Detection of circulating tumor cells by p75NTR expression in patients with esophageal cancer

Background The p75 neurotrophin receptor (p75NTR) is a cancer stem cell (CSC) marker in esophageal squamous cell carcinoma (ESCC). This study aimed to assess the use of p75NTR in detecting circulating tumor cells (CTCs) in ESCC. Methods Peripheral blood mononuclear cell expression of epithelial cell adhesion molecule (EpCAM) and p75NTR was detected in 23 ESCC patients (13 received chemo- or chemoradiotherapy and 10 received curative surgery) and 10 healthy controls by flow cytometry. Results EpCAM + p75NTR+ cell counts (average ± SD) were significantly higher in patients (n = 23, 16.0 ± 18.3) compared to controls (n = 10, 0.4 ± 0.9, p = 0.013). The sensitivity and specificity to differentiate ESCC patients from controls were 78.3 and 100 % (cut-off value 4.0), respectively. EpCAM + p75NTR+, but not EpCAM + p75NTR− cell counts, correlated with clinically diagnosed distant metastasis (n = 13, p = 0.006) and pathological venous invasion in resected primary tumors (n = 10, p = 0.016). Malignant cytology was microscopically confirmed in isolated EpCAM + p75NTR+ cells with immunocytochemical double staining. Conclusions p75NTR is suggested to be a useful marker for clinically significant CTCs, which exhibit highly metastatic features in ESCC.


Background
Esophageal squamous cell carcinoma (ESCC) is a leading cause of cancer-related deaths worldwide [1]. Despite advances in its diagnosis and multimodal therapies, the prognosis for ESCC patients remains poor with postoperative 5-year survival rates of 42 %, due to high incidences of metastasis [2,3].
Lately, attention has focused on the proportion of circulating tumor cells (CTCs) as an early marker of metastasis [4]. The CTC detection method most extensively used is based on immunomagnetic enrichment with epithelial cell adhesion molecule (EpCAM) and cytokeratin antibodies for subsequent immunological identification [5,6]. Quantification of CTCs using this method has established CTCs as an independent prognostic factor in patients with advanced colorectal [7], breast [8], and prostate cancer [9].
Other studies have identified cancer stem cells (CSCs), which are defined as a small fraction of tumor cells capable of self-renewal, tumorigenicity, and drug resistance [10,11]. CSCs are also responsible for metastasis through processes such as epithelial mesenchymal transition, invasion into vessels, circulation, and tumor initiation [12,13].
Moreover, reports suggest that a percentage of CTCs have characteristics reminiscent of CSCs, which are termed circulating tumor stem cells (CTSCs) [14]. Compared to CTCs, CTSCs are potentially a more accurate prognostic factor since cancer growth is dependent on CSCs [15].
In ESCC, several cell surface antigens have been reported to be CSC markers, such as p75 neurotrophin receptor (p75NTR) [16][17][18] and CD44 [19]. A recent report demonstrated that p75NTR-positive cells in ESCC cell lines showed significantly higher colony formation, enhanced tumor formation in mice, and greater chemoresistance, along with stronger expression of epithelial mesenchymal transition (EMT)-related genes [18]. However, detection of CTCs using CSC markers in ESCC has not previously been reported. The aim of this study was to detect CTCs in the peripheral blood of ESCC patients using combined expression of EpCAM, CD44, and p75NTR to assess its clinical significance.

Patients
This study included a series of 23 primary ESCC patients who consulted our hospital between January and December 2014. Nineteen of 23 patients were men, and the median age was 62.3 (range 38-88) years. Ten patients underwent R0 resection, while 13 patients received chemotherapy (CT) or chemoradiation therapy (CRT) without resection. Clinicopathological classifications were determined according to the criteria of the Union for International Cancer Control (UICC)-TNM Classification of Malignant Tumors Staging System (seventh edition). In the patients who underwent CT or CRT, 1, 4, 2, 1, and 5 patients were clinically diagnosed as cStageIB, cStageIIIA, cStageIIIB, cStageIIIC, and cStageIV (intramural metastasis to the stomach 3; bone 1; distant lymph node 1), respectively. In the patients who underwent surgery, 2, 5, 2, and 1 patients were pathologically diagnosed as stage IA, stage IIA, stage IIIA, and stage IV (T3N2M1: intramural metastasis to the stomach), respectively. Three patients presented with lymph node metastasis and lymph vessel invasion, and three presented with venous invasion.
Ten healthy volunteers were also recruited as negative controls. Subjects provided informed written consent prior to commencing the study, and all were enrolled following Institutional Review Board approval (no. 20-75).

