Improvement in cognitive function following a switch to ziprasidone from conventional antipsychotics, olanzapine, or risperidone in outpatients with schizophrenia

Schizophrenia Research 66 (2004) 101 – 113 www.elsevier.com/locate/schres

Improvement in cognitive function following a switch to ziprasidone from conventional antipsychotics, olanzapine, or risperidone in outpatients with schizophrenia Philip D. Harvey a,*, Herbert Meltzer b, George M. Simpson c, Steven G. Potkin d, Antony Loebel e, Cynthia Siu e, Steven J. Romano e b

a Mt. Sinai School of Medicine, New York University, 1425 Madison Avenue, Fourth Floor, New York, NY 10029, USA Vanderbilt Medical Center, Psychiatric Hospital at Vanderbilt, 1601 23rd Avenue South, Suite 306, Nashville, TN 37212-8645, USA c Department of Psychiatry, LAC+USC Medical Center, University of Southern California, IRD RM 204, 2020 Zonal Avenue, Los Angeles, CA, USA d Neuropsychiatric Center, University of California, Irvine, 101 City Drive South, Building 3, Irvine, CA, 92868-3298, USA e Pfizer Pharmaceutical Group, Pfizer Inc., 235 East 42nd Street, New York, NY 10017, USA

Received 17 April 2003; received in revised form 21 July 2003; accepted 31 July 2003

Abstract Objective: To assess changes in cognitive function in stable outpatients with schizophrenia switched to ziprasidone from conventional antipsychotics (n = 108), olanzapine (n = 104), or risperidone (n = 58) because of suboptimal efficacy or poor tolerability. Methods: In three separate 6-week trials, patients received ziprasidone 40 mg b.i.d. for 2 days, followed by 20 – 80 mg b.i.d. for the next 40 days. Before switching, and at endpoint, patients were evaluated with tests of working and secondary verbal memory, vigilance, visuomotor speed, verbal fluency, and executive functioning. Principal components factor analysis was performed to test for clustering of cognitive variables. Results: Significant improvements were seen at endpoint in secondary verbal memory (in all three groups), vigilance (in patients switched from conventional antipsychotics or risperidone), executive function (in patients switched from conventional antipsychotics or risperidone), and verbal fluency. Factor analysis on baseline scores suggested reduction of the cognitive variables to three factors: verbal skills, attention and short-term memory, and executive functioning. Analysis of z-transformed mean change in factor scores showed significant improvement in verbal skills and global score following the switch from conventional antipsychotics, olanzapine, or risperidone. Conclusions: Patients requiring a change in antipsychotic therapy may exhibit cognitive improvement following a switch to ziprasidone. D 2003 Elsevier B.V. All rights reserved. Keywords: Schizophrenia; Cognition; Ziprasidone; Atypical antipsychotics; Executive functioning; Vigilance; Attention; Verbal fluency

1. Introduction

* Corresponding author. Tel.: +1-212-659-8713; fax: +1-212860-3945. E-mail address: [email protected] (P.D. Harvey). 0920-9964/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.schres.2003.07.009

Cognitive impairment is a cardinal feature of schizophrenia, and patients with this disorder exhibit deficits in nearly all ability domains measured by standard clinical tests (Heinrichs and Zakzanis,

102

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

1998; Keefe et al., 1999). The cognitive deficits observed in patients with schizophrenia have an early onset, and antecedents of the impairments seen in adulthood are often present in children who ultimately develop the disease (Cornblatt et al., 1999). In most patients, cognitive dysfunction is apparent in the first episode of psychosis (Bilder et al., 2000), with the level of severity greater than that seen in first-episode patients with affective psychosis (Mojtabai et al., 2000) and similar in profile and severity to that seen in more chronically ill patients with schizophrenia (Saykin et al., 1994). Appreciation of the impact of cognitive dysfunction on functional outcome in patients with schizophrenia has increased markedly in recent years. An accumulating body of evidence has shown correlation between performance on cognitive tests and measurements of functional outcome (Green, 1996; Green et al., 2000). A recent review identified secondary verbal memory, immediate memory, vigilance, and executive function as most closely associated with functional outcomes (Green et al., 2000). The failure of conventional antipsychotic treatments to improve the functional outcome of patients with schizophrenia (Hegarty et al., 1994) may be due to lack of efficacy in the cognitive features of schizophrenia (Blyler and Gold, 2000). Moreover, the anticholinergic drugs used to treat extrapyramidal symptoms (EPS) resulting from conventional antipsychotic therapy can impair cognition, in particular memory (Spohn and Strauss, 1989; Strauss et al., 1990). One important feature of the newer, ‘‘atypical’’ antipsychotic medications is their apparent efficacy in improving cognitive features of the illness. Clozapine (Hagger et al., 1993; Potkin et al., 2001), risperidone (Green et al., 1997), and olanzapine (Purdon, 1999) have all been reported to exert beneficial effects on cognition compared with conventional antipsychotic treatment (see Meltzer and McGurk, 1999; Harvey and Keefe, 2001 for a review). When the effects of newer medications were compared with conventional treatment by meta-analysis, the overall effects of newer antipsychotics were found to confer a statistically significant greater improvement (Keefe et al., 1999). All of the atypical antipsychotic drugs previously shown to improve cognition, as well as ziprasidone (the focus of this study), are more potent serotonin (5-

