THANK YOU FOR SUBSCRIBING
Pharma Tech Outlook | Wednesday, August 18, 2021
Immunotherapy and machine learning combine to help the immune system detect hidden tumor cells in the human body.
Stay ahead of the industry with exclusive feature stories on the top companies, expert insights and the latest news delivered straight to your inbox. Subscribe today.
FREMONT, CA: Due to the lymphatic and immunological systems, the body depends on the attack for defense. Pathogenic villains are tracked down and eliminated by the immune system, acting as the body's police force.
The immune system of the body is highly good at detecting abnormally behaving cells. These are cells that have the potential to become tumors or cancer in the future. When the immune system detects the cells, it attacks and kills them.
[vendor_logo_first]Tumor cells can interfere with the immune system's natural reaction. Proteins on the surface of tumor cells can efficiently put immune cells to sleep by turning them off. But tumor cells can evolve ways of hiding from the immune system, so it's not always that simple.
Fortunately, immunotherapy is a viable option for resurrecting immune cells and restoring anti-cancer immunity.
Introducing immunotherapy
Immunotherapy is a kind of cancer treatment that helps the immune system combat cancer cells. Immune checkpoint blockers (ICB) are immunotherapy that tells immune cells to reject cancer cell shutdown orders.
The development of ICB has opened a new era in cancer treatment. Even though ICB has been used efficiently to treat many individuals and cancer types, only one-third of patients react to the medicine.
The effect of ICB has been significant, but it could be even greater if doctors could rapidly identify which patients are most likely to react to the medication. It will also be great if professionals could figure out why some people do not respond to ICB.
Searching the tumor microenvironment
The researchers wanted to identify specific biomarkers in tumor samples from the patients to determine whether they would respond to ICB.
The team used computational methods and records from previous clinical patient care to examine the microenvironment of tumors for mechanisms that can function as biomarkers to predict patients' response to ICB.
The solution
Rather than examining the actual biological response to ICB treatment, the researchers used the same datasets to create numerous alternative immune responses. Even though they are not the primary response to ICB, they can be used together to measure its effectiveness.
The team trained machine learning models using a large public dataset with numerous patient samples due to this method. The researchers next evaluate the effectiveness of the machine learning models on different datasets in which the actual response to ICB treatment was recorded.
More in News