`
`n=73
`
`n=44
`
`'.
`2
`
`2.3%_(1)
`9.6% (7L
`
`23ng)
`5.5% (4)
`23% (1)
`273/042)
`2.5
`
`
`1.4% (1 )_333.2: 23% i1)_
`
`Abdominal oain
`
`10.4%411)
`6.7% £1.51)
`Nausea
`anti/£19)
`4.1%(11)
`Pain
`1.2% (2)
`3.0%:(8)
`Constipation
`0.5% 1
`1.9% (5)
`,
`Sleep disorder
`From Sponsor‘s Table ISS.A10.8, pp.2270-2276.
`
`‘
`
`AEs from Child and Adolescent Placebo Controlled BID ADHD Database where the
`relative risk was >2 in at least one of the origin (race) groups and there was at least a
`two fold difference when comparin- the relative risks between orioin race carous
`Event
`Risk in Caucasians _Iil§n
`
`Im-
`
`(n=267)_
`(3:153
`_
`Emotional Labil
`6.0% 16
`1.2% 2
`>
`1%
`
`3.7% 10
`0.6%1
`_ .«
`
`:"=
`r.
`21.7% 58
`20.2% 33
`
`
`’_
`
`‘
`
`1.4% 1
`15.1% 11
`
`6.8% 3
`6.8% 3
`
`Adult ADHD Placebo Controlled Studies
`
`There were occasional differences in the RR for AEs in this stratified analysis. Using the
`criteria from above, the AEs with RRs that differed by origin are listed below. These
`results are based on a small number of non-Caucasians (atomoxetine=22, placebo=27).
`
`AEs from Adult Placebo Controlled BID ADHD Database where the relative risk was >2
`
`in at least one of the origin (race) groups and there was at least a two fold difference
`when com-oarin the relative risks between oriin race orous
`
`
`
`
`
`ATX
`
`ATX
`
`
`
`
`(n=247
`(n=236
`_(n_=22
`
`
`
`
`4.5°/o(1_L
`.
`.
`10
`i9”
`Nausea
`13.0% 32
`
`
`
`12.1%(3_0)
`.
`.
`4.5% 1
`3.7% 1
`t
`
`,-
`‘:
`Asthenia
`4.9% 12
`3.0% 7
`
`
`1
`.
`‘f
`.
`9.9% 16
`lm-otence‘
`.—
`
`
`
`..
`"
`Abnl E'aculation’
`6.8% 11
`
`
`
`
`
`._’
`'
`2.4% (2)
`o smenorrhea°
`7.1% (6)
`From Sponsor's Table ISS.A10.11, pp.2289—2298.
`* In males, °ln females
`'
`
`8.3% 1
`
`5.3% 1
`10.5% 2
`
`10.0% (1)
`
`12.5% (1)
`
`
`
`0.7% 1
`1.3% 2
`
`.
`4.9.5 Lab Outliers by Origin (Race)
`The sponsor found no statistically significant differences for lab outliers from Child and
`Adolescent Placebo Controlled BID ADHD studies or Adult placebo controlled studies
`when stratified by gender (ISS p.591, 599).
`
`4.9.6 Vital Signs, Weight, and QTc by Origin (Race)
`There were no notable differences In Vitals sign changes when stratified by origin (race).
`The vital sign mean changes from baseline compared to placebo were similar for
`pediatric Caucasians compared to pediatric others and adult Caucasians compared to
`adult others (ISS p. 595, 602).
`
`The data corrected QTc mean change from baseline compared to placebo for pediatric
`Caucasians was 0.62 compared to 4.0 for pediatric Others (ISS, p.598). The data
`corrected QTc mean change from baseline compared to placebo for adult Caucasians
`was 0.37 compared to —5.3 for adult Others (ISS, p.602).
`
`85
`
`

