**2. Schizophrenia**

Functional neuroimaging has been used to elucidate patterns of increased or decreased activity within the brains of schizophrenic and normal subjects during rest and various assigned tasks, revealing that the affected parts of the central nervous system are not contained within a single brain region, but rather lie within neural networks over several brain regions. Numerous structural brain researches studies employing CT and MRI have demonstrated significant volume reductions in key brain regions such as the lateral prefrontal cortex, anterior cingulate cortex (ACC), superior temporal cortex, hippocampus/parahippocampus, striatum and thalamus in patients with schizophrenia relative to normal subjects (Shenton et al., 2001). In support of these structural alterations, functional neuroimaging studies have produced representations of abnormalities in and across these regions. Taking these results together, a variety of symptoms, including

Resting State Blood Flow and Glucose Metabolism in Psychiatric Disorders 131

accompanied by other areas with hyperperfusion/hypermetabolism within the frontal cortex (Andreasen et al., 1997; Kim et al., 2000). The measurement conditions used under rest or the performance of a given task should also be taken into consideration. Whereas most of the studies with SPECT have been conducted under a resting state, many studies using FDG-PET have performed the comparison under a cognitive task such as continuous performance task (CPT) or California verbal learning task (CVLT). This is because of the possibility that a spontaneous fluctuation of mental state under a resting condition during scanning could result in varied distribution of rGMR in the participant group as a whole. Indeed, several PET studies using CPT (Potkin et al., 2002; Molina et al., 2005a, 2005b, 2009) or a visual attention task (Lehrer et al., 2005) showed a significant reduction of rGMR in the prefrontal cortex in patients compared to normal controls, very similar to the results obtained in almost all studies under a resting state. Then, reduction of rCBF/rGMR in the prefrontal cortex in patients relative to normal controls under a static state during the performance of cognitive tasks and under a resting state collectively

Although earlier studies have dealt this issue with dichotomous problem; presence or absence of hypofrontality, afterward, improvements in research design and analytic methods provide more detailed information such as distributed patterns within the frontal lobe within patients' brains or the degree of difference of the finding between patients and controls. In this context, in some meta-analysis studies (Davidson and Heinrichs, 2003; Hill et al., 2004), the finding of hypofrontality has been supported and thus established as a more

The hypoperfusion and hypometabolism in the frontal lobe have been presumed to be closely linked with negative symptoms and cognitive impairments in schizophrenia. These notions were demonstrated by the negative relationship between negative symptoms and blood flow/metabolism (Liddle et al., 1992; Wolkin et al., 1992; Ebmeier et al., 1993; Schröder et al., 1996; Andreasen et al., 1997; Erkwoh et al., 1997; Sabri et al., 1997; Ashton et al., 2000) and the significant reductions of blood flow/metabolism in the patients group with profound negative symptoms (Potkin et al., 2002; Gonul et al., 2003a), although several negative studies have existed (Vita et al., 1995; Min et al., 1999). On the other hand, whereas the cognitive dysfunctions that have recently received so much attention are closely related with negative symptoms, the reports exploring the relationship between the impairments and at rest blood flow/metabolism are very restricted (Penadés et al., 2002; Molina et al., 2009). A hypodopaminergic state in the prefrontal cortex is presumed to underlie the negative symptoms and cognitive impairments (Lynch, 1992; Remington et al., 2011) and thus, in this context, it is noted that hypofrontality strongly suggests an important part of

As for brain regions other than the frontal lobe, a number of previous studies have demonstrated substantial variations between the patients with schizophrenia and normal controls, with some reports observing increases in various activities and other reports

Both the lateral and medial phases in the temporal cortex have been closely related with positive symptoms, particularly hallucination and delusion. Based on accumulating evidence from fMRI studies, for example, the primary auditory cortex located in the

indicates hypofrontality.

convictive finding in the disease.

core pathophysiology in schizophrenia.

**2.2 rCBF/rGMR patterns in key regions other than the frontal lobe** 

documenting decreases, and thus no convincing consensus has been reached.

hallucination/delusion and negative symptoms, have been attributed not to abnormalities in a single brain region but to abnormalities in a distributed network of spatially distinct regions. Furthermore, functional neuroimaging studies have demonstrated that antipsychotics have substantial effects on brain functions, and have helped to elucidate the differences in action mechanisms among them.