HT)2A antagonists than they are dopamine (DA) D2 receptor blockers (Meltzer et al., 1989; Meltzer, 1999). This profile has been found to contribute to their ability to produce large increases in the release of DA in the prefrontal cortex, which may underlie an ability to improve cognition, as decreased availability of DA in the prefrontal cortex is believed to be the basis for cognitive impairment in schizophrenia (Weinberger and Gallhofer, 1997; Davis et al., 1991; Meltzer and McGurk, 1999). Ziprasidone also has a high affinity for 5-HT1A (where it acts as a partial agonist), 5-HT1D, and 5-HT2C receptors. 5-HT1A agonism also contributes to the increase in ziprasidone-induced DA release in the prefrontal cortex (Rollema et al., 1997; Ichikawam et al., 2001). The 5-HT1A agonism of ziprasidone is a key to its inhibitory effect on the firing of serotonergic neurons in the midbrain raphe (Sprouse et al., 1999). Ziprasidone has only modest a1-adrenergic and histaminergic activity, has negligible activity at muscarinic cholinergic receptors (Seeger et al., 1995; Schmidt et al., 2001; Roth et al., in press), and is a moderately potent blocker of both serotonin and norepinephrine reuptake into synaptic terminals (Seeger et al., 1995; Schmidt et al., 2001). All of the atypical agents that are 5HT2A/D2 antagonists also increase acetylcholine release in the prefrontal cortex, which may be a second factor in their ability to improve cognition in schizophrenia (Ichikawa et al., 2002). Thus, ziprasidone has all of the characteristics that would lead to the expectation of cognitive enhancement similar to other novel antipsychotic medications. Clinical trials have demonstrated ziprasidone’s efficacy in the treatment of patients with acute exacerbation of schizophrenia or schizoaffective disorder (Daniel et al., 1999; Keck et al., 1998; Goff et al., 1998; Hirsch et al., 2002; Arato et al., 2002). A favorable tolerability profile, including low incidence of EPS (Daniel et al., 1999; Keck et al., 1998; Goff et al., 1998; Hirsch et al., 2002) and minimal weight gain (Daniel et al., 1999; Keck et al., 1998; Arato et al., 2002), has also been established. In addition to efficacy against positive and negative symptoms, improvement in depressive symptoms has been reported (Daniel et al., 1999; Goff et al., 1998). Three identically designed 6-week prospective studies were conducted to determine the effects of switching stable psychiatric outpatients with schizo-

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

phrenia or schizoaffective disorder from conventional antipsychotics, olanzapine, or risperidone to ziprasidone. Data from these three studies on ziprasidone’s efficacy against positive and negative psychotic symptoms have been reported elsewhere (Weiden et al., 2003). Here we report results of the assessments of cognitive function performed in those studies, including learning and memory, attention, vigilance, executive function, and verbal fluency. Since ziprasidone improved positive and negative symptoms in these studies, even when patients were previously treated with newer antipsychotic medications, we were interested in whether cognitive improvement would occur, whether it would be consistent across medications, and whether the changes seen were related to or independent of the previously reported changes in clinical symptoms.

2. Patients and methods 2.1. Patient eligibility and screenings At screening, patients underwent medical and psychiatric examinations, and were considered eligible if their primary diagnosis was schizophrenia (any subtype) or schizoaffective disorder (any subtype), as defined in the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). Participants were outpatients aged 18 –55 years. All had at least an eighthgrade reading level (as determined by performance on a word-recognition reading test) and provided written, informed consent. They were required to be taking oral doses of antipsychotic medication at prescribed doses within 25% of the recommended daily dose for a minimum of 3 months, with no history of treatment resistance. Patients were also required to have normal laboratory findings and to have a negative urine toxicology screen for illicit drugs. Major exclusion criteria were the presence of other current DSM-IV Axis-I disorders not in remission. Patients were also excluded if they had more than moderate depressive symptoms (Montgomery – Asberg Depression Rating Scale score >16) or met the full DSM-IV criteria for major depression; had a history of psychoactive substance abuse or dependence not in current remission; or had received therapy with psychoactive agents other than antipsychotic medications or prn

103

benzodiazepines within the 4 weeks immediately preceding the study. Patients were recruited at multiple treatment sites in the United States. Patients who were receiving treatment with conventional antipsychotic medications, risperidone, or olanzapine were eligible. Patients were recruited into the study if their primary psychiatric clinician determined that they would benefit from a change in their medications, either because of suboptimal efficacy or because of side effects. Either or both of these reasons could apply and the reason for the switch was not recorded systematically. 2.2. Treatment Two hundred and seventy stable outpatients with schizophrenia or schizoaffective disorder were switched to ziprasidone from conventional antipsychotics (n = 108), olanzapine (n = 104), or risperidone (n = 58) because of suboptimal efficacy or tolerability, as determined by the investigator at each of the research sites. After screening and baseline assessments, patients were placed on oral ziprasidone (40 mg b.i.d. for 2 days, followed by flexible dosing between 20 and 80 mg b.i.d. for the next 40 days), and their previous antipsychotic was discontinued or tapered within 7 days. Anticholinergics and benzodiazepines were administered as needed but were prohibited in the 12 h before cognitive testing. 2.3. Cognitive assessments Cognitive function was assessed at baseline and at week 6 (or at early discontinuation). Baseline assessment was performed while patients were receiving their previous medication. Endpoint assessment was performed when patients had been receiving ziprasidone for at least 2 weeks. All patients were rated weekly with the Positive and Negative Syndrome Scale (PANSS) (Kay, 1991). The cognitive assessment battery evaluated cognitive functioning in five general domains: learning and memory, attention and vigilance, visuomotor speed, executive function, and verbal fluency. 2.3.1. Learning and memory The Rey Auditory Verbal Learning Test (RAVLT) (Spreen and Strauss, 1998) was used to assess learn-

