cost per quality-adjusted life-year (QALY) gained with atomox-
QT interval, or a family history of QT prolongation.[11] There etine versus the comparator strategies.[69-72] A survey of 83 parents
have been post-marketing spontaneous reports of prolonged QT of children with ADHD provided utility values for all four stu-
intervals during atomoxetine administration.[10,11] dies, and efficacy and safety of the involved treatments were based
In a pooled analysis of placebo-controlled trials, 2.5% (36 of on a review of controlled clinical trials and other clinical litera-
1434 patients) of atomoxetine and 0.2% (2 of 850) of placebo ture.[69-72] Costs and outcomes were estimated using a Monte
recipients demonstrated a heart rate of at least 110 bpm plus an Carlo simulation, with a time horizon of 1 year;[69-72] costs were
increase in heart rate of ‡25 bpm at endpoint.[10] A total of 4.8% estimated from the perspective of the National Health Service in
(59 of 1226) of atomoxetine and 3.5% (26 of 748) of placebo had England and Wales,[69] Dutch society,[71] the German health ser-
high SBP at endpoint; corresponding proportions of patients vice,[72] and the Norwegian healthcare system.[70] All four studies
with high DBP at endpoint were 4.0% (50 of 1262) and 1.1% included direct costs[69-72] (two studies only included study drug
(8 of 759).[10] Caution should be used when administering costs[69,72] ), with one study also including indirect costs.[71]
atomoxetine in patients with hypertension,[11] tachycardia,[11] All four studies divided patients into three groups: stimulant-
other conditions associated with abrupt heart rate or BP naive, stimulant-failure, and stimulant-contraindicated.[69-72] The

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
220 Garnock-Jones & Keating

Table IV. Pharmacoeconomic evaluation of atomoxetine vs other or no medication in children[69-72] and adolescents[70] with attention-deficit hyperactivity
disorder (ADHD). Incremental cost per quality-adjusted life-year (QALY) gained, estimated using a Markov model with a time horizon of 1 year
Study Perspective Year of Incremental cost per QALY gained

costing stimulant-naive pts stimulant-failed pts stimulant-contraindicated pts

IM MPH ER MPH DA IR no med TCA no med


Cottrell et al. [69]
NHS in England and Wales 2004 d15 224 d13 241 d14 945 d11 523/12 370a
Diamantopoulos et al.[72]b German health service NR h18 227 h7778 h14 385 h14 916
Laing et al.[71]b Dutch societal NR h18 831 h22 804 h13 120 Dominant
Tilden et al.[70]b Norwegian healthcare system 2005 h25 463c h19 162c h21 497c h22 385c
a Stimulant-naive/-exposed patients.
b Available as abstract plus poster.
c Converted from Norwegian kroner (NOK) using the exchange rate as at 13/10/2005 of NOK1 = h0.12784.
DA IR = dexamfetamine immediate release; ER MPH = extended-release methylphenidate; IM MPH = immediate-release methylphenidate; NHS = National
Health Service; no med = no medication; NR = not reported; pts = patients; TCA = tricyclic antidepressants.

fully published study also included stimulant-averse patients Pharmacoeconomic analyses of atomoxetine, in common with
(results for this group are not reported here), and further split the all pharmacoeconomic analyses, are subject to a number of lim-
stimulant-contraindicated patients into stimulant-naive and -ex- itations. Pharmacoeconomic analyses based on clinical trials ex-
posed.[69] Treatment algorithms differed slightly between studies, trapolate the results of such trials to the general population;
but involved an atomoxetine-based algorithm versus a com- however, patient populations, rates of compliance, and major
parator algorithm, with numbers of treatments in each algorithm outcomes in clinical trials may differ from those observed in real-
ranging from two to five for stimulant-naive patients, one to four life practice. Modeled analyses, such as those presented in this
for stimulant-failed patients, and one to three for stimulant- section, rely on a number of assumptions and use data from a
contraindicated patients. variety of sources. Results of pharmacoeconomic analyses may
Initial atomoxetine treatment algorithms appear cost effective not be applicable to other geographical regions because of dif-
versus algorithms involving initial methylphenidate (immediate- ferences in healthcare systems, medical practice, and unit costs.
or extended-release), dexamfetamine, tricyclic antidepressants, or
no treatment in stimulant-naive, -failed, and -contraindicated
children and adolescents with ADHD (table IV).[69-72] The in-
7. Dosage and Administration
cremental cost per QALY gained was below commonly accepted
cost-effectiveness thresholds.[69-72] Sensitivity analyses demon- Atomoxetine is indicated in various countries including the
strated that the models were robust to changes in most variables; US[10] and the UK[11] for the treatment of children and ado-
however, utility values were shown to be important indicators of lescents with ADHD;[10,11] atomoxetine is approved for use in
cost effectiveness.[69-72] children aged ‡6 years in the UK.[11]
An additional study (available as a poster) investigating the
cost effectiveness of atomoxetine versus placebo with regard to
societal costs in Sweden found that placebo was dominated by Table V. Summary of recommended atomoxetine dosage and administra-
tion in pediatric patients according to bodyweight.a Local prescribing in-
atomoxetine.[73] Costs were taken from 2005 databases, with an
formation should be consulted for further details
exchange rate of h1 = 9.2 Swedish kronor, and included both in-
direct and direct costs. Stimulant-naive patients with ADHD Atomoxetine £70 kg >70 kg
dosage (mg/kg/day) (mg/day)
(aged 7–15 years) were randomized to 10 weeks of treatment with
Initial dosage 0.5 40
atomoxetine or placebo; both groups received additional parental
training. Atomoxetine was associated with a significantly greater Target dosageb 1.2 80

least squares mean change from baseline in ADHD-RS score Maximum dosage 1.4 (US) 100

