This study describes an interim analysis of changes in cognitive function in young patients with a first episode of psychosis during treatment with quetiapine. The symptomatic relief afforded by second-generation antipsychotic medications (clozapine, risperidone, olanzapine and quetiapine) is well-described (Reference Kane, Honigfeld and SingerKane et al, 1988; Reference Marder and MeibachMarder & Meibach, 1994). The second generation agents reduce the positive symptoms but in contrast to the older (typical) antipsychotics also improve negative or deficit symptoms (Reference Kane, Honigfeld and SingerKane et al, 1988; Reference Hagger, Buckley and KennyHagger et al, 1993). As well, a distinctive side-effect profile, including fewer extrapyramidal side-effects, is a defining feature of these newer agents (Reference Kopala, Good and HonerKopala et al, 1997).

Although the efficacies of second-generation antipsychotics in reducing symptoms are reasonably well-documented, their effects on cognition are less clearly understood. Evidence to date, however, indicates that these newer compounds are able to target and improve specific aspects of cognition (Reference Hagger, Buckley and KennyHagger et al, 1993; Reference Hoff, Faustman and WienekeHoff et al, 1996; Reference Purdon, Jones and StipPurdon et al, 2000). Specifically, treatment with clozapine has been linked to possible improvements in verbal fluency and visuomotor tracking (Reference Hagger, Buckley and KennyHagger et al, 1993; Reference Lee, Thompson and MeltzerLee et al, 1994; Reference Hoff, Faustman and WienekeHoff et al, 1996), whereas risperidone treatment has been associated with probable enhancement of verbal working memory, new learning and attention (Reference Green, Marshall and WirshingGreen et al, 1997; Reference Purdon, Jones and StipPurdon et al, 2000). A more controversial profile of improvements (which requires replication) has been demonstrated with olanzapine treatment (learning, motor skills, executive abilities and visuospatial functioning; Reference Purdon, Jones and StipPurdon et al, 2000). A recent meta-analysis of the extant literature examining the effects of these medications on cognition indicated a modest effect size (Reference Keefe, Silva and PerkinsKeefe et al, 1999). However, in many reports, potentially confounding variables were not ruled out. Difficulties ascertaining drug-specific effects persist (Reference Harvey and KeefeHarvey & Keefe, 2001).

The effects of antipsychotic medication on cognitive function depend on accurate characterisation of the nature and magnitude of cognitive impairment in the group of patients to be studied. When examined longitudinally, patients who are in the early phases of a psychotic disorder (first-episode) demonstrate a remarkably consistent pattern of neuropsychological deficits over time, at least in the short-term (Reference Censits, Ragland and GurCensits et al, 1997). Longer-term follow-up studies are required to determine if further deterioration occurs with illness chronicity. In general, patients with non-affective psychotic disorders, such as schizophrenia, schizophreniform disorder and schizoaffective disorder are impaired on measures of memory, both verbal and non-verbal, attention, executive function, fluency and motor speed (Reference Saykin, Shtasel and GurSaykin et al, 1994).

Few studies have examined changes in cognitive functioning after treatment with quetiapine (Reference Sax, Strakowski and KeckSax et al, 1998; Reference Velligan and MillerVelligan & Miller, 1999; Reference Purdon, Malla and LabellePurdon et al, 2001). To date, no study has examined patients with a first-episode of psychosis before and after treatment with quetiapine. Despite the small number of chronically ill patients investigated, two double-blind studies provided complementary results, demonstrating significant improvements in verbal reasoning, fluency, executive function, motor speed and verbal memory (Reference Velligan and MillerVelligan & Miller, 1999; Reference Purdon, Jones and StipPurdon et al, 2000). These effects are in contrast to those observed in haloperidol-treated patients (Reference Velligan and MillerVelligan & Miller, 1999; Reference Purdon, Malla and LabellePurdon et al, 2001). A further study documented amelioration of attentional capacity after treatment with quetiapine (Reference Sax, Strakowski and KeckSax et al, 1998).