`

`
`
` CLINlCAL SAFETY REVIEW , ND '
`
`4.9.7 Adverse Events by Age
`Child and Adolescent Placebo Controlled BID ADHD Studies
`
`There were few AEs with different relative risks when comparing pediatric subjects <12
`years old (atomoxetine 258, placebo 172) to those >12 years old (atomoxetine 82,
`placebo 35). Only Hostility met the criteria used in these analyses. The relative risk for
`Hostility was 4.5 for pediatric subjects <12 (atomoxetine 2.7%, 7/258, placebo 06%
`1/172), and no Hostility AEs were reported in subjects >12 (ISS, pp.2318—2325).
`
`Adult ADHD Placebo Controlled Studies
`
`There were occasional differences in the RRs for AEs in this stratified analysis. Using
`the criteria from above, the AEs with RRs that differed in adults by age are listed below.
`
`AEs from Adult Placebo Controlled BID ADHD Database where the relative risk was >2
`
`in at least one of the age groups and there was at least a twofold difference when
`comparing the relative risks between agegroups
`
`Event
`Risk in Those<42
`RRQZ
`Risk in Those>42
`RR)”
`ATX
`PBO
`ATX
`PBO
`
`
`(n2128)
`(F146)
`(n=141)
`(n21?)
`
`Insomnia
`21.9% (28
`5.5% (8)
`.9
`4
`19.9% (28)
`12%15
`
`.
`5.1
`Dry mouth
`21.1% (27
`4.1%_(6)
`21.3% Q0
`10.3% 12
`
`
`.',
`.
`“21.3 ‘::
`Nausea
`8.6% 11
`6.8% 10
`15.6% 22
`2.6% 3
`
`iAnorexia
`8.5% 12 _Il * -’
`
`
`
`
`.
`7 Pain
`4.7% 6
`4.8% (7
`9.9% (14
`4.3% 5
`
`
`
`.
`,
`“4.731(5)
`512%49
`7.1%(10)
`3.4% (4)
`3 D sooesia
`
`
`E
`.
`i‘
`2.47%
`9.2% 13
`4.7%6
`0.9% 1
`i Libido decreased
`
`
`
`Dizziness
`’
`-
`
`
`
`4.3%)(6)
`4
`'
`Abnormal dreams
`6.3°/o((8)
`
`
`
`
`5. 4% j g
`14.6% (12
`'i lmpotence’
`
`
`
`
`
`5.5% (5)
`0 9% (1
`2.8% (4)
`Fever
`3.9%_(5)
`
`
`
`
`2. Parasthesia 2. 5% 5 5% (7) 2 10/053) 28% (4
`
`
`
`
`
`
`E Abnormal e‘aculation
`i Vomitin.
`
`
`
`Chest -ain
`3.5% 5
`2.6% 3
`’; . I
`.
`
`
`21%3 '2 Tf-I--_
`139% (5) — "
`.i —_ . 4
`
`
`
`From Sponsor’s Table lSS.A10.17, pp.2339-2348.
`* In males, °|n females
`
`Chills
`
`4.9.8 Lab Outliers by Age
`The sponsor found a statistically significant difference only for low urine specific gravity
`lab outlier (higher odds ratio compared to placebo in the <12 year old group) from Child
`and Adolescent Placebo Controlled BID ADHD studies (lSS, p.609). In the Adult placebo
`controlled studies when stratified by age, the <42 year old group had a higher odds ratio
`compared to placebo for urinalysis occult blood (ISS p.618).
`
`4.9.9 Vital Signs, Weight, and (He by Age
`There were no notable differences in Vitals sign changes when stratified by age. The
`vital sign mean changes from baseline cempared to placebo were similar for pediatric
`subjects <12 compared to pediatric subjects >12 and adult subjects <42 compared to
`adult subjects >42 (ISS p. 611, 619).
`
`86
`
`

`

`
`
`
`
`ll?
`
`The data corrected QTc mean change from baseline compared to placebo for pediatric
`subjects <12 was 0.45 compared to 4.43 for pediatric subjects >12 (ISS, p.614). The
`data corrected QTc mean change from baseline compared to placebo for adult subjects
`<42 was ~2.82 compared to ~2.56 for adult subjects >42 (lSS, p.621).
`
`4.10 Drug interaction
`The sponsor conducted drug interaction studies and results from those studies are
`summarized below.
`
`HFBO— looked at atomoxetine given with salbutamol. The sponsor reported an increase
`in heart rate with the combination. For example, at 2 hours post dosing, single dose, the
`mean increase in heart rate for salbutamol alone was i9bpm, for atomoxetine alone was
`2bpm, and for the combination was 46bpm. The sponsor did not find evidence of
`increase in blood pressure or SVR with the combination of salbutamol and atomoxetine
`compared to the either agent alone.
`
`LYAP-Iooked at atomoxetine given with methyphenidate in EM subjects. The sponsor
`FGDOTTed that there is no statistical difference in HR with methylphenidate in the presence
`or absence of atomoxetine.
`
`E002 —looked at atomoxetine given with ethanol. This was a two period double blind
`cross over study in healthy volunteers. During the first study period subjects took either
`atomoxetine (40mg bid), or placebo for 5 days. Ethanol was administered (2ml/kg) on the
`5‘“ day followed by measurements of psychomotor function. Patients were crossed over
`after a 4-week washout. The sponsor reported that there was no evidence of a greater
`interaction with alcohol among PM subjects compared to EM subjects. Additionally, the
`sponsor reported ”Following dosing with 40mg bid for 5 days, EM and PM subjects
`showed no difference in psychomotor performance in the absence of alcohol compared
`to placebo.
`
`HFBL— looked at atomoxetine given with paroxetine to EM subjects. This single blind
`sequential study administered paroxetine 20mg qd with atomoxetine, and the sponsor
`reported a substantial effect on PK of atomoxetine. The combination also resulted in
`greater orthostatic tachycardia compared to paroxetine alone in EM subjects or
`atomoxetine alone in PM subjects (data from study LYAE). EM subjects taking the
`combination had a greater decrease in orthostatic systolic BP compared to EM subjects
`taking paroxetine alone. The decrease in orthostatic SBP in EM subjects taking
`paroxetine and atomoxetine was similar to the decrease in orthostatic SBP in PM
`
`subjects taking atomoxetine alone. The paroxetine—atomoxetine interaction vital sign
`results are summarized in the following table:
`
`87
`
`

