A Double-Blind, Randomized, Placebo-Controlled Study of
`Quetiapine as Adj unctive Treatment for Adolescent Mania
`
`MELISSA P. DELBELLO. M.D.. MICHAEL L. SCHWIERS. M.S.. H. LEE ROSENBERG, 13.5..
`A.-‘JD STEPHEN M. STRAKOWSKI. M.D.
`
`ABSTRACT
`
`Objectives: This randomized. double-blind. placebo-controlled study examined the efficacy and lolerability ol quetiap-
`ine in combination with divalproex (DVP) for acute mania in adolescents with bipolar disorder. It was hypothesized that
`DVP in combination with quetiapine would be more ettective than DVP alone for treating rnania associated with adoles-
`cent bipolar disorder. Furthermore. it was hypothesized that quetiapine would be well tolerated. Method: Thirty manic or
`mixed bipolar I adolescents (12-18 years) received an initial DVP dose of 20 mgikg and were randomly assigned to 6 weeks
`of combination therapy with quetiapine. which was titrated to 450 mgiday (n = 15) or placebo (n = 15). Primary efficacy
`measures were change from baseline to endpoint in Young Mania Rating Scale (YMHS) score and ‘(MRS response rate.
`Salary and tolerability were assessed weekly. Results: The DVP + quetiapine group demonstrated a statistically signif-
`icantly greater reduction in ‘(MRS scores from baseline to endpoint than the DVP + placebo group (F127 = 5.04. p: .03).
`Moreover. YMRS response rate was significantly greater in the DVP + quetiapine group than in the DVP + placebo group
`(87% versus 53%: Fisher exact test, p = .05). No signilicant group differences from baseline to endpoint in safety mea-
`sures were noted. Sedation, rated as mild or moderate. was significantly more common in the DVP + quetiapine group
`than in the DVP + placebo group. conclusions: The findings of this study indicate that quetiapine in combination with
`DVP is more effective for the treatment of adolescent bipolar mania than DVP alone. In addition, the results suggest that
`quetiapine is well tolerated when used in combination with DVP iorthe treatment of mania. J. Am. Acad. Child Adoiasc.
`
`Psychiatry. 2002, 41 (10):1216—1223. Key Words: mania, bipolar disorder. quetiapine, adotescent.
`
`Although the onset of bipolar disorder typically occurs
`during adolescence (Lish et 31.. 1994). only one parallel-
`group. placebo-controlled study of adolescents or chil-
`dren with bipolar disorder has been published. Specifically,
`Geller and colleagues (1998) evaluated the efficacy of
`lithium in a 6-week, placebo-controlled study oF25 ado-
`
`Affffftd Ma)’ 10. 2017.}.
`From the Biprrlnr mm’ .”:yr'-iiniir Dirrtrrfrrr Rtfrfl-Pt’) Pmgmm. Dqbrrrrmrnt (if
`Pry-:i'sr'.-arty. Uni‘:-rrrity afCr'na'mmrr' C01/rgrafMra'irinr. Ct'nr':'i:nrttz'. Prrrrrm-rl
`in part at the 4:1: lnramatiaml Canfirmr: an Bipolar Dfrmzlcr. Pirrrbtcrgb. jun:
`.l'-‘tr. 200 I .
`This siiartji um rrippartrd by (2 grmrt from .-I:rmZrurrrz Pl:rrnrr.artn.tJ'm.!lr. Tb:
`itrm'mr.tgnrrgfir!i')- nrl-nnrtrirdgr the rmr':mru'r of:11:jiiilnrirfng r4's:n'n'.l'r rrafli /Ingeih
`i’-{ud':_.'.u'mI. Sirrtrw-m l»'t.7’i[i'Imr'r, Mirlitilr $a‘i'cn*. 1-‘ja_’yn Siicshndri. M. D.. and Krttbi
`Mnmrfiare. R. Pb. in rm’a':':i'r.-rr. they rtpprerirrtr the rripparr mm‘ rfliarrr of 11::
`Cirm‘m-mu‘ Ci:ii'a'reni Hrispirm'Mrr1'i'ml Crrmr Adnlzsrrnt ilrftdiral Pryrlziazrir
`U'm'l' «trending. xmrsing. rarinl want. and’ ariminirrmririe mrfl."
`Corn-rprmdrnrr to Dr. Drlflrfln. Drpiirrnrerrr of F'.r_yri';r'rrrijr. Um'urr.u'r_y nf
`Cinrivrrinri Colfegr affldrdirirtt. 2.31‘ Btlbrrdrr /litmus. .l'?O. Box 6.70559.
`Cinrinrmri. OH 45267-0559; e-mail: a’ti'br.inrp{@rinrrit'.mztdar.
