Diagnostic and clinical implications of functional neuroimaging in bipolar disorder.

Accepted Manuscript Diagnostic and Clinical Implications of Functional Neuroimaging in Bipolar Disorder John O. Brooks III, PhD, MD Nathalie Vizueta , PhD PII:

S0022-3956(14)00166-6

DOI:

10.1016/j.jpsychires.2014.05.018

Reference:

PIAT 2386

To appear in:

Journal of Psychiatric Research

Received Date: 2 December 2013 Revised Date:

15 April 2014

Accepted Date: 29 May 2014

Please cite this article as: Brooks III JO, Vizueta N, Diagnostic and Clinical Implications of Functional Neuroimaging in Bipolar Disorder, Journal of Psychiatric Research (2014), doi: 10.1016/ j.jpsychires.2014.05.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

RI PT

Abstract: 171 words Main Text: 6,464 words Number of Display Items: 5 Figures, 1 Table Number of References: 125 Supplementary Materials: 0

SC

Diagnostic and Clinical Implications of Functional Neuroimaging in Bipolar Disorder

John O. Brooks III PhD, MD

M AN U

Nathalie Vizueta, PhD

TE D

Department of Psychiatry & Biobehavioral Sciences, UCLA Semel Institute for Neuroscience & Human Behavior, Los Angeles, CA

AC C

EP

Correspondence: John O. Brooks, Ph.D., M.D. 760 Westwood Plaza, B8-267 Los Angeles, CA 90024 Voice: (310) 825-6179 Fax: (310) 206-2072 Email: [email protected]

KEY WORDS: Bipolar Disorder; fMRI; PET; biomarkers; treatment; mood disorders


Abstract Advances in functional neuroimaging have ushered in studies that have enhanced our

RI PT

understanding of the neuropathophysiology of bipolar disorder, but do not yet have clinical applications. We describe the major circuits (ventrolateral, dorsolateral, ventromedial, and anterior cingulate) thought to be involved in the corticolimbic

SC

dysregulation that can pervade mood states in patients with bipolar disorder. The

potential clinical application of functional neuroimaging in bipolar disorder is considered

M AN U

in terms of prognostic, predictive, and treatment biomarkers. To date, most prior research has focused on prognostic biomarkers to differentiate patients with bipolar disorder from those with other affective or psychotic diagnoses or healthy subjects. Work in the search for treatment biomarkers, which suggest mechanisms of pharmacodynamic or treatment response, and predictive work has involved only

TE D

pediatric patients diagnosed with bipolar disorder thus far. The results of work to date are encouraging and suggest that functional neuroimaging may be of eventual benefit in

EP

determining biomarkers of treatment response. Further refinement of biomarker identification and perhaps even illness characterization are needed to find prognostic

AC C

and predictive biomarkers of bipolar disorder.


Introduction Bipolar disorder, characterized by recurrent mood episodes, affects at least 3% of the population (Kessler, Chiu, Demler, & Walters, 2005). In bipolar I disorder,

RI PT

patients experience episodes of abnormally elevated mood (mania) and most often (although not required diagnostically) depression as well. Bipolar II disorder is similar, though patients experience milder episodes of elevated mood (hypomania) and

SC

depressive episodes are required. Over the course of the illness, patients with bipolar disorder spend much more time with depressive compared to mood elevation symptoms

M AN U

(Judd et al., 2003).

The underlying neurobiology of bipolar disorders has been investigated through the use of brain imaging, with functional neuroimaging playing a major role in helping to detect altered brain networks and regions. Links between clinical symptomology and

TE D

underlying neural changes raise the important issue of whether neuroimaging can contribute to the definition of biomarkers related to bipolar disorder. In this review, we focus primarily on functional neuroimaging findings that are relevant to the search for

EP

potential biomarkers of bipolar disorders. (Other reviews have provided more detailed general surveys of cerebral metabolic studies (Brooks, Wang, & Ketter, 2010b) and

AC C

fMRI findings (Altshuler & Townsend, 2012; Chen, Suckling, Lennox, Ooi, & Bullmore, 2011; Phillips & Swartz, 2014).) Savitz et al. (Savitz, Rauch, & Drevets, 2013) discussed important considerations

for using brain imaging to detect biomarkers in mood disorders. In particular, they noted the distinction between (a) prognostic biomarkers, which are baseline characteristics that identify risk for disease or disease progression, (b) predictive biomarkers, which


predict the likelihood of response to a treatment, and (c) treatment (or pharmacodynamic) biomarkers, which provide evidence of treatment effects. There is not necessarily overlap among the types of biomarkers, because a prognostic

RI PT

biomarker may meet requirements of a treatment biomarker. Indeed, treatment

biomarkers may vary according to type of treatment. Because verification of biomarkers requires longitudinal measures, reliability of functional brain imaging and patient-specific

SC

factors can affect the likelihood of biomarker detection.

We limit discussion to functional neuroimaging studies that have used either

M AN U

positron emission tomography in conjunction with 18flourodeoxyglucose (18FDG-PET) or functional magnetic resonance imaging (fMRI).

18

FDG-PET is an imaging modality that

involves measurement of regional uptake of radioactive glucose in the brain. A limitation of PET is that delay in isotope uptake precludes ‘real-time’ observation of

TE D

brain function. Moreover, the typical spatial resolution of PET scans is not high as in other imaging modalities. fMRI is a noninvasive procedure that provides an indirect

EP

measure of neural activity during task performance yielding fairly high-resolution images of blood oxygen level dependent (BOLD) activity. While advantageous for spatial

AC C

resolution, the BOLD signal of fMRI develops slowly, which results in poor temporal resolution. Often this tradeoff is reasonable to obtain real-time functional data. More detailed descriptions regarding imaging modalities and their application to the study of bipolar disorder may be found elsewhere (Adler, Cerullo, & Strakowski, 2012). We will first describe a prevailing model of the neural underpinnings of bipolar disorder to provide a context for functional neuroimaging findings. Next, we consider existing literature in terms of the three types of biomarkers. In our discussion of


research that could lead to prognostic biomarkers, we review changes in brain function associated with acute mood states of depression, mania, and normal (euthymic) mood as well as initial work in the area of differential diagnosis of bipolar disorder. Research

describe initial work that has been performed. Neural model of bipolar disorder

RI PT

related to predictive and treatment biomarkers in bipolar disorder is limited, but we

SC

Neural models of bipolar disorder have entailed variants of what has been

referred to as a corticolimbic (Anand, Li, Wang, Lowe, & Dzemidzic, 2009; Brooks,

M AN U

Hoblyn, Woodard, Rosen, & Ketter, 2009a) or anterior limbic (Adler, Delbello, & Strakowski, 2006; Strakowski, DelBello, & Adler, 2005) model, which is illustrated in Figure 1. The corticlolimbic model was used to explain altered emotional control after consensus meetings of researchers in bipolar disorder (Strakowski et al., 2012) and

TE D

further elaborated in a review of functional neuroimaging findings in emotion regulation (Phillips et al., 2014). The circuits described in the consensus model were proposed to account for internal and external emotional control along with cognition. Thus, a

EP

ventrolateral circuit appears to process external emotional stimuli automatically (Phillips et al., 2014) and a ventromedial circuit to process internally-generated emotion (Phillips,

AC C

Ladouceur, & Drevets, 2008). An additional anterior cingulate circuit was proposed to integrate emotional and cognitive output to modulate behavior. Although many of the structural components of the circuits are interconnected through the amygdala (a subcortical brain structure primarily involved with regulation of fear of potentially threatening stimuli (Fusar-Poli et al., 2009)), for ease of relating them to clinical


phenomena, we discuss each of these circuits separately in the context of proposed control functions. Insert Figure 1 here

RI PT

Ventrolateral circuit. The ventrolateral circuit, illustrated in Figure 2, includes the ventrolateral prefrontal cortex, generally defined as Brodmann’s areas (BA) 10 & 47. Output from the ventrolateral prefrontal cortex is routed to the ventromedial striatum

SC

(which includes the ventromedial caudate, ventral putamen, nucleus accumbens, and olfactory tubercle), then to the globus pallidus, and finally to the thalamus, which

M AN U

regulates emotional expression. The circuit is completed by fibers from the thalamus that project back to the ventrolateral prefrontal cortex (Almeida et al., 2009). The anterior temporal cortex, including BA 20 and 38, provides input to the ventrolateral circuit through its reciprocal connections with the ventrolateral prefrontal cortex and the

TE D

amygdala. Within the ventrolateral circuit, abnormalities of the globus pallidus and ventromedial striatum are thought to precede illness onset, whereas those in the

EP

ventrolateral prefrontal cortex may arise afterwards (Strakowski et al., 2012). Insert Figure 2 here

AC C

Ventromedial circuit. The ventromedial circuit, depicted in Figure 3, includes part of the ventromedial prefrontal cortex defined by BA 11, whose output is directed to the nucleus accumbens and onward to the thalamus. The thalamus completes a feedback loop through its projections back to the ventromedial prefrontal cortex. The ventromedial prefrontal cortex, nucleus accumbens, and thalamus all maintain reciprocal connections to the amygdala. The insula is involved in the ventromedial circuit through its reciprocal communications with the amygdala and the ventromedial


prefrontal cortex (Strakowski et al., 2012). Both the globus pallidus and the nucleus accumbens are thought to exhibit abnormalities antecedent to the onset of bipolar

Insert Figure 3 here

RI PT

disorder (Strakowski et al., 2012).

Dorsolateral circuit. As shown in Figure 4, the dorsolateral circuit includes BA 9

SC

and 10 in the dorsolateral prefrontal cortex, which project to the globus pallidus through the caudate nucleus. This region projects to the ventral anterior and mediodorsal

M AN U

thalamus, which project in turn back to BA 9 and 10 to complete the circuit (Almeida et al., 2009).

Insert Figure 4 here

Anterior cingulate circuit. The dorsal and ventral subdivisions of the anterior

TE D

cingulate appear related to cognitive and affective processing, respectively. The subgenual prefrontal cortex (BA 25) has reciprocal connections with the anterior cingulate cortex and the amygdala, and receives input from the ventromedial prefrontal

EP

cortex, which presumably accounts for its role in integrating cognitive and emotional information (Drevets et al., 1997; Strakowski et al., 2005). The anterior cingulate circuit,

AC C

which is illustrated in Figure 5, originates in BA 24 in the anterior cingulate, which projects to the ventral striatum (Mega & Cummings, 1994). Projections then continue to the ventral pallidum, which in turn projects to the mediodorsal thalamus and then back to the anterior cingulate to complete the circuit. Insert Figure 5 here Corticolimbic circuits and bipolar disorder. The corticolimbic network comprises circuits that can account for many clinical features of bipolar disorder (Strakowski et al.,


2012). The ventromedial circuit is involved in internal emotional control and thus mediates responses to internally generated emotional states. The ventrolateral circuit is involved in responding to external emotional stimuli and regulating emotional

RI PT

expression. The dorsolateral circuit is involved in the integration of higher-order

executive functions, such as judgment, decision-making, attention, reasoning and

decision-making, temporal organization of behavior, working memory, and organization

SC

of information. In healthy adult subjects, the anterior cingulate cortex has been shown to play a role in performance monitoring (Macdonald, Cohen, Stenger, & Carter, 2000),

M AN U

response inhibition (Mega et al., 1994), and some motivated behavior (apathetic states can arise from anterior cingulate dysfunction) (Mega et al., 1994). Taken together, it appears that, in bipolar disorder, dysfunction of prefrontal modulation of the amygdala and control of the striatum may contribute to mood states (Strakowski et al., 2012).

TE D

Functional deficits during euthymia are related to neural changes and resting state functional neuroimaging studies (performed when subjects are not engaged in a particular task) have revealed baseline alterations in cerebral function. For example, in

EP

a group of recovered older patients with bipolar disorder, Brooks et al. (Brooks et al., 2009a) measured resting glucose metabolism with 18FDG-PET and found that patients

AC C

with bipolar disorder, compared to healthy controls, exhibited decreased bilateral dorsolateral prefrontal metabolism in conjunction with increased bilateral metabolism in the amygdala and parahippocampal gyrus. The degrees of prefrontal hypometabolism and parahippocampal hypermetabolism were correlated with poorer memory performance (Brooks et al., 2009b). This pattern of results may reflect eventual


transformation of cerebral metabolic alterations during episodes of depression or mania into more stable trait differences. Potential prognostic biomarkers

RI PT

When patients are initially diagnosed with a depressive episode, it can be challenging, and at times impossible, to determine whether unipolar or bipolar

depression is the accurate diagnosis (Ghaemi, Sachs, Chiou, Pandurangi, & Goodwin,

SC

1999). At least half of patients with bipolar disorder present with an initial episode of depression (Perlis et al., 2004) and no history of hypomania or mania. A clinician is left

M AN U

with an inaccurate initial diagnosis of unipolar depression in such patients. Indeed, 60% of bipolar depressed adults who seek treatment are misdiagnosed as unipolar depression, and in 35% of patients with bipolar disorder it can take over ten years to receive an accurate diagnosis (Hirschfeld, Lewis, & Vornik, 2003). Misdiagnosis of

TE D

bipolar disorder as unipolar depression and the ensuing treatment decisions can lead to precipitation of hypomanic, manic, or mixed states (Altshuler et al., 1995) and result in a worse overall outcome (Goldberg & Ernst, 2002).

