Hippocampal pathology in individuals at ultra-high risk for psychosis: A multi-modal magnetic resonance study
Introduction
Numerous alterations of hippocampal structure, function and biochemistry have been noted in schizophrenia (Keshavan et al., 2008). Histological studies have demonstrated an overall decrease in hippocampal volume in schizophrenia patients compared to controls, largely attributable to reductions in neural size rather than neuronal loss (Harrison, 1999). This volume reduction has been demonstrated by magnetic resonance imaging in both chronic (Nelson et al., 1998) and first episode schizophrenia (Steen et al., 2006), suggesting that these may represent some of the earliest anatomical changes in the disorder. Further, neuropathological studies have observed an absence of gliosis in post-mortem tissue in schizophrenia, indicating that pathology has a neurodevelopmental rather than neurodegenerative basis (Wong and Van Tol, 2003) (although other explanations have been proffered such as Berger et al., 2003).
A key implication of these findings is that hippocampal volume reduction will be present prior to the onset of the disorder. Although there is some limited evidence for this (Witthaus et al., 2010), the vast majority of studies have failed to demonstrate that significantly smaller hippocampal volumes predict later psychosis onset in various at-risk samples (eg Johnstone et al., 2005, Velakoulis et al., 2006), and instead show that medial temporal volume reduction develops as the disease progresses, at least during the onset phase (Job et al., 2005, Pantelis et al., 2003).
The fact that previous volumetric studies have failed to show hippocampal differences between at-risk individuals who later develop psychosis and healthy controls does not necessarily mean there is no pathology present. Neuropathological and neuroimaging studies suggest that the nature of hippocampal pathology is complex and subtle (Harrison, 1999), and a more productive approach may be to use MRI techniques sensitive to changes at the microscopic level. One such approach is to study the concentrations of hippocampal metabolites using magnetic resonance spectroscopy (MRS). The most commonly studied metabolite is N-acetylaspartate (NAA), reductions in which are thought to be a marker of neuronal loss or dysfunction (Baslow, 2000). As with volumetric studies, there is strong evidence for lower NAA concentrations in the hippocampi of patients with established schizophrenia (Steen et al., 2005), although this is much less clear in the first episode (Fannon et al., 2003, Wood et al., 2008) and absent in populations at clinical high-risk for the disorder (Stone et al., 2009, Wood et al., 2003). Metabolites other than NAA may also be of interest, such as trimethylamines (choline-containing metabolites; TMA) or myo-inositol (mI), which both play roles in second messenger systems (Fisher et al., 1992, Hodgkin et al., 1998).
It is possible that the methods previously used to detect hippocampal differences prior to transition to psychosis were not sensitive enough, either because they were not measuring the appropriate substrate, or because the technique was not sufficiently accurate. For example, all but one of the previous imaging studies of at-risk individuals have been at a field strength of 1.5 T rather than 3 T, which results in reduced signal-to-noise ratio (particularly a problem for MRS) and increased variance in volume measurement (Briellman et al., 2001). Two recent publications have instead demonstrated alterations in hippocampal perfusion prior to transition to psychosis, either chronically (Schobel et al., 2009) or acutely in response to a cognitive task (Allen et al., in press). This provides evidence for hippocampal changes before the onset of first episode disorder.
Another alternative neuroimaging technique that is known to be highly sensitive to the presence of neuropathological changes is T2 relaxometry. T2 relaxometry offers a non-specific indicator of neuronal pathology, with relaxation times dependent on protons in macromolecules, iron concentration, mobile fatty acids, and bound and free water molecules (Bartlett et al., 2007, Whittall et al., 1997). Increases in T2 relaxation times have been associated with multiple types of neuropathology including demyelination, oedema, gliosis and axonal loss (Eriksson et al., 2007), as well as potentially being sensitive to chronic changes in perfusion (Teicher et al., 2000). This last point is particularly useful given the recent findings of chronic hyperperfusion in a population at clinical high-risk for psychosis (Schobel et al., 2009).
