P3.04-16 A Seven-Gene Expression Signature Reveals Unique Immune-Phenotypes Related to Major Oncogenic-Drivers in NSCLC

Abstract

In oncogenic-driven non-small cell lung cancer (NSCLC), programmed death ligand 1 (PD-L1) expression is the result of a constitutive oncogenic activation leading to an immunosuppressive microenvironment. However, the relationship between the major driver mutations (KRAS, EGFR and ALK) and PD-L1 and other immune markers remains unclear. Gene expression signatures incorporating not only PD-L1 but also other components of the stroma might better capture the immune-context of these oncogenic subgroups. A 7-gene ‘immune signature’ comprising CD4, CD8, PD-1, PD-L1, IFNG, GZMM and FOXP3 was included in a customized nCounter panel (NanoString Technologies), used in our clinical institution on a routine basis and designed to simultaneously screen for gene fusion drivers (ALK, ROS1, RET and NTRK1), MET overexpression and MET exon 14-skipping mutations in formalin-fixed paraffin embedded (FFPE) samples. A total of 296 advanced NSCLC patients from two different institutions were analyzed by the panel. Among them, 115 patients (38.9%) were also submitted to next-generation sequencing (NGS, Ion Torrent PGM® or GeneReader) . Analyses of variance (ANOVA) were used to describe statistical significance between immune response genes in the two major oncogenic groups: KRAS mutant (n=33) andALK rearranged (n=44), compared to wild-type (WT) tumor samples (n=38). Oncogenic genes (ALK, KRAS) were mutually exclusive. The analysis of the 7-gene signature revealed distinct expression patterns in the oncogenic biomarker groups. A significantly higher mRNA expression of CD4 and PD-L1 was found in ALK rearranged tumors compared to the KRAS mutant, and WT groups (p=0.0014 and p=0.0467, respectively). In addition, a trend was observed between GZMM mRNA levels and the oncogenic groups (p= 0.0665) whereas no association was found with the other immune genes (CD8, PD-1, IFNG, FOXP3). There was a significant linear correlation between CD4 and PD-L1 in ALK positive patients (p= 0.0214), but not in KRAS mutant samples (p= 0. 112). Unsupervised clustering across mRNA expression data from 296 samples using 7-immune-related genes showed two clusters, high expression for ALK-rearranged patients and low expression for KRAS mutant patients. The correlation between each of the immune genes was performed and a high correlation was found between PD-1 and FOXP3 (r=0.9) and PD-1 with GZMM (r=0.8). NSCLC tumors with ALK alterations show a distinct CD4 and PD-L1 immune profile when compared to KRAS and WT samples.