104

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

ing and memory. In this test, which is in common clinical use, the patient is presented with a 15-item list on five separate learning trials, and is instructed to recall as many of the words as possible after each of the trials, in any order. After five presentations of the first list (List A), a new list (List B) is presented once, with immediate recall of that list required. A new recall of List A is then required (Short-Delay Free Recall), followed by another recall of List A after a 20-min delay (Long-Delay Free Recall). The critical dependent measures are List A Trials 1 to 5 Recall (reflecting cumulative learning with exposure and practice), Long-Delay Recall Score, and recognition discriminability. In addition, a computerized Spatial Working Memory test was administered (Meltzer and McGurk, 1999). In this test, patients are shown a dot in one of eight locations described by a semicircular array. The stimulus is presented for 500 ms. After presentation of the stimulus, a delay of either 5 or 15 s ensues. After the delay, patients are asked to recall the location of the stimulus by pointing to the location of a target dot in the eight-location array. 2.3.2. Attention and vigilance This domain was examined with two measures, the Continuous Performance Test, Identical Pairs version (CPT-IP) (Cornblatt et al., 1989) and the Digit Span Distraction Test (DSDT) (Oltmanns and Neale, 1975). In the CPT-IP, a commonly used test of vigilance, patients are asked to press a computer response key whenever the same four-digit target stimulus occurs twice in a row. Four hundred and fifty stimuli are presented, with a 50-ms stimulus duration, at a 1000ms interstimulus interval. There are 90 target stimuli presented and 360 nontarget stimuli. The dependent variable for this test is the signal detection index, d V, which is the response sensitivity for discrimination of target and nontarget stimuli. The DSDT, another commonly used test of memory span and resistance to distractibility, has two conditions: nondistraction and distraction. In the nondistraction condition, six digits per trial are read in a tape-recorded female voice at a one-digit per 2-s rate. In the distraction condition, between the presentation of each of the five target digits per trial, a male voice reads four irrelevant digits. Patients are instructed to attend to the female voice, recall the digits in the order presented, and ignore anything that is presented in a male voice.

There are seven trials in each condition, and the two conditions had been matched for difficulty in normal subjects, indicating that poorer performance in distraction trials is abnormal. 2.3.3. Visuomotor speed This domain was examined with the Trailmaking Test, parts A and B (Spreen and Strauss, 1998). In part A, patients are instructed to connect, in ascending order and as quickly as possible, a series of circles containing numbers. In part B, where the circles contain numbers and letters, patients are instructed to alternate between numbers and letters. The dependent variable in this test is the time required to complete each condition. Since part B requires alternation between competing demands, we also considered this test to be an index of executive functioning. 2.3.4. Verbal fluency Two separate verbal fluency tests were administered (Spreen and Strauss, 1998). In the ‘‘category’’ fluency condition, patients are asked to produce the names of as many different animals as possible within a 1-min period, followed by as many fruits as possible and then as many vegetables as possible. In the ‘‘letter’’ fluency condition, patients are asked to produce as many words as possible that start with three different letters, F, A, and S, taking 1 min per letter and excluding proper names. Critical dependent variables are total scores for letter and category fluency. 2.3.5. Executive functioning This domain was examined with the Wisconsin Card Sorting Test (WCST) (Heaton et al., 1993). This commonly used test of executive functioning measures cognitive flexibility, maintenance of a cognitive set, and working memory. In this test, patients match a series of 128 stimulus cards presented by a computer to a set of four target cards. Dimensions of similarity are similar color, similar form (i.e., shape), and similar number of stimuli on the target cards and individual item cards. Patients are provided with feedback on an item-by-item basis after they sort each of the item cards. After they identify one of the correct dimensions, referred to as ‘‘Categories,’’ 10 correct responses are required before the correct category is shifted to the next one by the examiner. In these

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

analyses, the critical dependent variables are the number of categories completed and the total number of errors made. 2.4. Statistical analysis Inferential analyses on the cognitive variables were based on change from baseline to endpoint (week 6 or early discontinuation) for all patients. A paired t-test with mean and 95% confidence interval (CI) was applied to each cognitive variable in the change analyses. Additionally, we applied principal components analysis to test for clustering of the cognitive variables. Three factors were then extracted from the matrix of correlations among the baseline scores of these cognitive tests using a varimax (orthogonal) rotation. The adoption of three-factor structure was based on the rule of eigenvalues greater than 1 for each interpretable factor. We also performed an oblique rotation, and the results were consistent with those from varimax rotation. Each score was z-transformed based on baseline mean and standard deviation. Averaging z-scores on each of the contributing variables was used to develop a global scale and the factor scores. Means and 95% CIs for z-transformed mean change in factor scores were calculated. The factor scores were presented so that higher scores reflected better performance. To investigate the possibility that improved cognitive function might be associated with improvements in positive or negative symptoms or movement disorders, posthoc analyses were performed computing Pearson correlation coefficients between z-transformed mean changes in each neurocognitive factor score (including global score) and mean changes in PANSS Positive and Negative subscores and movement disorder scale (Abnormal Involuntary Movement Scale [AIMS] Severity, AIMS Movement, Simpson – Angus Rating Scale, and Barnes Akathisia Scale) scores.

105

Table 1 Demographic and baseline characteristics Conventional Olanzapine antipsychotics (n = 104) (n = 108) Men (%) Mean age in years (range) Schizophrenia (%) Receiving antiparkinsonian medications pre-study (%) Anticholinergic agents Dopaminergic agents Mean (S.D.) Simpson – Angus score Mean (S.D.) PANSS Total scorea

Risperidone (n = 58)

70 63 67 39.3 (18 – 61) 36.0 (19 – 57) 37.2 (18 – 61) 75 58

69 11

83 26

55

10

24

4

1

2

2.5 (3.1)

1.5 (2.7)

1.6 (2.3)

67.5 (16.3)

65.6 (16.7)

71.0 (18.7)

PANSS, Positive and Negative Syndrome Scale. a PANSS scores were available for 93, 88, and 41 patients, respectively, in the conventional, olanzapine, and risperidone groups.

switched to ziprasidone from conventional antipsychotics, olanzapine, or risperidone, respectively. Discontinuations due to insufficient clinical response occurred in 3.7%, 5.7%, and 1.7% of patients in the three respective groups. The mean duration of ziprasidone treatment in patients who had been switched from conventional antipsychotics, olanzapine, or risperidone was 35.1, 37.6, and 36.9 days, respectively. Among patients who completed the 6 weeks of treatment, ziprasidone dosages were similar across medication switching strategies and across the three studies. In the conventional antipsychotic study, the mean daily ziprasidone dose was 91.2 mg and the median was 80.0 mg; in the olanzapine study, the mean was 90.0 mg and the median was 84.3 mg; and in the risperidone study, the mean was 92.0 mg and the median was 85.8 mg.

