Researchers at UT MD Anderson Cancer Center presented Path-IO, an AI model using pathology slides to predict immunotherapy responses in metastatic non-small cell lung cancer (NSCLC) patients at the AACR Annual Meeting 202612.
Path-IO stratified patients into high-risk and low-risk groups, with high-risk patients having double the risk of death or disease progression compared to low-risk12.
Validated on over 1,000 patients across multiple institutions and countries, Path-IO outperformed PD-L1 testing in all datasets12.
The model uses routinely gathered pathology data, is explainable, and is designed for clinical translation; future steps include prospective studies and integration with radiomics, clinical, and genomic data12.
Presented by Rukhmini Bandyopadhyay, Ph.D., from Jia Wu's lab12.
Sources:
1. https://medicalxpress.com/news/2026-04-aacr-platform-machine-responses-patients.html
2. https://beta.hyper.ai/en/stories/91972f68b2450c133e0c5a343cddda87