`U890-8S67!02i'4l10—l2l6©2002 by the American Academy ufChild and
`Adolescent Psychiatry.
`DOI: 10.I097/0].CHl.0000024337.943lll.-‘il
`
`lescenrs with bipolar disorder and concurrent substance
`use disorders. They Found that lithium was more effec-
`tive than placebo for reducing global psychopathology
`scores. but. nonetheless, nearly half of the patients did
`not respond to lithium (Geller et al., 1998). This rate 0F
`lithium response is similar to that observed in adults
`(McElroy and Keck. 2000).
`In contrast to adults with bipolar disorder, children
`and adolescents with this illness are more likely to pre-
`sent with rapid cycling or in a mixed state (Geller et al.,
`2000). suggesting that anticonvulsants may be more eFFec—
`tive than lithium therapy (Swann er al., 1997). However.
`open-label treatment studies have Found that many chil-
`dren and adolescents with bipolar disorder do not respond
`to divalproex (DVP) (Kowatch er al., 2000; West er al.,
`1995). For example, Kowarch and colleagues (2000)
`assessed the comparative eFFectiveness of lithium, dival-
`proex sodium, and carbamazepine For the treatment oi:
`mania and hypomania in children and adolescents with
`bipolar disorder, types I and II. In this 6-week, open-
`
`1216
`
`_l. AM. a-'\C.-’\D. CHILD ADULESC. PSYCHMTRY. 4|-.lfl. OCTOBER 2on2
`
`1of8
`
`Alkermes, Ex .1019
`
`1 of 8
`
`Alkermes, Ex. 1019
`
`

`
`label. randomized study, they Found that although DVT’
`demonstrated the largest response rate oF the three treat-
`ments. 47% oF the patients Failed to respond to this ther-
`apy (Kowatch er al.. 2000).
`Together, these data suggest that alternative pharmaco-
`logical options For the treatment of pediatric mania are
`needed. Controlled investigations oiatypical antipsychotics
`suggest that they are eflicacious for the treatment of mania
`in adults (Segal et al.. 1998; Tohen et al.. 1999, 2000), and
`several case series suggest that these agents are also ePFec—
`tive for the treatment ofmania in children and adolescents
`
`(Chang and Ketter. 2000; Frazier et al.. 1999; Soutullo
`et al.. 1999). Thus the addition of an atypical antipsychotic
`to a mood stabilizer may decrease manic symptoms and
`improve response rates. indeed. Tohen and colleagues
`(2002) recently compared the efiicacy oicombined ther-
`apy with olanzapine and either DVP or lithium to DVP
`or lithium monotherapy For the treatment oiiacute mania
`in adults and Found that the response rate was significantly
`higher in the combination group (68 versus 45%).
`.Quetiapine Fumarate is an-atypical antipsychotic agent
`with a unique receptor binding profile. Quetiapine has a
`high affinity For histaminergic H1 and Oh-adrenergic neu-
`roreceptors. In addition. quetiapine exhibits aliinity For
`brain serotonin 5-HT; and 5-i-iT.,1 and dopamine D. and
`D3 receptors and has higher selectivity For 5-HT; relative
`to D3 receptors (Dev and Raniwalla. 2000: Jones et al..
`2001). Several case reports suggest that quetiapine is efi'ec-
`tive and well tolerated For the treatment of mania in adults
`
`(Dunayevich and Stml-<0WSl<i. 2000; Ghaemi and Katzow,
`1999: Zarate et al.. 2000), affective psychosis in adolescents
`(McConville et al.. 2000; Padla, 2001). and refractory bipo-
`lar disorder in children (Catapano-Friedman. 2001: Schaller
`and Behar. 1999). Furthermore. studies of patients with
`schizophrenia indicate that quetiapine does not diH:er From
`placebo in rates of extrapyramidal symptoms (EPS) or pro-
`lactin elevation (Kasper and Muller-Spahn. 2000).
`With these considerations in mind, the aim of this
`
`do_uble—blind. placebo-controlled augmentation study was
`to investigate the etiicacy and tolerability of quetiapine as
`an adjunct to DVP For the treatment of acute mania in
`hospitalized bipolar adolescents. To our knowledge, this is
`the first parallel-group, placebo—controlled study to com-
`pare mood stabilizer monotherapy with the combination
`of mood stabilizer plus an antipsychotic in adolescents with
`acute mania. Furthermore, this is the first controlled inves-
`
`tigation oF an atypical antipsychotic For the treatment of
`pediatric bipolar disorder and the first controlled study of
`
`QUETIAPINE l.\l ADOLESCENT MANIA
`
`quetiapine for the treatment of bipolar disorder. We hypoth-
`esized that the combination ofquetiapine and DVP would
`be more eilicacious For the treatment ol: adolescent mania
`
`than DVT’ alone. and that quetiapine would be well tol-
`erated as an adjunctive agent in this population.