### **2.1 Hypofrontality and negative symptoms in schizophrenia**

Ingvar and Franzen (1974) reported that patients with chronic schizophrenia showed significant reduction in the rCBF ratio of the frontal to occipital region compared to normal subjects and subjects with first-episode schizophrenia measured with 133Xe. This was the first study to report an abnormality in rCBF in schizophrenia. Following this work, several other studies examined the resting state blood flow and metabolism (Buchsbaum et al., 1982; Wolkin et al., 1985; Tamminga et al., 1992; Sachdev et al., 1997) and repeatedly reported significant decreases in patients with schizophrenia relative to normal participants. On the other hand, there have been studies showing no difference in this parameter between patients and normal controls (Gur et al., 1995; Sabri et al., 1997, Scottish Schizophrenia Research Group, 1998), or even an increase in rCBF/rGMR in patients compared to normal controls (Cleghorn et al., 1989; Ebmeier et al., 1993).

Early studies on this issue have presented very disparate results with respect to not only the presence or absence of hypoperfusion/hypomtabolism, but also, in cases in which it was present, the degree, relevant regions and correlation with symptoms of hypoperfusion/hypometabolism. The reason for these differences is presumed to be the large number of confounding factors, such as disease heterogeneity, treatment with antipsychotics, imcompleteness of results derived from the ROI method, measured value of absolute or relative data, different reference regions for relative data, measurement conditions under varied physiological states, and so on. Therefore, additional explorations with a more sophisticated study design for the drug-naïve subjects group, the same scanning conditions and reliable analytic methods are needed to reach a definitive conclusion on this issue.

As for the effects of antipsychotic medications, several studies on drug-naïve patients with first-episode schizophrenia demonstrated a significant reduction in blood flow and metabolism in the frontal cortex relative to age-matched normal controls under a resting condition (Buchsbaum et al., 1992a; Steinberg et al., 1995; Vita et al., 1995; Erkwoh et al., 1997) and task-related activation (Andreasen et al., 1992, 1997; Ashton et al., 2000) and suggested that the abnormal reduction in the prefrontal region occurs from a very early stage of the disease. With respect to the problem of analytic methods, ROI methods have been a mainstream from the 80s to late 90s, but voxel-wise methods representative of Statistic Parametric Mapping (SPM) have prevailed from the mid-90s and are the standard modality at present. This voxel-wise methods have successfully addressed two important problems in brain analyses: individual structural differences between the brains of participants and examiners' arbitress on target brain regions depending on *a priori* hypothesis. Numerical researches based on these methods have demonstrated a significant reduction in particularly the lateral, medial and orbital phases of the prefrontal cortex relative to normal controls (Andreasen et al., 1997; Ashton et al., 2000; Kim et al., 2000; Potkin et al., 2002; Lehrer et al., 2005; Molina et al., 2005a, 2005b, 2009), and these findings have shown that areas with hypoperfusion and hypometabolism were pervasive and further

hallucination/delusion and negative symptoms, have been attributed not to abnormalities in a single brain region but to abnormalities in a distributed network of spatially distinct regions. Furthermore, functional neuroimaging studies have demonstrated that antipsychotics have substantial effects on brain functions, and have helped to elucidate the

Ingvar and Franzen (1974) reported that patients with chronic schizophrenia showed significant reduction in the rCBF ratio of the frontal to occipital region compared to normal subjects and subjects with first-episode schizophrenia measured with 133Xe. This was the first study to report an abnormality in rCBF in schizophrenia. Following this work, several other studies examined the resting state blood flow and metabolism (Buchsbaum et al., 1982; Wolkin et al., 1985; Tamminga et al., 1992; Sachdev et al., 1997) and repeatedly reported significant decreases in patients with schizophrenia relative to normal participants. On the other hand, there have been studies showing no difference in this parameter between patients and normal controls (Gur et al., 1995; Sabri et al., 1997, Scottish Schizophrenia Research Group, 1998), or even an increase in rCBF/rGMR in patients compared to normal

