Pharmacotherapy impacts functional connectivity among affective circuits during response inhibition in pediatric mania

Contents lists available at Behavioural Brain Research Pharmacotherapy impacts functional connectivity among affective circuits during response inhibition in pediatric mania Mani N. Pavuluri , James A. Ellis , Ezra Wegbreit , Alessandra M. Passarotti , Michael C. Stevens a Pediatric Brain Research and Intervention Center, Institute for Juvenile Research, Berger-Colbeth Clinic, University of Illinois at Chicago, IL, USA b Olin Neuropsychiatry Research Center, The Institute of Living/Hartford Hospital, Yale University School of Medicine, CT, USA Objective: The aim of the current study was to determine the influence of implicated affective circuitry Received 17 August 2011 disturbance in pediatric bipolar disorder (PBD) on behavioral inhibition. The differential influence of Received in revised form an antipsychotic and an anti-epileptic medication on the functional connectivity across affective and 23 September 2011 cognitive neural operations in PBD was examined.
Accepted 3 October 2011 Methods: This was a six-week double blind randomized fMRI trial of risperidone plus placebo vs. dival- Available online 8 October 2011 proex plus placebo for patients with mania (n = 22; 13.6 ± 2.5 years). Healthy controls (HC; n = 14, 14.5 ± 2.8 years) were also scanned for normative comparison. Participants performed a response inhibi- tion fMRI task where a motor response, already ‘on the way' to execution, had to be voluntarily inhibited Functional connectivity on trials where a stop signal was presented. Independent component analysis was used to map functional connectivity across the whole brain.
Response inhibition Results: While there were no behavioral differences between the groups at pre- or post-drug trial, there was significant improvement on manic symptoms in the patient groups. All participants engaged an evaluative affective circuit (EAC: bilateral inferior frontal gyrus, middle frontal gyrus, anterior cingulate cortex (ACC), middle temporal gyrus, insulae, caudate and putamen) and a reactive affective circuit (RAC: bilateral occipital cortex, amygdala, medial frontal gyrus and insula) during task performance. Within the EAC, post-treatment and relative to HC, greater engagement was seen in left insula in risperidone group and left subgenual ACC in divalproex group. Within the RAC, greater baseline amygdala connectivity in patients did not alter with treatment.
Conclusion: EAC and RAC are two key circuits that moderate emotional influence on response inhibition in PBD. Risperidone and divalproex differentially engage the EAC. Limited change in amygdala activity with treatment in all patients indicates a likely trait deficit in PBD.
2011 Elsevier B.V. All rights reserved.
circuit consisting of the amygdala, posterior cingulate/precuneus, and fusiform/parahippocampal gyrus further dys- Overcoming the important challenges in understanding how function at multiple circuitry-wide levels. Similarly, decreased affective abnormalities contribute to disinhibition and impulsivity functional connectivity during mania between DLPFC and tem- in pediatric bipolar disorder (PBD) will help discover how interven- poral circuitry has even been found while participants were in a tions can alter these neural operations. To date, numerous studies of resting state Further, pharmacological fMRI studies of pedi- pediatric mania have focused on affective processing with or with- atric mania with risperidone, divalproex and lamotrigine have out cognitive challenge which have typically shown under activity shown increased activity during cognitive control under emo- in ventrolateral prefrontal cortex (VLPFC), medial PFC (MPFC), dor- tional challenge in VLPFC, DLPFC, MPFC, subgenual ACC, temporal solateral PFC (DLPFC) overactivity in anterior cingulate lobe, and striatum but the amygdala remained overac- cortex (ACC), amygdala, and striatum addition, reduced tive relative to healthy controls (HC) affective circuitry connectivity in PBD was found in an emotional face response level disturbance in PBD is likely to influence cognitive function, given our previous findings illustrating a strong interlink between affective and cognitive systems Therefore, in the cur- rent study, we sought to further examine how affective neural Corresponding author at: Pediatric Brain Research and Intervention Center, 1747 systems commonly involved in PBD influence impulsivity and dis- West Roosevelt Road, Department of Psychiatry, Chicago, IL 60608, USA.
Tel.: +1 312 413 0064, fax: +1 312 413 0063.
inhibition using cognitive paradigms without any affective stimuli E-mail address: (M.N. Pavuluri).
0166-4328/$ – see front matter 2011 Elsevier B.V. All rights reserved.
M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 There have been several fMRI studies of motor inhibition in this study was to use functional connectivity to map PBD patho- pediatric mania, including one treatment study. Leibenluft et al.
physiology within the hypothesized EAC and RAC. We predicted that failed inhibition during a Stop signal task in PBD that PBD patients would show less functional connectivity within involved decreased activation in right VLPFC and bilateral stria- each of these key networks during response inhibition.
tum when compared to healthy controls. In addition, a recent study A second goal was to link the hypothesized functional connec- employing a blocked Go/No-Go task similar to the Stop signal task tivity deficits to clinical improvement in response to two different used in the current study found increased activation in the DLPFC classes of medications known to stabilize affect in PBD. One of the in participants with PBD relative to controls However, the medications, risperidone, an antipsychotic that acts by serotonin Go/No-Go task employed in the Singh et al. study involved with- dopamine antagonism, known to reduce manic symptoms holding a prepotent response, rather than interrupting a response in PBD, improves VLPFC and MPFC activity pre- that was already on the way to completion, so the results of that dicted that risperidone would improve EAC functional connectivity study might not be directly comparable with the present study.
and that the treatment-induced reduction in manic symptoms will Recent studies by Passarotti et al. that emotion- correlate with the change in EAC connectivity. The other medi- ally linked brain regions such as the VLPFC and pregenual ACC are cation, divalproex sodium (divalproex), is an anti-epileptic that implicated both in the cognitive control of emotional processing modulates intracellular pathways also serves as a tradi- and in behavior inhibition in pediatric mania relative to atten- tional mood stabilizer known to reduce both manic and depressive tion deficit hyperactivity disorder (ADHD). Therefore, a particular symptoms that a similar anti-epileptic, lamotrigine, challenge in PBD studies is to unravel the relationship between led to greater subgenual and MPFC activity in pediatric mania dur- emotional systems and inhibitory control systems in manic states.
ing response inhibition predicted that divalproex would In addition, understanding the effects of medication treatments on also improve functional connectivity in EAC and that reduction the neural networks involved in response inhibition in PBD would in both the manic and depressive symptoms will correlate with help us understand the intervention effects on the neural func- the change in EAC. With regards to the RAC, previous functional tion. For example, lamotrigine monotherapy has been shown to imaging results are equivocal, with persistent increase in amygdala enhance underactive MPFC and temporal lobe regions during a activity, relative to HC, with treatment of mania decreased response inhibition task, suggesting that mood stabilizers which activity with reduction in depression within patients find- improve function in affective regions can also influence the brain ings will inform if functional connectivity in this circuit will be networks involved in response inhibition how dys- altered by either of the medications. Essentially, given that this function in the neural networks central to affective disturbance in is the first study comparing divalproex and risperidone effect on PBD motor inhibition will more conclusively link the behavior inhibition in pediatric mania, we began with the premise affective disturbance that is the hallmark of PBD to its behavioral that both medications would have an equal impact on EAC and RAC.