`
`METHOD
`
`Bipolar adolescents who were hospitalized For a manic or mixed
`episode were recruited from consecutive inpatient admissions to the
`Adolescent Pqrchiatric Unit at Cincinnati Children's Hospital Medical
`Center from May 2000 through May 2001. Patients were included in
`the study if they were 12-18 years old. met DSM-Wcriteria For bipo-
`larl disorder currently mixed or manic. and had :1 Young Mania Rating
`Scale (YMR5) (Fristad er al.. I992: Young et al.. 1978) score of 220.
`Patients were excluded if (1) they were pregnant; (2) their manic symp-
`toms were secondary to substance intoxication or withdrawal; (3) they
`had :1 substance use disorder within the prior 3 months; (4) they had
`a diagnosis of mental retardation (IQ < 70); (5) they had an unstable
`medical or neurological disorder. cataracts. or dinieally significant base-
`line laboratory abnormalities: or (6) they had a history of hypersensi-
`tivity. intolerance. or nonresponse to quetiapine or valproate. Nonresponse
`to valproate was defined as a 1-week trial with at least one therapeu-
`tic blood level of .280 mg1'L during the index mood episode without
`improvement in manic symptoms as determined by the subjects' and
`primary caregivers reports. Patients were also excluded if they had
`been treated with :1 depot neutoleptic within 3 months. an antide-
`pressant or antipsychotic within a week (Fluoxetine within :1 month).
`or a lmenzodiazepine or psychostirnulant within 72 hours. Patients pre-
`viously treated with lithium, valproate. or carbamazepine were required
`to have serum concentrations of <03 rnEq/l... 30 mg."L. and 3 mg/L.
`respectively. before receiving queriapine or valproate in this trial. to
`ensure that these medications were adequately "washed out." Patients
`were also excluded if they had been treated with other antiepileptic
`agents within 72 hours. Fifty potential study candidates were initially
`identified. However. 20 patients did not meet study inclusion and
`exclusion criteria because they had either congenital cataracts (11 = 3).
`a history ofintolerance or poor response to DVT’ (rt = 2). a substance
`use disorder (:1 = 3). or a primary psychiatric diagnosis other than bipo-
`lar disorder (Jr = 12). Therefore. 30 bipolar patients were randomized
`into this study (Fig. 1).
`This study was approved by the University of Cincinnati and the
`Children's Hospital Medical Center institutional review boards.
`Adolescent subjects provided written assent and their parents or legal
`guardians provided written informed consent For study participation
`and publication after study procedures were fully explained.
`Diagnostic interviews were performed with the Washington University
`in St. Louis Kiddie Schedule for Alifective Disorders and Schizophrenia
`(WASH-U-KSAD5) (Geller er al.. 200!) by trained raters (M.I‘.D..
`H.L.R.) with established diagnostic reliability (K = 0.94) (DelBello er al..
`2001). Adolescent subjects and their primary caregivers were interviewed
`separately. Primary caregiver and child responses were combined to ascer-
`tain diagnoses. Teachers and another primary caregiver were interviewed
`ifthere was :1 discrepancy between the primary caregivers and the ado-
`lescentis responses. All diagnoses were reviewed in a conference attended
`by the WASH-U-KSAD5 interviewer and at least one child and ado-
`lescent psychiatrist From which a consensus diagnosis was made.
`Demographic in Formation was obtained by interviewing the ado-
`lescenr and his or her primary caregivers. The Self-Rated Tanner Scale
`was used to assess the stage of adolescent sexual development (Morris
`and Udry. I980).
`
`1. AM. ACAD. CHILD ADOLESC. PSYCHIATRY. -nun. OCTOBER znnz
`
`I217
`
`2of8
`
`Alkermes, Ex. 1019
`
`2 of 8
`
`Alkermes, Ex. 1019
`
`

`
`DELBELLO ET AL.
`
`Eligible Patients
`
`(N=5tl)
`
`
`
`Not Randomized l'N=2D)
`
`
`
`Primary dlagnosls not BF! lN=12). cataracts (N=3}.