(-19.0 vs -6.3; p < 0.001), and total cost (excluding costs due to a According to both US[10] and UK[11] labeling unless otherwise specified.
loss of leisure time) decreased among atomoxetine recipients and b The minimum period before the first dose increment is 3 days in the US[10]
increased among placebo recipients (-h349 vs +h246; p = 0.002). or 7 days in the UK.[11]

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 221

Recommended atomoxetine dosages and the administration potential (but not US FDA-approved) therapies include bupro-
schedule are presented in table V. No additional benefit has been pion, tricyclic antidepressants, and a-adrenergic agonists.[74]
shown for atomoxetine dosages exceeding 1.2 mg/kg/day.[10,11] While the immediate-release formulations of methylpheni-
Atomoxetine can be taken with or without food, and can be date and amfetamines have been shown to be effective in
discontinued without tapering.[10,11] Capsules should be taken ADHD, they require multiple daily doses to achieve the optimal
whole.[10] Atomoxetine can be administered either as a single effect, and can potentially be associated with abuse or mis-
daily dose in the morning, or as evenly divided doses in the use.[77] Longer acting, once daily, extended-release formula-
morning and late afternoon or early evening.[10] tions of stimulants are just as effective as immediate release, and
The US prescribing information contains a black-box are easier to administer; however, they may be associated with
warning regarding suicidal ideation (section 5).[10] pharmacokinetic variability, as they rely on pH and gastro-
Initial and target atomoxetine dosages should be halved in intestinal transit time for delivery of the active ingredient.[77]
patients with moderate hepatic impairment (Child-Pugh class B) Atomoxetine is the first drug not classified as a stimulant to be
and reduced by three-quarters in patients with severe hepatic im- approved for ADHD, and, unlike stimulants, is not a controlled
pairment (Child-Pugh class C).[10,11] No dosage adjustment is re- substance.[78] It also has low to no abuse or misuse potential.[79]
quired for patients with end-stage renal disease or lesser degrees The issue of stimulant abuse is, however, controversial. It has
of renal insufficiency.[10,11] In the UK, in patients who are known been stated that patients who are being administered therapeutic
to be poor metabolizers or who are receiving concomitant CYP2D6 dosages of stimulants for the treatment of ADHD do not nor-
inhibitors, such as paroxetine, a lower starting dose and slower mally have a problem with abuse; in fact, there is the possibility
dosage up-titration than extensive metabolizers should be con- that stimulant treatment may reduce the chance of substance
sidered.[11] In the US, patients who are either poor metabolizers abuse in ADHD patients.[79] More emphasis is placed on the risk
or who are receiving concomitant strong CYP2D6 inhibitors are of diversion of immediate-release stimulants for recreation or
recommended to initiate treatment at the normal dosage, but performance enhancement (extended-release formulations can be
only increase to the normal target dosage if no response has oc- used, but it is more difficult to extract the drug).[79]
curred after 4 weeks and if the initial dosage is well tolerated.[10] In well designed clinical trials of 6–9 weeks’ duration, ato-
Local prescribing information should be consulted for con- moxetine was effective in pediatric patients (including stimulant-
traindications, precautions and warnings, drug interactions, naive patients) with ADHD, demonstrating greater improvements
dosage modifications, and patient monitoring requirements. from baseline in efficacy measures than placebo (sections 4.1.1 and
4.1.3). Atomoxetine was also effective in preventing relapse in
8. Place of Atomoxetine in the Management of longer term trials; its efficacy was maintained for at least 2 years
Attention-Deficit Hyperactivity Disorder (section 4.1.2). Atomoxetine, despite a half-life of just over 3 hours,
shows efficacy into the evening if given as a single morning dose
The main aim of therapy for ADHD patients is to improve (section 4.1.1), and can be administered either as a single dose or
psychologic functioning (including academic, family, and social two evenly divided doses (section 7).[10] However, a recent study
functioning),[74] i.e. the core symptoms of ADHD. Both behavioral has shown that the risk of some adverse events may be lower if
and pharmacologic therapies, as well as psychoeducation, are re- atomoxetine is initiated as a twice-daily regimen (section 5).[67]
commended.[74] Current treatment guidelines recommend the Prospective trials are required to confirm this possibility.
first-line use of psychostimulants (e.g. methylphenidate) and/or Co-morbid disorders (such as anxiety disorders, depressive
psychosocial intervention for the treatment of children (aged over disorders, tic disorders, and autism spectrum disorders), common
6 years) and adolescents with ADHD, particularly if no co-morbid in ADHD patients, were either not affected or improved on ad-
condition is present.[64,74,75] Second-line therapies include a dif- ministration of atomoxetine, and efficacy with regard to ADHD
ferent stimulant or atomoxetine.[64,75] In the case of patients with was not affected by the presence of co-morbidities (section 4.3).
ADHD and co-morbid anxiety, first-line treatment is either ato- However, four[53,54,56,57] of the six[52-57] trials in patients with co-
moxetine (to treat both disorders) or a stimulant (to treat the morbid disorders did not have ADHD symptom measures as their
ADHD) plus a serotonin reuptake inhibitor (to treat the anxi- primary endpoints, and thus definitive conclusions with regard to
ety).[64,74] Atomoxetine may also be a first-line therapy for patients the efficacy of atomoxetine cannot be drawn from these studies.
with co-morbid tics.[74] Patients unable to take stimulants because Comparisons with other ADHD medications revealed
of an active substance abuse disorder or previous adverse effects mixed results with regard to primary efficacy measures.
are advised to take atomoxetine as a first-line therapy.[74,76] Other Atomoxetine appeared more effective than standard current

