Forty years of structural imaging in psychosis: promises and truth

Objective: Since the first study published in the Lancet in 1976, structural neuroimaging has been used in psychosis with the promise of imminent clinical utility. The actual impact of structural neuroimaging in psychosis is still unclear. Method: We present here a critical review of studies involving structural magnetic resonance imaging techniques in patients with psychosis published between 1976 and 2015 in selected journals of relevance for the field. For each study, we extracted summary descriptive variables. Additionally, we qualitatively described the main structural findings of each article in summary notes and we employed a biomarker rating system based on quality of evidence (scored 1–4) and effect size (scored 1–4). Results: Eighty studies meeting the inclusion criteria were retrieved. The number of studies increased over time, reflecting an increased structural imaging research in psychosis. However, quality of evidence was generally impaired by small samples and unclear biomarker definitions. In particular, there was little attempt of replication of previous findings. The effect sizes ranged from small to modest. No diagnostic or prognostic biomarker for clinical use was identified. Conclusions: Structural neuroimaging in psychosis research has not yet delivered on the clinical applications that were envisioned.