3. Results

3.1. Cognitive battery scores

Baseline demographics and psychopathologic severity of patients enrolled in each of the three 6-week trials are shown in Table 1. The studies were completed by 72%, 79%, and 79% of patients who

Baseline scores, mean changes from baseline, and effect sizes (Cohen’s d) for cognitive battery tests for the three trials are shown in Table 2. A multivariate analysis of variance (MANOVA) was used to compare

106

Table 2 Changes in cognitive assessments following switch to ziprasidone from other antipsychotics Cognitive tests

Attention/vigilance DSDT Trailmaking A Executive function Trailmaking B WCST Categories attained, no. Total errors, no. Verbal fluency Category Fluency Letter Fluency

n

75 73 72

Baseline Mean change (S.E.) 37.2 7 0.84

Olanzapine to ziprasidone p

Effect n size

2.893 (0.866) 1.110 (0.321)

< 0.002 0.26 < 0.001 0.29

85 82

0.038 (0.014)

< 0.007 0.31

77

80 80

Baseline Mean change (S.E.) 38.4 6.7

Risperidone to ziprasidone p

Effect n size

4.494 (0.951) 1.256 (0.287)

< 0.001 0.39 < 0.001 0.34

46 45

0.037 (0.010)

< 0.001 0.36

44

23.9 23.5

0.342 (0.818) 0.388 (0.553)

0.667 0.06 0.485 0.08

42 42

78.8 76

2.114 (2.169) 0.224 (1.941)

0.333 0.10 0.908 0.01

42 43

0.84

Baseline Mean change (S.E.) 39.3 6.9

p

Effect size

3.935 (1.374) 1.422 (0.302)

< 0.007 0.32 < 0.001 0.37

0.038 (0.017)

< 0.030 0.33

22.4 22.7

1.763 (1.057) 0.738 (0.620)

0.134 0.21 0.240 0.11

74.4 46.0

4.422 (1.960) 1.047 (1.946)

< 0.030 0.22 0.593 0.06

0.84

70 70

25 23.3

1.169 (0.930) 0.086 (0.668)

0.222 0.28 0.898 0.02

63 69

70.1 51.7

5.442 (2.466) 6.638 (2.674)

0.031 0.23 < 0.020 0.21

73 44.2

73 131.3

18.630 (6.098)

< 0.004 0.23

83

100.2

0.482 (4.709)

0.918 0.00

46 109.0

6.826 (5.314)

0.205 0.12

69

3.8

0.246 (0.198)

0.218 0.09

77

4.0

0.273 (0.165)

0.102 0.09

45

3.6

0.756 (0.268)

0.007 0.34

69

47.4

0.696 (2.545)

0.785 0.02

76

43.6

3.118 (2.092)

0.140 0.12

45

47.2

9.711 (2.557)

< 0.001 0.35

85 76

34.6 27.9

1.263 (1.263) 2.658 (0.873)

0.140 0.12 < 0.004 0.24

85 46

35.3 31.4

1.576 (0.735) 0.262 (0.807)

< 0.040 0.12 0.746 0.02

47 46

34.5 29.0

0.617 (0.972) 2.370 (1.060)

0.528 0.04 < 0.040 0.21

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

Learning/memory Rey Total Learning Rey Long-Delay Recall Recognition/ discrimination Spatial Working Memory 5-s delay 15-s delay

Conventional to ziprasidone

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

the baseline cognitive functioning scores of the patients who did and did not complete a second cognitive assessment. There was no statistically significant overall difference between the groups, Wilks lambda = 0.52, Pillais Approx. F(11,269 = 1.24), p >0.40. 3.1.1. Learning and memory Improvements were observed in verbal but not spatial learning and memory for all three groups of patients switched to ziprasidone. Mean total learning scores on the RAVLT were significantly increased at endpoint for patients switched to ziprasidone from conventional antipsychotics ( p < 0.002), olanzapine ( p < 0.001), or risperidone ( p < 0.007). Improvements in delayed recall were also observed. Mean scores on the RAVLT for long-delay recall increased significantly for patients switched to ziprasidone from conventional antipsychotics ( p < 0.001), olanzapine ( p < 0.001), or risperidone ( p < 0.001). Recognition discrimination scores from this instrument showed the same overall pattern of significant improvement for patients switched to ziprasidone from conventional antipsychotics ( p < 0.007), olanzapine ( p < 0.001), or risperidone ( p < 0.027). Observed effects on performance of these tests were small to moderate. In contrast with the significant improvement noted in verbal learning and memory following the switch to ziprasidone, no significant change in the Spatial Working Memory Test was noted in either the 5- or 15-s delays in any of the three studies. 3.1.2. Attention and vigilance Improvement was observed in various assessments of attention and vigilance (Table 2). In the DSDT (distraction condition), significant improvement was noted in patients switched to ziprasidone from conventional antipsychotics ( p = 0.031) or from risperidone ( p < 0.030), but not in patients switched from olanzapine ( p = 0.333), who had the highest baseline scores on this test. Effects in patients switched from conventional antipsychotics or risperidone were small. Mean scores on the Trailmaking Test, part A, improved significantly in patients switched to ziprasidone from conventional antipsychotics ( p < 0.016). 3.1.3. Executive function For patients switched from conventional antipsychotics, mean scores were significantly improved on

