Background: Tumor cells with amesenchymal phenotype and/or cancer stem-like cells (CSCs) are known to contribute to metastasis. Circulating tumor cells (CTCs) undergoing epithelial-mesenchymal transition (EMT) may not be detected using an anti-EpCAM antibody. We have developedan antibody-independent CTC enrichment platform, Apostream that does not rely on EpCAM-based capture, assessing morphologic and dielectric properties gated on CD45- cells. We used this instrument to determine the feasibility and clinical relevancy of measuring EMT-CTCs in breast cancer patients who received primary systemic chemotherapy (NST). Methods: Blood samples from newly diagnosed breast cancer patients were prospectively collected before NST (T0), after NST (T1), and after definitive surgery (T2) and processed using the Apostream system. Isolated cells were stained with antibodies to leukocytes (anti-CD45), and the DAPI nuclear stain, to identify CTCs. These CTCs were also stained with additional markers and examined on a laser scanning cytometer to measure protein expression levels of various markers to define 4 CTC-phenotypes: epithelial (CK+, EpCAM+, E-cadherin+), EMT (β-catenin,+ vimentin+), combined epithelial and EMT (CK+, EpCAM+, E-cadherin+, vimentin+), and CSC (CD44+, CD24low). Pathological complete response (pCR) was correlated to CTCs and marker expression. Results: Of the 15 patients enrolled, 5 patients achieved pCR. Epithelial-CTCs were detected in 60%, 67%, and 77% of the T0, T1, and T2 samples, respectively. EMT-CTCs were detected in 80%, 87%, and 85% of these samples, respectively. The mean number of CTCs with epithelial and EMT phenotype was 32.8 (range, 0-434) and 68.7 (range, 0-687) respectively. Number ofCTCs with epithelial phenotype and CSC phenotype were more likely to decrease, after NST, if N stage (nodes) > or equal 3 (P = 0.014) or T stage (primary tumor) > or equal 4 (P = 0.066), respectively. Neither EMT (P = 0.505) nor epithelial (P = 0.580) phenotypes of CTCs seem to predict pCR. Conclusions: Apostream was successful in detecting both epithelial and EMT-CTCs. We will present the final data analysis of 50 patients to conclude on their predictive impact.