|
|
|
|
|
|
|
|
|
|
| Neuroimaging in Schizophrenia | |||||||||
| Historical Background | |||||||||
| Since Freud's early years as a young neurologist, the hope for objective diagnosis of psychiatric disorders has been alive and well. From Freud's earliest attempts to differentiate hysterical paralysis, loss of sensation, seizure-like episodes, amnesia, and other mental illnesses from certifiable neurologic disorders clinicians have been faced with critical challenges. Most articles have discussed the distinction between the so-called functional and organic disorders with considerable research energy being directed toward being able to differentiate these disorders by mental status examinations or psychological tests. This has been particularly noticed with schizophrenia in which syndromal qualities and the variance in the diagnostic category has been reduced significantly because dimensions, such as good-poor premorbid, paranoid-nonparanoid, positive-negative symptoms, and other such factors, have been considered. Requiring time periods for the manifestation of symptoms along with a better classification of the toxic states and the exclusion of many other biologic complications that can result from psychotic states have further tightened the diagnosis. In spite of these developments, the clinician has still looked forward to more objective and commonly accepted diagnostic criteria (Furlow B. 2000). | |||||||||
| As early as 1798, from observations made as a result of autopsy, John Haslam reported apparent structural changes in schizophrenic brains (Waddington, 1984). With the discovery of imaging technologies, his early observations have finally been confirmed and the genetic and the biologic substrate of schizophrenia seems generally unquestioned now. Although many current theories still point to apparent functional qualities of the disorder, neuroimaging findings of structural change along with dopenergic theories are most compelling. | |||||||||
| During the 1950s and 1960s it became increasingly popular to emphasize psychological and social factors in conceptualizations of schizophrenia. The fact that, despite many years of postmortem examinations, no single underlying neuropathological or neurophysiological factor common to most patients with schizophrenia had yet been identified compounded the prevailing skepticism with which the concept of biological contributions to mental illness was viewed during that period. Schizophrenia was held to be functional rather than organic, and rigorous attempts to link the illness to the brain languished (Arango C, Crespo-Facorro B, 1999). | |||||||||
| Over the past two decades several factors have resulted in a resurgence of interest in a neuropathological basis of schizophrenia and a renewed research effort to find it. One force behind the resurgence has been the emergence of medical treatments, especially pharmacological ones, that incontrovertibly ameliorate major psychotic symptoms. Another factor is the increasing availability of neuroimaging tools that have made it possible to study brain structure and function in great detail during life with little or no risk, discomfort, or inconvenience to the patient. No longer is the search for a neurobiological substrate of schizophrenia limited to indirect or peripheral measurements of trace substances in cerebrospinal fluid (CSF), urine, or serum--measurements that may or may not relate to brain physiology (Hendren RL, et al, 2000). | |||||||||
| Concurrent with the advent of new tools for the direct examination of the brain have been advances in the understanding of human functional neuroanatomy, which, in turn, have resulted in a more enlightened and refined approach to postmortem investigations. At the same time the development and acceptance of reliable, explicit diagnostic criteria that allow greater consensual definition of psychiatric populations helped overcome a major methodological problem that plagued earlier research efforts. | |||||||||
| Schizophrenia is increasingly believed to be based in organic conditions primarily affecting the CNS. Such features as peculiarities of gait and posture, disordered smooth-pursuit eye movements, minor (or soft) neurological signs, and subtle electroencephalographic (EEG) abnormalities all suggest a link between schizophrenia and CNS pathology. More compelling evidence is offered by neuropsychological, neuroradiological, and neurophysiological studies. Although no single pathognomonic neurostructural or neurophysiological abnormality has yet been delineated in schizophrenia, certain cerebral concomitants of the illness have been demonstrated in a consistent fashion. Increasingly, it appears that cerebral dysfunction of a system of interconnected cortical and subcortical brain regions may be present to a lesser or greater degree and may produce more or less psychopathology in individual patients (Bogerts B. 1999). | |||||||||
| Nevertheless, research efforts lead to significant findings and are totally altering our theoretic explanations of disorders that have historically been explained primarily by psychological conflict, object loss, early developmental patterns, or classical and operant learning. As newer and more available technologies make it possible to observe function, cognitive localization, drug effects, and other such factors. | |||||||||
| Neuroimaging In Schizophrenia | |||||||||
