Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

ICOSR 2017: A Functional Imaging Potpourri

11 Apr 2017

As part of our ongoing coverage of the 2017 International Congress on Schizophrenia Research (ICOSR), held March 25-28 in San Diego, we bring you session summaries from some of the participants in the Young Investigator program. We are, as always, grateful for the gracious assistance of YI program directors Laura Rowland and Scott Sponheim, as well as Michelle Tidwell of the ICOSR staff. For this report, we thank Sarah Haigh of the University of Pittsburgh Medical Center.


The Sunday, March 26, a functional brain imaging oral session at the International Congress on Schizophrenia Research focused on finding brain-related markers that reflect pathology in schizophrenia. Some of the talks framed the use of these markers to help classify who is at risk of developing schizophrenia, whereas other sought markers of altered mechanisms in the disorder.

The first talk, by Stephanie Hare of Georgia State University, focused on the relationship between symptoms in schizophrenia and their functional connectivity as measured in resting-state data. She split the schizophrenia group into those who had auditory hallucinations and those who had visual hallucinations. There were no significant differences between auditory and visual hallucinators in their functional connectivity, and so they were combined to compare against non-hallucinators and healthy controls. Both hallucinators and non-hallucinators showed elevated connectivity between superior temporal gyrus and hippocampus as measured with multiple regression analyses. Connectivity was primarily based in low-frequency bands (alpha and theta binding), suggesting altered functional connectivity in schizophrenia that is not specific to symptoms.

The second talk, by Susanna Fryer of the University of California, San Francisco, looked at error processing in individuals who are at risk of developing schizophrenia, those early in the disease course (within two years of diagnosis), and healthy controls. Individuals' ability to assess errors in their behavior helps with monitoring goal-directed actions and could be a key mechanism underlying the behavioral abnormalities. Using a go-no-go task, she found no significant effects of the group in fMRI responses in the anterior cingulate cortex or the dorsolateral prefrontal cortex―two areas strongly associated with error detection. However, she did find significant differences in activation in the left cerebellum (participants were right-handed), suggesting a motor-related slowing. There were no group differences in response accuracy, but patient groups did show slower and greater variability in their reaction times. Behavioral responses correlated with cerebellum responses in controls only. Early-course individuals were more impaired compared to at-risk individuals, highlighting the need to use protocols that are sensitive enough to detect even subtle abnormalities in behavioral and cortical responses before first break.

In a similar vein, Jamie Ferri, also of UCSF, replicated the findings that individuals with chronic schizophrenia demonstrate hyperconnectivity between sensory systems and thalamus, and hyperconnectivity among cerebellum, anterior cingulate, posterior cingulate, and thalamus. Early disease course and clinical high-risk individuals showed similar patterns of connectivity to those found in chronic patients. Clinical high-risk and early-course only differed from each other in the hyperconnectivity between sensory and thalamic regions, suggesting that thalamus and cerebellar connectivity deteriorates first, followed by sensory connections. This demonstrates widespread connectivity changes that are evident before first break and become worse with the duration of the disease.

Sameer Jauhar of King's College London then switched things up by focusing on changes in dopamine synthesis (measured using PET) before and after prescribing antipsychotics to individuals at their first episode of schizophrenia. He found that those who responded to medication (had lower positive symptoms with medication) showed significant improvements in dopamine but not glutamate synthesis. Together, this suggests that dopamine synthesis could be a predictor of response to antipsychotic medications.

The next speaker, Jennifer Forsyth of the University of California, Los Angeles, took a more mechanistic approach to investigate abnormal working memory in schizophrenia. She used D-cycloserine (DCS) to augment NMDA receptor activity in individuals with chronic schizophrenia during a zero-, one-, and two-back task while measuring gamma power. By subtracting responses from the zero-back condition, she was able to control for confounding effects of motor and muscle activity in gamma power. She found significantly better working memory performance in those who took DCS compared to those who took placebo, which was associated with reduced gamma power in frontal channels. These findings demonstrate that augmenting NMDA receptor activity in schizophrenia normalized cognitive abilities.

Eric Reavis, also of UCLA, used machine learning to ascertain potential differences in visual object recognition in schizophrenia, bipolar, and control participants. He measured five runs of fMRI responses to two example pictures of cups, chairs, and an outdoor scene. He used the first four runs to run multivariate pattern analysis (MVPA) to recognize the patterns of fMRI activity associated with each object category, and then tested the program to categorize the fMRI activity in the last run. Primarily, he focused on the lateral occipital complex, which is associated with object recognition. The program achieved good accuracy (over 40 percent for each participant group) but found no significant differences in MVPA accuracy among the three groups. He then used a searchlight analysis to find other parts of the brain that had reliable patterns of object categorization. Most of the visual cortex was identified, but there were still no group differences in classification accuracy. This suggests that there are no large-scale differences in object categorization networks in bipolar disorder or schizophrenia.

The next speaker, Deepak Sarpal of the University of Pittsburgh, investigated the possible mechanisms behind the association between longer duration of untreated psychosis (DUP) and poor treatment outcome. He measured the effects of risperidone and aripiprazole, or risperidone and omega-3 fatty acids, on functional connectivity in the striatum. Different subsections of the striatum correlated with DUP. He used principal component analysis to ascertain the main factor for the effect of DUP on functional connectivity and found that functional connectivity mediated the effect of DUP on treatment outcome. This suggests fundamental changes in neural responses associated with longer untreated psychosis.

Holly Hamilton of the San Francisco VA Healthcare System then brought the topic back to potential markers for abnormal neural functioning before a first psychotic episode. Individuals who are at clinical high risk for developing psychosis show smaller P3a and P3b responses—the two components of the P300 EEG response—when attending to oddball sounds. P3b was significantly smaller in converters compared to non-converters. Out of the non-converters, some will remain symptomatic, whereas others will go into remission. Those who go into remission have similar P3b to controls, suggesting that P3b can predict who will convert. This highlights the potential power of using P3 measures as a biomarker for susceptibility to psychosis.

Monte Buchsbaum of the University of California, San Diego, assessed dopaminergic binding in prefrontal cortex in chronic schizophrenia using PET imaging with fallypride and correlated binding with performance on the Wisconsin Card Sorting Task and the California Verbal Learning Task. All participants have never been on long-term antipsychotic medication. There was a significant positive correlation between fallypride binding and behavioral performance in healthy controls and a negative correlation in schizophrenia. This suggests that the natural course of schizophrenia leads to dopaminergic deficits in prefrontal areas that impact cognition.

Finally, Sarah Clark of Georgia State University investigated the neural correlates of self-effectiveness (belief in one's ability to succeed or to accomplish a task) and self-certainty as measures of insight in those at ultra-high risk for developing schizophrenia. People with schizophrenia often demonstrate low self-effectiveness and high self-certainty associated with lack of insight. She measured fMRI connectivity and found negative correlations between cerebellum and prefrontal areas and self-certainty, suggesting less flexible thinking, which is associated with thought disorder in schizophrenia.

Overall, this was a fascinating session summarizing the recent work on functional markers for identification or abnormal processing at various stages of schizophrenia. The emphasis on measuring these markers in those at risk of developing schizophrenia demonstrates the field's move toward identifying and treating these individuals before the onset of psychosis.