Psychiatry’s drug problem: why mental health care lacks the biological tests that transformed cancer treatment

For millions of people living with depression, schizophrenia, bipolar disorder or anxiety, the path to an effective medication is rarely a straight line. It is instead a sequence of educated guesses: a drug is prescribed, the patient waits weeks to see whether symptoms improve, and if nothing changes the process begins again with a different compound. This trial-and-error approach stands in stark contrast to fields such as oncology, where a biopsy and a molecular profile can determine, often within days, which therapy is most likely to work for a specific tumor. A new consensus position paper from the American College of Neuropsychopharmacology, co-led by a UCLA Health psychiatrist, argues that this gap is not an inevitable feature of mental illness but a solvable problem in how psychiatric drugs are developed and tested. The paper, published in the journal NPP – Digital Psychiatry and Neuroscience, lays out a detailed roadmap for bringing biomarkers into the core of psychiatric drug development.

The document emerged from the ACNP Precompetitive Stakeholder Task Force, a body deliberately constructed to span the boundaries that usually separate the players in psychiatric medicine. Its authors include academic researchers from institutions such as UCLA Health, representatives of the pharmaceutical industry, scientists from the National Institute of Mental Health, and members of the US Food and Drug Administration’s Office of Neuroscience. That breadth matters, the authors suggest, because the barriers to precision psychiatry are not purely scientific. They are also structural, involving how trials are designed, how data are shared or hoarded, and how regulators and payers evaluate evidence. A roadmap that only addressed one of those dimensions, they argue, would leave the others intact and progress stalled.

The central scientific problem the paper identifies is heterogeneity. Two people can receive the same diagnosis of major depression while having very different underlying biology driving their symptoms. One patient’s low mood may be tied to inflammatory processes, another’s to disrupted neural circuitry, and another’s to genetic vulnerabilities that shape how drugs are metabolized. Yet current psychiatric drug trials typically group participants by symptoms alone, enrolling everyone who meets diagnostic criteria into a single arm. When a medication that works well for one biological subgroup is tested in a mixed population, its effect can be diluted to the point of statistical invisibility. The drug may then be abandoned, not because it fails patients, but because the trial failed to find the patients it would have helped.

This is precisely the problem that biomarkers are designed to solve. Biomarkers are measurable biological signals, and in the psychiatric context they can take many forms: molecules detected in blood, patterns of electrical activity recorded by electroencephalography, genetic variants, or behavioral and physiological data streamed by wearable devices. By using such signals to stratify trial participants into biologically coherent subgroups, researchers can test whether a drug benefits the people whose biology matches its mechanism. The same tools, once validated, could eventually help clinicians match individual patients to treatments from the start, replacing months of sequential experimentation with an informed first choice.

The precedent from other areas of medicine is the paper’s strongest rhetorical weapon. In cancer care, molecular diagnostics now routinely determine whether a patient receives a targeted therapy, an immunotherapy, or a conventional chemotherapeutic regimen. Cardiac medicine uses troponin levels and imaging markers to guide acute decisions. Neurology has moved in the same direction, with fluid and imaging biomarkers of Alzheimer’s disease and Parkinson’s disease increasingly anchoring both diagnosis and trial enrollment. The authors argue that psychiatry can follow a comparable path, not by copying these fields wholesale, but by adapting their logic of measurable, decision-relevant biology to the specific challenges of brain and behavior.

One of the paper’s most consequential points concerns regulation, and specifically a misunderstanding the authors believe is slowing the field down. Many researchers and companies assume that using a biomarker in a drug trial requires lengthy FDA approval of that biomarker first. The authors state that this is not the case: biomarkers can already be incorporated into trial design without a protracted regulatory approval process. Clearing up this misconception matters because it removes what may be a significant, and unnecessary, deterrent. Companies that have been waiting for a regulatory green light that was never actually required can begin integrating biomarker strategies into their development programs now.

The roadmap also draws a careful boundary around what it does and does not attempt to accomplish. Its focus is precompetitive collaboration in drug development, meaning the shared scientific infrastructure that companies and academics can build together before products compete in the marketplace. Bringing validated biomarker tests into everyday clinical care is described as a separate and later step, with its own regulatory, economic and logistical hurdles. By scoping the effort this way, the authors aim to make the near-term goals concrete and achievable, rather than promising an immediate transformation of psychiatric practice that the evidence base cannot yet support.

The specific recommendations are pragmatic. The task force calls for the field to agree on common definitions and clear intended uses for psychiatric biomarkers, so that a term like treatment-response predictor means the same thing across laboratories and companies. It proposes small, focused studies of blood tests, genetics, EEG recordings and wearable device data that can reveal early signs of whether a treatment is working, before the full clinical picture emerges weeks later. Promising biomarkers should be tested against existing large datasets before expensive new trials are launched, extracting maximum value from data already collected. The paper urges a precompetitive framework for data sharing, explicitly including negative results that did not pan out, so that the whole field avoids repeating failed experiments. It further calls for standardized collection and analysis of biomarker data across research sites, early consultation with the FDA and European regulators, including on newer approaches that combine multiple data types, and a concerted effort to show industry and payers that biomarker-guided trials can lower costs while improving outcomes.

The human cost of the status quo animates the paper’s argument. Dr. Sahib Khalsa, the paper’s first author, a psychiatrist and director of Anxiety Disorders Research at the UCLA Semel Institute for Neuroscience and Human Behavior, framed the problem in terms of what patients currently endure. Right now, patients rely on repeated trials to learn whether a medication is working for them, he said, a process that is costly and time-consuming. The goal of the effort, he explained, is to give clinicians, researchers, drug developers and drug regulators better ways to collaborate and to see the biology behind a person’s symptoms, so that treatments can be matched more accurately from the start. In other words, the ambition is not merely faster drug approval but a fundamentally different experience of psychiatric care, in which the first prescription is informed by evidence rather than precedent.

The authors close with a sober assessment of what success will require. Precision psychiatry will not emerge from isolated breakthroughs, they conclude, but from sustained collaboration around shared standards and shared data. That framing is a deliberate challenge to a research culture in which laboratories and companies often guard their data and pursue proprietary paths. The task force’s composition, spanning academia, industry, the National Institute of Mental Health and the FDA, suggests that the stakeholders most able to change that culture have at least agreed to try. Whether the roadmap converts into validated tests that reach patients will depend on funding, on companies willing to share precompetitive data, and on regulators and payers recognizing biomarker-guided evidence as trustworthy. But the paper makes a clear claim: the tools to end trial-and-error prescribing in psychiatry are within reach, and the field’s failure to use them is a choice that can now be reversed.

Subject of Research: Biomarker-guided precision medicine in psychiatric drug development

Article Title: Cancer care uses biological tests to guide treatment. Why doesn't psychiatry?

Article References: Cancer care uses biological tests to guide treatment. Why doesn't psychiatry?. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: psychiatry, biomarkers, precision medicine, drug development, clinical trials, depression, schizophrenia, ACNP, UCLA Health, FDA, EEG, wearable devices