Schizophrenia: Neuroanatomical Theories

Neuroanatomical Theories

Enlarged Ventricles

The ventricles are the fluid-filled spaces in the center of the brain. The most consistent morphological finding in the literature of schizophrenia is enlarged ventricles which has been confirmed by a large number of CT and MRI studies. The effect size of ven-triculomegaly has been reported to be 0.7 (Raz and Raz, 1990). Seventy-nine percent of the well designed studies report enlarge-ment of lateral ventricles. Lawrie and Abukmeil (1998) report approximately 40% difference in volume between schizophrenia patients and controls across all volumetric MRI studies. It should be noted that although the ventricular increases are statistically significant, the ventricles are not grossly enlarged in most cases. In fact, radiologists most often read CT and MRI scans of patients with schizophrenia as normal. In addition, most studies of ven-tricular size demonstrate overlap between patients and normal control subjects, indicating that many patients have ventricles in the normal range. Nonetheless, enlargement of the ventricles is the first consistently reported finding confirming a brain abnor-mality in schizophrenia.

The pathophysiological significance of larger than normal ventricles is unclear. Enlarged ventricles are most likely a secon-dary manifestation of brain atrophy or some other process resulting in either focal or generalized reductions in brain mass. Indeed, there have been many reports of brain atrophy and reduced mass in the illness (Figure 45.1). Enlarged ventricles have also been reported in first-degree relatives of subjects with schizophrenia (Cannon et al., 1998; Seidman et al., 1997) and in persons suffering from schizotypal personality disorder (Buchsbaum et al., 1997) raising interesting speculations of whether ventriculomegaly may be an indicator of neurodevelopmental risk for schizophrenia (Lencz et al., 2001).

Limbic System

The limbic structures that have been implicated in schizophrenia are the hippocampus, entorhinal cortex, anterior cingulate and amygdala. These structures have important functions for memory (hippocampus), attention (anterior cingulate), and emotional ex-pression and social affiliation (amygdala). The entorhinal cortex serves as a “way station” between hippocampus and neocortex in that neurotransmissions between these regions synapse in the entorhinal cortex. The entorhinal cortex, hippocampus and other components of the parahippocampal gyrus are often considered “mesiotemporal” structures because of their close anatomical and functional relationship.

There are more reports of abnormalities in hippocampal and related mesiotemporal structures than other limbic struc-tures in schizophrenia. In fact, mesiotemporal pathology is con-sistently found in studies of schizophrenia and mesiotemporal structures are leading candidates for the neuroanatomical site of this condition. This region has been implicated by converging brain imaging and postmortem lines of evidence. One of the most consistent MRI morphological findings is reduction in size of the hippocampus. In addition, more than 25 postmortem studies have reported morphological and cytoarchitectural abnormalities in this structure. The findings have included reduced size and cellular number (white matter reductions, and abnormal cell ar-rangement. There is a bilateral reduction of approximately 4% hippocampal volume in schizophrenia. However, reduced hip-pocampal volume is not reported by all studies.

The anterior cingulate has been implicated in schizophre-nia largely because of postmortem findings of reduced gamma-aminobutyric acid (GABA) interneurons. In addition, functional imaging studies have demonstrated altered metabolic activity both at rest and during selective attention tasks in the anterior cingulate in patients with schizophrenia. Thirty-one studies evaluated one or more of the medial temporal lobe structures – hippocampus, amygdala, parahippocampal gyrus, entorhinal cortex – with 77% reporting positive findings; this is one of the higher percentages of abnormalities reported in all regions of interest throughout the brain.

Prefrontal Cortex

The prefrontal cortex is the most anterior portion of the neocor-tex, sitting behind the forehead. It has evolved through lower spe-cies to become one of the largest regions of the human brain, constituting approximately one-third of the cortex. It is respon-sible for some of the most sophisticated human functions. It contains a heteromodal association area that is responsible for integrating information from all other cortical areas as well as from several subcortical regions for the execution of purpose-ful behavior. Among its specific functions are working mem-ory, which involves the temporary storage (seconds to minutes) of information, attention and suppression of interference from internal and external sources. The most inferior portion of the prefrontal cortex, termed the orbital frontal cortex, is involved in emotional expression. Given its unique role, it is not surprising that the prefrontal cortex has been considered in the etiology of schizophrenia.

