Introduction
Schizophrenia often begins during adolescence or early adulthood and presents a broad impairment of cognitive or neuropsychological functions, including the executive functions, perceptual and motor processing, attentional skills, vigilance, verbal learning and memory, spatial working memory and verbal fluency Reference Bilder, Goldman and Robinson(1,Reference Bema, Kfksal and Haluk2). Cognitive dysfunction not only is a core syndrome but also is correlated with functional impairment of schizophrenia. Studies showed that the neurocognitive deficits appear predictive of social functioning, occupational functioning and the capacity for independent living in the community (Reference Green3–Reference Velligan and Miller5). Undoubtedly, the assessment of the level of cognitive functioning is regarded as a crucial dimension of the assessment of treatment outcomes in schizophrenia. Quetiapine is indicated for the treatment of schizophrenia. Its antipsychotic effect is thought by some to be mediated through antagonist activity at dopamine and serotonin receptors. Specifically, the D1 and D2 dopamine receptor, the alpha-1 and alpha-2 adrenergic receptor and 5-HT1A and 5-HT2 serotonin receptor subtypes are antagonised. Serial positron emission tomography scans evaluating the D2 receptor occupancy of quetiapine have showed that quetiapine very rapidly disassociates from the D2 receptor Reference Sharma(6). Theoretically, this allows for normal physiological surges of dopamine to elicit their normal effects in areas such as the nigrostriatal and tuberoinfundibular pathways, thus minimising the risk of side-effects such as pseudo-parkinsonism as well as minimising elevations in prolactin. Common side-effects of quetiapine include constipation, headache, dry mouth and weight gain (or loss). However, those effects may occur to a lesser degree compared to some other atypical antipsychotics such as olanzapine or clozapine. Some studies reported that quetiapine improves impaired cognitive functions in schizophrenia, including verbal fluency, memory and executive functions (Reference Purdon, Malla and Labelle7–Reference Meisenzahl, Scheuerecker and Zipse9). Two strategies have been used to evaluate the effects of quetiapine on cognitive functions in schizophrenia. First, neurocognitive and functional measures through the use of performance-based competence assessments as outcome measures in clinical trials were used to evaluate the effectiveness of quetiapine in improving cognitive and functional outcomes. Other studies reported that quetiapine had beneficial effects not only on cognitive skills but also particularly on verbal reasoning, fluency skills and immediate recall. An additional improvement in executive skills and visuomotor tracking and on the average of the six cognitive domains was also noted with sustained treatment by standardised neuropsychological measures Reference Purdon, Malla and Labelle(7). A cognitive function summary score improved in the quetiapine group relative to the group treated with conventional antipsychotics. Patients taking quetiapine did better with respect to both verbal fluency (initiation) and verbal memory. There were also statistically significant group differences with respect to quality of life favouring the quetiapine group Reference Velligan and Prihoda(10).
Second, the functional magnetic resonance imaging (fMRI) studies showed that in the verbal fluency task, there was significantly increased activation in the left inferior frontal cortex in the quetiapine-treated patients compared with the drug-naive sample. During auditory stimulation, the stably treated patients produced significantly greater activation in the superior temporal gyrus than did the drug-naive sample. The conclusion was that quetiapine treatment is associated with altered responses in blood oxygen level dependence in both the prefrontal and temporal cortex that cannot be accounted for by improved task performance subsequent to drug treatment Reference Jones, Brammer and Toole(11). In addition, it was concluded through fMRI assessment that the neuronal networks underlying working memory are clearly altered in schizophrenia Reference Meisenzahl, Scheuerecker and Zipse(9).
Cognitive deficits in schizophrenia not only include deficient performance in higher cognitive domains but also extend to information processing at the sensory and preattentive levels Reference Daniel, Umbricht and Bates(12). The event-related potential (ERP) mismatch negativity (MMN) is an effective measure of preattentive information processing by auditory change detection. MMN is a negative ERP component elicited by deviant stimulus, i.e. the standard sounds changed frequency, duration, intensity or location. Namely, the MMN is believed to be the outcome of a comparison process, commonly elicited when an incoming stimulus differs from the memory of repetitive tones (standards) occurring in the recent acoustic past. Because in the absence of such a memory trace, no MMN is generated, MMN generation indexes auditory sensory memory and reflects context-dependent information processing at the level of the auditory sensory cortex. There is no agreement to date regarding whether MMN reflects an attention-independent process. A previous study originally proposed that the MMN was unaffected by various attentional manipulations Reference Naatanen(13). The conclusion was based on evidence suggesting that it was the mismatch process between the neural trace of the standard (the repetitive sound) and the incoming deviant sound that generates the MMN response, which was unaffected. Several studies showed that MMN generation was affected by attention (Reference Alain and Woods14–Reference Woldorff, Hillyard and Gallen16). In addition, MMN likely reflects N-methyl-d-aspartate channel current influx in cortical layers II and III based on animal and human experiments Reference Lavoie, Murray and Deppen(17).