EP

Because different mood states characterize bipolar disorders, clues for potential prognostic biomarkers may be found when considering neural changes that are mood-

AC C

state-related as opposed to trait-related, which persist across mood states (Hariri, 2012). Several in-depth reviews of functional imaging in bipolar disorder have reviewed changes associated with mood states (Altshuler et al., 2012; Cerullo, Adler, Delbello, & Strakowski, 2009; Chen et al., 2011). Many of the early studies of functional brain changes associated with bipolar depression have used PET, though fMRI studies have increasingly been seen in recent investigations in bipolar depressed adults. The PET


studies we describe below reported differences in resting metabolism—measurements

the fMRI studies, which assess neural activity during a task.

RI PT

taken when subjects are not engaged in an experimental task—in contrast to most of

Neural changes associated with bipolar depression. Studies of bipolar

depression have detected decreased resting prefrontal metabolism with 18FDG-PET.

SC

Relative to healthy controls, a group of treatment-resistant, mostly rapid-cycling, medication-free patients diagnosed with bipolar depression exhibited decreased

M AN U

metabolism in the dorsolateral prefrontal cortex (Ketter et al., 2001). But, this pattern of results may have been specific to the use of a treatment-resistant subject sample, and thus not representative of bipolar depression. Brooks et al. (Brooks et al., 2009c) later studied a broader group of medication-free, depressed bipolar disorder patients who,

TE D

when compared to healthy controls, exhibited decreased metabolic rates in dorsolateral prefrontal cortex (BA 10, 46), medial orbital prefrontal cortex (BA 10, 11), anterior cingulate (BA 24, 32), and subgenual prefrontal cortex (BA 25). These findings were

EP

partially replicated in a sample of mostly medicated patients with bipolar depression (Hosokawa, Momose, & Kasai, 2009).

AC C

Some studies have found evidence of limbic (amygdala, hippocampus,

parahippocampal gyrus) hypermetabolism in depression, although these findings are inconsistent. For example, Ketter et al. (Ketter et al., 2001) reported amygdala hypermetabolism using 18FDG-PET in patients diagnosed with treatment-resistant and/or rapid-cycling bipolar disorder. Amygdala hypermetabolism has similarly been reported in other studies of unmedicated patients with bipolar depression who were


treatment resistant (Drevets, Bogers, & Raichle, 2002). In a study of patients with bipolar depression, Brooks et al. (Brooks et al., 2009c) did not find evidence of amygdala hypermetabolism in non-treatment resistant patients (although there was

RI PT

evidence of dorsolateral prefrontal hypometabolism as discussed above). Thus, limbic hypermetabolism may be unique to treatment-resistant or rapid-cycling bipolar depression.

SC

A meta-analysis of face emotion processing fMRI studies in bipolar subjects

M AN U

versus healthy adults found increased activity in the parahippocampal gyrus (extending to the amygdala) bilaterally and decreased ventrolateral prefrontal cortex (BA 47) activity in bipolar compared to healthy controls (Delvecchio et al., 2012). The finding that abnormal limbic activation in bipolar disorder centered on the parahippocampal gyrus is consistent with results from a previous meta-analysis (Chen et al., 2011), and

TE D

therefore may raise questions about current “amygdalocentric” models in bipolar disorder. However, one methodological consideration of the Delvecchio et al. (Delvecchio et al., 2012) meta-analysis is that the authors included studies of bipolar

EP

subjects in different mood states, whereas other emotion-processing reviews have

AC C

focused on depressed subjects (Townsend & Altshuler, 2012). Our group reported that, when engaged in a face-matching task subjects that

activates the amygdala, subjects with bipolar I depression, relative to healthy subjects, exhibited significant decreases in bilateral ventrolateral prefrontal cortex (BA 47) and right dorsolateral prefrontal cortex activation as well as increased activation of the left lateral orbitofrontal cortex (BA 10) (Altshuler et al., 2008). However, we found no significant between-group amygdala differences to fearful or angry faces, which is


consistent with other studies that similarly probed amygdala response to these same stimuli in acutely depressed bipolar subjects (Almeida, Versace, Hassel, Kupfer, & Phillips, 2010; Chen et al., 2006; Fournier, Keener, Almeida, Kronhaus, & Phillips, 2013;

RI PT

Van der Schot, Kahn, Ramsey, Nolen, & Vink, 2010). In contrast, emotion generation (i.e., viewing captioned pictures designed to produce positive or negative emotion) was associated with distinct patterns of regional activation were found for bipolar depressed

SC

as compared to healthy subjects entailing the activation of additional subcortical regions including the amygdala, thalamus, hypothalamus, and globus pallidus (Malhi et al.,

M AN U

2004). More recently, studies have demonstrated abnormally elevated right amygdala activity to negative (fearful or angry) faces (Hulvershorn et al., 2011; Perlman et al., 2012) and elevated left amygdala activity to mild sad and neutral expressions (Almeida et al., 2010) in bipolar I depressed versus healthy controls. Thus, while amygdala

TE D

reactivity to negative emotional stimuli is variable, abnormally decreased ventrolateral prefrontal cortex activity in bipolar depression is more consistent (Townsend et al., 2012).

EP

Amygdala response to positive emotional stimuli (i.e., happy facial expressions) has been explored in a few fMRI studies of bipolar depression (Almeida et al., 2010;

AC C

Almeida et al., 2009; Fournier et al., 2013; Grotegerd et al., 2013a) and all but one reported no significant-between group differences in amygdala activation. A recent review (Phillips et al., 2014) noted a pattern of abnormally increased amygdala, striatal, and medial prefrontal cortex activity to positively-valenced stimuli in bipolar disorder, but this was primarily observed in samples of remitted or partially remitted bipolar subjects (Blumberg et al., 2005; Keener et al., 2012; Lawrence et al., 2004; Surguladze et al.,


2010). Thus, differential amygdala response to positive emotional stimuli may depend on depression severity.

RI PT

There have been only two fMRI studies of emotional processing in bipolar II depression. One study required participants to view happy, fearful, and neutral faces and did not find any significant differences between depressed bipolar II subjects and healthy subjects (Marchand et al., 2011) although the inclusion of only male participants

SC

limits generalizablility. A more recent study that included unmedicated bipolar II

M AN U

depressed subjects reported significant bilateral hypoactivation in ventrolateral prefrontal cortex and right amygdala (Vizueta et al., 2012) compared to controls during an emotional face-matching task. Further, bipolar II depressed subjects demonstrated significantly reduced negative functional connectivity between the right amygdala and the right orbitofrontal cortex (BA 10) as well as the right dorsolateral prefrontal cortex

TE D

relative to healthy controls. This suggests that bipolar II depression is characterized by reduced regional orbitofrontal (BA 47) and limbic activation and altered connectivity in a fronto-temporal circuit. Increased positive functional connectivity between the amygdala

EP

and BA 47 region of the orbitofrontal cortex has been reported in bipolar I and II depressed patients (Versace et al., 2010; Vizueta et al., 2012), which may reflect

AC C

inappropriate inhibition of the amygdala in bipolar depression. Interestingly, there is evidence that, compared to euthymic bipolar I disorder subjects, euthymic bipolar II disorder subjects exhibit increased ventral striatal activity when anticipating a reward, thus suggesting other differences between the subtypes (Caseras, Lawrence, Murphy, Wise, & Phillips, 2013).


Neural changes associated with mania. The clinical challenges of studying mania have limited studies to relatively small sample sizes. Because mania has prominent effects on cerebral metabolism and blood flow, even functional neuroimaging studies 18

FDG-PET

RI PT

with small samples have been able to detect mood state-related changes.

imaging studies of patients with bipolar mania have revealed hypermetabolism in

anterior cingulate (Blumberg et al., 2000) and medial temporal structures (Rubinsztein

SC

et al., 2001). Hypermetabolism has also been observed in parahippocampal cortex in manic subjects in comparison to healthy controls (Brooks, Hoblyn, & Ketter, 2010a).

M AN U

Somewhat more commonly, studies have reported resting prefrontal hypometabolism in medial, dorsolateral, and subgenual regions of prefrontal cortex (al-Mousawi et al., 1996; Blumberg et al., 1999; Brooks et al., 2010a; Drevets et al., 1997). However, most of these studies have reported regions of either hypo- or hypermetabolism, though

TE D

one study captured simultaneous prefrontal hypometabolism and limbic hypermetabolism (Brooks et al., 2010a).

Studies using fMRI have reported network connectivity alterations associated with

EP

mania. In a study of 40 manic bipolar patients, Strakowski et al. (Strakowski et al., 2011) found that, in response to emotional cues, manic patients exhibited a blunted

AC C

response throughout a ventrolateral prefrontal emotional arousal network presumed to modulate increased amygdala output. Blunted prefrontal response was also observed in manic patients during an attention task (Fleck et al., 2012). In addition to ventrolateral prefrontal hypoactivation, orbitofrontal hypoactivation has been reported during mania (Altshuler et al., 2005a; Altshuler et al., 2005b).


Amygdala hyperactivation during mania has been reported in several studies (Altshuler et al., 2005a; Bermpohl et al., 2009; Blumberg et al., 2003; Strakowski et al., 2011). Findings of decreased functioning of ventrolateral prefrontal cortex (Altshuler et

RI PT

al., 2005a; Altshuler et al., 2005b; Blumberg et al., 2003; Elliott et al., 2004; Foland et al., 2008; Mazzola-Pomietto, Kaladjian, Azorin, Anton, & Jeanningros, 2009) in conjunction with reduced negative connectivity between the amygdala and the

SC

ventrolateral prefrontal cortex in bipolar manic subjects have led some researchers to speculate that deficits in prefrontal inhibition may result in increased amygdala reactivity

M AN U

(Foland et al., 2008).

Prefrontal hypoactivation and limbic hyperactivation during mania have been associated with cognitive deficits. For example, fMRI studies have suggested that attentional impairment during mania may reflect dorsolateral prefrontal cortex

TE D

hypoactivation, and decreased activation of the ventrolateral prefrontal network may result affect the ability to modulate increased amygdala output (Fleck et al., 2012). The orbitofrontal cortex also has reciprocal connections with the amygdala (see Figure 1),

EP

and may also modulate output. Studies using 18FDG-PET (Rubinsztein et al., 2001) or fMRI (Altshuler et al., 2005b; Elliott et al., 2004) have reported decreased orbitofrontal

AC C

metabolism or activation in manic patients. Distinguishing between bipolar and unipolar depression. Prognostic biomarkers

could be used to predict disease course, but could also aid in differential diagnosis. Most functional neuroimaging research in mood disorders has focused on either subjects with bipolar or unipolar depression and not compared the two, although recent reviews have highlighted the importance of this comparison (Cardoso de Almeida &


Phillips, 2013; Delvecchio et al., 2012). Table 1 includes findings from twelve taskbased fMRI studies that directly compared bipolar depressed and unipolar depressed individuals (Almeida et al., 2010; Almeida et al., 2009; Bertocci et al., 2012; Chase et

RI PT

al., 2013; Diler et al., 2013a; Diler et al., 2014; Fournier et al., 2013; Grotegerd et al.,

Radaelli et al., 2013; Taylor Tavares et al., 2008). Insert Table 1 here

SC

2013a; Lawrence et al., 2004; Marchand, Lee, Johnson, Gale, & Thatcher, 2013;

One study found that remitted bipolar subjects with elevated subsyndromal

M AN U

depressive symptoms exhibited greater amygdala and ventrolateral prefrontal cortex activation to fearful and happy expressions relative to healthy controls and unipolar depressed individuals (Lawrence et al., 2004). The remitted bipolar disorder subjects showed greater ventrolateral prefrontal cortex and dorsal anterior cingulate activity to

TE D

sad faces than unipolar depressed subjects, whereas the unipolar depressed group showed greater activity only in the putamen. However, another study that used an affect labeling found abnormally increased left amygdala activity to mild, sad, and

EP

neutral faces in acutely depressed bipolar subjects compared to unipolar depressed, healthy controls, and remitted bipolar subjects (Almeida et al., 2010). Amygdala

AC C

hypereactivity may be a depression-specific marker of bipolar but not unipolar depression, and may reflect greater attentional demands specifically for stimuli conveying internal distress or threat ambiguity. Abnormally increased amygdala reactivity itself is likely not specific to bipolar depression, as it has also been reported in anxiety disorders (Shin & Liberzon, 2010). Additionally, recent studies have failed to find increased amygdala reactivity in bipolar depressed compared to unipolar depressed


subjects to either naturalistic, dynamically changing sad faces (Fournier et al., 2013) or subliminally presented sad faces (Grotegerd et al., 2013a). Rather, greater amygdala reactivity specifically to angry (but not happy, sad, or fearful) faces was reported in

RI PT

unipolar depressed relative to bipolar depressed individuals (Fournier et al., 2013). This may mean that unipolar depressed subjects perceive angry faces as more threatening, whereas exaggerated amygdala reactivity in bipolar depression (Almeida et al., 2010)

SC

may reflect the amygdala’s sensitivity to ambiguity, particularly when a valence

judgment is required for resolution of ambiguity (Neta, Kelley, & Whalen, 2013).

