Quetiapine with lithium or divalproex for the
treatment of bipolar mania: a randomized,
double-blind, placebo-controlled study
Sachs G, Chengappa KNR, Suppes T, Mullen JA, Brecher M,
Devine NA, Sweitzer DE. Quetiapine with lithium or divalproex for
the treatment of bipolar mania: a randomized, double-blind,
Bipolar Disord 2004: 6: 213–223. ª Blackwell Munksgaard, 2004
Objective: Evaluate the efficacy and tolerability of quetiapine (QTP)
combined with lithium (Li) or divalproex (DVP) in the treatment of acute
Methods: Patients were randomized to 21 days of double-blind
treatment with QTP plus Li/DVP, or placebo (PBO) plus Li/DVP.
QTP was rapidly dosed up to a maximum of 800 mg/day; Li was dosed
to 0.7–1.0 mEq/L; or DVP to 50–100 lg/mL.
Results: Fifty-six of 91 (61.5%) individuals in the QTP + Li/DVP
group compared with 49 of 100 (49%) taking PBO + Li/DVP completed
the study. A significantly greater mean reduction in total Young Mania
Rating Scale (YMRS) score was observed at end-point in patients
receiving QTP + Li/DVP compared with those in the PBO + Li/DVP
group ()13.76 versus )9.93; p ¼ 0.021). The response rate (‡50% YMRS
improvement) was significantly higher in the QTP + Li/DVP group
than in PBO + Li/DVP-treated patients (54.3% versus 32.6%;
p ¼ 0.005), as was the proportion of patients achieving clinical remission
(YMRS < 12) (45.7% versus 25.8%; p ¼ 0.007). Patients receiving
QTP + Li/DVP also had a significantly greater improvement in Clinical
Global Impressions-Bipolar (CGI-BP) Severity of Illness scores ()1.38
versus )0.78; p ¼ 0.001). The mean last-week dose of QTP was 584 mg/
day in patients meeting response criteria. Common adverse events (at
least 10% and twice the rate of Li/DVP) in the QTP + Li/DVP group
included somnolence, dry mouth, asthenia, and postural hypotension.
Conclusions: Quetiapine combined with either Li or DVP has
superior efficacy compared with Li or DVP monotherapy for treating
patients with bipolar mania. Combination therapy was well-tolerated
and most adverse events were mild, withdrawal because of adverse
events being only 5% compared with 6% on Li or DVP
G Sachsa, KNR Chengappab,
T Suppesc, JA Mullend, M Brecherd,
NA Devinedand DE Sweitzerd
aHarvard Bipolar Research Program, Boston, MA,
bWestern Psychiatric Institute and Clinic, University
of Pittsburgh Medical Center, Pittsburgh, PA,
cUniversity of Texas Southwestern Medical Center,
Dallas, TX,dAstraZeneca, Wilmington, DE, USA
Key words: bipolar disorder – combination
therapy – divalproex – lithium – mania – quetiapine
Received 4 September 2003, revised and
accepted for publication 19 February 2004
Corresponding author: Gary Sachs, MD, Harvard
Bipolar Research Program, Massachusetts General
Hospital, 50 Staniford St, 5th Floor, Boston,
MA 02114, USA. Fax: 617-726-6768;
G Sachs: Consultant for Abbott Laboratories, Janssen Pharmaceutica, Novartis, Eli Lilly and Company, Bristol-Myers Squibb, Glaxo SmithKline
Pharmaceuticals, Elan, Sanofi, Sigma-Tau, AstraZeneca Pharmaceuticals; Grant Support from Abbott Laboratories, Janssen Pharmaceutica;
Honoraria from Abbott Laboratories, Glaxo SmithKline Pharmaceuticals, Janssen Pharmaceutica, Eli Lilly and Company, Bristol-Myers Squibb,
Solvay, Novartis, Sanofi, AstraZeneca Pharmaceuticals, Pfizer Inc; Stockholder – none. KNR Chengappa: has received grant/research support from
Ortho McNeil and Janssen Pharmaceutica, and serves on the speakers bureau for AstraZeneca, Eli Lilly, Janssen, and Ortho McNeil. T Suppes:
Sources of funding for clinical grants – Abbott Laboratories, AstraZeneca, Bristol-Myers Squibb, Glaxo SmithKline Pharmaceuticals, Janssen
Pharmaceutica, National Institute of Mental Health, Novartis, Robert Wood Johnson Pharmaceutical Research Institute, The Stanley Medical
Bipolar Disorders 2004: 6: 213–223
Copyright ª Blackwell Munksgaard 2004
Bipolar disorder is a prevalent and severe mental
illness characterized by debilitating mood swings
from intense euphoria to depression. The symp-
toms of mania associated with bipolar I disorder
are severe, interfere with the capacity to function
effectively in work or social settings, and often
require hospitalization (1).
While there is currently no cure for bipolar
disorder a number of agents are now available that
substantially decrease the morbidity of manic
episodes. Unfortunately, many patients do not
fully respond to these agents or are unable to
tolerate their side effects, contributing to poor
levels of adherence to medication (2). Effective
agents with improved tolerability that target a
broad range of symptoms associated with mania,
with substantial clinical efficacy, are needed.
Therefore, the ongoing development of new treat-
ment options to address the needs of patients and
physicians remains important.
The primary aim in the treatment of mania is
to rapidly control symptoms while minimizing
unwanted side effects. Currently, the first-line
approach to treatment of severe manic or mixed
episodes is the initiation of lithium (Li) or dival-
proex (DVP) in combination with an antipsychotic
(3, 4). For less ill patients, monotherapy with Li,
DVP, or an antipsychotic is recommended (3). For
those who do not fully respond to monotherapy,
the addition of an antipsychotic is the next-line
recommendation. In clinical practice this means
that combination or adjunct therapy is the norm,
not the exception, especially in the context of
short-duration hospital stays.