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
222 Garnock-Jones & Keating

therapy (any combination of medicines [immediate- or extended- Pharmacoeconomic analyses suggest that initial atomox-
release methylphenidate or clonidine] and/or behavioral coun- etine treatment appears cost effective compared with initial
seling, or no treatment) and as effective as, or noninferior to, methylphenidate (both immediate- and extended-release),
immediate-release methylphenidate, but less effective than os- dexamfetamine, tricyclic antidepressants or no treatment (sec-
motically released methylphenidate or extended-release mixed tion 6), demonstrating a cost per QALY that was below com-
amfetamine salts (section 4.2). However, the extended-release monly accepted cost-effectiveness thresholds.
mixed amfetamine salts study[26] and one of the osmotically Several of the most commonly occurring adverse events with
released methylphenidate studies[51] were of only3 weeks’ du- atomoxetine administration were generally consistent with in-
ration; full benefits of atomoxetine often take several weeks creased noradrenergic tone (e.g. somnolence, vomiting) [section 5];
(potentially up to 8 weeks)[8] to occur.[2,8,63,64] Studies involving other common adverse events included headache and decreased
a longer treatment period are required before conclusions can appetite. Adverse events were generally classified as mild or
be drawn in these comparisons. Also, one of the immediate- moderate, and serious adverse events were rare. An increase in
release methylphenidate studies[25] was not powered for a direct suicidal ideation among atomoxetine versus placebo recipients led
comparison, and should only be used as supporting data for the to a black-box warning for children and adolescents (section 5.1).
other immediate-release methylphenidate study.[24] Rarely, severe liver injury may also occur in atomoxetine re-
The studies had other potential limitations. Three of the cipients, with three reports of liver-related adverse events deemed
active comparator trials were open-label,[25,50,51] and only one probably related to atomoxetine treatment in the 4 years following
was placebo controlled.[23] The comparison of atomoxetine its market launch. There have also been reports of sudden death in
with extended-release mixed amfetamine salts[26] was con- patients with serious structural cardiac abnormalities and other
ducted in a laboratory classroom, and thus different from a serious heart problems receiving atomoxetine. Discontinuation of
‘normal’ classroom atmosphere; all children had ADHD (ra- atomoxetine is well tolerated, with a low incidence of dis-
ther than a small proportion), and the number of observers and continuation-emergent adverse events and no evidence of symp-
staff was higher than in an average non-laboratory classroom. tom rebound (sections 4.1.1 and 5). In contrast, there has been
This study also excluded patients with the inattentive subtype some anecdotal evidence of symptom rebound and discontinua-
of ADHD, thus not allowing comparison in these patients.[26] tion-emergent adverse events following withdrawal of stimulants,
The two studies investigating atomoxetine versus osmotically with dexamfetamine potentially causing more rebound than
released methylphenidate both excluded, for ethical reasons, methylphenidate.[80]
patients with no response to previous methylphenidate treat- Slight differences were noted in the adverse events profiles of
ment, thus introducing a possible methylphenidate-slanted atomoxetine and extended-release stimulants, probably reflecting
bias.[23,51] Patients with tic[23,51] or anxiety disorders[23] were differences in their mechanisms of action (e.g. atomoxetine ap-
also excluded, again for ethical reasons, which would also po- peared more commonly associated with somnolence, osmotically
tentially bias results in favor of methylphenidate (which is released methylphenidate with insomnia). The adverse event
contraindicated in these disorders), as atomoxetine may po- profile of atomoxetine also differed somewhat to that of im-
tentially be more beneficial in these populations. mediate-release methylphenidate; for example, atomoxetine was
Interestingly, the significant difference in efficacy observed associated with a significantly higher incidence of somnolence,
between atomoxetine and osmotically released methylphenidate nausea, vomiting, anorexia, and dizziness (section 5).
was not present when only the stimulant-naive patients in the In contrast, stimulants are commonly associated with appetite
population of one study were investigated;[23] stimulant-exposed suppression, stomach pain, insomnia, and weight loss.[4] There is
patients still demonstrated a significant treatment difference also a possible potential for tic disorders and growth effects, as
(section 4.2). This may reflect study design, as the trial excluded well as potential cardiovascular problems; the US prescribing
patients with no response to previous methylphenidate treatment. information for mixed amfetamine salts includes a black-box
Overall, the impact of atomoxetine treatment on HR-QOL warning regarding cardiovascular disorders, and other stimulants
appears positive. In shorter term, placebo-controlled trials, carry contraindications and warnings for patients with cardio-
HR-QOL either stayed constant or was improved, and longer vascular conditions.[4] However, with stimulants only two non-
term, placebo-controlled trials also demonstrated positive ef- fatal adverse cardiovascular events occur per million
fects (sections 4.1.1, 4.1.2, and 4.1.3). HR-QOL was also im- prescriptions, and less than one death occurs per million.[80] Sti-
proved to a greater extent with atomoxetine than with standard mulant treatment effects on growth is controversial.[80] There
current therapy (section 4.2). is evidence that growth is affected by many factors, and thus it is