Informed by the aforementioned fMRI studies of PBD, we have 2. Methods
proposed theoretical models of functional networks regions of higher cortical evaluation of emotional and behavioral control in the VLPFC, evaluation in the MPFC, This was an NIH-funded (1 K23 RR018638-01) six-week outpatient double blind executive function of emotion modulation in the DLPFC randomized controlled trial (DBRCT) of risperidone plus placebo (that resembled a divalproex capsule) vs. divalproex plus placebo (that resembled a risperidone tablet) emotional and cognitive control and mod- for manic and mixed episodes of bipolar disorder. This study was approved by the ulation is accompanied by greater activity in the ACC in PBD University of Illinois at Chicago's Institutional Review Board (IRB).
in both compensatory error correction and the complex interface of affective and cognitive processing collection of brain regions can be considered an evaluative affective Inclusion criteria were a DSM-IV diagnosis of mixed or manic bipolar disorder; circuit (EAC) that is likely to contribute to behavioral disinhibition 12–18 years old; and medication free or currently clinically unstable on medication, in PBD by interfering with oversight of behavioral control. Another justifying termination of the ineffective regimen (with consent, all subjects were proposed complementary posterior circuit was an occipito-limbic washed out and free of any medication for a week prior to baseline scanning, and associative circuit the occipital cortex and amygdala, four weeks in case of fluoxetine or aripiprazole). Prior exposure to SGAs and anti- epileptic medications was acceptable. Exclusion criteria included: active substance which is activated in response to incidental emotional process- abuse; serious medical problems; autism and non-affective psychotic disorders. Par- ing This reactive affective circuit (RAC) contributes to ticipants who had a diagnosis of ADHD preceding the onset of PBD were excluded to impulsive automatic response tendencies, which are moderated by reduce the confound of comorbid attentional disorders, relevant specifically in prob- evaluative MPFC region.
ing response inhibition. Using these criteria, we recruited 44 subjects into the study.
The regions within EAC and RAC inter-communicate and are After excluding subjects whose data were unusable due to motion artifacts (HC: n = 2; risperidone group: n = 3; divalproex group: n = 3), the final sample included important for task success. We implemented a novel approach to in the analyses consisted of 14 HC, and 22 patients randomized to either risperi- study functional connectivity among specific brain regions within done (n = 11) or divalproex (n = 11). No subjects dropped out of the study. Sample the EAC and RAC networks that influence activity in distal regions characteristics are summarized in within each network during task performance Independent 2.3. Assessment and efficacy measures component analysis (ICA), used in the current study, is a model-free technique that robustly identifies distinct spatiotemporal profiles Each child and their parent or legal guardian were interviewed using the of distributed brain function that closely correspond to known Washington University in St. Louis Kiddie Schedule for Affective Disorders and anatomical neural networks. This approach was used as a means to Schizophrenia (WASH-U-KSADS) supplemented by the episode characterization of test hypotheses regarding regional functional connectivity during bipolar disorder from the KSADS – Present and Lifetime version interviews were completed by doctoral-level clinicians with established inter-rater response inhibition. Although previous studies have documented reliability. The primary clinical efficacy measure was the Young Mania Rating Scale impairments in specific brain regions underlying response inhi- (YMRS) Child Depression Rating Scale-Revised was also administered bition in PBD and other studies have documented networks involved in response inhibition in normal adolescents, 2.4. Study dosing of risperidone and divalproex date, there have been no direct tests of whether brain regions in the EAC or RAC show disrupted functional connectivity in PBD The mean (standard deviation, SD) risperidone dose at endpoint was during demands for behavioral control. Therefore, the first goal of 1.43 (±0.35) mg/day in non-responders and 1.33 (±0.43) mg/day in responders M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 Demographic variables and clinical characteristics.
Age in years (Age range: 10–19) Estimated range: 70–139) 88.54, p < 0.001 16.73, p < 0.001 17.04, p < 0.001 DVPX = Divalproex; RISP = Risperidone; HC = Healthy control; YMRS = Young Mania Rating Scale; CDRS-R = Child Depression Rating Scale-Revised; SES = Socioeconomic status.
a Estimated with Wechsler Abbreviated Scale of Intelligence (WASI; Matrix Reasoning and Vocabulary Subtests).
(defined as improvement ≥50% on the YMRS scores). The mean (SD) divalproex displacement from the median head position was greater than 1.5 mm, or if head dose at endpoint was 863.64 (±210.54) mg/day in non-responders and 855.14 rotation from the median head position was greater than 0.5◦. There were no sig- (±245.23) mg/day in responders. The mean serum valproic level at end point was nificant group differences in the number of volumes retained after discarding those 98 ␮g/mL, and 95% of patients achieved a therapeutic serum valproic level of with motion artifact. Individual volumes were excluded from analyses if, relative to >75 ␮g/mL by the 5th day. No titration of medications was allowed after day 7.
median head position, head displacement was greater than 1.5 mm or head rotation One subject in the divalproex group received lorazepam as a rescue medication at was greater than 0.5◦. T-tests revealed no significant group differences in the num- a dose of 2 mg for severe agitation during the first week of the trial. No other rescue ber of volumes retained after discarding those with motion artifact. After motion medications or stimulants were used during the trial correction and de-trending using FIASCO, the functional images were preprocessed with SPM5 Slice timing correc- 2.5. fMRI session: response inhibition task tion was applied to the data to remove signal variation due to slice acquisition temporal onset differences. The first functional image volume of each participant was used to determine parameters for spatial normalization into Montreal Neu- The fMRI behavioral paradigm was a block design task in which a motor rological Institute (MNI) standardized space employed in SPM5 using non-linear response, already ‘on the way' from planning to execution, had to be voluntarily transformation. The normalization parameters determined for the first functional inhibited when a cue instructing subjects to stop an impending response was pre- volume were subsequently applied to all of the 240 functional image volumes for sented on some trials the beginning of each trial a fixation cross appeared for each participant. The normalized functional images were then smoothed with a 850 ms. On Go trials, a target stimulus (a green airplane) was presented for 800 ms.
