- Email: [email protected]
mixed episode remission
Journal of Psychiatric Research 44 (2010) 8–14
Contents lists available at ScienceDirect
Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/jpsychires
Rapid antipsychotic response with ziprasidone predicts subsequent acute manic/mixed episode remission Terence A. Ketter a,*, Ofer Agid b, Shitij Kapur c, Antony Loebel d, Cynthia O. Siu e, Steven J. Romano d a
Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA 94305-5723, USA Schizophrenia Program Center for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, Canada c Section on Schizophrenia, Imaging and Therapeutics, Department of Psychological Medicine and Psychiatry, Institute of Psychiatry, King’s College, London, United Kingdom d Pfizer Inc., New York, NY, USA e Data Power (DP) Inc., Ringoes, NJ, USA b
a r t i c l e
i n f o
Article history: Received 27 April 2009 Received in revised form 10 July 2009 Accepted 20 July 2009
Keywords: Rapid antipsychotic response Bipolar mania Remission
a b s t r a c t Objective: To assess rapid antipsychotic efficacy with oral ziprasidone monotherapy in bipolar acute manic/mixed episodes with psychotic features, and predictive value of rapid antipsychotic response for subsequent acute manic/mixed episode remission. Methods: Pooled analysis of two 3-week, randomized, double-blind, placebo-controlled trials of ziprasidone (40–160 mg/d) in inpatients with bipolar I disorder, and a current manic or mixed episode, with (n = 152) or without (n = 246) psychotic features. Psychosis improvement was evaluated by change in SADS-C psychosis score (sum of delusions, hallucinations, and suspiciousness items). Rapid antipsychotic response (P50% decrease in SADS-C psychosis score by Day 4) and acute manic episode response and remission (endpoint P50% MRS decrease, and a MRS score 6 12, respectively) were analyzed. Results: Significantly greater antipsychotic effects were observed by Day 4 with ziprasidone treatment (vs. placebo) and the magnitude of improvement increased significantly with time, in all subjects, in the subgroup of all psychotic subjects, and psychotic subjects with low baseline agitation (p < 0.05). Rapid antipsychotic response predicted subsequent acute manic episode remission independent of ziprasidone or placebo treatment received (p < 0.001, ROC AUC = 0.71) with significant improvement in accuracy of MRS remission prediction when compared to models using early changes in MRS score alone (p = 0.01). Limitations: Post hoc analysis, use of 3 SADS-C psychosis items to assess psychosis. Conclusions: The predictive value of rapid (Day 4) improvement in psychotic symptoms for subsequent (Day 21) remission of acute manic/mixed symptoms may facilitate enhanced therapeutics, in view of the current practice of brief hospitalization for patients with acute manic/mixed episodes with psychotic features. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Bipolar disorder is a common, serious psychiatric disorder that results in substantial financial and human burdens (Kleinman et al., 2003; Goodwin and Jamison 2007; Merikangas et al., 2007). Acute manic and mixed manic episodes constitute considerable clinical challenges, as by definition they entail hospitalization, psychosis, or severe impairment of psychosocial or occupational function (American Psychiatric Association 2000). Indeed, in aggregate, approximately one third of patients with such episodes in contemporary United States Food and Drug Administration (FDA) registration trials have psychotic features (Tohen et al., 1999,
* Corresponding author. Address: Psychiatry and Behavioral Sciences, 401 Quarry Road, Room 2124, Stanford, CA 94305-5723, USA. Tel.: +1 (650) 723 2515; fax: +1 (650) 723 2507. E-mail address: [email protected] (T.A. Ketter). 0022-3956/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jpsychires.2009.07.006
2000b, 2002; Sachs et al., 2002; Yatham et al., 2003; Hirschfeld et al., 2004; Yatham et al., 2004; Khanna et al., 2005; Vieta et al., 2005; Bowden et al., 2006; Suppes et al., 2008). Phase III studies for regulatory approval for the treatment of acute manic and mixed episodes commonly last 3–4 weeks. However, in contemporary clinical practice in the United States, durations of hospitalization for such episodes are commonly one week or less. Thus, clinical practice could be informed by knowledge regarding the degree to which rapid (within the first few days) improvement predicts subsequent (within the next few weeks) clinical status. A key goal in the pharmacologic treatment of acute bipolar manic and mixed episodes with psychotic features is rapid symptomatic improvement. Medications used for this condition commonly include mood stabilizers and/or first- or second-generation antipsychotics (Ketter et al., 2005). In recent years, five second-generation antipsychotics (olanzapine, risperidone, quetiapine, ziprasidone, and aripiprazole) have been approved by the
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
FDA for the monotherapy treatment of acute manic episodes (Tohen et al., 1999, 2000b; Keck et al., 2003a,b; Hirschfeld et al., 2004; Khanna et al., 2005; Potkin et al., 2005; Vieta et al., 2005; Sachs et al., 2006). With the exception of quetiapine immediate release formulation, these agents have also been approved for the monotherapy of acute mixed episodes. Although still controversial, rapid antipsychotic efficacy has been demonstrated in schizophrenia. However, such rapid treatment effects have not been well established in mania, where the time course of antipsychotic response remains to be established. It used to be assumed that early symptomatic improvement associated with antipsychotic treatments was related to non-specific sedative effects, while true antipsychotic response was delayed. This delayed action hypothesis for the treatment of schizophrenia spectrum disorders with antipsychotics has recently been challenged. A 24-h, double-blind, randomized trial of acutely agitated patients with schizophrenia (Kapur et al., 2005) and a meta-analysis of 42 double-blind active- or placebo-controlled trials (Agid et al., 2003) both found early antipsychotic responses with psychotropic treatments, distinct from drug-induced non-specific effects on behavior (e.g. agitation and excitement). Findings from our previous study further support the early antipsychotic action hypothesis, by demonstrating intramuscular ziprasidone produced rapid (within 4 h) effects on both psychosis and agitation in the treatment of subjects with psychotic disorders (Agid et al., 2008). Early studies also suggested that first-generation antipsychotics may yield rapid antipsychotic and antimanic effects in patients with acute mania. Garfinkel et al. (1980) found significant improvements compared to baseline in mean Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham, 1962) score in bipolar patients with acute mania after one week of treatment with haloperidol. Cookson et al. (1981) found that in bipolar patients with acute mania, both pimozide and chlorpromazine yielded significant decreases from baseline in mean Biegel-Murphy Mania Rating Scale (BMMRS) (Beigel et al., 1971) scores during each of the first 3 days after commencing treatment, and this improvement persisted for the remainder of the two-week study. Janicak et al. (1988) found that manic patients had improved mean BPRS and BMMRS scores with both thiothixene and chlorpromazine after 4 days of administration compared to baseline. Brown et al. (1989) found significant improvement in mean Young Mania Rating Scale (YMRS) (Young et al., 1978) score in manic patients after 3 days of taking haloperidol in a comparative study with carbamazepine. Segal et al. (1998) showed that in acute mania patients, haloperidol and risperidone yielded mean YMRS score reductions of 30% and 24%, respectively, by the end of the first week of treatment. Evidence of early onset of antimanic effects in acute mania has also accumulated for second-generation antipsychotics. Tohen et al. (1999) found olanzapine compared to placebo yielded a significantly greater decrease (33.3% vs. 5.21%) in mean YMRS score from baseline after one week of treatment. In studies by Keck et al. (2003b) and Potkin et al. (2005), ziprasidone-treated patients demonstrated significant improvement in mean Mania Rating Scale (MRS) scores as early as Day 2 ( 5.6 and 3.4 points, respectively) compared with placebo-treated patients ( 2.4 and 1.8 points, respectively) (p < 0.05), that was subsequently maintained over the course of the three week studies. Aripiprazole (n = 123) was found to reduce mean YMRS score by 5.8 points after 4 days of treatment, vs. a reduction of 2.4 points with placebo (n = 120, p < .005) (Keck et al., 2003a). Hirschfeld et al. (2004) presented evidence of a rapid antimanic effect for risperidone monotherapy, showing the mean YMRS score for the risperidone group (n = 124) was reduced by 6.5 points after 3 days of treatment, vs. 4.0 points for the placebo group (n = 119) (p < .001). Bowden et al. (2005) showed statistically significant changes from baseline
9
in mean YMRS score between quetiapine ( 8.03 points, n = 107) and placebo ( 4.89 points, n = 95) groups, beginning at Day 7 (p < 0.01) in bipolar acute mania patients. Tohen et al. (2000a) reviewed published controlled, blinded studies regarding the onset of action of antipsychotic medications in the treatment of acute mania. They found that first-generation (chlorpromazine, haloperidol) as well as second-generation (olanzapine, risperidone) antipsychotics had early onset of antimanic effects appearing after 2–6 days of treatment. Together, these data suggest rapid antimanic actions of antipsychotics in the treatment of acute manic/ mixed symptoms in patients with bipolar disorder. Accordingly, we conducted a post hoc pooled analysis of two similarly designed, placebo-controlled pivotal trials to assess the rapid antipsychotic and subsequent antimanic actions of the second-generation antipsychotic, ziprasidone, in the treatment of bipolar disorder patients with acute manic/mixed episodes, with psychotic features, and the clinical implications – specifically the degree to which rapid antipsychotic response predicted subsequent remission of manic symptoms. 2. Methods 2.1. Study subjects The full analysis dataset consisted of a total of 399 subjects pooled from two similarly designed, randomized, double-blind, placebo-controlled, 3-week trials of oral ziprasidone monotherapy (40–160 mg/d) in hospitalized patients with bipolar I disorder (Keck et al., 2003b; Potkin et al., 2005). At study baseline, all subjects were experiencing a current manic (n = 248) or mixed episode (n = 151), with (n = 152) or without (n = 246) psychotic features. Both studies were conducted in accordance with the principles of the Helsinki Declaration and were in compliance with Good Clinical Practice. The protocol and related informed consent form were approved by the local institutional review boards of the participating sites. Patients or their legal representatives provided written and verbal informed consent before any protocol-related procedures were performed. Additional details on the design of these two studies are provided in the primary study publications (Keck et al., 2003b; Potkin et al., 2005). Subjects were male (n = 209) or female (n = 190) inpatients, aged 18 years or older, with a primary diagnosis of bipolar I disorder and a current manic (296.4) or mixed (296.6) episode, defined according to the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (American Psychiatric Association 2000). Subjects had a Mania Rating Scale score (MRS) P 14, and scores of P2 on at least 4 MRS items, at both screening and baseline. 2.2. Clinical design and outcome measures Both studies were randomized, double-blind, placebo-controlled trials, comprising up to 10 days of single-blind, placebo screening/washout period and a 21-day treatment period. Subjects meeting all inclusion and exclusion criteria were randomized to receive up to 3 weeks (21 days) of double-blind treatment with ziprasidone or placebo, in a 2:1 ratio (ziprasidone:placebo). During the screening period, subjects discontinued prohibited concomitant psychotropic medications, as necessary. The duration of washout was at least 3–5 half-lives of the prohibited medications, and a minimum of 3 days for all antipsychotics. In both studies, ziprasidone was initiated at 40 mg twice daily with food (80 mg/d), then titrated to 80 mg twice daily with food (160 mg/d) on Day 2, and then flexibly adjusted by up to 40 mg/d in the range of 40–80 mg twice daily with food (80–160 mg/d) from Day 2 onward.
