Neil D. Woodward - US grants

DESCRIPTION (provided by applicant): Processing speed is among the most impaired neuropsychological domains in schizophrenia and is strongly associated with multiple dimensions of functional outcome, including self-care management and employment. Developing effective treatments for processing speed may improve functional outcome and reduce illness burden. Unfortunately, commonly used clinical neuropsychological tests lack the precision to delineate and measure specific cognitive mechanisms contributing to processing speed impairment. This has hampered efforts to understand the neural basis of processing speed dysfunction and is a significant barrier to the development of pro-cognitive interventions. This proposal seeks to address these barriers by using cognitive neuroscience-based tasks, namely single and dual-task choice reaction time (RT) experiments, to investigate processing speed impairment in schizophrenia. Cognitive models derived from these tasks posit that information processing speed is comprised of 3 stages: perceptual analysis, response selection, and response production. The available evidence from the handful of studies employing these paradigms suggests that response selection is differentially impaired in schizophrenia. However, it is not known if the impairment results from dysfunction of a unitary response selection process that cuts across sensory and motor modalities, or is modality specific. Resolving this issue is crucial to understanding the neural basis and treatment of the deficit. Generalized impairment would implicate dysfunction of a central, amodal response selection network and indicate that training programs shown to improve response selection and the speed of neural processing in amodal response selection brain regions should be attempted in schizophrenia. Conversely, modality-specific impairment would implicate dysfunction of modality-specific neural machinery and suggest that interventions targeting specific sensory and/or motor processes might be more effective at improving processing speed in schizophrenia. Evidence that auditory training improves verbal learning and memory provides a striking demonstration of a modality-specific approach to improving cognition in schizophrenia. The specific aims of this proposal are: 1) test the hypothesis that response selection impairment in schizophrenia generalizes across sensory and motor modalities; and 2) examine the integrity of multi-modal response selection brain networks and modality-specific brain regions in schizophrenia. The anticipated results of this work will further our understanding of the specific cognitive operations contributing to slowed processing speed in schizophrenia, identify candidate biomarkers of processing speed dysfunction, and lend guidance to the development and selection of pro-cognitive interventions. PUBLIC HEALTH RELEVANCE: Cognitive impairment in schizophrenia leads to profound disability and is a robust predictor of multiple dimensions of outcome, including self-care management, social functioning, and employment. This proposal will use cognitive neuroscience based methods to better understand the nature and neural basis of cognitive impairment in schizophrenia. The results of this work will help guide the selection of existing therapies to improving cognition and facilitate the development of novel pharmacological and non-pharmacological pro-cognitive interventions.

DESCRIPTION (provided by applicant): Autism spectrum disorders (ASD) are a group of pervasive developmental disorders characterized by deficits in social interaction and communication, and rigid or stereotypical behaviors. These complex behavioral problems arise, in part, from deficits in more basic sensory and motor systems. The thalamus plays a central role in processing sensory information and modulating overall cortical activity. Consequently, dysfunction of the thalamus and its connections with the cortex may account for some of the symptoms of ASD. Indeed, there is growing evidence that the structure and function of the thalamus is abnormal in ASD and contributes to ASD symptoms. Despite emerging evidence of thalamic abnormalities in ASD, several critical gaps in our knowledge about thalamic pathology in ASD remain. First, thalamocortical networks are organized topographically such that separate thalamic nuclei project to specific cortical areas raising the distinct possibility that thalamocortical networks may selectively affected in ASD. Second, we know virtually nothing about the post-natal developmental trajectories of thalamocortical networks in ASD. This is critical given mounting evidence that the developmental trajectories of the brain are altered in ASD. Moreover, the severity of ASD symptoms varies across the lifespan. Consequently, a complete understanding of thalamic circuitry in ASD and the relationship between thalamocortical network disturbances and ASD phenotypes will require systematic mapping of thalamocortical networks across the lifespan. This proposal will begin to address these critical gaps. Using an innovative resting-state fMRI method, we will determine if thalamocortical networks are altered in ASD (Aim 1); examine the developmental trajectories of thalamocortical networks in ASD (Aim 2); and establish the functional relevance of thalamocortical networks to the expression of ASD symptoms (Aim 3). The proposed work will contribute to our understanding of brain-basis of ASD.

