**3. Major Depressive Disorder**

Functional neuroimaging studies measuring at-rest brain perfusion and metabolism in patients with major depressive disorder (MDD) have demonstrated that the etiology of the disease is closely linked with multiple components of the frontal lobe, temporal lobe, parietal lobe, limbic/paralimbic regions, and basal ganglia. Recent knowledge on affection and perception acquired from multiple human and animal research fields strongly support the findings that have been observed within depressive patients' brains in neuroimaging studies. Although a number of functional neuroimaging studies for MDD have been conducted to date, the results were varied widely among the studies. However, a sequence of inconsistent findings on MDD has demonstrated that depressive patient groups consist of highly heterogeneous subtypes, and that the etiology of depression contains multiple symptoms.

Studies on the effects of antidepressants on brain perfusion and metabolism have reported the relatively consistent finding that abnormal activity in the key brain regions relevant to depression could be normalized by successful treatment. However, no reliable markers on response prediction have been available to date in the imaging studies. On the other hand, studies of electroconvulsive therapy (ECT), an established treatment modality for refractory depression, have suggested that its effective mechanism is involved in the inhibitory process within subjects' brains that occurred immediately following the ECT course.

#### **3.1 Abnormalities in multiple prefrontal cortex and limbic regions in MDD**

Earlier functional neuroimaging studies on depression have reported significant reduction in rCBF/rGMR in the frontal lobe or prefrontal cortex in patients with depression relative to normal subjects (Baxter et al., 1989; Martinot et al., 1990; Bench et al., 1992). However, several subsequent studies with the voxel based analyses have failed to confirm this finding (Skaf et al., 2002; Videbach et al., 2002; Bonne et al., 2003). Great progression made in research on human and animal emotion and perception has elucidated that the frontal lobe and limbic/paralimbic systems are tightly involved in affective and perceptive controls, including mood, attention, decision-making, anxiety, behaviors dependent on

et al., 2008). These complex patterns induced by clozapine have been suggested to be

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.

Functional neuroimaging studies measuring at-rest brain perfusion and metabolism in patients with major depressive disorder (MDD) have demonstrated that the etiology of the disease is closely linked with multiple components of the frontal lobe, temporal lobe, parietal lobe, limbic/paralimbic regions, and basal ganglia. Recent knowledge on affection and perception acquired from multiple human and animal research fields strongly support the findings that have been observed within depressive patients' brains in neuroimaging studies. Although a number of functional neuroimaging studies for MDD have been conducted to date, the results were varied widely among the studies. However, a sequence of inconsistent findings on MDD has demonstrated that depressive patient groups consist of highly heterogeneous subtypes, and that the etiology of depression contains multiple

Studies on the effects of antidepressants on brain perfusion and metabolism have reported the relatively consistent finding that abnormal activity in the key brain regions relevant to depression could be normalized by successful treatment. However, no reliable markers on response prediction have been available to date in the imaging studies. On the other hand, studies of electroconvulsive therapy (ECT), an established treatment modality for refractory depression, have suggested that its effective mechanism is involved in the inhibitory process

Earlier functional neuroimaging studies on depression have reported significant reduction in rCBF/rGMR in the frontal lobe or prefrontal cortex in patients with depression relative to normal subjects (Baxter et al., 1989; Martinot et al., 1990; Bench et al., 1992). However, several subsequent studies with the voxel based analyses have failed to confirm this finding (Skaf et al., 2002; Videbach et al., 2002; Bonne et al., 2003). Great progression made in research on human and animal emotion and perception has elucidated that the frontal lobe and limbic/paralimbic systems are tightly involved in affective and perceptive controls, including mood, attention, decision-making, anxiety, behaviors dependent on

within subjects' brains that occurred immediately following the ECT course.

**3.1 Abnormalities in multiple prefrontal cortex and limbic regions in MDD** 

strongly related to the drug's superior clinical characteristics.

**2.4 Conclusion** 

symptoms.

**3. Major Depressive Disorder** 

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

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

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 patients with MDD.

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

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

Resting State Blood Flow and Glucose Metabolism in Psychiatric Disorders 137

depression relative to normal. With respect to antidepressants and ECT, their mechanisms

Bipolar Disorder is characterized by distinctive affective labile episodes of manic/hypomanic state and/or depressive state. Concurrently, cognitive dysfunctions such as impairments of attention, working memory and executive function usually accompany the disease. Based on recent careful clinical observations, lifetime prevalence, including all bipolar II disorder, subthreshold bipolar disorder and drug-induced manic/hypomanic episode, is up to 5% (Merikangas et al., 2007). About 60% of patients with bipolar disorder are misdiagnosed as having MDD, and further, one-third of patients experience any psychiatric symptoms for more than 10 years before a correct diagnosis is made (Hirschfeld et al., 2003). Therefore, understanding the pathophysiology of bipolar disorder is very important for exact diagnosis and effective treatment. In neuroimaging studies on bipolar disorder, however, there have been a number of difficulties with the research, such as difficulty in recruiting patients with mania into the study and with safely scanning them, and the large heterogeneity within such patient groups in terms of affective state and disease subtype. Therefore, neuroimaging studies conducted to date have tended to have small sample sizes. Also, almost all studies on bipolar disorder have employed depressive patient groups combining cases of bipolar and unipolar depression, and the data acquired to date in manic and euthymic patients have been relatively restricted compared to the findings in depressive patients. In this context, resting state rCBF/rGMR studies on bipolar disorder have appeared to be inconsistent (Stoll et al., 2000; Strakowski et al., 2000). Still, recent resting state studies are providing a cortical-anterior subcortical dysfunction model of the disease pathology through several kinds of examination, including studies on mania and comparative studies between bipolar and unipolar depression (Keener and Phillips, 2007;