10
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
The 11-item Mania Rating Scale (MRS), derived from the Schedule for Affective Disorders and Schizophrenia-Change Bipolar (SADS-CB, May 2000 version) scale, and the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987) were assessed at screening, baseline (within 12 h prior to the first dose of study drug), and at Days 2 (SADS-CB only), 4 (SADS-CB only), 7, 14, and 21 (or at the time of early discontinuation). The SADS-CB, a diagnostic interview used to rate manic and depressive symptoms, is a modified version of the Schedule for Affective Disorders and Schizophrenia-Change (SADS-C) (Endicott and Spitzer, 1978). Manic-episode remission was defined as achieving a MRS score of 612 (Ketter et al., 2007). SADS-C psychosis total score was derived from the sum of SADS-C psychosis items (delusions, hallucinations, and suspiciousness). Early antipsychotic response rate was defined as the proportion of subjects achieving P50% improvement from baseline in the SADS-C Psychosis Score at Days 2 or 4. Agitation symptoms at Days 2 or 4 were assessed by the SADS-C agitation item (non-mania related). Primary analyses focusing on the MRS change score at Day 21 have previously been published (Keck et al., 2003b; Potkin et al., 2005).
2.4. Psychotic subgroup analyses Subgroups were based on baseline psychotic status and severity level of agitation symptoms. To reduce potential bias due to regression to the mean (RTM) (Allison et al., in press), the psychotic subgroup was defined based on PANSS positive items (delusions P1, conceptual disorganization P2, and hallucinatory behavior P3) and SADS-C psychosis change score was used as analysis endpoint. Specifically, the psychotic subgroup included all subjects with a baseline score of P4 on at least 1 of the following PANSS positive items: delusions P1, conceptual disorganization P2, and hallucinatory behavior P3. Baseline agitation severity levels were dichotomized, with the low agitation subgroup defined as having a PANSS Excitability Component (PANSS-EC) score of 614 and a maximum score of 3 (moderate) in two of the following 5 items (excitement, hostility, tension, uncooperativeness, and poor impulse). The PANSS-EC score was derived from the sum of 5 items (excitement P4, hostility P7, tension G4, uncooperativeness G8, and poor impulse control G14) (Lindenmayer et al., 2004).
3. Results 2.3. Statistical analysis Mixed model repeated measures (MMRM) analysis for change score from baseline was applied to estimate the time course of response, using all available data at each visit in the full analysis dataset (n = 399). The mixed effects models included fixed effects terms for treatment, visit, treatment-by-visit, baseline score, manic/mixed diagnosis, and study. Treatment and treatment-by-visit interaction terms provided test of ziprasidone effect at Day 2, Day 4, Weeks 1, 2, and 3. Random effects terms included center and subject nested within center. Unstructured covariance model was used to account for correlations among repeated measures within subjects. The global predictive accuracy of the early antipsychotic response model for subsequent acute manic/mixed episode remission in patients with acute bipolar mania, was evaluated using c-statistics in logistic regression analysis (including a treatment term in the model). C-statistic was defined as area under the Receiver Operating Characteristics curve (AUC-ROC) for sensitivity vs. 1-specificity across all possible cut-off values, range from 0.5 (non-informative) to 1 (perfect test discrimination) (Pencina and D’Agostino, 2004; Martinez-Aran et al., 2008). The predictive performance statistics which included positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity values were calculated for each treatment group using various thresholds in the early response criteria (i.e. 50%, 40%, 30%, and 20% improvement in Day 4 SADS-C Psychosis Score from baseline). Since PPV and NPV values are directly proportional to the prevalence of remission (e.g. PPV = 1 and NPV = 0 when all patients remitted in the sample), positive and negative likelihood ratios which overcame this problem were also provided (Altman and Bland, 1994). To test the hypothesis that Day-4 antipsychotic response offers significant incremental performance over early changes in MRS for remission prediction, the differences in average sensitivity (corrected by any potential increase in average ‘‘1-specificity”) between the two prediction models were evaluated, using the new integrated discrimination improvement (IDI) test (Pencina et al., 2008). Number needed to treat (NNT) (Laupacis et al., 1988) was calculated for early antipsychotic response and subsequent manic/mixed episode remission. A series of analysis of covariance (ANCOVA) models were performed to show the mediating relationships between the early effect on psychosis and reductions in MRS score at last visit.
Baseline clinical characteristics were comparable between the treatment groups within each study and in the combined analysis dataset (Table 1). The mean daily dose of ziprasidone (pooled data from both studies) was 122 mg/d. Overall, 43% of ziprasidone and 51% of placebo subjects discontinued study medication during the 3-week studies. In Study 1 (Keck et al., 2003b), 62% of ziprasidonetreated subjects received more than 14 days of treatment, compared with 51% of placebo-treated subjects. The median durations of treatment for the ziprasidone and placebo groups were 20 days and 15 days, respectively. In Study 2 (Potkin et al., 2005), 67% of ziprasidone-treated subjects received more than 14 days of treatment, Table 1 Patient baseline clinical characteristics. Characteristics
Ziprasidone (n = 268)
Placebo (n = 131)
Gender Male, n(%)
139 (52)
70 (53)
Race, n(%) White Black Other
191 (71) 42 (16) 35 (13)
88 (67) 27 (21) 16 (12)
Age Age, year (SD)
39.3 (11.1)
38.2 (11.0)
Manic symptoms, n(%) Manic Mixed
167 (62) 101 (38)
81 (62) 50 (38)
Psychotic symptoms, n(%) Psychoti ca Non-psychotic
101 (38) 167 (62)
51 (39)b 79 (61)
All subjects Mean MRS score (SD) Mean extracted HAM-D score (SD) Mean SADS-C psychosis total score (SD) Mean PANSS psychosis total score (SD)
26.8 (7.5) 7.7 (4.2) 3.4 (2.8) 12.4 (4.3)
26.5 (7.2) 7.1 (3.6) 3.3 (2.6) 12.2 (4.5)
Psychotic subgroup Mean SADS-C psychosis total score (SD) Mean PANSS psychosis total score (SD)
5.3 (2.9) 16.0 (3.9)
5.0 (2.8) 15.6 (3.9)
Includes all randomized subjects who had a baseline and at least one on-drug, postbaseline evaluation with the exception of 8 ziprasidone-treated and 3 placebotreated subjects from one site who were excluded due to sponsor’s inability to verify that data had been collected in a valid manner. a Psychotic = A score of P4 on at least 1 of the following items from the PANSS positive items (delusions, conceptual disorganization, and hallucinatory behavior). b One patient had missing baseline PANSS data.
11
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
compared with 58% of placebo-treated subjects. The median duration of treatment was 21 days for both treatment groups. For the psychotic subgroup in the combined analysis dataset (n = 152), 146 (96%) subjects completed Day-4 visit, 142 (93%) completed Day-7 visit, and 100 (66%) completed the 21-day study period. The time course of antipsychotic response to oral ziprasidone treatment in the psychotic subgroup of patients with bipolar mania (n = 152) is shown in Fig. 1. Mean baseline SADS-C psychosis total score was comparable between the ziprasidone and placebo groups (Table 1). No significant effect of treatment was found for Day 2 improvement in SADS-C psychosis total score (p = 0.486, n = 152, MMRM). However, significant improvement in the mean SADS-C psychosis total score was observed in the entire ziprasidone group (vs. placebo) as early as Day 4 (p = 0.004, n = 399, MMRM, not illustrated), as well as in the subgroup of patients with psychotic symptoms at baseline (p < 0.001, n = 152, MMRM, Fig. 1). The magnitude of the ziprasidone effect on psychosis scores increased by visit, with a significant treatment-by-visit interaction in both psychotic (p = 0.0016, n = 152, MMRM) and all subject populations (p < 0.001, n = 399, MMRM, not illustrated). Early antipsychotic response rate based on the proportion of patients with 50% or greater improvement in SADS-C psychosis total score from baseline to Day 4, was significantly higher with ziprasidone (compared to placebo) in all psychotic subjects (34% vs. 20%, p = 0.02, n = 149, logistic regression, Fig. 2, Left). In addition, ziprasidone treatment was associated with greater SADS-C psychosis total score improvement from baseline to Day 4, among psychotic subjects with low baseline agitation (39% vs. 23%, p = 0.03, n = 106, logistic regression) or high baseline agitation (24% vs. 9%, n = 40, p = 0.24) (Fig. 2). There was no significant interaction effect between baseline agitation level and treatment on Day 4 psychotic response (p = 0.955). NNTs for Day-4 antipsychotic response with ziprasidone treatment (vs. placebo) were 8 for all psychotic subjects, 7 for psychotic subjects with lower baseline agitation, and 7 for psychotic subjects with higher baseline agitation. This finding thus supported the specific early effect of ziprasidone on psychotic symptoms at Day 4. In addition, the ziprasidone effect on mean improvement (Day 4) in psychosis was significant (p = 0.023, ANCOVA), independent of reductions in non-mania related agitation symptoms (as assessed by the SADS-C agitation item) associated with ziprasidone treatment. At Day 21, the acute manic/mixed episode remission rates were significantly higher with ziprasidone (vs. placebo) in all subjects (49% vs. 36%, n = 399, p = 0.008, not illustrated), among all psychotic subjects (46% vs. 30%, p = 0.03, n = 151, Fig. 3, Left), and also
P<0.03* P=0.02 *
P<0.24
n=98
n=51
Psychotic Subjects NNT = 8
N
2 99 51
4
7
14
97 49
95 47
85 34
21 71 29
Improvement
*** *** Ziprasidone Placebo
*** **
***P<0.001; ** P<0.01; Treatment-by-visit interaction test: p=0.0016 (F=3.95, df=5, 607)
Fig. 1. In patients with psychotic manic/mixed episodes (n = 152), ziprasidone compared to placebo yielded increasing improvement in SADS-C psychosis total score over time. ***p < 0.001 or **p < 0.01 (ziprasidone vs. placebo). Treatment by visit interaction was statistically significant (p = 0.0016).
n=29
n=11
High Agitation NNT = 7
* P-value (Treatment-by-Agitation interaction) = 0.955. SADS-C Psychosis Response Rate was defined as >=50% improvement from baseline in SAD-C Psychosis Score Low agitation: PANSS- EC Component <= 14 with a maximum score of 3 (moderate) in two of the 5 items (excitement, hostility, tension, uncooperativeness, and poor impulse)
Fig. 2. SADS-C antipsychotic response rate at Day 4: psychotic subgroup (LOCF). In patients with psychotic manic/mixed episodes, ziprasidone compared to placebo yielded similar, increased Day 4 SADS-C antipsychotic response (P50% SADS-C psychosis score improvement) rates in all psychotic patients (left), and in psychotic patients with lower baseline agitation (middle). Rates in patients with higher baseline agitation (right) were lower for both ziprasidone and placebo, with a comparable mean ziprasidone vs. placebo difference that fell short of significance, perhaps due to limited statistical power. Thus, NNT for Day-4 antipsychotic response with ziprasidone treatment (vs. placebo) was 8 for all psychotic subjects, and 7 for psychotic subjects with lower agitation at baseline, and 7 for subjects with higher agitation at baseline. Treatment by agitation interaction was not statistically significant (p = 0.955). Numbers of subjects with agitation status data (ziprasidone 67 + 29 = 96, placebo 39 + 11 = 50) were slightly less than for psychotic status data, due to missing agitation status data in 3 subjects.