Early studies on this issue have presented very disparate results with respect to not only the presence or absence of hypoperfusion/hypomtabolism, but also, in cases in which it was present, the degree, relevant regions and correlation with symptoms of hypoperfusion/hypometabolism. The reason for these differences is presumed to be the large number of confounding factors, such as disease heterogeneity, treatment with antipsychotics, imcompleteness of results derived from the ROI method, measured value of absolute or relative data, different reference regions for relative data, measurement conditions under varied physiological states, and so on. Therefore, additional explorations with a more sophisticated study design for the drug-naïve subjects group, the same scanning conditions and reliable analytic methods are needed to reach a definitive

As for the effects of antipsychotic medications, several studies on drug-naïve patients with first-episode schizophrenia demonstrated a significant reduction in blood flow and metabolism in the frontal cortex relative to age-matched normal controls under a resting condition (Buchsbaum et al., 1992a; Steinberg et al., 1995; Vita et al., 1995; Erkwoh et al., 1997) and task-related activation (Andreasen et al., 1992, 1997; Ashton et al., 2000) and suggested that the abnormal reduction in the prefrontal region occurs from a very early stage of the disease. With respect to the problem of analytic methods, ROI methods have been a mainstream from the 80s to late 90s, but voxel-wise methods representative of Statistic Parametric Mapping (SPM) have prevailed from the mid-90s and are the standard modality at present. This voxel-wise methods have successfully addressed two important problems in brain analyses: individual structural differences between the brains of participants and examiners' arbitress on target brain regions depending on *a priori* hypothesis. Numerical researches based on these methods have demonstrated a significant reduction in particularly the lateral, medial and orbital phases of the prefrontal cortex relative to normal controls (Andreasen et al., 1997; Ashton et al., 2000; Kim et al., 2000; Potkin et al., 2002; Lehrer et al., 2005; Molina et al., 2005a, 2005b, 2009), and these findings have shown that areas with hypoperfusion and hypometabolism were pervasive and further

differences in action mechanisms among them.

controls (Cleghorn et al., 1989; Ebmeier et al., 1993).

conclusion on this issue.

**2.1 Hypofrontality and negative symptoms in schizophrenia** 

accompanied by other areas with hyperperfusion/hypermetabolism within the frontal cortex (Andreasen et al., 1997; Kim et al., 2000). The measurement conditions used under rest or the performance of a given task should also be taken into consideration. Whereas most of the studies with SPECT have been conducted under a resting state, many studies using FDG-PET have performed the comparison under a cognitive task such as continuous performance task (CPT) or California verbal learning task (CVLT). This is because of the possibility that a spontaneous fluctuation of mental state under a resting condition during scanning could result in varied distribution of rGMR in the participant group as a whole. Indeed, several PET studies using CPT (Potkin et al., 2002; Molina et al., 2005a, 2005b, 2009) or a visual attention task (Lehrer et al., 2005) showed a significant reduction of rGMR in the prefrontal cortex in patients compared to normal controls, very similar to the results obtained in almost all studies under a resting state. Then, reduction of rCBF/rGMR in the prefrontal cortex in patients relative to normal controls under a static state during the performance of cognitive tasks and under a resting state collectively indicates hypofrontality.

Although earlier studies have dealt this issue with dichotomous problem; presence or absence of hypofrontality, afterward, improvements in research design and analytic methods provide more detailed information such as distributed patterns within the frontal lobe within patients' brains or the degree of difference of the finding between patients and controls. In this context, in some meta-analysis studies (Davidson and Heinrichs, 2003; Hill et al., 2004), the finding of hypofrontality has been supported and thus established as a more convictive finding in the disease.