Cognitive performance is believed to be the single most important factor contributing to longer-term functional recovery (Reference Green, Kern and BraffGreen et al, 2000). Consequently, the effects of second-generation medications on cognition hold exciting promise for improving functional adaptability, particularly for those in the early stages of illness. The goal of this study was to examine first-episode patients and document the effects of quetiapine on their cognitive abilities. We hypothesised that improvements would be noted on measures of verbal and visuospatial memory, attention and executive functioning. Because of the low potential for extrapyramidal side-effects, no changes in motor/sensory performance were anticipated.

METHOD

Study design

Before initiation of quetiapine, baseline assessments were performed. These included the Positive and Negative Syndrome Scale (PANSS; Reference Kay, Fiszbein and OplerKay et al, 1987) and Global Assessment of Function (GAF; DSM—IV; American Psychiatric Association, 1994). These ratings were repeated weekly for 6 weeks, followed by monthly thereafter, until study completion. Diagnosis was re-evaluated at 6 months.

For all patients, exposure to antipsychotic medications was brief, so a medication wash-out period was not deemed necessary. Antipsychotic medications were crossed over. Quetiapine was initiated at 25 mg daily, then 25 mg twice daily and increased by 50-100 mg approximately every 3 days thereafter until therapeutic dose was reached. The maximum dose employed was 800 mg. (Maximum doses vary among different countries. According to Canadian Guidelines (the country in which this study was completed), 800 mg is the maximum recommended dose.) Therapeutic dose was defined as the level at which psychotic symptoms were adequately controlled in the absence of problematic side-effects.

Cognitive testing was performed before quetiapine was initiated and was repeated at 3 months, 6 months and 1 year. The neuropsychological test battery was chosen to assess cognitive domains that are impaired in patients with psychotic disorders and would have functional significance if performance were ameliorated. The battery included the following tasks (for more information see Reference LezakLezak, 1995).

1. (a) Rey Auditory Verbal Learning Test. This task is thought to measure immediate memory span, learning curves and strategies for verbally presented material.

2. (b) Benton Visual Retention Test. This task is used to measure recall of visually presented material.

3. (c) Wisconsin Card Sorting Test. This task has been extensively used in research in schizophrenia and appears to examine abstract behaviour and set-shifting ability.

4. (d) Continuous Performance Test. This task is thought to measure the capacity to sustain a focus of attention.

5. (e) Object Alternation. This is a task that has been borrowed from the non-human primate literature and modified to examine executive functioning, specifically response switching.

6. (f) Verbal Fluency. This task assesses verbal productivity by asking the subject to spontaneously generate words that begin with a given letter of the alphabet.

7. (g) The Trail Making Test. This is a two-part test of visuomotor and conceptual tracking.

8. (h) Grooved Pegboard. This task assesses motor speed and accuracy.

9. (i) Finger Tapping. This task measures motor speed.

The list of specific variables examined from each neuropsychological test can be found in Table 1. Alternate versions were employed for those tests for which they are available.

Table 1

Neuropsychological test battery and the specific cognitive variables investigated

Test	Variable examined
Auditory Verbal Learning Test1	Total recalled over 5 trials2
Benton Visual Retention Test1	Number of errors
Wisconsin Card Sorting Test	Number of perseverative errors
Object Alternation	Number of perseverative errors
Continuous Performance Test	Ability to discriminate target from non-target items, d' total2
Verbal Fluency1	Total number of words generated2
Trail Making Test	Time for Trails A and B
Grooved Pegboard	Time to completion (dominant and non-dominant hands)
Finger Tapping	Mean number of taps2 (dominant and non-dominant hands)

RESULTS

A total of 34 patients were enrolled into the study (6 females and 28 males). The sample at study entry was 24.1 (s.d.=6.5) years of age and had been educated, on average, for 12.4 (s.d.=2.3) years. At study entry, patients were mildly to moderately ill according to the positive and negative subscales of the PANSS. Symptomatic improvement in the positive and negative sub-scales was noted over time (see Fig. 1). At the time of interim analysis, 27 had completed the 3-month evaluation, 23 finished the 6-month assessment and 13 had reached the 1-year mark. Eight patients' data were excluded from the study. Three patients withdrew because of lack of compliance, three withdrew consent and two were discontinued for lack of efficacy.