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 223

possible that differences in height and weight are the result of baseline did not significantly differ between groups after »8[84]
other causes; however, the possibility cannot be ruled out that and 10[85] months’ therapy.[84,85] However, a significant
stimulants are associated with significantly stunted growth, as (p = 0.017) loss of benefit (in terms of ADHD-RS total scores)
several meta-analyses and trials appear to demonstrate.[80] versus baseline was shown in patients receiving a lower dosage;
Statistically, but not clinically, significant increases in heart patients receiving a higher atomoxetine dosage did not differ
rate and BP occur with atomoxetine administration (section 2.2), significantly from baseline in this measure.[84]
with few patients experiencing increases in heart rate or BP to Very young children (aged under 6 years) with ADHD have
abnormal levels (section 5.1); however, additional longer term not been extensively studied.[3] The majority of methylphenidate
data would be of use in expanding our knowledge of the cardiac studies have included school-age children; very few have involved
effects of atomoxetine. There is also an initial loss in expected preschoolers, and, despite an FDA warning against administra-
weight and height among atomoxetine recipients, although both tion to children under the age of 6 years, methylphenidate is often
measures return to expected measurements after a period of time prescribed off-label for these patients.[3] Atomoxetine appears
(section 2.2). However, it has been suggested that a weakness of potentially effective in this patient group. One noncomparative
this study was that growth was only monitored sporadically.[80] pilot study involving 22 children with ADHD, aged 5–6 years,
As atomoxetine is mainly metabolized via the CYP2D6 en- demonstrated that atomoxetine to a maximum dosage of
zymatic pathway, it remains in the system of poor metabolizers 1.8 mg/kg/day was generally effective with regard to core ADHD
of CYP2D6 substrates for longer than extensive metabolizers, symptoms.[86] A large, randomized, placebo-controlled study of
and at higher concentrations (section 3); thus, these patients are atomoxetine is ongoing in children with ADHD aged 5–6 years.[3]
more prone than extensive metabolizers to the adverse events Although atomoxetine has been extensively researched, addi-
associated with atomoxetine (section 5). However, this can be tional data are needed. Of interest would be a well designed, long-
managed satisfactorily by decreasing the dosage administered term safety study, prospectively investigating, for example, the
to these patients (section 7). A recent study has, interestingly, weight and height differences of atomoxetine recipients com-
shown that genotyping is not necessary to safely administer pared with placebo and active comparators. A greater number of
atomoxetine: investigators in the trial were able to administer trials to augment the discoveries in patients with co-morbid dis-
atomoxetine to similar efficacy and safety levels regardless of orders would also be desirable, in particular trials specifying
metabolizer status, without prior knowledge of whether the ADHD symptom measures as their primary endpoint, as would
patients were extensive or poor metabolizers.[81] more studies of atomoxetine in very young patients.
A common co-morbid problem among pediatric patients In conclusion, atomoxetine is an effective and generally well
with ADHD is disordered sleep, specifically reduced sleep time tolerated option for the treatment of ADHD in children and
and more total interrupted sleep time, as well as more daytime adolescents, and is not classified as a stimulant. Atomoxetine
sleepiness and difficulty getting up, compared with healthy showed efficacy in children and adolescents with ADHD in well
controls.[82] Atomoxetine appears less likely than methylphe- designed placebo-controlled trials, with mixed results seen in
nidate to exacerbate disordered sleep in pediatric patients with active comparator trials; atomoxetine did not differ significantly
ADHD (section 2.2). from or was noninferior to immediate-release methylphenidate
Higher dosages of atomoxetine are not associated with a and was significantly less effective than osmotically released
greater improvement in efficacy, and thus there appears to be no methylphenidate or extended-release mixed amfetamine salts. It
advantage to increasing atomoxetine dosages beyond the current can be administered either as a single daily dose or split into two
guidelines. Two randomized, double-blind studies involving evenly divided doses, has a negligible risk of abuse or misuse, and
atomoxetine nonresponders aged 6–16 years demonstrated no is not a controlled substance in the US. Atomoxetine is particu-
significant difference in ADHD-RS total scores between re- larly useful for patients at risk of substance abuse or misuse, as
cipients of atomoxetine 1.2 or 1.8 mg/kg/day (usual dosage) and well as those who have co-morbid anxiety or tics, or who do not
those receiving the increased dosage of 3.0 or 2.4 mg/kg/day.[83] wish to take a controlled substance. Thus, atomoxetine is a useful
The potential for lower (off-label) target dosages in the option in the treatment of ADHD in children and adolescents.
longer term treatment of ADHD has been investigated. In two
studies, responders to short-term atomoxetine treatment were Disclosure
randomized to lower (0.5[84] or 0.8[85] mg/kg/day) versus higher The preparation of this review was not supported by any external
(1.2–1.8[84] or 1.4[85] mg/kg/day) dosages of atomoxetine. Both funding. During the peer review process, the manufacturer of the agent
showed that the mean change in ADHD-RS total score from under review was offered an opportunity to comment on this article.

