Researchers from the City of Hope and Memorial Sloan Kettering (MSK) Cancer Center, US, have developed a tool named Inflammation Mixture Model (InflaMix) that can predict the likely response of non-Hodgkin lymphoma (NHL) patients to chimeric antigen receptor T-cell (CAR T) therapy.
Using machine learning, InflaMix was designed to evaluate inflammation, which may cause CAR T therapy to fail, by analysing a set of blood biomarkers in 149 subjects with NHL.
Go deeper with GlobalData
The tool utilises algorithms to identify patterns in blood test data, pinpointing an inflammatory biomarker.
With the analysis of the inflammatory signature detected by the tool, the researchers claim to have identified that it was related to the high CAR T treatment failure, including mortality or disease relapse risk.
The organisation noted that the tool is an “unsupervised” model, trained with no prior knowledge of clinical outcomes.
The team highlighted that the tool functioned well even with a reduced dataset from only six standard blood tests for lymphoma patients.
Validation of InflaMix’s predictive capability was conducted through studies of three independent cohorts, consisting of 688 NHL subjects with diverse clinical characteristics and subtypes of the disease, treated with various CAR T products.
Future research by the City of Hope and MSK teams will focus on whether blood inflammation identified by the tool directly affects the function of the CAR T and the origin of this inflammation.
The studies were partially funded by an MSK Support Grant, the National Institutes of Health and the National Cancer Institute.
The primary research was conducted at MSK, with the involvement of City of Hope Los Angeles and City of Hope National Medical Center president Dr Marcel van den Brink, who joined City of Hope in 2024 after over two decades at MSK.
Dr van den Brink said: “These studies demonstrate that by using machine learning and blood tests, we could develop a highly reliable tool that can help predict who will respond well to CAR T therapy.
“With a rigorous statistical approach, we demonstrated that this is one of the most thoroughly validated tests we have for predicting CAR T outcomes in lymphoma patients and could be used by oncologists everywhere to assess the risk of CAR T in an individual patient.”