12-mm full width at half-maximum Gaussian filter On Stop trials, a Stop signal (a man holding a Stop signal in his hands) replaced the airplane with equal probability 250, 350, or 450 ms after the airplane appeared and 2.8. Independent component analysis subjects had to inhibit their response. The task lasted 6.11 min and consisted of six experimental blocks, three of which were Go blocks (G) and three of which were fMRI time series from all participants for the response inhibition task were ana- Stop blocks (S), and there were 7 resting blocks (F) of 10 s fixation each. Each exper- lysed using a group ICA algorithm (GIFT v1.3h; imental block had 30 trials and lasted 49.5 s. The experimental and fixation blocks The fMRI time series data for all participants were concatenated, then subjected were pseudo-randomly interspersed as follows: (F) G (F) S (F) S (F) G (F) S (F) G (F).
to two principal component analysis data reduction stages data under- We adopted this 70/30 proportion of trials in both the Go and Stop blocks (e.g., 70% went a final ICA rotation using Infomax that produced 37 maximally independent Stop trials in Stop block and 70% of Go trials in Go block) so that subjects would not components number of components to estimate was determined using the habituate to fixed trial presentation within a certain block minimum description length criteria the ICA-derived group solution, data for each participant were then back-reconstructed that individual participant 2.6. MRI protocols variability was retained for hypothesis testing. For each component, this back- reconstruction method produced a spatial map representing brain regions within Gradient-echo echo-planar functional imaging and structural acquisitions were each component "network", and a time course of BOLD signal change across the performed with a 3.0 T whole body scanner (Signa, General Electric Medical System, fMRI paradigm. Group analyses of spatial maps determined differences in degree of Milwaukee, WI) at the MR Center within the UIC Hospital. To minimize head motion regional functional connectivity, while analyses of time-course information allowed we restricted the participants' head with foam cushions. T2*-weighted functional us to determine whether or not study groups engaged each network during the fMRI images were acquired with a gradient-echo echo-planar sequence (TR = 2500 ms, TE = 25 ms, flip angle = 90◦, FOV 20 × 20 cm2, 64 × 64 matrix, 3.125 × 3.125 mm in A systematic process was used on the 37 independent components to identify plane resolution, 5-mm slice thickness, 1-mm gap, 25 slices). Anatomical images those that would be retained for further analysis. The correlation of each compo- were also acquired in the axial plane (three-dimensional spoiled gradient recalled, nent's spatial map with a priori probabilistic maps of gray matter, white matter, 1.5 mm thick contiguous axial slices) and were later co-registered with the func- and cerebral spinal fluid (CSF) within MNI space (templates provided in SPM5) tional data. The experiment run consisted of 240 time points including a 5 s rest was calculated for all components. After discarding components with an r-squared session at the beginning that was collected to allow for T1 effects to stabilize. These value >0.025 with CSF or white matter, or with low correlation to gray matter that initial two images were not included in the analyses could be an artifact, 28 components were retained. This step primarily identified and excluded obvious signal artifacts (e.g., head motion, cardiac inflow pulsatile motion).
2.7. fMRI image processing and motion correction We then discarded components not engaged by the fMRI task. To assess task engage- ment, component time courses were parameterized using multiple regression to FIASCO software (Functional Imaging Analysis Software – Computational Olio) provide association coefficients (␤-weights) between component time courses and used to implement 3D motion estimation and correction, removal of slow an overall condition model of the response inhibition task (i.e., one condition model signal drift, and identification of images with artifacts such as high shot noise or for both Stop and Go blocks). One sample t-tests against zero were carried out on displacement that cannot be readily corrected by motion correction algorithms.
the ␤ weights (pooled across groups) to determine if the evidence for task engage- We excluded from the analyses individual volumes from the time series if head ment was greater than zero (i.e., whether or not the network was engaged by the M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 Influence of pharmacotherapy on response inhibition.
Region of interest Co-ordinates x, y, z in MNI Results in the evaluative affective circuit HC > PBD at baseline (pre-treatment) Risperidone > HC changes over treatment (Post- vs. pre-treatment) L inferior frontal gyrus R inferior frontal gyrus R medial frontal gyrus L medial frontal gyrus Divalproex > HC changes over treatment (Post- vs. pre-treatment) L inferior frontal gyrus R inferior frontal gyrus R medial frontal gyrus L medial frontal gyrus Results in the reactive affective circuit HC > PBD at baseline (pre-treatment) −23, −4, −19 Risperidone > HC changes over treatment (Post- vs. pre-treatment) −23, −4, −19 Divalproex > HC changes over treatment (Post- vs. pre-treatment) L = Left; R = Right; ACC = Anterior cingulate cortex; HC = Healthy controls, PBD = Pediatric bipolar disorder.
task). Only 14 components were significantly (p < 0.05) associated with the overall- 2.11. Whole-brain ANOVAS and region-of-interest (ROI) analyses of functional condition model and were retained. Estimates were then derived for Stop and Go blocks separately to investigate the effects of task set on brain networks. To identify and visualize which brain regions were significantly engaged in each component, As described above, ICA produced a spatial map for each participant × fMRI individual participants' spatial maps were entered into an SPM5 voxel-wise one- sessions (i.e., pre- vs. post-treatment) that depicted the voxel-wise strength of sample t-test for each component. Component spatial structure was visualized by functional connectivity. These were examined in a series of SPM5 2 × 2 (pre- vs.
overlaying these results on axial slices of representative brain anatomy. Significance post-treatment session by group) ANOVA models that contrasted PBD drug effects was evaluated using p < 0.05 family-wise error rate correction for the whole brain of divalproex vs. healthy controls, risperidone vs. healthy controls, and finally risperidone vs. divalproex. The defined ROIs from the EAC and RAC networks used two-tailed ˛ = 0.05, corrected for the number of ROIs searched (ROI masks 2.9. Clinical effects of treatment and behavioral analyses Participants' YMRS and CDRS-R values were analysed with 2 × 3 time (pre- vs.
2.12. Correlations of ROI-derived functional connectivity values and clinical post-trial) by group (risperidone, divalproex, HC) Analyses of variance (ANOVA).
Participants' reaction time and accuracy were examined using a series of 2 × 2 × 3 ANOVAs; block (Go vs. Stop) by time (pre- vs. post-trial) by group (risperidone, To investigate the effect of changes in brain regions on the clinical indices, cor- divalproex, HC). In accordance with signal detection theory, sensitivity (d-prime) relations were conducted within each drug treatment group between the areas that and criterion bias (c-bias) measures were computed from participants' hit (correct showed significant changes in the ANOVAs and the changes in YMRS and CDRS- Go) and false alarm (incorrect Stop) rates. The d-prime score represents participants' R scores. In all, eight correlations were conducted: for the risperidone group, the ability to adequately detect whether or not a signal is present – in this case, it is a YMRS and CDRS-R scores were correlated with the connectivity in the left insula proxy for their ability inhibit or reverse an already in progress response in response after treatment and the change in connectivity from pre- to post-treatment, and in to the Stop signal. The c-bias score represents participants' overall tendency to the divalproex group, the two clinical indices were each correlated with post and respond, with higher positive values indicating a greater likelihood to respond to change in connectivity in the left BA25, given that those were the areas that changed any trial. These measures were examined using separate 2 × 2 × 3 ANOVAs: block in each group.