M AN U

Studies comparing bipolar and unipolar depression have usually used supraliminal emotional faces, which may invoke an appraisal process involving prefrontal cortical regions that interact with the amygdala. Grotegerd et al. (Grotegerd et al., 2013a), used a backward-masking of emotional faces to examine automatic processing specific to the

TE D

amygdala. They reported increased amygdala reactivity to backward-masked sad faces and decreased amygdala reactivity to masked happy faces in unipolar compared to bipolar depressed subjects (Grotegerd et al., 2013a). This finding is consistent with an

EP

exaggerated amygdala reactivity to negative feedback reported in unipolar depressed (but not bipolar depressed subjects) relative to healthy controls (Taylor Tavares et al.,

AC C

2008), which may reflect a reduced ability of the ventrolateral prefrontal cortex to regulate the amygdala in unipolar depression. Because only one study has compared unipolar and bipolar subjects using a backward-masking technique, these results warrant replication.

In contrast to whole-brain and region-of-interest fMRI studies comparing unipolar and bipolar subjects, there is limited work with functional connectivity techniques.


Connectivity analyses assess whether task-dependent changes in one region predict activity of another, and may therefore help reconcile the divergent findings observed in the amygdala, a region that is highly interconnected with other brain structures

RI PT

subserving emotion regulation (Kim et al., 2011). To this end, Almeida et al. (Almeida et al., 2009) found decreased left ventromedial prefrontal cortical-amygdala effectivity connectivity with happy faces in bipolar compared to healthy controls and decreased

SC

right amygdala-ventromedial prefrontal positive connectivity. Conversely, unipolar depressed subjects demonstrated abnormally increased negative ventromedial-

M AN U

amygdala connectivity relative to healthy controls during processing of happy facial stimuli. Thus, overregulation of the amygdala by the ventromedial prefrontal cortex over the amygdala in unipolar (but not bipolar) depressed subjects, may lead to diminished amygdala responsiveness to happy faces in unipolar depressed patients.

TE D

A small-scale study comparing functional connectivity in depressed patients diagnosed with either bipolar II disorder or unipolar depression found differential connectivity between the right posterior cingulate and right parietal/insular regions for

EP

the two groups during performance of a motor activation task (Marchand et al., 2013). A resting state fMRI study found that bipolar and unipolar depressed subjects exhibited

AC C

decreased resting state corticolimbic connectivity between the pregenual anterior cingulate cortex and dorsomedial thalamus regions relative to healthy controls (Anand et al., 2009) but only included five bipolar depressed subjects and may have lacked sufficient statistical power to detect differences between the patient groups. As summarized in Table 1, recent fMRI studies have used cognitive or reward processing tasks to compare regional neural activation patterns in bipolar and unipolar


depression. During the more challenging 2-back memory load condition of an N-back paradigm, abnormally elevated dorsal anterior mid-cingulate cortical activity was found for unipolar depressed subjects, compared to both bipolar depressed and healthy

RI PT

subjects, which may suggest abnormal recruitment of attentional control circuitry in unipolar depression to sustain equivalent task performance (Bertocci et al., 2012).

A recent study that used a moral valence decision task reported abnormally

SC

increased ventrolateral prefrontal cortex activation for negative relative to positive

stimuli in unipolar depressed subjects, whereas bipolar subjects demonstrated the

M AN U

opposite pattern (Radaelli et al., 2013). Taken together with findings from Taylor Tavares et al. (Taylor Tavares et al., 2008) and Almeida et al. (Almeida et al., 2009), functional abnormalities in the ventrolateral prefrontal cortex that are modulated by the valence of emotional stimuli may distinguish bipolar and unipolar depressed subjects.

TE D

However, additional work employing functional connectivity analyses to examine ventrolateral prefrontal cortex activation to positive and emotional stimuli in both bipolar and unipolar depressed subtypes would be needed before reaching a definitive

EP

conclusion.

Summary and limitations. Amygdala findings to happy and fearful faces

AC C

presented using standard (visible) stimulus processing paradigms largely have not detected significant differences between depressed adults/adolescents with bipolar depression compared to those with unipolar depression. The studies that have been conducted further suggest that bipolar I depressed adults relative to unipolar depressed adults exhibit greater amygdala response to mood-congruent (sad) faces, and decreased abnormally reduced bottom-up amygdala-ventromedial prefrontal effective


connectivity to overt happy faces. The finding of increased amygdala reactivity in bipolar depression to sad faces is tempered by its restriction to potentially ambiguous (mild) sad faces, as this effect has not been found in several studies employing

RI PT

prototypical sad faces. In contrast, there is preliminary evidence that unipolar

depressed individuals relative to bipolar I depressed subjects show increased amygdala activation to negative feedback, angry or masked sad faces, and reduced amygdala

SC

activation to masked happy faces. Moreover, unipolar (but not bipolar) depressed

subjects exhibit abnormal recruitment of ventrolateral prefrontal cortical circuitry to

M AN U

negative feedback as well as abnormal recruitment of attentional control circuitry to direct attention away from ambiguous, neutral face distractors. Because most studies comparing unipolar depressed to bipolar depressed subjects did not include data from remitted subjects, it is unclear whether observed

TE D

between-group differences represent state or trait-related effects. There have been remarkably few prior fMRI studies that have directly compared bipolar II and unipolar

EP

depression. Because most patients with bipolar II disorder present for treatment when depressed rather than hypomanic, it is not surprising that bipolar II is

AC C

commonly misdiagnosed as unipolar major depressive disorder, leading to inappropriate treatment (Rastelli, Cheng, Weingarden, Frank, & Swartz, 2013). This highlights the need for work to identify prognostic biomarkers differentiating bipolar II from unipolar depression. Future research employing advanced connectivity with


larger sample sizes, including more evenly matched samples of males and females to explore possible gender differences, are needed to compare unipolar and bipolar

RI PT

depressed individuals.

Distinguishing between bipolar disorder and other illnesses. A psychiatric

SC

interview can detect the presence of depressive symptoms and thus a prognostic biomarker would not be needed in this sense. However, disease-related neural

M AN U

differences between subjects with a psychiatric illness and those without may lead to identification of targets for treatment personalization (cf. (Price, Paul, Schneider, & Siegle, 2013)), which would be of clinical importance. Recent work with trainable classification algorithms, such as support vector machines (SVM), has used functional

TE D

neuroimaging data with pattern classification to create discrimination algorithms to classify binary outcomes or diagnoses (for reviews, see (Fu & Costafreda, 2013; Orrù, Pettersson-Yeo, Marquand, Sartori, & Mechelli, 2012). The purpose of an SVM is to

EP

incorporate machine learning to determine the optimal boundary that discriminates between two groups.

AC C

A recent study using pattern classification techniques in adults found that the

pattern of neural activity for intense happy faces was significantly less distinct from that for neutral faces in the bipolar I depressed group compared to either healthy controls or unipolar depressed patients (Mourao-Miranda et al., 2012). Another SVM study found that neural activation during a verbal fluency task discriminated between healthy subjects and patients with schizophrenia 92% of the time and euthymic patients with


bipolar disorder 79% of the time (Costafreda et al., 2011). Finally, another study, albeit with a small sample size, applied an SVM approach to fMRI activation and achieved up to 90% correct classification of patients diagnosed with either bipolar I or unipolar

RI PT

depression (Grotegerd et al., 2013b). Thus, SVM approaches illustrate that functional neuroimaging data can be used to accurately classify patients into predefined groups, such as patients and controls or treatment responders and non-responders.

SC

Summary and limitations. When considered collectively, available evidence

suggests increases in amygdala activation during mania, and in some studies of bipolar

M AN U

I depression. Although amygdala activation may be relatively normal during euthymia in bipolar patients, there is evidence of amygdala hypermetabolism during euthymic periods—at least in older subjects with bipolar disorder. In contrast, dorsolateral and orbitofrontal regions of prefrontal cortex appear to be either hypoactive and/or

TE D

hypometabolic across mood states in bipolar I and II patients, which suggests that prefrontal cortex activation may be state-independent (i.e., decreased regardless of

EP

mood states) and thereby represent a trait feature of bipolar disorder. Studies using traditional fMRI task-based mapping analyses have yet to

AC C

determine a unique neural signature of bipolar depression, which precludes using this measure for differential diagnosis. The use of effective connectivity techniques that examine functional integration between brain regions and provide information about the direction of interactions may be a more promising approach to reveal a unique neural signature of bipolar disorder than standard region-of-interest or whole-brain analyses (see section entitled, Future directions of functional neuroimaging in bipolar disorder). There is evidence of between-group differences in amygdala activation, prefrontal


cortex activation, and amygdala-prefrontal connectivity in bipolar depressed relative to unipolar depressed individuals, it is unclear how these observed patterns of neural activation may relate to clinical or treatment measures.

RI PT

There are several limitations in functional neuroimaging studies conducted to date that temper the conclusions from these studies. First, the majority of studies of state-related neural changes in bipolar disorder included medicated patients. Although

SC

medication effects in neuroimaging studies have been a topic of debate (Hafeman,

Chang, Garrett, Sanders, & Phillips, 2012; Phillips, Travis, Fagiolini, & Kupfer, 2008),

M AN U

the use of unmedicated patients increases the risk of sampling bias because patients diagnosed with bipolar disorder who are able to function without medication for extended periods of time may not be representative of bipolar patients as a whole. Another limitation involves the cross-sectional approach used in most studies, in

TE D

which different groups of patients are compared in different mood states, and/or the subtypes of bipolar disorder have been conflated. Though cross-sectional studies of bipolar disorder are more feasible than longitudinal ones, they do not provide as much

EP

statistical power. Moreover, the inconsistent results observed in bipolar studies—

AC C

particularly regarding the amygdala in bipolar depression—may reflect task design differences coupled with the range of affective stimuli used. Within-subject longitudinal studies in bipolar disorder using the same well-validated experimental paradigm are needed to make definitive conclusions regarding trait- versus mood-state related neural findings in bipolar disorder (Hariri, 2012) and to elucidate the triggers for mood transitions as patients traverse manic, depressed, and euthymic episodes.


Potential predictive biomarkers Patients diagnosed with bipolar disorder do not respond equally to a given treatment, which suggests the presence of underlying dimensions that are not captured

RI PT

in current clinical nosology. Functional neuroimaging may help distinguish differences and similarities in underlying neurobiology among patients within and across diagnoses. Identification of such biomarkers can eventually provide future targets for disease

SC

treatment.

There have been few studies of potential predictive biomarkers in bipolar

M AN U

disorder. One study of adolescent patients diagnosed with bipolar mania reported that greater deactivation in right BA 47 at baseline was associated with subsequent treatment response to ziprasidone (Schneider et al., 2012). Another study, which used face labeling tasks, found that greater baseline fMRI activity in the ventral anterior

TE D

cingulate cortex in bipolar depressed adolescents was associated with greater improvement in depression after 6-weeks of open treatment (Diler et al., 2013b). To our knowledge, there are no such prognostic neuroimaging studies in adults

EP

with bipolar disorder. However, several studies in adults with unipolar depression suggest that predictive neuroimaging biomarkers may exist. Some of these studies

AC C

used 18FDG-PET and found regional brain metabolism could predict response to antidepressant (Kennedy et al., 2007; Little et al., 1996; McGrath et al., 2013; Saxena et al., 2003) or behavioral treatment (Konarski et al., 2009; Volk et al., 1997). Notably, a recent PET study of major depression reported that insula hypometabolism was associated with response to behavioral therapy, but not an antidepressant whereas the opposite response pattern held for insula hypermetabolism (McGrath et al., 2013).


While fMRI has not been employed to optimize treatment selection in unipolar depression, several studies have reported relations between baseline neural activation and positive treatment response (Costafreda, Khanna, Mourão-Miranda, & Fu, 2009;

al., 2009; Siegle, Carter, & Thase, 2006; Siegle et al., 2012).

RI PT

Davidson, Irwin, Anderle, & Kalin, 2003; Frodl et al., 2011; Fu et al., 2008; Salvadore et

Summary and limitations. At present, functional neuroimaging studies are in the

SC

early phases of predicting treatment response and do not provide measures that are sufficiently reliable for clinical use. Findings in pediatric bipolar and adult unipolar

M AN U

depression samples are encouraging in that predictive relations exist between neural activity and treatment outcome, and thus warrant further exploration in adult bipolar disorders as well. Potential treatment biomarkers

TE D

While fMRI studies have identified treatment-induced neural changes associated with mood improvement (Arnone et al., 2012; Fales et al., 2009; Fu et al., 2004; Heller et al., 2013a; Heller et al., 2013b; Light et al., 2011; Sheline et al., 2001; Victor, Furey,

EP

Fromm, Ohman, & Drevets, 2010), similar studies in bipolar adults during an acute depressive episode are lacking. One recent study (Ives-Deliperi, Howells, Stein,

AC C

Meintjes, & Horn, 2013) in adults with bipolar I and II disorder experiencing mild or subthreshold depressive symptoms found that mindfulness-based cognitive therapy increased activation in the medial prefrontal cortex during a mindfulness task, and signal changes in this region were strongly associated with increases in mindfulness (Ives-Deliperi et al., 2013). Another study with euthymic bipolar adults reported increased activation in the bilateral inferior frontal gyri during performance of a word-


face emotional Stroop task after three months of psychoeducation therapy (Favre et al., 2013). Several studies have directly explored medication effects with functional imaging in children or adolescents diagnosed with bipolar disorder (Chang, Wagner, Garrett,

RI PT

Howe, & Reiss, 2008; Diler et al., 2013c; Pavuluri, Passarotti, Fitzgerald, Wegbreit, & Sweeney, 2012; Yang et al., 2013).

