Cingulate gyrus volume and metabolism in the schizophrenia spectrum

Abstract

Background: The cingulate gyrus, which is involved in affect, attention, memory and higher executive functions, has been implicated as a dysfunctional region in schizophrenia. Postmortem studies report cytoarchitectural changes in the anterior cingulate gyrus (ACG) and functioning imaging studies show correlations between the degree of hypometabolism of the anterior cingulate and clinical symptoms in schizophrenia. Methods: Unmedicated patients with schizophrenia (n=27) and schizotypal personality disorder (SPD) (n=13), as well as sex- and age-matched control subjects (n=32), were studied with ¹⁸F-fluorodeoxyglucose positron emission tomography (PET) scans and magnetic resonance imaging (MRI). As a control over mental activity, all subjects performed a verbal working memory task during the PET protocol. The cingulate gyrus was first outlined on the MRI scans and, after coregistration, the coordinates were applied to the PET scans to yield a three-dimensional metabolic map of the cingulate gyrus for each subject. A statistical resampling method was used to analyze the metabolic differences between groups. Results: Compared with controls, patients with schizophrenia had lower relative glucose metabolic rates in the left anterior cingulate and the right posterior cingulate gyrus (PCG) assessed by 3-D significance probability mapping. SPD patients had higher glucose metabolic rates (GMRs) in the left posterior cingulate than did controls. Furthermore, volumetric measurement with MRI showed the left anterior cingulate and Brodmann area 24′ to be smaller in schizophrenic patients than controls. Conclusions: Compared with controls, patients with schizophrenia have metabolic and volumetric reductions in a cingulate gyrus area that is related to higher executive functions. Schizotypal patients rely more on sensory association areas to perform a cognitive task than do controls and seem to be a group that is partially distinct in its physiological and functional characteristics.

Introduction

In man, the cingulate cortex forms an arch, extending from the rostral subcallosal area anteriorly and following the curved superior surface of the corpus callosum bilaterally on the sagittal plane. The posterior region of the cingulate merges with the parahippocampal gyrus via the isthmus. The cingulate gyrus is a part of the limbic lobe, and the arch can be seen on a medial view of each cerebral hemisphere. Brodmann (Brodmann, 1909) designated the rostral region inferior to the genu of the corpus callosum as area 25, the region above the anterior part of the corpus callosum as area 24, and the caudal part as areas 23 and 29. Just anterior to area 24 is cingulo-frontal transition area 32, which lies between area 10, frontal pole, and area 24. The cingulate sulcus typically separates the two, but in some individuals there is one sulcus between areas 32 and 24, while in others there are two sulci with areas 10, 32, and 24 each separated (Vogt et al., 1995). The posterior cingulate gyrus is connected to the associative temporal, mediotemporal, and orbitofrontal cortices and to the medial pulvinar. The anterior cingulate gyrus is connected to intralaminar, mediodorsal, and ventral anterior thalamic nuclei, the amygdala, and the nucleus accumbens septi. Both areas of the cingulate gyrus are interconnected, and they have connections in common with other medial frontal areas, lateral frontal and posterior parietal cortices, the caudate nucleus, the claustrum, and the limbic thalamic nuclei Baleydier and Mauguiere, 1980, Bentivoglio et al., 1993, Vogt et al., 1993.

The anterior and posterior cingulate gyri have distinct functions. Brodmann areas 25 and 24 are involved in the recognition of affective states, execution of affect-related operations, and modulation of autonomic functions. Area 24′ (caudal part of area 24, as reviewed by Devinsky et al., 1995) is involved in the response selection to cognitively demanding tasks and has premotor functions (Devinsky et al., 1995). In contrast to the anterior division of the cingulate gyrus, the posterior division has been implicated in visuospatial behaviors involving post-saccadic neuronal responses and spatial memory Olson et al., 1993, Sutherland and Hoesing, 1993. The posterior cingulate gyrus is also involved in the retrieval of previously learned data regardless of its spatial content Maguire and Mummery, 1999, Van Horn et al., 1998.

