The quest to unlock precision psychiatry

• Proponents of precision psychiatry argue the approach can help — but how much is debated.

Background: The Diagnostic and Statistical Manual of Mental Disorders (DSM), first published by the American Psychiatric Association in 1952 with four subsequent updates, describes criteria for diagnosing psychiatric disorders, largely based on behavioral symptoms.

• The manual introduced a shared language to the field of psychiatry that continues to guide diagnosis and, just as importantly, insurance and billing codes, and also frames research questions.

But there's ongoing debate about how reliable the DSM's symptom criteria are for diagnosis and whether a diagnosis successfully predicts how someone will respond to treatment.

• A recent study from the nonprofit Sapien Labs used mental health symptoms self-reported by more than 100,000 adults to see whether the DSM meaningfully separates people by their symptoms.
• They found, for example, the symptoms of two people grouped as having autism spectrum disorder may be as different as those for a person in that group and one categorized as having ADHD.
• The fact that different symptoms patterns in different people can end up with different diagnoses is "a strength that recognizes the heterogeneity of the disorders we are talking about," says Paul Appelbaum, a professor of psychiatry at Columbia University who chairs the DSM steering committee.
• But,"recruiting people into clinical trials based on more homogeneous symptom profiles rather than the heterogeneous profiles of the DSM can help the search for biological markers of disorders," says Tara Thiagarajan, a neuroscientist and founder of Sapien Labs who is an author of the paper.

The National Institute of Mental Health invested in trying to find biomarkers that former director Tom Insel in 2017 said, in an often cited quote, didn't move the needle on outcomes for people with mental disorders.

• One issue, says Insel, who is an adviser to Alto Neuroscience, is that some biomarkers would be dismissed if they were found in only a fraction of people characterized by DSM criteria for disorders.
• "It's better to follow the biomarker data and create clusters based on that data than assume symptoms will map onto the biology," he says.

What's happening: In recent years, researchers have found some biological markers of schizophrenia, major depressive disorder and PTSD.

• "But none of them have yet reached the point where anyone has been able to demonstrate their value for psychiatric diagnosis per se," Appelbaum says.

• The DSM-5, published in 2013, explicitly says it hopes reliable and precise biomarkers can be incorporated into diagnostic criteria, he adds.

Aside from diagnosis, some biomarkers are now being assessed to try to predict better treatment outcomes for people, including by two startups that emerged in October —Neumora Therapeutics and Alto Neuroscience.

• They are using a variety of biomarkers — including imaging of the brain's networks, genetics, EEG brain wave measurements and behavioral data from tests or wearable devices — to try to determine who might have a better outcome from a drug or other treatment.
• The approach is nascent, but Amit Etkin, Alto's founder and CEO, hopes that over the next 10 years psychiatry will make the kind of progress oncology has by using genomics to better understand cancer mutations and harnessing that information to match therapies to individuals.
• "Psychiatry is stuck in the '90s and with the myth of the silver bullet" of a single drug or treatment that works for everyone determined to have a disorder, Etkin says.

The big picture: Critics say precision psychiatry puts too much emphasis on the biological underpinnings of disease and not enough on the experiences and environment that shape it.

• In his forthcoming book, "Healing: Our Path from Mental Illness to Mental Health," Insel argues "the problems are indeed medical but the solutions are social, relational, environmental, and political."
• "If the goal is to improve outcomes and increase the chances someone will recover, we need to know much more than their symptoms and biology," he says.