Although studies conducted in pediatric samples may not generalize to adult

SC

patients, they demonstrate the potential of functional neuroimaging to explain underlying treatment mechanisms. One study in pediatric bipolar disorder used fMRI to quantify

M AN U

brain response during an affective processing task before and after risperidone or divalproex treatment for a manic episode (Pavuluri et al., 2012). Risperidone treatment was associated with increased insula engagement during the affective task whereas treatment with divalproex was associated with increased subgenual prefrontal network

TE D

engagement. In a study of bipolar manic adolescents, baseline activation in right BA 47 was negatively correlated with improvement in manic symptoms (Schneider et al., 2012). Another study of children with bipolar disorder found that lamotrigine treatment

EP

was associated with decreased amygdala activation when viewing negative minus neutral stimuli as depressive symptoms subsided (Chang et al., 2008). In a study of

AC C

bipolar depressed adolescents completing a motor response inhibition task, patients demonstrated greater hyper-reactivity at baseline in bilateral ventrolateral prefrontal cortex and left superior temporal regions compared to healthy controls that did not change after treatment; however, treatment increased activity in subcortical regions (right hippocampus and left thalamus) in depressed adolescents, and lower thalamus activity at baseline was correlated with higher depression scores (Diler et al., 2013c).


Lastly, there is evidence that dorsolateral prefrontal cortex activation may be more readily altered by short-term pharmacotherapy in pediatric bipolar disorder than activation in the ventrolateral prefrontal cortex, anterior cingulate, amygdala, and ventral

RI PT

striatum (Yang et al., 2013). Specifically, normalization of increased baseline

dorsolateral prefrontal activation in response to emotional vs. neutral words in pediatric bipolar disorder subjects was observed by 16 weeks of pharmacotherapy, whereas

SC

increased amygdala, striatal, anterior cingulate and ventrolateral prefrontal activation normalized by three years of treatment (Yang et al., 2013).

M AN U

Summary and limitations. Investigations of pediatric bipolar disorder show promise in detecting neural alterations associated with pharmacotherapy (Fu et al., 2013), but these findings have not been replicated in adult populations. While studies in bipolar adults using fMRI to evaluate the effects of behavioral interventions are

TE D

underway, studies to date have been conducted in remitted or partially remitted bipolar subjects (Demant, Almer, Vinberg, Kessing, & Miskowiak, 2013; Favre et al., 2013; Ives-Deliperi et al., 2013). Nonetheless, they provide preliminary evidence that

EP

mindfulness-based cognitive therapy and psychoeducational therapy increase prefrontal cortex activation in bipolar disorder and thereby warrant replication with patients

AC C

experiencing acute bipolar depression. Regardless of whether treatment is pharmacological or behavioral, exploration of neural changes of response to treatment is confounded with change in mood. Future studies should examine both responders and non-responders to evaluate effects due to medication exposure alone versus mood improvement. Other factors, such as mood-state related changes, may moderate the


neural effects of treatment and must be accounted for in future work before functional neuroimaging finds its place as an index of therapeutic response. Future directions of functional neuroimaging in bipolar disorder

RI PT

Functional neuroimaging data in bipolar disorder has been interpreted within a corticolimbic network model and studies have provided largely compelling evidence of dysfunction of corticolimbic regions during different mood states of bipolar disorder.

SC

Given the important role that the prefrontal cortex plays in exerting top-down regulatory control over the amygdala (Hariri, Bookheimer, & Mazziotta, 2000; Hariri, Mattay,

M AN U

Tessitore, Fera, & Weinberger, 2003; Ochsner, Bunge, Gross, & Gabrieli, 2002), the observed prefrontal cortex hypoactivation in bipolar disorder may underlie vulnerability to experiencing future manic or depressed episodes as a result of an impaired emotion regulatory network. Here, the interplay between brain regions, or functional

TE D

connectivity, as it relates to biomarkers is important, because dysregulated network communications may explain observed regional changes in brain activation. Along these lines, a resting-state functional connectivity study revealed that, relative to healthy

EP

subjects, bipolar disorder subjects exhibited greater coupling between the right amygdala and ventrolateral prefrontal cortex that was mediated by the anterior cingulate

AC C

cortex (Torrisi et al., 2013). Further, connectivity findings in task-based fMRI studies highlight the importance of altered functional integration of emotion-regulatory brain circuits or networks and may be equally as important as knowledge of differences in regional brain activation.

Though functional connectivity studies employing emotion processing and emotion regulation tasks in bipolar disorder suggest altered connectivity in a fronto-


temporal circuit compared to healthy controls, these findings reflect correlations between activities in different neural regions over time, which preclude inferences about directionality. That is, significantly reduced negative connectivity between the amygdala

RI PT

and orbitofrontal cortex in bipolar mania may indicate that either higher activation in prefrontal/frontal cortical regions is associated with lower activation in the amygdala, suggesting top-down regulatory control of the orbitofrontal cortex over the amygdala, or

SC

greater amygdala activation is associated with lower activation in frontal regions.

Here, advanced effective connectivity techniques, which remain underutilized in

M AN U

fMRI studies with bipolar disorder, provide additional information regarding the direction of the inference or the impact that activity in one region exerts over that in another and can be used to estimate forward (bottom-up) versus backward (top-down) connectivity between regions (Friston, Harrison, & Penny, 2003; Roebroeck, Formisano, & Goebel,

TE D

2011). For example, relative to healthy participants, depressed bipolar disorder subjects exhibited significantly greater negative right amygdala-orbitomedial prefrontal effective connectivity and reduced positive left orbitomedial prefrontal-amygdala

EP

effective connectivity during facial expression labeling (Almeida et al. 2010). Conversely, patients with unipolar disorder, compared to healthy controls, showed

AC C

significantly greater negative left orbitomedial prefrontal–amygdala effective connectivity, possibly reflecting over-regulation of the amygdala by the orbitomedial prefrontal cortex. Thus, insufficient amygdala regulation by the orbitomedial prefrontal cortex (BA 10) may represent a predisposition to elevated mood in bipolar disorder. Lastly, multimodal neuroimaging studies to identify structure-function relationships would be useful as another avenue of future research in bipolar disorder to examine


whether the observed functional abnormalities in prefrontal cortical-amygdala circuity map on to structural abnormalities in these same regions. Summary

RI PT

Results of functional neuroimaging studies suggest the potential existence of biomarkers in bipolar disorders, but there are no established prognostic, predictive, or treatment biomarkers to date. Though some consensus emerges from prior work,

SC

reaching the level of qualification for a biomarker presents distinct challenges. One challenge, as Savitz et al. (Savitz et al., 2013) note, is that a biomarker should be able

M AN U

to achieve at least 80% positive predictive value (the number of true positives divided by the sum of true and false positives), a standard that current work has yet to meet with an independent cohort of subjects. The predictive threshold is also influenced by testretest reliability, which is not well-established for functional neuroimaging measures

TE D

(Lipp, Murphy, Wise, & Caseras, 2014; Sauder, Hajcak, Angstadt, & Phan, 2013). Similarly, a methodological challenge for connectivity studies of frontolimbic circuitry is that varying results may be obtained depending on the analytic connectivity strategy

EP

employed (i.e. a between-group psychophysiological interaction (PPI) analysis of task related functional connectivity versus global connectivity across all different conditions)

AC C

(Schwartz, Holt, & Rosenbaum, 2013). Before functional neuroimaging can augment differential diagnosis of bipolar disorder or distinguish between bipolar subtypes, additional research is needed. As we reviewed above, certain neuroimaging findings, such as altered regional

metabolism or activation, are related to treatment outcome. With further refinement and an increased focus on network connectivity, functional neuroimaging may eventually aid


selection of the optimal treatment as well as providing a physiological index of treatment response, but at present functional neuroimaging is not able to provide such an index. Newer approaches related to network connectivity, perhaps combined with behavioral

personalized psychiatric care in bipolar disorder.

RI PT

measures may improve the utility of neuroimaging as a critical component of

An additional challenge in the identification of biomarkers of bipolar disorder rests

SC

on the possibility that current diagnostic systems that rely on clinical symptoms include considerable heterogeneity within the same patient research groups and may not

M AN U

capture the dimensional nature of psychiatric illness. This notion is captured in the more general Research Domain Criteria (RDoC) approach adopted by the National Institute of Mental Health. The intention of the RDoC initiative is to lay the foundation for a new framework for conceptualizing psychiatric illness in terms of biobehavioral

TE D

dimensions, as opposed to clinical observation (Cuthbert & Insel, 2013; Insel et al., 2010). To the extent that current diagnoses do not correspond to neurophysiologically distinct disorders, searches for neural biomarkers will be challenging or impossible

EP

because the diagnostic classification does not reflect the underlying neuropathology. Defining dimensions of psychiatric illness through objective measures could eventually

AC C

increase the future utility of functional neuroimaging in clinical are.


References

AC C

EP

TE D

M AN U

SC

RI PT

Adler, C. M., Cerullo, M. A., & Strakowski, S. M. (2012). Brain imaging techniques and their application to bipolar disorder, The Bipolar Brain (pp. 3-16): Oxford University Press. Adler, C. M., Delbello, M. P., & Strakowski, S. M. (2006). Brain network dysfunction in bipolar disorder., CNS Spectr (pp. 312-320). al-Mousawi, A. H., Evans, N., Ebmeier, K. P., Roeda, D., Chaloner, F., & Ashcroft, G. W. (1996). Limbic dysfunction in schizophrenia and mania. A study using 18Flabelled fluorodeoxyglucose and positron emission tomography., Br J Psychiatry (pp. 509-516). Almeida, J. R. C., Versace, A., Hassel, S., Kupfer, D. J., & Phillips, M. L. (2010). Elevated amygdala activity to sad facial expressions: a state marker of bipolar but not unipolar depression., Biol. Psychiatry (pp. 414-421). Almeida, J. R. C. d., Versace, A., Mechelli, A., Hassel, S., Quevedo, K., Kupfer, D. J., & Phillips, M. L. (2009). Abnormal amygdala-prefrontal effective connectivity to happy faces differentiates bipolar from major depression., Biol. Psychiatry (pp. 451-459). Altshuler, L., Bookheimer, S., Proenza, M. A., Townsend, J., Sabb, F., Firestine, A., Bartzokis, G., Mintz, J., Mazziotta, J., & Cohen, M. S. (2005a). Increased amygdala activation during mania: a functional magnetic resonance imaging study., Am J Psychiatry (pp. 1211-1213). Altshuler, L., Bookheimer, S., Townsend, J., Proenza, M. A., Sabb, F., Mintz, J., & Cohen, M. S. (2008). Regional brain changes in bipolar I depression: a functional magnetic resonance imaging study., Bipolar Disord (pp. 708-717). Altshuler, L. L., Bookheimer, S. Y., Townsend, J., Proenza, M. A., Eisenberger, N., Sabb, F., Mintz, J., & Cohen, M. S. (2005b). Blunted activation in orbitofrontal cortex during mania: a functional magnetic resonance imaging study., Biol. Psychiatry (pp. 763-769). Altshuler, L. L., Post, R. M., Leverich, G. S., Mikalauskas, K., Rosoff, A., & Ackerman, L. (1995). Antidepressant-induced mania and cycle acceleration: a controversy revisited., Am J Psychiatry (pp. 1130-1138). Altshuler, L. L., & Townsend, J. D. (2012). Functional brain imaging in bipolar disorder, The bipolar brain (pp. 53-77): Oxford University Press. Anand, A., Li, Y., Wang, Y., Lowe, M. J., & Dzemidzic, M. (2009). Resting state corticolimbic connectivity abnormalities in unmedicated bipolar disorder and unipolar depression., Psychiatry Res (pp. 189-198). Arnone, D., McKie, S., Elliott, R., Thomas, E. J., Downey, D., Juhasz, G., Williams, S. R., Deakin, J. F., & Anderson, I. M. (2012). Increased amygdala responses to sad but not fearful faces in major depression: relation to mood state and pharmacological treatment. Am J Psychiatry 169, 841-50. Bermpohl, F., Dalanay, U., Kahnt, T., Sajonz, B., Heimann, H., Ricken, R., Stoy, M., Hägele, C., Schlagenhauf, F., Adli, M., Wrase, J., Ströhle, A., Heinz, A., & Bauer, M. (2009). A preliminary study of increased amygdala activation to positive affective stimuli in mania., Bipolar Disord (pp. 70-75).


