Cortical abnormalities in bipolar disorder investigated with MRI and voxel-based morphometry
Introduction
Bipolar disorder (BD) is a severe mental illness affecting approximately 1.3–1.6% of the population worldwide. BD is characterized by the presence of both depressive and manic episodes (Drevets and Todd, 1997). There are two diagnostic subtypes, BDI, in which the manic episodes result in pronounced impairment, and BDII, which is characterized by milder periods of hypomania, where elevated mood and activity are present without debilitation (American Psychiatric Association, 2000).
The etiology and pathophysiology of BD are unknown. Nevertheless, both in vivo neuroimaging and post mortem neuropathological studies have shown that individuals with BD have abnormal reductions in gray matter volume in parts of the medial prefrontal cortex (PFC), amygdala, hippocampal subiculum and ventral striatum (reviewed in Drevets et al., 2004, Strakowski et al., 2005). For example, within the medial PFC, morphometric MRI studies found reductions in gray matter volume in the anterior cingulate cortex (ACC) ventral and anterior to the genu of the corpus callosum (termed “subgenual” and “pregenual”, respectively) in BD subjects relative to healthy controls (Drevets et al., 1997, Lochhead et al., 2004, Lyoo et al., 2004, Sassi et al., 2004). The results of these in vivo MRI studies guided post mortem histopathological studies of the subgenual ACC to identify abnormally reductions in neuronal somal size and glial cell counts and increases in neuronal density in BD (Chana et al., 2003), suggesting that the corresponding decrease in gray matter volume was accounted for by reductions in dendritic arborization (Chana et al., 2003, Drevets, 2004). Additionally, in the pregenual ACC and the Brodmann Area (BA) 9 cortex situated anterior to the ACC (anteromedial and dorsal anterolateral PFC), post mortem studies showed abnormal reductions in neuronal and glial cell density in BD subjects relative to controls (Rajkowska et al., 2001, Todtenkopf et al., 2005).
Neuromorphometric abnormalities also have been identified in BD in basal ganglia and mesiotemporal lobe structures that are anatomically related to these medial PFC structures. In the striatum, a post mortem volumetric study found that the left accumbens area, bilateral external pallidum and right putamen were smaller in subjects with BD than in healthy controls (Baumann et al., 1999). In the hippocampal subiculum, BD samples showed abnormal reductions in the density of apical dendritic spines on pyramidal cells (Rosoklija et al., 2000), the concentrations of mRNA for synaptic proteins post mortem (Eastwood and Harrison, 2000) and the gray matter volume (Nugent et al., 2004), although the entire hippocampal volume generally has not been found to differ between BD and control samples. Finally, several neuroimaging studies reported that the amygdala volume is abnormal in BD, although disagreement exists regarding the direction of these differences with respect to healthy controls (reviewed in Drevets, 2004).
The regions implicated by these studies share extensive, monosynaptic, neuronal connections with each other and with the hypothalamus and periaqueductal gray to form a circuit that plays major roles in modulating emotional behavior (Ongur and Price, 2000). This circuit has been characterized as a “visceromotor network” based upon evidence that its projections function to modulate neuroendocrine, autonomic and behavioral responses to threatening, stressful and reward-related stimuli and contexts (Ongur and Price, 2000). Because BD is associated with disturbances in these domains of emotional behavior, the structural abnormalities found within this visceromotor circuit in BD conceivably could play an integral role in this condition's pathogenesis.
Anatomical studies in humans demonstrate that the medial PFC, ventral striatum, subiculum and amygdala regions involved in visceromotor function also share extensive, reciprocal, monosynaptic connections with areas of the orbitofrontal, posterior cingulate and superior temporal cortices (Ongur and Price, 2000, Kondo et al., 2004, Saleem et al., 2004, Price et al., 2004). These anatomical observations converge with neurophysiological data from human brain mapping studies of various emotional states (see Discussion) to suggest that specific orbitofrontal, posterior cingulate and superior temporal areas also may play roles in modulating emotional behavior. Whether morphometric abnormalities also exist in these regions in BD remains unclear, however.
Few post mortem neuropathological studies have assessed orbitofrontal, posterior cingulate or superior temporal cortices in BD, and preliminary results from morphometric MRI studies of BD either await replication or are in disagreement. In the ventral PFC, one voxel-based morphometry (VBM) study reported a reduction in GM density in the lateral orbital cortex situated on the inferior frontal gyrus (Lyoo et al., 2004), while another found a GM reduction located several centimeters away at the ventromedial pole of the superior frontal gyrus (Wilke et al., 2004). In the posterior cingulate cortex (PCC), one VBM study reported widespread deficits in gray matter “density” throughout the cingulate gyrus, including the PCC (Doris et al., 2004), of “poor outcome” BD, although this finding awaits replication and extension to BD subjects who are not poor outcome. Moreover, it remains unclear whether this difference in GM density is associated with a corresponding reduction in GM volume. In the superior temporal gyrus (STG), a post mortem histopathological study observed reduced glial cell density and reduced clustering of neurons into minicolumns in the planum temporale region in BD (Beasley et al., 2005). However, most MRI studies that employed manual segmentation of the entire STG found no significant volumetric differences between BD and control samples (Hirayasu et al., 2000, Schlaepfer et al., 1994, Brambilla et al., 2003, Kasai et al., 2003), although one reported abnormally decreased STG volume in children and adolescents with BD (Chen et al., 2004) and another found abnormally increased right anterior STG volume in adults with BD (Pearlson et al., 1997). These conflicting results regarding STG volume may reflect limitations in the sensitivity for detecting focally distributed differences between BD and control samples using neuromorphometric methods that sample an entire gyrus (see Discussion).
