Love Employee/iStock via Getty Images
Cancer immunotherapies are among the most expensive therapies available, yet for many patients, they simply don’t work well. Artificial intelligence might help.
New results from a study conducted by GE HealthCare Technologies (NASDAQ:GEHC) and Vanderbilt University Medical Center showed that using AI can predict patient responses with 70% to 80% accuracy.
The results were published recently in the Journal of Clinical Oncology Clinical Cancer Informatics.
The study used a retrospective analysis of responses to immunotherapy treatment from thousands of the medical center’s cancer patients and correlated them based on demographic, genomic, tumor, cellular, proteomic and imaging data.
AI models were designed to predict how effective certain immunotherapies would be, as well as the likelihood of a individual patient developing an adverse reaction.
Cancer immunotherapies can take on several forms, including CAR-T-cell therapies, checkpoint inhibitors, and monoclonal antibodies. CAR-Ts are the newest — and most promising — but also come with very high price tags.
A 2019 study in the journal Blood found that the the mean total cost of CAR-T treatment (41 days prior to through 154 days after CAR-T index) for relapsed/refractory (r/r) B cell non-Hodgkin lymphoma was $618,100.
Using AI to predict treatment response could have a serious impact on cancer treatment. By 2050, the number of people with cancer worldwide is projected to grow to over 35M compared to 20M in 2022, according to the World Health Organization’s International Agency for Research on Cancer.
GE HealthCare (GEHC) said it is evaluating plans to commercialize AI models as they could be used both in a clinical setting and for developments of future treatments.
“We aim to partner with pharmaceutical companies, researchers and clinicians to optimize and ultimately apply the AI models in therapy development and in clinical practice,” Jan Wolber, Global Product Leader – Digital at GE HealthCare’s Pharmaceutical Diagnostics division said in a statement.
Researchers noted that the models could be used for therapies in other areas, such as neurology or cardiology.