To date, there have been many studies on MMN in schizophrenia. Several studies reported that MMN amplitude in response to tone duration and tone frequency deviants was reduced in patients with schizophrenia Reference Umbricht and Krljes(18,Reference Anke, Indira and Ralf19). One study showed that the presence of MMN abnormalities was related to illness onset Reference Javitt, Shelley and Silipo(20). However, another study displayed normal MMN generation in first-episode patients Reference Salisbury, Shenton and Grigge(21). Thus, whether MMN deficits are present at illness onset remains controversial. Furthermore, MMN deficits are not observed in other major mental illnesses such as bipolar disorder or major depression, suggesting that they may be fairly specific to schizophrenia Reference Umbricht, Koller and Schmid(22). However, several studies had found that MMN amplitude diminished in dyslexia, aphasia, Asperger’s syndrome and Alzheimer’s disease (Reference Baldeweg, Richardson and Watkins23–Reference Pekkonen, Hirvonen and Jaaskelamen26). At this time, the data are not able to specify clearly the association between the diagnosis of schizophrenia and decreased MMN potentials involved in psychiatric illnesses.
Since changed MMN reflects abnormalities of early auditory processing in schizophrenia, we suppose that a sufficient treatment period (i.e. more than 8 weeks) with quetiapine may lead to the improvement of MMN. Up to now, no studies for the effects of quetiapine on MMN in schizophrenia have been reported. The goal of the present study was to investigate whether the effects of quetiapine on abnormalities of early auditory processing in patients with schizophrenia were reflected by auditory MMN.
Materials and methods
Study subjects
Subjects were 23 patients with Diagnostic and Statistical Manual of Mental Disorders (4th ed, DSM-IV) diagnosis criteria for schizophrenia and 23 matched age and gender controls with no personal or family history of schizophrenia. Patients with schizophrenia were recruited from Wuxi Mental Heath of Nanjing Medical University in Jiangsu, China. Controls were recruited from the employees of Wuxi Mental Health Center of Nanjing Medical University. Subjects and controls were excluded from the study if they were smokers; had a diagnosis of alcohol or substance dependence, neurological disorders and all kinds of head injury or had received electroconvulsive therapy in the past 6 months. All participants were Chinese. They all gave written informed consent to participate and all were paid. The protocol for the research project was approved by the Ethics Committee of Nanjing Medical University, China.
Clinical assessments
All participants underwent a clinical assessment by a psychiatrist to collect information on medication and sociodemographic data and to confirm/exclude a DSM-IV diagnosis. On the day of the ERP recording, psychopathology was rated in patients with the Positive and Negative Syndrome Scale (PANSS) Reference Kay, Fiszbein and Opler(27). Handedness was assessed using the Annett Handedness Scale Reference Annett(28). Ratings on this scale were recoded into the following definitions of handedness: Annett score 1 = right, 2–7 = mixed, 8 = left.
The demographic characteristics of the sample are detailed in Table 1.
Table 1 Demographic characteristics of the sample
L, left; M, mixed; R, right; SD, standard deviation.
Experimental procedure
Stimulation protocol and procedure ERP recordings were acquired during the presentation of auditory stimuli. Auditory stimuli consisted of 100 ms/1000 Hz standards intermixed with 100 ms/1500 Hz frequency deviants and 250 ms/1000 Hz duration deviants. All stimuli had a rise-fall time of 5 ms. Stimuli were presented in a fixed order (four standards, one frequency deviant, four standards and one duration deviant) with a stimulus onset asynchrony of 300 ms. The stimuli were presented through foam insert earphones at a nominal intensity of a 75-dB level. Stimuli were presented in four blocks with 1000 stimuli, each totalling 4000 stimuli including 3200 standards, 400 frequency deviants and 400 duration deviants. During presentation of the auditory test paradigm, subjects watched a silent self-selected video film to divert attention from the tones and to minimise boredom and reduce eye movement artefacts. Subjects were constantly monitored. Short breaks were offered to ensure full alertness and comfort during the recording session.
