Each year, about 100,000 Americans experience psychosis, a serious condition that disrupts thoughts and perceptions so profoundly that it can distort a person’s sense of reality. Now—just over a year after the first new schizophrenia drug in half a century was approved—a study in Nature Mental Health looks at how patients respond to it, offering early clues for more personalized treatment.

The study, led by Michael Halassa, professor of neuroscience at Tufts University School of Medicine, analyzed electronic medical records from 49 patients hospitalized for schizophrenia, schizoaffective disorder, or bipolar disorder with psychotic features. All patients were given the new drug, Cobenfy—a combination of xanomeline and trospium chloride—on top of their usual antipsychotic medications after standard treatments failed to adequately control their symptoms.

“The new drug targets different receptors in the nervous system than traditional antipsychotic drugs, which mainly block dopamine D2 receptors in the brain,” says Halassa. “Clinical trials showed the drug worked well compared with a placebo. But clinicians are still assessing how it performs as part of real-world care.”

Through statistical analyses of two small patient cohorts, Halassa’s exploratory study identified patterns that may help predict who may benefit from the xanomeline-trospium combination therapy and who might not. 

The study analyzed clinical data for 24 patients, identifying several predictive clinical features for individuals who responded, did not respond, or had mixed responses to the addition of the new drug. “A statistical analysis of clinical data from a second group of 25 patients independently replicated these findings, supporting the notion of biologically distinct psychosis subgroups,” says Halassa.

Patients with prominent “negative symptoms”—such as social withdrawal, low motivation, or reduced speech—showed the strongest improvements after receiving the new combination drug with their usual antipsychotic medications, including brighter mood and greater social engagement. Those with a history of stimulant use also tended to respond especially well to the new drug.

In contrast, patients who exhibited aggression or bipolar features (mainly manic symptoms) saw little benefit from adding the new drug to their treatment. Patients with intellectual disabilities also showed limited improvement, though Halassa notes this finding is tentative given the small number of patients with that diagnosis in the study.

The drug’s effects on other symptoms varied. Some patients with hallucinations improved, but not as consistently or dramatically as those with negative symptoms.

Halassa says the results suggest schizophrenia may be less a single disease than a collection of conditions—much like a fever or pain—that can stem from different causes and require different treatments. He hopes the findings mark an early step toward the development of precision psychiatry, in which treatment-response patterns help guide care as they already do in other areas of medicine, such as cancer and immunology. 

To shorten the long and uncertain path to recovery for patients with psychosis, he says researchers need to test whether these emerging patient subgroups can truly predict who responds to which treatments. That means running clinical trials that compare medications in people with specific cognitive or biological profiles and tracking the precise trajectories of these symptoms over time.

Clinicians will also need to track not just whether symptoms improve, but which ones—and under what kind of drug—to drive more personalized treatments. “If we start treating every symptom response and non-response as valuable data, we could save individuals and families years of trial and error in finding effective treatment,” Halassa says.