BACKGROUND:The essential roles of the tumor microenvironment (TME) have been recognized during the initiation and progression of primary lung adenocarcinoma (LUAD). The aim of the present study was to delineate the immune landscape in both primary cancer and matched lymph node metastasis from a cohort of locally advanced stage LUAD patients with distinct outcomes.METHODS:Formalin-fixed, paraffin-embedded samples were collected from 36 locally advanced LUAD patients. Transcriptome data of the tumor immune microenvironment were resolved using an immune oncology panel RNA sequencing platform. Bioinformatics approaches were used to determine the differentially expressed genes (DEGs), dysregulated pathways, and immune cell fraction between patients with early recurrence (ER) and late recurrence (LR).RESULTS:Here, we showed that in primary cancer tissues, 23 DEGs were obtained between patients with ER and LR. Functional analysis revealed that the LR in LUAD patients may be associated with enriched gene sets belonging to the antigen presentation and MHC protein complex, innate immune response, and IFN-γ signaling pathways. Next, the transcriptome data were adopted to quantify immune cell fractions, indicating that high infiltration of mast cells and neutrophils was correlated with ER. Interestingly, similar findings were observed in metastatic lymph nodes from patients suffering from ER or LR. By analyzing the shared immune features of primary cancers and lymphatic metastases, we unraveled the prognostic value and joint utility of two DEGs, CORO1A and S100A8.CONCLUSIONS:In LUAD, the enrichment in antigen presentation, MHC protein complex, and IFN-γ signaling, and low infiltration of neutrophils in primary or metastatic nodules may be indications for a favorable prognosis. Integrated with bioinformatics approaches, transcriptome data of immune-related genes from formalin-fixed, paraffin-embedded (FFPE) samples can effectively profile the landscape of the tumor immune microenvironment and help predict clinical outcomes.

e20625 Background: The resection of early stage NSCLC offers patients the best hope of a cure. However, the recurrence rate post-resection remains high. As the mechanisms involved in the process is still not clear due to the unavailability of accurate targets, our study was aimed to integrate the impact of different immune context present in lung adenocarcinoma (LUAD) microenvironment on patients’ prognosis. Methods: RNA targeted sequencing was performed on 24 primary tumor specimens from the resected local advanced LUADs . Transcripts of 395 immune related genes expressed in FFPE tumor samples were analyzed. The limma package was used to analyze the different expressed genes (DEGs) between patients with different prognosis. The gene set variance analysis (GSVA) analysis was performed to explore gene sets enrichment related to the prognosis (PFS, progression free survival) post-resection. Results: 23 DEGs were detected in primary tumor between the better (PFS > 18months, n = 12 ) and worse (PFS≤18months, n = 12 ) prognosis group. The combined prediction model containing MPO, IL-6, CXCR2, FCGR3B, ADGRE5 could identify the favorable prognosis of patients. GSVA and Log Rank test of survival data demonstrated that the antigen processing and lymphocyte activation pathway enrichment may associate with better prognosis (p = 0.01), whereas higher Neutrophils cell infiltration in primary tumor demonstrated a shorter PFS (p = 0.008). Conclusions: In LUAD, the immune related genes such as MPO, IL-6, CXCR2, FCGR3B, ADGRE5, can effectively profile the landscape of tumor immune microenvironment and predict the survival in early stage of lung adenocarcinoma. Accordingly immune pathways were correlated with prognosis of these patients. Our findings suggest that immune-related RNA expression pattern in locally advanced LUAD may provide a potential predictive marker for early recurrence after surgical resection.