P=0.03
P=0.09
P=0.07
n=100
n=50
Psychotic Subjects NNT = 7
Day
n=67 n=39 Low Agitation NNT = 7
n=69
n=39
Low Agitation NNT = 6
n=29
n=10
High Agitation NNT = 8
NNT for ziprasidone treatment (vs. placebo) was 8 for all subjects, 7 for all psychotic subjects, 6 for psychotic subjects with low agitation symptoms at baseline, and 8 for psychotic subjects with high agitation symptoms at baseline •Remission was defined as MRS score <= 12. P (Treatment-by-Agitation interaction)=0.287
Fig. 3. MRS remission at Day 21: psychotic subgroup (LOCF). In patients with psychotic manic/mixed episodes, ziprasidone compared to placebo yielded similar, increased Day 21 MRS mania remission (MRS 6 12) rates in all psychotic patients (left), and in psychotic patients with lower baseline agitation (middle). Rates in patients with higher baseline agitation (right) were lower for both ziprasidone and placebo, with a comparable mean ziprasidone vs. placebo difference that fell short of significance, perhaps due to limited statistical power. Thus, NNT for day 21 remission of manic symptoms with ziprasidone treatment (vs. placebo) was 7 for all psychotic subjects, 6 for psychotic subjects with lower agitation at baseline, and 8 for subjects with higher agitation at baseline. Treatment by agitation interaction was not statistically significant (p = 0.287). Numbers of subjects with agitation status data (ziprasidone 69 + 29 = 98, placebo 39 + 10 = 49) were slightly less than for psychotic status data, due to missing agitation status data in 3 subjects.
in psychotic subjects with a low level of baseline agitation (51% vs. 33%, p = 0.09, n = 108) or high level of baseline agitation (34% vs.
12
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
Sensitivity (True Positive Fraction)
1.0
0.8
0.6
0.4
AUC ROC = 0.71, 95%CI (0.6, 0.8) P-value for Day 4 SAD-C Psychosis Predictor<0.001
0.2 0.0
0.2
0.4
0.6
0.8
1.0
1 – Specificity (False Positive Fraction) Logistic Model: ziprasidone (vs. placebo) + Early SAD-C Psychosis Response
Fig. 4. SAD-C psychosis response rate (Day 4): Predictor of MRS remission in psychotic subgroup (n = 152). Receiver operating characteristic (ROC) area under the curve (AUC) = 0.71 (95% CI 0.6–0.8).
20%, n = 39, p = 0.07) (Fig. 3). There was no significant interaction effect between baseline agitation level and treatment on Day 21 remission outcome (p = 0.287). Attainment of remission in all subjects, as well as in the psychotic subgroup, demonstrated the consistently significant effect of ziprasidone on remission of manic symptoms at Day 21 (Fig. 3). NNTs for Day 21 remission of manic symptoms with ziprasidone treatment (vs. placebo) were 8 for all subjects, 7 for all psychotic subjects, 6 for psychotic subjects with lower baseline agitation, and 8 for psychotic subjects with higher baseline agitation. Fig. 4 shows the ROC plot obtained by calculating the sensitivity and specificity of every cut-off value from the multivariate logistic model used in the analysis. A global assessment of the performance of using Day-4 antipsychotic response as an early marker for predicting subsequent acute manic-episode remission shows an acceptable discriminatory power (AUC-ROC) in patients with acute bipolar mania and psychotic features independent of ziprasidone or placebo treatment received (p < 0.05, c-statistic = 0.71). The performance statistics (positive and negative predictive values, sensitivity, specificity, and likelihood ratios) for individual cut-off values based on 20%, 30%, 40% and 50% improvement criteria for Day 4 psychosis score were also presented in Table 2. It shows rapid antipsychotic response also predicted subsequent acute manic-episode remission for placebo subjects. Further, mediator analyses showed
that the effect of ziprasidone on early psychosis mediated improvement in the MRS score at the last visit (Day 21 or early termination), accounting for 23% (=0.56 2.02/4.99 100%) of the total effect on reduction in manic symptoms (LS mean change between treatment groups = 5.08). Among all subjects with non-zero baseline SADS-C psychosis score (n = 317), the percentage of improvement in SAD-C psychosis score at Day 4 was a significant predictor for subsequent manicepisode remission outcome (p = 0.009, logistic regression), above and beyond Day 4 MRS change score (p < 0.001). Comparing to the predictive model which included only Day 4 MRS change score as a covariate (Model M1), the addition of Day 4 psychotic response (Model M2: Day 4 MRS change score + Day 4 SAD-C psychosis change score) produced significant improvement in the accuracy of predicting mania MRS-derived remission at Day 21 (p = 0.01, n = 317, integrated discrimination improvement test). The added predictive ability of Day 4 psychotic response in Model M2 was attributable to balancing integrated sensitivity (across all cut-offs) with lowering false-positive error rate in mania remission prediction.
4. Discussion This post hoc analysis of pooled data from two similarly designed, randomized, placebo-controlled, pivotal trials showed that oral ziprasidone monotherapy produced a rapid and significant antipsychotic response after 4 days of treatment, in patients with acute manic/mixed episodes with psychotic features. Furthermore, we found that in subjects with psychotic features, early reduction in psychotic symptoms predicted subsequent (Day 21) acute manic-episode remission, and mediated overall improvement in manic symptoms at endpoint. Even a relatively modest (20%) degree of rapid improvement in psychosis could prove at least somewhat encouraging with respect to the chance of subsequent remission of mania, and such a threshold yielded a reasonable balance of sensitivity (76–80%) and specificity (58–69%) for both the ziprasidone and placebo groups. To our knowledge, this is the first demonstration that rapid improvement in psychotic symptoms associated with bipolar mania can occur after antipsychotic monotherapy treatment, independent of secondary effects on agitation-excitement (non-specific aspects of the acute presentation). These results carry important implications regarding the time course of antipsychotic action in
Table 2 Predictive Accuracy of early antipsychotic response test for acute manic/mixed episode remission among among psychotic subjects (n = 151). Day 4 Response criteria (% improvement in SAD-C psychosis score) (%)
Positive predictive valuea (%)
Negative predictive valueb (%)
Sensitivity (%)
Specificity (%)
Positive likelihood ratioc
Negative likelihood ratiod
(a) Ziprasidone group 20 30 40 50
61 58 62 61
73 66 65 61
76 64 58 44
58 60 69 75
1.81 1.60 1.87 1.76
0.41 0.60 0.61 0.75
(b) Placebo group 20 30 40 50
52 53 36 44
89 80 72 73
80 53 27 27
69 80 80 86
2.58 2.65 1.35 1.93
0.29 0.59 0.91 0.85
a Positive predictive value is the proportion of patients with the positive test results (based on Day 4 psychotic response) who are correctly predicted for subsequent remission (Day 21). b Negative predictive value is the proportion of patients with the negative test results (based on Day 4 psychotic response) who are correctly predicted for no remission (Day 21). c Positive likelihood ratio = sensitivity/(1-specificity), indicating how much the odds of the remission increase when the test is positive (‘‘ruling-in remission”, >1 for increase in the likelihood of remission). d Negative likelihood ratio = (1-sensitivity)/specificity, indicating how much the odds of the remission decrease when the test is negative (‘‘ruling-out remission”, <1 for decrease in the likelihood of remission).
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
the treatment of acute mania, as well as for clinical practice. Specifically, these data suggest that in patients with acute mania with psychotic features, the ability of an intervention to rapidly relieve psychosis may contribute substantively to the ability of such treatment to yield more comprehensive, subsequent improvement. This could merely be related to providing rapid relief in the severity of overall manic/mixed symptoms, or might involve a more complex mechanism specifically related to providing rapid relief of psychotic symptoms. Although no significant treatment effect was observed for the SADS-C psychosis total score at Day 2, significant improvement in this measure of psychosis was observed with ziprasidone (vs. placebo) on Day 4 and was maintained at all subsequent visits. We found the magnitude of the ziprasidone effect (vs. placebo) on psychosis increased by visit, with a significant treatment-byvisit interaction effect, indicating increasing responsiveness to treatment over the course of the 3-week study. These results were further supported by an analysis of the psychosis response rate (P50% improvement from baseline to Day 21 endpoint in SAD-C psychosis items), which significantly favored ziprasidone, in all subjects, in those with psychotic symptoms, and in the subset with lower levels of agitation at baseline. Since significant antimanic effects of ziprasidone were observed as early as Day 2 (Keck et al., 2003a,b; Potkin et al., 2005), it had therefore been assumed that early response to ziprasidone treatment might not be based solely on its antipsychotic features. Our findings confirmed that early antipsychotic effect was independent of non-specific behaviors associated with acute mania (e.g. agitation), based on covariate adjustment analyses, as well as analysis of the subgroup with lower agitation at baseline. Proportions of subjects receiving benzodiazepines (lorazepam, temazepam, diazepam) and their mean doses were similar between the two treatment groups during Days 1–7 and throughout the 3-week study period (Keck et al., 2003b; Potkin et al., 2005). Interestingly, the mediator analyses also showed that early antipsychotic effect was independent of improvement in agitation (as assessed by the SADS-C agitation item – non-mania related). These results are consistent with observations of early antimanic/antipsychotic effects in the treatment of acute mania, as reported in several previous case reports and blinded studies (Delay et al., 1952; Garfinkel et al., 1980; Cookson et al., 1981; Janicak et al., 1988; Brown et al., 1989; Segal et al., 1998; Shen, 1999; Tohen et al., 2000a; Keck et al., 2003b; Potkin et al., 2005). Our findings could help inform clinical therapeutics. In current clinical practice, the duration of hospitalization for acute manic/ mixed episodes is commonly one week or less. Clinicians encountering patients with particularly severe episodes (e.g. those with psychotic features) could potentially gain better understanding from early markers of subsequent manic-episode remission. In this regard, our results suggest that early improvement in psychotic symptoms associated with bipolar mania may be relevant to clinical decision on staying with, or modifying, current treatment, and to the likelihood of remission over a 3 week course of treatment. Limitations of this study include the use of post hoc, pooled analysis to demonstrate early antipsychotic response, although the two pivotal trials involved were similarly designed, thereby facilitating pooling for integrated analyses. Early antipsychotic response in this analysis was derived from the change score in the sum of the 3 SADS-C psychosis items (delusions, hallucinations, and suspiciousness). Other limitations include brevity of the inpatient program (three weeks), and high dropout rate during the double-blind treatment period. The symptom structures of early antipsychotic response in the treatment of acute mania warrant further investigation. In conclusion, our findings demonstrate that oral ziprasidone monotherapy is associated with a rapid antipsychotic response
13
(as early as Day 4) in patients with acute bipolar mania. This early reduction in psychotic symptoms had been found to be robust and maintained over the course of the 3-week study, independent of drug-induced reductions in agitation-excitement symptoms. Further investigation is warranted to better understand the mechanism and clinical implications of the predictive value of rapid antipsychotic response for subsequent remission in patients with bipolar disorder. Contributors All authors contributed to the analysis of data and writing of this manuscript. Analysis and interpretation of data: authors Ketter, Agid, Kapur, Loebel, Siu, and Romano. Drafting of the manuscript: authors Ketter, Agid, Loebel, and Siu. Statistical analysis: Siu. Critical revision of the manuscript for important intellectual content and approval of the final version for publication: authors Ketter, Ofer, Kapur, Loebel, Siu, and Romano. Role of funding source This study was sponsored by Pfizer Inc. The sponsor was involved in all stages from conception, analysis, to review of the manuscript. Data analysis was supported by Pfizer Inc., and interpreted collectively by all of the authors. Conflict of interest statement Terence Ketter, Ofer Agid, and Shitij Kapur did not receive any financial support from Pfizer Inc. in connection with the development of this manuscript. Terence Ketter has received grant/research support from Abbott Labs; AstraZeneca; Bristol-Myers Squibb; Cephalon; Eli Lilly; GlaxoSmithKline; Pfizer; Repligen; and Wyeth; consulting fees from Abbott Labs; AstraZeneca; Bristol-Myers Squibb; Eli Lilly; GlaxoSmithKline; Janssen; Jazz; Novartis; Organon; Solvay; Valeant; Vanda; Wyeth; and XenoPort; and lecture honoraria from Abbott Labs; AstraZeneca; Bristol-Myers Squibb; Eli Lilly; GlaxoSmithKline; Noven; Otsuka; and Pfizer. Ofer Agid has received speaker honorarium from Eli Lilly. Shitij Kapur has received grant support in the last 5 years from AstraZenec, Bristol Meyers Squibb, Eli Lilly, EMD, Darmstadt, Glaxo Smith Kline, Janssen, Neuromolecular Inc., and Pfizer. He is also a consultant or scientific advisor for AstraZeneca, Bristol Meyers Squibb, Eli Lilly, Glaxo Smith Kline, Janssen, Otsuka, Organon, Pfizer, Sanofi-Synthelabo, Servier, and Solvay Wyeth. Antony Loebel, and Romano were employees of Pfizer, Inc. during development and writing of the manuscript. Cynthia Siu was a paid consultant for Pfizer, Inc. in connection with the statistical analysis and development of this manuscript. Acknowledgement This study was previously presented in part as a poster at the APA Annual Meeting, Washington, DC, USA, May 3–8, 2008. References Agid O, Kapur S, Arenovich T, Zipursky RB. Delayed-onset hypothesis of antipsychotic action, a hypothesis tested and rejected. Archives of General Psychiatry 2003;60:1228–35. Agid O, Kapur S, Warrington L, Loebel A, Siu C. Early onset of antipsychotic response in the treatment of acutely agitated patients with psychotic disorders. Schizophrenia Research 2008;102:241–8. Allison DB, Loebel AD, Lombardo I, Romano SJ, Siu CO. Understanding the relationship between baseline BMI and subsequent weight change in antipsychotic trials: effect modification or regression to the mean? Psychiatry Research, in press. Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ 1994:102.
