BACKGROUND: Circulating tumor cells (CTCs) are expected to be utilized for decision-making in precision cancer medicine and further elucidation of cancer biology. However, it remains challenging to detect and isolate CTCs without any damages, that is critically important to analyze the molecular and biological properties of CTCs. Near infrared (NIR) imaging is one of the emerging technologies in the medical field and expected to be applied to visualization of blood vessels, nerves, and muscles under the skin or mucosa. Here we attempted to apply the NIR imaging technologies to distinguishing tumor cells from normal blood cells.

METHODS: To obtain NIR images, real-time composition imaging system Compovision® (Sumitomo Electric Industries, Ltd.) was used. This system is equipped with an NIR spectrographic camera and this camera processes hyperspectral data, in which each pixel has spectral data instead of a visible image in the wavelength band of 1,000-2,350 nm. The X-axis and Y-axis of the sensor chip correspond to the spatial direction and wavelength. Thus, hyperspectral data stores both image and spectral information. For this feasibility study, lung cancer cell lines with varying EpCAM expression levels such as A549, PC-9, PC-14, H23 were used. Lymphocytes and red blood cells were obtained from the peripheral blood of healthy volunteers. Phosphate-buffered saline (PBS) was used as a negative control. After we obtained NIR hyperspectral data, multivariate analysis such as principal component analysis (PCA) was performed for precise analysis.

RESULTS: We were able to obtain the hyperspectral data from all the lung cancer cell lines, lymphocytes, red blood cells and PBS alone and the spatially averaged spectrum was extracted from each data. Then the spectra were converted to absorption spectra and analyzed by PCA using the Unscrambler® X (CAMO Software) in the wavelength range of 1,000-2,200 nm. Two principal components were selected from results to separate data. Two-dimensional PCA score plot showed the clear segregation between tumor cells and normal blood cells such as lymphocytes and red blood cells, indicating the potential of non-invasive detection of tumor cells. In addition, obtained spectra from both EpCAM-expressing and EpCAM-non-expressing tumor cells were successfully segregated from those from normal blood cells. According to these results, it is suggested that segregating cells based on the difference in cytoskeletal components between epithelial cells and blood cells is feasible using the NIR imaging technology.

CONCLUSIONS: Using the NIR composition imaging technology, we were able to obtain the different spectral characteristics between tumor cells and normal blood cells, suggesting that this technology holds the potential to detect and isolate CTCs without any invasive labeling. This should be further explored for its application to the identification of CTCs.

Citation Format: Yasuhiro Koh, Satoshi Kambayashi, Hiroaki Akamatsu, Nahomi Tokudome, Hiroki Ueda, Nobuyuki Yamamoto. Non-invasive identification of tumor cells using near infrared composition imaging system [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3797. doi:10.1158/1538-7445.AM2017-3797