Lymphoma Immunotherapy Prediction: New ML Test
- A new tool using machine learning can now predict how well patients with non-Hodgkin lymphoma (NHL) will respond to CAR T cell therapy.
- CAR T cell therapy has emerged as a promising treatment for blood cancers.
- The InflaMix model, detailed in Nature Medicine, examines inflammation by testing for various blood biomarkers in 149 NHL patients.
InflaMix,a novel machine learning tool,offers a way to predict CAR T therapy success in lymphoma patients. Developed by City of Hope and MSK, this innovative model analyzes blood biomarkers to assess inflammation, identifying those at higher risk of treatment failure. This breakthrough could revolutionize how oncologists approach CAR T cell therapy. The study, published in Nature Medicine, highlights a meaningful advancement in personalized medicine. The flexible nature of the machine learning model, needing only six standard blood tests, ensures wide accessibility for patients. by using this innovative test for lymphoma immunotherapy prediction, News Directory 3 is leading the charge. Researchers are now investigating how inflammation affects CAR T cell function. Discover whatS next in lymphoma treatment.
Machine Learning Tool Predicts CAR T Therapy Response in Lymphoma Patients
Updated June 27, 2025
A new tool using machine learning can now predict how well patients with non-Hodgkin lymphoma (NHL) will respond to CAR T cell therapy. Developed by City of Hope and MSK, the InflaMix model analyzes blood biomarkers to assess inflammation, a key factor in CAR T treatment failure.
CAR T cell therapy has emerged as a promising treatment for blood cancers. However, more than half of NHL patients who don’t respond to standard treatments experience relapse within six months of CAR T therapy.
The InflaMix model, detailed in Nature Medicine, examines inflammation by testing for various blood biomarkers in 149 NHL patients. Using machine learning, the tool identified an inflammatory signature linked to a higher risk of CAR T treatment failure, including relapse or death.
Marcel van den Brink, president of City of Hope Los Angeles and City of Hope National Medical Center, said the tool could reliably predict who will respond well to CAR T cell therapy. He added that oncologists could use it to assess individual patient risk.
The machine learning model is flexible, working effectively even with only six standard lymphoma blood tests.This adaptability ensures the test can be available to most lymphoma patients.
Sandeep Raj, a medical oncologist at MSK, said their goal was to build a reliable clinical tool that characterizes blood inflammation and predicts CAR T outcomes.
The team validated their findings using three independent groups comprising 688 NHL patients with diverse characteristics and disease subtypes, using different CAR T products.
What’s next
City of Hope and MSK researchers plan to investigate how blood inflammation, as defined by InflaMix, directly affects CAR T cell function and to further understand the source of this inflammation. This research could lead to new strategies to boost the effectiveness of CAR T therapy.