`

` ”if; CLINICAL SAFETY REVIEW ;:s.;.;Np;§
`
`Table lSS.5.4.41.
`
`Comparison of Cardiovascular Variables (Least-Square
`Means) Averaged over Days 4 and 5 Between PM Subjects
`(Study LYAE) and EM Subjects on Paroxetine (Study HFBL)
`
`Least-Sguarc Means
`
`Paroxetine
`
`Placebo
`
`Least-Sguare Means
`Paroxetine and
`Tomoxctine
`
`Tomoxetine
`
`HFBL
`EM Subjectsa
`
`LYAE
`PM Subjects
`
`p-Valuc"
`(baseline)
`
`(20 mg BID)
`HFBL
`EM Subjects
`
`(30 mg BID)
`LYAE
`PM Subjects
`
`P-vaiue"
`(atomoxetine)
`
`79.5
`
`15.5
`
`71.5
`
`8.3
`
`0,072
`
`0.039
`
`l05.6
`
`33.9
`
`83.6
`
`18.6
`
`0.000]
`
`0.0006
`
`Variable
`SHR
`
`(bp;::
`OHR
`
`(bpm;
`088'?
`
`—S.5
`
`(mn: H2)
`
`~1.0
`0.068
`—16.4
`~l5.5
`0.82
`
`Abbfiziazions: Standing Heart Rate (SHR); Ortliostatic Heart Rate (OHR); Ortliostalic Systoiic BP (OSBP)
`B? = béood pressure, bpm = beats per minute. mmHg 2 millimeters of mercury. {Onliostatic refers to
`111: dit‘ierence between standing and supine measurements.)
`3 Befre administration oftomoxetine in Period 2.
`b P—xzéue ofthe indicated difi‘erence.
`
`Sourse Data: Data on file at Lilly Clinic.
`
`The sponsor reported iii/[Table ISS.5.'4.42, p.469, that during the co-administration
`period, there were more dizziness events (3.81% 4/105*) compared to paroxetine alone
`(0.68%, 1/147") or atomoxetine alone (0/175*).
`*#events/#intervals
`
`LYAJ- looked at atomoxetine given with midazolam. The sponsor reported that the
`combination did not result in more AEs than when atomoxetine was given alone.
`
`4.11 Overdose
`
`There were no large atomoxetine overdoses in human subjects in the ADHD
`development program. The sponsor identified 7 cases of overdose where patients took
`2-5 times the recommended dose. Five cases were dispensing mistakes and two were
`intentional. The five who were mistakenly dispensed the wrong amount of atomoxetine
`finished their respective studies. None of the 7 cases resulted in an SAE. Commonly
`reported symptoms among overdose subjects were anorexia, abdominal pain, and
`headache. The sponsor reported increases in heart rate but no changes in blood
`pressure or QTc in atomoxetine overdose patients (lSS, pp. 626-9).
`
`The sponsor described what appeared to be toxicity symptoms in a PM subject with the
`highest recorded serum concentration of atomoxetine in the NDA. That case was
`summarized above in the CP section and the summary is repeated here.
`
`Subject LYAE-1009 a 28-year-old PM male developed ataxia, myoclonlcjerking, of his legs at
`night and dizziness and had hyperreflexia on neurologic exam. He had been receiving
`atomoxetine 75mg bid (2.44mglkg/day) for five days and had a serum concentration of
`5596ng/mL, the highest recorded serum concentration. His symptoms resolved off atomoxetine
`over the next 48 hours.
`
`88
`
`

