Multiregional transcriptomic profiling provides improved prognostic insight in localized non-small cell lung cancer

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Lung cancer is a leading cause of cancer death globally.
• Non-small cell lung cancer (NSCLC) exhibits significant transcriptomic intratumor heterogeneity (RNA-ITH), hindering reproducible prognostic models.
• Understanding RNA-ITH is crucial for improving NSCLC prognostication.

Purpose of the Study:

• To investigate the impact of RNA-ITH on prognosis in localized NSCLC at gene, signature, and tumor microenvironment levels.
• To determine if multiregional expression profiling enhances the performance of prognostic signatures.
• To explore the role of immune cell infiltration heterogeneity in NSCLC prognosis.

Main Methods:

• Utilized multiregional RNA-sequencing data from public (TRACERx) and internal (MDAMPLC) cohorts (880 tumor samples, 350 individuals).
• Analyzed gene expression, prognostic signatures, and immune cell infiltration levels across different tumor regions.
• Applied survival analysis to assess the prognostic value of regional gene expression and immune cell infiltration.

Main Results:

• Maximal expression of hazardous genes and minimal expression of protective genes across tumor regions were more prognostic than average expression.
• Prognostic prediction accuracy of gene signatures significantly improved using regional expression extremes (max hazardous, min protective).
• Tumor microenvironment analysis revealed that regional immune cell infiltration extremes (lowest protective or highest hazardous) determined NSCLC patient prognosis.

Conclusions:

• RNA-ITH significantly impacts NSCLC prognostication at gene, signature, and tumor microenvironment levels.
• Multiregional assays and analysis of regional expression extremes can optimize expression-based prognostic biomarkers.
• Incorporating RNA-ITH into prognostic models holds great potential for improving patient stratification in NSCLC.