This summer was not only a season for travel, as I reported in my previous message, but also for reading. Among the items that caught my eye in the many hours I spent en route were a pair of studies on biomarkers. A biomarker is a biological or biobehavioral measurement that tells you something relevant about a patient. Think blood tests and electrocardiograms (EKGs). Some biomarkers are diagnostic: chest pain, for example, could mean indigestion or a heart attack. A blood test might reveal the presence of heart muscle enzymes in your blood, or an abnormal EKG could confirm damage to the heart’s electrical conduction system. Other biomarkers suggest at-risk states (like high blood pressure or high cholesterol pointing to risk of cardiac disease) or serve as a way to monitor disease progression (like tumor markers for certain types of cancer or hemoglobin A1c levels for diabetes). Biomarkers are most useful when they suggest a particular course of treatment, as these examples all do. Notably, psychiatry has no biomarkers, and many studies aim to remedy this gap in knowledge with the aim of helping doctors and patients make important clinical decisions.
The two studies both focused on identifying biomarkers for depression that might help refine the diagnosis and guide treatment. The first, Drysdale et al, by a large group of collaborators led by Dr. Conor Liston, used brain scans to study over 1000 people with depression.1 Measuring brain activity with functional magnetic resonance imaging (fMRI), they characterized each individual by how activity in different brain regions correlated with each other, creating individualized maps of so-called “functional connectivity,” since they reflect how each brain region seems to connect with the others. Using a mathematical clustering algorithm, the investigators could identify different types of connectivity maps in the sample of people with depression, which they called “biotypes.” Interestingly, even though symptoms didn’t enter into the clustering algorithms, the four different biotypes corresponded with different kinds of symptoms. Moreover, one of the biotypes responded better to a particular treatment for depression—transcranial magnetic stimulation (TMS)—suggesting the possibility that using fMRI functional connectivity maps could serve as a useful biomarker, one that suggests a particular treatment course. With NIH support, Dr. Liston is also using optogenetics and fMRI to investigate the mechanisms by which chronic stress can alter functional connectivity, an avenue towards understanding the role of stress—at the circuit, cell, and ultimately molecular level—in the development of disorders like depression and perhaps even these biotypes.
The second study, led by Dr. Helen Mayberg, also looked at functional connectivity in patients suffering from depression. Taking a more hypothesis-driven approach, this study built on previous work suggesting that changes in activity in a brain region called the anterior cingulate cortex were associated with effective treatment for depression. Interestingly, the changes in activity could be opposite in direction, depending on the treatment. Successful cognitive behavioral psychotherapy (CBT) increased activity in this area, while successful medication treatment decreased it. These contradictory effects of successful treatment led Dr. Mayberg to the hypothesis that there might be two types of depression—one in which too much cingulate connectivity to other regions drove depression and needed to be downregulated, and one in which too little drove depression and needed to be upregulated. The latter would presumably respond to medication, while the former would respond to CBT. Dunlop et al, the latest paper from the group, puts this hypothesis to the test, finding that connectivity between the cingulate cortex and other parts of the brain predicts differential response to medication vs. psychotherapy, just as Dr. Mayberg predicted.2
These two papers represent important initial steps towards clinically useful biomarkers that signal the need for further work. First, both need to be replicated, although Dr. Liston’s group did include a replication sample. The targeted approach of Dr. Mayberg resulted in a relatively specific pattern of connectivity being associated with treatment response, which may be challenging to replicate. Second, both approaches need to be refined to make them truly clinically useful. The biotypes discovered by Dr. Liston were only tested with one treatment; determining whether the biotypes that do not respond as robustly to TMS are better treated with other approaches (such as CBT or medication) is necessary if they are to help clinicians choose between treatments.
These papers presage an age of biomarker-driven precision medicine that will change the practice of psychiatry. But if you think about it, psychiatry has been doing precision medicine for a hundred years, just without the biomarkers. Psychiatrists are trained to really get to know the individual patient, what drives his or her difficulties here and now and how those difficulties arose, even tracing them back to early childhood. This deep connection to the individual is what sustained my interest in this richly rewarding medical specialty. Which brings me to a little “light” reading I’ve been squeezing in over the summer. I’ve started the Norwegian author Karl Ove Knausgård’s opus, “My Struggle,” a deeply personal, fictionalized autobiography in six volumes. In volume 1, he describes a young Karl Ove’s family life, spread over several isolated vignettes and narrated by the same character at nearly 40. The narrator is older, not clearly wiser, but clearly troubled. Reading it is like listening to a long patient narrative—you get selective, personal memories from the individual’s unique perspective, and also a powerful understanding of the effect of early life experiences on adult emotionality and function. Volume 1 is really about Karl Ove’s fraught relationship with his father, and in volume 2, which I’m about halfway through, you see how this influences his own relationship with his young family 30 years later. The brave, deeply honest prose is frighteningly open and startlingly real. And it reminds the reader that deep within all of us is a complex web of experience and biology that shapes how we behave. Unraveling that web to learn how to help a particular individual recover is a tall order indeed. But it is a challenge we must tackle if we are to truly transform psychiatry so we can better help those who suffer.
An interesting summer of travel and reading, but I’m looking forward to the coming academic year. New initiatives on suicide prevention, computational neuroscience, and a workshop on neural circuits await at NIMH. And lots of opportunities to learn about exciting new results coming from the many NIMH-funded investigators working hard to make things better for all of us.
1 Drysdale AT et al. Resting-state connectivity biomarkers define neurophysiological subtypes of depression. Nat Med. 2017 Jan;23(1):28-38. Doi: 10.1038/nm.4246.
2 Dunlop BW et al. Functional connectivity of the subcallosal cingulate cortex and differential outcomes to treatment with cognitive-behavioral therapy or antidepressant medication for major depressive disorder. American Journal of Psychiatry. 2017 Jun 1;174(6):533-545. doi: 10.1176/appi.ajp.2016.16050518.