Fig. 1

Positive ([UNK]) and negative ([UNK]) sub-scale ratings from the Positive and Negative Syndrome Scale at baseline, 3 months, 6 months and 1 year.

Twelve patients (35%) were antipsychotic medication naïve at the time of study entry. The remainder had been medicated for 7.4 (s.d.=6.2) weeks on average as follows: risperidone (n=11); quetiapine (n=6); haloperidol (n=2); loxapine (n=2); or olanzapine (n=1). Mean chlorpromazine equivalent dosage for initial medications x̄=288.3 (s.d.=172.4) mg/day.

On average, male patients were treated with higher doses of quetiapine (x̄=520.0 mg/day, s.d.=220.1) than females (x̄=512.5 mg/day, s.d.=275.0). Mean optimal dose for the first-episode patients studied was 517.9 mg/day (s.d.=225.8).

Mean neuropsychological test scores over the four study visits are shown in Table 2. With the exception of the timed motor tasks (Finger Tapping and Grooved Pegboard), performance was enhanced on all tasks during treatment with quetiapine. Statistically reliable improvements were noted on the Continuous Performance Test (CPT) d′ total score, i.e. the ability to discriminate target from non-target items), Object Alternation (number of perseverative errors) and Verbal Fluency Test (total score). Simple contrast post hoc tests indicated that CPT performance was improved at all three time points examined (3-month, 6-month and 1-year) relative to baseline, whereas Object Alternation and Verbal Fluency capacity was enhanced at 6 months and 1 year.

Table 2

Neuropsychological test scores at study test points (mean, 95% CI)

Test	Baseline	3 month	6 month	1 year	P
Auditory Verbal Learning Test	47.9	48.6	48.8	49.1	NS
	41.8-54.0	41.8-55.4	40.8-56.8	38.1-60.1
Benton Visual Retention Test	2.3	2.1	2.2	1.8	NS
	1.32-3.27	1.12-3.08	0.95-3.45	0.77-2.83
Wisconsin Card Sorting Test	10.3	21.6	11.2	8.2	NS
	6.46-14.1	8.42-34.8	4.34-18.1	5.01-11.4
Object Alternation	4.0	3.2	0.381	1.21	0.006
	1.88-6.12	0.92-5.48	0.03-0.73	0.17-2.32
Continuous Performance Test	1.2	1.71	1.91	1.81	0.001
	0.77-1.63	1.16-2.24	1.36-2.44	1.47-2.13
Verbal Fluency	34.3	37.0	38.81	41.51	0.02
	28.9-39.7	30.4-43.6	30.8-46.8	33.9-49.1
Trail Making Test
Trails A	28.0	26.2	22.5	22.9	NS
	21.7-34.3	19.1-33.3	19.0-26.0	18.7-27.1
Trails B	61.3	57.4	52.5	54.7	NS
	49.9-72.7	42.4-72.4	40.6-64.4	45.0-64.4
Grooved Pegboard2
Dominant	65.5	64.9	60.2	64.0	NS
	60.1-70.9	61.6-68.2	54.4-66.0	61.0-67.0
Non-dominant	71.6	71.2	67.4	68.2
	65.5-77.7	65.6-76.8	62.7-72.1	64.6-71.8
Finger Tapping2
Dominant	50.7	51.9	51.6	52.3	NS
	47.3-54.1	48.4-55.4	47.5-55.7	48.3-56.3
Non-dominant	47.5	46.8	46.8	47.7
	44.7-50.3	43.5-50.1	43.6-50.0	44.1-51.3

DISCUSSION

This study documents the interim analysis of a longer-term study of the effects on cognitive performance in a cohort of young patients with a first episode of schizophrenia and related disorders who were treated with quetiapine. To our knowledge, the data provide the first clear indication that cognitive test performance is improved in young patients with psychosis who were treated with quetiapine. Statistically significant improvements were observed on measures of attention, executive function and verbal fluency at 6 months and 1 year of treatment. Improvements in attention (CPT) were also noted at the 3-month testing session. These findings are consistent with studies of chronically ill, medicated patients.