`substance use disorders iM=3l. Intolerance or poor
`resoonse to DVP (N=‘2)
`
`
`
`Randomization
`(N=3|:I)
`
`
`
`
`Divalproex
`+
`
`Quetiafilne
`[hl=i5)
`
`
` Wtlhdnlwn (N‘-*1]
`
`Weakly folItrw—uo
`
`(N=15)
`
`Weekly follow-up
`(N=15i
`
`
`
`
`Withdrawn (N1)
`Lad: oloifioacy [N=1)
`Sack of efficacy (N=1)
`
`
`
`Poor parental oowpllmoe (N-2)
`Poor suhjecl compliance {N=l3
`
`Refusal ofblaod dram tN=1}
`Major depressive episode tN=1}
`Transfer In tialant mtitity [N=1)
`
`Comp leted Trlal
`N=14
`
`
`
`Fig.1 Diagram oF subject How by treatment group. BP I = bipolar I disorder: DVP = djvalprocx,
`
`Efiicacy and Safety Measures
`The primary cFficacy measure was the YM RS (Fristad et al.. 1992:
`Young et al.. 1978). Secondary efficacy measures included the Positive
`and Negative Syndrome Scaic-Positive Sl.llJS(:JiC (PANSS-I‘) and the
`Children's Depression Rating Scale (CD115) to assess the severity of
`psychotic (Kay ct al.. i989) and depressive symptoms (Poznanski et al..
`"1979. 1983). respectively. Overall level of Functioning was assessed at
`baseline and endpoint with Childrcns Global Assessment Scale (CGAS)
`scores (Shaffer et al.. 1933). A child and adolescent psychiatrist with
`previously established reliability for each rating scale (M.P.D.) com-
`pleted all ratings by interviewing the subject and his or her primary
`caregiver (intraclass correlation cot:Fiicient 2 0.9).
`EPS were assessed with the Simpson—Ang1.rs (Simpson and Angus.
`1970}, Barnes Altathisia (Barnes. 1989). and Abnormal Involuntary
`Movement Scales (Guy. 1976]. Laboratory tests obtained included a
`complete blood cell count (CBC]with differential and prolactin.
`thyroid-stimulating hormone (TSH). and valproic acid levels. In addi-
`tion. liver Function tests (LFTS), including alanine aminottansfetase,
`aspartate antinotransferase. and total bilitubin. were obtained. Viral
`signs obtained included weight and orthostatic blood pressure and pulse.
`Eiectrocardiograms (ECGS) were monitored throughout the study. in
`addition. physical and slit-lamp ocular examinations were performed
`on_cach subject at baseline and endpoint. Adverse events were assessed
`when ratings were obtained by asking the adolescents and their primary
`caregivers open-ended questions about potential side eiiccts.
`
`Study Protocol
`
`This study was a 6-week. randomized. parallel-group. cloubic-blind.
`placebo-controlled investigation of DVT’ monotherapy versus the
`combination of DVP plus quetiapine. After meeting all inclusion and
`exclusion criteria. subjects were randomly assigned to receive either
`placebo or adjunctive quetiapinc. Randomiution. which was assigned
`by invcstigational pharmacists, was Stratified by sex and the presence
`of psychosis using a random number generator. All inpatient and
`research staff were blind to subject treatment group.
`All subjects received an initial DVT’ dose of 20 mglltg per day on
`day 0. which was adjusted to achieve .1 therapeutic serum level of 80-130
`mg/dL. On day 0. subjects wcre also randomly assigned to receive
`placebo or an initial quctiapine close of 25 mg b.i.d.. which was titrated
`to a maximum of ISG mg t.i.d. by day 7. A maximum of 2 mg of
`lorazepam per day was permitted during the First 14 days of the study.
`Compliance was measured by pill count at each visit and by assess-
`ing valptoic acid serum levels. which were collected 10 to I4 hours after
`the last DVT’ dose on days 3, 7. lli. 21. and 42 (or termination from
`the study). In addition. each subject was asked to keep a medication
`log to encourage compliance and identify missed doses. Subjects were
`discontinued from the study if they missed more than 2 consecutive
`days oF study medication or more than six closes during any 7-day period.
`Efficacy and safety ratings were performed at baseline. clays 3 and
`7. and then weekly until day 42 or termination from the study. Vital
`signs were monitored at each visit. Scrum prolactin levels. LFTs.TSH.
`
`1218
`
`}. AM. AC.-H’). CHILD ADOLESC. PSYCHIATRY. 41:|0. OCTOBER znnz
`
`3of8
`
`Alkermes, Ex. 1019
`
`3 of 8
`
`Alkermes, Ex. 1019
`
`

`
`and CBC were assessed at baseline and day 42 or termination. ln addi-
`tion. LFTs and CBC were also assessed at days 7 and 21. ECGs were
`performed at baseline and days 7. 2]. and 42 or termination.
`Inpatient attending physicians (not associated with the study) dis-
`charged study participants from the inpatient psychiatry unit when
`they determined that the subjects were clinically stable. All subsequent
`visits were performed in an outpatient setting. The majority of patients
`were discharged 7 to 14 days after admission (93%). There was no sta-
`tistically significant group dihference in length of hospitalization.
