**4. Bipolar Disorder**

136 Neuroimaging – Cognitive and Clinical Neuroscience

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

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.

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

studies on this issue to date have failed to confirm conclusions.

**3.3 Conclusion** 

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; Pan et al., 2009).

#### **4.1 Bipolar mania**

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

Resting State Blood Flow and Glucose Metabolism in Psychiatric Disorders 139

disease, or whether reliable mood-state markers in the disease exist. In this context, studies

Some studies on patients with euthymic state compared to normal controls have reported a decrease of rCBF/rGMR in the lateral prefrontal (Culha et al., 2008; Brooks Ⅲ et al., 2009b) and ACC (Culha et al., 2008) at rest, and the lateral prefrontal (Krüger et al., 2003) and OFC (Blumberg et al., 1999; Krüger et al., 2003) during cognitive tasks or symptomprovocation. On the other hand, regions with increased perfusion/metabolism were observed in the subcortical areas such as the amygdala (Brooks Ⅲ et al., 2009b) and parahippocampus (Brooks Ⅲ et al., 2009b) at rest. Krüger et al (2003, 2006) in symptomprovocation studies demonstrated that although increased rCBF in the subgenual ACC seen in normal controls was deficit in euthymic patients, increased perfusion in the dorsal ACC was observed only in the patients. Though there have been very few studies conducted on euthymia, patients with euthymia appear to show a decrease of rCBF/rGMR in the prefrontal cortex and an increase in rCBF/rGMR in the subcortical areas, according to previous reports. These notions are comparable to recent clinical observations that patients in a euthymic state show significant cognitive impairments identical to the distinctive pathological states of mania and depression (Kessing, 1998; Elshahawi et al., 2011), and they are in preparatory stage to relapse fragile to stress

on euthymia will be more and more important for addressing these issues.

(Swann, 2010), but not asymptomatic state not meeting manic and depression.

Functional neuroimaging studies on bipolar disorder have demonstrated hypoactivity in the cortex, particularly the ventral prefrontal cortex, and concurrent hyperactivity in the subcortical or limbic/paralimbic regions. To data, however, this knowledge has not reflected the clinical bipolarity of mania and depression and thus remains a trait marker. Furthermore, these findings cannot be distinguished from those of other psychiatric disorders, including unipolar depression. Studies with more sophisticated designed and

Obsessive-compulsive disorder (OCD) has a lifetime prevalence of 2-3% (Weissman et al., 1994). OCD is characterized by persistent and recurrent thoughts that invade conscious awareness against a patient's will (obsessions) and is further usually accompanied by egodystonic, ritualistic behaviors that the patient is obliged to perform in order to prevent overwhelming anxiety (compulsions). Patients with OCD form a more homogeneous group than those with other psychiatric disorders, and this perhaps accounts for the fact that previous functional neuroimaging studies have provided relatively consistent findings on aberrant brain regions in this disorder, which include the OFC, ACC, caudate nuclei, thalamus and so on. That is, the etiology of OCD has been presumed to follow a cortico-subcortical model. Functional neuroimaging techniques have contributed substantially to the exploration of these areas relevant to the disorder. Furthermore, recent reports on treatment intervention for OCD have strongly suggested that selective serotonin reuptake inhibitors (SSRI) and cognitive behavior therapy (CBT), both established treatment approaches, raise some effects on patients' brain blood flow and metabolism, and further normalize aberrant regional perfusion

**4.4 Conclusion** 

larger sample size will be needed in the future.

and metabolism within these networks in treatment responders.

**5. Obsessive-Compulsive Disorder** 

lithium, increase of rCBF in the ACC was positively correlated with manic symptoms. These findings lead to and partly support the anatomical-functional hypothesis that while the orbitofrontal and lateral prefrontal impairments are related with affective/impulsive dysregulation and cognitive dysfunction, respectively, compensatory functional hyperactivity reflects the findings of increase in the ACC and limbic/paralimbic regions observed in resting-state studies (Keener and Phillips, 2007; Pan et al., 2009).