These conflicting results across studies also may reflect differences between the subject samples with respect to medication status or other clinical variables. In recent years, the mood stabilizing drug lithium was found to exert neurotrophic and neuroprotective effects in rodents and humans (Manji et al., 1999), and lithium treatment has been associated with increases in GM in longitudinal studies of BD (Moore et al., 2000a, Moore et al., 2000b). Other agents used for mood stabilization, such as valproic acid preparations, also exert neurotrophic and neuroprotective effects in rodents (Hao et al., 2004). Subjects with BD also are likely to receive antidepressant drugs, which have been associated with neurotrophic effects (Duman, 2004), and antipsychotic agents, which in some cases increase the GM volume of basal ganglia structures (Gur et al., 1998, Corson et al., 1999). The importance of taking such effects into account in neuromorphometric analyses has only recently been suggested by Sassi et al. (2004) and Drevets (2004), who found abnormal volumetric reductions in the ACC only in untreated BD subjects.
The current study is the first VBM study to assess morphometric abnormalities present in BD samples differentiated on the basis of recent medication status. Because the VBM approach is particularly useful for surveying large regions of cortex and revealing inherent differences in tissue volume between groups, this study specifically focused upon orbital and medial prefrontal cortical, PCC and STG regions that constitute the major cortical components of the visceromotor network.
Section snippets
Subjects
Thirty-six subjects (26 female, age = 39 ± 8.1) meeting Diagnostic and Statistical Manual of Mental Disorders-Version IV (DSM-IV) criteria for Bipolar Disorder Type I (n = 7) or II (n = 29) were imaged. Sixty-five healthy controls (46 female, age = 38 ± 11.8) with no personal or family history of psychiatric disorders in first-degree relatives also were imaged. Subjects were drawn from a larger pool of 39 BD subjects and 95 controls based upon having images of sufficient signal homogeneity to
Results
Demographic and clinical information about the subject samples is presented in Table 1. The mean age and gender composition did not differ significantly between groups. All 36 of the BD subjects met criteria for a current major depressive episode at the time of scanning, and all had experienced multiple episodes (>2) of hypomania or mania in the past. Three unmedicated BD subjects and four recently medicated BD subjects were classified as Type I, and the remaining subjects were Type II. The
Discussion
The VBM analysis showed left lateralized reductions in GM in BD patients relative to controls in the PCC, lateral orbital cortex and STG. In the PCC and orbital cortex, local increases in WM were identified in the vicinity of the areas where GM was reduced. The reciprocal pattern of differences between the GM and WM tissue compartments within the PCC and orbital cortex most likely reflects differences in gyral morphology between BD and healthy brains that would be expected to result from the GM
Summary
This study revealed morphometric abnormalities in BD in regions of the medial and orbital PFC, STG and PCC that, together with anatomically related areas of the amygdala, hippocampus, striatum, thalamus, hypothalamus and brain stem, form part of a visceromotor network that regulates the expression and experience of emotion. The separation of the BD subjects on the basis of recent medication status enabled exploration of psychotropic drug effects on these abnormalities that can facilitate the
Acknowledgments
Thanks to Michele Drevets, Joan Williams and Joana-Taylor Tavares for their work on this project. This research was supported by the Intramural Program of the NIH, NIMH.
References (78)
- et al.
Changes in gray matter volume in patients with bipolar disorder
Biol. Psychiatry
(2005) - et al.
Voxel-based morphometry—The methods
NeuroImage
(2000) - et al.
Evidence for altered neuronal organisation within the planum temporale in major psychiatric disorders
Schizophr. Res.
(2005) - et al.
Effect of sertraline on regional metabolic rate in patients with affective disorder
Biol. Psychiatry
(1997) - et al.
Two-dimensional assessment of cytoarchitecture in the anterior cingulate cortex in major depressive disorder, bipolar disorder, and schizophrenia: evidence for decreased neuronal somal size and increased neuronal density
Biol. Psychiatry
(2003) - et al.
Abnormal left superior temporal gyrus volumes in children and adolescents with bipolar disorder: a magnetic resonance imaging study
Neurosci. Lett.
(2004) - et al.
Reduction of cingulate gray matter density in poor outcome bipolar illness
Psychiatry Res.
(2004) - et al.
Functional anatomical correlates of antidepressant drug treatment assessed using PET measures of regional glucose metabolism
Eur. Neuropsychopharmacol.
(2002) - et al.
A voxel-based morphometric study of ageing in 465 normal adult human brains
NeuroImage
(2001) - et al.
Effects of mood and subtype on cerebral glucose metabolism in treatment-resistant bipolar disorder
Biol. Psychiatry
(2001)