`
`Statistical Analysis
`
`Prior to study initiation, sample size estimates were calculated by
`assuming a directional hypothesis (i.e.. that the combination therapy
`would he better than monotherapy) and a medium to large effect size.
`with 80% power and D! = .05 (Stevens. 1990).
`Statistical analyses were performed with the Statistical Analysis System
`for the PC (SAS Institute. Cary. NC. 1999). Clinical and demographic
`variables were identified as potential covariatcs using 1 tests or Fisher
`exact tests and a liberal p value of .2 for differences between groups.
`With the data from the intent-to-treat samples (H = lfilgroup), t
`tests were used to calculate differences from baseline to endpoint for
`each eflicacy measure within each treatment group. Primary efficacy
`measures were change from baseline to endpoint in YMRS and YMRS
`response. Response was defined as a '-250% reduction in ‘(MRS score
`from baseline to endpoint. Analysis of covariance (ANCOVA) was
`used to compare group differences in endpoint ‘(MRS score after con-
`trolling for baseline values. The effect size for each treatment group
`was calcuiated by using the mean change and standard deviation from
`baseline to endpoint in YMRS scores (Cohen. 1988). Group differ-
`ences in YMRS response rates were compared by using :1 one-tailed
`Fisher exact test. Secondary efficacy measures were change from base-
`line to endpoint in CGAS. CDRS, and PANSS-P scores. ANCOVAS
`were used to compare group differences in endpoint CGAS, CDRS.
`and PANSS-P scores after controlling for baseline values.
`In addition. likelihood-based mixed-model repeated-measures
`AN CO'\/As (pror niibccd) were conducted to evaluate group-by-day dif-
`ferences in YMRS. CDRS. and PANS5-P scores. with control for base-
`line scores. This analysis uses all available data and was selected to avoid
`biases that might be introduced with last observation carried forward
`or complerer analyses. As a follow-up analysis. least-squares means were
`calculated at each time point for each rating instrument to determine
`on which days statistically significant group differences occurred.
`Group differences in rates of side effects were assessed with two-
`tailed Fisher exact tests. ANCOVAs were used to compare endpoint
`laboratory measures benveen groups after controlling for baseline val-
`ues. Other analyses were performed as necessary.
`
`RESULTS
`
`Baseline Comparisons of Patient Characteristics
`
`Twenty-two (73%) of the 30 randomized subjects com-
`pleted the 6-week protocol. One patient in each group
`discontinued prematurely (at day 14in both cases) because
`of lack of efficacy for acute mania symptoms. The six
`remaining noncompleters were all in the DVT’ + queri-
`apine group. The reasons for these patients’ premature
`termination included refusal to participate in blood draws
`(ti = 1. day 7). parental treatment noncompliance (:2 = 2.
`
`QUETIAPINE IN :\DOLESCENT MANIA
`
`days 28 and 35). adolescent treatment noncompliance
`(n = 1. day 28). transfer to a distant residential treatment
`facility (11 = 1, day 28). and developing a major depres-
`sive episode after mania resolution (P2 = 1. day 21). No
`subjects in either group discontinued from the study
`because of medication side effects (Fig. I).
`There were no significant group differences in age, sex.
`race. socioeconomic status. Tanner stage. baseline CGAS,
`YMRS. CDRS. or PANSS-P scores or rates of mixed
`
`episodes. psychosis. and attention-deficit/hyperactivity
`disorder (Table 1). Age at onset of bipolar disorder was
`defined as the age at which a DSM-fVmood episode ini-
`tially occurred and was determined with the WASH-U-
`KSADS. Subjects in the DVP + quetiapine group had a
`younger age at onset of bipolar disorder compared with
`those in the DVP + placebo group (Table 1;}: = .01).
`Mean valproic acid level was 102 mgi'dl.. in the DVP +
`placebo group and 104 mg/dL in the DVP + quetiapine
`group. By day 3, 97% (29/30) of the subjects reached a
`therapeutic valproic acid level (mean i SD = 113 1 20
`mgi’dL) and by day 7. 100% had reached a therapeutic
`valproic acid level (114 t 26 mg/dL). Mean dosage of
`quetiapine was 432 mgfday in the DVP + quetiapine
`group. One subject in the DVP + quetiapine group was
`not titrated to the maximum dose of 450 mg/day because
`of excessive sedation and was treated with 250 mglday.
`
`Primary Efficacy Measures
`
`Analyses within each treatment group revealed a sta-
`tistically significant reduction from baseline to endpoint
`
`TABLE 1
`
`Demographic and Clinical Characteristics of Bipolar
`Adolescents by Treatment Group
`
`Variable
`
`Sex. 1! (Wu). female
`Age. mean (SD). yr
`Race. n (‘'/u). Caucasian
`Tanner stage. mean (SD)
`SE3. mean (SD)"
`Age onset bipolar disorder.