`

`
`
`4.12 Human Pregnancy
`There have been two pregnancies in subjects exposed to atomoxetine, both from the
`historical depression trials database. In one case, the subject was exposed for 6 weeks
`and when she had a positive pregnancy test, atomoxetine was stopped. The baby was
`born normal and healthy. The outcome of the second case is unknOWn (ISS, p.572)
`
`4.13 Withdrawal
`
`In the pediatric studies HFBD and HFBK, after 9 weeks of double blind treatment, the
`remaining atomoxetine subjects were observed on placebo for 1-week following abrupt
`atomoxetine discontinuation while placebo subjects continued their assigned treatment
`for that week. The sponsor compared AEs during this one—week discontinuation phase.
`There did not appear to be evidence of differences in the actual AEs or frequency of AEs
`reported during the discontinuation phase based on the small sample size (atomoxetine
`102, P80 92) (188, p.632). The atomoxetine discontinued group experienced decreases
`in DBP (-0.98mmHg), SBP (-O.43mmHg), and pulse (—3.8bpm) during the
`discontinuation phase compared to slight increases in these measurements in the
`placebo group (ISS, p.633). QTc, Fridericia and data corrected. decreased among
`atomoxetine withdrawn subjects (-O.157, —1 .505, respectively) and increased among
`placebo subjects (3.859. 4.442, respectively) (ISS, p.634)
`
`In the adult studies LYAA and LYAO, after the double blind treatment phase, 120mg/day
`and 90mg/day atomoxetine subjects were randomized to abrupt or tapered
`discontinuation and observed during a 4—week discontinuation phase. The taper reduced
`the dose by 30mg/week and then stopped atomoxetine once the subject reached
`60mg/day. The sponsor observed 73 abrupt and 94 tapered discontinuation subjects.
`There were no differences in the reasons for discontinuation during the discontinuation
`phase. One tapered and no abrupt d/c subjects discontinued during this phase for an
`AE. The following table summarizes selected AE risks by d/c assignment.
`
`Select AE risks following withdrawal of Atomoxetine, studies LYAA, LYAO
`Event
`Abrupt d/c (n=73)
`Taper (n=94)
`p—value
`Diainess
`5.5% (4)
`O
`.035
`Flu syndrome
`4.1% (3)
`2.1% (2)
`.654
`Somnolence
`4.1% (3)
`2.1% (2)
`.654
`Depression
`2.7% (2)
`1.1% (1)
`.581
`Insomnia
`2.7% (2)
`10.6% (10)
`.069
`
`During the discontinuation phase, both groups experienced decreases in SBP, DBP, and
`pulse with the declines in SBP and pulse greater among the abrupt d/c group. Both the
`abrupt dlc and the taper group experienced decreases in their data corrected QTc
`(-0.95, -1.25, respectively). (ISS, 636-43).
`
`4.14 Drug Disease Interaction
`The sponsor stated that hepatic impairment will have an impact on systemic exposure to
`atomoxetine since clearance is dependent on hepatic blood flow and hepatic function
`(482). Excretion of inactive metabolites in urine will not Change the effects of
`atomoxetine in patients withlend stage renal disease. The sponsor reported “Neither
`maximum serum concentrations nor total systemic exposure to atomoxetine parent
`differed significantly between patients with severe renal insufficiency and normal
`subjects” (p.482). Both the liver and hepatic insufficiency studies were single dose
`studies and therefore do not provide safety information about chronic dosing.
`
`89
`
`

