38eP AI-based identification of TKI-sensitive driver mutations using routine histopathology in lung adenocarcinoma

Tyrosine kinase inhibitors (TKIs) are a cornerstone of targeted therapy in lung adenocarcinoma, with TKI-sensitive driver mutations guiding treatment selection and predicting therapeutic response. Identifying these mutations currently relies on tumor genomic testing, which can be costly and may limit access in some clinical settings. Although prior studies have shown that histopathology images can predict specific mutations such as EGFR, they have not focused on clinically actionable TKI sensitivity and have often relied on task-specific end-to-end models.