Indeed, several lines of evidence have implicated the pre-frontal cortex in schizophrenia. CT studies have provided evi-dence for prefrontal atrophy, and some, although not all, MRI studies have found evidence for decreased volume of this struc-ture. One of the earliest observations from functional imaging studies of schizophrenia was reduced perfusion of the frontal lobes. This finding was subsequently replicated by several PET studies suggesting decreased frontal glucose utilization and blood flow, which came to be known as hypofrontality. Subsequent functional imaging studies provided further support for hypof-rontality by demonstrating that patients with schizophrenia failed to activate their frontal lobes to the same degree as normal control subjects when performing frontal cognitive tasks. This finding has been questioned because patients with schizophrenia typi-cally perform poorly in many cognitive paradigms, so it is unclear whether their lack of frontal activation is a primary frontal deficit or secondary to poor cognitive task performance related to factors such as lack of motivation, inattention, or cognitive impairment stemming from nonfrontal regions. Auditory hallucinations were found to be associated with increases in Broca’s area, a portion of the frontal cortex responsible for language production. This find-ing was of interest because it supported a hypothesis that auditory hallucinations were a form of abnormal “inner speech”.

MRI studies employing diffusion tensor imaging have re-ported changes suggestive of an abnormality in white matter con-nectivity possibly due to reduced myelination of fiber tracts in patients with schizophrenia (Buchsbaum et al., 1998). Magnetic resonance spectroscopy (MRS) studies have reported reduced levels of neuronal membrane constituents (phosphomonoesters) and/or increased levels of their breakdown products (phosphodi-esters) in patients with schizophrenia, primarily in the frontal cortex. Such abnormalities have been observed in treatment-na-ive first episode patients and have been correlated with trait-like negative symptoms and neurocognitive performance.

Though sometimes contradictory, the neuroimaging stud-ies consistently report abnormalities in the orbitofrontal region; often, these abnormalities tend to correlate with severity of schizophrenia symptomatology, show gender differences in rela-tion to spatial localization and the gray matter deficits may be more widespread in chronic, as compared with medication-naive first episode patients. Additional support for prefrontal cortical involvement in schizophrenia comes from postmortem studies with a range of findings. There have been reports of reduced cortical thickness, loss of pyramidal cells, malformed cellular architecture, loss of GABA interneurons and evidence of failed neuronal migration. A majority of the abnormalities represented a decline in function suggesting a widespread failure of gene ex-pression. Specifically, abnormalities involving the glycoprotein Reelin were observed in schizophrenia, a finding reported previ-ously by other postmortem studies. Reelin, an extracellular ma-trix glycoprotein secreted from different GABAergic interneu-rons during development and adult life, may be important for the transcription of specific genes necessary for synaptic plasticity and morphological changes associated with learning.

Temporal Lobe

The superior temporal gyrus is involved in auditory processing and, with parts of the inferior parietal cortex, is a heteromodalassociation area that includes Wernicke’s area, a language center. Because of the important role it plays in audition, it was hypoth-esized to be involved in auditory hallucinations. Indeed, MRI studies have found the superior temporal gyrus to be reduced in size in schizophrenia and have found a significant relation-ship between these reductions and the presence of auditory hal-lucinations. Similarly, Wernicke’s area, which is involved in the conception and organization of speech, has been hypothesized to mediate the thought disorder of schizophrenia, particularly conceptual disorganization. Support for this hypothesis comes from a report of a patient with vascular and other lesions of this region that produce Wernicke’s aphasia, a disruption in the or-ganization of speech that resembles the thought disorder of schizophrenia. MRI studies have found a relationship between morphological abnormalities in this region and conceptual dis-organization in schizophrenia. McCarley and colleagues (1999) reviewed 118 MRI studies published from 1988 to 1998; 62% of the 37 studies of whole temporal lobe showed volume reduction and/or abnormal asymmetry especially in the superior temporal gyrus, the highest percentage of any cortical region of interest.

Striatum

The striatum, consisting of the caudate, putamen, globus pal-lidus, substantia nigra and accumbens, is an output center for the cortex and has been traditionally thought to have a primary role in the execution of motor programs. Subsequent studies have demonstrated an important cognitive role for this structure as well. Moreover, in primary diseases of the striatum, such as Parkinson’s and Huntington’s diseases, clinical manifestations include psychosis and other schizophrenia type behavior, which has contributed to interest in this region in the pathophysiology of schizophrenia.

Two related bodies of data are most frequently cited re-garding the role of the striatum in schizophrenia; these concern the mechanism of antipsychotic drugs and postmortem studies of altered dopamine D₂ receptor numbers. The dorsal striatum (caudate and putamen) is the site of the vast majority of D₂ re-ceptors in the brain. All effective antipsychotic drugs antagonize this receptor and thus, by extrapolation, it was reasoned that this region might be central to the pathophysiology of schizophrenia. Moreover, the most consistent postmortem finding in the schizo-phrenia literature is an increased density of striatal D₂ receptors. However, neuroleptic exposure causes up-regulation of D₂ recep-tors, which may account for this postmortem finding. A current view of the antipsychotic mechanism is that the dorsal striatum is involved in mediating the extrapyramidal side effects of antipsy-chotic medications and, based on rodent studies of antipsychotic drug mechanisms, the ventral striatum (nucleus accumbens) may be involved in antipsychotic efficacy. Thus, attention has shifted toward the possible role of the accumbens in mediating the psy-chosis of schizophrenia.