107

the Trailmaking Test, part B ( p < 0.004). On the WCST, patients switched from olanzapine to ziprasidone showed a trend toward improvement in number of categories attained ( p = 0.102) and total errors ( p = 0.140). In patients switched from risperidone, mean total errors on the WCST declined significantly ( p < 0.001), as did number of categories attained ( p = 0.007). Effects for these patients were moderate. 3.1.4. Verbal fluency Mean changes in category fluency were significant in patients switched from olanzapine to ziprasidone ( p < 0.035). Mean changes in letter fluency were significant in patients switched from conventional

Table 3 Principal components factor analysis on 18 baseline scores with varimax rotation-factor loadings Cognitive function test

Spatial Working Memory (total no. of correct [5 s]) Spatial Working Memory (total no. of correct [15 s]) Continuous Performance (total signal detection) Continuous Performance (total reaction time) Trailmaking (part A time) Trailmaking (part B time) Category Fluency (no. of correct) Letter Fluency (no. of correct) Digit Span (% nondistraction) Digit Span (% distraction) Wisconsin Card Sorting (no. of categories attained) Wisconsin Card Sorting (total errors) Wisconsin Card Sorting (perseverative errors) Serial Verbal Learning (trial 1 performance) Serial Verbal Learning (trial 5 performance) Serial Verbal Learning (long-delay recall) Serial Verbal Learning (total performance) Serial Verbal Learning (recognition discrimination)

Verbal Attention and Executive skills short-term functioning memory 0.03

0.53

0.16

0.10

0.58

0.22

0.14

0.44

0.19

0.08

0.47

0.28

0.25 0.31 0.68

0.46 0.57 0.37

0.11 0.34 0.01

0.55 0.18 0.30 0.15

0.50 0.76 0.72 0.23

0.07 0.03 0.12 0.87

0.15

0.14

0.93

0.12

0.20

0.86

0.66

0.07

0.26

0.87

0.20

0.13

0.80

0.03

0.13

0.89

0.21

0.17

0.73

0.15

0.07

108

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

Fig. 1. Verbal skills domain, z-transformed mean change in dominant scale scores with 95% confidence intervals (CI).

agents ( p < 0.004) and from risperidone ( p < 0.031). Effects were small. 3.2. Factor analysis Principal components factor analysis on baseline scores suggested reduction of the 18 cognitive variables to three domains: verbal skills, attention/shortterm memory, and executive functioning (Table 3). For factor 1, verbal skills, the factor loadings showed seven contributing variables from serial verbal learning and fluency tests. The analysis of z-transformed mean change in factor scores showed significant improvement in verbal skills following the switch from conventional antipsychotics (mean, 0.38; 95% CI, 0.20 to 0.47), olanzapine (mean, 0.29; 95% CI, 0.15 to 0.43), or risperidone (mean, 0.36; 95% CI, 0.19 to 0.52) (Fig. 1). For factor 2, attention/short-

term memory, factor loadings showed eight contributing variables from the Spatial Working Memory, CPT, Trailmaking Test, and DSDT. All contributing variables other than trailmaking had the same sign; lower trailmaking scores mean better performance. The z-transformed mean change in factor scores showed improvement in attention/short-term memory following the switch to ziprasidone. The change was significant after the switch from risperidone (mean, 0.13; 95% CI, 0.02 to 0.25) but not from conventional agents (mean, 0.07; 95% CI, 0.04 to 0.17) or olanzapine (mean, 0.02; 95% CI, 0.11 to 0.17) (Fig. 2). For factor 3, executive functioning, factor loadings showed three contributing variables from the WCST. A lower total error and perseverative error variables score and a higher categories attained score reflect better performance. The z-transformed mean change in factor scale scores showed no significant

Fig. 2. Attention/short-term memory domain, z-transformed mean change in dominant scale scores, with 95% confidence intervals (CI).

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

109

Fig. 3. Executive functioning domain, z-transformed mean change in dominant scale scores, with 95% confidence intervals (CI).

change from baseline in patients switched from conventional antipsychotics (mean, 0.16; 95% CI, 0.41 to 0.09) or olanzapine (mean, 0.12; 95% CI, 0.13 to 0.38) but a significant change was found in those switched from risperidone (mean, 0.36, 95% CI, 0.092 to 0.63) (Fig. 3). For the global score, significant improvements were observed in patients switched from all three previous treatments: conventionals (mean, 0.13; 95% CI, 0.03 to 0.23), olanzapine (mean, 0.15; 95% CI, 0.01 to 0.25), or risperidone (mean, 0.26; 95% CI, 0.18 to 0.33) (Fig. 4). 3.3. Correlation analyses In the group switched from conventional antipsychotics, no significant positive or negative correla-

tions were observed between z-transformed mean change in any neurocognitive factors (three factors and the global score) and mean changes in abnormal movement scores or mean changes in PANSS Negative scores or PANSS Positive scores (r = 0.29 to 0.25; p > 0.05). Among patients switched from olanzapine, there were significant positive correlations between change in PANSS Positive score and change in factor 1 score (r = 0.43; p < 0.01) and between change in PANSS Positive scores and change in global score (r = 0.34; p < 0.05). Among patients switched from risperidone, there was a significant positive correlation between change in PANSS Negative score and change in factor 1 (r = 0.42; p < 0.05). No other significant positive or negative correlations were observed (r = 0.33 to 0.38; p >0.05).

Fig. 4. Global scale, z-transformed mean change, with 95% confidence intervals (CI).

110

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

There were no significant correlations between PANSS Positive scores and other neurocognitive factors, between PANSS Negative scores and any neurocognitive factor, or between movement disorder scores and any neurocognitive factor (r = 0.28 to 0.24; p> 0.05).

