AI Model Enhances Precision of Immunotherapy in Gastric and Colorectal Cancer (Image supported by ChatGPT)
SEOUL, May 28 (Korea Bizwire) — A South Korean-led research team has developed an artificial intelligence (AI) model capable of predicting a cancer patient’s response to immunotherapy, potentially offering a new level of precision in tailoring treatments for gastric and colorectal cancers.
Professor Jae Ho Jeong of Yonsei University College of Medicine, in collaboration with researchers from the Mayo Clinic and Vanderbilt University Medical Center in the United States, announced the development of the model on Tuesday.
The AI tool analyzes cancer cell images to determine whether patients are likely to benefit from immune checkpoint inhibitors, a class of drugs that activate the immune system to attack tumors.
While immunotherapy has emerged as a transformative cancer treatment, its effectiveness varies significantly depending on genetic factors — particularly the presence of microsatellite instability-high (MSI-H), a biomarker indicating better response in certain gastrointestinal cancers.
Traditional detection methods, such as immunohistochemical staining, often miss MSI-H when it is located in less visible areas of tumor tissue.
Yonsei Cancer Center. (Image courtesy of Yonhap)
The newly developed AI model addresses that limitation by breaking down pathological images into thousands of micro-segments and analyzing each for MSI-H probability.
It not only detects the presence of MSI-H with high precision but also maps its exact location and distribution within the tumor — offering physicians objective, visual guidance to support treatment decisions. The model also provides confidence scores for its predictions.
Notably, in clinical validation trials, the AI model identified MSI-H in several patients previously deemed ineligible for immunotherapy under standard testing methods. These patients went on to receive immune checkpoint inhibitors and responded positively to the treatment.
“This AI model enhances our ability to personalize cancer care by identifying patients who are likely to benefit from immunotherapy,” said Prof. Jeong. “It provides a scientific basis for prescriptions, increasing accuracy in clinical decisions.”
The study was published in the latest issue of npj Digital Medicine, a peer-reviewed journal by Nature Publishing Group.
M. H. Lee (mhlee@koreabizwire.com)