type (Go vs. Stop) by session (pre- vs. post-trial) by group (risperidone, divalproex, 2.13. Correlations of brain and clinical data to behavioral indices 2.10. Identification of task-engaged networks affected by treatment To investigate the effect of changes in brain regions on the participants' behav- ioral performance, similar correlations were conducted within each drug treatment Given that the focus of this study was to examine differential changes between group between the areas that showed significant changes in the ANOVAs and the medication groups in the engagement of brain circuits during the response inhi- changes in sensitivity (d-prime) and response threshold (c-bias) scores. In all, eight bition task over the course of treatment, with the HC group serving as a control additional correlations were conducted: for the risperidone group, the d-prime and for practice effects and development, our analyses focused on two networks that c-bias scores were correlated with the connectivity in the left insula after treatment showed a significant group by time interaction. ANOVA compared the beta weights and the change in connectivity from pre- to post-treatment, and in the divalproex between groups in SPSS using group (risperidone, divalproex, HC) as the between- group, the two behavioral indices were each correlated with post and change in subjects factor, and time (pre vs. post) as the within-subjects factor. Because the Stop connectivity in the left BA25, given that those were the areas that changed in each blocks and Go blocks likely engaged different, albeit related cognitive processes, group. Finally, four correlations between the changes in clinical data and the changes separate ANOVAs were conducted for Stop and Go blocks ( in behavioral data were also conducted, but no significant results were found M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 3. Results
divalproex group showed greater connectivity across treatment, relative to HC, in the left subgenual ACC (BA25) (t(23) = 0.98, Sample demographic and clinical data are summarized in p = 0.05), and the left insula (t(23) = 1.84, p = Comorbid diagnosis included separation anxiety disorder When compared directly with each other, the risperidone group, (n = 1) and conduct disorder (n = 1) in risperidone group, and gen- relative to the divalproex group, showed a trend toward greater eralized anxiety disorder (n = 1) and a history of substance abuse functional connectivity in the left insula ROI (t(20) = 1.91, p = 0.07).
(n = 1) in divalproex group.
In the RAC, risperidone group showed decreased change in func- tional connectivity in the left amygdala, relative to HC (t(23) = 2.47, 3.1. Behavioral results p = 0.01) Similarly, the divalproex group showed decreased change in functional connectivity, relative to HC, in the Patient groups and HC did not differ in reaction time (RT) at right amygdala (t(23) = 2.44, p = 0.01) With regards baseline or in degree of change with time between groups, and all to occipital cortex, relative to HC, the risperidone group showed an subjects slowed down with time. Accuracy did not differ between increase in the right middle occipital gyrus (t(23) = 2.96, p = 0.003) groups in the Go blocks or the Stop blocks and did not change with and the divalproex group showed an increase in functional connec- tivity in the right superior occipital gyrus (t(23) = 2.14, p = 0.02).
Patient groups and HC did not show any significant differ- ences on the 2 × 2 × 3 block type by session by group ANOVA for 3.3. Correlations of ROI values and clinical indices d-prime scores, except for a significant difference between the groups' overall d-prime scores, F(2,30) = 3.59, p = 0.040. Post hoc For the risperidone group, there was a significant correlation tests revealed that although the two patient groups did not dif- between the change in YMRS scores and the level of functional con- fer from each other, they both showed significantly less sensitivity nectivity of the insula after treatment (r(9) = −0.629, p = 0.038). For overall than the HC. All groups, however, did show increases in the divalproex group, there was a significant correlation between sensitivity on the Go blocks – albeit nonsignificantly – and the the change in functional connectivity in the left subgenual ACC HC showed an increase on the Stop Blocks, whereas the patients between sessions and the decrease in CDRS-R score (r(9) = −0.713, showed slight decreases. In addition, although there was a signifi- p = 0.014). No other correlations between clinical indices and activ- cant difference in sensitivity between the HC and the two patient ity changes in these ROIs were found.
groups in the first session, F(2,30) = 5.07, p = 0.013, there was no such difference in the second session, F(2,30) = 1.95, p = 0.160. Thus 3.4. Correlations of ROI values and behavioral indices there is some evidence that the PBD group became slightly better at the task, relative to HC, before and after medication treatment.
For the risperidone group, there was a significant correla- The criterion bias scores suggest a possible mechanism for these tion between the change in participants' response threshold and slight increases in sensitivity. All of the participants shifted from the level of functional connectivity of the insula after treatment a more liberal responding threshold (higher c-bias) in the first (r(9) = 0.84, p = 0.009), as well as a trend toward a correlation session to a more conservative one (lower c-bias) in the second between participants' threshold change and the change in left session, F(2,30) = 9.65, p = 0.004, with no other effects of group or insula connectivity (r(9) = −0.690, p = 0.05). For the divalproex block type. Taken together, the behavioral results indicate that group, none of the correlations were significant.
both patients' and controls were able to strategically adjust their response criteria in the second session to attempt to improve their performance above the first session, but that HC were slightly more successful than patients in actually doing so. (A table and additional detail that summarizes the behavioral data is available There are four key findings in this study. First, the ICA meth- two hypothesized functional circuits engaged during response inhibition, the "appraising" EAC (bilateral inferior frontal gyrus, middle frontal gyrus, ACC, middle temporal 3.2. fMRI whole brain analyses results gyrus, insulae, striatum) and the "automatic" RAC (bilateral occip- ital lobe, parietal lobe, and amygdala). Second, ROI analysis within Only two networks survived our selection process that identi- the EAC and RAC confirmed our hypothesis that medications altered fied non-artifactual, task-associated components with evidence for the functional connectivity. Patients, relative to HC and within EAC, significant effects of treatment in the PBD group. The first compo- showed greater engagement of left insula with risperidone and left nent (henceforth known as the Evaluative Affective Circuit (EAC)) subgenual ACC with divalproex. Third, as predicted, reduction in showed functionally-integrated activity in the bilateral inferior manic symptoms correlated with increased connectivity of insula frontal gyrus, middle frontal gyrus, anterior cingulate cortex, mid- within EAC with risperidone. The reduction in depressive symp- dle temporal gyrus, insulae, and striatum (FDR p < 0.001, toms correlated with increased left subgenual engagement with The second component (henceforth known as the Reactive Affec- divalproex. Fourth, amygdala functional connectivity did not alter tive Circuit (RAC)) showed functionally-integrated activity in the with either of the medications. This could be a trait abnormality in bilateral occipital lobe, parietal lobe, and amygdala (FDR p < 0.001, patients where prefrontal effects, rather than subcortical activity The increase or decrease in activity within the networks sig- changes, are seen during treatment, during motor inhibition. An nifies their connectivity within the component in all subjects, PBD alternate explanation may be that the amygdala activity reduces and HC included, during the task performance.
with treatment, but relative to HC, it remains high.
Next, the change in ROI within the EAC and RAC were examined Both groups engaged in the behavioral strategy of becoming to determine the change in connectivity among these networks more conservative in their responses, but were only slightly suc- within each patient group in response to risperidone or dival- cessful in translating this strategy into improved performance to proex, relative to HC. In the EAC, the risperidone group showed perform the task in the second session than the patients were, but greater increase in functional connectivity across treatment, rel- both groups showed a similar shift in strategy. Thus, these results ative to HC, in the left and (t(23) = 3.22, p = 0.003) and the right suggest that as patients' mood symptoms abated with treatment, insula (t (23) = 1.90, p = 0.07) (Also in the EAC, the they were able to employ the same response strategy as HC.

M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 Fig. 1. Evaluative affective circuit during response inhibition.