4. Discussion Switching patients with schizophrenia from either conventional or atypical antipsychotic agents to ziprasidone was associated with enhanced cognitive functioning in these three open-label studies. Improvements were seen in all cognitive domains tested, including verbal fluency, learning and memory, executive functioning, and attention, although not all domains improved in all three studies. Importantly, no deterioration in any domain was observed after the switch to ziprasidone. The problem of making multiple comparisons across three studies was addressed through factor analysis, by which 18 cognitive variables were reduced to three domains: verbal skills, attention/short-term memory, and executive function. Mean changes in factor scores showed improvement in the verbal skills domain in all three groups and in attention/short-term memory and executive function in patients switched from risperidone. Additionally, the change in global score calculated from 18 cognitive variables indicated improvement in all three groups. Although functional outcome in schizophrenia is determined by numerous factors, the domains used to aggregate cognitive improvements in this study are known to exert significant influence on the daily lives of patients with schizophrenia (Green et al., 2000). Thus, while the observed changes in certain cognitive tests in our study were modest (with effect sizes in the small to moderate range), their cumulative effect across different domains may in reality be sufficient to enhance functional outcome. This observation is supported by the effect sizes in individual tests, which exceeded 0.20 (small improvement) (Harvey and Keefe, 2001) in 8 of 12 cognitive tests in patients switched from either conventional antipsychotics (range, 0.02 to 0.31) or risperidone (range, 0.04 to 0.37) and in 3 of 12 tests in patients switched from olanzapine (range, 0.00 to 0.39). Although small, such

improvements are hardly inconsequential if one considers that they represent means, and that for some patients in each group actual improvements may have been quite large. Identifying which subsets of patients experience large cognitive improvements after switching antipsychotic medication would be clinically valuable. It is interesting to speculate why numerically more cognitive assessments showed improvement in patients switched from conventional agents or risperidone than in those switched from olanzapine. The most likely reason is that patients switched from olanzapine exhibited less severe psychopathology at baseline (mean PANSS Total, 65.6) than did patients switched from conventional agents (mean PANSS Total, 67.5) or risperidone (mean PANSS Total, 71.0). Although mean PANSS Total scores improved significantly in all three groups, the magnitude of improvement was also least in patients switched from olanzapine. On the other hand, mean baseline weight was greatest in the olanzapine group (93.4 kg), as was the degree of weight loss after the switch to ziprasidone ( p < 0.0001 versus baseline). Together, these data suggest that olanzapine patients were more likely to have been switched to ziprasidone because of poor tolerability than because of inadequate clinical effect. However, these studies were not designed to distinguish patients with inadequate symptom control from those experiencing tolerability issues and these data were not collected systematically. This may be an important topic for later research. Moreover, cognitive impairment may covary with severity of negative symptoms (Addington et al., 1991). Improvements in PANSS Negative Subscale scores were observed in all three studies (Weiden et al., 2003). Only among patients switched from risperidone did we find a significant correlation between improvement in PANSS Negative score and a change in a neurocognitive factor (factor 1, verbal skills). Thus, the majority of cognitive changes cannot be attributed to improvements in negative symptoms. Among patients switched from olanzapine, there were significant correlations between changes in PANSS Positive scores and changes in factor 1 (verbal skills) and global score, although these correlations accounted for only 10% of the variance in cognitive changes.

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

Conceivably, cognitive improvements in these patients could have been due to improvement of EPS and an associated reduced need for anticholinergic drugs, as both EPS and anticholinergic medication may be associated with impaired cognition in patients with schizophrenia (Spohn and Strauss, 1989; Cleghorn et al., 1990; Gelenberg et al., 1989). However, we found no significant correlation between improvements in neurocognitive factors and changes in movement disorder severity rating scales. Furthermore, significant improvement in the verbal skills domain was observed among patients switched from olanzapine, as well those switched from risperidone or conventional agents. Limitations to be borne in mind when interpreting these results include those inherent to an open-label switch design, such as rater or patient expectations, changes in patient motivation, and practice effects (Harvey and Keefe, 2001). In the current multi-group design, some of these effects are possibly obviated relative to single-group switches. Expectations, both patient and tester, regarding improvements associated with a switch to newer medications are likely to be similar across medications from which the patients were switched, while there was considerable variation in the extent to which patients improved across cognitive domains across the medications from which they switched. Practice effects are a potential problem in switch studies, particularly in verbal learning and problem solving tests. In several recent studies, however, schizophrenic subjects who were retested while receiving conventional antipsychotic medications failed to improve in their performance (Bilder et al., 2002; Harvey et al., 2002). If treatment with newer medications enhanced practice effects, then patients switched from risperidone or olanzapine should have improved more than patients switched from conventional medications, which was not consistent with the data. The improvement noted in patients switched from olanzapine or risperidone was found despite the fact that patients had been taking these other atypical agents for at least 3 months, at conventional doses, and did not show resistance to them in terms of symptom control. This suggests that the additional improvement in cognition noted in this study did not simply result from more time on newer medications. In conclusion, switching patients with schizophrenia to ziprasidone from other antipsychotics was

111

associated with improvements across a wide range of cognitive domains, including learning and memory, attention and vigilance, and executive function. If these results can be replicated in a study of more rigorous design, it would suggest that different atypical agents may be more beneficial for cognition than others in specific patients. These data suggest that ziprasidone therapy may modify the types of cognitive deficits that limit rehabilitation and predict poor functional outcome. Although accumulating data suggest that atypical antipsychotics, including ziprasidone, reduce cognitive impairment, specific inferences relating cognitive change to specific agents are yet to be replicated in large trials (Harvey et al., in press; Bilder et al., 2002). The improvements observed in these trials require confirmation in larger prospective, double-blind studies.

