P1.04-58 Uncovering the Tumor Microenvironment of KRAS-Driven Lung Adenocarcinoma: The Link Between Th17 Signaling and B Cell

Non small cell lung cancer, been histologically classified into adenocarcinoma (AD) and squamous cell carcinoma, is one of the most deadly malignancy worldwide. Lung AD (LUAD) could benefit of a plethora of target therapies and, in the last few years, also of immunotherapies. Here we focused on a cohort of LUAD aiming to gain insights into the immune contexture of such a malignancy. 20 patients affected by advanced LUAD, previously analyzed through the CE-IVD Oncomine solid tumor DNA kit, have been included in our cohort. DNA and RNA were isolated from 6 μm thick FFPE sections using the QIAamp DNA/RNA FFPE Tissue Kit (Qiagen). Two custom panels were designed using Ion Ampliseq Designer Tool. DNA and RNA libraries were prepared according to manufacturer’s instructions. Torrent suite variant caller and Vardict tool were used to call variants, subsequently annotated with Annovar. RNA raw read counts were analyzed by DESeq2 R package. TIMER web-tool was used to deconvolve immune-cytotype composition and LUAD-TCGA dataset has been used to validate our findings. Immune infiltration results were validated with immunohistochemistry in an independent cohort. We explored the mutational status of 41 genes and the expression of 94 genes, related to immune-checkpoint, inflammation and stromal microenvironment. Surprisingly, we found that our cohort has a very low mutational burden if we consider our panel as its surrogate. Regarding gene expression data, we identified 31 genes significantly deregulated in tumor tissues compared with a pool of normal pleura samples. Unsupervised hierarchical clustering of the deregulated genes is able to identify two clusters of tumor samples, differently enriched in alterations in actionable. In particular, we identified a cluster enriched in patients carrying KRAS alterations. GO/KEGG enrichment displayed terms related, as expected, to T cell differentiation but more interestingly term linked to Th17 lymphocytes. Thus, we perform in silico deconvolution through TIMER algorithm. Estimation performed on our gene expression matrix showed that, after stratification based both on cluster and KRAS mutational status, B cell infiltration is lower in KRAS-mutated enriched cluster. Notably, also in LUAD-TCGA dataset, B cell infiltration is significantly low in KRAS mutated patients. Such a finding has been validated in situ through immunohistochemistry in an independent cohort. Moreover, cases in LUAD-TCGA with low B cell infiltration have a significantly worse overall survival than those with higher levels. In our cohort we observed that cases belonging to cluster enriched in KRAS-mutated patients have a poor outcome. LUAD driven by KRAS mutation represents an unmet clinical need, being refractory to pharmacological inhibition. Our results link KRAS mutations to composition and in particular to B cell infiltration. The role of B cell in tumor microenvironment of lung cancer has been previously explored, demonstrating that low level of infiltration is related to short survival. Interestingly, we found that deregulated genes are enriched in GO/KEGG terms related to Th17, which, through CXCL13 signaling, support B cell recruitment. Thus, the present findings could be helpful in a better definition of immunotherapeutic approaches for KRAS mutated patients.