Biomarkers for Phenotypic Classification of Patients with Acute Psychiatric Symptoms: The Relevance of White Matter Hyperintensities

The investigation of biomarkers capable of identifying subgroups of psychiatric patients based on a combination of clinical and neurobiological features, such as white matter hyperintensities (WMH), will enable more accurate medical practice, with a better prediction of clinical presentation, progression, and response to treatment. The aim of this study was to investigate whether patients with acute psychiatric symptoms could be grouped according to clinical and neurobiological features. A k-means cluster analysis of 90 acute psychiatric inpatients, aged 45 to 75 years, who met criteria for schizophrenia, bipolar disorder, and major depressive disorder was performed. Multidimensional clinical and neuroimagological data were included in the analysis. Validity of the clusters was tested by re-clustering the sample and comparing the silhouette coefficient. The cluster analysis extracted three clinical phenotypes that can be identified based on the burden of WMH. A “vascular-cognitive” phenotype (cluster 1, n=13), a “low vascular-psychiatric” phenotype (cluster 2, n=58), and a “vascular-depressive/apathy” phenotype (cluster 3, n=19). The silhouette coefficient was higher in the clusters including the WMH variable, reflecting a structuring effect on the clusters. The cluster analysis extracted subgroups of subjects that can be distinguished by WMH burden, cognitive performance, and psychiatric symptoms.