References Addington, J., Addington, D., Maticka-Tyndale, E., 1991. Cognitive functioning and positive and negative symptoms in schizophrenia. Schizophr. Res. 5, 123 – 134. Arato, M., O’Connor, R., Bradbury, J.E., Meltzer, H., 2002. A 1-year, double-blind, placebo-controlled trial of ziprasidone 40, 80 and 160 mg/day in chronic schizophrenia: the Ziprasidone Extended Use in Schizophrenia (ZEUS) study. Int. Clin. Psychopharmacol. 17, 207 – 215. Bilder, R.M., Goldman, R.S., Robinson, D., Reiter, G., Bell, L., Bates, J.A., Pappadopulos, E., Willson, D.F., Alvir, J.M., Woerner, M.G., Geisler, S., Kane, J.M., Lieberman, J.A., 2000. Neuropsychology of first-episode schizophrenia: initial characterization and clinical correlates. Am. J. Psychiatry 157, 549 – 559. Bilder, R.M., Goldman, R.S., Volavka, J., Czobor, P., Hoptman, M., Sheitman, B., Lindenmayer, J.P., Citrome, L., McEvoy, J., Kunz, M., Chakos, M., Cooper, T.B., Horowitz, T.L., Lieberman, J.A., 2002. Neurocognitive effects of clozapine, olanzapine, risperidone, and haloperidol in patients with chronic schizophrenia or schizoaffective disorder. Am. J. Psychiatry 159, 1018 – 1028. Blyler, C.R., Gold, J.M., 2000. Cognitive effects of typical antipsychotic medication treatment: another look. In: Sharma, T., Harvey, P.D. (Eds.), Cognition in Schizophrenia. Oxford Univ. Press, Oxford, pp. 241 – 265. Cleghorn, J.M., Kaplan, R.D., Szechtman, B., Szechtman, H., Brown, G.M., 1990. Neuroleptic drug effects on cognitive function in schizophrenia. Schizophr. Res. 3, 211 – 219. Cornblatt, B.A., Lenzenweger, M.F., Erlenmeyer-Kimling, L., 1989. The continuous performance test, identical pairs version: II. Contrasting attentional profiles in schizophrenic and depressed patients. Psychiatry Res. 29, 65 – 85.

112

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113

Cornblatt, B., Obuchowski, M., Roberts, S., et al., 1999. Cognitive and behavioral precursors of schizophrenia. Dev. Psychopathol. 11, 487 – 508. Daniel, D.G., Zimbroff, D.L., Potkin, S.G., Reeves, K.R., Harrigan, E.P., Lakshminarayanan, M., 1999. Ziprasidone 80 mg/ day and 160 mg/day in the acute exacerbation of schizophrenia and schizoaffective disorder: a 6-week placebo-controlled trial. Ziprasidone Study Group. Neuropsychopharmacology 20, 491 – 505. Davis, K.L., Kahn, R.S., Ko, G., Davidson, M., 1991. Dopamine in schizophrenia: a review and reconceptualization. Am. J. Psychiatry 148, 1474 – 1486. Gelenberg, A.J., Van Putten, T., Lavori, P.W., Wojcik, J.D., Falk, W.E., Marder, S., Galvin-Nadeau, M., Spring, B., Mohs, R.C., Brotman, A.W., 1989. Anticholinergic effects on memory: benztropine versus amantadine. J. Clin. Psychopharmacol. 9, 180 – 185. Goff, D.C., Posever, T., Herz, L., Simmons, J., Kletti, N., Lapierre, K., Wilner, K.D., Law, C.G., Ko, G.N., 1998. An exploratory haloperidol-controlled dose-finding study of ziprasidone in hospitalized patients with schizophrenia or schizoaffective disorder. J. Clin. Psychopharmacol. 18, 296 – 304. Green, M.F., 1996. What are the functional consequences of neurocognitive deficits in schizophrenia? Am. J. Psychiatry 153, 321 – 330. Green, M.F., Marshall Jr., B.D., Wirshing, W.C., Ames, D., Marder, S.R., McGurk, S., Kern, R.S., Mintz, J., 1997. Does risperidone improve verbal working memory in treatment resistant schizophrenia? Am. J. Psychiatry 154, 799 – 804. Green, M.F., Kern, R.S., Braff, D.L., Mintz, J., 2000. Neurocognitive deficits and functional outcome in schizophrenia: are we measuring the ‘‘right stuff’’? Schizophr. Bull. 261, 119 – 136. Hagger, C., Buckley, P., Kenny, J.T., Friedman, L., Ubogy, D., Meltzer, H.Y., 1993. Improvement in cognitive function and psychotic symptoms in treatment-refractory schizophrenic patients. Biol. Psychiatry 34, 702 – 712. Harvey, P.D., Keefe, R.S.E., 2001. Studies of cognitive change in patients with schizophrenia following novel antipsychotic treatment. Am. J. Psychiatry 158, 176 – 184. Harvey, P.D., Heaton, R.K., Palmer, B.A., Mohamed, S., Kennedy, J., Brickman, A., 2002. Stability of cognitive impairment in elderly schizophrenic patients receiving conventional antipsychotic treatment. J. Clin. Psychiatry (Abstract) 63, 1065. Harvey, P.D., Green, M.F., Meltzer, H.Y., et al., 2003. The cognitive effects of risperidone and olanzapine in patients with schizoaffective disorder or schizophrenia. Psychopharmacology (in press). Heaton, R.K., Chelune, C.J., Talley, J.L., Kay, G.G., Curtiss, G., 1993. Wisconsin Card Sorting Test Manual—Revised and Expanded Psychological Assessment Resources, Odessa, FL. Hegarty, J.D., Baldessarini, R.J., Tohen, M., 1994. One hundred years of schizophrenia: a meta-analysis of the outcome literature. Am. J. Psychiatry 151, 1409 – 1416. Heinrichs, R.W., Zakzanis, K.K., 1998. Neurocognitive deficit in schizophrenia: a quantitative review of the evidence. Neuropsychology 12, 426 – 445. Hirsch, S., Kissling, W., Bauml, J., Power, A., O’Connor, R.,

