Biomarkers for Psychosis Risk Prediction

Background: Although neurodevelopmental processes figure prominently in the etiology of psychosis, onset of overt symptoms typically does not occur until late adolescence or early adulthood. It is particularly critical to identify early (pre-onset) biomarkers of psychotic illness, given that overt diagnosis may be a late manifestation of the underlying disease processes. As such, there is intense interest in enhancing methods for early detection, which may ultimately lead to prevention strategies. 

 Objectives: Here I will present new findings regarding neurobiological risk factors predictive of conversion to psychosis, and their implications for potential pathophysiologic mechanisms. Potential convergence with biomarkers relevant to autism spectrum disorders (ASD) will be highlighted.

 Methods: Longitudinal studies of putative neuroimaging and blood-based biomarkers were conducted in both clinically and genetically defined at-risk cohorts.  The North American Prodrome Longitudinal Study (NAPLS) is a prospective longitudinal investigation of ~275 youth at clinical high risk (CHR) for psychosis, defined by the presence of moderate to severe attenuated positive symptoms. Neuroimaging and blood assays were conducted over a 2-year follow-up period, during which clinical outcomes were monitored. Gene expression and neuroimaging findings indexing psychosis risk in a highly penetrant genetic subtype of psychosis (i.e., patients with 22q11.2 microdeletions; 22q11DS) will also be presented.

 Results: Findings indicate that CHR subjects who converted to psychosis showed a steeper rate of gray matter loss in prefrontal cortical regions, as well as greater ventricular expansion, relative to non-converting CHR subjects and healthy controls. Differential tissue loss was predicted by an aggregate measure of pro-inflammatory cytokines in plasma.  Parallel structural neuroanatomic findings in 22q11DS patients indicated that cortical thickness in the medial prefrontal cortex uniquely predicted psychotic symptoms. Further, gene expression studies in this cohort indicated highly significant overlap between a module of co-expressed genes associated with psychosis in 22q11DS and transcriptional changes in an idiopathic schizophrenia cohort.

Conclusions:   Structural neuroanatomic changes precede the onset of overt psychotic symptoms; convergent findings from clinical and genetic high risk cohorts implicate prefrontal cortical changes in symptom onset. Further, findings implicate neuroinflammatory processes in psychosis etiology, possibly by activating synaptic pruning processes, leading to observed gray matter tissue loss. Given the long-hypothesized role of neuroinflammatory factors in the etiology of autism, these findings suggest a possible point of etiologic overlap.  Future work to determine temporal precedence of peripheral blood and brain biomarkers of psychosis is warranted. Relevance to risk prediction in ASD is highlighted throughout this presentation.