Multiregion gene-expression profiling to reveal heterogeneity in molecular subtypes and immunotherapy response signatures in lung cancer.

Abstract

e20077 Background: Intra-tumor heterogeneity (ITH) may be present in all molecular levels. Genomic ITH at the exome level has been reported in many cancer types, but comprehensive gene expression ITH has not been well studied. Methods: We collected 35 samples from 10 NSCLC patients (3 or 4 regions/tumor) including lung adenocarcinoma (n = 6), squamous cell carcinoma (n = 2), large cell carcinoma (n = 1) and pleomorphic carcinoma (n = 1). Using Affymetrix Gene 1.0 ST arrays, we generated the gene expression data for each sample. Diverse gene expression signatures associated with clinical outcomes were tested for ITH. Results: Inter-tumor heterogeneity was generally higher than intra-tumor heterogeneity, but some tumors showed a substantial level of ITH. The analysis of various clinically relevant gene expression signatures including molecular subtype, EMT and immunotherapy response signatures also revealed heterogeneity between different regions of the same tumor. The gene expression ITH we observed was associated with heterogeneous tumor microenvironments represented by stromal and immune cells infiltrated. Conclusions: Our data suggest that RNA-based prognostic or predictive molecular tests should be carefully conducted in consideration of the gene expression ITH.