DESCRIPTION (provided by applicant): This project will investigate thalamocortical networks in schizophrenia (SZ) and psychotic bipolar disorder (BD). Despite considerable evidence that the thalamus is abnormal in psychosis, several knowledge gaps must be addressed before thalamic pathology can be established as a biomarker. First, although attention has focused on the mediodorsal (MD) nucleus, conflicting findings from post-mortem studies and limitations of conventional imaging approaches has made it difficult to establish the anatomical specificity of thalamic pathology. Second, it is not known if thalamocortical dysconnectivity extends to psychotic BD. Overlapping deficits in cognitive functions supported by the thalamic circuitry, including working memory (WM), suggests this may be the case. However, differences in the pathways leading to cognitive impairment also predict there will be differences. In SZ, cognitive impairment is superimposed on a background of compromised pre-morbid functioning and remains stable across illness stages. In BD, pre-morbid functioning is intact and cognitive impairment is relatively modest in the early stage of the illness; whereas chronic patients are virtually indistinguishable from SZ. This has led to different etiological models of psychosis: SZ is conceptualized as a neurodevelopment disorder and BD a neuroprogressive illness. These models predict overlapping abnormalities in thalamocortical circuitry in chronic patients, but differential impairment in the early stage of psychosis. Finally, despite compelling evidence from animal models linking thalamocortical dysconnectivity to WM, the functional consequences of thalamocortical network dysfunction are poorly understood. In separate studies of chronic (Aim 1) and early stage psychosis (Aim 2), we will determine if thalamocortical network pathology varies in accordance with the different trajectories of cognitive impairment in psychotic disorders, and is related to WM impairment (Aim 3). We will test the following specific hypotheses: 1) in chronic psychosis, both SZ and psychotic BP patients will exhibit reduced connectivity between the PFC and MD thalamus; and 2) in early stage psychosis, SZ, but not psychotic BP, will exhibit reduced PFC-MD thalamus connectivity. Additionally, we will test the hypothesis that somatomotor hyper-connectivity observed in prior studies of chronic SZ, which we proposed is due to atypical pre-morbid brain maturation, is present in early stage SZ, but not early stage or chronic psychotic BP. If confirmed, these hypotheses will provide a powerful approach to differentiating psychotic disorders at both early and chronic stages of the illness, and further support the different etiological models of SZ and BD. Alternatively, the results might indicate there is greater overlap in thalamocortical pathology between SZ and BD than appreciated and challenge the different etiological models of these disorders. Moreover, examining functional connectivity during WM may provide translational evidence supporting a mechanistic thalamocortical dysconnectivity model of WM impairment and identify potential treatment targets for ameliorating cognitive impairment.

Project Summary: Brain networks linking the cortex to thalamus are critical for cognitive, sensory, and motor functioning. Dysruption of thalamocortical networks has been implicated in the pathophysiology of neurodevelopmental disorders, including psychosis, and mechanisms of clinical phenotypes, especially cognitive impairment. This view is supported by neuroimaging, including a series of studies by the Co-PIs, which consistently find a combination of reduced thalamic connectivity with the prefrontal cortex (PFC) and sensorimotor-thalamic hyper-connectivity in schizophrenia and bipolar disorder. While significant progress has been made, critical knowledge gaps remain with respect to the normal developmental trajectory of thalamocortical networks and onset of connectivity disturbances in psychosis; relationship between cognitive functions supported by thalamocortical circuits and thalamocortical connectivity biomarkers; and clinical utility of imaging thalamocortical networks. The availability of several large-scale cross-sectional datasets containing multi-modal neuroimaging data and extensive phenotypic data on healthy and at-risk individuals has created an unprecedented opportunity to address these critical knowledge gaps. They include the Cambridge Center for Ageing and Neuroscience (Cam-CAN: n=656, ages 18-88); Nathan Kline Institute-Rockland Sample (NKI- RS: n=932, ages 6-85); Pediatric Imaging, Neurocognition, and Genetics dataset (PING: N=1239, ages 3-20); Philadelphia Neurodevelopmental Cohort (PNC: n=1601, ages 8-21), which includes psychosis spectrum (PS) youth; and the North American Prodromal Longitudinal Study (NAPLS: n=397). Building on our prior work in clinical populations and leveraging the considerable resources of these datasets, we propose to chart the development of thalamocortical networks in healthy subjects (Aim 1) and youth with PS symptoms (Aim 2), characterize the associations between thalamocortical biomarkers and cognition (Aim 3), and investigate the clinical utility of thalamocortical connectivity biomarkers at identifying atypical brain development in individual subjects across the psychosis continuum and predicting conversion to psychosis in clinical high risk individuals (Aim 4). The proposed Aims will: 1) establish critical normative lifespan development data for refining and testing etiological models of not just psychosis, but other neurodevelopmental disorders, as well as aging- related disorders; 2) inform the pathophysiology of psychotic disorders and contribute to dimensional models of psychosis; 3) help define the neural basis of executive cognitive abilities thereby providing the necessary foundation for mechanistic models of normal cognitive function and cognitive impairment in psychosis; and 4) potentially provide risk biomarkers and intervention targets for youth at risk for developing a psychotic disorder.