`

`
`
`5. Review of Systems
`5.1 Cardiovascular
`
`There were no deaths due to cardiovascular causes in the atomoxetine development
`program.
`
`The sponsor identified the following four CV SAEs in atomoxetine subjects: murmur
`(HFBE-O23—0894), peripheral shutdown (LYAF-570-1882), unstable angina (LYAR-083—
`6419), and chest pain (LYAB-103-5786). A syncopal event (LYAB—063—5565) was listed
`under the neurological body system but will be considered here with the CV SAEs. A
`pediatric cardiologist evaluated the murmur SAE, and an echocardiogram documented a
`small patent foramen ovale not considered hemodynamically significant. The peripheral
`shutdown SAE narrative described symptoms of feeling cold 1/2-1 hour following
`atomoxetine dosing in a 9—year-old male. The subject was found to have “peripheral
`shutdown" in hands and feet and was admitted for observation. Atomoxetine was
`
`stopped with resolution of symptoms. The chest pain SAE involved a pediatric subject
`with a history of pulmonary artery stenosis and the patient improved following
`discontinuation. The unstable angina SAE was a 46 year old male with a history of
`hypefiension, moderate obesity and a family history of CAD. He developed chest pain
`and was diagnosed with unstable angina. He discontinued from the study. The syncopal
`event occurred in a 7-year—old EM male and was incompletely described. Work up
`included a normal EEG, normal labs, negative drug screen, and normal ECG and head
`CT. Apparently the subject experienced two additional syncopal episodes after stopping
`atomoxetine. Atomoxetine was subsequently restarted with no additional syncopal
`episodes.
`
`The historical database (depression, urinary incontinence trials) identified 22 CV SAEs in
`1,275 subjects. Extrasystole, hypertension, and tachycardia occurred in 3 subjects each.
`Atrial arrhythmia, bundle branch block, and syncope occurred in 2 subjects each. The
`following CV events were reported for one subject: angina, arrhythmia, ECG abnormal,
`hemorrhage, myocardial infarction, ST elevated, and vascular disorder. The narratives
`for these events provided little information about the events.
`
`CV AEs did not commonly lead to discontinuation from the pediatric or adult ADHD
`studies. Palpitation and tachycardia were the only CV AEs leading to discontinuation of
`more than two subjects in the pediatric phase ”It” ADHD trials. The narratives for these
`events did not suggest an association with dizziness or syncope. One pediatric subject
`discontinued for hypertension and had a baseline BP of 117/81mmHg and a highest
`recorded BP of 120/94mmHg. In the adult placebo controlled trials, chest pain (n=2) and
`palpitations (n=2) were the events leading to discontinuation of more that 1 subject.
`Neither palpitation event was associated with dizziness or syncope, but one case was
`associated with parasthesias. One adult discontinued for hypertension (baseline BP
`153/94mmHg. highest on study 141/99mmHg) and one for hypotension (baseline BP
`129/81 mmHg, on day of d/c 120/80mmHg).
`
`Four atomoxetine subjects discontinued from clinical pharmacology trials for syncope.
`Nan’ative descriptions for these events were summarized in the clinical pharmacology
`section of this review (n=2) and in the QT review section (n=2). All four subjects who
`experienced syncope were EMS and the range of atomoxetine doses was 10-60mg. Two
`subjects were also taking fluoxetine, a CYPZDG inhibitor, at the time of the event.
`Although there was insufficient evidence to be certain of the causes of these syncope
`events, the sponsor's narratives suggested orthostatic blood pressure and vasovagal
`
`90
`
`

`

`
`
`etiologies. In one case, the event occurred while standing for morning V8 and was
`associated with bradycardia and the subject felt faint prior to the event. One syncope
`event followed rising from a squatting position in a subject who may have been
`dehydrated from vomiting and diarrhea. Another syncope event occurred after voiding
`and was preceded by lightheadedness. The last event occurred during a fast and after
`sitting up for a blood draw. The sponsor identified 5 subjects (1.6%, 5/316) with syncope
`AEs within 12 hours of an atomoxetine dose in CP studies (ISS p.417). Although no
`events coded as postural hypotension led to discontinuation, 10 subjects (32%, 10/316)
`had a postural hypotension AE within 12 hours of an atomoxetine dose in GP studies
`(ISS p.417).
`
`In a post hoc analysis using a case definition for symptomatic orthostatic hypotension
`(SOH), 8 clinical pharmacology subjects experienced 17 SOH events. The sponsor
`reported that SOH was more common among PM subjects (13.3%, 4/30) than EM
`subjects (27%, 4/150). None of the SOH events occurred within 24 hours of first
`atomoxetine dose. Four subjects had events 48-60 hours after their first atomoxetine
`dose. Seven of the subjects had an SOH event at the 40-60mg dose range.
`
`In the historical depression placebo controlled trials, vasodilatation and tachycardia were
`the only CV AEs leading to discontinuation of at least 3 atomoxetine subjects and at
`least twice as frequent compared to placebo. In these studies, the risk for vasodilatation
`leading to d/c was 0.8% (9/1153) in the atomoxetine group compared to 0 in the placebo
`group (n=654). The risk/for tachycardia leading to d/c was 0.3% (4/1153) in the
`atomoxetine group compared to 0 in the placebo group (n=654). One atomoxetine
`subject (0.1%, 1/1153) and no placebo subjects discontinued for syncope.
`
`In the pediatric ADHD studies, no CV AEs occurred in at least 5% of subjects.
`Tachycardia was reported in 2.5% (n=49) subjects, chest pain in 2.3% (n=44) subjects,
`postural hypotension in 1.2% (n=23) subjects, hypertension in 0.7% (n=13) subjects and
`syncope in 0.4% (n=7) subjects. The following table summarizes the risks for these
`events in the placebo controlled pediatric ADHD bid trials allowing a comparison of risk
`by treatment.
`
`Risks for CV AEs reported in Pediatric and Adolescent Placebo Controlled ADHD
`studies using BID Dosing
`ATX N=340
`
`PBO N=207
`
`CV AE
`
`Postural Hypotension
`Tachycardia
`Chest Pain
`Hypertension
`Pal
`itation
`
`1.8% (6)
`1.5% (5)
`1.2% (4)
`0.6% (2)
`0.6% 2
`
`Syncope was not reported in this safety sub-group
`
`0.5% (1)
`0.5% (1)
`0
`0
`0.5% 1
`
`The risks for tachycardia (2.2% PM, 1.1% EM), syncope (1.1% PM, 0.2% EM), and
`hypertension (0.6% PM, 0.2% EM) were greater among pediatric PM subjects than EM
`subjects, suggesting a potential exposure level response relationship.
`
`In the Adult ADHD studies, no CV AEs occurred in at least 5% of subjects. Vasodilitation
`and palpitation were reported in 4.1% (n=11) of subjects, chest pain in 3% (n=8) of
`subjects, tachycardia in 3% (n=8) subjects, hypertension in 0.7% (n=2) subjects, and
`postural hypotension in 0.4% (n=1). Each of these CV events occurred at least twice as
`
`9]
`
`