Flow cytometric detection of cultured cells mixed with blood samples
Human ESCC cell lines (KYSE30 and KYSE140) were established by Shimada and colleagues and cultured in Ham's F12/RPMI-1640 (Wako, Osaka, Japan) with 2 % fetal calf serum (Gibco; Grad Island, NY, USA) according to a previously published method [20]. For the accuracy assay, 0, 10, 100, and 500 of KYSE cells were mixed with 3 mL each of peripheral blood samples obtained from the healthy volunteers.

Cytological examination of the sorted cells
Sorted EpCAM + p75NTR+ cells were centrifuged at 250× g for 5 min using a Cytospin™ chamber (Stat Spin, Beckman Coulter, Tokyo, Japan) and evaluated using a transmitted light microscope (BX61/DP70, Olympus, Tokyo, Japan).

Culture of the sorted cells
Sorted EpCAM + p75NTR+ cells were resuspended with 100 μl of Ham's F12/RPMI-1640 with 2 % fetal calf serum and seeded onto one of the 96-well cell culture plates and cultured at 37°C in a humidified atmosphere of 5 % carbon dioxide in air for 14 days.

Statistical analysis
Statistical analyses were performed using JMP v.11 (SAS Institute Inc., Cary, NC, USA). Chi-square and Fisher's exact tests were performed, and probability (p) values of less than 0.05 were deemed statistically significant. Receiver operating characteristic (ROC) curves and the area under the curve (AUC) were used to assess the feasibility of using EpCAM+ cells and EpCAM+/p75NTR+ cell counts as a measure of CTC numbers in patients with ESCC.

Accuracy assay for detecting CTCs
EpCAM expression was detected in 100 and 100 % of KYSE30 and KYSE140, respectively. Within the EpCAM+ cells, 14.6 and 35.5 % of KYSE30 and KYSE140 cells, respectively, were p75NTR+ (Fig. 1a). EpCAM+ cells were detected in the peripheral blood of healthy volunteers, in which 0, 10, 100, and 500 of KYSE30 and KYSE140 cells were mixed beforehand (Fig. 1b). Regression analysis on the numbers of observed EpCAM+ versus the expected  (Fig. 1c).

Detection of CTCs in ESCC patients based on EpCAM expression
EpCAM+ cells were detected in 5 of 10 (50.0 %) healthy controls and 22 of 23 (95.7 %) ESCC patients with cell counts (average ± SD) of 2.3 ± 2.5 and 34.0 ± 35.8, respectively (p = 0.011, Fig. 2). Based on the ROC curve analysis to differentiate ESCC patients from controls, the largest AUC value for EpCAM+ cell counts was 7.0 with optimal sensitivity and specificity measurements of 86.9 and 100 %, respectively.
The proportion of EpCAM + p75NTR+ cells in EpCAM+ cells (average ± SD) was 56.7 ± 39.6 % (range 5.6-100.0 %). Based on the ROC curve analysis to differentiate ESCC patients from controls, the largest AUC value for EpCAM + p75NTR+ cell counts was 4.0 with optimal sensitivity and specificity measurements of 78.3 and 100 %, respectively. The relationship between CTC counts and various clinicopathological features in the 13 patients who underwent CT or CRT is presented in Table 1. The mean EpCAM + p75NTR+ cell counts correlated with clinically diagnosed distant metastasis (p = 0.003), while EpCAM + p75NTR− cell counts did not show relationship with any of these pathological factors. The relationships between CTC counts and various clinicopathological factors in the ten patients who underwent R0 resection are presented in Table 2. The mean EpCAM + p75NTR+ cell counts correlated with pathological venous invasion (p = 0.016), while EpCAM + p75NTR− cell counts did not show relationship with any of the pathological factors.

Cytology of sorted EpCAM + p75NTR+ cells
We evaluated sorted EpCAM + p75NTR+ cells from the peripheral blood of an ESCC patient microscopically. A representative figure of cells staining positive for EpCAM-APC (red) and p75NTR-FITC (green), in the cell membranes, is provided in Fig. 4. The cell represented is a mononuclear cell, 35 μm in diameter with a high nucleocytoplasmic ratio, which led cytopathologists to establish a diagnosis of malignant cytology.
We could not establish primary culture of the EpCAM + p75NTR+ cells sorted from any of the 23 examined patients (data not shown).