#### **4.2 Bipolar depression**

Although there have been more reports on bipolar depression than on mania, the findings from this body of work are rather confusing. This may be due, at least in part, to the design of these studies. That is, earlier studies have frequently used a disease group combining cases of unipolar and bipolar depression, and when they have compared bipolar depression with other conditions, they have alternatively used normal healthy subjects, patients with unipolar depression and subjects with mania/euthymia as the comparison group. Moreover, the different studies have different target regions (ACC, subgenual prefrontal cortex and amygdala). With respect to the cortex, although few reports demonstrated any regions with hyperperfusion and hypermetabolism in bipolar depression relative to normal controls, areas with hypoperfusion/hypometabolosm in the patients compared to normal controls spread very broader in the lateral prefrontal (Baxter et al., 1985, 1989; Ketter et al., 2001; Brooks III et al., 2009a), medial prefrontal (Baxter et al., 1985; Bauer et al., 2005; Brooks III et al., 2009a), subgenual ACC (Drevets et al., 1997; Brooks III et al., 2009a), temporal lobe (Baxter et al., 1985; Ketter et al., 2001; Bhardwaj et al., 2010), occipital lobe (Baxter et al., 1985; Ketter et al., 2001) and parietal lobe (Baxter et al., 1985; Ketter et al., 2001). On the other hand, hyperperfusion/hypermetabolism have also been observed in the subcortical or limbic/paralimbic areas, including the amygdala (Ketter et al., 2001; Drevets et al., 2002; Bauer et al., 2005; Mah et al., 2007), subgenual ACC (Drevets et al., 1997; Bauer et al., 2005; Mah et al., 2007), ventral striatum (Bauer et al., 2005), caudate nucleus (Ketter et al., 2001; Mah et al., 2007), and putamen (Ketter et al., 2001; Mah et al., 2007), nucleus accumbens (Ketter et al., 2001; Mah et al., 2007), thalamus (Ketter et al., 2001; Bauer et al., 2005) and cerebellum (Bauer et al., 2005).

There have been a few reports comparing patients with bipolar depression and bipolar mania within the same study. Examination of the subgenual ACC (Brodmann area 25) by Drevets et al. (1997) demonstrated clear distinction of increased activity when mania and decreased activity when depression, and growing attention has been paid to this area as a mood-state marker in bipolar disorder. However, some subsequent studies showed higher metabolism in the depressive state (Bauer et al., 2005; Mah et al., 2007), indicating a failure to conform. The inconsistency among studies on small anatomical area such as the subgenual ACC may be related to shortcomings in the characteristics of nuclear imaging, such as insufficient spatial resolution of the scanner or inaccurate normalization to the standard brain (Drevets et al., 2002).

#### **4.3 Euthymia**

Although manic state and depressive state represent clinically extreme and opposite symptoms, neuroimaging findings on the two states are relatively similar. Thus, a corticalsubcortical model raises some questions as to whether this model means trait marker in the disease, or whether reliable mood-state markers in the disease exist. In this context, studies on euthymia will be more and more important for addressing these issues.

Some studies on patients with euthymic state compared to normal controls have reported a decrease of rCBF/rGMR in the lateral prefrontal (Culha et al., 2008; Brooks Ⅲ et al., 2009b) and ACC (Culha et al., 2008) at rest, and the lateral prefrontal (Krüger et al., 2003) and OFC (Blumberg et al., 1999; Krüger et al., 2003) during cognitive tasks or symptomprovocation. On the other hand, regions with increased perfusion/metabolism were observed in the subcortical areas such as the amygdala (Brooks Ⅲ et al., 2009b) and parahippocampus (Brooks Ⅲ et al., 2009b) at rest. Krüger et al (2003, 2006) in symptomprovocation studies demonstrated that although increased rCBF in the subgenual ACC seen in normal controls was deficit in euthymic patients, increased perfusion in the dorsal ACC was observed only in the patients. Though there have been very few studies conducted on euthymia, patients with euthymia appear to show a decrease of rCBF/rGMR in the prefrontal cortex and an increase in rCBF/rGMR in the subcortical areas, according to previous reports. These notions are comparable to recent clinical observations that patients in a euthymic state show significant cognitive impairments identical to the distinctive pathological states of mania and depression (Kessing, 1998; Elshahawi et al., 2011), and they are in preparatory stage to relapse fragile to stress (Swann, 2010), but not asymptomatic state not meeting manic and depression.