4.1. ICA reveals EAC and RAC engagement in PBD during response networks were significantly engaged differently in patients, rela- tive to HC. On tasks where performance might be challenging and elicit emotional reactions to success or failures, demanding the We were able to map two distinct functional networks, the EAC need to modulate responses in the context of performing an ardu- and the RAC, in PBD and HC. Interestingly, these putative emotional ous task, evaluative higher cortical regions have come into play

M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 Fig. 2. Reactive affective circuit during response inhibition.
is now well established that inferior frontal gyrus engages of insula, ACC, and temporo-striatal regions a in affective and response inhibition control, has liaison between affective and cognitive regions of decision con- strong connectivity to the "cognitive" middle frontal gyrus trol, error correction, and emotional moderation, as in this study.
These cortical regions of EAC work in concert with opercular regions For example, in previous work, the pregenual cingulate-insular

M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 Fig. 3. Region of interest within functional circuits: medication effects over treatment period.
cortex elements of the EAC network described in this study have connectivity between visual association cortex to the amygdale been linked to cognitive set maintenance they increase the occipito-temporal inferior longitudinal fasciculus in response to errors during ongoing performance of a similar the high level of frustration associated with having to Go/No-Go response inhibition task suggests that impair- inhibit responses in this study, amygdala was engaged on the left ment of processing due to abnormal integration of insular cortex side The amygdala is found to play a key role in the inter- might disrupt these executive control abilities, with inability to action of emotion and cognition such as emotion's influence on monitor inhibition in the light of manic symptoms. The RAC is a attention and perception findings support the notion of posterior affective circuit that also emerged during response inhi- emotional impulsivity this early automatic bition task performance. The functional connectivity that emerged response is proportional to the attention directed to the stimuli in RAC is in line with our previous findings in support of direct have yet to establish the sequence of signal pathways M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 to determine whether EAC activity is preceded by that in RAC or 4.2.3. Risperidone and divalproex fail to dampen amygdala connectivity within RAC The amygdala showed significantly increased functional con- nectivity within RAC, at baseline that failed to change with 4.2. Medication-specific effects on functional connectivity treatment. There is consistent evidence from functional studies in PBD showing increased amygdala activity remained While biomarkers of medication outcome and prognosis are an hyperactive, relative to HC, regardless of treatment for mania ideal goal, given the preliminary nature and the complexity of our and during euthymic illness state addition to functional studies, it may be optimal to provide a bio-signature based on a neuroimaging abnormalities in the PFC and amygdala, structural cluster of biological findings that maps any given drug's operation.
neuroimaging studies indicate smaller amygdala volumes in PBD Also, findings on fMRI studies of drug mechanism must always patients relative to HC, contrasts with adult stud- be interpreted based on the paradigm chosen to probe a specific ies that report larger normal volumes.
domain, the illness status, and the class of drug.
Larger amygdala in adult studies in fact been hypothesized to result from hypertrophy due to chronic and exces- 4.2.1. Risperidone increases insular engagement within EAC sive activation in manic patients While findings of altered Our main finding with risperidone is of critical importance as size do not necessarily imply intrinsic primary abnormalities in the risperidone is often administered to address aggression and neg- amygdala, they may correlate with functional abnormalities in the ative reactivity within acute mania parallel, insula amygdala such as increased connectivity in response to the emo- increased activation while processing negative emotions in healthy tional and cognitive challenges posed by the task (e.g. frustration, humans with increased serotonin neurotransmission well effortful inhibition of responses), regardless of dysfunction in PFC as in PBD the insula showed hypometabolism input to the amygdala. Given the central role that the amygdala during omission errors on an attentional task in euthymic bipolar plays in arousal and emotional reactivity its role in the adults a finding that aligns with structural studies showing emotion-cognition interface is meaningful that amygdala decreased grey matter in left insula in patients and their rela- did not alter its task-engagement following treatment with either tives Risperidone appears to engage this critical region in medication in PBD as was seen in controls. The change in amyg- the EAC to a greater extent, likely by the virtue of its serotonin- dala connectivity found in HC might be the result of the increase in dopamine antagonism, due to greater serotonin HT 1A receptors automaticity due to task familiarity in the RAC and decreased need in the insula emotion and attention during task performance. The increased functional connectivity of insula in the EAC with risperidone treatment and its correlation with change 4.3. Clinical symptoms influence response inhibition in manic symptoms underscores the mechanism behind the effect of risperidone in mania. However, given that HT 1A receptors are The novel findings are that both risperidone and divalproex abundantly present in hippocampus and amygdala, these results of groups showed greater connectivity in two distinct regions in EAC, insular hyperconnectivity must be interpreted as potentially spe- relative to HC, and that both of these regions correlated signifi- cific to risperidone and during behavior inhibition. Alternatively, cantly and differentially with the reduction in manic and depressive risperidone administered in a larger sample may illustrate a more symptoms, respectively. The increased engagement of insula in the wide spread and significantly increased engagement of these other EAC after treatment with risperidone illustrates the alterations in regions that are rich in HT 1A receptors.
neural connectivity underlying the reduction in manic symptoms.
The increase in the functional connectivity in subgenual ACC pro- vides insight into the mechanisms of the traditional mood stabilizer 4.2.2. Divalproex increases subgenual ACC engagement within divalproex, which additionally addresses depressive symptoms subsyndromal depressive symptoms commonly Within the EAC, left subgenual ACC showed greater functional coexist with pediatric mania. Depressive symptoms paired with connectivity in divalproex patients, relative to HC. This finding expending attentional resources during response inhibition task is consistent with the underactive subgenual ACC in untreated performance would lead to "emotional impulsivity" in PBD, patients with adult bipolar disorder while performing an atten- especially because cognitive conflicts require effortful processing tional task key region modulates autonomic responses and, therefore, can be aversive and evoke negative emotions and neurotransmission in animals Indeed, stimulation of If the patients were dealing with mood symptoms, they could have subgenual ACC ameliorated symptoms in adult depression interfered with their ability to implement a more conservative and perfusion was normalized with treatment in the same region response strategy similar to that found by the HC. Thus, our study Subgenual ACC also exhibited decreased grey matter in is significant in exhibiting the effects of risperidone and divalproex adult mood disorders where there was reduction in glia during motor inhibition, with significant reductions in manic and Further, we were able to show that subgenual ACC engagement depressive mood symptoms and engagement of the EAC through significantly correlated with reduced depressive symptoms in the insula and subgenual ACC, respectively, to moderate patients' emo- current study, although we failed to show such correlation with tional impulsivity.
manic symptoms. Finally, while divalproex engaged subgenual ACC A major strength of this study is that the ICA method is a data- greater than HC, risperidone did not differ from divalproex when driven (i.e. model-blind) technique to measure dynamic changes we compared the two patient groups. These results are in line in brain network circuits working in concert, rather than isolated with our findings from a previous study of risperidone illustrating changes in activity. A weakness of this study is the relatively small increased subgenual ACC activity during a cognitive control task sample size, although couched in a strong study design. The use performed under emotional challenge alternate explana- of a block design for the task also did not permit us to compare tion for this lack of difference on direct comparison between the the networks that were engaged for Stop vs. Go trials or correct two medications may be due to the fact that the brain circuitry vs. incorrect trials, unlike some previous studies engagement may be specific to a mood state than a specific med- given that that the study is based on a model of the neural circuits ication. The brain functional signature may be more specific to a that takes into account overall performance during the response manic episode vs. a depressive episode or a euthymic state inhibition task, the fact that the study was a block design is less M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 relevant. Moreover, given that we found comparable levels of per- of cognitive and emotional brain systems in pediatric bipolar disorder. J Child formance between patients and HCs, we avoided the potential Adolesc Psychopharmacol 2010;20(October (5)):395–406.
