KRAS mutations are among the most prevalent oncogenic drivers in cancer, yet treatment with first-generation KRAS inhibitors (KRASi), such as sotorasib, offers limited and often short-lived benefit. We analyzed RNA-sequencing data from 72 non-small cell lung cancer (NSCLC) patient biopsies using Genialis™ krasID, a predictive RNA-based biomarker, for stratifying sotorasib response. krasID integrates modular biological signatures relevant to KRAS biology to stratify patients by predicted response and duration of benefit. krasID achieved high predictive performance and accurately identified patients likely to remain on therapy at six months, a key milestone aligned with clinical progression benchmarks. Longitudinal analysis of pre- and post-treatment samples revealed distinct resistance mechanisms, some of which preserved therapeutic modulation of KRAS biology.

Published for ASCO 2025.