Abstract A35: Single cell genomic profiling of circulating tumor cells (CTCs) from metastatic colorectal cancer (mCRC) identify tumor heterogeneity and rare somatic driver alterations

• Published in: Molecular cancer therapeutics 2015-12, Vol.14 (12_Supplement_2), p.A35-A35
• Main Authors: Greene, Stephanie, Lee, Jerry, Landers, Mark, Sanga, Sandeep, Jendrisak, Adam, Graf, Ryon, Louw, Jessica, Werner, Shannon, Wang, Yipeng, Dittamore, Ryan, Marrinucci, Dena
• Format: Article
• Language: English
• ISSN: 1535-7163; 1538-8514
• DOI: 10.1158/1535-7163.TARG-15-A35

Background: Mostly asymptomatic until late stage, colorectal cancer is driven by the successive accumulation of genetic alterations resulting in genomic instability within subclonal tumor populations. mCRC often progresses as a subclonally diverse multifocal disease due to selective therapeutic pressure, the surrounding tumor microenvironment, and underlying genomic heterogeneity. Targeted therapies against EGFR, VEGF or BRAF have shown increased response in a subset of patients; however, patient stratification using standard population analysis of DNA markers from tumor biopsy, (i.e. chromosomal instability, microsatellite instability, promoter methylation, resistance mutations), is problematic due to tumor heterogeneity. CTCs reflect the active metastatic subclonal populations at any given time, making single cell analysis of DNA markers a more accurate, real-time picture of cumulative metastatic diversity. Using Epic's enrichment-free CTC analysis platform, we characterized individual CTCs from a mCRC patient to understand the extent of intra-patient genomic heterogeneity, including the presence genomic instability and point mutations. We compared the prevalence of clinically relevant subclonal alterations within patient CTCs to CRC TCGA data, offering insights into identification of therapeutic opportunities and potential mechanisms of resistance. Methods: Blood was collected from a heavily pretreated mCRC patient and was processed for CTC analysis using the Epic Platform. 34 CTCs were individually recovered, lysed, whole genome amplified, constructed into shotgun libraries and target enriched for all coding regions of 500 pan-cancer genes. Enriched libraries were sequenced to an average depth of 697X coverage by 2×150 PE sequencing. Sequences were aligned and somatic mutations were determined using VarScan with the patient's WBC as germline reference. Variants were filtered for functional gain- or loss-of-function mutations by SIFT/PolyPhen2 and selected based on low frequency in 1000g database. Genomic instability and loss of heterozygosity (LOH) was also assessed. Somatic variants deriving from the patient CTC cohort and TCGA CRC cohort of 302 patients were annotated, analyzed, and compared using GenePool™ software (Station X). Results: MLL3 alterations, frequently observed in primary CRC biopsies (14%), were identified in 70% of all CTCs sequenced. Previously cited somatic variants were detected in minor subclonal populations of CTCs, including APC