Brain Circuit Analysis Reveals Six Distinct Subtypes of Depression and Anxiety

Researchers at Stanford University and their colleagues employed functional magnetic resonance imaging (fMRI) to evaluate task-free and task-evoked data of brain circuit activity in a sample of 801 patients with depression and anxiety of varying severities along with 137 control individuals without these disorders. Almost all the participants with depression or anxiety were unmedicated. Of this group, 250 patients who subsequently enrolled in a clinical trial where they received either behavioral or pharmacotherapy were imaged again after completing treatment.

While prior studies have relied largely on a whole-brain, task-free use of connectivity biomarkers with thousands of features, the current study employed a theoretically driven approach that limited inputs by having participants conduct cognitive tasks in areas known to be affected by depression or anxiety, such as sustained attention and identifying emotions. The researchers said that narrowing the scope may ultimately help guide clinical therapies going forward.

The underlying thesis of the study was that depression and anxiety are disorders of brain circuit function. “Patients with depression and anxiety exhibit dysfunction in the activity and connectivity of brain circuits in response to specific probes of general and emotional cognition,” the authors wrote.

The researchers employed a new standardized imaging technology—the Stanford Et Cere Image Processing System—which analyzes six different brain circuits: three controlling intrinsic activities that are always active and three that turn on and respond to external stimuli.

By comparing the fMRI data with patient symptom profiles and behavioral measures, the authors were able to identify and validate six different depression/anxiety biotypes, which they named after the specific circuits affected. For example, biotype NSA+PA+ involved increased (+) activation in both the negative affect circuit and the positive affect circuit during an emotion-processing task, particularly for sad stimuli. Individuals with this biotype generally had severe anhedonia and ruminative brooding. The biotype was transdiagnostic and found in people with major depression, generalized anxiety, and social anxiety.

The authors also found that some of these biotypes were more likely to respond to some treatments compared with others. For example, patients with the CA+ biotype—hyperactivation of the circuit for cognitive control—were more likely to respond to venlafaxine than other antidepressants; this biotype’s symptoms include anhedonia but also anxious arousal and alterations in threat response.

The authors concluded that their identification of six biotypes is just one of many possible solutions to disentangling heterogeneity within psychiatric disorders, noting that the biotypes point to the existence of multiple neural pathways that show up as depression and anxiety. Symptom differences between biotypes were relatively small, which highlights the need for finer-grained clinical measures that can be used consistently in future studies.

This research is already being initiated into practice, corresponding author Leanne Williams, Ph.D., director of the Stanford Center for Precision Mental Health and Wellness, told Psychiatric News. “We’ve established a translational clinic at Stanford using our Stanford Et Cere Imaging system to biotype patients referred to the clinic,” Williams said. “Reimbursement for scans has been demonstrated. A goal for this clinic is to serve as the flagship for demonstrating the utility of biotyping and [eventually] disseminating to other sites.”

Jacobus F.A. Jansen, Ph.D., is a professor at Maastricht University in the Netherlands who specializes in the use of fMRI as a means for improving treatment and care of people with cognitive and neurological disorders. He told Psychiatric News that the current study shows great promise, but more research is required. “I would be most convinced if their methodology would hold up in a prospective study where these biotypes would, indeed, predict treatment outcome,” Jansen said.

For most psychiatric conditions, regular MRI scans are negative, and hence not incorporated in the clinical workup, he continued. While fMRI may reveal more, routine use of it in clinical evaluations would be costly and hard to justify at this time. Use of fMRI in psychiatric diagnosis and treatment prediction “remains tricky,” Jansen noted. Functional MRI is “by nature very noisy, and its signal is confounded by several physiological processes that are likely not relevant.”

Beata Godlewska, M.B.B.S., Ph.D., is a clinical researcher and honorary consultant psychiatrist at the University of Oxford who has studied the use of neuroimaging as a tool for individualized treatment in depression. While she agreed the current study needs replication, she thinks it should also encourage future explorations of biotypes based on unique patterns of brain function during task performance and at rest, and their relationship to response to pharmacological treatments.

“While MRI is still considered costly,” Godlewska said, “if response prediction is reliable, the costs of the scan will certainly [be less than] the costs of failed treatments and work time lost to illness, not to mention personal suffering.”

Depression and anxiety disorders continue to constitute major causes of premature death and long-term disability, with an estimated 280 million people worldwide suffering from depression alone. The World Health Organization projects that by 2030 depression will rank as the leading contributor to the global burden of disease.

The study findings point to the need for clinicians to focus more on the precise clustering of symptoms rather than a generic diagnosis of depression and anxiety, given that this may have practical relevance to treatment choice, Godlewska added. Even with its practical limitations, the study “offers a glimmer of hope that precision psychiatry—the delivery of personalized treatments for individual patients—based on data that clinicians can easily interpret and use in everyday practice, may one day be practical.”

This research was supported by the National Institutes of Health, the Gustavus and Louise Pfeiffer Research Foundation, and Brain Resource Ltd. ■