Lana Ilankovic-Kambeitz1; 1Ludwig-Maximilian-University
The growing knowledge about enduring functional impairment and disability in psychosis and in high risk (CHR) patients encouraged us to search for prognostic biomarkers to individually predict functional and clinical outcome independent of a subsequent transition to full-blown psychosis or remission of symptoms after psychosis. We aim to elucidate structural and functional brain determinants and potential modifiers of functional outcome trajectories in early stages of psychosis. We have employed Multivariate Pattern Analysis (MVPA) in the PRONIA discovery sample to predict functional outcomes of recent onset psychosis (ROP) patients and CHR patients after 12 months of naturalistic treatment, based on the patients' structural neuroimaging measures. The Global Functioning: Social and Role scale (GF:S, GF:R) were used to determine good or poor outcome status at the follow up examination. Volume-based pattern classification predicted good vs. poor outcome status at follow-up in CHR patients with accuracy levels of up to 77% as determined by leave-site out cross-validation. The neuroanatomical prediction signature mapped to volume increments in cortical areas pertaining to the default-mode and central executive network whereas volume reductions were located in hub regions of the salience network and the temporal cortices. The detection of CHR and psychosis individuals with specific brain alterations associated with poorer outcome at follow-up may help identifying a critical group of at-risk persons, who irrespective of diagnostic thresholds require clinical treatment and therapeutic support. In the next step, we are planning to employ functional resting state (rs) analysis to predict functional outcome in the same group of patients.
