The present report is an interim analysis of part of our prospective study of CTCs as GFP+ cells in patients with treatment-naïve gastric adenocarcinoma who underwent surgery at the Digestive Disease Center of Showa University Northern Yokohama Hospital (Tsuzuki-ku, Japan) between April 2010 and May 2011. Peripheral blood samples (three samples per time point) were obtained from 37 patients before surgery and at 4 and 24 weeks after surgery. The inclusion criteria were as follows: i) Histologically confirmed adenocarcinoma of the stomach by endoscopic biopsy; ii) clinical solitary tumor; iii) no prior endoscopic resection, chemotherapy or radiotherapy; iv) age, 20–80 years; v) Eastern Cooperative Oncology Group performance status (28) of 0 or 1; vi) sufficient organ function; and vii) written informed consent. The exclusion criteria were as follows: i) Synchronous or metachronous malignancy; ii) pregnancy or breast-feeding; iii) active or chronic viral hepatitis; iv) active bacterial or fungal infection; v) diabetes mellitus; vi) systemic administration of corticosteroids; and vii) unstable hypertension. The pathological stage of the disease was determined according to the 7th edition of the American Joint Committee on Cancer/International Union Against Cancer tumor-node-metastasis classification system (29). The depth of the tumor invasion in 4 patients without gastrectomy and the regional lymph node status of 7 patients without sufficient lymphadenectomy were surgically diagnosed.

All patients were followed up at our hospital every 3 months after surgery. The patients underwent endoscopy and computed tomography at least once per year depending on their disease stage and course. Postoperative therapy was allowed if required.

Peripheral blood samples were collected from healthy volunteers to establish the cut-off size of GFP+ cells. All volunteers were employees of Sysmex Corporation (Kobe, Japan), and were interviewed individually prior to sample collection. The volunteer group consisted of 7 men (mean age, 31.4 years; range, 24–39 years) and 3 women (mean age, 33.7 years; range, 26–48 years). All volunteers underwent medical check-ups upon employment and annually thereafter. Check-ups included medical interviews, auscultation, chest radiography, and blood and urine analyses. None of the volunteers was receiving medical treatment, pregnant or breast-feeding, or had donated blood within the month previous to the initiation of the study.

The present study was approved by the Institutional Review Board of Showa University Northern Yokohama Hospital (Tsuzuki-ku, Japan; approval number 0903-03). The study protocol was explained to patients and volunteers prior to obtention of written informed consent. Our study was registered with the University Hospital Medical Information Network in Japan (Tokyo, Japan; registration number 000004026).

OBP-401, a telomerase-specific, replication-selective adenovirus variant in which the TERT promoter drives the expression of the viral EIA and EIB genes, and into which the GFP gene is integrated (22), was used. The sensitivity and specificity of the OBP-401 assay have been reported by Kim et al (30). Briefly, the assay was evaluated five times using MDA-MB-468 breast carcinoma cells as positive controls. The numbers of GFP+ cells in samples containing 1 MDA-MB-468 cell and 7.5 ml blood were 1, 1, 1, 2 and 3, respectively, while the numbers of GFP+ cells in samples containing 20 MDA-MB-468 cells and 7.5 ml blood were 15, 17, 19, 22 and 24, respectively. Viral samples were stored at −80°C.

Details of sample preparation and assay are described in our previous report (23). A 7.5-ml peripheral vein blood sample was obtained from each patient before surgery and at 4 and 24 weeks after surgery. Blood was collected in tubes containing citric acid, phosphoric acid and dextrose, and stored at 4°C. Samples were assayed within 48 h of collection. Samples were centrifuged for 5 min at 540 × g, and the plasma phase was removed. Cell pellets were washed four times with PBS (Sigma-Aldrich; Merck Millipore, Darmstadt, Germany) and twice with RPMI medium (Sigma-Aldrich; Merck Millipore). Upon suspension in RPMI medium, cells were infected with 4×10⁸ plaque-forming units of OBP-401 for 24 h at 37°C. Dead cells were stained with the red fluorescent reactive dye L23102 (Thermo Fisher Scientific, Inc., Waltham, MA, USA). Upon inactivation of OBP-401, cells were fixed with 2% paraformaldehyde for 20 min at room temperature and treated with a surface-active agent (Emalgen 2025 G; Kao Chemicals, Tokyo, Japan) for 10 min at 40°C to degrade red blood cells. Cells were incubated with phycoerythrin-labeled anti-human cluster of differentiation (CD) 45 antibody (catalogue number 368510; BioLegend, San Diego, CA, USA; 1:5 dilution) for 30 min at 25°C, which was diluted into PBS containing 2% fetal bovine serum (FBS; Sigma-Aldrich; Merck Millipore). Cells were washed with PBS containing 2% FBS and mounted on glass slides for microscopic analysis (two slides per sample).

