Novel CTC Detection Method in Patients with Pancreatic Cancer Using High-Resolution Image Scanning

Abstract

Background/objectives: Appropriate biomarkers are necessary for early diagnosis and multidisciplinary treatment of pancreatic ductal adenocarcinoma (PDAC). In recent years, the clinical utility of circulating tumor cells (CTC) as biomarkers for various can-cers has been reported; however, their detection rate in PDAC remains low, and clinical evidence is not yet established. CTC detection methods with high reliability and per-formance are essential for clarifying the importance of CTC in patients with PDAC.

Methods: A total of 5 mL peripheral blood samples were collected from 38 patients newly diagnosed with PDAC and 17 healthy controls. Negatively enriched cells were immunofluorescently stained with EpCAM-phycoerythrin and cell surface vi-mentin-fluorescein isothiocyanate (CSV). Images were automatically captured using an all-in-one fluorescence microscope. Cellular regions were detected from these images, and the average luminance of the cellular regions was calculated. A total of 9086 and 1071 cell images were obtained from patients with PDAC and healthy controls, respec-tively.

Results: In the EpCAM assay, a threshold that included 95% of healthy individuals was optimal for distinguishing patients with PDAC from healthy controls, with a sensi-tivity, specificity, and area under the curve of 0.74, 0.76, and 0.84, respectively. At this threshold, the CTC-positivity rate in patients with PDAC was 76.3%. Conversely, the CSV assay failed to demonstrate a valid threshold to distinguish patients with PDAC from healthy controls. No significant differences were found between CTC and clini-copathological features among patients with PDAC.

Conclusions: The method using high-resolution image scanning has the potential to identify CTC with greater objectiv-ity by quantifying cell luminance values.

Keywords: circulating tumor cells; high-resolution image scanning; liquid biopsy; negative enrichment; pancreatic ductal adenocarcinoma.

Figure 1

Schematic of overall design. (a) Flowchart of experimental procedures. (b) Image capture and pre-processing. Images are automatically captured with an all-in-one fluorescence microscope (BZ-X800) using a objective with a Z step size of 5 μm. (c) Cut-out images of CTC candidate cells (scale bar: 20 μm). Red fluorescence indicates EpCAM staining, green fluorescence indicates cell-surface vimentin (CSV), and blue fluorescence indicates nuclear staining with DAPI. CTC, circulating tumor cells.

Figure 2

Distribution of luminance values in the test set. (a) Luminance values in EpCAM assay (1~38: patients with PDAC, A~G: Healthy controls). (b) Luminance values in CSV assay (1~38: patients with PDAC, A~G: Healthy controls). CSV, cell surface vimentin; PDAC, pancreatic ductal adenocarcinoma.

Figure 3

Number of candidate CTC (a) Number of candidate CTC in EpCAM (99). (b) Number of candidate CTC in EpCAM (95). (c) Number of candidate CTC in EpCAM (90). (d) Number of candidate CTC in CSV (99). (e) Number of candidate CTC in CSV (95). (f) Number of candidate CTC in CSV (90). CTC, circulating tumor cells; CSV, cell surface vimentin. Colored dots represent individual data points from each sample (blue: healthy controls; red: patients with PDAC). Statistical significance was evaluated using the Mann–Whitney U test. Data are presented as median (interquartile range). The box represents the interquartile range, the horizontal line indicates the median, the whiskers denote the minimum and maximum values within 1.5 × IQR, and outliers are shown as individual points.

Figure 4

Receiver operating characteristic (ROC) curves comparing the diagnostic performance of EpCAM and CSV assays for distinguishing pancreatic ductal adenocarcinoma (PDAC) patients from healthy controls.

Figure 5

Association between the number of CTC and clinicopathological data in patients with PDAC. (a) Early stage (stage I–II) and advanced stage (stage III–IV). (b) CY positivity. (c) Tumor size. Statistical differences were assessed using the Mann–Whitney U test or Fisher’s exact test, as appropriate. Data are presented as median (interquartile range). The box represents the interquartile range, the horizontal line indicates the median, the whiskers denote the minimum and maximum values within 1.5 × IQR, and outliers are shown as individual points.