Characterization of the tumor transcriptomic differences by smoking history in locally advanced or metastatic urothelial carcinoma (UC).

Abstract

561 Background: Understanding the influences of cigarette smoke exposure at the gene expression level may have clinical significance and therapeutic implications. Investigating variations in the gene expression patterns could provide insights into differences in disease progression, survival, and therapeutic response based on smoking history amongst smokers vs. non-smokers (PMID: 29050337). Our objective was to correlate smoking with the gene expression of pts with locally advanced or metastatic UC. Methods: In this IRB-approved retrospective study, eligibility criteria included pts with locally advanced or metastatic UC with available tumor-based comprehensive transcriptomic profiling from a CLIA-certified lab. DeSeq2 analysis was implemented in Bioconductor software to analyze differentially expressed genes based on smoking history. DeSeq2 results included the Log2 Fold change, Wald-Test p-values, and Benjamini-Hochberg adjusted p-values for each differentially expressed gene. These results were then subjected to Gene Set Enrichment Analysis (GSEA) to identify pathways upregulated or downregulated in each cohort. All bioinformatic analysis was conducted in R-Studio, version 4.1.1. Statistical significance was predetermined at p < 0.05. Results: 38 pts were eligible and included: Smokers (n=18) and non-smokers (n=20). The median age at diagnosis for smokers vs. non-smokers (67 vs.71 years). The three most upregulated significant pathways amongst smokers (vs. non-smokers) were the MYC, E2F target, and cholesterol homeostasis pathways (q for all <0.01) [Table]. The three most downregulated significant pathways were interferon-alpha, interferon-gamma, and inflammatory response pathways (q for all <0.01). Differential individual gene expression profiles will be presented at the meeting. Conclusions: Pts with advanced UC with smoking history demonstrated a different transcriptomic profile than those without smoking. Increased expression of proto-oncogene MYC and E2F could explain an aggressive disease amongst smokers. Future trials could utilize these transcriptomic differences by smoking history in guiding drug development. These hypothesis-generating results, upon external validation, may provide the rationale for personalized therapy in pts with UC. [Table: see text]