14
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed., text revision (DSM-IV-TR). Washington: American Psychiatric Association; 2000. Beigel A, Murphy DL, Bunney WE. The manic-state rating scale, scale construction, reliability, and validity. Archives of General Psychiatry 1971;25:256–62. Bowden CL, Grunze H, Mullen J, Brecher M, Paulsson B, Jones M, et al. A randomized, double-blind, placebo-controlled efficacy and safety study of quetiapine or lithium as monotherapy for mania in bipolar disorder. Journal of Clinical Psychiatry 2005;66:111–21. Bowden CL, Swann AC, Calabrese JR, Rubenfaer LM, Wozniak PJ, Collins MA, et al. A randomized, placebo-controlled, multicenter study of divalproex sodium extended release in the treatment of acute mania. Journal of Clinical Psychiatry 2006;67:1501–10. Brown D, Silverstone T, Cookson J. Carbamazepine compared to haloperidol in acute mania. International Clinical Psychopharmacology 1989;4:229–38. Cookson J, Silverstone T, Wells B. Double-blind comparative clinical trial of pimozide and chlorpromazine in mania: a test of the dopamine hypothesis. Acta Psychiatrica Scandinavia 1981;64:381–97. Delay J, Deniker P, Harl JM. Traitement des états d’excitation et d’agitation per une méthode medicamentense derivée de l’hibernothérapie. Annals Médicine Psychology (Paris) 1952;110:267–73. Endicott J, Spitzer RL. A diagnostic interview, the schedule for affective disorders and schizophrenia. Archives of General Psychiatry 1978;35:837–44. Garfinkel PE, Stancer HC, Persad E. A comparison of haloperidol, lithium carbonate and their combination in the treatment of mania. Journal of Affective Disorder 1980;2:279–88. Goodwin FR, Jamison K. Manic-depressive illness, bipolar disorders and recurrent depression. 2nd ed. New York: Oxford University Press; 2007. Hirschfeld RM, Keck Jr PE, Kramer M, Karcher K, Canuso C, Eerdekens M, et al. Rapid antimanic effect of risperidone monotherapy, a 3-week multicenter, doubleblind, placebo-controlled trial. American Journal of Psychiatry 2004;161:1057–65. Janicak PG, Bresnahan DB, Sharma R, Davis JM, Comaty JE, Malinick C. A comparison of thiothixene with chlorpromazine in the treatment of mania. Journal of Clinical Psychopharmacology 1988;8:33–7. Kapur S, Arenovich T, Agid O, Zipursky R, Lindborg S, Jones B. Evidence for onset of antipsychotic effects within the first 24 hours of treatment. American Journal of Psychiatry 2005;162:939–46. Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophrenia Bulletin 1987;13:261–76. Keck Jr PE, Marcus R, Tourkodimitris S, Ali M, Liebeskind A, Saha A, et al. A placebocontrolled, double-blind study of the efficacy and safety of aripiprazole in patients with acute bipolar mania. American Journal of Psychiatry 2003a;160:1651–8. Keck Jr PE, Versiani M, Potkin S, West SA, Giller E, Ice K. Ziprasidone in the treatment of acute bipolar mania, a three-week, placebo-controlled, double-blind, randomized trial. American Journal of Psychiatry 2003b;160:741–8. Ketter TA, Wang PW, Nowakowska C, Marsh WK, Bonner JC. Treatment of acute mania in bipolar disorder. In: Ketter TA, editor. Advances in the treatment of bipolar disorder. Washington, DC: American Psychiatric Publishing Inc.; 2005. p. 11–55. Ketter TA, Jones M, Paulsson B. Rates of remission/euthymia with quetiapine monotherapy compared with placebo in patients with acute mania. Journal of Affective Disorder 2007;100(Suppl. 1):S45–53. Khanna S, Vieta E, Lyons B, Grossman F, Eerdekens M, Kramer M. Risperidone in the treatment of acute mania, double-blind, placebo-controlled study. British Journal of Psychiatry 2005;187:229–34. Kleinman L, Lowin A, Flood E, Gandhi G, Edgell E, Revicki D. Costs of bipolar disorder. Pharmacoeconomics 2003;21:601–22. Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful measures of the consequences of treatment. New England Journal of Medicine 1988;318:1728–33.
Lindenmayer JP, Brown E, Baker RW, Schuh LM, Shao L, Tohen M, et al. An excitement subscale of the positive and negative syndrome scale. Schizophrenia Research 2004;68:331–7. Martinez-Aran A, Vieta E, Chengappa KNR, Gershon S, Mullen J, Paulsson B. Reporting outcomes in clinical trials for bipolar disorder: a commentary and suggestions for change. Bipolar Disorder 2008;10:566–79. Merikangas KR, Akiskal HS, Angst J, Greenberg PE, Hirschfeld RM, Petukhova M, et al. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication. Archives of General Psychiatry 2007;64:543–52. Overall JE, Gorham DR. The brief psychiatric rating scale. Psychological Report 1962;10:799–812. Pencina MJ, D’Agostino Sr RB. Overall C as a measure of discrimination in survival analysis: model specific population value and confidence interval estimation. Statistics in Medicine 2004;23:2109–23. Pencina MJ, D’Agostino Sr RB, D’Agostino Jr RB, Vasan SR. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Statistics in Medicine 2008;27:157–72. Potkin SG, Keck Jr PE, Segal S, Ice K, English P. Ziprasidone in acute bipolar mania, a 21-day randomized, double-blind, placebo-controlled replication trial. Journal of Clinical Psychopharmacology 2005;25:301–10. 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. American Journal of Psychiatry 2002;159:1146–54. Sachs G, Sanchez R, Marcus R, Stock E, McQuade R, Carson W, et al. Aripiprazole in the treatment of acute manic or mixed episodes in patients with bipolar I disorder, a 3-week placebo-controlled study. Journal of Psychopharmacology 2006;20:536–46. Segal J, Berk M, Brook S. Risperidone compared with both lithium and haloperidol in mania, a double-blind randomized controlled trial. Clinical Neuropharmacology 1998;21:176–80. Shen WW. A history of antipsychotic drug development. Comparative Psychiatry 1999;40:407–14. Suppes T, Eudicone J, McQuade R, Pikalov 3rd A, Carlson B. Efficacy and safety of aripiprazole in subpopulations with acute manic or mixed episodes of bipolar I disorder. Journal of Affective Disorder 2008;107:145–54. Tohen M, Sanger TM, McElroy SL, Tollefson GD, Chengappa KN, Daniel DG, et al. Olanzapine versus placebo in the treatment of acute mania. Olanzapine HGEH Study Group. American Journal of Psychiatry 1999;156:702–9. Tohen M, Jacobs TG, Feldman PD. Onset of action of antipsychotics in the treatment of mania. Bipolar Disorder 2000a;2:261–8. Tohen M, Jacobs TG, Grundy SL, McElroy SL, Banov MC, Janicak PG, et al. Efficacy of olanzapine in acute bipolar mania, a double-blind, placebo-controlled study: The Olanzapine HGGW Study Group. Archives of General Psychiatry 2000b;57:841–9. Tohen M, Chengappa KN, Suppes T, Zarate Jr CA, Calabrese JR, Bowden CL, et al. Efficacy of olanzapine in combination with valproate or lithium in the treatment of mania in patients partially nonresponsive to valproate or lithium monotherapy. Archives of General Psychiatry 2002;59:62–9. Vieta E, Mullen J, Brecher M, Paulsson B, Jones M. Quetiapine monotherapy for mania associated with bipolar disorder: Combined analysis of two international, double-blind, randomised, placebo-controlled studies. Current Medical Research and Opinions 2005;21:923–34. Yatham LN, Grossman F, Augustyns I, Vieta E, Ravindran A. Mood stabilisers plus risperidone or placebo in the treatment of acute mania: an international, doubleblind, randomised controlled trial. British Journal of Psychiatry 2003;182:141–7. Yatham LN, Paulsson B, Mullen J, Vagero AM. Quetiapine versus placebo in combination with lithium or divalproex for the treatment of bipolar mania. Journal of Clinical Psychopharmacology 2004;24:599–606. Young RC, Biggs JT, Ziegle VE, Meyer DA. A rating scale for mania, reliability, validity and sensitivity. British Journal of Psychiatry 1978;133:429–35.