`mean (SD). yr”
`Mixed episode, )2 (%)
`Psychosis. rt (%)
`ADHD. It (We)
`
`DVI’ 4- Placebo DVI’ 4- Quetiapine
`(r1 = 15)
`(71 = 15)
`
`(47)
`7
`(2)
`I45
`(37)
`13
`3.9 (1.3)
`3.6 (1.9)
`
`11
`13
`7
`8
`
`(3)
`(37)
`(47)
`(53)
`
`(47)
`7
`(2)
`l4.l
`(80)
`I2
`3-3 (1.1)
`3.0 (1.5)
`
`3
`10
`7
`i0
`
`l3)
`(67)
`(47)
`(67)
`
`Nate: DVI’ = divalproex: SE5 = socioeconomic status; ADHD =
`attcn tion-deficit/hyperactivitydisorder.
`“ Range = 1-7. rating of 3 = parental yearly income ofS2fl.O00—
`535.000.
`
`"' Significant difference between groups: in = 2.75, p = .0].
`
`J. AM. ACAD. CHILI) ADOLESC. l"5\'CHlATR‘t'. 4i;l0. ocrosea 2on2
`
`I219
`
`4of8
`
`Alkermes, Ex. 1019
`
`4 of 8
`
`Alkermes, Ex. 1019
`
`

`
`I lliueluu E] F1-Jr---inl
`
`kl
`
`DI-‘.i BELLO ET AL.
`
`no
`‘:5
`30
`' 25
`20
`15
`10
`
`Score
`YMR5
`
`50
`
`Di-'I‘+[’i.:cebo
`
`LJ\‘P+Quoriapmc
`
`Fig. 2 Manic adolescents in the divalproex (DVP) + quetiapine group (ti =
`I5) had a greater reduction in Young Mania Rating Scale (YM RS) scores from
`baseline to endpoint compared with those in the DVT’ + placebo group in =
`I5): analysis of covariance: F”: = 5.04.}: = .05. ‘p = .003: "p < .0001.
`
`in YMRS score (Fig. 2). However, the DVP + quetiap-
`ine group demonstrated a significantly greater reduction
`in YMRS score from baseline to endpoint than the DVP +
`placebo group (Fm; = 5.04, p =: .03) (Fig. 2).
`The YMRS response rate was significantly greater in
`the DVP + quetiapine group than in the DVP + placebo
`group (87% versus 53%; Fisher exact test,p = .05). YMRS
`responders did not differ from nonresponders in length
`of time in the study (mean length of time in the study
`was 5.3 and 5.1 weeks, respectively, p = .7).
`
`Secondary Efficacy Measures
`
`‘Within each treatment group, CDRS (DVP + placebo,
`r: 4.7,}: = .0004 and DV1’ + quetiapine. r: 3.0,}? = .01).
`PANSS-P (DVP + placebo. t: 3.9,}: = .002 and DVP +
`quetiapine. r= 3.1,}? = .009), and CGAS (DVP + placebo,
`r = 8.6, p < .0001 and DVP + quetiapine, r= 11.0,}: <
`.0001) scores were significantly reduced from baseline to
`endpoint. However. there were no significant differences
`between groups in change from baseline to endpoint in
`CDRS (Fm-; = 0.0.p = 1.0), PANSS-P(F1'27 = 0.1, p =
`.8), and CGAS (F137 = 1.5,}: = .2) scores.
`
`Response Over Time
`
`Subjects in the DVI’ + quetiapine group demonstrated
`an overall greater reduction over time in YMRS scores than
`did subjects in the DVP + placebo group (Fig? = 3.5, p <
`.01) (Fig. 3). Specifically, statistically significant group dif-
`ferences were found on days 14. 21. and 42 (p = .009,p =
`.005, p = .01 , respectively). No statistically significant group
`differences were found for change in CDRS (Fm = 0.1,
`p = .7) or PAN5S—P (Fm; = 0.5.19 = .4) scores over time.
`
`Lorazepam Use
`
`Three subjects in the DVP + placebo group and two
`subjects in the DVP + quetiapine group required lorazepam
`
`Fig. 3 Manic adolescents in the clivalproex 4 quetiapitic group (squares; rt
`= 15) had a statistically significantly greater I.'L'(lllCrlDl1 in Young Mania Rating
`Scale (YM RS) scores over time than those in the DVI’ + placebo group (dia-
`monds: n = 15): 2|'|2ll)'5l.\‘ nfI:m'.1riancc:F,,;- = 8.3.}: < .01. ‘p < .0].
`
`during the first 14 days of the study. Four of the subjects
`required only one dose of lorazepam (0.5—l mg) and one
`subject required three doses (total dose = 1.5 mg). There
`was no significant group difference in amount oflorazepam
`used (p = .6).