AC C

EP

TE D

M AN U

SC

RI PT

Bertocci, M. A., Bebko, G. M., Mullin, B. C., Langenecker, S. A., Ladouceur, C. D., Almeida, J. R. C., & Phillips, M. L. (2012). Abnormal anterior cingulate cortical activity during emotional n-back task performance distinguishes bipolar from unipolar depressed females., Psychol Med (pp. 1417-1428). Blumberg, H. P., Donegan, N. H., Sanislow, C. A., Collins, S., Lacadie, C., Skudlarski, P., Gueorguieva, R., Fulbright, R. K., McGlashan, T. H., Gore, J. C., & Krystal, J. H. (2005). Preliminary evidence for medication effects on functional abnormalities in the amygdala and anterior cingulate in bipolar disorder. Psychopharmacology (Berl) 183, 308-13. Blumberg, H. P., Leung, H.-C., Skudlarski, P., Lacadie, C. M., Fredericks, C. A., Harris, B. C., Charney, D. S., Gore, J. C., Krystal, J. H., & Peterson, B. S. (2003). A functional magnetic resonance imaging study of bipolar disorder: state- and traitrelated dysfunction in ventral prefrontal cortices., Arch. Gen. Psychiatry (pp. 601609). Blumberg, H. P., Stern, E., Martinez, D., Ricketts, S., de Asis, J., White, T., Epstein, J., McBride, P. A., Eidelberg, D., Kocsis, J. H., & Silbersweig, D. A. (2000). Increased anterior cingulate and caudate activity in bipolar mania., Biol. Psychiatry (pp. 1045-1052). Blumberg, H. P., Stern, E., Ricketts, S., Martinez, D., de Asis, J., White, T., Epstein, J., Isenberg, N., McBride, P. A., Kemperman, I., Emmerich, S., Dhawan, V., Eidelberg, D., Kocsis, J. H., & Silbersweig, D. A. (1999). Rostral and orbital prefrontal cortex dysfunction in the manic state of bipolar disorder., Am J Psychiatry (pp. 1986-1988). Brooks, J. O., Hoblyn, J. C., & Ketter, T. A. (2010a). Metabolic evidence of corticolimbic dysregulation in bipolar mania., Psychiatry Res (pp. 136-140). Brooks, J. O., Hoblyn, J. C., Woodard, S. A., Rosen, A. C., & Ketter, T. A. (2009a). Corticolimbic metabolic dysregulation in euthymic older adults with bipolar disorder., J Psychiatr Res (pp. 497-502). Brooks, J. O., Rosen, A. C., Hoblyn, J. C., Woodard, S. A., Krasnykh, O., & Ketter, T. A. (2009b). Resting prefrontal hypometabolism and paralimbic hypermetabolism related to verbal recall deficits in euthymic older adults with bipolar disorder., Am J Geriatr Psychiatry (pp. 1022-1029). Brooks, J. O., Wang, P. W., Bonner, J. C., Rosen, A. C., Hoblyn, J. C., Hill, S. J., & Ketter, T. A. (2009c). Decreased prefrontal, anterior cingulate, insula, and ventral striatal metabolism in medication-free depressed outpatients with bipolar disorder., J Psychiatr Res (pp. 181-188). Brooks, J. O. I. I. I., Wang, P. W., & Ketter, T. A. (2010b). Functional brain imaging studies in bipolar disorder: Focus on cerebral metabolism and blood flow, Bipolar disorder: Clinical and neurobiological foundations (pp. 200-210): Oxford. Cardoso de Almeida, J. R., & Phillips, M. L. (2013). Distinguishing between unipolar depression and bipolar depression: current and future clinical and neuroimaging perspectives. Biol Psychiatry 73, 111-8. Caseras, X., Lawrence, N. S., Murphy, K., Wise, R. G., & Phillips, M. L. (2013). Ventral Striatum Activity in Response to Reward: Differences Between Bipolar I and II Disorders., Am J Psychiatry.


AC C

EP

TE D

M AN U

SC

RI PT

Cerullo, M. A., Adler, C. M., Delbello, M. P., & Strakowski, S. M. (2009). The functional neuroanatomy of bipolar disorder., Int Rev Psychiatry (pp. 314-322). Chang, K. D., Wagner, C., Garrett, A., Howe, M., & Reiss, A. (2008). A preliminary functional magnetic resonance imaging study of prefrontal-amygdalar activation changes in adolescents with bipolar depression treated with lamotrigine., Bipolar Disorders (pp. 426-431). Chase, H. W., Nusslock, R., Almeida, J. R., Forbes, E. E., Labarbara, E. J., & Phillips, M. L. (2013). Dissociable patterns of abnormal frontal cortical activation during anticipation of an uncertain reward or loss in bipolar versus major depression., Bipolar Disorders (pp. 839-854). Chen, C.-H., Lennox, B., Jacob, R., Calder, A., Lupson, V., Bisbrown-Chippendale, R., Suckling, J., & Bullmore, E. (2006). Explicit and implicit facial affect recognition in manic and depressed states of bipolar disorder: a functional magnetic resonance imaging study., Biol. Psychiatry (pp. 31-39). Chen, C.-H., Suckling, J., Lennox, B. R., Ooi, C., & Bullmore, E. T. (2011). A quantitative meta-analysis of fMRI studies in bipolar disorder., Bipolar Disord (pp. 1-15). Costafreda, S. G., Fu, C. H. Y., Picchioni, M., Toulopoulou, T., McDonald, C., Kravariti, E., Walshe, M., Prata, D., Murray, R. M., & McGuire, P. K. (2011). Pattern of neural responses to verbal fluency shows diagnostic specificity for schizophrenia and bipolar disorder., BMC Psychiatry (pp. 18). Costafreda, S. G., Khanna, A., Mourão-Miranda, J., & Fu, C. H. Y. (2009). Neural correlates of sad faces predict clinical remission to cognitive behavioural therapy in depression., Neuroreport (pp. 637-641). Cuthbert, B. N., & Insel, T. R. (2013). Toward the future of psychiatric diagnosis: the seven pillars of RDoC. BMC Med 11, 126. Davidson, R. J., Irwin, W., Anderle, M. J., & Kalin, N. H. (2003). The neural substrates of affective processing in depressed patients treated with venlafaxine., Am J Psychiatry (pp. 64-75). Delvecchio, G., Fossati, P., Boyer, P., Brambilla, P., Falkai, P., Gruber, O., Hietala, J., Lawrie, S. M., Martinot, J. L., McIntosh, A. M., Meisenzahl, E., & Frangou, S. (2012). Common and distinct neural correlates of emotional processing in Bipolar Disorder and Major Depressive Disorder: a voxel-based meta-analysis of functional magnetic resonance imaging studies. Eur Neuropsychopharmacol 22, 100-13. Demant, K. M., Almer, G. M., Vinberg, M., Kessing, L. V., & Miskowiak, K. W. (2013). Effects of cognitive remediation on cognitive dysfunction in partially or fully remitted patients with bipolar disorder: study protocol for a randomized controlled trial. Trials 14, 378. Diler, R. S., de Almeida, J. R. C., Ladouceur, C., Birmaher, B., Axelson, D., & Phillips, M. (2013a). Neural activity to intense positive versus negative stimuli can help differentiate bipolar disorder from unipolar major depressive disorder in depressed adolescents: a pilot fMRI study., Psychiatry Res (pp. 277-284). Diler, R. S., Ladouceur, C. D., Segreti, A., Almeida, J. R. C., Birmaher, B., Axelson, D. A., Phillips, M. L., & Pan, L. A. (2013b). Neural correlates of treatment response


AC C

EP

TE D

M AN U

SC

RI PT

in depressed bipolar adolescents during emotion processing., Brain Imaging Behav (pp. 227-235). Diler, R. S., Pan, L. A., Segreti, A., Ladouceur, C. D., Forbes, E., Cela, S. R., Almeida, J. R. C., Birmaher, B., Axelson, D. A., & Phillips, M. L. (2014). Differential Anterior Cingulate Activity during Response Inhibition in Depressed Adolescents with Bipolar and Unipolar Major Depressive Disorder., J Can Acad Child Adolesc Psychiatry (pp. 10-19). Diler, R. S., Segreti, A. M., Ladouceur, C. D., Almeida, J. R. C., Birmaher, B., Axelson, D. A., Phillips, M. L., & Pan, L. (2013c). Neural correlates of treatment in adolescents with bipolar depression during response inhibition., Journal of Child and Adolescent Psychopharmacology (pp. 214-221). Drevets, W. C., Bogers, W., & Raichle, M. E. (2002). Functional anatomical correlates of antidepressant drug treatment assessed using PET measures of regional glucose metabolism., Eur Neuropsychopharmacol (pp. 527-544). Drevets, W. C., Price, J. L., Simpson, J. R., Todd, R. D., Reich, T., Vannier, M., & Raichle, M. E. (1997). Subgenual prefrontal cortex abnormalities in mood disorders., Nature (pp. 824-827). Elliott, R., Ogilvie, A., Rubinsztein, J. S., Calderon, G., Dolan, R. J., & Sahakian, B. J. (2004). Abnormal ventral frontal response during performance of an affective go/no go task in patients with mania., Biol. Psychiatry (pp. 1163-1170). Fales, C. L., Barch, D. M., Rundle, M. M., Mintun, M. A., Mathews, J., Snyder, A. Z., & Sheline, Y. I. (2009). Antidepressant treatment normalizes hypoactivity in dorsolateral prefrontal cortex during emotional interference processing in major depression. J Affect Disord 112, 206-11. Favre, P., Baciu, M., Pichat, C., De Pourtales, M. A., Fredembach, B., Garcon, S., Bougerol, T., & Polosan, M. (2013). Modulation of fronto-limbic activity by the psychoeducation in euthymic bipolar patients. A functional MRI study. Psychiatry Res 214, 285-95. Fleck, D. E., Eliassen, J. C., Durling, M., Lamy, M., Adler, C. M., DelBello, M. P., Shear, P. K., Cerullo, M. A., Lee, J.-H., & Strakowski, S. M. (2012). Functional MRI of sustained attention in bipolar mania., Mol Psychiatry (pp. 325-336). Foland, L. C., Altshuler, L. L., Bookheimer, S. Y., Eisenberger, N., Townsend, J., & Thompson, P. M. (2008). Evidence for deficient modulation of amygdala response by prefrontal cortex in bipolar mania., Psychiatry Res (pp. 27-37). Fournier, J. C., Keener, M. T., Almeida, J., Kronhaus, D. M., & Phillips, M. L. (2013). Amygdala and whole-brain activity to emotional faces distinguishes major depressive disorder and bipolar disorder. (pp. 741-752). Friston, K. J., Harrison, L., & Penny, W. (2003). Dynamic causal modelling., Neuroimage (pp. 1273-1302). Frodl, T., Scheuerecker, J., Schoepf, V., Linn, J., Koutsouleris, N., Bokde, A. L., Hampel, H., Moller, H. J., Bruckmann, H., Wiesmann, M., & Meisenzahl, E. (2011). Different effects of mirtazapine and venlafaxine on brain activation: an open randomized controlled fMRI study. J Clin Psychiatry 72, 448-57. Fu, C. H., & Costafreda, S. G. (2013). Neuroimaging-based biomarkers in psychiatry: clinical opportunities of a paradigm shift. Can J Psychiatry 58, 499-508.


AC C

EP

TE D

M AN U

SC

RI PT

Fu, C. H., Williams, S. C., Cleare, A. J., Brammer, M. J., Walsh, N. D., Kim, J., Andrew, C. M., Pich, E. M., Williams, P. M., Reed, L. J., Mitterschiffthaler, M. T., Suckling, J., & Bullmore, E. T. (2004). Attenuation of the neural response to sad faces in major depression by antidepressant treatment: a prospective, event-related functional magnetic resonance imaging study. Arch Gen Psychiatry 61, 877-89. Fu, C. H. Y., Williams, S. C. R., Cleare, A. J., Scott, J., Mitterschiffthaler, M. T., Walsh, N. D., Donaldson, C., Suckling, J., Andrew, C., Steiner, H., & Murray, R. M. (2008). Neural responses to sad facial expressions in major depression following cognitive behavioral therapy., Biol. Psychiatry (pp. 505-512). Fusar-Poli, P., Placentino, A., Carletti, F., Landi, P., Allen, P., Surguladze, S., Benedetti, F., Abbamonte, M., Gasparotti, R., Barale, F., Perez, J., McGuire, P., & Politi, P. (2009). Functional atlas of emotional faces processing: a voxel-based metaanalysis of 105 functional magnetic resonance imaging studies., J Psychiatry Neurosci (pp. 418-432). Ghaemi, S. N., Sachs, G. S., Chiou, A. M., Pandurangi, A. K., & Goodwin, K. (1999). Is bipolar disorder still underdiagnosed? Are antidepressants overutilized?, J Affect Disord (pp. 135-144). Goldberg, J. F., & Ernst, C. L. (2002). Features associated with the delayed initiation of mood stabilizers at illness onset in bipolar disorder., J Clin Psychiatry (pp. 985991). Grotegerd, D., Stuhrmann, A., Kugel, H., Schmidt, S., Redlich, R., Zwanzger, P., Rauch, A. V., Heindel, W., Zwitserlood, P., Arolt, V., Suslow, T., & Dannlowski, U. (2013a). Amygdala excitability to subliminally presented emotional faces distinguishes unipolar and bipolar depression: An fMRI and pattern classification study., Hum Brain Mapp. Grotegerd, D., Suslow, T., Bauer, J., Ohrmann, P., Arolt, V., Stuhrmann, A., Heindel, W., Kugel, H., & Dannlowski, U. (2013b). Discriminating unipolar and bipolar depression by means of fMRI and pattern classification: a pilot study., Eur Arch Psychiatry Clin Neurosci (pp. 119-131). Hafeman, D. M., Chang, K. D., Garrett, A. S., Sanders, E. M., & Phillips, M. L. (2012). Effects of medication on neuroimaging findings in bipolar disorder: an updated review., Bipolar Disord (pp. 375-410). Hariri, A. R. (2012). The highs and lows of amygdala reactivity in bipolar disorders. Am J Psychiatry 169, 780-3. Hariri, A. R., Bookheimer, S. Y., & Mazziotta, J. C. (2000). Modulating emotional responses: effects of a neocortical network on the limbic system., Neuroreport (pp. 43-48). Hariri, A. R., Mattay, V. S., Tessitore, A., Fera, F., & Weinberger, D. R. (2003). Neocortical modulation of the amygdala response to fearful stimuli., Biol. Psychiatry (pp. 494-501). Heller, A. S., Johnstone, T., Light, S. N., Peterson, M. J., Kolden, G. G., Kalin, N. H., & Davidson, R. J. (2013a). Relationships between changes in sustained frontostriatal connectivity and positive affect in major depression resulting from antidepressant treatment., Am J Psychiatry (pp. 197-206). Heller, A. S., Johnstone, T., Peterson, M. J., Kolden, G. G., Kalin, N. H., & Davidson, R. J. (2013b). Increased prefrontal cortex activity during negative emotion regulation