In order to detect the treatment effects on MMN, auditory ERPs were recorded at baseline, 4 weeks and 8 weeks of quetiapine treatment. For healthy controls, ERPs were recorded once. At baseline, 17 patients were neuroleptic naive and 6 neuroleptic free (2 for at least half a year and 4 for at least 1 month). After 2 weeks of follow-up, patients received quetiapine 600–800 mg/day (mean value 721.7, standard deviation 90.2).
Electroencephalographic recordings According to the 10/20 International System, electroencephalography (EEG) was recorded with the Stellate Harmonie EEG device (Physiotec Electronics Ltd., Montreal, Quebec, Canada) from Fz, left mastoid and right mastoid sites using Electro-Cap Electrode System (ECI™ Electro-Caps; Electro-cap International, Eaton, OH, USA). Ear electrodes served as a reference, and the ground electrode was attached to the forehead. Eye movement artefacts were monitored by recording vertical and horizontal electro-oculogram from electrodes placed above and below the right eye and at the left outer canthus. Electrode impedance was kept below 5 kΩ. System band pass was 0.1–30 Hz and digitalised continuously at a sampling rate of 250 Hz. Digital tags were obtained for all auditory stimuli.
Data analysis
Brain Electrical Source Analysis (BESA) program (version 5.1.8, software) was used to perform data analysis. Epochs were constructed that consisted of a 100-ms pre-stimulus baseline and a 500-ms post-stimulus interval. All epochs with amplitudes exceeding ±75 μV at any electrode were excluded automatically. Epochs were averaged offline for each subject and stimulus type and digitally filtered with a low-pass filter of 15 Hz (24 dB down). MMN waveforms were obtained by subtracting waveforms elicited by standards from those elicited by frequency- or duration-deviant stimuli. Frequency MMN amplitude was defined as the peak negativity within a 100- to 225-ms latency window, and duration MMN amplitude was defined as the peak negativity within the 200–300 ms range.
Statistical analyses
Data were analysed using SPSS (version 10.0). Comparisons of PANSS scores (PANSS total scores, Positive Symptom Scale scores, Negative Symptom Scale scores and Total Psychopathology Scale scores) and mean amplitudes and latencies of frequency and duration MMN between neuroleptic-naive and neuroleptic-free patients were done using independent-sample t-tests. Comparisons of amplitudes of MMN between healthy controls and patients were done using paired-sample t-tests. PANSS scores were analysed by one-way repeated measure analysis of variances (ANOVA) with session (baseline and 4- and 8-week treatments) as a within-subject factor. The effects of quetiapine treatment on amplitudes of MMN were analysed by repeated measure ANOVA with session (baseline and 4- and 8-week treatments) and MMN type (frequency versus duration) as within-subject factors. Least square difference (LSD) tests were performed as post hoc analyses if indicated. Correlation coefficients between MMN and PANSS scores were calculated by the Pearson test. Alpha values of 0.05 were considered significant throughout.
Results
Comparisons of PANSS, MMN between neuroleptic-naive and neuroleptic-free patients
No differences in PANSS, mean amplitudes and latencies of frequency and duration MMN at baseline and after 4- and 8-week treatments between neuroleptic-naive and neuroleptic-free patients were observed (p> 0.05) (Tables 2 and 3).
Table 2 PANSS scores [presented as mean (standard deviation)] before and after quetiapine treatment of neuroleptic-naive (naive = 17) and neuroleptic-free patients (free, n = 6)
Table 3 Amplitudes and latencies of MMN [presented as mean (standard deviation)] of neuroleptic-naive (naive, n = 17) and neuroleptic-free patients (free, n = 6) at baseline and 4- and 8-week treatments
Comparisons of PANSS before and after quetiapine treatment
PANSS scores were analysed by one-way repeated measure ANOVA with session (baseline and 4- and 8-week treatments) as a within-subject factor revealing significant session main effect for PANSS scores (for PANSS total scores: F = 255, df = 2, p = 0.000; for Positive Symptom Scale scores: F = 155, df = 2, p = 0.000; for Negative Symptom Scale scores: F = 50, df = 2, p = 0.000; for Total Psychopathology Scale scores: F = 20, df = 2, p = 0.000). Quetiapine decreased all PANSS, Total Psychopathology, Positive Symptom and Negative Symptom Scale scores. There was no significant correlation between changes in amplitudes of frequency and duration MMN and changes in PANSS scores at baseline and follow-up periods (Table 4).
