AI tool predicts response to cancer therapy (2024)

Different immune cells, shown in yellow, red, and blue, attacking a mouse tumor after treatment with radiation and an immune checkpoint therapy. Steve Seung-Young Lee / Univ. of Chicago Comprehensive Cancer Center / National Cancer Institute

See Also

Kettering Health hiring Credentialed Medical Asst - Pre-Employment Services - Miamisburg - FT/Days in Miamisburg, Ohio, United States | LinkedIn

Chemotherapy, radiation, and surgical removal of tumors have long been the standard approaches for treating different types of cancer. But in recent decades, different immunotherapies have become available. These rely on the body’s immune system to find and destroy cancer cells. One type of immunotherapy, called checkpoint inhibition, has greatly improved the treatment of many types of cancer. Immune checkpoint inhibiting drugs can make cancer cells more vulnerable to immune system attack. But they don’t work for everyone.

Scientists have been looking forbetter ways to identify patients most likely to respond to these drugs. People who probably wouldn’t benefit from them could avoid unnecessary treatments and side effects and be given different treatments. To date, two biomarkers have been approved by the U.S. Food and Drug Administration to identify patients most likely to benefit from these medications. One measures tumor mutational burden, which is the number of DNA mutations in cancer cells. But results of these tests are not always accurate. Other predictive tests depend on tumor molecular data that are costly to obtain and not routinely collected.

A research team led by Dr. Eytan Ruppin of NIH’s National Cancer Institute and Dr. Luc Morris of Memorial Sloan Kettering Cancer Center set out to create a more accurate predictive tool based on readily available biomarkers. To do this, they first analyzed a large data set that included information on more than 2,880 cancer patients with 18 different types of solid tumors. All had been treated with immune checkpoint inhibitors.

The team assessed over 20 different clinical, pathologic, and genomic features. They also examined patient outcomes, such as response to therapy and survival. Using machine learning, they tried to identify which combination of features could best predict a patient’s response to immune checkpoint inhibitors. Results were published on June 3, 2024, in Nature Cancer.

After developing and testing different machine learning models, the team created a new type of AI scoring system, termed LORIS (logistic regression-based immunotherapy-response score). It is based on the tumor mutational burden along with five clinical features that are routinely collected from patients. These include the patient’s age, cancer type, history of cancer therapy, blood albumin (a protein made by the liver), and blood NLR (a measure of inflammation).

Further testing showed that LORIS was better than other methods at predicting a patient’s chance of responding to immune checkpoint inhibitors. This included predictive models based on many more clinical features. LORIS could also consistently predict short-term and long-term survival after immunotherapy. The scientists note that this approach could help guide treatment decisions and maximize benefits to patients. But larger studies are needed to evaluate the tool in clinical settings.

“We were able to develop a new predictive model for immunotherapy response across many different cancer types using only six simple variables,” Morris says. “In contrast to prior models, some of which are very complex, this model is very accessible to clinicians.”

Ruppin adds, “this study provides another example for the importance and benefit of building collaborations between clinicians and data scientists across different centers in our nation to collect and analyze large patient data cohorts to advance patient care.”

The LORIS tool is publicly available at https://loris.ccr.cancer.gov.