[11] Pavuluri MN, Passarotti AM, Mohammed T, Carbray JA, Sweeney JA. Enhanced confound that large differences in task difficulty between groups working and verbal memory after lamotrigine treatment in pediatric bipolar caused the neural changes we observed. We were able to compare disorder. Bipolar Disord 2010;12(March (2)):213–20.
two types of medications in an unmedicated manic patient sample, [12] Pavuluri MN, Passarotti AM, Lu LH, Carbray JA, Sweeney JA. Double-blind ran- domized trial of risperidone versus divalproex in pediatric bipolar disorder: using monotherapy, with a normative HC sample for comparison. It fMRI outcomes. Psychiatry Res 2011;193(July (1)):28–37.
is compelling to see that previous findings can now be incorporated [13] Passarotti AM, Sweeney JA, Pavuluri MN. Fronto-limbic dysfunction in mania into more sophisticated circuit-level models that serve as potential pre-treatment and persistent amygdala over-activity post-treatment in pedi- bio-signatures of treatment outcome. These results are beginning atric bipolar disorder. Psychopharmacology 2011;216(August (4)):485–99.
[14] Passarotti AM, Sweeney JA, Pavuluri MN. Differential engagement of cognitive to inform why clinicians would need to combine two classes of and affective neural systems in pediatric bipolar disorder and attention deficit medications because they exhibit complementary mechanisms of hyperactivity disorder. J Int Neuropsychol Soc 2010;16(January (1)):106–17.
action within specific functional neural circuits.
[15] Passarotti AM, Sweeney JA, Pavuluri MN. Neural correlates of incidental and directed facial emotion processing in adolescents and adults. Sog Cogn Affect Neurosci 2009;4(January (4)):387–98.
[16] Passarotti AM, Pavuluri MN. Brain functional domains inform therapeutic interventions in attention-deficit/hyperactivity disorder and pediatric bipolar disorder. Expert Rev Neurother 2011;11(June (6)):897–914.
Behavioral response inhibition in pediatric mania engages two [17] Leibenluft E, Rich BA, Vinton DT, Nelson EE, Fromm SJ, Berghorst LH, et al. Neu- functionally connected circuits, the EAC and the RAC. Risperidone ral circuitry engaged during unsuccessful motor inhibition in pediatric bipolar disorder. Am J Psychiatry 2007;164(January (1)):52–60.
works by engaging insula to a greater degree than HC, and dival- [18] Singh MK, Chang KD, Mazaika P, Garrett A, Adleman N, Kelley R, et al. Neural proex works by engaging subgenual ACC relative to HC within the correlates of response inhibition in pediatric bipolar disorder. J Child Adolesc EAC. Reduction in manic symptoms for the risperidone group and [19] Pavuluri MN, Passarotti AM, Harral EM, Sweeney JA. Enhanced prefrontal func- depressive symptoms in the divalproex group distinctly showed tion with pharmacotherapy on a response inhibition task in adolescent bipolar correlations with change in the functional connectivity of EAC.
disorder. J Clin Psychiatry 2010;71(November (11)):1526–34.
However, amygdala activity did not alter with treatment within [20] Pavuluri MN, Sweeney JA. Integrating functional brain neuroimaging and devel- opmental cognitive neuroscience in child psychiatry research. J Am Acad Child RAC with either risperidone or divalproex, suggesting a probable Adolesc Psychiatry 2008;47(November (11)):1273–88.
trait of dysfunctional emotional reactivity in PBD, relative to HC.
[21] Pavuluri MN, O‘connor MM, Harral EM, Sweeney JA. An fMRI study of the These patterns of activity contribute to the development of biosig- interface between affective and cognitive neural circuitry in pediatric bipolar disorder. Psychiatry Res 2008;162(February (3)):244–55.
natures of treatment response in PBD.
[22] Devinsky O, Morrell MJ, Vogt BA. Contributions of anterior cingulate cortex to behaviour. Brain 1995;118(February (Pt 1)):279–306.
[23] Catani M, Jones DK, Donato R, Ffytche DH. Occipito-temporal connections in the human brain. Brain 2003;126(September (Pt 9)):2093–107.
[24] Morris JS, Ohman A, Dolan RJ. Conscious and unconscious emotional learning This research was funded by NIH 1 K23 RR018638-01 and NIH- in the human amygdala. Nature 1998;393(June (6684)):467–70.
[25] Friston K. Beyond phrenology: what can neuroimaging tell us about distributed circuitry? Annu Rev Neurosci 2002;25:221–50. Epub 2002 March 19.
[26] Passarotti AM, Sweeney JA, Pavuluri MN. Neural correlates of response inhibi- tion in pediatric bipolar disorder and attention deficit hyperactivity disorder.
Appendix A. Supplementary data
Psychiatry Res 2010;181(January (1)):36–43.
[27] Stevens MC, Kiehl KA, Pearlson GD, Calhoun VD. Functional neural networks Supplementary data associated with this article can be found, in underlying response inhibition in adolescents and adults. Behav Brain Res the online version, at [28] Konarski JZ, Kennedy SH, Segal ZV, Lau MA, Bieling PJ, McIntyre RS, et al.
Predictors of nonresponse to cognitive behavioural therapy or venlafaxine using glucose metabolism in major depressive disorder. J Psychiatry Neurosci [29] Pandina G, DelBello M, Kushner S, Van Hove I, Augustyns I, Kusumakar V, et al.
[1] Chang K, Adleman NE, Dienes K, Simeonova DI, Menon V, Reiss A. Anoma- Risperidone for the treatment of acute mania in bipolar youth [poster]. In: lous prefrontal-subcortical activation in familial pediatric bipolar disorder: Presented at the 54th annual meeting of the American Academy of Child and a functional magnetic resonance imaging investigation. Arch Gen Psychiatry Adolescent Psychiatry (AACAP). 2007.
[30] Monti B, Polazzi E, Contestabile A. Biochemical, molecular and epigenetic mech- [2] Pavuluri MN, Schenkel LS, Aryal S, Harral EM, Hill SK, Herbener ES, et al. Neu- anisms of valproic acid neuroprotection. Curr Mol Pharmacol 2009;2(January rocognitive function in unmedicated manic and medicated euthymic pediatric bipolar patients. Am J Psychiatry 2006;163(February (2)):286–93.
[31] Kowatch RA, Findling R, Scheffer R, Stanford KE. Placebo controlled trial of dival- [3] Pavuluri MN, O‘Connor MM, Harral EM, Sweeney JA. Affective neural circuitry proex versus lithium for bipolar disorder. In: Presented at the Annual Meeting during facial emotion processing in pediatric bipolar disorder. Biol Psychiatry of the American Academy of Child and Adolescent Psychiatry. 2007.