AC C

EP

TE D

M AN U

SC

RI PT

as a predictor of depression symptom severity trajectory over 6 months. JAMA Psychiatry 70, 1181-9. Hirschfeld, R. M. A., Lewis, L., & Vornik, L. A. (2003). Perceptions and impact of bipolar disorder: how far have we really come? Results of the national depressive and manic-depressive association 2000 survey of individuals with bipolar disorder., J Clin Psychiatry (pp. 161-174). Hosokawa, T., Momose, T., & Kasai, K. (2009). Brain glucose metabolism difference between bipolar and unipolar mood disorders in depressed and euthymic states., Prog Neuropsychopharmacol Biol Psychiatry (pp. 243-250). Hulvershorn, L. A., Karne, H., Gunn, A. D., Hartwick, S. L., Wang, Y., Hummer, T. A., & Anand, A. (2011). Neural Activation During Facial Emotion Processing in Unmedicated Bipolar Depression, Euthymia, and Mania., Biol. Psychiatry (pp. 603-610). Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K., Sanislow, C., & Wang, P. (2010). Research domain criteria (RDoC): toward a new classification framework for research on mental disorders. Am J Psychiatry 167, 748-51. Ives-Deliperi, V. L., Howells, F., Stein, D. J., Meintjes, E. M., & Horn, N. (2013). The effects of mindfulness-based cognitive therapy in patients with bipolar disorder: A controlled functional MRI investigation., J Affect Disord. Judd, L. L., Akiskal, H. S., Schettler, P. J., Coryell, W., Endicott, J., Maser, J. D., Solomon, D. A., Leon, A. C., & Keller, M. B. (2003). A prospective investigation of the natural history of the long-term weekly symptomatic status of bipolar II disorder., Arch. Gen. Psychiatry (pp. 261-269). Keener, M. T., Fournier, J. C., Mullin, B. C., Kronhaus, D., Perlman, S. B., LaBarbara, E., Almeida, J. C., & Phillips, M. L. (2012). Dissociable patterns of medial prefrontal and amygdala activity to face identity versus emotion in bipolar disorder. Psychol Med 42, 1913-24. Kennedy, S. H., Konarski, J. Z., Segal, Z. V., Lau, M. A., Bieling, P. J., McIntyre, R. S., & Mayberg, H. S. (2007). Differences in brain glucose metabolism between responders to CBT and venlafaxine in a 16-week randomized controlled trial., Am J Psychiatry (pp. 778-788). Kessler, R. C., Chiu, W. T., Demler, O., & Walters, E. E. (2005). Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the National Comorbidity Survey Replication., Archives of General Psychiatry (pp. 617-627). Ketter, T. A., Kimbrell, T. A., George, M. S., Dunn, R. T., Speer, A. M., Benson, B. E., Willis, M. W., Danielson, A., Frye, M. A., Herscovitch, P., & Post, R. M. (2001). Effects of mood and subtype on cerebral glucose metabolism in treatmentresistant bipolar disorder., Biol. Psychiatry (pp. 97-109). Kim, M. J., Loucks, R. A., Palmer, A. L., Brown, A. C., Solomon, K. M., Marchante, A. N., & Whalen, P. J. (2011). The structural and functional connectivity of the amygdala: from normal emotion to pathological anxiety., Behav. Brain Res. (pp. 403-410). Konarski, J. Z., Kennedy, S. H., Segal, Z. V., Lau, M. A., Bieling, P. J., McIntyre, R. S., & Mayberg, H. S. (2009). Predictors of nonresponse to cognitive behavioural therapy or venlafaxine using glucose metabolism in major depressive disorder., J Psychiatry Neurosci (pp. 175-180).


AC C

EP

TE D

M AN U

SC

RI PT

Lawrence, N. S., Williams, A. M., Surguladze, S., Giampietro, V., Brammer, M. J., Andrew, C., Frangou, S., Ecker, C., & Phillips, M. L. (2004). Subcortical and ventral prefrontal cortical neural responses to facial expressions distinguish patients with bipolar disorder and major depression., Biol. Psychiatry (pp. 578587). Light, S. N., Heller, A. S., Johnstone, T., Kolden, G. G., Peterson, M. J., Kalin, N. H., & Davidson, R. J. (2011). Reduced right ventrolateral prefrontal cortex activity while inhibiting positive affect is associated with improvement in hedonic capacity after 8 weeks of antidepressant treatment in major depressive disorder. Biol Psychiatry 70, 962-8. Lipp, I., Murphy, K., Wise, R. G., & Caseras, X. (2014). Understanding the contribution of neural and physiological signal variation to the low repeatability of emotioninduced BOLD responses. Neuroimage 86, 335-42. Little, J. T., Ketter, T. A., Kimbrell, T. A., Danielson, A., Benson, B., Willis, M. W., & Post, R. M. (1996). Venlafaxine or bupropion responders but not nonresponders show baseline prefrontal and paralimbic hypometabolism compared with controls., Psychopharmacol Bull (pp. 629-635). Macdonald, A. W., Cohen, J. D., Stenger, V. A., & Carter, C. S. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control., Science (pp. 1835-1838). Malhi, G. S., Lagopoulos, J., Ward, P. B., Kumari, V., Mitchell, P. B., Parker, G. B., Ivanovski, B., & Sachdev, P. (2004). Cognitive generation of affect in bipolar depression: an fMRI study., Eur J Neurosci (pp. 741-754). Marchand, W. R., Lee, J. N., Garn, C., Thatcher, J., Gale, P., Kreitschitz, S., Johnson, S., & Wood, N. (2011). Aberrant emotional processing in posterior cortical midline structures in bipolar II depression., Prog Neuropsychopharmacol Biol Psychiatry (pp. 1729-1737). Marchand, W. R., Lee, J. N., Johnson, S., Gale, P., & Thatcher, J. (2013). Differences in functional connectivity in major depression versus bipolar II depression., J Affect Disord. Mazzola-Pomietto, P., Kaladjian, A., Azorin, J.-M., Anton, J.-L., & Jeanningros, R. (2009). Bilateral decrease in ventrolateral prefrontal cortex activation during motor response inhibition in mania., J Psychiatr Res (pp. 432-441). McGrath, C. L., Kelley, M. E., Holtzheimer, P. E., Dunlop, B. W., Craighead, W. E., Franco, A. R., Craddock, R. C., & Mayberg, H. S. (2013). Toward a neuroimaging treatment selection biomarker for major depressive disorder. JAMA Psychiatry 70, 821-9. Mega, M. S., & Cummings, J. L. (1994). Frontal-subcortical circuits and neuropsychiatric disorders., J Neuropsychiatry Clin Neurosci (pp. 358-370). Mourao-Miranda, J., Reinders, A. A. T. S., Rocha-Rego, V., Lappin, J., Rondina, J., Morgan, C., Morgan, K. D., Fearon, P., Jones, P. B., Doody, G. A., Murray, R. M., Kapur, S., & Dazzan, P. (2012). Individualized prediction of illness course at the first psychotic episode: a support vector machine MRI study., Psychol. Med. (pp. 1037-1047).


AC C

EP

TE D

M AN U

SC

RI PT

Neta, M., Kelley, W. M., & Whalen, P. J. (2013). Neural responses to ambiguity involve domain-general and domain-specific emotion processing systems. J Cogn Neurosci 25, 547-57. Ochsner, K. N., Bunge, S. A., Gross, J. J., & Gabrieli, J. D. E. (2002). Rethinking feelings: an FMRI study of the cognitive regulation of emotion., J Cogn Neurosci (pp. 1215-1229). Orrù, G., Pettersson-Yeo, W., Marquand, A. F., Sartori, G., & Mechelli, A. (2012). Using Support Vector Machine to identify imaging biomarkers of neurological and psychiatric disease: a critical review., Neurosci Biobehav Rev (pp. 1140-1152). Pavuluri, M. N., Passarotti, A. M., Fitzgerald, J. M., Wegbreit, E., & Sweeney, J. A. (2012). Risperidone and divalproex differentially engage the fronto-striatotemporal circuitry in pediatric mania: a pharmacological functional magnetic resonance imaging study., J Am Acad Child Adolesc Psychiatry (pp. 157170.e5). Perlis, R. H., Miyahara, S., Marangell, L. B., Wisniewski, S. R., Ostacher, M., Delbello, M. P., Bowden, C. L., Sachs, G. S., Nierenberg, A. A., & Investigators, S.-B. (2004). Long-Term implications of early onset in bipolar disorder: data from the first 1000 participants in the systematic treatment enhancement program for bipolar disorder (STEP-BD). Biol. Psychiatry (pp. 875-881). Perlman, S. B., Almeida, J. R. C., Kronhaus, D. M., Versace, A., Labarbara, E. J., Klein, C. R., & Phillips, M. L. (2012). Amygdala activity and prefrontal cortex-amygdala effective connectivity to emerging emotional faces distinguish remitted and depressed mood states in bipolar disorder., Bipolar Disord (pp. 162-174). Phillips, M. L., Ladouceur, C. D., & Drevets, W. C. (2008). A neural model of voluntary and automatic emotion regulation: implications for understanding the pathophysiology and neurodevelopment of bipolar disorder., Mol Psychiatry (pp. 829-833-57). Phillips, M. L., & Swartz, H. A. (2014). A Critical Appraisal of Neuroimaging Studies of Bipolar Disorder: Toward a New Conceptualization of Underlying Neural Circuitry and a Road Map for Future Research., Am J Psychiatry. Phillips, M. L., Travis, M. J., Fagiolini, A., & Kupfer, D. J. (2008). Medication effects in neuroimaging studies of bipolar disorder., Am J Psychiatry (pp. 313-320). Price, R. B., Paul, B., Schneider, W., & Siegle, G. J. (2013). Neural Correlates of Three Neurocognitive Intervention Strategies: A Preliminary Step Towards Personalized Treatment for Psychological Disorders., Cognit Ther Res (pp. 657-672). Radaelli, D., Dallaspezia, S., Poletti, S., Smeraldi, E., Falini, A., Colombo, C., & Benedetti, F. (2013). Different neural responses to a moral valence decision task in unipolar and bipolar depression., ISRN Psychiatry (pp. 568617-10): Hindawi Publishing Corporation. Rastelli, C. P., Cheng, Y., Weingarden, J., Frank, E., & Swartz, H. A. (2013). Differences between unipolar depression and bipolar II depression in women. J Affect Disord 150, 1120-4. Roebroeck, A., Formisano, E., & Goebel, R. (2011). The identification of interacting networks in the brain using fMRI: Model selection, causality and deconvolution., Neuroimage (pp. 296-302).


AC C

EP

TE D

M AN U

SC

RI PT

Rubinsztein, J. S., Fletcher, P. C., Rogers, R. D., Ho, L. W., Aigbirhio, F. I., Paykel, E. S., Robbins, T. W., & Sahakian, B. J. (2001). Decision-making in mania: a PET study., Brain (pp. 2550-2563). Salvadore, G., Cornwell, B. R., Colon-Rosario, V., Coppola, R., Grillon, C., Zarate, C. A., Jr., & Manji, H. K. (2009). Increased anterior cingulate cortical activity in response to fearful faces: a neurophysiological biomarker that predicts rapid antidepressant response to ketamine. Biol Psychiatry 65, 289-95. Sauder, C. L., Hajcak, G., Angstadt, M., & Phan, K. L. (2013). Test-retest reliability of amygdala response to emotional faces. Psychophysiology 50, 1147-56. Savitz, J. B., Rauch, S. L., & Drevets, W. C. (2013). Clinical application of brain imaging for the diagnosis of mood disorders: the current state of play. Mol Psychiatry 18, 528-39. Saxena, S., Brody, A. L., Ho, M. L., Zohrabi, N., Maidment, K. M., & Baxter, L. R. (2003). Differential brain metabolic predictors of response to paroxetine in obsessive-compulsive disorder versus major depression., Am J Psychiatry (pp. 522-532). Schneider, M. R., Adler, C. M., Whitsel, R., Weber, W., Mills, N. P., Bitter, S. M., Eliassen, J., Strakowski, S. M., & Delbello, M. P. (2012). The effects of ziprasidone on prefrontal and amygdalar activation in manic youth with bipolar disorder., Isr J Psychiatry Relat Sci (pp. 112-120). Schwartz, C. E., Holt, D. J., & Rosenbaum, J. F. (2013). Methodological challenges of moving beyond DSM categories: transdiagnostic studies of frontolimbic connectivity changes in vulnerable individuals. Biol Psychiatry 74, 240-1. Sheline, Y. I., Barch, D. M., Donnelly, J. M., Ollinger, J. M., Snyder, A. Z., & Mintun, M. A. (2001). Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study. Biol Psychiatry 50, 651-8. Shin, L. M., & Liberzon, I. (2010). The neurocircuitry of fear, stress, and anxiety disorders., Neuropsychopharmacology (pp. 169-191). Siegle, G. J., Carter, C. S., & Thase, M. E. (2006). Use of FMRI to predict recovery from unipolar depression with cognitive behavior therapy. Am J Psychiatry 163, 735-8. Siegle, G. J., Thompson, W. K., Collier, A., Berman, S. R., Feldmiller, J., Thase, M. E., & Friedman, E. S. (2012). Toward clinically useful neuroimaging in depression treatment: Prognostic utility of subgenual cingulate activity for determining depression outcome in cognitive therapy across studies, scanners, and patient characteristics, Archives of General Psychiatry (pp. 913-924). Strakowski, S. M., Adler, C. M., Almeida, J., Altshuler, L. L., Blumberg, H. P., Chang, K. D., Delbello, M. P., Frangou, S., McIntosh, A., Phillips, M. L., Sussman, J. E., & Townsend, J. D. (2012). The functional neuroanatomy of bipolar disorder: a consensus model., Bipolar Disord (pp. 313-325). Strakowski, S. M., DelBello, M. P., & Adler, C. M. (2005). The functional neuroanatomy of bipolar disorder: a review of neuroimaging findings., Mol Psychiatry (pp. 105116). Strakowski, S. M., Eliassen, J. C., Lamy, M., Cerullo, M. A., Allendorfer, J. B., Madore, M., Lee, J.-H., Welge, J. A., Delbello, M. P., Fleck, D. E., & Adler, C. M. (2011). Functional magnetic resonance imaging brain activation in bipolar mania:


AC C

EP

TE D

M AN U

SC

RI PT

evidence for disruption of the ventrolateral prefrontal-amygdala emotional pathway., Biol. Psychiatry (pp. 381-388). Surguladze, S. A., Marshall, N., Schulze, K., Hall, M. H., Walshe, M., Bramon, E., Phillips, M. L., Murray, R. M., & McDonald, C. (2010). Exaggerated neural response to emotional faces in patients with bipolar disorder and their firstdegree relatives. Neuroimage 53, 58-64. Taylor Tavares, J. V., Clark, L., Furey, M. L., Williams, G. B., Sahakian, B. J., & Drevets, W. C. (2008). Neural basis of abnormal response to negative feedback in unmedicated mood disorders. Neuroimage 42, 1118-26. Torrisi, S., Moody, T. D., Vizueta, N., Thomason, M. E., Monti, M. M., Townsend, J. D., Bookheimer, S. Y., & Altshuler, L. L. (2013). Differences in resting corticolimbic functional connectivity in bipolar I euthymia., Bipolar Disorders (pp. 156-166). Townsend, J., & Altshuler, L. L. (2012). Emotion processing and regulation in bipolar disorder: a review. Bipolar Disord 14, 326-39. Van der Schot, A., Kahn, R., Ramsey, N., Nolen, W., & Vink, M. (2010). Trait and state dependent functional impairments in bipolar disorder., Psychiatry Res (pp. 135142). Versace, A., Thompson, W. K., Zhou, D., Almeida, J. R. C., Hassel, S., Klein, C. R., Kupfer, D. J., & Phillips, M. L. (2010). Abnormal left and right amygdalaorbitofrontal cortical functional connectivity to emotional faces: state versus trait vulnerability markers of depression in bipolar disorder., Biol. Psychiatry (pp. 422431). Victor, T. A., Furey, M. L., Fromm, S. J., Ohman, A., & Drevets, W. C. (2010). Relationship between amygdala responses to masked faces and mood state and treatment in major depressive disorder. Arch Gen Psychiatry 67, 1128-38. Vizueta, N., Rudie, J. D., Townsend, J. D., Torrisi, S., Moody, T. D., Bookheimer, S. Y., & Altshuler, L. L. (2012). Regional fMRI Hypoactivation and Altered Functional Connectivity During Emotion Processing in Nonmedicated Depressed Patients With Bipolar II Disorder., Am J Psychiatry (pp. 831-840). Volk, S. A., Kaendler, S. H., Hertel, A., Maul, F. D., Manoocheri, R., Weber, R., Georgi, K., Pflug, B., & Hor, G. (1997). Can response to partial sleep deprivation in depressed patients be predicted by regional changes of cerebral blood flow? Psychiatry Res 75, 67-74. Yang, H., Lu, L. H., Wu, M., Stevens, M., Wegbreit, E., Fitzgerald, J., Levitan, B., Shankman, S., & Pavuluri, M. N. (2013). Time course of recovery showing initial prefrontal cortex changes at 16 weeks, extending to subcortical changes by 3 years in pediatric bipolar disorder. J Affect Disord 150, 571-7.


Figure Captions Figure 1. Key regions of the corticolimbic network and their major network connections

RI PT

relevant to bipolar disorder (Brooks et al. 2009a) ACC = Anterior cingulate, AMG = Amygdala, ATC = Anterior Temporal Cortex, CV = Cerebellar vermis, DLPFC = Dorsolateral prefrontal cortex, HYPTH = Hypothalamus, MOFC = Medio-orbital

SC

prefrontal cortex, PHG = Parahippocampal gyrus, SGPFC = subgenual prefrontal cortex, THAL = Thalamus, VLPFC = Ventrolateral prefrontal cortex.

M AN U

Figure 2. Regions involved in the ventrolateral prefrontal circuit.

Figure 3. Regions involved in the ventromedial prefrontal circuit. Figure 4. Regions involved in the dorsolateral prefrontal circuit.

AC C

EP

TE D

Figure 5. Regions involved in the anterior cingulate circuit.

ACCEPTED MANUSCRIPT

Page 1 of 6 Table 1. Task-based functional neuroimaging studies directly comparing individuals with bipolar or unipolar depression Depression severity (mean, SD)

Paradigm

Method

Emotion processing and emotion regulation paradigms CDRS-R: BP = 74.5 (12.8), MDD = 65.8 (13.3), HC = 19.1 (1.8)

Adolescents: (age range: 12-17) BP age: 15.6, MDD age: 15.9, HC age: 15.6 Unmedicated: 2 BP and 4 MDD

BP age: 42, MDD age: 41.2, HC age: 41.1 Unmedicated: 2 22 BP (86% F), 30 MDD (73% F), 29 HC (59% F) BP age: 34.0, MDD age: 30.3, HC age: 32.5 Medication status unspecified

HRSD: BP = 23.3 (4.6), MDD = 25.1 (7.1), HC = 1.2 (1.8) All patients were acutely depressed inpatients.

Sad, happy, and neutral face primes masked by neutral faces

SPM8: whole-brain

EP

Fournier et al. (2013) (Fournier, Keener, Almeida, Kronhaus, & Phillips, 2013)

22 BPId (11F), 22 MDDd (11F), 22 HC (11F)

SPM5: whole-brain

17HRSD: BP = 19.7 (6.1), MDD = 21.1 (3.7), HC = 1.4 (2.2)

Dynamically changing faces (morph from 0% neutral to 100% emotional display) and shape control task

AC C

Grotegerd et al. (2013)(Gro tegerd et al., 2013)

Gender labeling task with intense happy or fearful (H100%, F100%), mild happy or fearful (H50%, F50%), and neutral (Hn or Fn) faces

L/R amygdala ROIs created using WFU PickAtlas SPM8: whole-brain L/R amygdala ROIs created using WFU PickAtlas

Results

Happy: experiment: each condition (neutral, mild happy, intense happy) > fixation;

No significant group differences in amygdala for either happy or fear experiment

Fear: experiment: each condition (neutral, mild fear, intense fear) > fixation Masked sad > Neutral,

MDD > HC: L inferior parietal (Hn), L occipital (Hn), L insula (F100), L insula (F50), R superior temporal (F50), L postcentral (F50), R occipital (F50), L VLPFC BA 45 (Fn).

M AN U

10 BP (5 BPId, 5 BPIId) (8F), 10 MDD (8F), 10 HC (8F)

TE D

Diler et al. (2013)(Dil er et al., 2013)

Key condition or contrasts

RI PT

Group Characteristics

SC

Study

Masked happy > Neutral

Anger > shape, Fear > shape, Sad > shape, Happy > shape

MDD < HC: L superior prefrontal (H100), L superior temporal (H100), R parahippocampus (H50), R precuneus (Hn), R insula (Fn), R frontal precentral (F50), L middle frontal (BA 6 and 10) (F50)

MDD > BP: bilateral amygdala (sad faces) MDD > HC: bilateral amygdala (sad faces) HC > BP: L amygdala (sad faces) BP > MDD: R amygdala (happy faces) HC > MDD: R amygdala (happy faces) BP > HC: n.s. amygdala difference (happy)

No significant group differences in amygdala to fear, sad or happy faces. MDD > BP: L amygdala (angry faces); L fusiform, R occipital, parietal, ACC (BA32), bilateral insula, and bilateral temporal cortical regions (angry and fear). BP > MDD: R ACC (BA32), bilateral regions of the insula, middle and superior temporal, and parietal regions (sad faces). MDD > HC: L amygdala (angry faces); bilateral

ACCEPTED MANUSCRIPT

Page 2 of 6 temporal-parietal regions and the R ACC (angry faces), L middle and superior temporal gyri and R parietal regions (happy faces).

BPd age: 36.6, BPr age: 33.3, MDDd age: 32.7, HC age: 32.7 Unmedicated: 2 BPd, 2 MDDd, 1 BPr

BPd age: 36.6; MDDd age: 32.3; HC age: 28.3 Unmedicated: 2 BPId and 2 MDD

Lawrence

25HRSD score: BPd = 21.5 (6.4) MDD = 24.6 (6)

EP

15 BPId (14F); 16 MDDd (13F); 16 HC (12F)

12 BPI (5F);

AC C

Almeida et al. (2009) (Almeida et al., 2009b)

Affect labeling task with intense happy, sad or fearful (H100%, S100%, F100%), mild happy, sad or fearful (H50%, S50%, F50%), and neutral faces Affect labeling task with intense happy, sad or fearful (H100%, S100%, F100%), mild happy, sad or fearful (H50%, S50%, F50%), and neutral faces

BDI score:

SPM5: whole-brain L/R amygdala ROIs created using WFU PickAtlas

Gender

intense happy [or sad, or fear] > baseline;

SC

25HRSD: BPd = 21.5 (6.4), BPr = 1.5 (1.1), MDDd = 24.5 (6.1)

mild happy [or sad, or fear] > baseline;

M AN U

15 BPId (14F), 15 BPIr (10F), 15 MDDd (13F), 15 HC (12F)

TE D

Almeida et al. (2010) (Almeida, Versace, Hassel, Kupfer, & Phillips, 2010)

RI PT

HC > BP: n.s. amygdala difference (angry); R temporal-parietal regions (angry faces), L temporal and R occipital regions (fear faces), L fusiform (happy faces).

SPM5: whole-brain R/L amygdala and R/L OMPFC (BA 11) created using WFU PickAtlas

Not

BP > HC: L middle and superior temporal regions and R ACC (sad faces). No significant group differences in amygdala for intense sad or intense/mild happy or intense/mild fear faces. BPId > BPIr: L amygdala (mild sad, neutral) BPId > MDDd: L amygdala (mild sad, neutral)

neutral > baseline

intense happy (or sad) > baseline; mild happy (or sad) > baseline; neutral > baseline DCM: R/L bottom-up amygdala– OMPFC and top-down OMPFC– amygdala EC intense happy

Standard ROI analyses: Main effect of group for left amygdala in the sad experiment. Group x condition interaction in left amygdala in the happy experiment. DCM Analysis: Significant group difference in left-sided top-down OMPFC–amygdala EC during the happy experiment: HC > MDDd, HC > BPId, MDDd = BPId Significant group difference in right-sided bottom-up amygdala–OMPFC EC in the happy experiment: HC BPId, HC = MDDd, MDDd = BPId Trend between-group difference in left-sided topdown OMPFC–amygdala EC in the sad experiment: HC < MDDd, HC < BPId, MDDd = BPId Fear:

ACCEPTED MANUSCRIPT

Page 3 of 6

functional ROIs included clusters of significant activation in amygdala for each of the six emotional > neutral contrasts

mild happy [or sad, or fear] > neutral

BP > MDD: L amygdala, L VLPFC (BA 47), R globus pallidus/anterior thalamus (intense fear); R globus pallidus/anterior thalamus, L medial prefrontal cortex (mild fear);

RI PT

MDD: 2 mildly depressed, 2 moderately depressed, 5 severely depressed

[or sad, or fear] > neutral

BP > HC: L amygdala, L VLPFC (BA 47) (intense fear); R globus pallidus/anterior thalamus (mild fear); HC > BP: R amygdala, R DLPFC, R hippocampus (mild fear);

SC

BP: 3 euthymic, 7 mildly depressed, 2 moderately to severely depressed

Specified: whole-brain

M AN U

All BP and MDD were medicated.

labeling task with intense happy, sad or fearful (H100%, S100%, F100%), mild happy, sad or fearful (H50%,S50 %, F50%), and neutral faces

TE D

Age (all groups): 41

MDD = 31.8 (11.8), BP = 15.3 (9.2), HC = 2.27 (2.3)

EP

9 MDD (4F); 11 HC (4F)

AC C

et al. (2004)(La wrence et al., 2004)