Table 4 PANSS scores [presented as mean (standard deviation)] before and after quetiapine treatment
* p < 0.01.
Comparison at baseline of patients and healthy controls
Patients showed smaller mean amplitudes of frequency and duration MMN at baseline than did controls (in frequency MMN, t = 2.067, p = 0.045; in duration MMN, t = 2.263, p = 0.029; df for all electrodes = 22). The mean amplitudes of frequency and duration MMN were reduced in patients at 8-week treatments compared to controls however, did not reach significance (in frequency MMN, t = 1.649, p = 0.113; in duration MMN, t = 1.861, p = 0.091; df for all electrodes = 22). No differences in latencies between patients and controls were observed (Table 5; Fig. 1).
Fig. 1 Grand averages of frequency and duration MMN (MMN waveforms were obtained by subtracting waveforms elicited by standards from those elicited by frequency- or duration-deviant stimuli) in patients at baseline and controls. RM, right mastoid; LM, left mastoid.
Table 5 Amplitudes and latencies of MMN [presented as mean (standard deviation)] of healthy controls (n = 23) and patients (n = 23) at baseline and 4- and 8-week treatments
* Patients, after 8-week treatments versus baseline and after 4-week treatments, p< 0.05.
Effects of quetiapine treatment
A repeated measure ANOVA with session (baseline and 4- and 8-week treatments) and MMN type (frequency versus duration) as within-subject factors revealed no significant MMN type or MMN type × session main effect for MMN amplitudes (for MMN type: F = 0.704, df = 1, p = 0.403; for MMN type × session: F = 0.299, df = 2, p = 0.796). Session main effect, however, was significant (F = 3.576, df = 2, p = 0.031). LSD tests were performed as post hoc analyses and showed significant differences between MMN amplitudes at 8 weeks and those at both baseline (p = 0.025) and 4 weeks (p = 0.020). However, the test did not show a significant difference between MMN amplitudes at 4 weeks and those at baseline (p = 0.935). MMN amplitudes at 8 weeks were higher than those at 4 weeks and those at baseline (Table 5; Fig. 2).
Fig. 2 Grand averages of frequency and duration MMN (MMN waveforms were obtained by subtracting waveforms elicited by standards from those elicited by frequency- or duration-deviant stimuli) in patients at baseline and 4- and 8-week treatments of quetiapine.
Discussion
This study is the first to use electrophysiological indices of automatic auditory information processing, i.e. MMN, to assess cognitive improvement in patients with schizophrenia treated by atypical neuroleptic quetiapine. Our trial results authenticate previous hypothesis that treatment with quetiapine leads to the improvement of MMN in schizophrenia. The atypical neuroleptic quetiapine not only is an effective 5HT2a and D2 receptor antagonist but also is a 5HT1a receptor partial agonist, increasing dopamine and acetylcholine release in the prefrontal cortex Reference Sharma(6). Because MMN amplitude is highest in frontal channels (Reference Sato, Yabe and Hiruma29–Reference Dittmann, Juptner and Jentzen31), MMN amplitudes at Fz electrodes present the status of cognitive function. Our study showed that quetiapine improved the amplitudes of MMN after 8-week treatments, which proves that quetiapine has an effect on passive attention in patients with schizophrenia. From a neuroelectrophysical standpoint, MMN offers objective evidence that treatment with the quetiapine ameliorates preattentive deficits in schizophrenia.