The hypoperfusion and hypometabolism in the frontal lobe have been presumed to be closely linked with negative symptoms and cognitive impairments in schizophrenia. These notions were demonstrated by the negative relationship between negative symptoms and blood flow/metabolism (Liddle et al., 1992; Wolkin et al., 1992; Ebmeier et al., 1993; Schröder et al., 1996; Andreasen et al., 1997; Erkwoh et al., 1997; Sabri et al., 1997; Ashton et al., 2000) and the significant reductions of blood flow/metabolism in the patients group with profound negative symptoms (Potkin et al., 2002; Gonul et al., 2003a), although several negative studies have existed (Vita et al., 1995; Min et al., 1999). On the other hand, whereas the cognitive dysfunctions that have recently received so much attention are closely related with negative symptoms, the reports exploring the relationship between the impairments and at rest blood flow/metabolism are very restricted (Penadés et al., 2002; Molina et al., 2009). A hypodopaminergic state in the prefrontal cortex is presumed to underlie the negative symptoms and cognitive impairments (Lynch, 1992; Remington et al., 2011) and thus, in this context, it is noted that hypofrontality strongly suggests an important part of core pathophysiology in schizophrenia.

#### **2.2 rCBF/rGMR patterns in key regions other than the frontal lobe**

As for brain regions other than the frontal lobe, a number of previous studies have demonstrated substantial variations between the patients with schizophrenia and normal controls, with some reports observing increases in various activities and other reports documenting decreases, and thus no convincing consensus has been reached.

Both the lateral and medial phases in the temporal cortex have been closely related with positive symptoms, particularly hallucination and delusion. Based on accumulating evidence from fMRI studies, for example, the primary auditory cortex located in the

Resting State Blood Flow and Glucose Metabolism in Psychiatric Disorders 133

A number of previous studies have shown that typical neuroleptics such as haloperidol reduce blood flow and metabolism in the frontal lobe. These effects were repeatedly replicated in studies of both acute (Bartlett et al., 1998; Lahti et al., 2005) and chronic administration (Bartlett et al., 1991; Buchsbaum et al., 1992b; Miller et al., 1997, 2001; Lahti et al., 2003). Further, whereas haloperidol was reported to be related with hypoperfusion and hypometabolism in the hippocampus in terms of amelioration of positive symptoms (Lahti et al., 2003), increases of rCBF/rGMR in the motor cortex induced by haloperidol were presumed to be related with extra-pyramidal symptoms (Molina et al., 2003; Buchsbaum et al., 2007), and the decrease in activity in the occipital cortex following haloperidol treatment might be related with sedative effects (Bartlett et al., 1991; Desco et

An increase in rCBF/rGMR in the basal ganglia in patients with schizophrenia by neuroleptics, in particular haloperidol, is the most consistent finding among numerous reports on antipsychotics. This has been replicated very well in the acute effect (Lahti et al., 2005) as well as the chronic effect (Buchsbaum et al., 1987, 1992a, 2007a; Miller et al., 1997, 2001; Scottish Schizophrenia Research Group, 1998; Corson et al., 2002; Desco et al., 2003; Lahti et al., 2003). The increase of blood flow and metabolism in this area is presumed to be due to increases of activity in the post synapses through upregulation of DA D2 receptors induced by a potent blocking action of the receptor by haloperidol (Miller et al., 1997; Corson et al., 2002). This notion is in line with the increase of volume in this area following

Studies on the effects of atypical antipsychotics on brain perfusion/metabolism have become to be examined based on more detailed neuronal substrates than studies on typical antipsychotics by appearance of voxel wise analysis. Although risperidone has less effect on the reduction of blood flow in the frontal lobe than haloperidol (Miller et al., 2001), the drug induces a significant reduction in the prefrontal cortex relative to baseline (Berman et al., 1996; Liddle et al., 2000; Ngan et al., 2002; Molina et al., 2008). In the basal ganglia, the degree of increase in blood flow/metabolism by risperidone is likely smaller than that by haloperidol (Liddle et al., 2000; Miller et al., 2001). Liddle et al. (2000) demonstrated that treatment with risperidone for 6 weeks showed a significant positive relation between decrease in the hippocampus and decrease in reality distortion, suggesting that the

