Several studies identified changes of CMRGlu and rCBF after cognitive behavior therapy (CBT) in OCD patients. Elevated baseline rCBF and CMRGlu was observed in OCD patients in the orbitofrontal cortex, anterior cingulate cortex, basal ganglia, and thalami (Baxter et al. 1987; Perani et al. 1995; Nakatani et al. 2003; Nordahl et al. 1989; Yamanishi et al. 2009). The decrease of metabolism in connection to response to psychotherapy was observed in the prefrontal cortex (Brody et al. 2001; Kennedy et al. 2007; Yamanishi et al. 2009), the bilateral caudate (Schwartz et al. 1996), the head of the right caudate (Baxter et al. 1992; Nakatani et al. 2003), and the bilateral thalamus (Saxena et al. 2009), replicating the results of several studies of OCD treatment using neurosurgery (Biver et al. 1995; Sachdev et al. 2001) and pharmacotherapy (Baxter et al. 1992; Saxena et al. 2002). Good response to psychotherapy was associated with greater decrease in the right caudate metabolism compared to nonresponders (Schwartz et al. 1996; Nakatani et al. 2003). In contrast, Apostolova et al. (2010) reported increase in the right caudate nucleus in both CBT and paroxetine group of OCD patients. This was attributed to the effect of therapy on concomitant affective and anxiety disorders and the large proportion of early-onset OCD in the sample.

Posttreatment changes of activity in the dorsal anterior cingulate cortex are somewhat dubious: increase of the activity in this area was reported by Saxena et al. (2002), while decrease was observed by Brody et al. (2001) and some other pharmacotherapy studies (Perani et al. 1995; Baxter et al. 1992; Saxena et al. 2001). No changes after CBT were seen in the study by Baxter et al. (1992). Decrease in cingulate activity may not be a necessary mechanism of action for the improvement of OCD symptoms, but may still be associated with treatment response. Increase was also reported in the temporal lobes (Brody et al. 2001; Kennedy et al. 2007).

Response to CBT and antidepressant pharmacotherapy in patients with panic disorder resulted in decrease of right frontal and temporal regions in both treatment groups (Prasko et al. 2004).

Pretreatment metabolic predictors of response to CBT and pharmacotherapy were studied and the results are dubious: positive correlation was found between treatment response in the CBT-treated group and pretreatment left orbitofrontal cortex metabolism (Brody et al. 1998), the left and right lateral orbitofrontal cortex, the left dorsomedial prefrontal cortex, and thalamus (Kennedy et al. 2007). A tendency to a negative correlation exists between the relative response to treatment and the pretreatment glucose metabolism in the left orbitofrontal cortex (Apostolova et al. 2010) and in the right inferior occipital cortex and left inferior temporal cortex (Kennedy et al. 2007). Yamanishi et al. reported significant correlation of the baseline rCBF in the bilateral orbitofrontal cortex and the change in the severity score among responders (Yamanishi et al. 2009). Schwartz et al. (1996) reported significant bilateral decrease in caudate glucose metabolic rates in the CBT responders group in comparison to nonresponders.

Brody et al. (2001) compared DMM patients treated with either CBT or selective serotonin reuptake inhibitors (SSRI) and healthy controls. Regional brain metabolic abnormalities seen at baseline tended to normalize with both treatments. Decrease in prefrontal cortex and left anterior cingulated gyrus and increase in left temporal lobe were seen. Martin et al. (2001) compared changes in rCBF after venlafaxine, an antidepressant of the serotonin–norepinephrine reuptake inhibitor class and CBT. Increase in the right basal ganglia was seen in both types of therapy, and increase in the right posterior cingulate followed CBT. Goldapple et al. (2004) compared CMRGlu changes after CBT and SSRI therapy. Widespread decrease of activity in frontal regions was associated with both types of therapy, and increase in hippocampus, parahippocampus, and dorsal cingulate was seen in CBT group.

