Making an IMPACT on Precision Medicine in Psychiatry

Cheryl* was like many of my patients, struggling to juggle the competing demands of her personal life and her life as a busy professional while simultaneously struggling with severe depression that left her with crippling exhaustion and hopelessness. I listened as she told her story, which included her husband who was battling cancer, her successes and challenges in the workplace, her complex history of experiencing both abuse and support as a child, and an anger-filled—if amicably ended—first marriage.

As I ended our intake session with a set of standard questions about risk factors and symptoms, I felt like I understood her, at least as well as a trained active listener can after an hour-and-a-half appointment. It was then her turn to ask me a question: “Will I ever get better?” “Yes,” I assured her, “nearly everyone with depression responds to one of the many treatments we can offer.” “But which one will work for me?” she asked. That was always a harder question to answer.

Cheryl was asking for a precision medicine approach to treating her mental illness. Practicing precision medicine  means using data about a patient to make precise treatment recommendations that apply to that patient’s particular circumstance. For much of medicine, including psychiatry, precision medicine is an aspirational goal of the future. But for some illnesses—like some forms of cancer—precision medicine is already here. Many women with breast cancer will receive a specific treatment regimen based on their type of cancer's specific molecular signature. Although a precision medicine approach is available today for cancer treatment, we just don’t have one for treating mental illnesses. Yet.

NIMH hopes to change the game for precision medicine in psychiatry with a groundbreaking new initiative. NIMH’s Individually Measured Phenotypes to Advance Computational Translation in Mental Health (IMPACT-MH) initiative will support research that tests new ways of adding data, such as performance on computerized behavioral tasks or information about activity levels, to traditional clinical information in ways that could help mental health providers and their patients make informed decisions about the future. Studies supported through the IMPACT-MH initiative will test whether combining these different kinds of data improves predictions about mental health treatment responses and outcomes, enabling a vision for precision psychiatry of the future.

IMPACT-MH will focus first on novel behavioral measures that might help us learn more about what makes individual patients like Cheryl unique—setting the stage for new, personalized treatments. Why start with behavior? Because properly designed behavioral tasks can get at specific brain-based functions in a way that is inexpensive and easy to do “at scale”—meaning these behavioral tests can be carried out on large numbers of people cost-effectively. Behavioral tests enable research that reaches more—and more diverse—groups of people, because the tests can be remotely delivered (with smartphones or other internet-linked devices) and easily repeated (to detect differences among people and changes within an person over time). This reach will help ensure that the results can improve mental health care as well as enhance mental health equity for all.

NIMH has already taken the first steps to implement IMPACT-MH by publishing two funding opportunities that will launch the initiative. The first funding opportunity  will support U01 projects that will develop these behavioral measures, distribute them to large cohorts of research participants, and integrate the results from these measures with clinical and demographic information gathered from participants’ clinical records. Funded projects will also analyze the integrated behavioral and clinical data and ask whether the behavioral measures improve the ability to predict significant information in the clinical data—like whether a patient will respond to a given medication or whether their illness improves or worsens with time.

The second funding opportunity  will establish a data coordinating center that will accumulate data from across the different U01 projects, developing a data storage infrastructure and data analysis pipelines that will be open to the broader scientific community. The data coordinating center will perform analyses on these coordinated datasets. Even more importantly, they will be charged with helping other scientists test their own ideas using the IMPACT-MH data.

Combining behavioral data with data from clinical records will yield complex datasets, so NIMH is encouraging researchers to use computational approaches to discover patterns in the data that aren’t obvious but might help discover new ways of understanding individual differences among people seeking treatment for mental health problems or who are at risk of developing mental health problems. 

Finally, NIMH is also emphasizing the need to design these studies with an approach that is trans-diagnostic (not restricted to a single mental illness diagnosis), dimensional (measuring function along a continuum rather than as the presence or absence of an illness or ability), and integrative (involving data analyses that consider multiple streams of information at once), all principles consistent with NIMH’s Research Domain Criteria (RDoC) approach to understanding mental illnesses.

By combining computational approaches and RDoC principles, these initiatives are the first steps for IMPACT-MH, which we hope to grow over the next few years into a multi-faceted, concerted effort to bring precision psychiatry to patients like Cheryl. IMPACT-MH will enable the community of researchers, clinicians, and people who experience mental disorders to view mental illnesses in an entirely new way, allowing us all to see the important individual differences that make each patient unique and that point the way to personalized, effective interventions.

*I have changed the name and other biographical details to protect my former patient’s privacy.