The present study was a prospective study performed in patients with pancreatic cancer at the University of Tsukuba Hospital (Tsukuba, Japan). A total of 15 whole-blood specimens from patients with pancreatic cancer were collected between October 2021 and April 2022. The study protocols were approved by the Institutional Review Board of the University of Tsukuba Hospital (Tsukuba, Japan; approval no. H30-150) and all patients provided written informed consent.

From each patient, 5 ml whole blood was collected into an EDTA blood collection tube (Venoject II vacuum blood collection tube; Terumo) and aspirated with a 10-ml syringe (SS-10SZ; Terumo), and the syringe was attached to an automatic pump (KDS100; KD Scientific, Inc.). The following operations were all performed at a flow rate of 50 ml/h. A 13-mm diameter 15±0.5-µm pore filter (φ15-RM-P30d-t10; Optnics Precision) and a 8±0.5-µm pore filter (φ8-RM-P17d-t10; Optnics Precision) were respectively mounted on the filter holders (CSS-00352; Optnics Precision), and these two filter devices were connected in tandem to the syringe containing the sample. Blood samples were passed through the 15-µm pore filter and 8-µm pore filter, and cells were collected by each filter. Subsequently, the filter was washed by flushing with 10 ml PBS/2 mM EDTA. In cases 1 to 6, 2 ml of staining solution [PE-conjugated mouse anti-human epithelial cellular adhesion molecule (EpCAM; 1:200 dilution; IgG2b clone 9C4; cat. no. 324206; BioLegend) and Hoechst 33342 (1:2,000 dilution; cat. no. H3570; Invitrogen; Thermo Fisher Scientific, Inc.) in PBS], was added and incubated at room temperature for 30 min. Subsequently, the filter was washed by flushing with 10 ml PBS. Finally, the filter device was removed from the syringe and placed on the 35-mm dish (MS-10350; Sumitomo Bakelite Co., Ltd.) in all cases. The cells on the filter were cultured at 37°C in a humidified atmosphere containing 5% CO₂ in Iscove's Modified Dulbecco's Media (IMDM; cat. no. 098-06465; Fujifilm) 1:1 with Nutrient Mixture F-12 Ham (cat. no. N6658; Sigma-Aldrich; Merck KGaA), 20% FBS (cat. no. 35-079-CV; Corning, Inc.), Matrigel^® (cat. no. 354234; Corning, Inc.), recombinant mouse EGF (cat. no. 053-07751) and recombinant human basic fibroblast growth factor (FGF; cat. no. 064-04541; both from Fujifilm) and B27 (cat. no. 17504-044; Gibco; Thermo Fisher Scientific, Inc.). The culture medium was made to flow between the inside and outside of the filter device. The culture medium was exchanged by aspirating from the outside of the filter device and then adding the medium to the inside of the filter device (Fig. 1). After being cultured, the cells were washed 3 times with 200 µl PBS and 200 µl staining solution (PE-conjugated mouse anti-human EpCAM and Hoechst 33342, as specified above) was added and the mixture was incubated at room temperature for 30 min. For cases 14 and 15, 200 µl of staining solution (Calcein-AM; 1:2,000 dilution; cat. no. 341-07901; Dojindo; and PE-conjugated mouse anti-human EpCAM-PE; 1:200 dilution; cat. no. 324206; clone 9C4; Biolegend; in PBS) was added and cells were incubated at room temperature for 30 min. After the above-mentioned staining, the cells were washed 3 times with 200 µl PBS.

The collected filter was placed on a 35-mm dish. For observation, an automatic fluorescence microscope (MSZ25; Nikon Corporation) was used and an image of the entire area of the filter was acquired with a ×10 objective lens (excitation wavelengths of 395, 472 and 545 nm and emission wavelengths of 460, 520 and 615 nm), before automatically generating a single image file (software: Nikon NIS-Elements BR v.5.21; Nikon Corp.).

The clinicopathological characteristics of the patients are presented in Table I. The age range of the 15 cases included was 44–87 years with a median age of 74 years. The cohort comprised 6 males and 9 females. The histologic type was invasive ductal adenocarcinoma in 13 cases and intraductal papillary mucinous carcinoma in 2 cases. The blood was collected from all patients just before the surgery. Among the 15 patients, 4 patients with borderline resectable or unresectable locally advanced pancreatic cancer received pre-operative treatment, 2 patients received systemic chemotherapy and 14 underwent a course of gemcitabine/nab-paclitaxel and 2 cycles of gemcitabine (1,000 mg/m²) with S1 oral intake prior to surgery. A total of 2 patients received chemoradiotherapy, including 2 cycles of gemcitabine infusion with concurrent proton beam radiation (67.5 Gy) (Table I).

CTCs isolated on the precision microfiltration membrane with 8- and 15-µm pore size were cultured, and enumeration and evaluation of the state of the cancer cells on the filter was performed by fluorescent labeling in the real-time (approximate culture time: 0, 25, 45 and 100 days). CTCs captured on the filter were cultured using IMDM-F12 medium containing 20% FBS, −2% Matrigel, EGF, bFGF and B27. A CTC was defined as a single, intact round oval cell with a visible nucleus (Hoechst 33342-positive) that stained positive with anti-EpCAM (Fig. 2A and B). Clusters were defined as aggregates of two or more CTCs and <50 µm in the long diameter (Fig. 2C-E), and colonies were defined as aggregates of >50 µm in the long diameter (Fig. 2F-H). The numbers of single CTCs, clusters and colonies for cases 1–6, which were observed at three sequential time-points, are provided in Table II. In several cases, there were no CTCs immediately after filtration, but single cells, clusters and colonies were detected after cultivation in each case. The numbers of single CTCs, clusters and colonies for cases 7–13, after culturing for different durations, at one time-point, are presented in Table III.

The number and state of cancer cells on the filter were evaluated by staining with Calcein AM to confirm cultured CTC activity. The cells were observed after exposing them to Calcein AM in cases 14 and 15. The EpCAM-positive cells cultured on the filter stained with Calcein AM were considered viable CTCs (Table IV, Fig. 3).