Olanzapine is likely that its effect of blood flow/metabolism in the frontal lobe is lesser than that by risperidone (Gonul et al., 2003b; Molina et al., 2005c; Buchsbaum et al., 2007b). Clozapine, the gold standard among the atypical neuroleptics, has a pharmacological profile with weaker blockade of DA D2 receptors and broader actions for multiple receptors than other atypical antipsychotics, and these characteristics are presumed to be related to its superior clinical efficacy relative to other neuroleptics. Interestingly, several previous studies have reported that clozapine induced a significant reduction in blood flow/metabolism in the prefrontal cortex (Potkin et al., 1994, 2003; Cohen et al., 1997; Lahti et al., 2003; Molina et al., 2005d, 2008). On the other hand, increases in several parts of the prefrontal cortex, including the ACC (Lahti et al., 2003) and decreases in the hippocampus (Lahti et al., 2003; Potkin et al., 2003) have been shown by some studies, supporting the drug's clinical actions such as ameliorations of delusions/hallucinations and cognitive impairments. Indeed, responders to clozapine exhibited more prominent changes in blood flow/metabolism above mentioned rather than non-responders (Potkin et al., 2003; Molina

haloperidol treatment in structural MRI studies (Shenton et al., 2001).

hippocampus is an important target area of risperidone.

al., 2003; Lahti et al., 2003).

superior temporal cortex has been demonstrated to be closely related to auditory hallucination (Dierks et al., 1999; Lennox et al., 2000). Indeed, the first-episode and drugnaïve patients with auditory hallucinations presented higher (Horga et al., 2011) and lower metabolism (Cleghorn et al., 1992; Vita et al., 1995) compared with normal controls. Further, activity in this region was reported to be negatively associated with disorganization as a form of thought disorders (Ebmeier et al., 1993; Erkwoh et al., 1997; Sabri et al., 1997). The hippocampal and/or parahippocampal gyrus are also related with hallucination/delusion and disorganization. PET studies have shown an increase (Gur et al., 1995; Molina et al., 2005b) and decrease (Tamminga et al., 1992; Kim et al., 2000; Horga et al., 2011) in rCBF/rGMR of the regions in schizophrenia compared with normal controls, and positive (Liddle et al., 1992) and negative correlations (Schröder et al., 1996) between metabolism in the regions and hallucinations. Although these reports have very conflicting results and do not reach a definitive conclusion, they do suggest that both the lateral and medial parts of the temporal lobe are closely related with the positive symptoms.

The findings of activity within other key brain regions in schizophrenia have been very controversial. As for the striatum, several reports on drug-naïve patients have shown a significant reduction relative to normal controls (Buchsbaum et al., 1987, 1992a; Shihabuddin et al., 1998), suggesting a relation with putative neurological soft signs in the very early stage (Dazzan et al., 2004). The thalamus has a function of filtering all sensory signals from input to the cortex, and is known to play a primary role in the etiology of schizophrenia- namely, dysfunction in the correct perception of information from the external world. The activity in the thalamus has been alternatively reported to increase (Andreasen et al., 1997; Jacobsen et al., 1997; Kim et al., 2000; Clark et al., 2001) or decrease (Vita et al., 1995; Hazlett et al., 1999, 2004; Buchsbaum et al., 1996; Lehrer et al., 2005). Moreover, increases of rCBF/rGMR in the cerebellum (Andreasen et al., 1997; Kim et al., 2000; Desco et al., 2003) and the subcortical regions (Buchsbaum et al., 1998, 2007a; Desco et al., 2003) have been observed. As described above, attempts to clarify the pathophysiology of schizophrenia have focused on brain regions from the frontal and temporal cortex to the subcortical regions including the striatum, thalamus, hippocampus and cerebellum. It appears that the approach of elucidating the pathophysiology requires an integrative interpretation based on the putative aberrant networks and their correlation with symptoms. Taken together, these findings suggest that resting blood flow and metabolism studies contribute to the elucidation of the disease pathophysiology by macroscopic investigation over the whole brain and microscopic investigation focusing on key regions.

