Abstract 1561: Increased phenotypic depth for automated visual identification of biomarkers on circulating tumor cells by cocktailing epithelial markers EpCAM and cytokeratin

Abstract

Abstract Background: Canonical methods for CTC identification (such as CellSearch®) have typically relied on co-expression of surface (EpCAM) and cytoplasmic (cytokeratin, CK) epithelial markers to separate them from white blood cells. RareCyte has developed AccuCyte®-CyteFinder®, a platform for automated visual identification and retrieval of rare cells in blood by immunofluorescence (IF). For CTC detection, EpCAM and CK are placed in individual fluorescence channels, allowing for identification of CTCs that may be EpCAM(-) /CK(+) as well as EpCAM(+)/CK(-). Recently we have developed technology that increases the number of markers from 4 to 6 to allow for biomarker phenotypic analysis in the additional two channels. Given that our platform does not require co-expression of EpCAM and CK, we hypothesized that we could gain a third biomarker channel by combining detection reagents for EpCAM and CK into a cocktail within a single channel. We tested feasibility of epithelial marker cocktailing with an assay to detect androgen receptor (AR) in prostate CTCs. Methods: Normal human whole blood samples were spiked with prostate cancer lines PC3, LNCaP, and 22RV1 as model CTCs (mCTCs). Blood samples from University of Washington patients with advanced cancer were collected under an IRB-approved protocol. Blood was processed onto microscope slides and stained on an automated stainer using a CTC assay for detection of AR in which anti-cytokeratin and anti-EpCAM antibodies were combined in a reagent cocktail into a single channel; the CD45 WBC marker and a nuclear dye were included in the assay. After development using spike-in models, the assay was applied to a patient sample. Results: No CTCs were identified in non-spiked normal donor blood samples. In spike-in models, the epithelial cocktail assay performed as well as the non-cocktail assay in identifying CTCs. AR fluorescence signal was present in LNCaP and 22RV1 mCTCs (which express AR) and absent in PC3 mCTCs (which do not express AR), confirming specificity of AR staining. When the assay was applied to a blood sample from a prostate cancer patient with known high CTC count, 160 CTCs per mL were identified; of these, 114 were observed to be AR-positive (71%). Conclusions: Combining the epithelial markers cytokeratin and EpCAM into a single-wavelength cocktail is feasible for IF identification of CTCs, allowing an additional channel for phenotypic investigation. In a patient with advanced prostate cancer with high CTC count, there was high incidence (71%) of AR expression in identified CTCs. Citation Format: Yao Sun, Arturo Ramirez, Daniel Campton, Tanisha Mojica, Alisa Clein, Celestia Higano, Daniel Sabath, Eric Kaldjian. Increased phenotypic depth for automated visual identification of biomarkers on circulating tumor cells by cocktailing epithelial markers EpCAM and cytokeratin [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1561.