Contents lists available at ScienceDirect
Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/jpsychires
Rapid antipsychotic response with ziprasidone predicts subsequent acute manic/mixed episode remission Terence A. Ketter a,*, Ofer Agid b, Shitij Kapur c, Antony Loebel d, Cynthia O. Siu e, Steven J. Romano d a
Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA 94305-5723, USA Schizophrenia Program Center for Addiction and Mental Health, Department of Psychiatry, University of Toronto, Toronto, Canada c Section on Schizophrenia, Imaging and Therapeutics, Department of Psychological Medicine and Psychiatry, Institute of Psychiatry, King’s College, London, United Kingdom d Pfizer Inc., New York, NY, USA e Data Power (DP) Inc., Ringoes, NJ, USA b
a r t i c l e
i n f o
Article history: Received 27 April 2009 Received in revised form 10 July 2009 Accepted 20 July 2009
Keywords: Rapid antipsychotic response Bipolar mania Remission
a b s t r a c t Objective: To assess rapid antipsychotic efficacy with oral ziprasidone monotherapy in bipolar acute manic/mixed episodes with psychotic features, and predictive value of rapid antipsychotic response for subsequent acute manic/mixed episode remission. Methods: Pooled analysis of two 3-week, randomized, double-blind, placebo-controlled trials of ziprasidone (40–160 mg/d) in inpatients with bipolar I disorder, and a current manic or mixed episode, with (n = 152) or without (n = 246) psychotic features. Psychosis improvement was evaluated by change in SADS-C psychosis score (sum of delusions, hallucinations, and suspiciousness items). Rapid antipsychotic response (P50% decrease in SADS-C psychosis score by Day 4) and acute manic episode response and remission (endpoint P50% MRS decrease, and a MRS score 6 12, respectively) were analyzed. Results: Significantly greater antipsychotic effects were observed by Day 4 with ziprasidone treatment (vs. placebo) and the magnitude of improvement increased significantly with time, in all subjects, in the subgroup of all psychotic subjects, and psychotic subjects with low baseline agitation (p < 0.05). Rapid antipsychotic response predicted subsequent acute manic episode remission independent of ziprasidone or placebo treatment received (p < 0.001, ROC AUC = 0.71) with significant improvement in accuracy of MRS remission prediction when compared to models using early changes in MRS score alone (p = 0.01). Limitations: Post hoc analysis, use of 3 SADS-C psychosis items to assess psychosis. Conclusions: The predictive value of rapid (Day 4) improvement in psychotic symptoms for subsequent (Day 21) remission of acute manic/mixed symptoms may facilitate enhanced therapeutics, in view of the current practice of brief hospitalization for patients with acute manic/mixed episodes with psychotic features. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Bipolar disorder is a common, serious psychiatric disorder that results in substantial financial and human burdens (Kleinman et al., 2003; Goodwin and Jamison 2007; Merikangas et al., 2007). Acute manic and mixed manic episodes constitute considerable clinical challenges, as by definition they entail hospitalization, psychosis, or severe impairment of psychosocial or occupational function (American Psychiatric Association 2000). Indeed, in aggregate, approximately one third of patients with such episodes in contemporary United States Food and Drug Administration (FDA) registration trials have psychotic features (Tohen et al., 1999,
* Corresponding author. Address: Psychiatry and Behavioral Sciences, 401 Quarry Road, Room 2124, Stanford, CA 94305-5723, USA. Tel.: +1 (650) 723 2515; fax: +1 (650) 723 2507. E-mail address: [email protected] (T.A. Ketter). 0022-3956/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jpsychires.2009.07.006
2000b, 2002; Sachs et al., 2002; Yatham et al., 2003; Hirschfeld et al., 2004; Yatham et al., 2004; Khanna et al., 2005; Vieta et al., 2005; Bowden et al., 2006; Suppes et al., 2008). Phase III studies for regulatory approval for the treatment of acute manic and mixed episodes commonly last 3–4 weeks. However, in contemporary clinical practice in the United States, durations of hospitalization for such episodes are commonly one week or less. Thus, clinical practice could be informed by knowledge regarding the degree to which rapid (within the first few days) improvement predicts subsequent (within the next few weeks) clinical status. A key goal in the pharmacologic treatment of acute bipolar manic and mixed episodes with psychotic features is rapid symptomatic improvement. Medications used for this condition commonly include mood stabilizers and/or first- or second-generation antipsychotics (Ketter et al., 2005). In recent years, five second-generation antipsychotics (olanzapine, risperidone, quetiapine, ziprasidone, and aripiprazole) have been approved by the
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
FDA for the monotherapy treatment of acute manic episodes (Tohen et al., 1999, 2000b; Keck et al., 2003a,b; Hirschfeld et al., 2004; Khanna et al., 2005; Potkin et al., 2005; Vieta et al., 2005; Sachs et al., 2006). With the exception of quetiapine immediate release formulation, these agents have also been approved for the monotherapy of acute mixed episodes. Although still controversial, rapid antipsychotic efficacy has been demonstrated in schizophrenia. However, such rapid treatment effects have not been well established in mania, where the time course of antipsychotic response remains to be established. It used to be assumed that early symptomatic improvement associated with antipsychotic treatments was related to non-specific sedative effects, while true antipsychotic response was delayed. This delayed action hypothesis for the treatment of schizophrenia spectrum disorders with antipsychotics has recently been challenged. A 24-h, double-blind, randomized trial of acutely agitated patients with schizophrenia (Kapur et al., 2005) and a meta-analysis of 42 double-blind active- or placebo-controlled trials (Agid et al., 2003) both found early antipsychotic responses with psychotropic treatments, distinct from drug-induced non-specific effects on behavior (e.g. agitation and excitement). Findings from our previous study further support the early antipsychotic action hypothesis, by demonstrating intramuscular ziprasidone produced rapid (within 4 h) effects on both psychosis and agitation in the treatment of subjects with psychotic disorders (Agid et al., 2008). Early studies also suggested that first-generation antipsychotics may yield rapid antipsychotic and antimanic effects in patients with acute mania. Garfinkel et al. (1980) found significant improvements compared to baseline in mean Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham, 1962) score in bipolar patients with acute mania after one week of treatment with haloperidol. Cookson et al. (1981) found that in bipolar patients with acute mania, both pimozide and chlorpromazine yielded significant decreases from baseline in mean Biegel-Murphy Mania Rating Scale (BMMRS) (Beigel et al., 1971) scores during each of the first 3 days after commencing treatment, and this improvement persisted for the remainder of the two-week study. Janicak et al. (1988) found that manic patients had improved mean BPRS and BMMRS scores with both thiothixene and chlorpromazine after 4 days of administration compared to baseline. Brown et al. (1989) found significant improvement in mean Young Mania Rating Scale (YMRS) (Young et al., 1978) score in manic patients after 3 days of taking haloperidol in a comparative study with carbamazepine. Segal et al. (1998) showed that in acute mania patients, haloperidol and risperidone yielded mean YMRS score reductions of 30% and 24%, respectively, by the end of the first week of treatment. Evidence of early onset of antimanic effects in acute mania has also accumulated for second-generation antipsychotics. Tohen et al. (1999) found olanzapine compared to placebo yielded a significantly greater decrease (33.3% vs. 5.21%) in mean YMRS score from baseline after one week of treatment. In studies by Keck et al. (2003b) and Potkin et al. (2005), ziprasidone-treated patients demonstrated significant improvement in mean Mania Rating Scale (MRS) scores as early as Day 2 ( 5.6 and 3.4 points, respectively) compared with placebo-treated patients ( 2.4 and 1.8 points, respectively) (p < 0.05), that was subsequently maintained over the course of the three week studies. Aripiprazole (n = 123) was found to reduce mean YMRS score by 5.8 points after 4 days of treatment, vs. a reduction of 2.4 points with placebo (n = 120, p < .005) (Keck et al., 2003a). Hirschfeld et al. (2004) presented evidence of a rapid antimanic effect for risperidone monotherapy, showing the mean YMRS score for the risperidone group (n = 124) was reduced by 6.5 points after 3 days of treatment, vs. 4.0 points for the placebo group (n = 119) (p < .001). Bowden et al. (2005) showed statistically significant changes from baseline
9
in mean YMRS score between quetiapine ( 8.03 points, n = 107) and placebo ( 4.89 points, n = 95) groups, beginning at Day 7 (p < 0.01) in bipolar acute mania patients. Tohen et al. (2000a) reviewed published controlled, blinded studies regarding the onset of action of antipsychotic medications in the treatment of acute mania. They found that first-generation (chlorpromazine, haloperidol) as well as second-generation (olanzapine, risperidone) antipsychotics had early onset of antimanic effects appearing after 2–6 days of treatment. Together, these data suggest rapid antimanic actions of antipsychotics in the treatment of acute manic/ mixed symptoms in patients with bipolar disorder. Accordingly, we conducted a post hoc pooled analysis of two similarly designed, placebo-controlled pivotal trials to assess the rapid antipsychotic and subsequent antimanic actions of the second-generation antipsychotic, ziprasidone, in the treatment of bipolar disorder patients with acute manic/mixed episodes, with psychotic features, and the clinical implications – specifically the degree to which rapid antipsychotic response predicted subsequent remission of manic symptoms. 2. Methods 2.1. Study subjects The full analysis dataset consisted of a total of 399 subjects pooled from two similarly designed, randomized, double-blind, placebo-controlled, 3-week trials of oral ziprasidone monotherapy (40–160 mg/d) in hospitalized patients with bipolar I disorder (Keck et al., 2003b; Potkin et al., 2005). At study baseline, all subjects were experiencing a current manic (n = 248) or mixed episode (n = 151), with (n = 152) or without (n = 246) psychotic features. Both studies were conducted in accordance with the principles of the Helsinki Declaration and were in compliance with Good Clinical Practice. The protocol and related informed consent form were approved by the local institutional review boards of the participating sites. Patients or their legal representatives provided written and verbal informed consent before any protocol-related procedures were performed. Additional details on the design of these two studies are provided in the primary study publications (Keck et al., 2003b; Potkin et al., 2005). Subjects were male (n = 209) or female (n = 190) inpatients, aged 18 years or older, with a primary diagnosis of bipolar I disorder and a current manic (296.4) or mixed (296.6) episode, defined according to the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) (American Psychiatric Association 2000). Subjects had a Mania Rating Scale score (MRS) P 14, and scores of P2 on at least 4 MRS items, at both screening and baseline. 2.2. Clinical design and outcome measures Both studies were randomized, double-blind, placebo-controlled trials, comprising up to 10 days of single-blind, placebo screening/washout period and a 21-day treatment period. Subjects meeting all inclusion and exclusion criteria were randomized to receive up to 3 weeks (21 days) of double-blind treatment with ziprasidone or placebo, in a 2:1 ratio (ziprasidone:placebo). During the screening period, subjects discontinued prohibited concomitant psychotropic medications, as necessary. The duration of washout was at least 3–5 half-lives of the prohibited medications, and a minimum of 3 days for all antipsychotics. In both studies, ziprasidone was initiated at 40 mg twice daily with food (80 mg/d), then titrated to 80 mg twice daily with food (160 mg/d) on Day 2, and then flexibly adjusted by up to 40 mg/d in the range of 40–80 mg twice daily with food (80–160 mg/d) from Day 2 onward.