In most cases, atypical antipsychotics as a class
are preferred over typical antipsychotics because
they offer equivalent efficacy with a more benign
side effect profile than typical agents (3). However,
it cannot be assumed that newer atypical anti-
psychotic agents are effective either as mono-
therapy or in combination with other agents.
Tolerability is of particular concern when combi-
nation therapy is used, as an increased risk from
cumulative adverse effects, although not always
observed, might be expected (5, 6).
A growing body of literature and clinical experi-
ence suggests that the atypical antipsychotic que-
tiapine (QTP) is efficacious and well-tolerated for
the treatment of mania. Several open-label and
retrospective studies indicate that QTP is effective
and well-tolerated in the treatment of bipolar
mania in adults both as monotherapy and in
combination with standard agents (7–11). One
small, double-blind, randomized, controlled study
has provided preliminary evidence for the efficacy
of QTP, when added to ongoing DVP treatment, in
significantly improving manic, depressive, and
psychotic symptoms in adolescent patients with
bipolar disorder (12).
Despite being a recommended approach (3, 4),
few large, randomized, and adequately controlled
studies have examined the safety and efficacy of
atypical antipsychotics in combination with stan-
dard agents for the treatment of acute mania (13,
14). These studies support the hypothesis that an
atypical antipsychotic in combination with Li or
DVP is more effective than Li or DVP monothera-
py in the treatment of acute mania. The trial
presented here evaluates the efficacy of QTP as
combination therapy with Li or DVP in bipolar
This 3-week, double-blind, randomized, parallel-
group, placebo (PBO)-controlled study of QTP in
combination with Li or DVP was conducted at
27 sites in the United States. The study protocol
was reviewed and approved by the appropriate
institutional review board in accordance with the
standards and guidelines established in the current
amendment of the Declaration of Helsinki, and
consistent with good clinical practice and applic-
able regulatory requirements. Written informed
consent was obtained from all patients prior to any
Adult patients (‡18 years) hospitalized for a DSM-
IV diagnosis of bipolar I disorder, most recent
episode manic, who had been treated with Li or
DVP for at least 7 of the 28 days immediately prior
to randomization (day 1) were eligible for selection.
A history of at least one documented manic or
mixed episode prior to the episode responsible for
the current hospitalization was required for selec-
tion. At screening and randomization, subjects
were selected who had a Young Mania Rating
Scale (YMRS) score of at least 20, with a score of
Research Institute; Consulting agreements/advisory boards – AstraZeneca, Bristol-Myers Squibb, Eli Lilly Research Laboratories, Glaxo
SmithKline Pharmaceuticals, Janssen Pharmaceutica, Johnson & Johnson Pharmaceutical Research & Development, Pfizer, Pharmaceutical
Research Institute (PRI), Ortho McNeil Pharmaceutical, UCB Pharma, Novartis Pharmaceuticals; Financial interests/stock ownership – none. JA
Mullen, M Brecher, NA Devine, DE Sweitzer are all employees of AstraZeneca Pharmaceuticals.
Sachs et al.
at least 4 on two of the four core YMRS items of
irritability, speech, content, and disruptive/aggres-
sive behavior. Patients were also required to have a
score of at least 4 for overall bipolar illness on the
Clinical Global Impressions-Bipolar (CGI-BP)
Severity of Illness scale.
Pregnant or lactating women and those of child-
bearing potential not using a reliable method of
contraception were excluded from participating in
the study. Patients whose current manic episode
was due to a medical condition were also excluded.
Other patients who were excluded were those
meeting DSM-IV criteria for rapid cycling, those
who had required hospitalization for 3 or more
weeks for the index manic episode, or those with
known intolerance or lack of response to QTP or
clozapine. The continuous daily use of benzodiaze-
pines, in excess of 4 mg/day of lorazepam or the
equivalent, was also not allowed during the month
preceding screening. Patients requiring the use of
antihypertensive medications, unless stable for at
least 1 month, or the use of antidepressants during
the screening period (day )7 to 1) or within a
period of five half-lives of the drug prior to study
randomization, were also ineligible. The use of
depot haloperidol and fluphenazine (within one
injection cycle), and certain cytochrome P450 3A4
inhibitors and inducers, thioridazine, or any
experimental drugs within 2 weeks prior to ran-
domization was not permitted. Also excluded were
patients who had a history of clinically significant
Eligible subjects underwent assessments in the
7 days prior to randomization that included a
complete medical and psychiatric history, inclu-
ding YMRS and the CGI-BP Severity of Illness
assessments, a complete physical examination, vital
signs including weight, hematologic and clinical
chemistry testing, thyroid function tests, and a
12-lead electrocardiogram (ECG).
Patients who met selection criteria began or
continued treatment with Li or DVP on or before
day 1 (randomization). Assignment and dose
adjustments to either Li or DVP were based on
medical criteria, including previous response if
known, and clinician discretion in order to achieve
symptom control, minimize side effects, and
achieve target trough serum concentrations of Li
and DVP within the established therapeutic range
(0.7–1.0 mEq/L Li or 50–100 lg/mL DVP).
Quetiapine was administered twice daily (morn-
PBO, in a double-blind fashion. Investigators
remained blinded to study treatment throughout
thetrial.Initial target doseswere100 mg/dayatday
1, 200 mg/day at day 2, 300 mg/day at day 3, and
400 mg/day at day 4. The dose of QTP was adjusted
600 mg/day at day 5, and 200 and 800 mg/day at
days 6to 21. Guidelines discouraged clinicians from
of at least 600 mg/day QTP had been attempted.
