مدولاسیون تحت قشری از کنترل توجه توسط داروهای ضد روان پریشی نسل دوم در روان پریشی اپیزود اول

Abstract Psychotic disorders are characterized by significant deficits in attentional control, but the neurobiological mechanisms underlying these deficits early in the course of illness prior to extensive pharmacotherapy are not well understood. Moreover, little is known regarding the symptom and brain changes associated with amelioration of attentional impairments through antipsychotic pharmacotherapy. In this study 14 male patients experiencing a first-episode of psychosis with minimal prior antipsychotic treatment completed an attentional control task while undergoing functional magnetic resonance imaging at the onset of treatment with a second generation antipsychotic (risperidone or aripiprazole) in a double blind randomized clinical trial and then again following approximately 12 weeks of treatment. In addition, 14 age-, and performance-matched healthy male volunteers who were not treated completed the same task at a baseline timepoint and then again following 12 weeks. Patients showed significantly greater activation than healthy volunteers in the right globus pallidus, left thalamus, and right thalamus at the time of the baseline scan. Among patients there was a significant reduction in right globus pallidus blood-oxygen level dependent (BOLD) response following antipsychotic treatment that correlated significantly with improvement in response accuracy and reductions in thought disturbance. No changes in globus pallidus activation were observed in healthy volunteers over this time period. These preliminary findings suggest that improvement in attentional control and concomitant reductions in thought disturbance in first-episode psychosis may be associated with reductions in subcortical activity following administration of second-generation antipsychotics early in the course of illness. These findings have implications for understanding how changes in basal ganglia activity may be linked to improvements in attentional control through antipsychotics.

1. Introduction Given their strong dopaminergic inneveration, the basal ganglia have strong relevance to neurobiological models of schizophrenia (Perez-Costas et al., 2010) and may play a role in successful treatment intervention (Molina et al., 2011). Furthermore, both mitochondrial and receptor density abnormalities within the basal ganglia have been reported in schizophrenia (Kung and Roberts, 1999, Meisenzahl et al., 2007 and Murray et al., 1995), further supporting their relationship to mechanisms underlying antipsychotic medications. Magnetic resonance (MR) imaging studies have demonstrated structural alterations involving the caudate nucleus, globus pallidus and putamen in patients with schizophrenia (Bilder, 1992, Brandt and Bonelli, 2008, Corson et al., 1999 and Hokama et al., 1995) and unaffected siblings (Mamah et al., 2008). Several literature reviews suggest that administration of typical (in contrast to atypical) antipsychotics may be associated with volumetric increases in the basal ganglia (Navari and Dazzan, 2009 and Smieskova et al., 2009). Thus, examination of regions comprising the basal ganglia early in the course of illness and prior to extensive pharmaocologic intervention may best address questions regarding their role in the neurobiology of schizophrenia and how antipsychotics may mediate associated changes in neuropsychological functioning. Attentional control is the ability for an individual to decide what should be acknowledged in the environment compared to what can be ignored. Deficits in attentional control are considered one of the hallmark features of schizophrenia and related psychotic disorders (Carter et al., 1992, Lalanne et al., 2012 and Reilly et al., 2008). Successful attentional control (or executive attention) is related to the ability to conduct top-down processing and is mainly under the individual's control (in contrast to bottom-up processing). Aside from frontal regions, empirical work indicates that attentional control is mediated by structures comprising the basal ganglia, including the globus pallidus (Bočková et al., 2011, Muir et al., 1993 and Scott et al., 2002), caudate nucleus (Canavero and Fontanella, 1998) and putamen (Max et al., 2002). Moreover, hyperactivation in basal ganglia regions has been reported in rats during attention related tasks (Sotoyama et al., 2011) and lesions to the basal ganglia yield attention deficits both in rats (Muir et al., 1993 and Thompson et al., 1985) and humans (Max et al., 2002 and Scott et al., 2002). Also, involvement of the globus