There are presently no published T2 relaxometry studies of individuals at-risk for psychosis. The few published studies have all been conducted with chronic schizophrenia cohorts, and on the whole suggest higher T2 relaxation times in various structures in the left hemisphere, including the hippocampus (Andreasen et al., 1991, Pfefferbaum et al., 1999, Supprian et al., 1997, Williamson et al., 1992). Furthermore, these studies have small sample sizes and have not related T2 data to more commonly reported indices of brain integrity, such as volume or NAA. In the current study, we focused on the hippocampi of an at-risk cohort (defined on the basis of current symptoms and decline in function (Yung et al., 2004b)) using three different methodologies; hippocampal volume, T2 relaxation time and proton MRS. At-risk individuals were followed up over a 2-year period, and compared based on whether or not they made a transition to frank psychosis. Based on our previous studies, we predicted that subjects who subsequently made the transition to psychotic disorder would demonstrate subtle hippocampal pathology at baseline compared to the other groups. Specifically, we expected to see elevation of hippocampal T2 without significant reductions in hippocampal volume.
Section snippets
Ultra-high risk (UHR) group
66 UHR participants (25 male; 38%) were recruited through the Personal Assessment and Crisis Evaluation (PACE) Clinic, ORYGEN Youth Health (Yung et al., 2004a). UHR participants met criteria for at least one of the following groups: (i) State and trait risk factors, (ii) Attenuated psychotic symptoms, and (iii) Brief Limited Intermittent Psychotic Symptoms (BLIPS), as assessed by the Comprehensive Assessment of At-Risk Mental State (CAARMS; Yung et al., 2005). The presence and severity of
Participant demographics
There were no differences in gender ratio between the UHR and comparison groups (χ²(1,94) = 0.000, p = 0.996). The comparison group was significantly older (M = 21.14 years, SD = 4.74 years) than the UHR group (M = 19.18 years, SD = 3.17 years); t(93) = 2.03, p = 0.049. All subsequent analyses included age as a covariate, although the two UHR groups did not significantly differ from each other on age. The two UHR groups did not differ on the total negative symptoms subscale (UHR-NP M = 14.8, SD = 4.7; UHR-P M = 15.7, SD =
Discussion
This is the first study to combine multiple modalities of imaging to investigate the integrity of the hippocampi in schizophrenia research, and also the first to examine T2 relaxometry in individuals at high risk for psychosis. We showed that, while hippocampal volume was reduced in the UHR group, neither medial temporal metabolite concentrations nor hippocampal T2 relaxation times were significantly different from controls. When we divided the UHR group on the basis of later transition to
Conflict of interest statement
No funding body had any role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript. A/Prof Wood had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Acknowledgments
This work was supported by a Project Grant (ID: 145627) and two Program Grants (IDs: 350421, 566529) from the National Health and Medical Research Council Australia (NHMRC), and by the Colonial Foundation. A/Profs Wood and Yücel were supported by Clinical Career Development Awards, and Profs Yung and Pantelis were supported by Senior Research Fellowships, both from the NHMRC. Drs Nelson and Seal were supported by Ronald Phillip Griffith Fellowships from the University of Melbourne, and both Dr
References (65)
- et al.
T1 and T2 relaxation times in schizophrenia as measured with magnetic resonance imaging
Schizophrenia Research
(1991) Does stress damage the brain?
Biological Psychiatry
(1999)- et al.
The hippocampus in families with schizophrenia in relation to obstetric complications
Schizophrenia Research
(2008) - et al.
Correlation between quantitative MRI and neuropathology in epilepsy surgical retration specimens—T2 correlates with neuronal tissue gray matter
NeuroImage
(2007) - et al.
Selective deficit of hippocampal N-Acetylaspartate in antipsychotic-naive patients with schizophrenia
Biological Psychiatry
(2003) - et al.
Pituitary volume predicts future transition to psychosis in individuals at ultra-high-risk of developing psychosis
Biological Psychiatry
(2005) - et al.
Diacylglycerols and phosphatidates: which molecular species are intracellular messengers?
Trends in Biochemical Sciences
(1998) - et al.
Proton magnetic resonance spectroscopy in subjects at risk for schizophrenia
Schizophrenia Research
(2006) - et al.
Grey matter changes over time in high risk subjects developing schizophrenia
NeuroImage
(2005) - et al.
Schizophrenia, “just the facts”: what we know in 2008 Part 3: Neurobiology
Schizophrenia Research
(2008)