Several treatment-dependent differences in limbic and cortical activity may reflect treatment modalities’ specific mechanisms, which are yet to be explained in future studies.

Several authors explored the dopamine/serotonin system changes after treatment with CBT: psychodynamic psychotherapy in MDD patients had no effect on dopamine D2 receptor binding in the striatum in ¹¹C-raclopride PET study (Hirvonen et al. 2011), but midbrain serotonin transporter density, marked by 5-HT_1a receptor binding of ¹¹CWAY-100635, significantly increased (Karlsson et al. 2010). Lehto et al. (2008) observed midbrain serotonin (SERT) and striatum dopamine transporter (DAT) densities after 12 months of psychodynamic psychotherapy in depressive patients. Midbrain SERT density increased significantly in atypical but not in nonatypical depression patients, while there were no changes in the levels of DAT.

These findings are consistent with previous studies that provided evidence of decrease in density of serotonin receptors in depression (Drevets et al. 1999; Hirvonen et al. 2008). Antidepressive effect of enhanced serotonergic activity is widely used in therapy with SSRI. According to similar clinical improvement in both types of therapy, we can assume that CBT enhances serotonergic activity through increased levels of SERT.

Cervenka et al. (2012) showed a direct relationship between symptom change after CBT and extrastriatal binding of ¹¹C FLB457, a high-affinity dopaminergic D₂/D₃ antagonist in nine patients with social anxiety disorder (SAD). Negative correlation between change in D₂ receptor binding potential and the anxiety symptoms change was found for the medial prefrontal cortex and hippocampus.

Anterior cingulotomy is currently the most common neurosurgical procedure for treatment-refractory psychiatric syndromes. In this procedure, anterior portion of the cingulate gyrus is stereotactically lesioned, interrupting tracts between the cingulate gyrus and the frontal lobes. This eliminates the efferent projections of the anterior cingulate cortex to the orbitofrontal cortex and to the limbic system, resulting in a 30–68 % response rate in OCD and depression (Ballantine et al. 1987; Dougherty et al. 2002; Jenike et al. 1991; Jung et al. 2006).

Two studies involving FDG PET identified cerebral metabolic correlates as potential predictors of treatment response to anterior cingulotomy.

Rauch et al. (2001) identified one locus of significant correlation of higher preoperative metabolism and better treatment response for OCD within right posterior cingulate cortex, approximate Brodmann area 31, while Dougherty et al. (2003) identified two loci, associated with a greater postoperative improvement of major depression – the left subgenual prefrontal cortex and left thalamus.

Greenspan et al. (2008) reported a case report on H₂ ¹⁵O PET findings of pain-evoked blood flow response in which perception of pain and temperature was assessed before and after cingulotomy for OCD. The preoperative pain-evoked activation of the bilateral middle cingulate cortex diminished postoperatively, but the activation was seen in the ipsilateral parasylvanian cortex. The study gives evidence to the functional connectivity of the hierarchical pain network.

Subcaudate tractotomy is primarily used for the treatment of refractory depression. It targets a region of white matter localized beneath the head of the caudate known as the substantia innominata, interrupting the loop between the cortex and thalamus via the striatum (Shah et al. 2008).

A bifrontal stereotactic tractotomy, a procedure that destroys bifrontal pathways located beneath and in front of the head of the caudate nucleus, was used by Biver et al. (1995) to treat a 37-year-old patient with refractory OCD. Hypometabolism was seen on a posttreatment FDG PET scan in comparison to the baseline scan in areas known to be associated with depression: the medial part of the orbital cortex, thalamus, caudate nucleus, and subgenual cingulate. Surgical intervention resulted in partial clinical improvement.

Limbic leukotomy is the combination of stereotactic lesions created in the anterior cingulotomy and subcaudate tractotomy. A response of 36–50 % was reported in 21 patients who underwent limbic leukotomy for OCD or depression (Montoya et al. 2002).