`

`
`
` ’ CLINICAL SAFETY REVIEW;
`
`commonly among atomoxetine subjects compared to placebo subjects except for
`hypertension which occurred more frequently among placebo subjects. Syncope was not
`reported as a treatment emergent AE in these studies.
`
`The sponsor's analyses of vital sign data support an atomoxetine-related increase in
`systolic blood pressure, diastolic blood pressure and pulse. The following tables
`summarize the mean changes from the pediatric and adult ADHD placebo controlled BID
`groups.
`
`Diastolic Blood Pressure, Systolic Blood Pressure and Pulse Mean Change from
`Baseline to Endpoint, Child and Adolescent acute placebo controlled ADHD studies and
`
`Adult acute placebo controlled ADHD studies usin- BID dosin
`[ Parameter
`_L
`Treatment (9) j Mean Chance
`
`
`Child and Adolescent acuteplacebo controlled ADHD studies usin BID dosing__
`i Diastolic BP
`Atomoxetine (335)
`2.060
`.002
`Placebo (204)
`-O.453
`
`Systo!.: BP
`Atomoxeline (335)
`2.791
`Placebo (204)
`1.184
`
`Pulse
`Atomoxetine (335)
`7.816
`Placebo (20/1)
`1.532
`
`
`
`
`Adult acute pliebo controlled ADHD studies usino BID dosin-
`
`
`
`1.771
`Diastolic BP
`Atomoxetine (258)
`Placebo (258L
`
`Systolic BP
`Atomoxetine‘(258)
`Placebo (25§_)
`
`l Pulse
`,
`Atomoxetine (258)
`Placebo (258L
`
`
`
`
`
`The SBP, DBP and pulse mean changes observed in the once daily dose study were
`comparable to the changes illustrated above. The mean increases in diastolic blood
`pressure, systolic blood pressure, and pulse were higher among PM subjects compared
`to EM subjects suggesting a potential exposure level response relationship.
`
`The vital sign outlier analyses were consistent with the mean change analyses. In
`pediatric ADHD BID subjects, 19% of atomoxetine subjects had a DBP high outlier
`compared to 11% of placebo subjects (RR=1.7) and 18% of atomoxetine subjects had a
`SBP high outlier compared to 9% of placebo subjects (RR=2). Nine percent of
`atomoxetine subjects had a high pulse outlier compared to 4% of placebo subjects
`(RR=2.25).
`
`The relative risks for DBP and SBP outliers in the pediatric ADHD QD studies were
`similar to the risks in the pediatric BID studies. The data demonstrated a higher relative
`risk for high pulse outliers in the once daily dosed pediatric studies compared to the BID
`studies. In the QD studies, the risk for a high pulse outlier was 7.1% in the atomoxetine
`group and 1.2% in the placebo group, RR=5.9. In the pediatric BID studies the high
`pulse outlier RR was 2.25.
`
`In the adult studies, there was little difference in risk for high DBP between atomoxetine
`and P80 but 5% of atomoxetine and 3.5% of placebo subjects had a SBP high outlier
`(RR=1.4). Almost 11% of atomoxetine subjects had a high pulse outlier compared to 3%
`of placebo subjects (RR=3.7).
`
`92
`
`