Lorazepam, zolpidem, chloral hydrate, and zale-
plon within specified dose limitations were allowed
during the study. The maximum permissible doses
for lorazepam were 6 mg/day from screening to the
day prior to randomization, 4 mg/day from days 1
to 4, 2 mg/day from days 5 to 7, and 1 mg/day from
days 8 to 10, after which lorazepam was not
allowed. Only one sleep medication was allowed
per day throughout the trial, with a maximum dose
of 10 mg/day zolpidem; chloral hydrate 2 g/day
from days 1 to 7, then 1 g/day; and 20 mg/day
zaleplon. If necessary, intramuscular haloperidol
could be used for severe agitation, but only during
the screening period.
Patients were required to remain in the hospital
for the first 7 days of the randomized period. After
this time, they could be treated as either inpatients
or outpatients as clinically indicated.
Assessments and vital sign measurements were
performed at baseline (day 1) and at days 4, 7, 10,
14, and 21. The primary efficacy variable was the
change from baseline in YMRS total score at the
final assessment (15).
Secondary efficacy variables included YMRS
response rate, defined as the percentage of patients
with ‡50% decrease from baseline in the YMRS
score; clinical remission, defined as an end-point
YMRS score £12 (16); change from baseline in
CGI-BP Severity of Illness score (17); CGI-BP
Global Improvement scale score; MADRS total
score (18); Positive and Negative Syndrome Scale
(PANSS) total score (19); PANSS Activation
subscale score; PANSS Supplemental Aggression
Risk subscale score (20); and Global Assessment
Scale (GAS) score (21).
Scales for the assessment of neurologic adverse
events included the modified Simpson-Angus Scale
(SAS) (22) and the Barnes Akathisia Rating
(BARS) Scale (23).
The rate of treatment-emergent depression was
also monitored, defined as a Montgomery-Asberg
Quetiapine combination therapy in bipolar mania
Depression Rating Scale (MADRS) score ‡18, with
an increase from baseline of ‡4 at any two
consecutive assessments or at the last observation.
Patients were examined and questioned on all
study days regarding any adverse events. Safety
concomitant medication records, change from
baseline to day 21 in clinical laboratory assessments
(including hematology and chemistry), vital signs,
ECG, physical examination, and weight. Adverse
events included any treatment-emergent symptoms
clinically significant changes in laboratory tests,
vital signs, weight, or ECG.
Statistical analyses were performed using the
efficacy and safety populations. The safety popu-
lation included all randomized patients who
received at least one dose of study drug. All
patients who received at least one dose of study
medication and at least one postbaseline efficacy
assessment were included in the intent-to-treat
(ITT) population. Data were analyzed using a last
observation carried forward (LOCF) method.
ANCOVA was used to compare changes from
baseline on the YMRS, MADRS, and PANSS
scales, with the baseline values and clinician-
outcomes were analyzed with Cochran-Mantel-
Haenszel techniques or logistic regression (strati-
fied by clinician-assigned Li or DVP). All statistical
tests were 2-tailed with a significance level of 0.05.
One hundred ninety-one patients were randomly
assigned to QTP (n ¼ 91) or PBO (n ¼ 100). One
patient in the QTP group discontinued prior to
receiving treatment. Overall discontinuations after
beginning treatment, including discontinuations
for adverse events, were more frequent in the
PBO group (51.0%) than the QTP group (37.8%)
(p ¼ 0.073). Of the 191 randomized patients, 81
and 89 in the QTP and PBO groups, respectively,
were included in the ITT population (Fig. 1).
Patients from one complete center were exclu-
ded from the ITT analysis because of a center-
specific protocol violation. Baseline demographics
(Table 1) and episode characteristics (Table 2)
were comparable in each of the ITT groups. The
mean age at onset of bipolar disorder for all
patients was 23.3 years. The two treatment groups
had a similar proportion of patients diagnosed as
having bipolar disorder, characterized as moder-
ate, severe without psychotic features, or severe
with psychotic features.
Almost all patients were treated with a mood-
stabilizer prior to randomization. In each treat-
ment combination, approximately half of the
patients had taken their trial mood-stabilizer for
22 or more days prior to randomization (50%
QTP + Li; 48.6% PBO + Li; 51.0% QTP +
DVP; 55.8% PBO + DVP). Few patients began
mood-stabilizer treatment at randomization: one
patient in the QTP + Li arm; one in the PBO +
Li arm; two in the QTP + DVP arm; and four in
the PBO + DVP arm. Only one patient (receiving
PBO + DVP) had not been treated with mood-
stabilizer prior to randomization.
A comparable number of subjects in each
randomized arm received either Li or DVP: 35
patients received Li and 55 received DVP in the
QTP group, while in the PBO group 42 patients
received Li and 58 DVP.
The mean last-week dose in all patients receiving
QTP was 504 mg/day. Among responders, the
received ‡400 mg/day and 59% received ‡600 mg/
was 584 mg/day:91%
Fig. 1. Disposition of study population.
Sachs et al.
day. Mean serum concentrations of DVP and Li in
the QTP-treated (day 21: DVP 65 lg/mL; Li
0.78 mEq/L) and PBO-treated groups (day 21:
DVP 65 lg/mL; Li 0.71 mEq/L) were clinically
Over the 3-week double-blind period, median Li
between 0.67 (day 4) and 0.85 mEq/L (day 21) in
the QTP group, and between 0.70 (day 4) and
0.80 mEq/L (days 10 and 14) in the PBO group.
Median valproate serum levels ranged between 68
(days 10 and 14) and 72 lg/mL (day 21) in the QTP
group, and between 82 (day 10) and 66 lg/mL (day
21) in the PBO group. There was no significant
difference in mean Li or DVP serum levels between
the QTP and PBO groups at any visit.