#### **2.3 Impacts of antipsychotics on blood flow and metabolism**

Antipsychotics have some significant effects on brain blood flow and metabolism, and are presumed to be closely related to the potency of neuroleptics. All antipsychotics commonly induce dopamine (DA) D2 receptor antagonistic actions, resulting in the most direct action for improvement of delusions and hallucinations. Traditionally, typical antipsychotics such as haloperidol, an almost pure DA D2 blocker, had been widely used. But more recently, atypical antipsychotics have become the mainstay in the clinical practice. These atypical antipsychotics can reduce the extra-pyramidal symptoms and improve the negative symptoms and cognitive impairments by an antagonistic action on the 5-HT 2A receptors. Functional neuroimaging studies have provided important insights about the differences in pharmacological action and treatment effect among a diverse range of antipsychotics, and the subsequent functional changes in the central nervous system.

superior temporal cortex has been demonstrated to be closely related to auditory hallucination (Dierks et al., 1999; Lennox et al., 2000). Indeed, the first-episode and drugnaïve patients with auditory hallucinations presented higher (Horga et al., 2011) and lower metabolism (Cleghorn et al., 1992; Vita et al., 1995) compared with normal controls. Further, activity in this region was reported to be negatively associated with disorganization as a form of thought disorders (Ebmeier et al., 1993; Erkwoh et al., 1997; Sabri et al., 1997). The hippocampal and/or parahippocampal gyrus are also related with hallucination/delusion and disorganization. PET studies have shown an increase (Gur et al., 1995; Molina et al., 2005b) and decrease (Tamminga et al., 1992; Kim et al., 2000; Horga et al., 2011) in rCBF/rGMR of the regions in schizophrenia compared with normal controls, and positive (Liddle et al., 1992) and negative correlations (Schröder et al., 1996) between metabolism in the regions and hallucinations. Although these reports have very conflicting results and do not reach a definitive conclusion, they do suggest that both the lateral and medial parts of

The findings of activity within other key brain regions in schizophrenia have been very controversial. As for the striatum, several reports on drug-naïve patients have shown a significant reduction relative to normal controls (Buchsbaum et al., 1987, 1992a; Shihabuddin et al., 1998), suggesting a relation with putative neurological soft signs in the very early stage (Dazzan et al., 2004). The thalamus has a function of filtering all sensory signals from input to the cortex, and is known to play a primary role in the etiology of schizophrenia- namely, dysfunction in the correct perception of information from the external world. The activity in the thalamus has been alternatively reported to increase (Andreasen et al., 1997; Jacobsen et al., 1997; Kim et al., 2000; Clark et al., 2001) or decrease (Vita et al., 1995; Hazlett et al., 1999, 2004; Buchsbaum et al., 1996; Lehrer et al., 2005). Moreover, increases of rCBF/rGMR in the cerebellum (Andreasen et al., 1997; Kim et al., 2000; Desco et al., 2003) and the subcortical regions (Buchsbaum et al., 1998, 2007a; Desco et al., 2003) have been observed. As described above, attempts to clarify the pathophysiology of schizophrenia have focused on brain regions from the frontal and temporal cortex to the subcortical regions including the striatum, thalamus, hippocampus and cerebellum. It appears that the approach of elucidating the pathophysiology requires an integrative interpretation based on the putative aberrant networks and their correlation with symptoms. Taken together, these findings suggest that resting blood flow and metabolism studies contribute to the elucidation of the disease pathophysiology by macroscopic investigation over the whole brain and microscopic investigation focusing on key regions.

Antipsychotics have some significant effects on brain blood flow and metabolism, and are presumed to be closely related to the potency of neuroleptics. All antipsychotics commonly induce dopamine (DA) D2 receptor antagonistic actions, resulting in the most direct action for improvement of delusions and hallucinations. Traditionally, typical antipsychotics such as haloperidol, an almost pure DA D2 blocker, had been widely used. But more recently, atypical antipsychotics have become the mainstay in the clinical practice. These atypical antipsychotics can reduce the extra-pyramidal symptoms and improve the negative symptoms and cognitive impairments by an antagonistic action on the 5-HT 2A receptors. Functional neuroimaging studies have provided important insights about the differences in pharmacological action and treatment effect among a diverse range of antipsychotics, and

the temporal lobe are closely related with the positive symptoms.

**2.3 Impacts of antipsychotics on blood flow and metabolism** 

the subsequent functional changes in the central nervous system.

