One study found strong evidence that the current trial-and-error approach used in clinical practice for selecting the right antidepressant can be replaced by this new approach to precision medicine. The research has been published in the ‘Biological Psychiatry Journal’.
“This is a significant breakthrough. It is non-invasive. It can and should be used immediately,” said Madhukar Trivedi, MD, professor of clinical psychiatry and director of the Center for Depression Research and Clinical Care, one of the pillars from the Peter O’Donnell Jr. Brain Institute. Dr Trivedi said the new biomarkers could prevent patients with severe depression from taking the wrong medication for two to three months. Severe depression for that long can lead to job loss, the loss of a marriage, and even death by suicide. The study tested the common antidepressant drug sertraline with a control group taking a placebo. Patients who did not respond to sertraline after eight weeks were switched to the antidepressant bupropion. The researchers measured changes in the reactions of brain circuits as study participants performed a rewarding task in the scanner. Non-invasive functional magnetic resonance imaging (fMRI) was performed in more than 300 participants to assess changes in brain function at rest as well as during the rewarding task.
The study used this data and new innovations to build new machine learning models that tell scientists and clinicians which specific regions and circuits of the brain are associated with predicting treatment response for each drug. “The signatures we found are unique to the response of each antidepressant,” said Albert Montillo, PhD, assistant professor in the bioinformatics department at Lyda Hill, whose lab produced the 10,000 lines of code to effectively tune new ones. predictive models and sophisticated data cleaning methods to suppress fMRI head movement and achieve levels of precision never seen in other lab tests.
“Due to the inherent complexity of the human brain, neuroscientists generally find that brain activity can account for 15% of the variance in symptom relief. That would be an important scientific finding. Twenty percent is huge,” he said. said Dr Montillo. “In this study, we are able to explain 48% of the variance in symptom relief with sertraline, 34% for bupropion and 28% for placebo,” added Dr. Montillo.
Dr Trivedi said the results are very credible because the underlying data used by the research is broadly representative of the heterogeneity of clinical data, including data from Massachusetts General Hospital in Boston, Columbia University in New York and the University of Michigan, as well as the rigor of the analytical approach with the use of deep learning models. The study is one of the first adaptations of deep machine learning to predicting antidepressant outcomes, for which Dr Montillo developed methods to multiply the original fMRI data tenfold. His work has built models that reliably predict outcomes, especially on patients who were not used to train the models.
“This is a marked improvement over standard prediction approaches currently in use. We have also reached a point where our results are stable and may pave the way for future work,” said Dr Trivedi. Dr Montillo added, “The analytical approach we have developed can be easily adapted to identify biomarker signatures and predict outcomes for other treatments for depression, both pharmacological and non-pharmacological.
With a non-invasive approach and a wealth of evidence, Drs. Trivedi and Montillo said clinicians should take this approach now. They will seek additional funds to advance the research and see if it is compatible with the blood biomarkers developed by Dr. Trivedi. Drs. Trivedi and Montillo have no conflicts of interest related to this work. (ANI)
(This story was not edited by Devdiscourse staff and is auto-generated from a syndicated feed.)