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
224 Garnock-Jones & Keating

Changes resulting from any comments received were made on the basis of 22. Wernicke JF, Faries D, Girod D, et al. Cardiovascular effects of atomoxetine in
scientific and editorial merit. children, adolescents, and adults. Drug Saf 2003; 26 (10): 729-40
23. Newcorn JH, Kratochvil CJ, Allen AJ, et al. Atomoxetine and osmotically
released methylphenidate for the treatment of attention deficit hyperactivity
disorder: acute comparison and differential response. Am J Psychiatry 2008;
References 165 (6): 721-30
1. Greydanus DE, Pratt HD, Patel DR. Attention deficit hyperactivity disorder
across the lifespan: the child, adolescent, and adult. Dis Mon 2007; 53 (2): 70-131 24. Wang Y, Zheng Y, Du Y, et al. Atomoxetine versus methylphenidate in pae-
diatric outpatients with attention deficit hyperactivity disorder: a random-
2. Katragadda S, Schubiner H. ADHD in children, adolescents, and adults. Prim
ized, double-blind comparison trial. Aust NZ J Psychiatry 2007 Mar; 41 (3):
Care 2007 Jun; 34 (2): 317-41
222-30
3. Greenhill LL, Posner K, Vaughan BS, et al. Attention deficit hyperactivity
25. Kratochvil CJ, Heiligenstein JH, Dittmann R, et al. Atomoxetine and me-
disorder in preschool children. Child Adolesc Psychiatr Clin North Am 2008;
thylphenidate treatment in children with ADHD: a prospective, randomized,
17 (2): 347-66
open-label trial. J Am Acad Child Adolesc Psychiatry 2002 Jul; 41 (7): 776-84
4. Smoot LC, Boothby LA, Gillett RC. Clinical assessment and treatment of
26. Wigal SB, McGough JJ, McCracken JT, et al. A laboratory school comparison
ADHD in children. Int J Clin Pract 2007; 61 (10): 1730-8
of mixed amphetamine salts extended release (Adderall XR) and atomoxetine
5. American Psychiatric Association. Diagnostic and statistical manual of mental (Strattera) in school-aged children with attention deficit/hyperactivity dis-
disorders. 4th ed., text revision. Washington, DC: American Psychiatric order. J Atten Disord 2005 Aug; 9 (1): 275-89
Association, 2000
27. Heil SH, Holmes HW, Bickel WK, et al. Comparison of the subjective, phy-
6. Biederman J, Faraone SV. Attention-deficit hyperactivity disorder. Lancet siological, and psychomotor effects of atomoxetine and methylphenidate in
2005; 366 (9481): 237-48 light drug users. Drug Alcohol Depend 2002 Jul 1; 67 (2): 149-56
7. Ray GT, Levine P, Croen LA, et al. Attention-deficit/hyperactivity disorder in 28. Lile JA, Stoops WW, Durell TM, et al. Discriminative-stimulus, self-reported,
children: excess costs before and after initial diagnosis and treatment cost performance, and cardiovascular effects of atomoxetine in methylphenidate-
differences by ethnicity. Arch Pediatr Adolesc Med 2006 Oct; 160 (10): 1063-9 trained humans. Exp Clin Psychopharmacol 2006 May; 14 (2): 136-47
8. Harpin VA. Medication options when treating children and adolescents with 29. Spencer TJ, Kratochvil CJ, Sangal RB, et al. Effects of atomoxetine on growth
ADHD: interpreting the NICE guidance 2006. Arch Dis Child Educat Pract in children with attention-deficit/hyperactivity disorder following up to five years
2008; 93: 58-65 of treatment. J Child Adolesc Psychopharmacol 2007 Oct; 17 (5): 689-700
9. Blick SK, Keating GM. Lisdexamfetamine. Paediatr Drugs 2007; 9 (2): 129-35 30. Spencer TJ, Trzepacz PT, Witte MM, et al. Effects of atomoxetine on sexual
10. Eli Lilly and Company. Atomoxetine (Strattera) prescribing information [online]. development in children and adolescents with ADHD [abstract no. NR6-037].
Available from URL: http://pi.lilly.com/us/strattera-pi.pdf [Accessed 2009 Feb 9] 161st Annual Meeting of the American Psychiatric Association; 2008 May
11. Eli Lilly and Company. Atomoxetine (Strattera) summary of product char- 3-10; Washington, DC
acteristics [online]. Available from URL: http://emc.medicines.org.uk/printfriend 31. Sangal RB, Owens J, Allen AJ, et al. Effects of atomoxetine and methylphe-
lydocument.aspx?documentid=14482&companyid=46 [Accessed 2009 Feb 9] nidate on sleep in children with ADHD. Sleep 2006 Dec 1; 29 (12): 1573-85
12. Wong DT, Threlkeld PG, Best KL, et al. A new inhibitor of norepinephrine up- 32. Witcher JW, Long A, Smith B, et al. Atomoxetine pharmacokinetics in children
take devoid of affinity for receptors in rat brain. J Pharmacol Exp Ther 1982; and adolescents with attention deficit hyperactivity disorder. J Child Adolesc
222 (1): 61-5 Psychopharmacol 2003 Spring; 13 (1): 53-63
13. Barton J. Atomoxetine: a new pharmacotherapeutic approach in the man- 33. Witcher J, Kurtz D, Heathman M, et al. Population pharmacokinetic analysis
agement of attention deficit/hyperactivity disorder. Arch Dis Child 2005 Feb; of atomoxetine in pediatric patients [abstract no. OII-B-1]. Clin Pharmacol
90 Suppl. 1: i26-9 Ther 2004; 75 (2): 46
14. Ledbetter M. Atomoxetine: a novel treatment for child and adult ADHD. 34. Sauer JM, Ring BJ, Witcher JW. Clinical pharmacokinetics of atomoxetine.
Neuropsych Dis Treat 2006; 2 (4): 455-66 Clin Pharmacokinet 2005; 44 (6): 571-90
15. Bolden-Watson C, Richelson E. Blockade of newly-developed antidepressants 35. Chalon SA, Desager J-P, DeSante KA, et al. Effect of hepatic impairment on
of biogenic amine uptake into rat brain synaptosomes. Life Sci 1993; 52 (12): the pharmacokinetics of atomoxetine and its metabolites. Clin Pharmacol
1023-9 Ther 2003 Mar; 73 (3): 178-91
16. Bymaster FP, Katner JS, Nelson DL, et al. Atomoxetine increases extracellular 36. Belle DJ, Ernest CS, Sauer J-M, et al. Effect of potent CYP2D6 inhibition
levels of norepinephrine and dopamine in prefrontal cortex of rat: a potential by paroxetine on atomoxetine pharmacokinetics. J Clin Pharmacol 2002 Nov;
mechanism for efficacy in attention deficit/hyperactivity disorder. Neuro- 42 (11): 1219-27
psychopharmacology 2002 Nov; 27 (5): 699-711
37. Michelson D, Faries D, Wernicke J, et al. Atomoxetine in the treatment of
17. Cusack B, Nelson A, Richelson E. Binding of antidepressants to human brain children and adolescents with attention-deficit/hyperactivity disorder: a ran-
receptors: focus on newer generation compounds. Psychopharmacology domized, placebo-controlled, dose-response study. Pediatrics 2001 Nov;
(Berl) 1994; 114 (4): 559-65 108 (5): E83-91
18. Gehlert DR, Schober DA, Gackenheimer SL. Comparison of (R)-[3H]to- 38. Michelson D, Allen AJ, Busner J, et al. Once-daily atomoxetine treatment for
moxetine and (R/S)-[3H]nisoxetine binding in rat brain. J Neurochem 1995; children and adolescents with attention deficit hyperactivity disorder: a ran-
64 (6): 2792-800 domized, placebo-controlled study. Am J Psychiatry 2002 Nov; 159 (11):
19. Gehlert DR, Gackenheimer SL, Robertson DW. Localization of rat brain 1896-901
binding sites for [3H]tomoxetine, an enantiomerically pure ligand for nor- 39. Spencer T, Heiligenstein JH, Biederman J, et al. Results from 2 proof-of-
epinephrine reuptake sites. Neurosci Lett 1993; 157 (2): 203-6 concept, placebo-controlled studies of atomoxetine in children with attention-
20. Zerbe RL, Rowe H, Enas GG, et al. Clinical pharmacology of tomoxetine, deficit/hyperactivity disorder. J Clin Psychiatry 2002 Dec; 63 (12): 1140-7
a potential antidepressant. J Pharmacol Exp Ther 1985; 232 (1): 139-43 40. Gau SS, Huang YS, Soong WT, et al. A randomized, double-blind, placebo-
21. Gilbert DL, Ridel KR, Sallee FR, et al. Comparison of the inhibitory and controlled clinical trial on once-daily atomoxetine in Taiwanese children and
excitatory effects of ADHD medications methylphenidate and atomoxetine adolescents with attention-deficit/hyperactivity disorder. J Child Adolesc
on motor cortex. Neuropsychopharmacology 2006 Feb; 31 (2): 442-9 Psychopharmacol 2007 Aug; 17 (4): 447-60