| The clinical relevance of structural neuropathological findings lies in their effect on brain function and behavior. At the turn of the century the brain areas targeted for neuropathological examination in schizophrenia were in the neocortex, a reasonable first choice based on Emil Kraepelin's notion of a profound deterioration of personality (be it called dementia precox or schizophrenia) and the fact that the neocortex was the primary site of pathology in the other dementias to which schizophrenia was likened at the time. However, targeting brain areas on the basis of clinical correlates depends on a knowledge of the precise function of the areas being studied. Until relatively recently little was known about the anatomical sites of higher cognitive and psychological functions. For example, the functions of and interrelationships between the largest and most complex areas of the neocortex, the frontal and temporal lobes, are only beginning to be elucidated. Also, knowledge of the limbic areas and diencephalon had been limited. New tools for examining the human brain directly during life have helped to refine current concepts of the roles of some of those brain regions in human behavior and have offered new impetus for their study in postmortem specimens of patients with schizophrenia (Allen DN. et al, 2000). | |||||||||
| Currently, there are three main modalities which are used in brain-imaging techniques in Psychiatry and hence in schizophrenia : computerized tomography (CT), Magnetic Resonance Imaging and various types of perfusion studies, typically examining regional cerebral blood flow (rCBF). | |||||||||
| Computerized Tomography In Schizophrenia | |||||||||
| In CT imaging, CT had some significant advantage over the superior images obtained with MRI. In addition to being less expensive, very accessible, and less likely to provoke anxiety symptoms in patients, CT has a disadvantage of being unable to detect changes in important structures, such as hippocampi and basal ganglia. Aside from the ventricular structures that are easiest to measure, few reliable volumetric measurements have been accessible with CT. It does have the additional advantage of having fewer contraindications, such as those found with MRI (i.e., met al clips, pacemakers, intubation, and so forth). CT imaging does offer good correspondence with gross brain structures, just as with MRI. (Bogerts B,1999). | |||||||||
| In 1990, Raz & Raz summarized much of the neuroimaging literature (1993 studies) with a metanalysis. They clearly found increased lateral ventricular volume with an effect size (Fig 1). This corresponds to a 43% non overlap between the distributions of schizophrenics and controls. Findings also suggested that differences in the third ventricle had a similar effect size. They further found that males were at greater risk for abnormalities, and "the average cumulative length of hospitalization, adjusted for patient's age and duration of illness, predicted ventriculomegaly in schizophrenia". | |||||||||
| Fig (1): Dilation of the lateral ventricles in a CT brain of female (a)and a male (b) Schizophrenic Patients. Radiological examination showed evidence of the patient's increased cerebrospinal fluid in ventricles and sulci. | |||||||||
| In 1991 Hyde et al reported more than 100 controlled studies for which both CT and MRI were used and that showed mild to moderate dilation of the lateral and third ventricles. From review of these studies it was concluded that dilation was not the result of medication taken by patients, since several of these studies control for premorbid brain volumes. It was further asserted from the studies that the ventricular dilation appeared to be static and not progressive in its deterioration. These findings appear relevant to theories of schizophrenia since the diagnostic criteria lead us to at least three different assumptions: (1) schizophrenia must have an origin with a discrete or identifiable pathognomic lesion; (2) because the disorder results in disturbances of reason and higher cortical activity, the cortex must be implicated; and (3) schizophrenia characteristically starts in late adolescence and early adulthood, results in social and economic deterioration and, therefore, is presumably a neurodegenerative illness. The preponderance of studies to now seems to provide broad confirmation of numbers one and two while number three remains questionable. If gliosis is the most expected finding in active neurodegeneration, then recent studies are unconvincing (Denihan A, et al, 2000). | |||||||||
| These assumptions have led to structural studies of the frontal and the temporal lobes as well as the limbic system, since all seem theoretically important. Further elaboration of the enlarged ventricular findings have been reported in a well-controlled twin study by Torrey et al (1994). In monozygotic twins who are discordant for schizophrenia, the ventricular enlargement is virtually always found in the schizophrenic twin but not in the genetically identical well twin. This strongly argues for an environmental insult or an undetermined slow-developing pathognomic process. | |||||||||
| Another study (Shenton ME, et al, 1992) had shown a reduction in the volume of gray matter in the left anterior hippocampus amygdala. This represented an average 19% reduction from normal subjects. The left parahippocampal gyrus was also reduced by 13%. However, in this study, they did not find the overall volume of the temporal lobe to be reduced. They still concluded that the left temporal lobe is clearly implicated by localized reduction of gray matter | |||||||||
| In the reported study, (Buchsbaum MS et al, 1996) volumetric studies of subcortical structures found the thalamus of the patients to be significantly smaller than that of the volunteers. This appeared to be the greatest in the left anterior region. They concluded from their PET findings that differences were greater for metabolism in the weighted thalamic area (rate X area) than for rate per area, a finding consistent with reported greater decreases in total neuron number than in neuron density in the thalami of schizophrenic patients. In another finding, (Breier A, et al, 1992) the limbic, prefrontal cortex and the caudate structures were specifically evaluated with MRI. Breier et al report that schizophrenic patients had significantly reduced volume in both the right and the left amygdala-hippocampal complex. They also found smaller right and left prefrontal volumes and larger left caudate size. Reductions in the right and the left amygdala and the left hippocampus were also reported. An incidental finding was that the right white matter volume in schizophrenic patients was significantly related to right amgdala/hippocampal volumes, data that provide preliminary support for the hypothesis of abnormal limbic-cortical connection in schizophrenia. | |||||||||