A previous study showed reduced MMN in stable chronic patients with schizophrenia and deduced that chronic patients represent a more homogeneous sample concerning the genetics of MMN deficits Reference Daniel, Umbricht and Bates(12). Another study displayed that MMN amplitude was reduced in patients with schizophrenia and relatives compared with controls, and there were no significant differences between patients and relatives; therefore, the results suggest that reduced MMN amplitude may be an endophenotype marker of the predisposition to schizophrenia Reference Michie, Innes-Brown and Todd(32). The above two studies support that MMN deficits may represent a trait marker. However, a study reported that the MMN has no significant familial influence and is normal among the unaffected relatives. The researchers concluded that although the MMN is abnormal in patients with schizophrenia, it is a weak or unreliable marker of vulnerability when applied to subclinical populations. Therefore, it is unlikely to be an endophenotype for the disorder Reference Elvira, Croft and McDonald(33). In our study, the mean amplitudes of frequency and duration MMN were reduced in patients at 8-week treatments compared to controls however, did not reach significance, which deduces that MMN abnormalities are state dependent. However, because of the small sample, our results have to be considered preliminary. As a matter of fact, whether MMN deficits represent a trait marker or state marker remains controversial. The probable reason for inconsistencies is that different stimulus condition designs were used and insufficient numbers of subjects were recruited for these studies.
Neurocognitive impairment is now recognised as a fundamental symptom of schizophrenia Reference Saykin, Gur and Gur(34,Reference Gold and Harvey35). Improvement in cognitive function is increasingly recognised as an important goal of therapy. To some extent, the degree of neurocognitive impairment has been shown to be a much stronger predictor of community functioning than either positive or negative symptom severity Reference Velligan, Mahurin and Diamond(36). In this trial, quetiapine had no significant effects on amplitudes of MMN after 4-week treatments. It might be deduced that improvement in cognitive function during treatment with quetiapine in schizophrenia relies on a sufficient treatment period. Previous studies failed to show the effects of atypical antipsychotics such as clozapine, risperidone and olanzapine on MMN after a 4-week treatment (Reference Umbricht, Javitt and Novak37–Reference Korostenskaja, Dapsys and Siurkute39). The probable reason for this failure is that the treatment period was insufficient.
In the past, MMN for both duration and frequency deviants was investigated in patients with schizophrenia, and the results showed that patients with schizophrenia showed significantly smaller mean MMN than did healthy control subjects Reference Umbricht, Koller and Schmid(22). Consistent with previous research, the present study showed that MMN amplitudes were significantly reduced in the patient group, which showed that MMN might be an abnormal index of preattentive automatic auditory information processing.
MMN is considered an index of the auditory sensory memory because MMN generation depends on the context in which a specific stimulus is presented. Namely, it is only generated when a memory trace of the regularity or invariance has been established against which any incoming stimulus is being compared. MMN provides a measure of cognitive deficits associated with schizophrenia Reference Naatanen, Paavilainen and Alho(40), being the first measurable brain response component that differentiates between frequent and deviant auditory stimuli and reflects the properties of an automatic, memory-based, comparison process.
We observed no significant correlation between changes in amplitudes of frequency and duration MMN and changes in PANSS scores at baseline and follow-up periods. The probable reason is that PANSS scores mainly reflect obvious symptoms, such as both positive and negative symptoms. Furthermore, few of the cognitive symptoms of automatic auditory information processing were reflected. Our results show that the MMN amplitudes at Fz reflect the degree of neurocognitive impairment. MMN amplitude improvement may be a possible biomarker of treatment efficacy. The improvement of this functional marker may indicate an important pathway towards new therapeutic strategies that target cognitive dysfunction in schizophrenia. It is important that clinicians understand the benefits and limitations of modern neuroimaging techniques and are also suitably equipped to appraise future developments Reference Malhi and Lagopoulos(41). The use of MMN in evaluating psychopathology and therapeutic effects is helpful in the clinical management of schizophrenic patients. Therefore, it is necessary to validate this study effect using similar parameters in future studies.
In this study, during presentation of the auditory test paradigm, the attention of the subjects was diverted from the tones, and the results may exclude a practice effect. However, the limitation of this study is the failure to retest the control group at 4 and 8 weeks for further checking the validity.
In conclusion, MMN offers the ability to map the neuroelectrophysical response to treatment intervention within the brain and may enable improved understanding of the neuropharmacology of schizophrenia. Our study showed that quetiapine was useful for the treatment of schizophrenic patients, especially on improvement of abnormalities of early auditory processing, as assessed by MMN, a non-invasive technique with good temporal resolution.
Acknowledgements
This study was supported by the Key Technologies Research and Development Program of Nanjing Medical University, China (no. 20060587) .We thank Professor Helen Arnott for the English corrections.