Known prior therapeutic responses to Li and
DVP indicated that 78.2% of patients in the QTP
group and 81.0% of PBO-treated patients receiving
Li during the study had previously shown minimal
to much improvement with Li. Among these
Li-treated patients, 18.8% in the QTP group and
16.2% in the PBO group had an unknown prior
response to Li. Of patients chosen to receive DVP
at randomization, 71.5% and 76.9% previously
showed at least a minimal response to DVP, while
prior response was unknown in 22.4% and 17.3%
in the QTP and PBO groups, respectively. No
patients were known to have deteriorated on Li or
DVP in the past and few (four in each treatment
group) were known to have not responded to their
study mood-stabilizer in the past.
Lorazepam use was limited and consistent with
protocol requirements. The trend for greater use of
lorazepam in the PBO + Li/DVP group from days
3 to 10 did not reach statistical significance
(p ¼ 0.371). Overall, there was no statistical dif-
ference between groups in the use of sleep medi-
cations during the study (p ¼ 0.283).
Patients treated with QTP and Li/DVP showed
significantly greater improvement (p ¼ 0.021) at
day 21 in YMRS score from baseline to end-point
compared with patients treated with PBO + Li/
DVP (Table 3). Mean scores on 10 of the 11
individual YMRS item scores were reduced more
in the QTP than in the PBO group and showed at
least a 1-point change from baseline. Items on the
YMRS that showed the greatest change from
baseline at end-point in the QTP + Li/DVP
group, and the most differentiation versus PBO,
included irritability (mean change )1.6 versus
)0.7), disruptive/aggressive behavior ()1.6 versus
)0.7), sleep ()1.2 versus )0.6), and increased
motor activity ()1.4 versus )0.9), respectively.
The YMRS response rate at day 21 was signi-
ficantly greater for QTP-treated patients than for
those receiving Li/DVP alone (54.3% versus
32.6%; p ¼ 0.005). Response rates in patients
receiving a combination of QTP and Li (56.3%)
or QTP and DVP (53.1%) were similar. In patients
randomized to PBO, the response rate in those
receiving DVP versus Li was 36.5% versus 27.0%.
The differences were not analyzed statistically as
the study was not prospectively powered to com-
pare Li with DVP within each treatment group.
There were no statistically significant differences
in the percentage of patients achieving a YMRS
responsewith QTP combination
patients who reported somnolence compared with
those who did not (59% and 51%, respectively).
In addition, there was no statistically signifi-
cant interaction between the treatment effect, as
Table 1. Baseline demographics and clinical characteristics (ITT, LOCF)
(n ¼ 81)
(n ¼ 89)
(n ¼ 170)
BMI ¼ body mass index; ITT ¼ intent-to-treat; LOCF ¼ last
observation carried forward.
Table 2. Baseline disease criteria (ITT, LOCF)
(n ¼ 81)
(n ¼ 89)
(n ¼ 170)
Episode type (%)a
Manic severe without
Manic severe with
Known duration of
illness (mean years)
Number of manic/mixed
past year (median)
Number of depressive
past year (median)
17.6 18.0 17.8
aEpisode type by DSM-IV criteria.
bIncludes three patients with mixed episodes.
YMRS ¼ Young
LOCF ¼ last observation carried forward.
Mania RatingScale; ITT ¼ intent-to-treat;
Quetiapine combination therapy in bipolar mania
measured by YMRS score, and baseline psychotic
symptoms (p > 0.25) measured by both the
PANSS-positive subscale and three PANSS items
that were summed to provide a measure of baseline
psychosis (delusions, conceptual disorganization,
and hallucinatory behavior).
More patients in the QTP + Li/DVP-treated
group (45.7%) than the group receiving Li/DVP
alone (25.8%) achieved remission of symptoms on
the YMRS by day 21, and the difference in favor of
QTP was statistically significant (p ¼ 0.007). The
combination of QTP and Li (46.9%) or QTP and
DVP (44.9%) achieved similar rates of remission.
Improvement in CGI-BP Severity of Illness
scores was observed for both treatment groups,
although QTP-treated patients consistently dem-
onstrated greater decreases, which reached statis-
tical significance at day 10 and all subsequent
assessments. The change in CGI-BP Severity of
Illness scores at day 21 was )1.38 in QTP-treated
patients and )0.78 in those receiving Li/DVP
monotherapy (p ¼ 0.001) (Table 3).
A significantly greater number of patients in the
QTP group were also rated as ?much improved? or
?very much improved? on the CGI-BP Global
Improvement scale (50.6%)
patients in the PBO group (31.5%) (p ¼ 0.012).
Repeat analysis of the primary end-point inclu-
ding the center that was excluded from the ITT
population because of center-specific protocol
violations did not alter the conclusions of the
study or their interpretation.
Scores on the PANSS Supplemental Aggression
Risk subscale weresimilar
decreased for both treatment groups during the
3-week trial. However, there was a statistically
significant treatment effect at day 21 (p ¼ 0.020) in
favor of QTP + Li/DVP ()4.64) versus the
Table 3. Summary of efficacy results at primary end-point/day 21 (ITT, LOCF)
Quetiapine (n ¼ 81)Placebo (n ¼ 89)
Mean baseline valueMean change Mean baseline valueMean change
YMRS total scoreLi/DVP
YMRS response (%)b
YMRS remission (%)c
CGI-BP Severity of
Improvement response (%)d
Global Assessment Scale
aChange from baseline to end-point Li/DVP + QTP versus Li/DVP + PBO.
b‡50% decrease from baseline in the YMRS total score.
cYMRS score £12.
dProportion of patients ?much? or ?very much? improved.