A number of previous studies have shown that typical neuroleptics such as haloperidol reduce blood flow and metabolism in the frontal lobe. These effects were repeatedly replicated in studies of both acute (Bartlett et al., 1998; Lahti et al., 2005) and chronic administration (Bartlett et al., 1991; Buchsbaum et al., 1992b; Miller et al., 1997, 2001; Lahti et al., 2003). Further, whereas haloperidol was reported to be related with hypoperfusion and hypometabolism in the hippocampus in terms of amelioration of positive symptoms (Lahti et al., 2003), increases of rCBF/rGMR in the motor cortex induced by haloperidol were presumed to be related with extra-pyramidal symptoms (Molina et al., 2003; Buchsbaum et al., 2007), and the decrease in activity in the occipital cortex following haloperidol treatment might be related with sedative effects (Bartlett et al., 1991; Desco et al., 2003; Lahti et al., 2003).

An increase in rCBF/rGMR in the basal ganglia in patients with schizophrenia by neuroleptics, in particular haloperidol, is the most consistent finding among numerous reports on antipsychotics. This has been replicated very well in the acute effect (Lahti et al., 2005) as well as the chronic effect (Buchsbaum et al., 1987, 1992a, 2007a; Miller et al., 1997, 2001; Scottish Schizophrenia Research Group, 1998; Corson et al., 2002; Desco et al., 2003; Lahti et al., 2003). The increase of blood flow and metabolism in this area is presumed to be due to increases of activity in the post synapses through upregulation of DA D2 receptors induced by a potent blocking action of the receptor by haloperidol (Miller et al., 1997; Corson et al., 2002). This notion is in line with the increase of volume in this area following haloperidol treatment in structural MRI studies (Shenton et al., 2001).

Studies on the effects of atypical antipsychotics on brain perfusion/metabolism have become to be examined based on more detailed neuronal substrates than studies on typical antipsychotics by appearance of voxel wise analysis. Although risperidone has less effect on the reduction of blood flow in the frontal lobe than haloperidol (Miller et al., 2001), the drug induces a significant reduction in the prefrontal cortex relative to baseline (Berman et al., 1996; Liddle et al., 2000; Ngan et al., 2002; Molina et al., 2008). In the basal ganglia, the degree of increase in blood flow/metabolism by risperidone is likely smaller than that by haloperidol (Liddle et al., 2000; Miller et al., 2001). Liddle et al. (2000) demonstrated that treatment with risperidone for 6 weeks showed a significant positive relation between decrease in the hippocampus and decrease in reality distortion, suggesting that the hippocampus is an important target area of risperidone.

Olanzapine is likely that its effect of blood flow/metabolism in the frontal lobe is lesser than that by risperidone (Gonul et al., 2003b; Molina et al., 2005c; Buchsbaum et al., 2007b).

Clozapine, the gold standard among the atypical neuroleptics, has a pharmacological profile with weaker blockade of DA D2 receptors and broader actions for multiple receptors than other atypical antipsychotics, and these characteristics are presumed to be related to its superior clinical efficacy relative to other neuroleptics. Interestingly, several previous studies have reported that clozapine induced a significant reduction in blood flow/metabolism in the prefrontal cortex (Potkin et al., 1994, 2003; Cohen et al., 1997; Lahti et al., 2003; Molina et al., 2005d, 2008). On the other hand, increases in several parts of the prefrontal cortex, including the ACC (Lahti et al., 2003) and decreases in the hippocampus (Lahti et al., 2003; Potkin et al., 2003) have been shown by some studies, supporting the drug's clinical actions such as ameliorations of delusions/hallucinations and cognitive impairments. Indeed, responders to clozapine exhibited more prominent changes in blood flow/metabolism above mentioned rather than non-responders (Potkin et al., 2003; Molina

Resting State Blood Flow and Glucose Metabolism in Psychiatric Disorders 135

reward/punishment, and so on. It is, therefore, very reasonable that hypoactivity in the frontal lobe is observed in subjects with depression relative to normal subjects. Inconsistent results among the previous studies mentioned above, suggest great heterogeneity of patients with the disease. Therefore, a number of confounding factors, such as age, sex, brain organic condition (ischemia and atrophy), pharmacotherapy (drug class, dose and duration), and disease stage (acute or remit), could easily affect brain activity, leading to a

