It is well established that pediatric bipolar disorder (PBD) not only presents with affected dysregulation but also cognitive difficulties in working memory, attention, verbal memory, and executive functional domains (Pavuluri et al, 2009). Neural circuitry abnormalities that explain this multi-domain dysfunction (Pavuluri, in press) have led to the pursuit of how such abnormal pathophysiology can be reversed with pharmacotherapy (Mayanil et al, 2011). We conducted the first series of fMRI studies in pediatric mania by examining the effects of lamotrigine, risperidone, and divalproex sodium (DVPX) individually as well as through a pharmacotherapy algorithm. The pharmacological fMRI studies used tasks to probe the domains of emotion processing, response inhibition, and the interface of emotion and cognition. Biomarkers of pharmacotherapy response, especially the increase in ventrolateral prefrontal cortex (VLPFC) activity in PBD relative to the healthy controls (HC), have been tied to the reduction in manic symptoms regardless of any task (Mayanil et al, 2011). Most studies summarized here were conducted on adolescents suffering from hypomanic state at baseline and are illustrated in Figure 1.

Figure 1
figure 1

Medication works on brain to stabilize mood regulation. DLPFC, dorsolateral prefrontal cortex; MPFC, medial prefrontal cortex; MTL, medial temporal lobe; VLPFC, ventrolateral prefrontal cortex; VStr, ventral striatum; PACC, pregenual anterior cingulate cortex; PCC, posterior cingulate cortex; SACC, subgenual anterior cingulate cortex.

PowerPoint slide

Full size image

In the response inhibition and impulse control domain, using the stop signal task, lamotrigine increased the medial prefrontal cortex (MPFC), pregenual, subgenual, and posterior cingulate activity (PCC) in PBD over 14 weeks, while striatal activity was increased in HC. Lamotrigine’s glutametergic attenuating action is associated with reduction in bipolar depression, where MPFC, DLPFC, striatum, and PCC showed increased activation from the baseline (Chang et al, 2008). Another double-blind randomized trial using independent-component analysis, compared the change in connectivity of brain circuits with DVPX vs risperidone in PBD and HC. Here, DVPX increased the connectivity of the subgenual cortex within the affective evaluation and inhibition circuitry during response inhibition. This is similar to the earlier outcome with lamotrigine, confirming the role of subgenual cortex during impulse control with an antiepileptic agent. However, antipsychotic risperidone is associated with increased engagement of insula relative to DVPX group or HC.

At the interface of working memory and affective processing domains, using the N-Back working memory task, relative to HC, patients with PBD showed increased activation response in subgenual cingulate and ventral striatum when risperidone was administered. The VLPFC–MPFC and medial temporal regions showed greater activity with lamotrigine as well as DVPX (the two antiepileptic agents) relative to baseline while patients were remembering angry faces.

At the interface of executive function and affective processing domains, evaluative MPFC and ACC were engaged during negative word matching with all drugs, that is, risperidone, DVPX, and lamotrigine. However, under the negative emotional challenge, cognitive DLPFC activation response was reduced with lamotrigine and risperidone. This explains the possibility that cognitive enhancers may be needed post-mood stabilization, based on medications used or to improve the residual attention difficulties intrinsic to PBD (Pavuluri et al, 2009). Furthermore, prognostic markers of clinical outcome began to emerge, with greater amygdala activity at baseline being a marker for poor outcome in the case of risperidone (possibly involving greater effort to subdue the amygdala), whereas greater MPFC activity was a marker for a good outcome (a sign of greater deployment of higher cortical region) in the case of DVPX, on direct comparison of these medication groups. Overall, these results indicated three things: (1) negative or angry stimuli were more successful in eliciting drug-related activity than happy or neutral faces in probing group differences; (2) medications differentially engaged the brain circuitry based on neurochemistry and tasks; and (3) state vs trait: with all medications, the amygdala showed reduced activity from baseline along with recovery from manic state, but remained active relative to that of HC (trait marker), while the PFC regions were normalized with mood stabilization (state marker). However, long-term treatment for at least 4 months with a standardized algorithm led to increased amygdala connectivity (Wegbreit et al, 2011) in the affective circuitry and normalized subcortical activity (Yang et al, 2013).

In summary, in PBD, multiple domains are malfunctioning at baseline. Combination of treatments such as coupling mood stabilizers with antipsychotics for severe episodes of mania may be useful as each type of medication engages different brain circuits.

FUNDING AND DISCLOSURE

Dr Pavuluri receives or has received grant support from the following sources: National Institutes of Health (NIH), the National Institute of Mental Health, the National Alliance for Research on Schizophrenia and Depression (NARSAD), the American Foundation for Suicide Prevention, and the Marshall Reynolds Foundation. She is the recipient of the Berger-Colbeth Term Chair in Child Psychiatry and participated in the Otsuka Pharmaceuticals National Advisory Board meeting once.