15 Apr 2017
by Lesley McCollum
Neuroinflammation is a hot topic in schizophrenia research, but in a Monday, March 27, morning symposium at the International Congress on Schizophrenia Research meeting in San Diego, scientists called into question the use of positron emission tomography imaging with the radiotracer targeting translocator protein 18 kDa (TSPO). It is considered a standard in the field as a biomarker of inflammation, but recent findings challenge some widely held assumptions about what TSPO measures.
TSPO expression has often been equated with microglial activity, and inflammation in schizophrenia is thought to be reflected in increased TSPO levels. However, Tina Notter of the University of Zurich in Switzerland presented research published earlier this year revealing decreased TSPO expression in first-episode patients (Notter et al., 2017). She also reported that TSPO changes were present not only in microglial cells but also in astrocytes and endothelial cells in an immune activation mouse model. These findings have a game-changing implication—that TSPO may not be a specific microglial marker—and raise the question of what TSPO expression changes in schizophrenia actually represent.
Session chair Romina Mizrahi of the Center for Addiction and Mental Health in Canada further rattled the assumption that increased TSPO indicates increased inflammation in schizophrenia. She presented new findings in a clinical high-risk population, in addition to published findings in untreated first-episode patients (Hafizi et al., 2016) and treated schizophrenia patients experiencing psychosis (Kenk et al., 2015): All three groups revealed no significant difference in TSPO compared with controls, suggesting no increase in inflammation in the disorder.
Mizrahi’s findings also suggested a trend for decreased TSPO expression in first-episode and clinical high-risk patients, in agreement with the findings presented by Notter. In the question portion after the talk, co-chair Urs Meyer of the University of Zurich reminded the group that although increased inflammation is assumed to be a bad thing in schizophrenia, it could also be that not enough activation has detrimental effects. Meyer noted that a reduction in TSPO challenges the predominant idea in the field that TSPO is a marker of neuroinflammation.
An important caveat to PET TSPO studies was presented by Tiago Reis Marques of Imperial College London, who reported findings from a new meta-analysis of nine PET studies on TSPO in schizophrenia. In the meta-analysis, different outcome measures told different stories—measures of relative TSPO binding potential (the ratio of the binding potential in a target region to that in the brain overall) revealed an increase in schizophrenia, whereas the more traditional measure of volume of distribution (the ratio of the radioligand concentration in a target region to that in plasma) revealed no differences.
Akira Sawa of Johns Hopkins University, discussant for the symposium, summed up the sobering message of the morning that we really have yet to understand what TSPO is. Although neurodegenerative diseases such as Alzheimer’s disease reliably show increases in TSPO, the radiotracer may be inadequate as a marker in conditions of mild inflammation. Researchers should be much more cautious when interpreting what differences in TSPO mean, Sawa said, also referring to its multiple roles in the brain and the variation in genotypes that affect TSPO binding from person to person. Ultimately, concluded Meyer, the discussion comes down to determining whether the methods we use are meaningful for measuring inflammation in schizophrenia and if they can be used for informing us about individual patients.