YMRS ¼ Young Mania Rating Scale; CGI-BP ¼ Clinical Global Impressions-Bipolar; ITT ¼ intent-to-treat; LOCF ¼ last observation
carried forward; N/A ¼ not applicable; – ¼ data not available; Li ¼ lithium; DVP ¼ divalproex; QTP ¼ quetiapine; PBO ¼ placebo.
Sachs et al.
PBO + Li/DVP
numerical difference favoring QTP + Li/DVP for
improvement in activation score approached sta-
tistical significance (p ¼ 0.057).
Total scores on the PANSS were similar at
baseline and decreased for both treatment groups
by day 21 ()12.47 in the QTP + Li/DVP group
versus )10.14 in the PBO + Li/DVP group,
p ¼ 0.323).
In patients with psychotic features at screening,
QTP treatment resulted in a statistically significant
improvement in PANSS total ()17.17 versus
)5.93; p ¼ 0.006) and positive ()7.56 versus
)3.25; p ¼ 0.002) subscale scores by day 21.
Although patients meeting DSM-IV criteria for
a full mixed episode at baseline were excluded,
mean baseline MADRS scores were moderate (13.7
in the QTP + Li/DVP group and 14.2 in those
randomized to PBO). Total MADRS scores
decreased in both groups, and at day 21 an
improvement of )3.36 in the QTP-treated group
versus )2.79 in those receiving Li/DVP mono-
therapy was observed (p ¼ 0.650).
By final assessment, there was a greater improve-
ment in GAS scores for the QTP-treated patients
(15.32) than those treated with Li/DVP alone
(11.49), and the difference approached statistical
significance (p ¼ 0.075).
group()2.84) (Fig. 2).The
Safety and tolerability
Most adverse events were mild or moderate and
withdrawal rates associated with adverse events
were similar in both groups (5.5% for QTP versus
6.0% for PBO) (Table 4). The incidence of adverse
events occurring in at least 10% of patients in the
QTP + Li/DVP group were somnolence (40%
versus 10%; p < 0.001), headache (27% versus
21%; p ¼ 0.655), dry mouth (19% versus 4%;
p ¼ 0.005), asthenia (11% versus 3%; p ¼ 0.052),
postural hypotension (11% versus 3%; p ¼ 0.052),
and dizziness (10% versus 6%; p ¼ 0.439) versus
the PBO + Li/DVP group, respectively (Table 5).
Patients treated with QTP + Li/DVP exhibited
low levels of extrapyramidal symptoms (EPS)
(including akathisia) that were similar to those
reported by the PBO + Li/DVP group. The mean
changes in the modified SAS were )1.0 and )0.3
for QTP versus PBO, respectively. BARS scores
were low at baseline for both the QTP-treated
patients (0.7) and controls (0.6), and the mean
change was )0.4 and 0, respectively.
The mean weight at baseline of QTP-treated
patients in the safety population was similar to the
control group (88.4 kg versus 86.6 kg), as was the
body mass index (BMI) (30.0 versus 29.8). At day
p = 0.020
p = 0.057
QTP + Li/DVP
PBO + Li/DVP
Fig. 2. Change in Positive and Negative Syndrome Scale
(PANSS) Supplemental Aggression Risk and PANSS Activa-
tion subscale scores (intent-to-treat, ITT; last observation car-
riedforward,LOCF).Li ¼ lithium;DVP ¼ divalproex;QTP ¼
quetiapine; PBO ¼ placebo. Baseline scores: Aggression Risk –
QTP 18.3, PBO 17.7; Activation – QTP 17.1, PBO 16.5.
Table 4. Withdrawals from study (safety population)
(n ¼ 91)
(n ¼ 100)
Progression of disease
Lost to follow-up
Lack of efficacy
AE ¼ adverse event.
Table 5. Incidence of most common adverse events occurring in ‡10% of
patients (safety population)
(n ¼ 90)
(n ¼ 100)p-value
Event type (%)
COSTART, Coding Symbols for a Thesaurus of Adverse Reac-
tion Terms are used.
Patients with multiple events of the same type are counted once.
p-values derived from chi-squared test versus placebo.
Quetiapine combination therapy in bipolar mania
21, there was a modest increase in weight in QTP-
treated patients (mean increase 1.60 kg; 3.53 lbs)
compared with the PBO group (0.36 kg; 0.79 lbs).
Patients in the low to normal BMI categories at the
start of the trial gained more weight (2.60 kg;
5.72 lbs) than those already overweight or obese at
entry (0.10 kg; 0.22 lbs). There was no difference in
the proportion of patients with a 7% or greater
increase in weight (day 21: 3.9% versus 1.2%,
respectively; p ¼ 0.335; LOCF). Comparison of
the separate effects of Li or DVP on BMI and body
weight revealed no significant differences.
There were no notable differences between
treatment groups in mean change from baseline
patients exhibited a greater mean change in total
thyroxine ()18.46 nmol/L versus 3.23 nmol/L) and
a slightly greater mean increase in thyroid-stimu-
0.42 mIU/L) than patients receiving PBO + Li/
DVP. The mean change in free thyroxine was
0.22 pmol/L inQTP-treated
)0.65 pmol/L in patients receiving PBO + Li/
DVP. There were no clinically significant differ-
ences in glucose laboratory data from random
sampling between the QTP and PBO treatment
groups, and all other chemistry findings were
unremarkable. There were no apparent clinically
significant systematic differences between the two
treatment groups in changes in pulse, blood
pressure, or ECG parameters over the course of
There was also no statistically significant differ-
ence between groups in the rate of emergent
depression, which was 17.3% for the QTP group
versus 13.5% for the group receiving Li or DVP
alone (p ¼ 0.469).