Memory and attention deficits are common in patients with psychotic disorders (Reference Lussier and StipLussier & Stip, 2001) and could represent the domains of cognition that are most severely disturbed (Reference O'CarrollO'Carroll, 1992; Reference Saykin, Shtasel and GurSaykin et al, 1994). Our data clearly demonstrated impairment in these cognitive domains at baseline as compared with published normative data. Of these domains, performance on a measure of attention attained normal levels and also underwent statistically significant improvement. Performance on verbal and non-verbal memory measures did not reach significance, although there did appear to be an indication of improved functioning over time.

The number of perseverative errors on Object Alternation (a task used to assess response switching) was reduced to below baseline levels for a sample of healthy controls (details available from author on request). In contrast, the number of perseverative errors from the Wisconsin Card Sorting Test did not reliably decrease. However, an unusual pattern of performance was observed on this measure: better scores were observed at baseline than on subsequent sessions, suggesting a non-representative estimate of baseline functioning. A larger sample could permit better understanding of this finding.

There are several advantages to the study design employed. The open label design ensured that patients were treated optimally and thus, potential symptomatic confounds were minimised. Also, examining patients with a first episode of illness is beneficial as the pre-drug status was known (naïve or short-term prior exposure) and is likely to contribute minimally to the observed effects. Finally, the frequency of cognitive assessments and adequacy of the neuropsychological test battery allowed for the examination of early and longer-term effects on cognition (Reference Keefe, Silva and PerkinsKeefe et al, 1999). At the same time, we were able to diminish the motivational challenges that occur with too frequent or elaborate assessments.

The lack of a control group (or a group of patients treated with another agent) leaves unanswered whether the improvements noted were simply a result of facilitation because of repeated exposure to the tasks. Several factors argue against this interpretation. First, alternate forms were employed for those tasks for which they were available (see Table 1). Although this measure would not mitigate procedural practice effects, content-specific memory would not be affected. The greatest influence of practice may reasonably be expected to occur on initial readministration with inter-test intervals of 3 months or less. Performance in our sample continued to improve beyond the second testing session. Second, patients who are treated with conventional antipsychotic medications do not demonstrate cognitive improvement with repeated administration of neuropsychological tests (Reference Kern, Green and BarringerKern et al, 1998; Reference Purdon, Jones and StipPurdon et al, 2000). Nevertheless, without a comparator group, the effects of practice cannot be ruled out. Regardless of the basis for improvement, patients in the current study did improve over time and this observation may be the most meaningful.

Despite modest sample size, statistical improvement over time was observed in a quarter of the cognitive test variables studied. For all other test variables (with the exception of the timed motor/sensory tasks, which were predicted to be unchanged), all mean scores showed trends towards better performance with treatment. These values supported our hypotheses and will be carefully examined in future analyses.

The results of this preliminary study are encouraging but remain tentative. Clearly, larger sample sizes, along with more rigorous study designs (including patients treated with other antipsychotic medications) will allow for better characterisation of the drug-specific effects of quetiapine on various domains of cognition. Nevertheless, this study indicates that, at least, quetiapine is not likely to have adverse effects on cognitive function and could well have beneficial effects. Given the strong relationship between cognition and functional adaptation, continued intensive study of the effects of quetiapine on cognition is highly desirable. Further data on patients' functional adaptability after treatment with quetiapine and other aspects of the study will be forthcoming.

CLINICAL IMPLICATIONS

1. ▪ In young patients with first-episode psychosis treated with quetiapine, statistically significant improvement was noted on measures of attention (Continuous Performance Test; CPT), verbal productivity (Verbal Fluency Test) and executive function (Object Alternation Test) after 6 months and 1 year of treatment.

2. ▪ For the CPT, improvement was also noted after 3 months of treatment.

3. ▪ Positive and negative symptoms, as measured by the Positive and Negative Syndrome Scale, appear to be well-managed by the doses of quetiapine used in this study.

LIMITATIONS

1. ▪ Without a control group, the effects of practice cannot be ruled out.

2. ▪ Larger sample sizes are required to confirm the validity of these findings.

3. ▪ Further study is required to determine whether these putative improvements in cognition translate into enhanced functional adaptability (outcome).