Label-Free Morphological Discrimination between Different Cancer Cell Lines and Leukocytes by Digital Holographic Microscopy: a Model for Enumeration of Circulating Tumors Cells in Blood

Itay Barnea 1 Alex Barbul 1 Pinhas Girshovitz 2 Nathan Shaked 2 Rafi Korenstein 1
1Department of Physiology and Pharmacology, Tel Aviv University, Israel
2Department of Biomedical Engineering, Tel Aviv University, Israel

The isolation and enumeration of circulating tumor cells (CTCs) in blood are used to monitor metastatic disease progression and guide cancer therapy. The common accepted approach of enriching and enumeration of CTCs from the large cell population of leukocytes is based on immune-labeling of epithelial markers such as the epithelial cell adhesion molecule (EpCAM) and cytokeratins (CK), as well as the absence of the leukocyte marker, CD45. We suggest an alternative label-free methodology for discriminating between CTCs and leukocytes, based on label-free 2D and 3D morphological parameters acquired by Digital Holographic Microscopy (DHM). DHM is based on measuring an interference pattern composed of a superposition of the light field which has interacted with the cell and a mutually-coherent reference field. From the recorded complex field, it is possible to digitally reconstruct the quasi-three-dimensional distribution of the cell quantitative phase image, which is proportional both to the spatial cell thickness (with nanometer accuracy) and its refractive index. Thus DHM provides label-free quantitative information on a spectrum of parameters defining cell morphology and content.

Several types of metastatic cell-lines WM-266-4, MCF-7, SW-640, and A549, isolated from different organs (skin, breast, colon and lung, respectively), were used as cellular models for CTC. Three main subsets of leukocytes (monocytes, neutrophils and T cells) were isolated from venous human blood, using negative selection. About 50 cells of each cell type were imaged by DHM. The morphological parameters of the different cells were analyzed and compared using ANOVA. Results demonstrate statistically significant differences in several morphological parameters, consisting of the area, mass and surface roughness of the cells, demonstrating the ability of DHM to discriminate the model CTC types from leucocytes.









Powered by Eventact EMS