The study focuses on the prospective validation of the ORACLE prognostic signature in lung adenocarcinoma (LUAD), utilizing multiregion RNA sequencing (RNA-seq) data from an expanded cohort of patients within the TRACERx study. ORACLE was benchmarked against six other RNA-seq-based prognostic signatures for LUAD, showing a lower discordant risk classification rate and higher stability in expression across tumor regions. The research highlights ORACLE's significant association with overall survival (OS), lung-cancer-specific survival, and disease-free survival (DFS), independent of known clinicopathological risk factors.
ORACLE's effectiveness was further demonstrated in a pseudo-single biopsy cohort, where it maintained a significant association with OS across 1,000 iterations, suggesting its reliability even with limited biopsy samples. In patients with stage I LUAD, ORACLE risk classes showed a significant association with OS, outperforming current clinical standards in prognostic utility.
The study also explores ORACLE's potential as a biomarker for invasive and metastatic potential, indicating that high-risk clonal expression changes are detectable in preinvasive lesions and primary tumors with early systemic dissemination. Additionally, ORACLE's utility in identifying chemosensitivity in treatment-naive patients was investigated, revealing a positive correlation between ORACLE risk score and sensitivity to certain chemotherapies, particularly cisplatin.
Overall, the findings suggest that ORACLE serves as a comprehensive metric of lung cancer evolution, capturing genetic evolutionary metrics and providing complementary prognostic information alongside other biomarkers. The study underscores the importance of clonal expression signals in understanding tumor evolution and their potential in guiding personalized treatment strategies for LUAD patients.