Quetiapine demonstrated superior efficacy in the
control of the symptoms of mania when used in
combination with Li or DVP compared with the use
of Li or DVP alone. In addition to the positive
result on the primary efficacy measure (change from
baseline in YMRS score at day 21; p ¼ 0.021) QTP
combination therapy was statistically and clinically
superior on several other efficacy parameters.
Quetiapine combination therapy resulted in a
statistically superior clinical response, defined as a
50% reduction on the primary efficacy variable
(YMRS). This definition of clinical response is an
established criterion for validating that efficacy on
a primary end-point is clinically relevant (24).
The CGI-BP Severity of Illness scale is widely
recognized as a measure of disease severity and
global improvement in bipolar illness and a close
correlate of clinical effectiveness (17). In this study,
the statistically significant improvement in clinical
response was supported by significant improve-
ment in CGI-BP Severity of Illness and Global
Improvement scale scores in patients receiving
combination therapy with QTP.
A statistically significant improvement in the
proportion of patients in remission by the end of
the study, and consistently favorable changes in the
other efficacy variables, were also observed. Taken
together, these findings strongly suggest that a
clinically valid treatment effect associated with
QTP treatment was seen during the trial. The data
are also consistent with previous clinical studies of
QTP undertaken in smaller populations of patients
with bipolar disorder (7–12, 25, 26).
The treatment effect observed during the study
was independent of somnolence, as there was no
correlation between the proportion of patients who
reported somnolence and those who achieved a
clinical response. Efficacy in the treatment of
mania was also independent of the antipsychotic
effects of QTP, as improvement on the YMRS was
not correlated with the presence of psychotic
symptoms at baseline.
Significant improvement in patients with psy-
choticfeatures at baseline was observed, includinga
highly significant effect on positive symptoms
(p ¼ 0.002). Overall, QTP-treated patients showed
statistically and clinically significant improvement
in supplemental aggression-risk assessment com-
pared with controls. These findings are likely to be
clinically relevant as aggression, especially in the
form of violence directed toward self, family, and
health care workers, is an important therapeutic
target in the effective management of mania asso-
ciated with bipolar disorder. Psychotic symptoms
and aggression are also highly distressing, and may
necessitate the use of physical restraint. The
improvement in aggression and psychotic symp-
toms seen in this study may facilitate treatment in
less restrictive settings and is consistent with the
known effects on hostility and aggression of QTP
treatment in patients with schizophrenia (27).
This study was not designed to establish efficacy
in bipolar depression, but it was important to
determine whether the combination of QTP and Li
or DVP precipitated depression. No significant
effect on treatment-emergent depression was ob-
served on the MADRS with QTP combination
treatment over the course of the study. The
MADRS and CGI-BP Severity of Illness scale
findings indicate that improvements in the symp-
toms of mania achieved by the addition of QTP to
Li or DVP were not associated with a worsening of
Sachs et al.
the other symptoms of bipolar disorder. These
findings are of interest as they indicate clinically
relevant effects of QTP in the treatment of bipolar
mania and justify further studies of its potential
role in bipolar depression and long-term mainten-
Conventional dose–response analyses are not
possible in studies of a flexible-dose design. How-
ever, an estimate of the QTP dose in responders
gives strong guidance for its effective therapeutic
range: 91% of patients who responded to QTP
were administered at least 400 mg/day, and most
responders received about 600 mg/day.
The simple initial dosing regimen, and target
dose (600 mg/day) supported by this study were
generally well-tolerated and may have advantages,
particularly in acute treatment, where the goal is
rapid stabilization. Overall, a greater proportion of
patients treated with Li or DVP alone discontinued
therapy than those receiving QTP combination
therapy. The same pattern was observed in patients
who discontinued due to adverse events.
Weight gain in the QTP group was consistent
with that seen in schizophrenia patients treated
with QTP monotherapy (28–30). The weight
increase in QTP-treated patients was not associ-
ated with any withdrawals. The weight gain in
QTP-treated patients was lowest among those
already overweight at baseline for whom the risk
of weight-related complications is of greater con-
cern, i.e. only 0.1 kg (0.22 lbs) in those overweight
by baseline BMI category. There were few adverse
events potentially related to diabetes mellitus
observedin this study.
between the treatment groups were detected in
terms of changes in glucose values.
Although already low at baseline, the BARS and
modified SAS scores indicated a lower magnitude
and incidence of drug-induced neurologic side
effects, such as rigidity, tremor, salivation, and
akathisia in QTP-treated patients. The addition of
QTP to Li or DVP did not result in an increased
incidence of EPS, including akathisia, across the
full range of doses used in this study.
There were no clinically important changes in
laboratory, vital sign, or ECG parameters in
patients treated with QTP in this study, with the
exception of changes in thyroxine concentrations
that were not associated with clinical hypothyroid-
The results reported here are generally consis-
tent with mania combination therapy studies with
other atypical antipsychotics (13, 14), supporting
a possible class effect for the atypical antipsych-
otics in combination therapy. However, the pos-
sibility of clinically significant differences between
individual atypical agents cannot be ruled out.
Tolerability profiles represent an important area
of differentiation between the atypical antipsy-
chotic therapies (13, 14) and further head-to-head
studies to confirm any differences between the
tolerability profiles of these atypical antipsychot-
ics are needed.
The protocol inclusion and exclusion criteria
used in this study were selected to ensure that the
findings were generalizable to a broad range of
patients with bipolar mania, and to control for
characteristics that could confound the results.
That the study drew patients from private practice,
community health centers, Veterans Affairs cen-
ters, and community hospitals suggest the findings
will be applicable to a broad range of practice
Higher minimum YMRS symptom criteria,
including the four double-weighted items of irrita-
bility, speech, content, and disruptive/aggressive
behavior, may have contributed to the total mean
YMRS at baseline in all patients being higher in
this study than in the similarly designed combina-
tion therapy studies of Sachs et al. (13) and Tohen
et al. (14). The protocol requirement for patients to
be hospitalized on entering the study and addi-
tional minimum CGI-BP criteria may also have
tended to select for patients with more severe
illness. As such, the findings of this study may be
more generalizable to the population of hospital-
ized patients seen in routine clinical practice, who
are often more severely ill than those studied under
controlled conditions in clinical trials.