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
Atomoxetine: A Review 225

41. Brown RT, Perwien A, Faries DE, et al. Atomoxetine in the management 59. Svanborg P, Thernlund G, Gustafsson PA, et al. Efficacy and safety of ato-
of children with ADHD: effects on quality of life and school functioning. moxetine as add-on to psychoeducation in the treatment of attention defi-
Clin Pediatr (Phila) 2006 Nov; 45 (9): 819-27 cit/hyperactivity disorder: a randomized, double-blind, placebo-controlled
42. Weiss M, Tannock R, Kratochvil C, et al. A randomized, placebo-controlled study in stimulant-naive Swedish children and adolescents. Eur Child Adolesc
study of once-daily atomoxetine in the school setting in children with ADHD. Psychiatry 2009 Apr; 18 (4): 240-9
J Am Acad Child Adolesc Psychiatry 2005 Jul; 44 (7): 647-55 60. Escobar R, Montoya A, Polavieja P, et al. Evaluation of patients’ and parents’
43. Kelsey DK, Sumner CR, Casat CD, et al. Once-daily atomoxetine treatment quality of life in a randomized placebo-controlled atomoxetine study in
for children with attention-deficit/hyperactivity disorder, including an as- ADHD. J Child Adolesc Psychopharmacol. In press
sessment of evening and morning behavior: a double-blind, placebo- 61. Michelson D, Read HA, Ruff DD, et al. CYP2D6 and clinical response to
controlled trial. Pediatrics 2004 Jul; 114 (1): e1-8 atomoxetine in children and adolescents with ADHD. J Am Acad Child
44. Wilens TE, Kratochvil C, Newcorn JH, et al. Do children and adolescents with Adolesc Psychiatry 2007 Feb; 46 (2): 242-51
ADHD respond differently to atomoxetine? J Am Acad Child Adolesc Psy- 62. Kratochvil CJ, Milton DR, Vaughan BS, et al. Acute atomoxetine treatment of
chiatry 2006 Feb; 45 (2): 149-57 younger and older children with ADHD: a meta-analysis of tolerability and
45. Wernicke JF, Adler L, Spencer T, et al. Changes in symptoms and adverse efficacy. Child Adolesc Psychiatry Ment Health 2008; 2 (25): 1-9
events after discontinuation of atomoxetine in children and adults with at- 63. Montoya A, Escobar R, Gilaberte I, et al. Continued improvement in newly
tention deficit/hyperactivity disorder: a prospective, placebo-controlled as- diagnosed children and adolescents with attention-deficit/hyperactivity dis-
sessment. J Clin Psychopharmacol 2004 Feb; 24 (1): 30-5 order: a randomized, double-blind, placebo-controlled trial of atomoxetine
46. Buitelaar JK, Michelson D, Danckaerts M, et al. A randomized, double-blind [abstract no. poster]. 8th congress of the European Society for Child and
study of continuation treatment for attention-deficit/hyperactivity disorder Adolescent Psychiatry; 2007 Aug 25-29; Florence
after 1 year. Biol Psychiatry 2007 Mar 1; 61 (5): 694-9 64. Pliszka SR, Crismon ML, Hughes CW, et al. The Texas Children’s Medication
47. Michelson D, Buitelaar JK, Danckaerts M, et al. Relapse prevention in pe- Algorithm Project: revision of the algorithm for pharmacotherapy of atten-
diatric patients with ADHD treated with atomoxetine: a randomized, double- tion-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 2006
blind, placebo-controlled study. J Am Acad Child Adolesc Psychiatry 2004 Jun; 45 (6): 642-57
Jul; 43 (7): 896-904 65. Bangs ME, Jin L, Zhang S, et al. Hepatic events associated with atomoxetine
48. Kratochvil CJ, Wilens TE, Greenhill LL, et al. Effects of long-term atomox- treatment for attention-deficit hyperactivity disorder. Drug Saf 2008; 31 (4):
etine treatment for young children with attention-deficit/hyperactivity dis- 345-54
order. J Am Acad Child Adolesc Psychiatry 2006 Aug; 45 (8): 919-27
66. Bangs ME, Tauscher-Wisniewski S, Polzer J, et al. Meta-analysis of suicide-
49. Wilens TE, Newcorn JH, Kratochvil CJ, et al. Long-term atomoxetine treat- related behavior events in patients treated with atomoxetine. J Am Acad
ment in adolescents with attention-deficit/hyperactivity disorder. J Pediatr Child Adolesc Psychiatry 2008 Feb; 47 (2): 209-18
2006 Jul; 149 (1): 112-9
67. Greenhill LL, Newcorn JH, Gao H, et al. Effect of two different methods of