10
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
The 11-item Mania Rating Scale (MRS), derived from the Schedule for Affective Disorders and Schizophrenia-Change Bipolar (SADS-CB, May 2000 version) scale, and the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987) were assessed at screening, baseline (within 12 h prior to the first dose of study drug), and at Days 2 (SADS-CB only), 4 (SADS-CB only), 7, 14, and 21 (or at the time of early discontinuation). The SADS-CB, a diagnostic interview used to rate manic and depressive symptoms, is a modified version of the Schedule for Affective Disorders and Schizophrenia-Change (SADS-C) (Endicott and Spitzer, 1978). Manic-episode remission was defined as achieving a MRS score of 612 (Ketter et al., 2007). SADS-C psychosis total score was derived from the sum of SADS-C psychosis items (delusions, hallucinations, and suspiciousness). Early antipsychotic response rate was defined as the proportion of subjects achieving P50% improvement from baseline in the SADS-C Psychosis Score at Days 2 or 4. Agitation symptoms at Days 2 or 4 were assessed by the SADS-C agitation item (non-mania related). Primary analyses focusing on the MRS change score at Day 21 have previously been published (Keck et al., 2003b; Potkin et al., 2005).
2.4. Psychotic subgroup analyses Subgroups were based on baseline psychotic status and severity level of agitation symptoms. To reduce potential bias due to regression to the mean (RTM) (Allison et al., in press), the psychotic subgroup was defined based on PANSS positive items (delusions P1, conceptual disorganization P2, and hallucinatory behavior P3) and SADS-C psychosis change score was used as analysis endpoint. Specifically, the psychotic subgroup included all subjects with a baseline score of P4 on at least 1 of the following PANSS positive items: delusions P1, conceptual disorganization P2, and hallucinatory behavior P3. Baseline agitation severity levels were dichotomized, with the low agitation subgroup defined as having a PANSS Excitability Component (PANSS-EC) score of 614 and a maximum score of 3 (moderate) in two of the following 5 items (excitement, hostility, tension, uncooperativeness, and poor impulse). The PANSS-EC score was derived from the sum of 5 items (excitement P4, hostility P7, tension G4, uncooperativeness G8, and poor impulse control G14) (Lindenmayer et al., 2004).
3. Results 2.3. Statistical analysis Mixed model repeated measures (MMRM) analysis for change score from baseline was applied to estimate the time course of response, using all available data at each visit in the full analysis dataset (n = 399). The mixed effects models included fixed effects terms for treatment, visit, treatment-by-visit, baseline score, manic/mixed diagnosis, and study. Treatment and treatment-by-visit interaction terms provided test of ziprasidone effect at Day 2, Day 4, Weeks 1, 2, and 3. Random effects terms included center and subject nested within center. Unstructured covariance model was used to account for correlations among repeated measures within subjects. The global predictive accuracy of the early antipsychotic response model for subsequent acute manic/mixed episode remission in patients with acute bipolar mania, was evaluated using c-statistics in logistic regression analysis (including a treatment term in the model). C-statistic was defined as area under the Receiver Operating Characteristics curve (AUC-ROC) for sensitivity vs. 1-specificity across all possible cut-off values, range from 0.5 (non-informative) to 1 (perfect test discrimination) (Pencina and D’Agostino, 2004; Martinez-Aran et al., 2008). The predictive performance statistics which included positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity values were calculated for each treatment group using various thresholds in the early response criteria (i.e. 50%, 40%, 30%, and 20% improvement in Day 4 SADS-C Psychosis Score from baseline). Since PPV and NPV values are directly proportional to the prevalence of remission (e.g. PPV = 1 and NPV = 0 when all patients remitted in the sample), positive and negative likelihood ratios which overcame this problem were also provided (Altman and Bland, 1994). To test the hypothesis that Day-4 antipsychotic response offers significant incremental performance over early changes in MRS for remission prediction, the differences in average sensitivity (corrected by any potential increase in average ‘‘1-specificity”) between the two prediction models were evaluated, using the new integrated discrimination improvement (IDI) test (Pencina et al., 2008). Number needed to treat (NNT) (Laupacis et al., 1988) was calculated for early antipsychotic response and subsequent manic/mixed episode remission. A series of analysis of covariance (ANCOVA) models were performed to show the mediating relationships between the early effect on psychosis and reductions in MRS score at last visit.
Baseline clinical characteristics were comparable between the treatment groups within each study and in the combined analysis dataset (Table 1). The mean daily dose of ziprasidone (pooled data from both studies) was 122 mg/d. Overall, 43% of ziprasidone and 51% of placebo subjects discontinued study medication during the 3-week studies. In Study 1 (Keck et al., 2003b), 62% of ziprasidonetreated subjects received more than 14 days of treatment, compared with 51% of placebo-treated subjects. The median durations of treatment for the ziprasidone and placebo groups were 20 days and 15 days, respectively. In Study 2 (Potkin et al., 2005), 67% of ziprasidone-treated subjects received more than 14 days of treatment, Table 1 Patient baseline clinical characteristics. Characteristics
Ziprasidone (n = 268)
Placebo (n = 131)
Gender Male, n(%)
139 (52)
70 (53)
Race, n(%) White Black Other
191 (71) 42 (16) 35 (13)
88 (67) 27 (21) 16 (12)
Age Age, year (SD)
39.3 (11.1)
38.2 (11.0)
Manic symptoms, n(%) Manic Mixed
167 (62) 101 (38)
81 (62) 50 (38)
Psychotic symptoms, n(%) Psychoti ca Non-psychotic
101 (38) 167 (62)
51 (39)b 79 (61)
All subjects Mean MRS score (SD) Mean extracted HAM-D score (SD) Mean SADS-C psychosis total score (SD) Mean PANSS psychosis total score (SD)
26.8 (7.5) 7.7 (4.2) 3.4 (2.8) 12.4 (4.3)
26.5 (7.2) 7.1 (3.6) 3.3 (2.6) 12.2 (4.5)
Psychotic subgroup Mean SADS-C psychosis total score (SD) Mean PANSS psychosis total score (SD)
5.3 (2.9) 16.0 (3.9)
5.0 (2.8) 15.6 (3.9)
Includes all randomized subjects who had a baseline and at least one on-drug, postbaseline evaluation with the exception of 8 ziprasidone-treated and 3 placebotreated subjects from one site who were excluded due to sponsor’s inability to verify that data had been collected in a valid manner. a Psychotic = A score of P4 on at least 1 of the following items from the PANSS positive items (delusions, conceptual disorganization, and hallucinatory behavior). b One patient had missing baseline PANSS data.
11
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
compared with 58% of placebo-treated subjects. The median duration of treatment was 21 days for both treatment groups. For the psychotic subgroup in the combined analysis dataset (n = 152), 146 (96%) subjects completed Day-4 visit, 142 (93%) completed Day-7 visit, and 100 (66%) completed the 21-day study period. The time course of antipsychotic response to oral ziprasidone treatment in the psychotic subgroup of patients with bipolar mania (n = 152) is shown in Fig. 1. Mean baseline SADS-C psychosis total score was comparable between the ziprasidone and placebo groups (Table 1). No significant effect of treatment was found for Day 2 improvement in SADS-C psychosis total score (p = 0.486, n = 152, MMRM). However, significant improvement in the mean SADS-C psychosis total score was observed in the entire ziprasidone group (vs. placebo) as early as Day 4 (p = 0.004, n = 399, MMRM, not illustrated), as well as in the subgroup of patients with psychotic symptoms at baseline (p < 0.001, n = 152, MMRM, Fig. 1). The magnitude of the ziprasidone effect on psychosis scores increased by visit, with a significant treatment-by-visit interaction in both psychotic (p = 0.0016, n = 152, MMRM) and all subject populations (p < 0.001, n = 399, MMRM, not illustrated). Early antipsychotic response rate based on the proportion of patients with 50% or greater improvement in SADS-C psychosis total score from baseline to Day 4, was significantly higher with ziprasidone (compared to placebo) in all psychotic subjects (34% vs. 20%, p = 0.02, n = 149, logistic regression, Fig. 2, Left). In addition, ziprasidone treatment was associated with greater SADS-C psychosis total score improvement from baseline to Day 4, among psychotic subjects with low baseline agitation (39% vs. 23%, p = 0.03, n = 106, logistic regression) or high baseline agitation (24% vs. 9%, n = 40, p = 0.24) (Fig. 2). There was no significant interaction effect between baseline agitation level and treatment on Day 4 psychotic response (p = 0.955). NNTs for Day-4 antipsychotic response with ziprasidone treatment (vs. placebo) were 8 for all psychotic subjects, 7 for psychotic subjects with lower baseline agitation, and 7 for psychotic subjects with higher baseline agitation. This finding thus supported the specific early effect of ziprasidone on psychotic symptoms at Day 4. In addition, the ziprasidone effect on mean improvement (Day 4) in psychosis was significant (p = 0.023, ANCOVA), independent of reductions in non-mania related agitation symptoms (as assessed by the SADS-C agitation item) associated with ziprasidone treatment. At Day 21, the acute manic/mixed episode remission rates were significantly higher with ziprasidone (vs. placebo) in all subjects (49% vs. 36%, n = 399, p = 0.008, not illustrated), among all psychotic subjects (46% vs. 30%, p = 0.03, n = 151, Fig. 3, Left), and also
P<0.03* P=0.02 *
P<0.24
n=98
n=51
Psychotic Subjects NNT = 8
N
2 99 51
4
7
14
97 49
95 47
85 34
21 71 29
Improvement
*** *** Ziprasidone Placebo
*** **
***P<0.001; ** P<0.01; Treatment-by-visit interaction test: p=0.0016 (F=3.95, df=5, 607)
Fig. 1. In patients with psychotic manic/mixed episodes (n = 152), ziprasidone compared to placebo yielded increasing improvement in SADS-C psychosis total score over time. ***p < 0.001 or **p < 0.01 (ziprasidone vs. placebo). Treatment by visit interaction was statistically significant (p = 0.0016).
n=29
n=11
High Agitation NNT = 7
* P-value (Treatment-by-Agitation interaction) = 0.955. SADS-C Psychosis Response Rate was defined as >=50% improvement from baseline in SAD-C Psychosis Score Low agitation: PANSS- EC Component <= 14 with a maximum score of 3 (moderate) in two of the 5 items (excitement, hostility, tension, uncooperativeness, and poor impulse)
Fig. 2. SADS-C antipsychotic response rate at Day 4: psychotic subgroup (LOCF). In patients with psychotic manic/mixed episodes, ziprasidone compared to placebo yielded similar, increased Day 4 SADS-C antipsychotic response (P50% SADS-C psychosis score improvement) rates in all psychotic patients (left), and in psychotic patients with lower baseline agitation (middle). Rates in patients with higher baseline agitation (right) were lower for both ziprasidone and placebo, with a comparable mean ziprasidone vs. placebo difference that fell short of significance, perhaps due to limited statistical power. Thus, NNT for Day-4 antipsychotic response with ziprasidone treatment (vs. placebo) was 8 for all psychotic subjects, and 7 for psychotic subjects with lower agitation at baseline, and 7 for subjects with higher agitation at baseline. Treatment by agitation interaction was not statistically significant (p = 0.955). Numbers of subjects with agitation status data (ziprasidone 67 + 29 = 96, placebo 39 + 11 = 50) were slightly less than for psychotic status data, due to missing agitation status data in 3 subjects.