The threshold for GFP fluorescence intensity was determined as previously reported (23). Briefly, blood samples (7.5 ml) from healthy volunteers were mixed with ~30,000 A549 (lung carcinoma), HepG2 (hepatocellular carcinoma), HEC-1 (endometrial carcinoma), KATO-III (gastric carcinoma), SBC-3 (small cell lung carcinoma), LNCaP (prostate adenocarcinoma), MDA-MB-468 (breast carcinoma) or OVCAR-3 (ovarian carcinoma) cells. The A549, HepG2, HEC-1, KATO-III, SBC-3 and MDA-MB-468 cell lines were obtained from the Health Science Research Resources Bank (Osaka, Japan), while the LNCap and OVCAR-3 cell lines were obtained from the RIKEN Cell Bank (Tokyo, Japan). Cell lines were cultured according to the vendor's specifications. The samples were assayed using OBP-401, and GFP+ cells were visualized by fluorescence microscopy and counted. The GFP signal intensity threshold was determined to be 2.85×10⁷ molecules of equivalent fluorochrome on the basis of the minimal GFP intensity level observed in the blood samples mixed with the cells. There was no significant difference in cell size prior and subsequent to OBP-401 infection.

In our previous study (23), the various sizes of GFP+ cells in each sample made it difficult to determine which GFP+ cells best represented CTCs. To establish a constant value, the optimum threshold derived from a receiver operating characteristic curve analysis based on cell size was used to define GFP+ CTCs; this threshold was 7.735 µm. Thus, two groups of GFP+ cells were analyzed: Small (S)-GFP+ cells, measuring <7.735 µM in diameter, and large (L)-GFP+ cells, measuring ≥7.735 µm in diameter (27).

Cell counting and analysis. In a blind analysis, GFP+ cells on slides were counted on a computer-controlled fluorescence microscope (IX71; Olympus Corporation, Tokyo, Japan). Cells with fluorescent emission counts of ≥2.85×10⁷ molecules of equivalent fluorochrome were counted as GFP+ cells. The criteria for a GFP+ cell did not include the presence of epithelial markers such as EpCAM or cytokeratin, since tumor cells undergoing EMT have been reported to lose epithelial markers (31). CD45+ cells were excluded from the analysis.

All statistical analyses were performed using JMP Pro 11.1.1 software (SAS Institute, Cary, NC, USA). The Kaplan-Meier analysis and the Wilcoxon test were used to calculate relapse-free survival rates and for non-parametric comparisons of the numbers of GFP+ cells. The relapse-free survival rate of the patients who underwent non-curative surgery declined to 0 months. P≤0.05 was considered to indicate a statistically significant difference.

The clinicopathological characteristics of 37 patients (26 men and 11 women; mean age, 60.7 years; range, 33–76 years) are summarized in Table I. The median follow-up period of the surviving patients was 39 months. A total of 33 patients underwent pathological curative surgery, and 5 of these patients experienced disease recurrence. In total, 6 patients succumbed to disease; 16 patients underwent distal gastrectomy; 18 patients underwent total gastrectomy; and 3 underwent exploratory laparotomy. Of the 37 patients, 17 received chemotherapy post-surgery; 13 patients received oral chemotherapy (S-1); and 4 received oral chemotherapy combined with infusion (S-1/cisplatin and S-1/docetaxel).

The association between the number of L-GFP+ cells and the relapse-free survival rate is shown in Fig. 1. Of the 37 patients enrolled in our study, 33 and 30 were relapse free at 4 and 24 weeks, respectively, after surgery. There were no significant differences in the relapse-free survival rates of patients with ≥6 L-GFP+ cells and patients with <6 L-GFP+ cells, either before (P=0.072; Fig. 1A) or at 4 (P=0.109; Fig. 1B) or 24 (P=0.342; Fig. 1C) weeks after surgery. At 4 weeks after surgery, the relapse-free survival rate was, however, relatively higher in patients with ≥6 GFP+ cells than in patients with <6 GFP+ cells.

The numbers of L-GFP+ cells in peripheral blood samples from patients without relapse following curative surgery (R-), patients with relapse following curative surgery (R+) and patients who underwent non-curative surgery (Non-C) were compared at three time points. The mean numbers of L-GFP+ cells at 0, 4 and 24 weeks after surgery were as follows: R-, 5.6, 8.3 and 6.1, respectively; R+, 6.0, 4.0 and 7.2, respectively; and Non-C, 6.8, 0.8 and 4.0, respectively (Fig. 2A). There was a significant difference between the number of L-GFP+ cells in samples from the R- patients and the Non-C patients at 4 weeks after surgery (P=0.034).

The mean numbers of S-GFP+ cells at 0, 4 and 24 weeks after surgery were as follows: R-, 8.7, 11.3 and 11.6, respectively; R+, 11.8, 10.4 and 13.0, respectively; and Non-C, 10.3, 10.8 and 20.0, respectively (Fig. 2B). There were no significant differences between the three groups either prior or subsequent to surgery. The number of S-GFP+ cells in samples from Non-C patients at 24 weeks after surgery was, however, relatively high.

The associations between the number of GFP+ cells, postoperative chemotherapy and outcome are represented in Fig. 3. Of the 37 patients, 6 and 17 were receiving postoperative chemotherapy 4 and 24 weeks after surgery. The number of L-GFP+ cells was similar prior and subsequent to surgery, regardless of chemotherapy (Fig. 3A). Although there was no significant difference, the number of S-GFP+ cells was relatively high in the samples from the patients who were receiving postoperative chemotherapy 24 weeks after surgery (Fig. 3B).