[32] Wagner KD, Redden L, Kowatch RA, Wilens TE, Segal S, Chang K, et al. A double- [4] Rich BA, Vinton DT, Roberson-Nay R, Hommer RE, Berghorst LH, McClure blind, randomized, placebo-controlled trial of divalproex extended-release in EB, et al. Limbic hyperactivation during processing of neutral facial expres- the treatment of bipolar disorder in children and adolescents. J Am Acad Child sions in children with bipolar disorder. Proc Natl Acad Sci USA 2006;103(June Adolesc Psychiatry 2009;48(May (5)):519–32.
[33] Chang KD, Wagner C, Garrett A, Howe M, Reiss A. A preliminary functional [5] Blumberg HP, Martin A, Kaufman J, Leung HC, Skudlarski P, Lacadie C, et al.
magnetic resonance imaging study of prefrontal-amygdalar activation changes Frontostriatal abnormalities in adolescents with bipolar disorder: preliminary in adolescents with bipolar depression treated with lamotrigine. Bipolar Disord observations from functional MRI. Am J Psychiatry 2003;160(July (7)):1345–7.
[6] Blumberg HP, Kaufman J, Martin A, Charney DS, Krystal JH, Peterson BS. Sig- [34] Kaufman J, Birmaher B, Brent DA, Ryan ND, Rao U. K-SADS-PL. J Am Acad Child nificance of adolescent neurodevelopment for the neural circuitry of bipolar Adolesc Psychiatry 2000;39(October (10)):1208.
disorder. Ann N Y Acad Sci 2004;1021(June):376–83.
[35] Young RC, Biggs JT, Ziegler VE, Meyer DA. A rating scale for mania: reliability, [7] Pavuluri MN, Passarotti AM, Harral EM, Sweeney JA. An fMRI study of the neural validity and sensitivity. Br J Psychiatry 1978;133(November):429–35.
correlates of incidental versus directed emotion processing in pediatric bipolar [36] Poznanski EO, Mokros HB. Children‘s Depression Rating Scale Revised (CDRS-R).
disorder. J Am Acad Child Adolesc Psychiatry 2009.
Los Angeles, CA: Western Psychological Services; 1995.
[8] Rich BA, Fromm SJ, Berghorst LH, Dickstein DP, Brotman MA, Pine DS, et al.
[37] Eddy WF, Fitzgerald M, Genovese CR, Mockus A, Noll DC. Functional image Neural connectivity in children with bipolar disorder: impairment in the analysis software – computational olio. In: Prat A, editor. Proceedings in Com- face emotion processing circuit. J Child Psychol Psychiatry 2008;49(January putational Statistics. Heidelberg: Physica-Verlag; 1996. p. 39–49.
[38] Calhoun VD, Adali T, Pearlson GD, Pekar JJ. A method for making group infer- [9] Dickstein DP, Gorrostieta C, Ombao H, Goldberg LD, Brazel AC, Gable CJ, et al.
ences from functional MRI data using independent component analysis. Hum Fronto-temporal spontaneous resting state functional connectivity in pediatric Brain Mapp 2001;14(November (3)):140–51.
bipolar disorder. Biol Psychiatry 2010;68(November (9)):839–46.
[39] Calhoun VD, Kiehl KA, Pearlson GD. Modulation of temporally coherent brain [10] Pavuluri MN, Passarotti AM, Parnes SA, Fitzgerald JM, Sweeney JA. A pharma- networks estimated using ICA at rest and during cognitive tasks. Hum Brain cological functional magnetic resonance imaging study probing the interface Mapp 2008;29(July (7)):828–38.
M.N. Pavuluri et al. / Behavioural Brain Research 226 (2012) 493–503 [40] Bell AJ, Sejnowski TJ. An information-maximization approach to blind separa- [64] Kodaka F, Ito H, Takano H, Takahashi H, Arakawa R, Miyoshi M, et al. Effect of tion and blind deconvolution. Neural Comput 1995;7(November (6)):1129–59.
risperidone on high-affinity state of dopamine D2 receptors: a PET study with [41] Li YO, Adali T, Calhoun VD. Estimating the number of independent com- agonist ligand [11C](R)-2-CH3O-N-n-propylnorapomorphine. Int J Neuropsy- ponents for functional magnetic resonance imaging data. Hum Brain Mapp [65] Strakowski SM, Adler CM, Holland SK, Mills N, DelBello MP. A preliminary [42] Erhardt EB, Rachakonda S, Bedrick EJ, Allen EA, Adali T, Calhoun VD. Compar- FMRI study of sustained attention in euthymic, unmedicated bipolar disorder.
ison of multi-subject ICA methods for analysis of fMRI data. Hum Brain Mapp [66] Drevets WC, Savitz J, Trimble M. The subgenual anterior cingulate cortex in [43] Worsley KJ, Marrett S, Neelin P, Vandal AC, Friston KJ, Evans AC. A unified mood disorders. CNS Spectr 2008;13(August (8)):663–81.
statistical approach for determining significant signals in images of cerebral [67] Mayberg HS, Lozano AM, Voon V, McNeely HE, Seminowicz D, Hamani C, activation. Hum Brain Mapp 1996;4(1):58–73.
et al. Deep brain stimulation for treatment-resistant depression. Neuron [44] Leibenluft E, Rich BA. Pediatric bipolar disorder. Annu Rev Clin Psychol [68] Drevets WC, Bogers W, Raichle ME. Functional anatomical correlates of antide- [45] Rubia K, Russell T, Overmeyer S, Brammer MJ, Bullmore ET, Sharma T, et al. Map- pressant drug treatment assessed using PET measures of regional glucose ping motor inhibition: conjunctive brain activations across different versions metabolism. Eur Neuropsychopharmacol 2002;12(December (6)):527–44.
of go/no-go and stop tasks. Neuroimage 2001;13(February (2)):250–61.
[69] Mayberg HS, Brannan SK, Tekell JL, Silva JA, Mahurin RK, McGinnis S, et al.
[46] Durston S, Thomas KM, Worden MS, Yang Y, Casey BJ. The effect of preceding Regional metabolic effects of fluoxetine in major depression: serial changes and context on inhibition: an event-related fMRI study. Neuroimage 2002;16(June relationship to clinical response. Biol Psychiatry 2000;48(October (8)):830–43.
[70] Mayberg HS, Liotti M, Brannan SK, McGinnis S, Mahurin RK, Jerabek PA, [47] Petrides M, Pandya DN. Comparative cytoarchitectonic analysis of the human et al. Reciprocal limbic-cortical function and negative mood: converging PET and the macaque ventrolateral prefrontal cortex and corticocortical connection findings in depression and normal sadness. Am J Psychiatry 1999;156(May patterns in the monkey. Eur J Neurosci 2002;16(July (2)):291–310.