P=0.03
P=0.09
P=0.07
n=100
n=50
Psychotic Subjects NNT = 7
Day
n=67 n=39 Low Agitation NNT = 7
n=69
n=39
Low Agitation NNT = 6
n=29
n=10
High Agitation NNT = 8
NNT for ziprasidone treatment (vs. placebo) was 8 for all subjects, 7 for all psychotic subjects, 6 for psychotic subjects with low agitation symptoms at baseline, and 8 for psychotic subjects with high agitation symptoms at baseline •Remission was defined as MRS score <= 12. P (Treatment-by-Agitation interaction)=0.287
Fig. 3. MRS remission at Day 21: psychotic subgroup (LOCF). In patients with psychotic manic/mixed episodes, ziprasidone compared to placebo yielded similar, increased Day 21 MRS mania remission (MRS 6 12) rates in all psychotic patients (left), and in psychotic patients with lower baseline agitation (middle). Rates in patients with higher baseline agitation (right) were lower for both ziprasidone and placebo, with a comparable mean ziprasidone vs. placebo difference that fell short of significance, perhaps due to limited statistical power. Thus, NNT for day 21 remission of manic symptoms with ziprasidone treatment (vs. placebo) was 7 for all psychotic subjects, 6 for psychotic subjects with lower agitation at baseline, and 8 for subjects with higher agitation at baseline. Treatment by agitation interaction was not statistically significant (p = 0.287). Numbers of subjects with agitation status data (ziprasidone 69 + 29 = 98, placebo 39 + 10 = 49) were slightly less than for psychotic status data, due to missing agitation status data in 3 subjects.
in psychotic subjects with a low level of baseline agitation (51% vs. 33%, p = 0.09, n = 108) or high level of baseline agitation (34% vs.
12
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
Sensitivity (True Positive Fraction)
1.0
0.8
0.6
0.4
AUC ROC = 0.71, 95%CI (0.6, 0.8) P-value for Day 4 SAD-C Psychosis Predictor<0.001
0.2 0.0
0.2
0.4
0.6
0.8
1.0
1 – Specificity (False Positive Fraction) Logistic Model: ziprasidone (vs. placebo) + Early SAD-C Psychosis Response
Fig. 4. SAD-C psychosis response rate (Day 4): Predictor of MRS remission in psychotic subgroup (n = 152). Receiver operating characteristic (ROC) area under the curve (AUC) = 0.71 (95% CI 0.6–0.8).
20%, n = 39, p = 0.07) (Fig. 3). There was no significant interaction effect between baseline agitation level and treatment on Day 21 remission outcome (p = 0.287). Attainment of remission in all subjects, as well as in the psychotic subgroup, demonstrated the consistently significant effect of ziprasidone on remission of manic symptoms at Day 21 (Fig. 3). NNTs for Day 21 remission of manic symptoms with ziprasidone treatment (vs. placebo) were 8 for all subjects, 7 for all psychotic subjects, 6 for psychotic subjects with lower baseline agitation, and 8 for psychotic subjects with higher baseline agitation. Fig. 4 shows the ROC plot obtained by calculating the sensitivity and specificity of every cut-off value from the multivariate logistic model used in the analysis. A global assessment of the performance of using Day-4 antipsychotic response as an early marker for predicting subsequent acute manic-episode remission shows an acceptable discriminatory power (AUC-ROC) in patients with acute bipolar mania and psychotic features independent of ziprasidone or placebo treatment received (p < 0.05, c-statistic = 0.71). The performance statistics (positive and negative predictive values, sensitivity, specificity, and likelihood ratios) for individual cut-off values based on 20%, 30%, 40% and 50% improvement criteria for Day 4 psychosis score were also presented in Table 2. It shows rapid antipsychotic response also predicted subsequent acute manic-episode remission for placebo subjects. Further, mediator analyses showed
that the effect of ziprasidone on early psychosis mediated improvement in the MRS score at the last visit (Day 21 or early termination), accounting for 23% (=0.56 2.02/4.99 100%) of the total effect on reduction in manic symptoms (LS mean change between treatment groups = 5.08). Among all subjects with non-zero baseline SADS-C psychosis score (n = 317), the percentage of improvement in SAD-C psychosis score at Day 4 was a significant predictor for subsequent manicepisode remission outcome (p = 0.009, logistic regression), above and beyond Day 4 MRS change score (p < 0.001). Comparing to the predictive model which included only Day 4 MRS change score as a covariate (Model M1), the addition of Day 4 psychotic response (Model M2: Day 4 MRS change score + Day 4 SAD-C psychosis change score) produced significant improvement in the accuracy of predicting mania MRS-derived remission at Day 21 (p = 0.01, n = 317, integrated discrimination improvement test). The added predictive ability of Day 4 psychotic response in Model M2 was attributable to balancing integrated sensitivity (across all cut-offs) with lowering false-positive error rate in mania remission prediction.
4. Discussion This post hoc analysis of pooled data from two similarly designed, randomized, placebo-controlled, pivotal trials showed that oral ziprasidone monotherapy produced a rapid and significant antipsychotic response after 4 days of treatment, in patients with acute manic/mixed episodes with psychotic features. Furthermore, we found that in subjects with psychotic features, early reduction in psychotic symptoms predicted subsequent (Day 21) acute manic-episode remission, and mediated overall improvement in manic symptoms at endpoint. Even a relatively modest (20%) degree of rapid improvement in psychosis could prove at least somewhat encouraging with respect to the chance of subsequent remission of mania, and such a threshold yielded a reasonable balance of sensitivity (76–80%) and specificity (58–69%) for both the ziprasidone and placebo groups. To our knowledge, this is the first demonstration that rapid improvement in psychotic symptoms associated with bipolar mania can occur after antipsychotic monotherapy treatment, independent of secondary effects on agitation-excitement (non-specific aspects of the acute presentation). These results carry important implications regarding the time course of antipsychotic action in
Table 2 Predictive Accuracy of early antipsychotic response test for acute manic/mixed episode remission among among psychotic subjects (n = 151). Day 4 Response criteria (% improvement in SAD-C psychosis score) (%)
Positive predictive valuea (%)
Negative predictive valueb (%)
Sensitivity (%)
Specificity (%)
Positive likelihood ratioc
Negative likelihood ratiod
(a) Ziprasidone group 20 30 40 50
61 58 62 61
73 66 65 61
76 64 58 44
58 60 69 75
1.81 1.60 1.87 1.76
0.41 0.60 0.61 0.75
(b) Placebo group 20 30 40 50
52 53 36 44
89 80 72 73
80 53 27 27
69 80 80 86
2.58 2.65 1.35 1.93
0.29 0.59 0.91 0.85
a Positive predictive value is the proportion of patients with the positive test results (based on Day 4 psychotic response) who are correctly predicted for subsequent remission (Day 21). b Negative predictive value is the proportion of patients with the negative test results (based on Day 4 psychotic response) who are correctly predicted for no remission (Day 21). c Positive likelihood ratio = sensitivity/(1-specificity), indicating how much the odds of the remission increase when the test is positive (‘‘ruling-in remission”, >1 for increase in the likelihood of remission). d Negative likelihood ratio = (1-sensitivity)/specificity, indicating how much the odds of the remission decrease when the test is negative (‘‘ruling-out remission”, <1 for decrease in the likelihood of remission).
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
the treatment of acute mania, as well as for clinical practice. Specifically, these data suggest that in patients with acute mania with psychotic features, the ability of an intervention to rapidly relieve psychosis may contribute substantively to the ability of such treatment to yield more comprehensive, subsequent improvement. This could merely be related to providing rapid relief in the severity of overall manic/mixed symptoms, or might involve a more complex mechanism specifically related to providing rapid relief of psychotic symptoms. Although no significant treatment effect was observed for the SADS-C psychosis total score at Day 2, significant improvement in this measure of psychosis was observed with ziprasidone (vs. placebo) on Day 4 and was maintained at all subsequent visits. We found the magnitude of the ziprasidone effect (vs. placebo) on psychosis increased by visit, with a significant treatment-byvisit interaction effect, indicating increasing responsiveness to treatment over the course of the 3-week study. These results were further supported by an analysis of the psychosis response rate (P50% improvement from baseline to Day 21 endpoint in SAD-C psychosis items), which significantly favored ziprasidone, in all subjects, in those with psychotic symptoms, and in the subset with lower levels of agitation at baseline. Since significant antimanic effects of ziprasidone were observed as early as Day 2 (Keck et al., 2003a,b; Potkin et al., 2005), it had therefore been assumed that early response to ziprasidone treatment might not be based solely on its antipsychotic features. Our findings confirmed that early antipsychotic effect was independent of non-specific behaviors associated with acute mania (e.g. agitation), based on covariate adjustment analyses, as well as analysis of the subgroup with lower agitation at baseline. Proportions of subjects receiving benzodiazepines (lorazepam, temazepam, diazepam) and their mean doses were similar between the two treatment groups during Days 1–7 and throughout the 3-week study period (Keck et al., 2003b; Potkin et al., 2005). Interestingly, the mediator analyses also showed that early antipsychotic effect was independent of improvement in agitation (as assessed by the SADS-C agitation item – non-mania related). These results are consistent with observations of early antimanic/antipsychotic effects in the treatment of acute mania, as reported in several previous case reports and blinded studies (Delay et al., 1952; Garfinkel et al., 1980; Cookson et al., 1981; Janicak et al., 1988; Brown et al., 1989; Segal et al., 1998; Shen, 1999; Tohen et al., 2000a; Keck et al., 2003b; Potkin et al., 2005). Our findings could help inform clinical therapeutics. In current clinical practice, the duration of hospitalization for acute manic/ mixed episodes is commonly one week or less. Clinicians encountering patients with particularly severe episodes (e.g. those with psychotic features) could potentially gain better understanding from early markers of subsequent manic-episode remission. In this regard, our results suggest that early improvement in psychotic symptoms associated with bipolar mania may be relevant to clinical decision on staying with, or modifying, current treatment, and to the likelihood of remission over a 3 week course of treatment. Limitations of this study include the use of post hoc, pooled analysis to demonstrate early antipsychotic response, although the two pivotal trials involved were similarly designed, thereby facilitating pooling for integrated analyses. Early antipsychotic response in this analysis was derived from the change score in the sum of the 3 SADS-C psychosis items (delusions, hallucinations, and suspiciousness). Other limitations include brevity of the inpatient program (three weeks), and high dropout rate during the double-blind treatment period. The symptom structures of early antipsychotic response in the treatment of acute mania warrant further investigation. In conclusion, our findings demonstrate that oral ziprasidone monotherapy is associated with a rapid antipsychotic response
13
(as early as Day 4) in patients with acute bipolar mania. This early reduction in psychotic symptoms had been found to be robust and maintained over the course of the 3-week study, independent of drug-induced reductions in agitation-excitement symptoms. Further investigation is warranted to better understand the mechanism and clinical implications of the predictive value of rapid antipsychotic response for subsequent remission in patients with bipolar disorder. Contributors All authors contributed to the analysis of data and writing of this manuscript. Analysis and interpretation of data: authors Ketter, Agid, Kapur, Loebel, Siu, and Romano. Drafting of the manuscript: authors Ketter, Agid, Loebel, and Siu. Statistical analysis: Siu. Critical revision of the manuscript for important intellectual content and approval of the final version for publication: authors Ketter, Ofer, Kapur, Loebel, Siu, and Romano. Role of funding source This study was sponsored by Pfizer Inc. The sponsor was involved in all stages from conception, analysis, to review of the manuscript. Data analysis was supported by Pfizer Inc., and interpreted collectively by all of the authors. Conflict of interest statement Terence Ketter, Ofer Agid, and Shitij Kapur did not receive any financial support from Pfizer Inc. in connection with the development of this manuscript. Terence Ketter has received grant/research support from Abbott Labs; AstraZeneca; Bristol-Myers Squibb; Cephalon; Eli Lilly; GlaxoSmithKline; Pfizer; Repligen; and Wyeth; consulting fees from Abbott Labs; AstraZeneca; Bristol-Myers Squibb; Eli Lilly; GlaxoSmithKline; Janssen; Jazz; Novartis; Organon; Solvay; Valeant; Vanda; Wyeth; and XenoPort; and lecture honoraria from Abbott Labs; AstraZeneca; Bristol-Myers Squibb; Eli Lilly; GlaxoSmithKline; Noven; Otsuka; and Pfizer. Ofer Agid has received speaker honorarium from Eli Lilly. Shitij Kapur has received grant support in the last 5 years from AstraZenec, Bristol Meyers Squibb, Eli Lilly, EMD, Darmstadt, Glaxo Smith Kline, Janssen, Neuromolecular Inc., and Pfizer. He is also a consultant or scientific advisor for AstraZeneca, Bristol Meyers Squibb, Eli Lilly, Glaxo Smith Kline, Janssen, Otsuka, Organon, Pfizer, Sanofi-Synthelabo, Servier, and Solvay Wyeth. Antony Loebel, and Romano were employees of Pfizer, Inc. during development and writing of the manuscript. Cynthia Siu was a paid consultant for Pfizer, Inc. in connection with the statistical analysis and development of this manuscript. Acknowledgement This study was previously presented in part as a poster at the APA Annual Meeting, Washington, DC, USA, May 3–8, 2008. References Agid O, Kapur S, Arenovich T, Zipursky RB. Delayed-onset hypothesis of antipsychotic action, a hypothesis tested and rejected. Archives of General Psychiatry 2003;60:1228–35. Agid O, Kapur S, Warrington L, Loebel A, Siu C. Early onset of antipsychotic response in the treatment of acutely agitated patients with psychotic disorders. Schizophrenia Research 2008;102:241–8. Allison DB, Loebel AD, Lombardo I, Romano SJ, Siu CO. Understanding the relationship between baseline BMI and subsequent weight change in antipsychotic trials: effect modification or regression to the mean? Psychiatry Research, in press. Altman DG, Bland JM. Diagnostic tests 2: predictive values. BMJ 1994:102.