HC > MDD: R globus pallidus/anterior thalamus (intense fear); R amygdala, R DLPFC, L medial prefrontal cortex, R hippocampus (mild fear) Happy: BP > MDD: R DLPFC (intense happy); L amygdala, caudate/putamen, R VLPFC (mild happy) BP > HC: L amygdala, caudate/putamen, R VLPFC (mild happy) HC > MDD: L amygdala, R parahippocampal gyrus, bilateral regions of thalamus, midbrain, caudate nucleus, and L amygdala (intense happy); R DLPFC (intense happy); L amygdala, caudate/putamen (mild happy) HC > BP: L amygdala, R parahippocampal gyrus, bilateral regions of thalamus, midbrain, caudate nucleus, and L amygdala (intense happy) Sad: No significant group differences in amygdala to sad faces. MDD > BP: R putamen (mild sad) BP > MDD: R ventral prefrontal cortex, ventral and dorsal R ACC (intense sad); L hippocampus, L VLPFC (mild sad) BP > HC: R ventral prefrontal cortex, ventral and dorsal R ACC (intense sad); L hippocampus, L VLPFC (mild sad)

ACCEPTED MANUSCRIPT

Page 4 of 6

HC > BP: bilateral DLPFC (intense sad); OFC, R putamen, R DLPFC (mild sad)

RI PT

HC > MDD: bilateral DLPFC (intense sad); OFC (mild sad)

25HRSD: BP = 33.9 (8.0), MDD = 26.6 (5.7), HC = 1.9 (2.3)

Cardguessing game 4 Trial types: expectation of a possible win, followed by a win outcome (win trials) or a no change outcome (disappointm ent trials); expectation of a possible loss, followed by a loss outcome (loss trials) or no change (relief trials)

Reward expectancy, anticipation per se (the anticipation regressor minus the baseline regressor), and prediction error

8-mm spheres centered at coordinates derived from other studies for 4 ROIs: dACC, L VLPFC, R/L VS

MDD > HC: L medial PFC

ACC reward expectancy: HC > BP and MDD together, HC > BP, HC = MDD, BP = MDD L VLPFC reward expectancy: No significant group difference. L VLPFC anticipation per se—baseline: BP > MDD, BP > HC, MDD=HC VS prediction error: No significant group difference.

Cognitive paradigms Diler et al. 12 BP (6 BPId, 6 (2014)(Dil BPIId) (10F), er et al., 10 MDDd (8F), 2014) 10 HC (8F)

AC C

EP

TE D

19 BP and 31 MDD were taking at least 1 medication.

SPM8: whole-brain

M AN U

Reward processing paradigms Chase et 24 BPId (19F), 42 MDDd (31F), al. (2013)(Ch 37 HC (25F) ase et al., 2013) BP age: 33.94 MDD age: 31.04 HC age: 33.09

SC

Neutral: No significant group differences in amygdala to neutral faces > fixation.

CDRS-R: BP = 73.8 (12.5), MDD = 65.8 (13.3), HC = 19.1 (1.8)

Letter Go/NoGo task

SPM5: whole-brain

NoGo versus Go blocks

No significant group differences between BP and MDD, in comparison to HC. BP (but not MDD) > HC: L ACC (BA 32) (NoGo)

ACCEPTED MANUSCRIPT

Page 5 of 6 Adolescents: (age range: 12-17) BP age: 15.5, MDD age: 15.9 HC age: 15.6

MDD > HC: L Caudate, L Superior Temporal (BA 41), L Occipital (NoGo)

Go/NoGo task with pos and neg adjective words

BPId age: 46.0 MDDd age: 46.1 HC age: 40.7

BPd age: 33.4 MDDd age: 38.3 HC age: 33.9

AC C

16 BPd (9F), 17 MDDd (10F), 18 HC (11F)

Emotional Face N-Back task, attend to letter and ignore neutral, fear, happy face flankers or no face distractor

EP

BPId age: 31.9, MDDd age: 29.7, HC age: 32.8 Psychotropic medication use: 87% of MDDd females and 77.8% of BPId females.

Taylor Tavares et al. (2008)(Tay lor Tavares et al., 2008)

25HRSD score: BPId = 24.2 (8.4), MDDd = 26.3 (6.3)

MADRS score: BPd = 26 (2.8), MDDd = 23 (1.7), HC = 0.3 (0.3)

SPM5: whole-brain

TE D

18 BPId (18F), 23 MDDd (23F), 16 HC (16F)

SPM2: whole-brain

Probabilistic Reversal Learning Task

(no-go > fixation) > (go > fixation with moral valence of stimuli (pos/neg) and diagnosis as factors

SC

Patients had HDRS score >18

no treatment with long acting neuroleptic drugs in the last three months Bertocci et al. (2011)(Ber tocci et al., 2012)

RI PT

BP > HC: L Caudate, L Superior Temporal (BA 41), L Occipital, L DLPFC, L Medial Frontal (BA 10) (NoGo)

M AN U

Radaelli et al. (2013)(Ra daelli et al., 2013)

Unmedicated: 3 BP, 4 MDD 14 BPId (9F), 11 MDDd (6F), 11 HC (6F)

MDD (but not BP) > HC: L Caudate (Go)

L/R amygdala ROIs created using WFU PickAtlas

Differences between BPId and MDDd in L VLPFC (BA 47): higher activity for neg in MDDd, higher activity for pos in BPId Significant interaction of diagnosis (MDDd versus HC) and moral valence of the stimuli (neg-pos) in L VLPFC (BA 47): HC (higher activity for pos > neg), MDDd (higher activity for neg > pos)

2-back: fear face-no face;

Differences between BPId and HC in bilateral DLPFC (BA 10), bilateral medial PFC (BA 9), R ACC (BA24), temporal cortex and insula, and parietal and occipital cortex: HC (higher activity for neg > pos), BPId (higher activity for pos > neg) No significant group differences in the amygdala, DLPFC or in the rACC/dACC.

2-back: happy face-no face;

MDDd > BPId: L dAMCC (BA32) (2-back:neutral-2back:no face condition)

2-back: neutral face-no face.

MDDd > HC: L putamen (2-back:happy-2-back:no face condition), bilateral dAMCC (BA32) and R putamen (2-back:neutral-2-back:no face condition) BPId > HC: R putamen (2-back:neutral-2-back:no face condition)

SPM5: whole-brain L/R amygdala ROIs from the AAL map; 10mm

1. Error switches (ES) 2: Error nonswitches (ENS) 3: Error switches-nonswitches (ESENW).

ROI Results: BPd > MDDd: DMPFC (Error switches, Error nonswitches, All neg feedback); R VLPFC (Error switches, Reversal); bilateral VLPFC (Error switchesnon-switches) BPd > HC: DMPFC (Error non-switches)

ACCEPTED MANUSCRIPT

Page 6 of 6

MADRS score: BPIId = 27.8 (2.7), MDDd = 26.6 (4.0)

Motor activation task

4: All neg feedback (ANF). 5: Reversal (Rev)

SPM5: whole-brain

Between-group differences in R PCC FC

BPIId > MDDd: stronger FC between R PCC and R parietal/insular region (portions of the R inferior parietal lobule, precentral gyrus and insula)

Correlations between PCC FC and depression severity (as measured by MADRS score)

MDDd > BPIId: no regions of stronger PCC FC

R PCC seed-region connectivity analysis using CONN software (www.nitrc. org/projects /conn)

HC > MDDd: DMPFC (Error switches); R VLPFC (Error switches, Reversal), bilateral VLPFC (Error switches-non-switches);

RI PT

MDDd > HC: less of a reduction in R amygdala in MDD compared to HC (All neg feedback) (p=0.05)

SC

Other paradigms Marchand 14 BPIId, et al. 26 MDDd (2013)(Ma rchand, M: 35 subjects Lee, (88%), F: 5 (12%) Johnson, were female. Gale, & Thatcher, BPIId age: 29.5 2013) MDDd age: 27.8

radius spheres centered at MNI xyz coordinates : DMPFC (8, 32,52), L VLPFC (−32, 24, −4), R VLPFC (38, 24, −2)

M AN U

No psychotropic medication treatment within 3 weeks (8 weeks for fluoxetine)

Correlational analyses: MDDd (but not in BPIId): depression severity (MADRS score) positively correlated with FC between R PCC and several regions (bilateral inferior frontal gyri and left regions of insula, anterior cingulate, medial frontal, middle frontal and superior frontal gyrus)

AC C

EP

TE D

No psychotropic medication treatment within 3 months Abbreviations: AAL, Automated Anatomical Labeling; ACC, anterior cingulate cortex; ANF, All negative feedback (all negative feedback events > correct response baseline); BA, Broadmann area; BDI, Beck Depression Inventory; BP, bipolar disorder; BPd, bipolar disorder in depressed state; BPId, bipolar type 1 disorder in depressed state; BPIr, bipolar type 1 disorder in euthymic (remitted); BPIId, bipolar type II disorder in depressed state; CDRS-R, Children’s Depression Rating Scale-Revised; CRB, correct response baseline; dACC, dorsal ACC; dAMCC, Dorsal anterior midcingulate cortex; DCM, Dynamic Causal Modeling; DLPFC, Dorsolateral Prefrontal Cortex; DMPFC, Dorsomedial Prefrontal Cortex; EC, effective connectivity; ES, Error switches (probabilistic switch errors > correct response baseline); ENW, Error non-switches (probabilistic non-switch errors > correct response baseline); ES-ENW, Error switches-non-switches (probabilistic switch errors > probabilistic non-switch errors); F, female; FC, functional connectivity; F100, 100% intense (prototypical) fear facial expressions; F50, mild fear (50% intensity) facial expressions; Fn, neutral facial expressions presented in fear experiment; HC, healthy control; H100, 100% intense (prototypical) happy facial expressions; H50, mild happy (50% intensity) facial expressions; Hn, neutral facial expressions presented in happy experiment; L, left; M, male; MADRS, Montgomery-Asberg Depression Rating Scale; MDD, unipolar major depressive disorder; MDDd, unipolar major depressive disorder in depressed state; rMDD, unipolar major depressive disorder in remitted (euthymic) state; MNI, Montreal Neurological Institute; neg, negative; n.s., non significant; OFC, orbitofrontal cortex; OMPFC, orbitomedial prefrontal cortex; PCC, posterior cingulate cortex; pos, positive; PFC, prefrontal cortex; R, right; rACC, rostral ACC; Rev, Reversal (errors where subject subsequently reversed responding > errors where subject did not subsequently reverse responding). ROI, region of interest; VLPFC, ventrolateral PFC; VS, ventral striatum; WFU, Wake Forest University; 17HDRS, 17-item Hamilton Depression Rating Scale; 21HDRS, 21-item Hamilton Depression Rating Scale; 25HDRS, 25-item Hamilton Depression Rating Scale.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Acknowledgments

We thank Drs. Christopher J. Patrick and Pavel Blagov for scholarly comments on an

AC C

EP

TE D

M AN U

SC

RI PT

earlier version of this manuscript.

ACCEPTED MANUSCRIPT

Conflict of Interest Dr. Brooks is a member of the Speaker’s Bureau for Sunovion and has received

AC C

EP

TE D

M AN U

SC

RI PT

research support from Pfizer, Inc. Dr. Vizueta has no disclosures.

ACCEPTED MANUSCRIPT

Contributors

AC C

EP

TE D

M AN U

SC

RI PT

Both Dr. Brooks and Dr. Vizueta participated in the conceptualization, literature review, and writing of the manuscript.

ACCEPTED MANUSCRIPT

Role of Funding Source Dr. Vizueta was supported in part by a fellowship from the UCLA Integrative Study

AC C

EP

TE D

M AN U

SC

RI PT

Center in Mood Disorders.

Bipolar disorder: Functional neuroimaging markers in relatives.

Functional neuroimaging research in bipolar disorder.

Neuroimaging and neuropathological findings in bipolar disorder.

Can neuroimaging disentangle bipolar disorder?

Neuroimaging findings in late-onset schizophrenia and bipolar disorder.

Staging bipolar disorder: clinical, biochemical, and functional correlates.

Historical Underpinnings of Bipolar Disorder Diagnostic Criteria.

Diagnostic potential of multimodal neuroimaging in posttraumatic stress disorder.

Diagnostic Precursors to Bipolar Disorder in Offspring of Parents With Bipolar Disorder: A Longitudinal Study.

Wnt and GSK3 Signaling Pathways in Bipolar Disorder: Clinical and Therapeutic Implications.

[Differential diagnosis of bipolar disorder: historical and clinical implications and perspectives].

Migraine headache and bipolar disorder comorbidity: A systematic review of the literature and clinical implications.

Clinical endophenotypes for bipolar disorder.

Prevalence rates and clinical implications of bipolar disorder "with mixed features" as defined by DSM-5.

Clinical correlates of age at onset distribution in bipolar disorder: a comparison between diagnostic subgroups.

Functional impairment and previous suicide attempts in bipolar disorder.

Bipolar disorder in primary care: clinical characteristics of 740 primary care patients with bipolar disorder.

Uncovering the etiology of conversion disorder: insights from functional neuroimaging.

Empirical evidence for discrete neurocognitive subgroups in bipolar disorder: clinical implications.

Factor structure of cognition and functional capacity in two studies of schizophrenia and bipolar disorder: Implications for genomic studies.

Neurocognitive predictors of performance-based functional capacity in bipolar disorder.

Glial pathology in bipolar disorder: potential therapeutic implications.

functional outcomes?

Functional neuroimaging correlates of thinking flexibility and knowledge structure in memory: Exploring the relationships between clinical reasoning and diagnostic thinking.