Single cell RNA-sequence analysis to identify transcriptomic differences associated with treatment outcome and ethnicity in circulating tumor cells (CTCs) from patients (pts) with metastatic colorectal cancer (mCRC).

Abstract

3041 Background: CTCs are a promising diagnostic and prognostic biomarker in colorectal cancer. The analysis of CTCs at the single-cell level may allow to capture tumor heterogeneity and identify novel prognostic and predictive markers in mCRC. Our aim was to evaluate whether the transcriptome profile of CTCs was associated with progression-free survival (PFS) in pts receiving treatment for mCRC and whether we could identify significant differences based on pts’ ethnicity. Methods: Single CTCs were prospectively collected from 25 pts undergoing treatment for mCRC at the USC/Norris Comprehensive Cancer Center between April and October 2019. Pre-treatment peripheral blood was processed using CTC-FIND comprised filter separation and immunomagnetic depletion of CD45/50-expressing cells to collect ultra-pure CTCs samples. After isolation to single cell, CTCs were processed by single cell RNA-sequence (scRNAseq). Principal component analysis was used for clustering expression data. DESeq2 was used to identify differentially expressed genes between pts with short ( < 150 days) and long term ( > 150 days) PFS, as well as between Hispanic and non-Hispanic ethnicity with control for FDR ( Q < 0.1). Ingenuity pathway analyses (IPA) of enriched pathway networks were performed. Results: Main pts characteristics were as follow: median age = 52 years; M/F = 18/7; Hispanics/non-Hispanics = 7/18; treatment line (n pts) = 1st (5), 2nd (11), ≥ 3rd (9). We identified 59 single CTC and CTC cluster from 22/25 pts (range 1-9, median = 2/pt). scRNAseq analysis identified two separate clusters of pts based on PFS (short term vs long term). Hispanic pts were mainly distributed within the short term PFS cluster. The IPA of the network of top 40 enriched pathways showed several pathways related with metabolism, such as Sirtuin signaling, mitochondrial dysfunction and oxidative phosphorylation, and IL6/JAK/STAT signaling pathways in pts with short term PFS ( Q < 0.05). When comparing Hispanic vs non-Hispanic pts, we detected enrichment of neuroinflammation signaling pathway in Hispanics, including CXCL8 (fold change, FC: 11.04), GAD2 (FC: 4.77), IRAK3 (FC: 17.06), PLCG2 (FC: 4.65) and TYROBP (FC: 8.43) (all Q < 0.05). Conclusions: In our study CTCs transcriptome profiles showed an association with PFS in pts receiving treatment for mCRC, with an enrichment in mitochondria-related pathways and IL6/JAK/STAT signaling in the CTCs of pts with shorter PFS. Additionally, CTCs scRNAseq identified differentially expressed genes in Hispanic pts, displaying enrichment in neuroinflammation signaling. These results highlight the potential of CTCs molecular profiling as a tool to identify novel prognostic and predictive biomarkers in mCRC and actionable molecular pathways which may impact tumor spread and treatment response