`

` CLlNlCAL SAFETY REVIEW éNp
`
`
`
`PM subjects had higher mean increases in DBP, SBP, and pulse compared to EM
`subjects, although the high outlier risks were similar for both groups.
`
`Orthostatic blood pressure data were not collected during phase ”II” ADHD trials. The
`sponsor examined orthostatic blood pressure changes in GP multi-dose studies. At most
`of the doses and time points for PM subjects the orthostatic systolic blood pressure
`changes were more negative than placebo with the greatest difference —29.4mmHg,
`75mg dose pre dose. The orthostatic systolic blood pressure change differences
`between EM and placebo subjects were smaller with the greatest difference -12.6mmHg,
`40mg at 1—hour post dose (Table lSS.5.4.27, p.445). At most of the doses and time
`points for PM subjects the orthostatic diastolic blood pressure changes were more
`negative than placebo (greatest difference —18.2mmHg, 60mg dose at time 0). The
`orthostatic systolic blood pressure change differences between EM and placebo
`subjects were smaller with the greatest difference —9.2mmHg, 20mg at pre dose (Table
`68.54.31, p.449).
`
`The ECG data from a clinical pharmacology study demonstrated QTc prolongation in PM
`subjects at the 60 BlD and 75 BlD dosages with the greatest increases at the pre—dose
`time point for the 75mg BID dose. A second clinical pharmacology study that used
`iluoxetine to create ”phenotypic" PM subjects did not demonstrate QTc prolongation.
`The ECG data collected during the phase ll/lll studies did not find atomoxetine related
`QTc prolongation although the ECGs were not collected in the same careful manner
`used in the clinical pharmacology studies. Considering clinical events that could be
`potentially related to arrhythmia, there were no sudden deaths or documented adverse
`events of torsades de pointes, ventricular tachycardia or ventricular fibrillation in the
`development program. There were six atomoxetine subjects with convulsions and there
`were occasional syncopal episodes in atomoxetine subjects. Atomoxetine was
`associated with increased risk of dizziness and palpitations compared to placebo but any
`relationship between these events and effect on QTc is speculative. Pre-clinical data
`demonstrate that atomoxetine blocks lK, providing a potential mechanism for the QTc
`results observed in the clinical pharmacology study.
`
`5.2 Digestive System
`There were no deaths due to digestive causes in the atomoxetine development program.
`
`The sponsor identified 14 subjects with digestive system SAEs. Nine of these events
`were cases coded as appendicitis, two were acute abdominal pain, one was a GI
`infection, one was increased LFTs and the last was vomiting in a subject who took an
`atomoxetine overdose. The appendicitis cases were summarized above. One of the
`abdominal pain cases (LYAl—055-5048) occurred in an 11—year~old male who had a prior
`history of abdominal pain. The pain stopped after discontinuing atomoxetine but then
`later recurred off drug. The second abdominal pain SAE occurred in a 49-year-old
`female (LYAR-081-5952) and was attributed to diverticulitis although the subject was not
`treated with antibiotics making this diagnosis suspect. The GI infection case (LYAF-601-
`7009) was an 8-year-old male who developed vomiting and diarrhea and was admitted
`to work up a possible diagnosis of celiac disease. The increased LFT case (LYAF-652-
`9053) was discussed above and was notable for hives, confusion and ALT of 169 and
`AST of 136. LFTs improved off drug and hepatitis serology was reportedly negative.
`Subject 004-1125 a 13 year old male with a history of thalassemia minor took twice the
`prescribed amount of atomoxetine for 3 weeks and developed lightheadedness and
`intermittent vomiting that persisted for 1 week following atomoxetine discontinuation.
`
`93
`
`

