Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

ICOSR 2017: Technology Drives New Directions in Schizophrenia Imaging

30 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 Giulio Pergola of the University of Bari, Aldo Moro, in Italy.


Ever since postmortem studies identified brain changes in patients with schizophrenia, the idea that brain neuropathology may be a distinctive feature of the disorder, and perhaps may be even used to differentiate schizophrenia patients from healthy controls and subgroups of patients from each other, has fascinated many investigators. With multiple sessions on brain imaging, ICOSR 2017 has shown that neuroimaging studies in the field of schizophrenia research keep increasing. The technological advancements and the growing number of computational resources now allow researchers to combine multiple datasets and delve deeper not only into brain structure but also into its function.

Some of the presenters adopted a classic “biomarker” approach, striving to identify neuroimaging features that differentiate schizophrenia patients from healthy controls―especially those associated with social cognition and emotions. For example, Amy Jimenez from the University of California, Los Angeles, studied memory retrieval in “self” and “other” conditions involving verbal stimuli. The “self” condition yielded functional magnetic resonance imaging activity patterns that differed among groups, and the authors interpreted the finding as possibly reflecting a different balance of inward/outward attention. Further detailing affective memory retrieval in schizophrenia, Lauri Tuominen of Massachusetts General Hospital in Boston investigated brain response in schizophrenia patients and healthy controls using morphed faces. Interestingly, the researchers adjusted their morphing step, in which neutral faces were morphed with an emotional face (fear) to match discrimination accuracy across participants at 75 percent. In their fear conditioning paradigm, they found altered activation patterns in schizophrenia patients relative to controls, particularly in the anterior insula and the dorsal anterior cingulate cortex, and the deficit appeared related to impaired generalization in the clinical population.

Other studies capitalized on the temporal resolution of electro- and magnetoencephalography to derive temporally accurate proxies of brain processing. Abbie Popa from the University of California, Davis, studied the N2 component of event-related potentials in adolescents with 22q11.2 deletion syndrome (22q), a genetic condition strongly associated with schizophrenia risk and possibly with some of its neural mechanisms. Using a Go-NoGo task, which included a standard and an emotional variant, they found that adolescents with 22q showed different amplitudes of N2 compared to controls in both Go-NoGo conditions. While previous studies had investigated N2 in schizophrenia, this study suggests that altered N2 amplitudes may be an outcome of genetically altered neurodevelopment. Consistently, Caitlyn Kruiper from the Center for Neuropsychiatric Schizophrenia Research at the University of Copenhagen found that P3 differences between schizophrenia patients and controls during selective attention performance are present at onset in untreated first-episode patients. Electrophysiology seems to yield even more general differences between schizophrenia patients and healthy individuals, as highlighted by Petra Rupert of the University of Pennsylvania. She and her co-authors investigated whether candidate electrophysiological biomarkers of schizophrenia extend to clinical risk subjects (CR). They found that occipital alpha power collected with eyes closed differentiated both schizophrenia patients and CR from healthy controls, and were associated with early cognitive impairment.

In keeping with the focus on electrophysiology, Seung Suk Kang of the University of Missouri presented thought-provoking findings on the claustrum, a brain region abnormal in psychosis, famously known for Francis Crick's hypothesis about its role in consciousness and, more recently, for the discovery of giant neurons linking the claustrum with the entire brain in mice (Reardon, 2017). Kang and his co-author Scott Sponheim manually segmented the claustrum in schizophrenia patients and healthy controls using T1 MR images and used boundary element method models of the cerebral cortex to study the brain network supporting a visual perception task. They found abnormal claustral-visual cortex connectivity during the visual perception task, with the intriguing finding that schizophrenia patients with prominent visual hallucinations presented an even larger discrepancy from healthy controls than patients with auditory hallucinations.

Despite the promising findings obtained in relatively small samples, Molly Erickson of Rutgers University delivered a more sobering account of electrophysiological alterations in schizophrenia. She and her group examined mismatch negativity in association with symptom severity in a very large meta-analysis and reported a negative result. In line with the idea that symptom severity may be associated with brain function, Matilda Azis of King’s College London aimed to understand the heterogeneity in ultra-high-risk (UHR) individuals using arterial spin labeling (ASL), a technique to study cerebral blood flow. This is especially relevant because only some of the UHR patients will transition to psychosis, and ASL could be a predictor of the transitioning. In the largest multisite study of UHR to date, the authors found that symptoms, combined in five factors through a dimension reduction step, were associated with cerebral blood flow in the thalamus and hippocampus bilaterally (disorganized factor symptoms) and in the bilateral prefrontal and cingulate cortices (anxiety symptoms).

Can we apply what we learn from neuroimaging to the pharmacological treatment of schizophrenia, especially regarding the negative symptoms, which are the most difficult to treat? Stephan Taylor of the University of Michigan reported that we are not there yet. His group associated GABA availability, investigated via magnetic resonance spectroscopy, with negative affect, based on the link of both GABA and negative affect with benzodiazepine administration in chronic schizophrenia. However, they found no group differences, warranting more detailed investigations aimed at clinical translation.

The session closed on a positive note with the report by David White of the University of Alabama at Birmingham that pretreatment glutamate levels and resting-state connectivity in the hippocampus (investigated via magnetic resonance spectroscopy and functional magnetic resonance imaging, respectively) are associated with treatment response monitored after six weeks.

In summary, functional brain imaging studies at ICOSR have shown two emerging trends: first, the understanding of neuroimaging features as dimensions to inform diagnosis, hence as a step toward dimensional diagnoses that take into account both symptom severity and neuroimaging patterns, and second, the increasingly widespread effort to differentiate patients from one another at an early stage, particularly regarding clinical outcome prediction. In line with other fields in schizophrenia research, it is a dire necessity to employ larger sample sizes from multiple sites to demonstrate the generalizability of the findings.