The rationale for requiring a minimum period
of mood-stabilizer treatment in the inclusion
criteria was to reflect a common clinical practice
supported by recent guidelines (3, 4) in which a
treating physician might consider adding QTP to
the treatment regimen of a patient who had
already initiated traditional
monotherapy. Approximately half of the patients
within each treatment combination had been
treated for at least 3 consecutive weeks with the
same mood-stabilizer that they received at rand-
Although the average serum Li levels at day 21
were at the lower end of the optimal dose range,
higher levels were achieved during the course of the
study and differences in treatment outcome within-
and between-groups did not appear to be affected
by Li levels. Moreover, Li levels in the present
study reflect the clinical judgment of the study
clinicians, who were encouraged to adjust the dose
to maximize response and tolerability within the
range of 0.7–1.0 mEq/L. In a similarly designed
monotherapy study, mean Li levels of 0.77 mEq/L
Quetiapine combination therapy in bipolar mania
were shown to be therapeutic and significantly
superior to PBO (31). It is possible that patients
with bipolar mania who tolerate higher serum Li
levels (0.8–1.2 mEq/L) may achieve greater efficacy
improvement than noted in this study. However,
the short duration of hospital stays, and long-term
adherence issues with Li make combination treat-
ment more attractive and pragmatic for many
Patients with rapid cycling and mixed episodes
were excluded from the present study in order to
control for characteristics with the potential to
confound the study findings. The exclusion of
patients with mixed episodes was to ensure that
treatment-emergent depression was not obscured
by high baseline MADRS scores. The inclusion of
patients with mixed episodes may also have
introduced a greater imbalance in the clinician-
assignment of Li and DVP (32). Similarly, the
inclusion of patients with rapid cycling might
have artificially inflated remission rates, as higher
rates of spontaneous remission can occur in these
individuals. Preliminary data from other studies
suggest that QTP may be also be effective in the
treatment of mixed mania (12), and rapid cycling
(7, 25), although further study is warranted in
In summary, this study has addressed the effi-
cacy and safety of QTP in combination with
standard therapies (Li and DVP) in patients with
bipolar mania. The efficacy findings, across a range
of end-points, support the use of QTP as combi-
nation therapy in bipolar mania. The findings also
indicate that QTP was generally well-tolerated, and
most adverse events were mild, transient, and
similar to those in schizophrenia trials where
QTP was used as monotherapy (28–30).
The research presented was supported by AstraZeneca Phar-
maceuticals, Wilmington, Delaware, USA.
1. Woods SW. The economic burden of bipolar disease.
J Clin Psychiatry 2000; 61 (Suppl. 13): 38–41.
2. Keck PE Jr, McElroy SL, Strakowski SM et al. 12-month
outcome of patients with bipolar disorder following
hospitalization for a manic or mixed episode. Am J
Psychiatry 1998; 155: 646–652.
3. American Psychiatric Association. Practice guideline for
the treatment of patients with bipolar disorder (revision).
Am J Psychiatry 2002; 159: 1–50.
4. Suppes T, Dennehy EB, Swann AC. Report of the Texas
Consensus Conference Panel on Medication Treatment of
Bipolar Disorder 2000. J Clin Psychiatry 2002; 63: 288–
5. Young AH, Macritchie KA, Calabrese JR. Treatment of
bipolar affective disorder. BMJ 2000; 321: 1302–1303.
6. Suppes T, Rush AJ Jr, Kraemer HC, Webb A. Treatment
algorithm use to optimize management of symptomatic
patients with a history of mania. J Clin Psychiatry 1998;
59: 89–96; quiz 97–98.
7. Ghaemi S, Goldberg J, Ko J, McNally E. Quetiapine
treatment of rapid-cycling bipolar disorder: an open
prospective study. Int J Neuropsychopharmacol 2002; 5
(Suppl. 1): S110.
8. Dunayevich E, Tugrul RN, Strakowski SM. Quetiapine in
the treatment of mania. Poster presented at the 7th World
Congress of Biological Psychiatry, July 2001, Berlin,
9. Zarate CA Jr, Rothschild A, Fletcher KE, Madrid A,
Zapatel J. Clinical predictors of acute response with
quetiapine in psychotic mood disorders. J Clin Psychiatry
2000; 61: 185–189.
10. Chisholm K, Dennehy E, Suppes T. Clinical response to
quetiapine add-on for treatment of refractory bipolar
disorder. Paper presented at 2nd International Forum on
Mood and Anxiety Disorders; November 28–December 1,
2001, Monte Carlo, Monaco.
11. Sajatovic M, Brescan DW, Perez DE et al. Quetiapine
alone and added to a mood stabilizer for serious mood
disorders. J Clin Psychiatry 2001; 62: 728–732.
12. DelBello MP, Schwiers ML, Rosenberg HL, Strakowski
SM. A double-blind, randomized, placebo-controlled
study of quetiapine as adjunctive treatment for adolescent
mania. J Am Acad Child Adolesc Psychiatry 2002; 41:
13. Sachs GS, Grossman F, Ghaemi SN, Okamoto A, Bowden
CL. Combination of a mood stabilizer with risperidone or
haloperidol for treatment of acute mania: a double-blind,
placebo-controlled comparison of efficacy and safety. Am J
Psychiatry 2002; 159: 1146–1154.