`
`Tolerability and Side Effects
`
`There were no significant group differences in change
`from baseline to endpoint in QTC interval. TSH, white
`blood cell count. hematocrit, platelet count, prolactin
`level, weight. EPS ratings, or LFTS (Table 2). In addition,
`there were no subjects who had an abnormally elevated
`prolactin level at endpoint. No subjects had orthostatic
`hypotension during this study. No subjects developed
`cataracts or a serious adverse event during this study.
`The most common side effects in both treatment groups
`were sedation. nausea, headache, and gastrointestinal irri-
`tation (Table 3). Sedation was significantly more common
`in the DVP + quetiapine group than in the DVP + placebo
`group (Fisher exact test. p = .03). However, within the
`DVP + quetiapine group, there was no significant differ-
`ence in rate ofsedation between responders and nonre-
`sponders (Fisher exact test, p = .4). All side effects were rated
`as mild to moderate by the subjects and their caregivers.
`
`DISCUSSION
`
`The results of this study indicate that quetiapine in
`combination with DVP is more effective at reducing manic
`symptoms associated with bipolar disorder than DVP
`monotherapy. Furthermore. the results suggest that que-
`tiapine is well tolerated when used in combination with
`
`1220
`
`J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY. shun. OCTOBER 2002
`
`5of8
`
`Alkermes, Ex. 1019
`
` D
`
`7‘
`
`14
`
`21
`Day
`
`28
`
`35
`
`-12
`
`5 of 8
`
`Alkermes, Ex. 1019
`
`

`
`TABLE 2
`Safety and Laboratory Measures by Treatment Group
`DVT’ + Placebo
`DVP + Quetiapine
`
`Variable
`(n = 15)
`(n = 15)
`
`QUETIAPINE IN ADOLESCENT MANIA
`
`Change in QTc. mean (SD). rnsec
`Change in TSH. mean (SD). mlUJ'L
`Change in \Vl3C. mean (SD). 10‘i’t1L
`Change in HCT. mean (SD). %
`Change in platelets. mean (SD). 10"/|.lL
`Change in prolactin. mean (SD), ng/ml.
`Prolactin elevation. in (Win)
`Change in weight. mean (SD). kg
`Change in EPS ratings. mean (SD)
`(0)
`0
`(0)
`0
`AIMS
`-0.] (0.3)
`(0.3)
`0.1
`Barnes Akathisia Scale
`
`
`
`
`-0.1 (1.1) 0Simpson—Angus Scale (0.8)
`
`(17.9)
`1
`(L7)
`0.3
`(0.6)
`-0.6
`(0.4)
`—1.l
`-58.4 (27.2)
`-5.7
`(3.7)
`9
`(60)
`2.5
`(2.1)
`
`(2l.6)
`7
`0.6 (1.8)
`-2.)
`(0.7)
`-2.2 (0.3)
`-63.4 (4.2)
`-1.6 (6.9)
`7
`(47)
`4.2 (3.2)
`
`Nate: DVT’ = divalproeit: TSH = thyroid-stimulating hormone: WBC = white blood cell count: HCT = hernatocrit: EPS =
`estrapyramidal symptom: AIMS = Abnormal lnvoluntary Movement Scale.
`
`DVT3 Despite differences in study populations and method-
`ologies. the response rate with DVP monotherapy was
`consistent with that reported in bipolar children and ado-
`lescents by Kowatch and colleagues (53%) (Kowatch er al..
`2000). as well as in manic adults in other studies
`
`(Bowden et al.. 1994). These Findings suggest that the
`increased response obtained by adding quetiapine is not
`simply due to an arypically poor response to DVI’ alone.
`Seventy-th ree percent of patients completed this study.
`which is consistent with completion rates from other
`studies of patients with acute mania (Kowatch et ai..
`2000: Tohen er al., 2002). The percentage of patients
`completing this study was greater in the DVP + placebo
`group (93%) than in the DVT’ + quetiapine group (53%).
`However. most of the reasons for dropout were unrelated
`to treatment and were due to psychosocial factors. The
`rate of dropout due to lack of efficacy was the same in
`both groups (7%). The analyses performed were based
`
`TABLE 3
`
`Adverse Events by Treatment Group
`
`DVP + Quetiapine
`DVP + Placebo
`('7 = 15)
`('1 = 15)
`"‘d"'='5~‘= Eve"?
