Prediction of transition to psychosis from an at-risk mental state using structural neuroimaging, genetic, and environmental data

Abstract

IntroductionPsychosis is usually preceded by a prodromal phase in which patients are clinically identified as being at in an “At Risk Mental State” (ARMS). A few studies have demonstrated the feasibility of predicting psychosis transition from an ARMS using structural magnetic resonance imaging (sMRI) data and machine learning (ML) methods. However, the reliability of these findings is unclear due to possible sampling bias. Moreover, the value of genetic and environmental data in predicting transition to psychosis from an ARMS is yet to be explored.MethodsIn this study we aimed to predict transition to psychosis from an ARMS using a combination of ML, sMRI, genome-wide genotypes, and environmental risk factors as predictors, in a sample drawn from a pool of 246 ARMS subjects (60 of whom later transitioned to psychosis). First, the modality-specific values in predicting transition to psychosis were evaluated using several: (a) feature types; (b) feature manipulation strategies; (c) ML algorithms; (d) cross-validation strategies, as well as sample balancing and bootstrapping. Subsequently, the modalities whose at least 60% of the classification models showed an balanced accuracy (BAC) statistically better than chance level were included in a multimodal classification model.Results and discussionResults showed that none of the modalities alone, i.e., neuroimaging, genetic or environmental data, could predict psychosis from an ARMS statistically better than chance and, as such, no multimodal classification model was trained/tested. These results suggest that the value of structural MRI data and genome-wide genotypes in predicting psychosis from an ARMS, which has been fostered by previous evidence, should be reconsidered.

Prognostic models predicting transition to psychotic disorder using blood-based biomarkers: a systematic review and critical appraisal

“…Two studies were developed models in participants of the Vienna omega-3 randomised-controlled trial [ 37 , 39 ]. There was one model developed in the Personalised Prognostic Tools for Early Psychosis Management (PRONIA) study [ 17 ], one model developed in a Korean cohort study [ 40 ], one model developed in participants recruited from the Outreach and Support in South London high-risk service [ 43 ], and one validated in the ICAAR (Influence of Cannabis in the emergence of psychopathological symptoms in Adolescents and Adults at-Risk) study [ 47 ].…”

“…Of the 16 development studies, nine internally validated model performance, of which five accounted for optimism in their performance estimates [ 15 , 17 , 41 – 43 ]. Four studies which used internal validation did not include the predictor selection process within the internal validation procedure.…”

“…Three studies which adjusted their performance estimates for optimism used polygenic risk scores (PRS) as a blood-based predictors [ 17 , 41 , 43 ]. Perkins et al [ 41 ] investigated adding PRS to an established cox regression model comprised of clinical predictors.…”

Prognostic models predicting transition to psychotic disorder using blood-based biomarkers: a systematic review and critical appraisal

“…Of the 17 included studies, 16 were prognostic model development studies [15][16][17][18][19][36][37][38][39][40][41][42][43][44][45][46] and one study was a prognostic model external validation study [47]. All studies were conducted in outpatients.…”

“…None of the included studies assessed model calibration, a key metric for which a model must perform acceptably if it is to be used to determine whether an intervention should be offered Three studies which adjusted their performance estimates for optimism used polygenic risk scores (PRS) as a blood-based predictors [17,41,43]. Perkins et al [41] investigated adding PRS to an established cox regression model comprised of clinical predictors.…”