Post-mortem studies of schizophrenia indicated cytoarchitectural changes in the anterior cingulate gyrus characterized by a reduced number of non-pyramidal GABAergic neurons, an increased number of excitatory axonal bundles, and a relative increase in the number of dopaminergic inhibitory input to the remaining non-pyramidal cells Benes, 1998, Benes et al., 1992. It is postulated that this “miswiring” may be the cause of “overinclusive thinking,” attentional difficulties and an overall conceptual disorganization in schizophrenia patients (Benes, 1998). Reduced neuronal somal size has also been observed in the anterior cingulate of patients with schizophrenia (Chana et al., 2003).

A relative reduction in frontal metabolic rate and blood flow Andreasen et al., 1992b, Buchsbaum, 1995 among schizophrenia patients has been reported, but these studies have not always differentiated among frontal areas. Large geometrically shaped regions of interest, as were used in the first metabolic positron emission tomography (PET) studies of schizophrenia (Buchsbaum et al., 1982), may include both medial frontal and cingulate areas in the same region of interest. Even when medial areas are specifically assessed with stereotaxic methods, both with glucose metabolic rate (GMR) (Siegel et al., 1993) and cerebral blood flow (Andreasen et al., 1992b), observations of lower activity might include the cingulo-frontal transitional region (Brodmann area 32), which is adjacent to the cingulate region (Brodmann area 24). Tamminga et al. (1992) reported lower GMR in selected regions of the anterior cingulate cortex of schizophrenia patients than controls, but regions of interest were identified on the PET scans themselves rather than anatomically, and no comparison with medial frontal areas was reported. More recently, alterations in the cingulate functions of schizophrenia patients doing word-recall and attentional tasks emphasizing the involvement of this area in cognitive processing difficulties have been described Carter et al., 1997, Crespo-Facorro et al., 1999, Dehaene et al., 2003, Hofer et al., 2003, Nohara et al., 2000, Yucel et al., 2002. The anterior cingulate gyrus is a part of a circuitry that modulates the startle reflex, and this function is reportedly impaired in schizophrenia (Hazlett et al., 1998b). Fletcher et al. (1999) hypothesize that one of the core elements in schizophrenia is the disconnectivity of the dorsal–frontal/temporal circuitries modulated by a functionally impaired anterior cingulate. Supporting this hypothesis, a recent PET study showed that schizophrenia patients performing a working memory task had altered fronto-temporal connections compared with the control group and that this metabolic pattern was a trait marker for schizophrenia in the whole cohort (Meyer-Lindenberg et al., 2001).

Many investigators have found correlations between psychotic processing and altered blood flow, metabolism, or event related potentials in the cingulate gyrus. Liddle et al. (1992) reported a positive correlation between perfusion in pixels in the region of the anterior cingulate cortex and behavioral disorganization as assessed from factor scores of the Manchester scale (items loading for inappropriate affect, thought disorder, and poverty of speech), but the significant region extended forward from area 24 through area 32 into area 10 (their Fig. 1). Silbersweig et al. (1995) found anterior cingulate blood flow to be maximal during hallucinations (Talairach coordinates 12, 10 and 32; Talairach and Tournoux, 1988), but the peak was just anterior to the cingulate sulcus and centered in area 9, not 32. Also, formal thought disorder in schizophrenia patients is associated with decreased blood flow in cingulate cortex (McGuire et al., 1998). Low resolution electromagnetic tomography also demonstrated significantly diminished brain activity in the anterior cingulate of schizophrenic patients compared with controls (Fallgatter et al., 2003).