An increasing body of evidence on the structure of psychopathology indicates the presence of a general factor of psychopathology (GF), also referred to as a P-factor, that explains a significant amount of the common variance in the expression of multiple psychopathological symptoms and disorders. The presence of a GF has substantial implications for understanding psychopathology because it suggests that a comprehensive understanding of mental illness requires an untangling of nonspecific risk factors from more narrow, dimension- specific risk factors or features. Caspi et al., (2014) suggest that schizophrenia symptoms are so highly correlated with the GF, that they are mostly an expression of the GF. However, methodological limitations, such as the use of community and youth samples, limit conclusions from the existing literature. We propose to test the hypothesis that schizophrenia spectrum (SS) symptoms load heavily on both a GF AND a separate higher-order psychosis factor using a sample of 1000 adult psychiatric and medical treatment seeking subjects that includes a substantial group of patients with SS disorders as well as patients with significant externalizing and internalizing symptoms. The potential utility of this hierarchical dimensional approach is that it allows quantification of the extent to which each individual possesses a high GF score (reflected in the overall breadth of symptoms), or symptoms that are more narrowly constrained to a specific 2nd order factor (such as externalizing, internalizing, and psychosis factors) or even more narrow, first-order symptom dimensions. Using this quantitative approach, we will test the extent to which neuropsychological and structural and functional MRI measures that have previously been observed in patients with schizophrenia are more strongly related to the GF versus a psychosis factor. This will allow us to test the hypothesis that some neural correlates, such as the volume of the anterior cingulate area, are nonspecific correlates of the GF, while others, such as temporal cortical sensory processing abnormalities, are specific to the psychosis, and remain even after controlling for the GF. In order to test the prognostic significance of the GF, we will additionally test whether scores on the GF are predictive of course of illness in 50 SS patients experiencing first-episode psychosis. Taken together, the study will provide the most comprehensive test to date of the relevance of GF model to understanding the expression and neural correlates of severe psychopathology,

Schizophrenia is a severe and heterogeneous mental disorder that impacts most domains of function including behavior, cognition, and emotion. Recent models have highlighted important alterations of the emotion brain networks in schizophrenia that contribute to schizophrenia symptoms, like paranoia and delusions. To date, the studies of emotion in schizophrenia have primarily focused on fear processing and have shown heightened amygdala responses to neutral stimuli and altered amygdala-prefrontal cortex connectivity. However, recent research suggests that another brain region?the bed nucleus of the stria terminalis (BNST)?may play a critical role in anxiety and that BNST-mediated anxiety is distinct from amygdala-mediated fear. The RDoC?s Negative Valence System recognizes this fear-anxiety distinction and has separate constructs for Response to Acute Threat (amygdala) and Response to Potential Harm (BNST). To our knowledge, the BNST has yet to be examined in individuals with schizophrenia. Using methods pioneered by our lab to study the human BNST, we have collected preliminary data in schizophrenia. Our pilot data provides initial evidence for BNST connectivity differences in both response to unpredictable threat, a measure of the response to potential harm construct, and during a resting state in individuals with schizophrenia compared to healthy controls. Further, we found evidence that BNST alterations in schizophrenia differ for those who do or do not have comorbid anxiety. Individuals with schizophrenia and anxiety disorders demonstrated stronger connectivity between BNST and multiple brain regions involved in threat detection, uncertainty, and anxiety relative to those with schizophrenia and no anxiety disorder. The current study will investigate BNST connectivity in three groups: individuals with schizophrenia with a comorbid anxiety disorder (SZ+ANX), individuals with schizophrenia without a comorbid anxiety disorder (SZ-ANX), and healthy controls (HC). We hypothesize that individuals with schizophrenia will have altered BNST connectivity in response to unpredictable threat and altered BNST intrinsic connectivity relative to HC. In addition we predict that SZ+ANX group will show BNST hyperconnectivity relative to SZ- ANX. We will test these hypotheses with three specific aims. (1) Investigate BNST connectivity in response to unpredictable threat in individuals with schizophrenia; (2) Determine whether there are differences in BNST intrinsic connectivity in individuals with schizophrenia; (3) Test for relationships among BNST connectivity, stress responses (skin conductance and cortisol), and clinical symptoms in schizophrenia. Given the prevalence of anxiety in schizophrenia, BNST alterations within schizophrenia are likely and may shed new light on the neurobiological mechanisms underlying emotion alterations in schizophrenia. The results from the proposed study can provide a foundation for future studies of emotion in schizophrenia, determine whether there are neurobiological differences in anxiety subgroups, and guide the development of novel neuroscientifically-informed treatments.