Psychiatric disorders are well known to have a significant heritable component and this has been demonstrated conclusively in family, twin and adoption studies. Despite this, it has proven difficult to identify any individual or multiple biologically linked genetic parameters that explain a substantial proportion of this heritable variance.
Modern genetics technology allows characterization of genetic variants across the entire human genome. These approaches have not only confirmed the role of genetic factors for risk of psychiatric disorders and highlighted the fact that such risk is likely conferred by a multitude of factors with small individual effects, but also demonstrated that a substantial portion of genetic risk is shared between different psychiatric illnesses (Kendler, 2013).
While some of the illness linked variants were found to aggregate in certain biological pathways, it is currently not possible to predict genetic illness risk at the individual patient level to an extent that would be useful clinically. IMAGEMEND, however, has a clear multi-modal focus, aiming to integrate genetic readouts with those of brain structure and function as well as those of environmental risk. This will give an unprecedented opportunity to bridge the gap between genetic underpinning, environmental influences and alterations at the neural systems level and may lead to a new generation of clinical tools to aid in the clinical management of psychiatric illnesses.
The IMAGEMEND consortium is also in an ideal position to evaluate the alignment of such illness-associated, multi-modal biological patterns with currently established diagnostic constructs, which may uncover biologically defined and potentially trans-diagnostic subgroups of patients and highlight biological mechanisms that may be the basis for development of novel therapeutics.
It is widely believed that psychiatric illnesses are disorders of the brain. In the past 30 years many neuroimaging techniques have been successfully developed to investigate the brain of the living patient. Among those, magnetic resonance imaging (MRI) represents a unique noninvasive technology for imaging of brain structure and function without exposing patients to potentially harmful radiation. Structural MRI quantifies shape and size of brain structures while functional MRI explores differences in brain blood oxygenation as an indirect readout of activity.
Studies that have used MRI technology extensively reported both structural and functional alterations in various psychiatric disorders. However, these brain alterations are subtle and distributed in nature, thus limiting the utility of MRI in the clinical management of these disorders. IMAGEMEND will attempt to address this problem through statistical integration of neuroimaging data with genetic, clinic and environmental data in order to obtain “multimodal profiles” to differentiate diagnosis and predict the course of psychiatric illnesses.
Finally, IMAGEMEND will be utilizing neuroimaging technology to advance the development of a real-time functional MRI based neurofeedback that represents an entirely new treatment modality for psychiatric disorders. Specifically, using this integrated imaging system, patients will be able to directly and non-invasively modify illness relevant brain circuits. This will open a new avenue for psychiatric intervention in addition to the currently available pharmacotherapy and psychotherapy.
Despite the broad availability of neuroimaging facilities such as MRI and their vast utilization to study psychiatric disorders, no neuroimaging applications have reached clinical practice in psychiatry except for excluding somatic, especially neurological disorders and quantification of atrophies in some dementias. In general, no diagnostic of predictive test for psychiatric disorders has reached clinical practice. This also includes genetic tests, which have, despite the volume of genome-wide association studies performed on patients with schizophrenia and bipolar disorder, insufficient classification performance to be useful for individual subject predictions. IMAGEMEND is targeted at the development of diagnostic and predictive tools aimed at several central issues in clinical management of mental illness and addresses an enormous public heath need. Specifically, an objective tool to discriminate schizophrenia, bipolar disorder and ADHD is especially in early disease phases are urgently required for more accurate differential diagnosis. Similarly, a neuroimaging tool to predict response and side-effect occurrence would be invaluable in clinical practice to objectify the process of treatment selection and achieve better patient outcomes through stratification at the biological level.
Numerous genetic studies have demonstrated that schizophrenia, bipolar disorder and ADHD are highly heritable disorders. For schizophrenia, illness risk exceeds 40% in monocygotic twins and 6.5% in first degree relatives (Cardno et al. 1999; Kendler et al. 1993). Similar heritability estimates have been found for bipolar disorder (Kieseppä et al. 2004) and ADHD has been reported to have even higher heritability between 60% and 90% (Todd 2000; Faraone et al. 2005).
Despite this, studies investigating genetic markers of these illnesses at a whole-genome level have had limited success in explaining substantial parts of this heritable variance, likely in part because the genetic architecture is complex and includes pleiotropic, epistatic as well as gene-environment interaction contributions. At the individual marker level, observed genetic variation is partially consistent with changes in neurotransmitter (e.g. dopamine) systems hypothesized to be involved in the diseases’ aetiologies. Additionally, genetic association studies have implicated other biological processes in the aetiology of mental illnesses, such as neurodevelopment, neuron differentiation and cell structural processes, including synaptic scaffolding (Addington and Rapoport 2011). At the brain level, a broad spectrum of illness associated differences including structural and functional changes have been observed using imaging technologies.
Most of these studies have, however, been performed in small cohorts and have not been validated in independent datasets. In particular for schizophrenia, numerous studies have also investigated changes in peripheral molecular systems and found abnormalities in glucose metabolism, immunological changes and altered growth factor levels (Insel 2010).
Nevertheless, this evidence has so far been correlative without a clear link to potential aetiological processes affecting brain function. This is likely due to lack of multi-modal investigations where genetic, imaging and other data are combined in a single analytic framework.
