Biomarkers Outperform Symptoms in Parsing Psychosis Subgroups
Three biomarker-based categories, called biotypes, outperformed traditional diagnoses, such as schizophrenia and bipolar disorder with psychosis, in sorting psychosis cases into distinct subgroups on the basis of brain biology…
“Just as fever or infection can have many different causes, multiple psychosis-causing disease processes – operating via different biological pathways – can lead to similar symptoms, confounding the search for better care.”
The results lend support to the institute’s Research Diagnostic Criteria (RDoC) initiative, which frees scientists from designing research based on traditional diagnostic categories, encouraging them to explore groupings based on genomics, behavioral dimensions, physiological traits, or brain imaging findings. More precise diagnosis is expected to lead to improved treatments.
They add to increasing evidence of biological overlap between traditional, symptom-based diagnostic categories for disorders in which patients experience psychotic symptoms. The BSNIP study used sophisticated statistical approaches to sort such cases into homogeneous subgroups, based on multiple neurobiological measures and levels of analysis.
Three distinct psychosis-related biotypes emerged that cut across clinical diagnosis boundaries. These subsets of biomarkers differentiated groupings of psychosis cases from each other considerably better than did traditional clinical diagnoses. External measures – social functioning, brain structure, and rates of psychosis-related illness and biomarker patterns in patients’ first-degree relatives – also validated the neurobiological distinctiveness of the biotype subgroups more than they did the distinctiveness of symptom-based categories.
People with different biotypes differed in their mix of psychosis-related impairments.
Biotype 1 showed the most severe impairment in a set of brain functions that the researchers distilled into a construct they call “cognitive control” – the ability to flexibly exert control over attention and information processing to meet one’s goals…59 percent had a schizophrenia diagnosis, 20 percent had bipolar disorder with psychosis.
Biotype 1 cases were also the most socially impaired.
Biotype 2 cases showed intermediate levels of impaired cognitive control, but had normal to accentuated brain responses to sensory inputs and fast visual orienting, a set of brain functions called “sensorimotor reactivity” – the ability to detect and process sensory stimuli…44 percent had a bipolar disorder with psychosis diagnosis.
Biotype 3 showed normal cognitive control, modestly impaired sensorimotor reactivity, were the least socially impaired and had the lowest positive (e.g., hallucinations and delusions) and negative (e.g., blunted emotion) symptoms…32 percent had schizophrenia
Biotype 1 cases – and to a lesser degree Biotype 2 cases – showed reduced gray matter, the brain’s working tissue, across several areas of the cortex, or outer mantle, known to process higher-order information. By contrast, in Biotype 3 cases, the largest of the groups, such reduced gray matter was mostly localized in emotion-processing areas in deeper brain regions. No brain structural differences distinguished the traditional categories of schizophrenia and schizoaffective disorder from each other.
Unaffected first-degree relatives of psychosis cases tended to show similar, but reduced, abnormalities in cognitive control, sensorimotor reactivity, social functioning and brain structure, suggesting that they may harbor “constitutional” risk factors for the biotypes…So even if patients have the same nominal diagnosis, they might tend to have different underlying illness processes.
“The biotypes outcome provides proof-of-concept that structural and functional brain biomarker measures can sort individuals with psychosis into groups that are neurobiologically distinctive and appear biologically meaningful,”