50. Prasad S, Harpin V, Poole L, et al. A multi-centre, randomised, open-label initiating atomoxetine on the adverse event profile of atomoxetine. J Am
study of atomoxetine compared with standard current therapy in UK children Acad Child Adolesc Psychiatry 2007 May; 46 (5): 566-72
and adolescents with attention-deficit/hyperactivity disorder (ADHD). Curr
68. Mosholder AD, Gelperin K, Hammad TA, et al. Hallucinations and other
Med Res Opin 2007 Feb; 23 (2): 379-94
psychotic symptoms associated with the use of attention-deficit/hyperactivity
51. Kemner JE, Starr HL, Ciccone PE, et al. Outcomes of OROS methylpheni- disorder drugs in children. Pediatrics 2009 Feb; 123 (2): 611-6
date compared with atomoxetine in children with ADHD: a multicenter,
69. Cottrell S, Tilden D, Robinson P, et al. A modeled economic evaluation
randomized prospective study. Adv Ther 2005 Sep 31; 22 (5): 498-512
comparing atomoxetine with stimulant therapy in the treatment of children
52. Geller D, Donnelly C, Lopez F, et al. Atomoxetine treatment for pediatric with attention-deficit/hyperactivity disorder in the United Kingdom. Value
patients with attention-deficit/hyperactivity disorder with comorbid anxiety Health 2008 May 30; 11 (3): 376-88
disorder. J Am Acad Child Adolesc Psychiatry 2007 Sep; 46 (9): 1119-27
70. Tilden D, Richardson R, Nyhus K, et al. A modelled economic evaluation of
53. Kratochvil CJ, Newcorn JH, Arnold LE, et al. Atomoxetine alone or combined atomoxetine (Strattera) for the treatment of three patient groups with at-
with fluoxetine for treating ADHD with comorbid depressive or anxiety tention deficit hyperactivity disorder [abstract no. PMH5]. Value Health 2005
symptoms. J Am Acad Child Adolesc Psychiatry 2005 Sep; 44 (9): 915-24 Dec 6; 8: A197. Plus poster presented at the International Society for Phar-
54. Bangs ME, Hazell P, Danckaerts M, et al. Atomoxetine for the treatment of macoeconomics and Outcomes Research Eighth Annual European Congress,
attention-deficit/hyperactivity disorder and oppositional defiant disorder. 2005, Florence
Pediatrics 2008 Feb; 121 (2): e314-20
71. Laing A, Cottrell S, Robinson P, et al. A modelled economic evaluation
55. Bangs ME, Emslie GJ, Spencer TJ, et al. Efficacy and safety of atomoxetine in comparing atomoxetine with current therapies for the treatment of children
adolescents with attention-deficit/hyperactivity disorder and major depres- with attention deficit/hyperactivity disorder in The Netherlands [abstract no.
sion. J Child Adolesc Psychopharmacol 2007 Aug; 17 (4): 407-20 PMH10]. Value Health 2005 Dec 6; 8: A198. Plus poster presented at the
56. Arnold LE, Aman MG, Cook AM, et al. Atomoxetine for hyperactivity in International Society for Pharmacoeconomics and Outcomes Research
autism spectrum disorders: placebo-controlled crossover pilot trial. J Am Eighth Annual European Congress, 2005, Florence
Acad Child Adolesc Psychiatry 2006 Oct; 45 (10): 1196-205 72. Diamantopoulos A, Lorenzo M, Happich M, et al. A cost-utilty analysis of
57. Allen AJ, Kurlan RM, Gilbert DL, et al. Atomoxetine treatment in children atomoxetine against current stimulant therapies for the treatment of children
and adolescents with ADHD and comorbid tic disorders. Neurology 2005 with attention deficit/hyperactivity disorder in Germany [abstract no.
Dec 27; 65 (12): 1941-9 PMH36]. Value Health 2007 Nov 31; 10: A298. Plus poster presented at the
International Society for Pharmacoeconomics and Outcomes Research Tenth
58. Eli Lilly. Clinical study summary: study B4Z-MC-LYAC. A phase 3 random-
Annual European Congress, 2007, Dublin
ized, double-blind, placebo-controlled efficacy and safety comparison of
fixed-dose ranges (mg/kg/day) of tomoxetine with placebo in child and ado- 73. Myrén K-J, Poole L, Svanborg P. Atomoxetine’s effect on societal costs in
lescent outpatients with ADHD, aged 8 to 18 years. [online]. Available from Sweden: a 10 week double blind randomized placebo control clinical trial
URL: http://pdf.clinicalstudyresults.org/documents/company-study_163_0. [poster]. Economic and Social Commission for Asia and the Pacific. 2007 Aug
pdf [Accessed 2009 Feb 9] 26, Florence