There have been few studies on manic patients, and those that have been performed have been largely biased by very small sample size, patients with manic level that can cooperate with study, and continuous pharmacotherapy consisting of a mixture of mood stabilizers, antidepressants and antipsychotics. In these studies, rCBF/rGMR reduction in the prefrontal cortice, particularly the ventral prefrontal cortex and increase in the subcortical areas compared to normal controls have been relatively consistent, providing corticalsubcortical or cortical–limbic/paralimbic regions impairment as a disease model in bipolar disorder. Decrease in brain perfusion/metabolism in the frontal cortex has been reported in the lateral prefrontal cortex at rest (al-Mousawi et al., 1996; Bhardwaj et al., 2010; Brooks III et al., 2010) and during cognitive tasks (Blumberg et al., 1999; Rubinsztein et al., 2001) and in the orbitofrontal cortex at rest (Blumberg et al., 1999) and during cognitive tasks (Blumberg et al., 1999; Rubinsztein et al., 2001). On the other hand, increases of rCBF/rGMR have been reported in the dorsal ACC (Rubinsztein et al., 2001), caudal ACC (Blumberg et al., 2000) and ventral/subgenual ACC (Drevets et al., 1997; Blumberg et al., 2000; Brooks III et al., 2010) and the head of the caudate (Blumberg et al., 2000; Brooks III et al., 2010). Goodwin et al. (1997) reported that in patients with relapsed manic episodes following withdrawal of

have been under examination.

**4. Bipolar Disorder** 

Pan et al., 2009).

**4.1 Bipolar mania** 

of antidepressants are available in clinical practice at present, studies on the effect of specific antidepressants on brain perfusion or metabolism and the studies on the relationship between clinical improvement and the brain activity induced by antidepressants have been very restricted, and, further, the few such studies that exist usually have very small sample sizes. According to previous studies on these issues, aberrant regions at baseline prior to initial treatment in subjects with MDD appear to be normalized, particularly in responders to the agent. However, it is very uncertain whether the abnormalities can be recovered to a level similar to that in normal subjects (Baxter et al., 1985, 1989; Tutus et al., 1998; Ishizaki et al., 2008) or remain to a certain degree (Hurwitz et al., 1990; Martinoti et al., 1990). The discrepancies among these studies might be due to differences in class, dose of antidepressant, diverse treatment durations, different definitions of effectiveness or recovery of symptoms, or small sample sizes. Several selective serotonin reuptake inhibitors (SSRIs; paroxetine and citalopram) and serotonin and noradrenaline reuptake inhibitors (SNRIs; venlafaxine) in some well-designed studies have been examined most extensively in terms of their effects on brain perfusion/metabolism in patients with MDD. However, although several key regions, such as the frontal, temporal, parietal, and limbic regions and the basal ganglia, have been widely found to be relevant areas affected by the depressants studied, consistent findings on the combination of the relevant areas or their change directions have been very scarce. With respect to the prediction of the response to antidepressants, the greater the perfusion in the ACC (Mayberg et al., 1997), rectul gyrus (Buchsbaum et al., 1997), and lateral prefrontal cortex (Joe et al., 2006; Brockmann et al., 2009) prior to treatment was, the better the expected response. On the other hand, a decrease in rCBF/rGMR prior to treatment in the ACC (Brody et al., 1999; Konarski et al., 2009), lateral prefrontal cortex (Navarro et al., 2004) and hippocampus/basal ganglia/thalamus (Milak et al., 2009) led to a good treatment response. Therefore, the studies on this issue to date have failed to confirm conclusions.

ECT is usually indicated the patients with MDD who have been treatment-resistant to antidepressants. While this modality provides a relatively high rate of response for these patients, the understanding of its mechanism of action remains very poor. During seizures induced by ECT, evident reductions in rCBF/rGMR occurred over large brain areas (Takano et al., 2007). Afterwards, hypoperfusion and hypometabolism, to a lesser degree than during the seizure, in several brain regions, including the prefrontal region, have continued for a maximum of several months. This findings is presumed to be related to clinical responsiveness (Prohovnik et al., 1986; Rosenberg et al., 1988; Guze et al., 1991). However, some studies have demonstrated significant increases in rCBF in several brains (Bonne et al., 1996; Kohn et al., 2007). These discrepancies might be due to several confounding factors, such as procedural-related factors including anesthetics and electrode replacements, or to varying durations between the termination of the ECT course and imaging scanning.

#### **3.3 Conclusion**

The etiology of depression is strongly suggested to be related to the frontal lobe and limbic/paralimbic regions. However, the highly heterogeneity of patients with depression could lead to inconsistent results observed among studies. In addition, assessing the results in anatomically small areas or components with obscure boundaries, such as the subgenual ACC, amygdala, and OFC, is very difficult, and this serious problem in the interpretations of these regions stems from the effects of volume reduction in these regions in patients with depression relative to normal. With respect to antidepressants and ECT, their mechanisms have been under examination.