14. Tohen M, Baker RW, Altshuler LL et al. Olanzapine
versus divalproex in the treatment of acute mania. Am J
Psychiatry 2002; 159: 1011–1017.
15. Young RC, Biggs JT, Ziegler VE, Meyer DA. A rating
scale for mania: reliability, validity and sensitivity. Br J
Psychiatry 1978; 133: 429–435.
16. Tohen M, Jacobs TG, Grundy SL et al. Efficacy of
olanzapine in acute bipolar mania: a double-blind, pla-
cebo-controlled study. The Olanzapine HGGW Study
Group. Arch Gen Psychiatry 2000; 57: 841–849.
17. Spearing MK, Post RM, Leverich GS, Brandt D, Nolen
W. Modification of the Clinical Global Impressions (CGI)
Scale for use in bipolar illness (BP): the CGI-BP. Psychi-
atry Res 1997; 73: 159–171.
18. Montgomery SA, Asberg M. A new depression scale
designed to be sensitive to change. Br J Psychiatry 1979;
19. Kay SR, Fiszbein A, Opler LA. The positive and negative
syndrome scale (PANSS) for schizophrenia. Schizophr Bull
1987; 13: 261–276.
20. White L, Harvey PD, Opler L, Lindenmayer JP. Empirical
assessment of the factorial structure of clinical symptoms
in schizophrenia. A multisite, multimodel evaluation of the
factorial structure of the Positive and Negative Syndrome
Scale. The PANSS Study Group. Psychopathology 1997;
21. Endicott J, Spitzer RL, Fleiss JL, Cohen J. The global
assessment scale. A procedure for measuring overall
severity of psychiatric disturbance. Arch Gen Psychiatry
1976; 33: 766–771.
Sachs et al.
22. Simpson GM, Angus JW. A rating scale for extrapyram-
idal side effects. Acta Psychiatr Scand Suppl 1970; 212:
23. Barnes TR. A rating scale for drug-induced akathisia. Br J
Psychiatry 1989; 154: 672–676.
24. Montgomery DB. ECNP Consensus Meeting March 2000
Nice: guidelines for investigating efficacy in bipolar disor-
der. European College of Neuropsychopharmacology. Eur
Neuropsychopharmacol 2001; 11: 79–88.
25. Vieta E, Parramon G, Padrell E et al. Quetiapine in the
treatment of rapid cycling bipolar disorder. Bipolar Disord
2002; 4: 335–340.
26. Ghaemi SN, Rosenquist K, Ko J, Goodwin FK, Baldessa-
rini RJ. Effects of newer and older antidepressants in
bipolar depression (abstract). Bipolar Disord 2003; 5
(Suppl. 1): 49.
27. Chengappa KN, Goldstein JM, Greenwood M, John V,
Levine J. A post hoc analysis of the impact on hostility and
agitation of quetiapine and haloperidol among patients
with schizophrenia. Clin Ther 2003; 25: 530–541.
28. Arvanitis LA, Miller BG. Multiple fixed doses of ?Seroquel?
(quetiapine) in patients with acute exacerbation of schizo-
phrenia: a comparison with haloperidol and placebo. The
Seroquel Trial 13 Study Group. Biol Psychiatry 1997; 42:
29. Copolov DL, Link CG, Kowalcyk B. A multicentre,
double-blind, randomized comparison of quetiapine (ICI
204,636, ?Seroquel?) and haloperidol in schizophrenia.
Psychol Med 2000; 30: 95–105.
30. Small JG, Hirsch SR, Arvanitis LA, Miller BG, Link CG.
Quetiapine in patients with schizophrenia. A high- and
low-dose double-blind comparison with placebo. Seroquel
Study Group. Arch Gen Psychiatry 1997; 54: 549–557.
31. Paulsson B, Huizar K. Quetiapine monotherapy versus
placebo for acute bipolar mania (STAMP 2) (abstract).
Bipolar Disord 2003; 5: 74.
32. Bowden CL, Brugger AM, Swann AC et al. Efficacy of
divalproex versus lithium and placebo in the treatment of
mania. The Depakote Mania Study Group. JAMA 1994;
Appendix: Study Investigators
Claudia Baldassano MD, Philadelphia, PA, USA
Mohammid Bari MD, Chula Vista, CA, USA
Louise Beckett MD, Oklahoma City, OK, USA
Ronald Brenner MD, Lawrence, NY, USA
David Brown MD, Austin, TX, USA
Franca Centorrino MD, Belmont, MA, USA
James C-Y. Chou MD, New York, NY, USA
William Coryell MD, Iowa City, IA, USA
John Crayton MD, Hines, IL, USA
Andrew Cutler MD, Winter Park, FL, USA
Lori Davis MD, Tuscaloosa, AL, USA
Louis Fabre MD, Houston, TX, USA
David Feifel MD, San Diego, CA, USA
Arman Goenjian MD, Long Beach, CA, USA
Michael Lesem MD, Bellaire, TX, USA
Joseph McEvoy MD, Butner, NC, USA
Charles Merideth MD, San Diego, CA, USA
Michael Plopper MD, San Diego, CA, USA
William Privitera MD, Austin, TX, USA
Joachim Raese MD, Riverside, CA, USA
Samuel C. Risch MD, Charleston, SC, USA
Alan Rosenbaum MD, Farmington Hills, MI, USA
David Schnur MD, Elmhurst, NY, USA
Robert T. Segraves MD, PhD, Cleveland, OH, USA
Joyce Small MD, Indianapolis, IN, USA
Patricia Suppes MD, PhD, Dallas, TX, USA
Kathleen Toups MD, Walnut Creek, CA, USA
Richard Weisler MD, Raleigh, NC, USA
Craig Wronski MD & Himasiri DeSilva MD, Santa Ana, CA,
Quetiapine combination therapy in bipolar mania