Over the course of a lifetime, as many as one in five people will experience depression. For many, psychotherapy or first-line antidepressants are highly effective and allow people to get back on their feet in a few weeks or months. Yet others try treatment after treatment, struggling for years before they find something that alleviates their symptoms. Researchers have long clashed over precise efficacy rates, especially since we all respond differently to drugs and therapy. After all, the brain is our most mysterious organ — and it’s not always easy to discern what’s going on inside our skulls.
But a solution may be in the works: Scientists think that by carefully studying the brains of people with depression, they could eventually predict which treatments could work for specific patients, therefore eliminating months or years of frustrating trial-and-error. Although this research is still in its early stages, labs have already identified patterns of brain activity that seem to be associated with responses to certain antidepressant medications, cognitive behavioral therapy, and repetitive transcranial magnetic stimulation, a treatment that involves stimulating specific regions of the brain with a series of short magnetic pulses.
Looking Deeper
To find predictors of antidepressant response, scientists peek at brain function. They accomplish this using techniques like electroencephalography (EEG), which measures electrical activity generated by firing neurons, and functional magnetic resonance imaging (fMRI), which measures blood flow to different parts of the brain and shows which areas are most active.
Amit Etkin, a psychiatrist and neuroscientist who founded the biopharmaceutical company Alto Neuroscience, believes measurements of brain activity are likely to be more useful for predicting antidepressant response than genetic tests, another method suggested to decode the brain. “EEG is a much more specific measure of brain activity that tells you about circuits involved in cognition and emotion and so forth at the time that you’re trying to make the diagnosis … as opposed to genetics, which is there when you’re born and it’s there when you’re eighty,” he says.
Recently, Etkin and a team of researchers developed an algorithm that used EEG data to predict how patients would respond to a common antidepressant medication called sertraline. While the algorithm isn’t yet completely accurate for each patient, predicted improvements in depression scores were positively correlated with actual boosts on the Hamilton Depression Rating Scale, and generalized to people tested at different facilities.
While EEG is relatively inexpensive and more accessible, fMRI enables researchers to home in on tiny regions of the brain that can’t be visualized with EEG, says Vince Calhoun, an engineer and neuroscientist who directs the Center for Translational Research in Neuroimaging and Data Science.
Calhoun’s research group used fMRI to identify connectivity patterns between two brain networks, the default mode network and the cognitive control network, in people with depression. Some individuals with depression spent relatively little time in a state where these two networks were particularly anti-correlated — when activity in one network increased, activity in the other network decreased. These subjects, says Calhoun, were more likely to respond to treatment with electroconvulsive therapy (ECT), a treatment generally reserved for those with depression that is unresponsive to less invasive therapies. His team published their findings in Frontiers in Human Neuroscience this July.
Calhoun says an important next step for his research will be to replicate these findings, and show they hold true in various patients tested in various research centers.
Depression as an Indicator
But why do people with the same disorder respond so differently to the same treatment?
This is because depression may not ultimately be a single disorder. While two people could have similar symptoms, Etkin says, the underlying biological cause of the condition could vary. “There’s almost no way that something as common as depression could really have just one cause,” he says. Instead, he likens depression to a fever: “It’s an indicator of disease, as opposed to the disease itself.” Like a fever, there may be multiple underlying causes, and certain causes may respond best to certain treatments.
Finding predictive biomarkers will be important to understand the efficacy of currently available depression treatments, he explains, but it could also help guide development of new antidepressants, or the repurposing of existing drugs for other disorders as treatments.
The way many clinical trials are conducted — accepting patients based on symptoms of depression rather than underlying biomarkers — may actually be causing us to miss out on effective treatments, Etkin notes.
For example, he says, “Maybe [a new drug] can treat depression extremely well in 20 percent of people. But because you’re diluting them with the rest of the 80 percent of the sample that doesn’t respond, you can’t actually detect that the drug works.” In the same way that clinical trials of certain cancer therapies only accept patients whose cancer exhibits specific biomarkers, depression biomarkers could help scientists test new therapies in people who are most likely to respond to that specific treatment.
Etkin is more optimistic than most regarding how soon these technologies will reach a broader population. “At Alto, we have three drugs in phase two trials, where the goal is to identify responders … we hope to get a drug on the market by 2026.”
This piece has been updated with the correct spelling of Alto Neuroscience.