pallidus and caudate nucleus in attentional control has been demonstrated using electrophysiological recording in Parkinson's disease (Bočková et al., 2011 and Kropotov et al., 1997) as well as in healthy human positron emission tomography studies (Corbetta et al., 1991). There is increasing data that basal ganglia dysfunction may be associated with attentional control deficits in schizophrenia (Manoach et al., 2000). For example, measures of attention/vigilance have been linked to volumetric alterations within the caudate nucleus and putamen (Mamah et al., 2008). Additional data suggest that attentional control in schizophrenia may improve with antipsychotics (McGurk et al., 2004) and that this could be related to their significant D2 dopaminergic antagonism efficacy. In this regard Cohen et al. (1998) reported that the basal ganglia play a role in sustained attention, likely contribute to psychotic symptoms and mediate antipsychotic response. More specifically, these authors reported that patients had greater regional cerebral metabolic rates in the posterior putamen during an attention task, which was associated with worse antipsychotic treatment response. Few functional magnetic resonance imaging (fMRI) studies have assessed the potential impact of antipsychotic medications on attentional control early in the course of schizophrenia and prior to extensive pharmacotherapy. In one study Keedy et al. (2009) reported less dorsal striatal activation using fMRI in patients compared to healthy controls during visual tracking and attention following antipsychotic treatment, which the authors interpreted as a possible adverse effect of treatment that could relate to dopamine blockade. To clarify the role of second-generation antipsychotics in mediating attentional control in patients with psychotic disorders, we conducted a longitudinal fMRI study examining the relationship between antipsychotic pharmacotherapy on basal ganglia activity during performance of the Multisource Interference Test (Bush and Shin, 2006) and its relationship to clinical improvement. We hypothesized that patients would demonstrate functional abnormalities in the basal ganglia at the onset of treatment when performing an attentional task in line with the rich dopaminergic innervations of the basal ganglia (Hall et al., 1994 and Richtand et al., 2007), and that there would be significant changes in BOLD activity following antipsychotic treatment. An additional study goal was to investigate whether changes in BOLD response following pharmacotherapy were associated with symptom improvement.

3. Results 3.1. Clinical improvement There was significant improvement after 12 weeks in overall BPRS score (mean change=−19.14, range=−3 to −38, t=6.39, d.f.=22.89, p<0.001). Among the four clinical domains, thought disturbance (mean change=−6.14, range=−1 to −11, t=7.33, d.f.=22.75, p<0.001), hostility-suspiciousness (mean change=−4.57, range=0 to −9, t=6.11, d.f.=25.28, p<0.001) and anxiety-depression (mean change=−4.07, range=0 to −10, t=3.42, d.f.=18.80, p=0.029) showed significant improvement ( Table 2). Table 2. Brief psychiatric rating scale clinical evaluation. Total Thought disturbance Hostility-suspiciousness Anxiety-depression Baseline (S.D.) 51.86 11.00 9.71 9.50 10.10 2.60 4.01 2.18 12 week (S.D.) 31.00 4.86 5.64 4.86 6.86 1.75 1.95 1.83 Table options 3.2. Behavioral results As control subjects were selected based on their baseline accuracy as well as their age and gender there was no significant difference between groups in baseline accuracy. There was no significant group difference in baseline reaction time or followup reaction time. There was, however, a significant difference between the two groups at the time of the follow up scan for response accuracy (t=2.60, d.f.=20.88, p=0.016; Fig. 1A) with healthy volunteers performing better than patients. Among patients there was an improvement in response accuracy (t=2.35, d.f.=13, p=0.035; Fig. 1A), but not in reaction time ( Fig. 1B) following 12 weeks of antipsychotic treatment ( Table 3). Healthy volunteers demonstrated an improvement in response accuracy (t=3.55, d.f.=13, p=0.0036, Fig. 1A) and reaction time (t=3.32, d.f.=13, p=0.0056, Fig. 1B) across the 2 timepoints. Accuracy and Reaction Time (with standard error bars) in Patients and Healthy ... Fig. 1. Accuracy and Reaction Time (with standard error bars) in Patients and Healthy Volunteers at Baseline and 12 Week Followup Timepoints. Note: Asterisks denote the following: blue=significant change in controls, red=significant increase in patients, black=significant difference between patients and controls at 12 week scan. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Figure options Table 3. Behavioral results. Accuracy (% correct) Reaction time (ms) Patients Baseline 82.79 ⁎ t(13)=2.35 p=0.035 882.10 t(13)=1.47 p=0.16 12 Week 89.14 825.71 Controls Baseline 88.57 ⁎⁎ t(13)=3.55 p=0.0036 864.52 ⁎⁎ t(13)=3.32 p=0.0056 12 Week 94.64 703.32 ⁎p<0.05. ⁎⁎p<0.01. Table options 3.3. fMRI results At the baseline scan, patients showed significantly greater activation than healthy volunteers in the basal ganglia (Fig. 2A), including the right globus pallidus (t=3.24, d.f.=25.30, p=0.0034; Fig. 2B), left thalamus (t=2.65, d.f.=25.96, p=0.014; Fig. 2C), and right thalamus (t=3.56, d.f.=25.98, p=0.0015; Fig. 2D). Baseline activation at Z=0 slice; (A) Red/Yellow: patients higher than healthy ... Fig. 2. Baseline activation at Z=0 slice; (A) Red/Yellow: patients higher than healthy volunteers; Blue: patients lower than healthy volunteers; (B) Right Globus Pallidus ROI; (C) Right Thalamus; (D) Left Thalamus. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Figure options The whole brain voxelwise revealed significant changes in patients from the baseline to the 12 week scans. Patients demonstrated significantly decreased activation in the globus pallidus, putamen, and thalamus, as well as several other cortical regions (Fig. 3 and Table 4). Healthy volunteers did not show significant decreases in any of these regions (Fig. 3). Regions demonstrating Significant Changes in BOLD Activation in Patients and ... Fig. 3. Regions demonstrating Significant Changes in BOLD Activation in Patients and Healthy Volunteers. Figure options Table 4. Regions demonstrating 12 week changes in BOLD activation in patients. Number of voxels P Z-MAX MNI Region Z-MAX X (mm) Z-MAX Y (mm) Z-MAX Z (mm) Decrease in 12 weeks 48,996 3.64E-15 7.88 27 0 57 Superior Frontal Gyrus 12,049 1.37E-06 7.03 45 −31 51 Post Central Gyrus 9078 1.26E-05 6.21 −12 −18 1 Left Thalamus 5.74 11 −20 1 Right Thalamus 4.83 26 −6 0 Right Globus Pallidus 4.79 1 −27 −12 Brain Stem 4.76 −4 −28 −12 Brain Stem 4.54 15 −12 9 Right Thalamus 8544 0.0000191 5.91 50 2 35 Right Precentral Gyrus 7245 0.000055 5.95 15 −52 −19 Cerebellum 6939 0.0000712 6.7 36 35 24 Right Frontal Pole 6486 0.000105 6.13 −53 0 −8 Left Central Opercular Cortex 1205 0.0309 5.53 −36 −53 −31 Cerebellum Increase in 12 weeks 2887 0.00356 5.32 −49 −63 25 Left Lateral Occipital Cortex 1932 0.0113 5.57 −10 27 59 Left Superior Frontal 1729 0.0147 5.05 −14 59 28 Left Frontal Pole Table options Among subcortical ROIs, the change in right globus pallidus activation was significantly correlated with response accuracy improvement (r=−0.50, d.f.=14, p=0.047) in patients. The reduction in right globus pallidus activation was also significantly correlated with decreases in the BPRS thought disturbance factor (r=0.54, d.f.=11, p=0.032, Fig. 4B). Specifically, the post-hoc analysis revealed that the reduction in right pallidal activity was significantly correlated with the unusual thought content item (r=0.52, d.f.=11, p=0.042, Fig. 4C). No other BPRS composite or individual scores were significantly correlated with reductions in right pallidal activity. There was a significant correlation between improvements in response accuracy and decreases in thought disturbance (r=−0.64, d.f.=12, p=0.013, Fig. 4D). (A) Contrast Z score in the Right Globus Pallidus. (B) Association between ... Fig. 4. (A) Contrast Z score in the Right Globus Pallidus. (B) Association between baseline to 12 week change in the BPRS Thought Disturbance factor (where negative values indicate improvement) and baseline to 12 week change in contrast Z score (where negative values indicate reduced activation) in the Right Globus Pallidus. (C) Association between baseline to 12 week improvement in the BPRS Unusual Thought Content item and baseline to 12 week change in contrast Z score in the Right Globus Pallidus. (D) Association between baseline to 12 week change in the BPRS Thought Disturbance factor and baseline to 12 week change in response accuracy in the Interference condition. Figure options The causal steps approach showed a trend towards significance for the fully mediated model in which the effect of right globus pallidus activation on thought disturbance was mediated by accuracy in attention. Using the production-of-coefficients approach, the mediation effect size was calculated to be 1.58. A post-hoc voxelwise analysis did not detect any differences between two medications. Moreover, an ROI analysis of the right globus pallidus comparing the two groups was not statistically significant (t=0.43, d.f.=11.64, p=0.67).