Validation of a single-cell sequencing assay for the detection of circulating tumor cells in metastatic breast cancer.

Abstract

1053 Background: Blood Biopsy and the evaluation of Circulating Tumor Cells (CTC) has provided valuable clinical information for patients with metastatic breast cancer, specifically where a tissue biopsy is not feasible. A critical problem in the field is that most assays cannot distinguish true CTCs from other types of rare cells. By combining established protein biomarkers associated with circulating epithelial cells with single-cell whole-genome sequencing to assess copy number variations (CNVs, a signature for genomic instability - GI), we differentiate rare cells from true CTCs. Additionally, we can determine the presence or absence of regions that contain gene amplifications (e.g., ERBB2/HER2). Here we report the validation of a clinical blood biopsy CTC assay that can enumerate CTCs, determine if those CTCs have GI and if selected CTCs are expressing HER2 or ER antigens. Methods: Relevant cell line controls of epithelial origin and known expression levels of ER or HER2 were spiked into healthy donor whole blood at varying concentrations. The nucleated cells were deposited on microscope glass slides at high density. Immunofluorescence staining of slides was performed to identify the presence and localization of nuclei (via DAPI), pan-CK, HER2, ER, CD45 and CD31. Nucleated cells were classified as CTC candidates if they were pan-cytokeratin(+) and CD45/CD31(-), and were further characterized for expression of ER and HER2 protein. Defined CTC candidates were imaged and located by specific coordinates allowing single-CTC isolation for whole genome amplification. Single-cell sequencing data was then analyzed with the Epic CTC DNA copy number analysis pipeline to detect GI and amplification at the ERBB2 locus. Results: The limit of detection for CTCs was determined to be 1 CTC in 6 million white blood cells with a linear range of 1- 300 CTC per slide. The sensitivity and specificity for HER2 detection was determined to be 94% and 97%, respectively. ER assay sensitivity and specificity were 91% and 100%, respectively. Overall accuracy was 100% for HER2 and 94% for ER detection. Overall precision across all relevant variables produced %CVs <10% for HER2 and ER. A genome possessed GI if more than 12 large-scale transitions (LSTs) were detected. For detection of 2-fold amplification at the ERBB2 locus, gene amplification was detected with 85% sensitivity and 94% specificity. Tested healthy donor blood samples were absent of detected CTCs and lacked LST/GI or ERBB2 amplification in all (46) isolated WBCs. Conclusions: These results reflect the validation of a novel clinical assay combining CTC detection with single-cell GI and amplification of the ERBB2 locus. This assay better discriminates CTC detection with the addition of an assessment of GI. Additionally, along with protein expression measures of ER and HER2, it provides the ability to assess if HER2 expression is driven by ERBB2 amplification.