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)
226 Garnock-Jones & Keating

74. Pliszka S. Practice parameter for the assessment and treatment of children and 82. Owens J, Sangal B, Sutton V, et al. Subjective and objective measures of sleep in
adolescents with attention-deficit/hyperactivity disorder. J Am Acad Child children with attention-deficit/hyperactivity disorder. Sleep Med. Epub 2008
Adolesc Psychiatry 2007 Jul; 46 (7): 894-921 Aug 6
75. Remschmidt H. Global consensus on ADHD/HKD. Eur Child Adolesc Psy- 83. Kratochvil CJ, Michelson D, Newcorn JH, et al. High-dose atomoxetine
chiatry 2005 May; 14 (3): 127-37 treatment of ADHD in youths with limited response to standard doses. J Am
Acad Child Adolesc Psychiatry 2007 Sep; 46 (9): 1128-37
76. Dopheide JA. ASHP therapeutic position statement on the appropriate use of
medications in the treatment of attention-deficit/hyperactivity disorder in 84. Newcorn JH, Michelson D, Kratochvil CJ, et al. Low-dose atomoxetine for
pediatric patients. Am J Health Syst Pharm 2005; 62 (14): 1502-9 maintenance treatment of attention-deficit/hyperactivity disorder. Pediatrics
2006 Dec; 118 (6): e1701-6
77. Findling RL. Evolution of the treatment of attention-deficit/hyperactivity
disorder in children: a review. Clin Ther 2008 May; 30 (5): 942-57 85. Wietecha L, Williams D, Herbert M, et al. Adolescent ADHD: atomoxetine
maintenance effects on age-appropriate developmental/adaptive functioning
78. Simpson D, Plosker GL. Atomoxetine: a review of its use in adults with [abstract; poster]. 54th Annual Meeting of the American Academy of Child
attention deficit hyperactivity disorder. Drugs 2004; 64 (2): 205-22 and Adolescent Psychiatry; 2007 Oct 23-28; Boston (MA)
79. Wolraich ML, McGuinn L, Doffing M. Treatment of attention deficit hyper- 86. Kratochvil CJ, Vaughan BS, Mayfield-Jorgensen ML, et al. A pilot study of
activity disorder in children and adolescents: safety considerations. Drug Saf atomoxetine in young children with attention-deficit/hyperactivity disorder.
2007; 30 (1): 17-26 J Child Adolesc Psychopharmacol 2007 Apr; 17 (2): 175-85
80. Graham J, Coghill D. Adverse effects of pharmacotherapies for attention-
deficit hyperactivity disorder: epidemiology, prevention and management.
CNS Drugs 2008; 22 (3): 213-37 Correspondence: Karly P. Garnock-Jones, Wolters Kluwer Health | Adis,
81. Trzepacz PT, Williams DW, Feldman PD, et al. CYP2D6 metabolizer status 41 Centorian Drive, Private Bag 65901, Mairangi Bay, North Shore 0754,
and atomoxetine dosing in children and adolescents with ADHD. Eur Neu- Auckland, New Zealand.
ropsychopharmacol 2008 Feb; 18 (2): 79-86 E-mail: demail@adis.co.nz

ª 2009 Adis Data Information BV. All rights reserved. Pediatr Drugs 2009; 11 (3)

You might also like