Expression of p75NTR in the primary tumor of resected ESCC specimens
The expression of p75NTR was detected in five (50.0 %) tumors, in which the first few layers nearest to the tumor's infiltrative margin exhibited positive staining (Fig. 5). There was no relationship between p75NTR

Discussion
In this study, we detected CTCs by flow cytometry based on the combined expression of EpCAM and p75NTR in patients with ESCC. Our results demonstrated that the EpCAM+ cell count in the peripheral blood of patients was significantly higher than that of controls, reflecting a tumor-bearing state, and thus indicating successful detection of CTCs using this method. Our results were consistent with those in previously published reports of patients with metastatic lung cancer (CK + CD45− CTCs detected by flow cytometry in all examined patients, with significantly higher cell counts than controls) [21,22]. We revealed that detecting a combination of EpCAM and p75NTR expression was more accurate as a diagnostic marker than EpCAM detection alone.
In addition, our results highlighted that EpCAM + p75NTR+, but not EpCAM + p75NTR− CTC counts, Recent reports have identified CTSCs, which have CSC-like potential and enhanced metastatic capacity in various solid tumors, such as breast [23], colon [24], and hepatocellular carcinoma [25], using a combination of EpCAM and CSC markers in the each tumor (e.g., CD44 or CD133). Further reports have also demonstrated CTSCs to be useful markers for the diagnosis, treatment responsiveness, and prognosis of patients with breast [26] and gastric cancer [27]. To the best of our knowledge, this is the first report to detect CTCs expressing a CSC marker in ESCC patients using flow cytometry.
In this study, we also detected CTCs with combined expression of EpCAM and CD44, which is one of the other proposed CSC markers in ESCC [19]; however, almost all the EpCAM+ cells were CD44+, indicating that CD44 is almost equivalent to EpCAM as a CTC marker and therefore it is not useful to differentiate subpopulations of CTCs in ESCC.
The most widely studied CTC detection method is an immunomagnetic-based enrichment, requiring fixation of the cells, using EpCAM and cytokeratin antibodies for subsequent immunological identification [5,6]. Previous studies on the detection of CTCs in ESCC have also used a CELLSEARCH system, which allows immunomagnetic capture of fixed EpCAM+ cells in combination with cytokeratin and CD45 immunofluorescence staining [28,29]. In contrast, we used flow cytometry to detect CTCs, which enabled us to analyze the expression of multiple cell surface markers in viable cells. We sorted an EpCAM + p75NTR+ CTC and cytologically identified an atypical cell as well.
However, we could not obtain primary culture of sorted EpCAM + p75NTR+ CTC from any of the 23 examined patients and thus could not assess the biological properties of the cells. A recent report from our laboratory demonstrated that 100 and 3000 cells were minimally required to establish colonies in vitro and mouse xenograft tumors, respectively, using p75NTR+ cells isolated from cultured KYSEs by flow cytometric cell sorting, suggesting decreased cell survival and recovery due to cell damage during isolation procedure [18].
Another possibility is that a small part (1/3000-1/100) of p75NTR+ cells are true CSCs which can establish colonies or xenograft tumors. In addition, a previous report demonstrated that even using resected primary tumor, only 21 esophageal cancer cell lines (KYSEs) were   [20]. Taken together, improvement of devices, procedure, and culture condition, which can enrich and recover target cells without causing cell damage or cell death, is needed to establish CTC isolation with higher yield and higher viability. Thereafter, characterization of the isolated EpCAM + p75NTR+ CTCs may provide us with a better understanding of their biological features in ESCC. A proportion of EpCAM+ cells was detectable, even in controls, although the cell counts were significantly lower than in ESCC patients. The presence of these cells could possibly be explained by a nonspecific immunological reaction or contamination of skin cells [30]. In addition, a previous study in mice demonstrated the presence of EpCAM+ bone marrow-derived mesenchymal stem cells with expression of epithelial phenotypic markers [31], suggesting that detection of EpCAM+ non-cancer cells is possible in the peripheral blood. Again, characterization of isolated EpCAM+ cells from the peripheral blood of healthy controls may also provide us with a better understanding of the biological significance of EpCAM+ cells.
The presence of EpCAM + p75NTR+ CTCs did not correlate with p75NTR expression in the primary ESCC tumors, demonstrating the heterogeneity of cell surface marker expression, as has been shown in various other types of tumors [32]. In addition, regulation of p75NTR expression by various environmental stimuli, including tissue injury and pro-inflammatory cytokines, has been reported [33]. This may suggest possible induction of p75NTR expression after mobilization of the metastatic cancer cells into systemic circulation.

Conclusions
Our flow cytometric detection of CTCs in patients with ESCC has highlighted the clinical significance of EpCAM + p75NTR+ CTCs in ESCC, providing us with a basis from which to establish large-scale prospective studies. Further investigations based on the isolation of viable EpCAM + p75NTR+ CTCs may shed light on the molecular and biological roles of these cells in tumor metastasis in ESCC as well.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests.