• Sustainability
  • DE&I
  • Pandemic
  • Finance
  • Legal
  • Technology
  • Regulatory
  • Global
  • Pricing
  • Strategy
  • R&D/Clinical Trials
  • Opinion
  • Executive Roundtable
  • Sales & Marketing
  • Executive Profiles
  • Leadership
  • Market Access
  • Patient Engagement
  • Supply Chain
  • Industry Trends

Study Shows Potential of Artificial Intelligence in Predicting Responses to Treatment for Major Depressive Disorder

News
Article

Investigators harness machine learning to evaluate response to sertraline treatment for major depressive disorder.

Image credit: photon_photo | stock.adobe.com. Artificial intelligence (AI), machine learning and modern computer technologies concepts. Business, Technology, Internet and network concept.

Image credit: photon_photo | stock.adobe.com

Researchers are beginning to realize the potential of artificial intelligence to predict the safety and efficacy of drugs across different indications and disease states. In a recent study published by Amsterdam University Medical Center, investigators found that a multimodal machine learning approach accelerated the time in which they could determine the efficacy of sertraline treatment in major depressive disorder (MDD).1

"The algorithm suggested that blood flow in the anterior cingulate cortex, the area of brain involved in emotion regulation, would be predictive of the efficacy of the drug. And at the second measurement, a week after the start, the severity of their symptoms turned out to be additionally predictive," said Eric Ruhé, psychiatrist, Radboudumc, in press release.2

In order to achieve this, the study authors relied on a secondary analysis of data from the EMBARC study, a double-blind, randomized clinical trial involving patients with MDD. Additionally, MR neuroimaging and clinical data were collected before and after one week of treatment. Assessment of performance in predicting remission after eight weeks was achieved through balanced accuracy and area under the receiver operating characteristic curve scores.

The trial, which included 296 adult outpatients with unmedicated recurrent or chronic MDD, found that the machine learning approach significantly outperformed chance in predicting response to sertraline treatment. External validation on placebo nonresponders and patients switched from placebo to sertraline indicated specificity for sertraline treatment compared to placebo.1

Results also displayed that a majority of patients involved had no response to the drug, with only one-third reporting positive results. The authors suggest that as a next step, prescribers should focus on finding other forms of treatment, especially considering that there are several adverse effects associated with the treatment.1,2

"This is important news for patients. Normally, it takes six to eight weeks before it is known whether an antidepressant will work," said one of the study authors Liesbeth Reneman, professor of Neuroradiology, Amsterdam UMC, in a press release.2

Despite the success of the study, the authors also found that there were several limitations. They stressed that in the future, there needs to be more of an emphasis on validation on larger datasets, different antidepressants, and populations with more clinical heterogeneity. Furthermore, they noted a lack of information on responses to previous antidepressant treatments.1

“We found no significant difference in performance between pretreatment and early-treatment prediction. If externally validated, early-treatment response prediction would likely not require a second session of MRI scanning, reducing cost and lowering patient burden. Because our results show a performance drop in unimodal ASL models at early treatment but not at pretreatment, and we decided a priori to use relative changes at early treatment, which are sensitive to physiological variability, we suspect that this choice was suboptimal for ASL predictors,” the study authors wrote.1

In order to overcome this limitation, the authors suggested combining absolute ASL predictors with early-treatment clinical predictors to improve performance.1

The study contributes to understanding individual variability in response to antidepressant treatment and highlights the potential of predictive modeling in personalized medicine. Personalized medicine approaches, informed by predictive modeling, could optimize treatment outcomes and reduce patient burden and costs. It also suggests that opportunities will arise to individualize clinical sertraline treatment for MDD patients, also leading to more tailored plans for patients with MDD.

References

1. Treatment Response Prediction in Major Depressive Disorder Using Multimodal MRI and Clinical Data: Secondary Analysis of a Randomized Clinical Trial. The American Journal of Psychiatry. February 7, 2024. Accessed February 15, 2024. https://ajp.psychiatryonline.org/doi/abs/10.1176/appi.ajp.20230206?download=true&journalCode=ajp

2. Artificial intelligence helps predict whether antidepressants will work in patients. EurekAlert. February 7, 2024. Accessed February 15, 2024. https://www.eurekalert.org/news-releases/1033410

Recent Videos