Studies with careful sample selection, in which subjects who were, for example, in a drugnaïve state or in withdrawal from antidepressants for several weeks, were careful selected in order to reduce the heterogeneity have reported significant hypoperfusion and hypometabolism in the dorsolateral prefrontal cortex in subjects with depression relative to normal controls (Kimbrell et al., 2002; Gonul et al., 2004). The reduction in activity in this region was the most consistent finding among those in the frontal lobe as a whole. Additionally, rCBF and rGMR in the dorsolateral prefrontal cortex were negatively correlated with the severity of depression (Baxter et al., 1989; Martinot et al., 1990; Hurwitz et al., 1990; Bonne et al., 1996; Kimbrell et al., 2002; Gonul et al., 2004). Subanalyses of each symptom have shown the degree of psycho-motor retardation and the activity in the prefrontal cortex to be negative correlated (Bench et al., 1993; Dolan et al., 1993; Videbach et al., 2002). Although increased activities in the ventrolateral prefrontal cortex and OFC have been suggested by a sequence of studies by Drevets (Drevets et al., 1992, 1997; Drevets, 1999, 2000), other studies did not sufficiently examine these areas. With respect to the medial prefrontal cortex and ACC, although most studies with relatively large ROIs in this area, observed hypoperfusion and hypometabolism (Hurwitz et al., 1990; Bench et al., 1992, 1993; Bonne et al., 1996; Mayberg et al., 1997; Videbach et al., 2002; Gonul et al., 2004), several detailed studies on these regions demonstrated decreased activities in the dorsal medial prefrontal and dorsal ACC (Kimbrell et al., 2002; Fitzgerald et al., 2008) and increased activities in the rostral ACC (Drevets, 1999; Konarski et al., 2007). In particular, the latter region was suggested that the greater perfusion and metabolism was, the better clinical

varied distribution of rCBF/rGMR in the patient group as a whole.

response to antidepressant treatment was predicted (Mayberg et al., 1997).

**3.2 Change of rCBF/rGMR induced by antidepressants and ECT** 

patients with MDD.

As for the limbic region, increases in rCBF/rGMR in the amygdala (Drevets et al., 1992; Abercrombie et al., 1998; Videbach et al., 2002) and caudate (Gonul et al., 2004; Périco et al., 2005) were observed in patients with depression relative to normal subjects. The subgenual ACC, a component within the paralimbic system, was hypoactive in patients with unipolar depression (Drevets et al., 1997; Skaf et al., 2002; Fitzgerald et al., 2008), but also in patients with bipolar depression (Drevets et al., 1997). The caudate was also reported to show hypometabolism (Baxter et al., 1985; Drevets et al., 1992). These reductions in activity in anatomically small areas, such as the subgenual ACC and caudate, might be due to the partial volume effects (Krishnan et al., 1992; Drevets, 2000). The ventrolateral prefrontal cortex, including the subgenual ACC, has closely reciprocal connectivities with the amygdala, hypotharamus and brain stem, and disturbances of these networks could lead to the hypersensitivity to failure, pathological guilt and exaggeration of self-esteem shown in

Antidepressant agents are shown to be effective for 50-60% patients with MDD (Hirschfeld et al., 2002), and only 20-35% of patients reach remission (Mann, 2005). While diverse classes

et al., 2008). These complex patterns induced by clozapine have been suggested to be strongly related to the drug's superior clinical characteristics.

#### **2.4 Conclusion**

Functional neuroimaging studies performed in schizophrenic subjects under a resting state have made progress in the accumulation of findings on hypoperfusion/hypometabolism in the frontal lobe. It is noted that the hypofrontality is closely related with negative symptoms. On the other hand, the brain regions relevant to positive symptoms are still clearly unknown. The studies performed thus far have well explored the effects of various antipsychotics on the brain blood flow and metabolism, but neuroleptic-induced reductions in blood flow/metabolism in the prefrontal cortex have been obscure in terms of their relationship with the improvement of positive symptoms or secondary negative symptoms. By contrast, alteration in the limbic regions or the medial phase of the temporal cortex, such as the hippocampus, has been shown to be related with positive symptoms, and functional neuroimaging studies have contributed to detection of the origin of positive symptoms.