#### **4.4 Conclusion**

138 Neuroimaging – Cognitive and Clinical Neuroscience

lithium, increase of rCBF in the ACC was positively correlated with manic symptoms. These findings lead to and partly support the anatomical-functional hypothesis that while the orbitofrontal and lateral prefrontal impairments are related with affective/impulsive dysregulation and cognitive dysfunction, respectively, compensatory functional hyperactivity reflects the findings of increase in the ACC and limbic/paralimbic regions

Although there have been more reports on bipolar depression than on mania, the findings from this body of work are rather confusing. This may be due, at least in part, to the design of these studies. That is, earlier studies have frequently used a disease group combining cases of unipolar and bipolar depression, and when they have compared bipolar depression with other conditions, they have alternatively used normal healthy subjects, patients with unipolar depression and subjects with mania/euthymia as the comparison group. Moreover, the different studies have different target regions (ACC, subgenual prefrontal cortex and amygdala). With respect to the cortex, although few reports demonstrated any regions with hyperperfusion and hypermetabolism in bipolar depression relative to normal controls, areas with hypoperfusion/hypometabolosm in the patients compared to normal controls spread very broader in the lateral prefrontal (Baxter et al., 1985, 1989; Ketter et al., 2001; Brooks III et al., 2009a), medial prefrontal (Baxter et al., 1985; Bauer et al., 2005; Brooks III et al., 2009a), subgenual ACC (Drevets et al., 1997; Brooks III et al., 2009a), temporal lobe (Baxter et al., 1985; Ketter et al., 2001; Bhardwaj et al., 2010), occipital lobe (Baxter et al., 1985; Ketter et al., 2001) and parietal lobe (Baxter et al., 1985; Ketter et al., 2001). On the other hand, hyperperfusion/hypermetabolism have also been observed in the subcortical or limbic/paralimbic areas, including the amygdala (Ketter et al., 2001; Drevets et al., 2002; Bauer et al., 2005; Mah et al., 2007), subgenual ACC (Drevets et al., 1997; Bauer et al., 2005; Mah et al., 2007), ventral striatum (Bauer et al., 2005), caudate nucleus (Ketter et al., 2001; Mah et al., 2007), and putamen (Ketter et al., 2001; Mah et al., 2007), nucleus accumbens (Ketter et al., 2001; Mah et al., 2007), thalamus (Ketter et al., 2001; Bauer et al., 2005) and

There have been a few reports comparing patients with bipolar depression and bipolar mania within the same study. Examination of the subgenual ACC (Brodmann area 25) by Drevets et al. (1997) demonstrated clear distinction of increased activity when mania and decreased activity when depression, and growing attention has been paid to this area as a mood-state marker in bipolar disorder. However, some subsequent studies showed higher metabolism in the depressive state (Bauer et al., 2005; Mah et al., 2007), indicating a failure to conform. The inconsistency among studies on small anatomical area such as the subgenual ACC may be related to shortcomings in the characteristics of nuclear imaging, such as insufficient spatial resolution of the scanner or inaccurate normalization to the

Although manic state and depressive state represent clinically extreme and opposite symptoms, neuroimaging findings on the two states are relatively similar. Thus, a corticalsubcortical model raises some questions as to whether this model means trait marker in the

observed in resting-state studies (Keener and Phillips, 2007; Pan et al., 2009).

**4.2 Bipolar depression** 

cerebellum (Bauer et al., 2005).

standard brain (Drevets et al., 2002).

**4.3 Euthymia** 

Functional neuroimaging studies on bipolar disorder have demonstrated hypoactivity in the cortex, particularly the ventral prefrontal cortex, and concurrent hyperactivity in the subcortical or limbic/paralimbic regions. To data, however, this knowledge has not reflected the clinical bipolarity of mania and depression and thus remains a trait marker. Furthermore, these findings cannot be distinguished from those of other psychiatric disorders, including unipolar depression. Studies with more sophisticated designed and larger sample size will be needed in the future.