The impairment of dopaminergic modulation of the anterior cingulate gyrus (ACG) activation is well demonstrated in schizophrenia. While performing a verbal fluency task, patients fail to activate the ACG, but low-dose apomorphine administration, presumably by blocking presynaptic dopamine transmission, normalizes the task activation of the structure (Fletcher et al., 1996). Patients with schizophrenia also respond poorly to neuroleptic agents if they have lower GMR in the ACG while performing an attentional task (Cohen et al., 1998).

Magnetic resonance imaging (MRI) studies with volumetric measurements of the cingulate gyrus are few in number and inconclusive. Measured on several comparable MRI slices, the anterior cingulate, bilaterally, was found to be smaller in schizophrenia patients than in controls Goldstein et al., 1999, Noga et al., 1995. On the basis of significance probability maps, several investigators reported shrinkage of the anterior cingulate gyrus gray matter in patients compared with controls Job et al., 2002, Kubicki et al., 2002, Shapleske et al., 2002, Velakoulis et al., 2002. Only a few studies traced consecutive cingulate slices for volumetric measurement: one reported smaller anterior cingulate (Takahashi et al., 2003) in patients, two others reported smaller anterior cingulate only in female patients Goldstein et al., 2002, Takahashi et al., 2002a, while others reported no differences in the volume of the anterior cingulate in patients compared with controls (Convit et al., 2001). Anisotropy maps from diffusion tensor MRI were used to identify cingulate volume and also showed reduced cingulate fasciculus volume (Kubicki et al., 2003). In a study of schizotypal personality disorder, patients did not differ from controls in the volume of the anterior cingulate gyrus, but female schizotypal patients, like patients with schizophrenia, failed to show the normal pattern of hemispheric gray–white matter asymmetries (Takahashi et al., 2002b).

Previously we developed a stereotaxic method using MR images available in a subset of subjects and examined the metabolic rate in the anterior and posterior cingulate cortex of a cohort of schizophrenia patients during the performance of an attentional task. We reported lower metabolic rate in the anterior cingulate cortex in the schizophrenia patients (sample from University of California, Irvine) compared with controls (Haznedar et al., 1997a). Patients from our current cohort (Mount Sinai School of Medicine) were entered into an earlier analysis using a stereotaxic method based on average brain location to measure glucose metabolism in the frontal cortex, and we reported lower GMR in Brodmann area 24 Buchsbaum et al., 2002, Hazlett et al., 2000, but the entire cingulate gyrus was not assessed nor were intracingulate statistical contrasts made. We have now improved our methodology by employing the gold standard method of tracing the cingulate on individual MR images coregistered to FDG-PET. This allows both the volume and the activity of the cingulate to be assessed. We now use a three-dimensional tracing technique that involves tracing the entire cingulate gyrus on contiguous 1.2-mm MR images for each subject and a three-dimensional statistical parametric mapping technique with resampling. In addition to schizophrenia, this enhanced analysis is extended to schizotypal personality disorder. Schizotypal personality disorder (SPD) is genetically related to schizophrenia (Battaglia and Torgersen, 1996), shows neurophysiological impairments similar to those in schizophrenia Cadenhead et al., 1996, O'Driscoll et al., 1998 and may share with schizophrenia some unique structural and/or functional changes in the brain Buchsbaum et al., 1997, Downhill et al., 2001. Compared with schizophrenia patients, SPD patients are mostly drug-naı̈ve and do not suffer from the effects of institutionalization.

We hypothesized that schizophrenia patients would have smaller anterior cingulate gyrus volumes compared with controls and that SPD patients would have cingulate volumes intermediate between those of the schizophrenia and control groups. We also hypothesized that compared with controls, schizophrenia patients would have reduced GMR in areas of the anterior cingulate related to higher executive function and that SPD patients would have intermediate values but might activate alternative regions from controls as a compensatory mechanism. This study provides an entirely independent analysis of cingulate volume from our recent report (Mitelman et al., 2004) using a completely nonoverlapping patient and normal cohort and an anatomical rather than stereotaxic assessment of cingulate volume.