`

`
` CLINICAL SAFETY REVIEW NDA721ri-47111%».r,'=.9:1. 2
`
`
`
`Liver function test abnormal (n=5) and nausea (n=2) were the only digestive SAE
`occurring in more than 1 atomoxetine subject enrolled in depression and urinary
`incontinence trials in the Historical database.
`
`Digestive AEs did not commonly lead to discontinuation from clinical trials for
`atomoxetine subjects. Nausea led to the discontinuation of 3 subjects in the pediatric
`ADHD trials using BlD dosing. Pediatric subject LYAB—084-4924 discontinued for
`abnormal LFTs. The highest recorded ALT for this subject was 87U and highest total
`bilirubin was 0.3mg/dL. ln the adult ADHD studies, no atomoxetine subjects discontinued
`for digestive AEs. Subject LYAT-023-3409 discontinued from the once daily dosed
`pediatric ADHD study for vomiting. This 6-year—old male developed nausea and vomiting
`after 14 days of atomoxetine. The sponsor did not provide an outcome in this sub,ect’s
`narrative.
`
`in the historical data base depression trials, nausea (0.5%, 6/1 ,153) and constipation
`(0.3%, 3/1 ,153), were the only AEs leading to discontinuation of more than 1
`atomoxetine subject.
`
`Digestive AEs were commonly reported by atomoxetine subjects in the ADHD
`development program and data from placebo controlled trials suggest a relationship
`between atomoxetine and these AEs. Abdominal pain (21%, 411/1933) anorexia (15%,
`282/1,933), vomiting (14%, 272/1,933) nausea (11%, 212/1,933), diarrhea (6%,
`123/1 ,933) and constipation (5%, 96/1933) were the most commonly reported digestive
`AEs in the pediatric and adolescent BlD ADHD studies. Anorexia, dyspepsia,
`constipation, weight loss, and gastroenteritis occurred in at least 1% of atomoxetine
`subjects and at least twice as frequently compared to placebo on the pediatric ADHD
`BID placebo controlled trials. In the adult ADHD BlD placebo controlled trials, nausea,
`anorexia, constipation, flatulence, and rectal disorder occurred in at least 1% of
`atomoxetine subjects and at least twice as frequently compared to placebo. in the
`pediatric once daily ADHD study, anorexia, abdominal pain, vomiting, nausea,
`dyspepsia, and diarrhea occurred in at least 1% of atomoxetine subjects and at least
`twice as frequently compared to placebo. Data from study LYAC, the pediatric ADHD
`BID study that randomized to fixed doses suggests a dose response for 3 digestive AEs.
`Those data are provided below.
`
`Event
`
`Selected LYAC Adverse Events by Dose (mg/kg/day)*
`ATX 0.5 n=44
`ATX 1.2 n=84
`ATX1.8 n=83
`
`,
`PBO n=83
`
`Abdominal Pain
`11.4% (n=5)
`14.3% (n=12)
`14.5% (n=5)
`10.8% (n=9)
`Anorexia
`6.8% (n=3)
`11.9% (n=10)
`12% (n=10)
`4.8% (n=4)
`
`
`
`n=5 Vomitin . .
`
`The analyses of lab data did not suggest a relationship between atomoxetine and
`increases in transaminases or total bilirubin.
`
`5.3 Hemic and Lymphatic System
`There were no deaths due to hemic or lymphatic causes in the atomoxetine
`development program. There were no hemic or lymphatic SAEs in the ADHD database.
`in the historical database there was one leukopenia SAE and one leukemia SAE.
`Hemic and Lymphatic System AEs did not commonly lead to discontinuations from
`ADHD trials. One pediatric subject (LYAB-064-5616) discontinued for lymphocytosis,
`which was diagnosed as mononucleosis. No adult subjects discontinued for hemic or
`
`94
`
`

`

`
` CLINICAL SAFETY REVIEW
`
`
` ,NDA’V21-41gz1; . "
`
`lymphatic system AEs. No subjects discontinued for hemic or lymphatic system AEs in
`the historical database depression trials.
`
`Ecchymosis (2 6% 51/1 ,933) was the only Hemic and Lymphatic System AEs occurring
`in more than 1% of pediatric ADHD subjects from BID trials. Anemia (0. 5%, 10/1 9,33)
`and leukopenia (0 5%, 10/1,933) were rarely reported in these studies. Data from
`placebo controlled pediatric trials did not suggest increased risks for hemic and
`lymphatic system AEs among atomoxetine subjects.
`
`Aside from small mean increases in platelet counts in atomoxetine subjects but not
`present in placebo subjects, lab data analyses did not suggest atomoxetine-related
`changes in hematologic parameters.
`
`5.4 Metabolic and Nutritional
`There were no deaths due to Metabolic and Nutritional causes in the atomoxetine
`
`develcpment program. There was one SAE in the Metabolic and Nutritional body system
`in the ADHD database. Subject LYAR-081-5953, a 35-year-old male with a baseline
`glucose of 105mg/dL, was hospitalized for newly diagnosed diabetes mellitus (blood
`glucose 451mg/dL). In the historical database, the only Metabolic and Nutritional SAE
`was hypoglycemia.
`
`Weight loss led to the discontinuation of one pediatric subject from the ADHD studies.
`No adults discontinued from placebo controlled ADHD trials for a metabolic or nutritional
`AE. Two subjects from the historical database depression trials discontinued for weight
`loss, the only Metabolic and Nutritional AEs leading to discontinuation.
`
`Weight loss was the only metabolic and nutritional AE reported by more than 1% of
`atomoxetine subjects (26%, 51/1 ,933) in pediatric ADHD trials. ln the placebo controlled
`pediatric ADHD trials, weight loss was reported by 2.4% (8/340) of atomoxetine subjects
`compared to 0 placebo subjects. Similarly, 2.2% (6/269) of adult ADHD subjects had a
`weight loss AE compared to 0.8% (2/263) of placebo subjects. In study LYAC, a
`pediatric ADHD study where subjects were randomized to 3 fixed doses of atomoxetine
`or placebo, there were no weight loss AEs in the placebo (n=83) or 0.5mg/kg/day (n=44)
`groups. In the 1 .2mg/kg/day group 1.2% (1/84) reported a weight loss AE compared to
`2.4% (2/83) of th