| Although some studies (Buchanan RW, et al, 1993) illustrate changes in some structures that are related directly to symptom constellations (negative symptoms seen in enlarged right caudate, but not in the hippocampus and the amygdala), other research fails to include the existence of specific clinical subgroups or point to brain structural changes that are inherently symptomatic of the schizophrenia disease process (Allen DN, et al, 2000). A continuum hypothesis that would account for the diversity of findings may be more justified. The studies showing differences related to specific structures have not been particularly robust and again points to the research dilemma of using behavioral diagnostic criteria to predict brain morphology, and so forth. | |||||||||
| The most defensible observations appear to be the following: (1) there are structural changes in the brains of schizophrenics that appear to affect frontal and temporal lobes more than posterior brain, and the left is generally more implicated than the right. These changes in volume are generally illustrated by increased lateral ventricular size as well as the third ventricle. Inconsistent changes in the size of hippocampi are also commonly observed. (2) Consistent with observed frontal structural changes, which appears to be more white matter than gray, hypofrontality is observed when the patient is required to perform a task that might require activation of frontal-lobe-linked neuronetworks. (3) One of the more consistent findings is a disruption of the limbic system and complex circuitry involving thalamocortical projections that affects patients filtering, organizing, and formulating concepts. It appears to involve subcortical systems and not discreet lesions or related anomolies. (4) The abnormalities in the neurologic substrate cannot be accounted for by medication effect or developmental history, and there is no evidence of neurodegenerative effects (Zakzanis KK et al, 2000). | |||||||||
| MAGNETIC RESONANCE IMAGING IN SCHIZOPHRENIA | |||||||||
| MRI is based on the principle that atoms with an odd number of nucleons have a magnetic dipole and tend to align or spin within a directional polar axis in a normal magnetic field. When stimulated with a radio-frequency (RF) wave, the energy absorption is manifested by a change in the orientation of the original alignment. This is the process called resonance. When the RF source is turned off, the energy is released in a process called realization. This release of energy can be detected as RF signals which are measured and processed into images. Traditionally, there are different weightings that have to do with the way and time duration of the application of the magnetic field. These are typically referred to as T1, T2, and mixed-weighted or proton-density images (Born J,1999). | |||||||||
| Studies of MRI scans were searched for signs of early developmental anomaly, and the scans did not differ from normal in this regard. MRI has also been used extensively to replicate the finding of ventricular enlargement in both acute (DeLisi LE, et al,1991) as well as chronic schizophrenics (Fig 2) (Gur RE, et al, 1994). | |||||||||
| Fig (2) : MRI imaging done in schizophrenia showing ventricular dilation and prominent cortical CSF with distinct enlargement in sucal width. | |||||||||
| Recent well-controlled representative studies of these findings illustrate structural changes. Turetsky et al (1995) obtained quantitative MRI studies to examine specifically the left-frontal and the temporal-lobe CSF volumes. Abnormal brain asymmetry was reported with reduced volume in both left and right frontal lobes. In the subjects who showed greater left parenchymal volume reduction with attendant CSF volume increase, an increased number of negative symptoms were reported. This led to conclusions that there are selected structural deficits in schizophrenia rather than the diffusely undifferentiated CNS abnormalities that are inferred by a preponderance of studies that did not specifically look for volumes (fig 3). This also suggests that the negative symptoms are more associated with temporal-lobe rather than frontal-lobe abnormality. This was reported as consistent with suggestions of temporolimbic prefrontal network abnormality in schizophrenia (Henn FA, Braus DF,1999). | |||||||||
| Fig (3) : left and Right Hemisphere Views of Three-Dimensional Renderings of Significant Differences in Gray Matter, White Matter, and CSF Between Patients With Early-Onset Schizophrenia and Normal Comparison Subjects, as determined by Statistical Parametric Mapping following MRI. | |||||||||
| MRI has made it possible to obtain volumetric measures of lobular differences and look at more specific temporal and limbic structures. In a comprehensive review, Hyde et al (1991) pointed to an established finding of reduced size of the hippocampus and amygdala (fig 4). These are known to be critical components of the limbic system. Several different research labs were reported to have found changes in both neuropathology and neurochemistry related to the hippocampus and the amygdala both through MRIs and autopsy studies (Beckmann H, 1999). | |||||||||
| Fig (4) : reduced size of the hippocampus and amygdala in MRI scan | |||||||||