14
T.A. Ketter et al. / Journal of Psychiatric Research 44 (2010) 8–14
American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed., text revision (DSM-IV-TR). Washington: American Psychiatric Association; 2000. Beigel A, Murphy DL, Bunney WE. The manic-state rating scale, scale construction, reliability, and validity. Archives of General Psychiatry 1971;25:256–62. Bowden CL, Grunze H, Mullen J, Brecher M, Paulsson B, Jones M, et al. A randomized, double-blind, placebo-controlled efficacy and safety study of quetiapine or lithium as monotherapy for mania in bipolar disorder. Journal of Clinical Psychiatry 2005;66:111–21. Bowden CL, Swann AC, Calabrese JR, Rubenfaer LM, Wozniak PJ, Collins MA, et al. A randomized, placebo-controlled, multicenter study of divalproex sodium extended release in the treatment of acute mania. Journal of Clinical Psychiatry 2006;67:1501–10. Brown D, Silverstone T, Cookson J. Carbamazepine compared to haloperidol in acute mania. International Clinical Psychopharmacology 1989;4:229–38. Cookson J, Silverstone T, Wells B. Double-blind comparative clinical trial of pimozide and chlorpromazine in mania: a test of the dopamine hypothesis. Acta Psychiatrica Scandinavia 1981;64:381–97. Delay J, Deniker P, Harl JM. Traitement des états d’excitation et d’agitation per une méthode medicamentense derivée de l’hibernothérapie. Annals Médicine Psychology (Paris) 1952;110:267–73. Endicott J, Spitzer RL. A diagnostic interview, the schedule for affective disorders and schizophrenia. Archives of General Psychiatry 1978;35:837–44. Garfinkel PE, Stancer HC, Persad E. A comparison of haloperidol, lithium carbonate and their combination in the treatment of mania. Journal of Affective Disorder 1980;2:279–88. Goodwin FR, Jamison K. Manic-depressive illness, bipolar disorders and recurrent depression. 2nd ed. New York: Oxford University Press; 2007. Hirschfeld RM, Keck Jr PE, Kramer M, Karcher K, Canuso C, Eerdekens M, et al. Rapid antimanic effect of risperidone monotherapy, a 3-week multicenter, doubleblind, placebo-controlled trial. American Journal of Psychiatry 2004;161:1057–65. Janicak PG, Bresnahan DB, Sharma R, Davis JM, Comaty JE, Malinick C. A comparison of thiothixene with chlorpromazine in the treatment of mania. Journal of Clinical Psychopharmacology 1988;8:33–7. Kapur S, Arenovich T, Agid O, Zipursky R, Lindborg S, Jones B. Evidence for onset of antipsychotic effects within the first 24 hours of treatment. American Journal of Psychiatry 2005;162:939–46. Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophrenia Bulletin 1987;13:261–76. Keck Jr PE, Marcus R, Tourkodimitris S, Ali M, Liebeskind A, Saha A, et al. A placebocontrolled, double-blind study of the efficacy and safety of aripiprazole in patients with acute bipolar mania. American Journal of Psychiatry 2003a;160:1651–8. Keck Jr PE, Versiani M, Potkin S, West SA, Giller E, Ice K. Ziprasidone in the treatment of acute bipolar mania, a three-week, placebo-controlled, double-blind, randomized trial. American Journal of Psychiatry 2003b;160:741–8. Ketter TA, Wang PW, Nowakowska C, Marsh WK, Bonner JC. Treatment of acute mania in bipolar disorder. In: Ketter TA, editor. Advances in the treatment of bipolar disorder. Washington, DC: American Psychiatric Publishing Inc.; 2005. p. 11–55. Ketter TA, Jones M, Paulsson B. Rates of remission/euthymia with quetiapine monotherapy compared with placebo in patients with acute mania. Journal of Affective Disorder 2007;100(Suppl. 1):S45–53. Khanna S, Vieta E, Lyons B, Grossman F, Eerdekens M, Kramer M. Risperidone in the treatment of acute mania, double-blind, placebo-controlled study. British Journal of Psychiatry 2005;187:229–34. Kleinman L, Lowin A, Flood E, Gandhi G, Edgell E, Revicki D. Costs of bipolar disorder. Pharmacoeconomics 2003;21:601–22. Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful measures of the consequences of treatment. New England Journal of Medicine 1988;318:1728–33.
Lindenmayer JP, Brown E, Baker RW, Schuh LM, Shao L, Tohen M, et al. An excitement subscale of the positive and negative syndrome scale. Schizophrenia Research 2004;68:331–7. Martinez-Aran A, Vieta E, Chengappa KNR, Gershon S, Mullen J, Paulsson B. Reporting outcomes in clinical trials for bipolar disorder: a commentary and suggestions for change. Bipolar Disorder 2008;10:566–79. Merikangas KR, Akiskal HS, Angst J, Greenberg PE, Hirschfeld RM, Petukhova M, et al. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication. Archives of General Psychiatry 2007;64:543–52. Overall JE, Gorham DR. The brief psychiatric rating scale. Psychological Report 1962;10:799–812. Pencina MJ, D’Agostino Sr RB. Overall C as a measure of discrimination in survival analysis: model specific population value and confidence interval estimation. Statistics in Medicine 2004;23:2109–23. Pencina MJ, D’Agostino Sr RB, D’Agostino Jr RB, Vasan SR. Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Statistics in Medicine 2008;27:157–72. Potkin SG, Keck Jr PE, Segal S, Ice K, English P. Ziprasidone in acute bipolar mania, a 21-day randomized, double-blind, placebo-controlled replication trial. Journal of Clinical Psychopharmacology 2005;25:301–10. 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. American Journal of Psychiatry 2002;159:1146–54. Sachs G, Sanchez R, Marcus R, Stock E, McQuade R, Carson W, et al. Aripiprazole in the treatment of acute manic or mixed episodes in patients with bipolar I disorder, a 3-week placebo-controlled study. Journal of Psychopharmacology 2006;20:536–46. Segal J, Berk M, Brook S. Risperidone compared with both lithium and haloperidol in mania, a double-blind randomized controlled trial. Clinical Neuropharmacology 1998;21:176–80. Shen WW. A history of antipsychotic drug development. Comparative Psychiatry 1999;40:407–14. Suppes T, Eudicone J, McQuade R, Pikalov 3rd A, Carlson B. Efficacy and safety of aripiprazole in subpopulations with acute manic or mixed episodes of bipolar I disorder. Journal of Affective Disorder 2008;107:145–54. Tohen M, Sanger TM, McElroy SL, Tollefson GD, Chengappa KN, Daniel DG, et al. Olanzapine versus placebo in the treatment of acute mania. Olanzapine HGEH Study Group. American Journal of Psychiatry 1999;156:702–9. Tohen M, Jacobs TG, Feldman PD. Onset of action of antipsychotics in the treatment of mania. Bipolar Disorder 2000a;2:261–8. Tohen M, Jacobs TG, Grundy SL, McElroy SL, Banov MC, Janicak PG, et al. Efficacy of olanzapine in acute bipolar mania, a double-blind, placebo-controlled study: The Olanzapine HGGW Study Group. Archives of General Psychiatry 2000b;57:841–9. Tohen M, Chengappa KN, Suppes T, Zarate Jr CA, Calabrese JR, Bowden CL, et al. Efficacy of olanzapine in combination with valproate or lithium in the treatment of mania in patients partially nonresponsive to valproate or lithium monotherapy. Archives of General Psychiatry 2002;59:62–9. Vieta E, Mullen J, Brecher M, Paulsson B, Jones M. Quetiapine monotherapy for mania associated with bipolar disorder: Combined analysis of two international, double-blind, randomised, placebo-controlled studies. Current Medical Research and Opinions 2005;21:923–34. Yatham LN, Grossman F, Augustyns I, Vieta E, Ravindran A. Mood stabilisers plus risperidone or placebo in the treatment of acute mania: an international, doubleblind, randomised controlled trial. British Journal of Psychiatry 2003;182:141–7. Yatham LN, Paulsson B, Mullen J, Vagero AM. Quetiapine versus placebo in combination with lithium or divalproex for the treatment of bipolar mania. Journal of Clinical Psychopharmacology 2004;24:599–606. Young RC, Biggs JT, Ziegle VE, Meyer DA. A rating scale for mania, reliability, validity and sensitivity. British Journal of Psychiatry 1978;133:429–35.