Thalamus

The thalamus is a nucleus that receives subcortical input and out-puts it to the cortex. One theory posits that the thalamus provides a filtering function for sensory input to the cortex. A deficit in thalamic filtering was proposed to account in part for the experi-ential phenomena of being overwhelmed by sensory stimuli re-ported by many patients with schizophrenia. Preclinical studies have demonstrated that antipsychotic drugs modulate thalamic input to the cortex, which has been offered as a model for antip-sychotic drug action. Several MRI studies have reported reduced volume, and functional abnormalities of the thalamus in patients with schizophrenia. Postmortem studies have also found cell loss and reductions in tissue volume in thalamic nuclei. This thalamic tissue reduction is considered as a possible evidence of abnormal circuitry linking the cortex, thalamus and cerebellum.

Neural Circuits

Because of the large number of different neuroanatomical find-ings in studies of schizophrenia and the appreciation that brain function involves integration of several brain regions, current thinking about the neuroanatomy of this illness is centered on neural circuits. It is conceivable that an isolated lesion anywhere in a neural circuit could result in dysfunction of the entire net-work, and therefore spurious conclusions could be drawn by in-vestigating only one component of a neural network. Evidence suggests that schizophrenia may be associated with a decrease in synaptic connectivity of the dorsal prefrontal cortex though this is not reported by all studies. McGlashan and Hoffman (2000) have proposed the Developmentally Reduced Synaptic Connec-tivity (DRSC) model which proposes that cortical gray matter deficits may arise from either reduced baseline synaptic density due to genetic and/or perinatal factors, or excessive pruning of synapses during adolescence and early adulthood or both. There is regionally specific decreased neuronal size in cortical layer III with cytoplasmic atrophy and generally reduced neuropil. The reduced size and increased density of neurons or glia and de-creased cortical thickness suggest that cell processes and synap-tic connections are reduced in schizophrenia. This is consistent with reports of decreased concentrations of synaptic proteins (e.g., synaptophysin). These cell processes and synapses could be lost as a consequence of a neurochemically mediated (through dopamine and or glutamate) synaptic apoptosis that would com-promise cell function and alter brain morphology without, how-ever, producing serious cell injury (and thus inducing glial reac-tions). However, McCarley and colleagues (1999) suggest that the main neural abnormality in schizophrenia involves neural con-nectivity (dendrite/neuropil/gray matter changes) rather than the number of neurons or network size. They suggest that a “failure of inhibition” on the cellular level is present in schizophrenia and may be linked to a “failure of inhibition” at the cognitive level. According to Lafargue and Brasic (2000) abnormalities involv-ing the temporolimbic–prefrontal cerebral circuitry is postulated to underlie the organizational and memory deficits commonly observed in schizophrenia patients. Furthermore, as reviewed by these authors, a possible insult or injury to the mechanism of GABAergic and glutamatergic influence during early corticogen-esis may largely contribute to the later manifestation of clinical schizophrenia. Malfunction of the cooperating sensory systems of excitation and inhibition during the early stages of develop-ment of the brain could result in the failure of “pioneer neurons” properly to differentiate and migrate to their appropriate cerebral locations. Consequently the later migrating projection neurons may fail to reach or invade their preselected area-specific brain sites. A disturbance of the proper GABAergic and glutamatergic influences would upset NMDA mechanisms and normal corti-cal development. If such disturbance is actively occurring from the onset of cerebral ontogeny, the affected individual may suf-fer from the signs and symptoms observed in schizophrenia. A challenge for the future is developing new approaches to examin-ing the brain as an integrated and highly interactive system. An unanswered question is whether the morphological differences reflect hypoplasia (failure to develop) or atrophy (shrinkage)

Electrophysiology

Electroencephalogram (EEG) records the electrical activity of brain, which may reflect the mental functions carried out by the neurons possibly in “real time”. However, the precise localization of this event in the specific brain region is poor. When EEG activ-ity from repeated presentations of a specific stimulus is summed across trials, some potentials related to the specific processing of the target stimulus can be extracted from the EEG and are referred as event-related potentials (ERPs).

The P300 ERP, a positive deflection occurring approxi-mately 300 milliseconds after the introduction of a stimulus, is regarded as a putative biological marker of risk for schizophrenia. The P300 amplitudes are smaller in patients with schizophrenia and is one of the most replicated electrophysiological findings.