[48] Fair DA, Cohen AL, Power JD, Dosenbach NU, Church JA, Miezin FM, et al. Func- [71] Dickstein DP, Milham MP, Nugent AC, Drevets WC, Charney DS, Pine DS, et al.
tional brain networks develop from a "local to distributed" organization. PLoS Frontotemporal alterations in pediatric bipolar disorder: results of a voxel- Comput Biol 2009;5(May (5)):e1000381.
based morphometry study. Arch Gen Psychiatry 2005;62(July (7)):734–41.
[49] Dosenbach NU, Fair DA, Cohen AL, Schlaggar BL, Petersen SE. A dual-networks [72] Chang K, Karchemskiy A, Barnea-Goraly N, Garrett A, Simeonova DI, Reiss A.
architecture of top-down control. Trends Cogn Sci 2008;12(March (3)): Reduced amygdalar gray matter volume in familial pediatric bipolar disorder.
J Am Acad Child Adolesc Psychiatry 2005;44(June (6)):565–73.
[50] Stevens MC, Kiehl KA, Pearlson GD, Calhoun VD. Brain network dynamics during [73] DelBello MP, Zimmerman ME, Mills NP, Getz GE, Strakowski SM. Magnetic res- error commission. Hum Brain Mapp 2009;30(January (1)):24–37.
onance imaging analysis of amygdala and other subcortical brain regions in [51] Amaral DG, Price JL, Pitkanen A, Carmichael ST. Anatomical organization of adolescents with bipolar disorder. Bipolar Disord 2004;6(February (1)):43–52.
the primate amygdaloid complex. In: Aggleton J, editor. The Amygdala: Neu- [74] Frazier JA, Ahn MS, DeJong S, Bent EK, Breeze JL, Giuliano AJ. Magnetic reso- robiological Aspects of Emotion, Memory, and Mental Dysfunction. New York: nance imaging studies in early-onset bipolar disorder: a critical review. Harv Wiley-Liss; 1992. p. 1–66.
Rev Psychiatry 2005;13(May (3)):125–40.
[52] Armony JL, Servan-Schreiber D, Cohen JD, LeDoux JE. An anatomically [75] Altshuler L, Bartzokis G, Grieder T, Curran J, Mintz J. Amygdala enlargement constrained neural network model of fear conditioning. Behav Neurosci in bipolar disorder and hippocampal reduction in schizophrenia: an MRI study demonstrating neuroanatomic specificity. Arch Gen Psychiatry 1998;55:663–4.
[53] Phelps EA. Emotion and cognition: insights from studies of the human amyg- [76] Strakowski SM, DelBello MP, Sax KW, Zimmerman ME, Shear PK, Hawkins JM, dala. Ann Rev Psychol 2006;57:27–53.
et al. Brain magnetic resonance imaging of structural abnormalities in bipolar [54] Whalen PJ, Bush G, McNally RJ, Wilhelm S, Mcinerney SC, Jenike MA, et al.
disorder. Arch Gen Psychiatry 1999;56(March (3)):254–60.
The emotional counting stroop paradigm: a functional magnetic resonance [77] Swayze VW, Andreasen NC, Alliger RJ, Yuh WT, Ehrhardt JC. Subcortical and imaging probe of the anterior cingulate affective division. Biol Psychiatry temporal structures in affective disorder and schizophrenia: a magnetic reso- nance imaging study. Biol Psychiatry 1992;31(February (3)):221–40.
[55] Pessoa L, McKenna M, Gutierrez E, Ungerleider LG. Neural processing of [78] Strakowski SM, Adler CM, DelBello MP. Volumetric MRI studies of mood dis- emotional faces requires attention. Proc Natl Acad Sci USA 2002;99(August orders: do they distinguish unipolar and bipolar disorder? Bipolar Disord [56] Davis M, Whalen PJ. The amygdala: vigilance and emotion. Mol Psychiatry [79] Brambilla P, Harenski K, Nicoletti M, Sassi RB, Mallinger AG, Frank E, et al. MRI investigation of temporal lobe structures in bipolar patients. J Psychiatr Res [57] Biederman J, Mick E, Hammerness P, Harpold T, Aleardi M, Dougherty M, et al.
Open-label, 8-week trial of olanzapine and risperidone for the treatment of [80] Altshuler LL, Bartzokis G, Grieder T, Curran J, Jimenez T, Leight K, et al. An MRI bipolar disorder in preschool-age children. Biol Psychiatry 2005;58(October study of temporal lobe structures in men with bipolar disorder or schizophre- nia. Biol Psychiatry 2000;48(July (2)):147–62.
[58] Pavuluri MN, Henry DB, Devineni B, Carbray JA, Naylor MW, Janicak PG.
[81] Gorno-Tempini ML, Pradelli S, Serafini M, Pagnoni G, Baraldi P, Porro C, et al.
A pharmacotherapy algorithm for stabilization and maintenance of pedi- Explicit and incidental facial expression processing: an fMRI study. Neuroimage atric bipolar disorder. J Am Acad Child Adolesc Psychiatry 2004;43(July (7)): [82] Critchley H, Daly E, Phillips M, Brammer M, Bullmore E, Williams S, et al. Explicit [59] Pavuluri MN, Henry DB, Findling RL, Parnes S, Carbray JA, Mohammed T, et al.
and implicit neural mechanisms for processing of social information from facial Double-blind randomized trial of risperidone versus divalproex in pediatric expressions: a functional magnetic resonance imaging study. Hum Brain Mapp bipolar disorder. Bipolar Disord 2010;12(September (6)):593–605.
[60] Takano H, Ito H, Takahashi H, Arakawa R, Okumura M, Kodaka F, et al. Sero- [83] Keightley ML, Winocur G, Graham SJ, Mayberg HS, Hevenor SJ, Grady CL. An tonergic neurotransmission in the living human brain: A positron emission fMRI study investigating cognitive modulation of brain regions associated with tomography study using [(11)C]dasb and [(11)C]WAY100635 in young healthy emotional processing of visual stimuli. Neuropsychologia 2003;41(5):585–96.
men. Synapse 2010;(November).
[84] Geller B, Fox LW, Clark KA. Rate and predictors of prepubertal bipolarity during [61] Pavuluri MN, O'Connor MM, Harral EM, Sweeney JA. Lamotrigine effects on follow-up of 6- to 12-year-old depressed children. J Am Acad Child Adolesc neurocognition in pediatric bipolar disorder. [Poster]. Florida: Am College of Psychiatry 1994;33(May (4)):461–8.
[85] Findling RL, Youngstrom EA, McNamara NK, Stansbrey RJ, Demeter CA, Bedoya [62] Brooks III JO, Hoblyn JC, Ketter TA. Metabolic evidence of corticolimbic dysreg- D, et al. Early symptoms of mania and the role of parental risk. Bipolar Disord ulation in bipolar mania. Psychiatry Res 2010;181(February (2)):136–40.
[63] Matsuo K, Kopecek M, Nicoletti MA, Hatch JP, Watanabe Y, Nery FG, et al. New [86] Botvinick MM. Conflict monitoring and decision making: reconciling two structural brain imaging endophenotype in bipolar disorder. Mol Psychiatry perspectives on anterior cingulate function. Cogn Affect Behav Neurosci

Source: http://jaellis.com/PharmacotherapyAffectiveCircuits.pdf