2002. A 28-week comparison of ziprasidone and haloperidol in outpatients with stable schizophrenia. J. Clin. Psychiatry 63, 516 – 523. Ichikawa, J., Dai, J., O’Laughlin, I.A., Fowler, W.L., Meltzer, H.Y., 2002. Atypical, but not typical, antipsychotic drugs increase cortical acetylcholine release without an effect in the nucleus accumbens or striatum. Neuropsychopharmacology 26, 325 – 339. Ichikawam, J., Ishii, H., Bonaccorso, S., Fowler, W.L., O’Laughlin, I.A., Meltzer, H.Y., 2001. 5-HT(2A) and D(2) receptor blockade increases cortical DA release via 5-HT(1A) receptor activation: a possible mechanism of atypical antipsychotic-induced cortical dopamine release. J. Neurochem. 76, 1521 – 1531. Kay, S.R., 1991. Positive and Negative Syndromes in Schizophrenia. Brunner-Mazel, New York. Keck Jr., P., Buffenstein, A., Ferguson, J., Feighner, J., Jaffe, W., Harrigan, E.P., Morrissey, M.R. 1998. Ziprasidone 40 and 120 mg/day in the acute exacerbation of schizophrenia and schizoaffective disorder: a 4-week placebo-controlled trial. Psychopharmacology 140, 173 – 184. Keefe, R.S., Silva, S.G., Perkins, D.O., Lieberman, J.A., 1999. The effects of atypical antipsychotic drugs on neurocognitive impairment in schizophrenia: a review and meta-analysis. Schizophr. Bull. 25, 201 – 222. Meltzer, H.Y., 1999. The role of serotonin in antipsychotic drug action. Neuropsychopharmacology 21, 106S – 115S. Meltzer, H.Y., McGurk, S.R., 1999. The effects of clozapine, risperidone, and olanzapine on cognitive function in schizophrenia. Schizophr. Bull. 25, 233 – 255. Meltzer, H.Y., Matsubara, S., Lee, J.C., 1989. Classification of typical and atypical antipsychotic drugs on the basis of dopamine D-1, D-2 and serotonin2 pKi values. J. Pharmacol. Exp. Ther. 251, 238 – 246. Mojtabai, R., Bromet, E., Harvey, P.D., Carlson, G.A., Craig, T.J., Fennig, S., 2000. Are there neuropsychological differences between first episode patients with schizophrenia and affective disorder? Am. J. Psychiatry 157, 1453 – 1460. Oltmanns, T.F., Neale, J.M., 1975. Schizophrenic performance when distractors are present: attentional deficit or differential task difficulty? J. Abnorm. Psychology 84, 205 – 209. Potkin, S.G., Fleming, K., Jin, Y., Gulasekaram, B., 2001. Clozapine enhances neurocognition and clinical symptomatology more than standard neuroleptics. J. Clin. Psychopharmacol. 21, 479 – 483. Purdon, S.E., 1999. Cognitive improvement in schizophrenia with novel antipsychotic medications. Schizophr. Res. 35, S51 – S60. Rollema, H., Lu, Y., Schmidt, A.W., Zorn, S.H., 1997. Clozapine increases dopamine release in prefrontal cortex by 5-HT1A receptor activation. Eur. J. Pharmacol. 338, R3 – R5. Roth, B.L., Sheffler, D., Potkin, S.G., 2003. Atypical antipsychotic drug actions: unitary or multiple mechanisms for multiple mechanisms for ‘atypicality’? Clin. Neurosci. Res. (in press). Saykin, A.J., Shtasel, D.L., Gur, R.E., Kester, D.B., Mozley, L.H., Stafiniak, P., Gur, R.C., 1994. Neuropsychological deficits in neuroleptic naive patients with first-episode schizophrenia. Arch. Gen. Psychiatry 51, 124 – 131. Schmidt, A.W., Lebel, L.A., Howard Jr., H.R., Zorn, S.H. 2001.

P.D. Harvey et al. / Schizophrenia Research 66 (2004) 101–113 Ziprasidone: a novel antipsychotic agent with a unique human receptor binding profile. Eur. J. Pharmacol. 425, 197 – 201. Seeger, T.F., Seymour, P.A., Schmidt, A.W., Zorn, S.H., Schulz, D.W., Lebel, L.A., McLean, S., Guanowsky, V., Howard, H.R., Lowe III, J.A., et al., 1995. Ziprasidone (CP-88,059): a new antipsychotic with combined dopamine and serotonin receptor antagonist activity. J. Pharmacol. Exp. Ther. 275, 101 – 113. Spohn, H.E., Strauss, M.E., 1989. Relation of neuroleptic and anticholinergic medication to cognitive functions in schizophrenia. J. Abnorm. Psychology 98, 367 – 380. Spreen, O., Strauss, E., 1998. A Compendium of Neuropsychological Tests and Norms, 2nd ed. Oxford Univ. Press, New York. Sprouse, J.S., Reynolds, L.S., Braselton, J.P., Rollema, H., Zorn,

113

S.H., 1999. Comparison of the novel antipsychotic ziprasidone with clozapine and olanzapine: inhibition of dorsal raphe cell firing and the role of 5-HT1A receptor activation. Neuropsychopharmacology 21, 622 – 631. Strauss, M.E., Reynolds, K.S., Jayaram, G., Tune, L.E., 1990. Effects of anticholinergic medication on memory in schizophrenia. Schizophr. Res. 3, 127 – 129. Weiden, P.J., Simpson, G., Potkin, S.G., Romano, S.J., 2003. Effectiveness of switching to ziprasidone for stable but symptomatic outpatients with schizophrenia. J. Clin. Psychiatry 64, 580 – 588. Weinberger, D.R., Gallhofer, B., 1997. Cognitive function in schizophrenia. Int. Clin. Psychopharmacol. 12, S29 – S36.