Doctors may be able to assess if immunotherapy medications help treat a patient’s cancer with the aid of the machine-learning model. Researchers at Memorial Sloan Kettering Cancer Center in New York and the National Cancer Institute’s (NCI) Center for Cancer Research led the study, which was published June 3, 2024, in Nature Cancer. The National Institutes of Health includes NCI.
The Food and Drug Administration has currently approved the use of two predictive biomarkers to help identify patients who might benefit from immune checkpoint inhibitor medication. Tumor mutational load, or the quantity of mutations in cancer cells’ DNA, is the first. The other is PD-L1, a protein in tumor cells that suppresses the immune system and is targeted by various immune checkpoint inhibitors.
These biomarkers do not, however, always correctly indicate how the body will react to immune checkpoint inhibitors. Although response can be accurately predicted by recent machine-learning algorithms that incorporate molecular sequencing data, this type of data is costly to acquire and is not frequently collected.
A distinct type of machine-learning model is described in the new study, which bases its predictions on five clinical features routinely obtained from patients: blood albumin level, age, cancer type, history of systemic therapy, and blood neutrophil-to-lymphocyte ratio, which is a measure of inflammation.
The tumor mutational burden, as determined by sequencing panels, is also taken into account by the model. Data from several different data sets, comprising 2,881 individuals treated with immune checkpoint inhibitors spanning 18 solid tumor types, were used to build and assess the model.
The model correctly predicted the total survival time as well as the time until the disease recurred for each patient, the time until the disease recurred for each patient and the anoint inhibitor. Interestingly, according to the researchers, the approach could also identify patients who could still benefit from immunotherapy despite having a low tumor mutational burden.
Larger prospective trials are required, according to the researchers, to fully assess the AI model in clinical settings. The Logistic Regression-Based Immunotherapy-Response Score (LORIS), an AI model, is now accessible to the general public. The program uses information on the six variables mentioned above to assess the probability that a patient would respond to immune checkpoint inhibitors.
Luc G. T. Morris, M.D., of Memorial Sloan Kettering Cancer Center, and Eytan Ruppin, M.D., Ph.D., of NCI’s Center for Cancer Research co-led the study. Tiangen Chang, Ph.D., and Yingying Cao, Ph.D., from Dr. Ruppin’s group at the NCI’s Center for Cancer Research, led the effort.
For more information: LORIS robustly predicts patient outcomes with immune checkpoint blockade therapy using common clinical, pathologic, and genomic features, nature cancer, DOI: 10.1038/s43018-024-00772-7
more recommended stories
-
Coffee and Tea May Lower Dementia Risk in Hypertensives
Researchers used data from 453,913 participants.
-
Machine Learning Links Sleep, Anxiety to Depression Risk
Researchers utilized machine learning alongside brain.
-
Unique Gut Bacteria Found in Singapore Seniors
Researchers used deep shotgun metagenomic sequences.
-
Disrupted Glucose Transport Linked to Myelin Thinning and Aging
Researchers looked at how oligodendroglial lipid.
-
Study Links Nighttime Light Exposure to Increased Alzheimer’s Risk
Researchers looked at the connection between.
-
Cruciferous Vegetables Lower Blood Pressure in Adults
Researchers conducted a crossover-design randomized controlled.
-
Machine Learning: Enhancing Flu Vaccine Development
A machine learning algorithm that can.
-
Immunotherapy Reduces Death & Recurrence in Lung Cancer
According to a new analysis by.
-
Hormones and Gender Differences in Immune Response
Scientists have discovered that hormones have.
-
Lifestyle Interventions Post-Colorectal Cancer Treatment
A randomized controlled trial (RCT) was.
Leave a Comment