`I2 (30)
`5 (33)
`Sedation“
`4 (27)
`6 (40)
`Nausea/vomiting
`5 (33)
`3 (20)
`D5“‘i“E5-‘
`7 (47)
`7 (47)
`Headache
`7 (47)
`5 (33)
`Gastrointestinal irritation
`Joint pain
`2 (13)
`2 (13)
`
`DP)’ 1'“°ufl'I
`2 (13)
`S (33)
`,V,,,,_. V_.|uE5 repmen. ,, (%)_
`" Fishcrcxact test./1 = .03.
`
`solely on the data that were observed and despite the
`reduced sample size, significant differences were still Found.
`Dropout limits the power to detect significant difierences.
`Nonetheless. the observed data do indicate a significant
`treatment effect.
`
`Clinical Impiications
`
`Although there are several case series suggesting that
`the atypical antipsychotics clozapine. olanzapine. and
`risperidone are effective as adju nctive treatments to mood
`stabilizers for the treatment of pediatric mania, the side
`effects of these agents limit their utility in children and
`adolescents. Specifically. clozapine is associated with agran—
`ulocytosis and seizures, olanzapinc is associated with
`increased appetite and weight gain. and risperidone is
`associated with prolonged weight gain. prolactin eleva-
`tion. and EPS (Chang and Ketter. 2000; Frazier et al.,
`1999. 2001; Kearns et al., 2000; Kowatch et al.. 1995:
`Selva and Scott, 2001; Soutullo er al.. 1999: Zuddas et al..
`
`2000). In contrast to these other agents. the results of
`this study suggest that quetiapine is well tolerated as
`adjunctivc treatment for mania in bipolar adolescents.
`Specifically, no subjects developed laboratory. ECG, or
`viral sign abnormalities. Moreover, in this study queri-
`apine was not associated with prolactin elevation or BPS.
`Although sedation was the most common side effect in
`h DVP
`.
`.
`.
`d I
`t e
`+ quetiapine group. manic a o escents com-
`monly present with poor sleep. so short-term somnolence
`is often therapeutic. Further analysis revealed that within
`each treatment group. antimanic response was not asso-
`ciated with sedation. suggesting that the improvement
`
`J. AM. scan. CHILD ADOLESC. PSYCHIATRY. 41:10. OCTOBER 2002
`
`1221
`
`6of8
`
`Alkermes, Ex. 1019
`
`6 of 8
`
`Alkermes, Ex. 1019
`
`

`
`DI-‘.l1§El..l_O ET AL.
`
`in our sample was due to improvement in mood and
`behavior and not excessive sedation.
`
`Weight gain is one of the major side effects of most anti-
`manic agents. In this study. all subjects were treated with
`DVP. which is associated with weight gain (Demir and
`Aysun. 2000). Therefore. most adolescents gained weight
`during the course oithis study. Although the combination
`of queriapine and DVP resulted in a larger weight gain
`than DVT’ alone. the addition of quetiapine did not sig-
`nificantly increase the amount of weight gained. Nonethelas.
`investigations examining weight changes associated with
`quetiapine monotherapy in children and adolescents are
`necessary. Minimizing the weight gain associated with most
`of the pharmacological treatments for bipolar disorder
`might increase medication compliance in manic adoles-
`cents and decrease the long-term risks associated with sig-
`nificant weight gain, such as hyperglycemia and type II
`diabetes mellitus (McIntyre et al., 2001).
`
`Limitations
`
`Several limitations of this study need to be considered
`when interpreting the results. First. the small sample size
`might limit the power of this study to detect group dif-
`ferences in some of the secondary efficacy, safety mea-
`sures. and demographic and clinical variables. Second,
`although this 6-week study demonstrated that quetiap-
`ine is effective and well tolerated as adjunctive treatment
`to DVP for (rear: mania, studies of longer duration are
`necessary to assess whether long-term treatment with this
`combination is safe. as well as effective. for preventing
`recurrent affective episodes. Third. the dosing regimen
`implemented in this study was selected primarily on the
`basis of clinical experience of several child and adoles-
`cent psychiatrists and previous studies of quetiapine in
`patients with schizophrenia. Although no patients dis-
`continued treatment because of the adverse effects of
`
`medication. future studies are needed to investigate other
`closing strategies, which might maximize efficacy and tol-
`erability. Finally. the patients in this study were all hos-
`pitalized and were predominantly Caucasian; therefore,
`the results may not generalize to other sociodemographic
`or clinical groups. In addition. nonresponders to val-
`proate were excluded from this study. which further lim-
`its our ability to generalize the findings. However, to our
`lcnowledge. quetiapine is the first treatment for bipolar
`disorder to have its efficacy initially demonstrated in a
`pediatric. rather than adult, population. Despite these
`limitations, the results of this study indicate that queri-
`
`apine is well tolerated as acljunctive treatment to UV?
`for adolescent mania. In addition. this study suggests that
`the combination of DVT’ and quetiapine is more effec-
`tive th