This study included gastric adenocarcinoma patients who underwent gastrectomy at Tottori University Hospital (Yonago, Japan). The patients who had preoperative treatment, such as radiotherapy, chemotherapy, or other medical interventions, were excluded. Adjuvant S-1 was performed in 34 patients who had stage II or III GC. The Japanese Classification of Gastric Cancer was used to determine the clinicopathologic findings . This study was approved by the Institutional Review Board at Tottori University Hospital (18A108).
Isolation of tumor-infiltrating mononuclear cells
Tumor-infiltrating mononuclear cells were isolated form 15 GC patients who underwent gastrectomy as previously described . In brief, fresh cancer tissues and non-cancerous gastric mucosa (at least 5 cm apart from the tumor in the resected specimen) were cut into small pieces with a size of approximately 1mm, and digested with 0.002% DNase I, 0.08% collagenase IV, and 0.01% hyaluronidase (all from Worthington, Lakewood, NJ, USA) at 37°C for 60 min. After filtering through 70-μm cell strainers (BD Falcon, Franklin Lakes, USA), density-gradient centrifugation using Ficoll-Paque (Pharmacia, Uppsala, Sweden) was performed to obtain the mononuclear cells.
Flow cytometry analysis
The antibodies used in this study are follows: anti-PD-1-phycoerythrin (PE) (Biolegend, San Diego, USA), anti-PD-1-peridinin-chlorophyll-protein complex (PerCP) (Biolegend), anti-CD45-PE-Cyanin 5 (PE-Cy5) (BD PharMingen, San Jose, USA), anti-CD11b-fluorescein isothiocyanate (FITC) (BD PharMingen), anti-CD11b-Allophecocyanin (APC) (BD PharMingen), anti-CD11c-APC (BD PharMingen), and anti-CD206-APC (BD PharMingen). The BD LSRFortessa™ cell analyzer (BD Biosciences, San Jose, CA, USA) was used for the analysis.
CD11b-positive cells were isolated from mononuclear cells obtained from GC tissue using a Magnetic Cell Sorting System (Milteny Biotec, Bergisch Gladbach, Germany). Cells were resuspended into RPMI 1640 (Thermo Fisher Scientific, Tokyo, Japan) in 96 well plate (Corning, NY, USA) and incubated at 37°C with Texas red conjugated Zymosan A (FUJIFILM, Tokyo, Japan) for 4 hours. After washing with phosphate buffered salts (PBS; FUJIFILM), cells were stained with anti-CD11b-FITC, anti-PD-1-PerCP, and DAPI (Cell Biolabs, San Diego, CA, USA). The numbers of PD-1+ macrophages that phagocytosed Zymosan A were determined by flow cytometry analysis.
Immunohistochemistry was carried out using samples from 102 patients with stage I–III gastric adenocarcinoma as previously described . Four µm-thick paraffin sections were dewaxed, deparaffinized in xylene, and rehydrated through a graded alcohol series. The sections were boiled for 20 min in a microwave oven in 10 mM citrate buffer (pH 6.0) to retrieve PD-1 and CD68 antigen. The slides were subsequently incubated with rabbit anti-PD-1 antibody (Clone EPR4877(2), Abcam plc, Cambridge, UK; 1:500 dilution) and mouse anti-CD68 antibody (Clone PG-M1, Dako, Santa Clara, CA, USA; 1:100 dilution) overnight at 4°C. The slides were then incubated with the conjugated goat anti-mouse polymer horseradish peroxidase (HRP) and the conjugated goat anti-rabbit polymer alkaline phosphatase (AP) secondary antibodies
(MACH 2 double stain®; Biocare Medical, Pacheco, CA, USA) for 30 min. Staining was visualized with peroxidase substrate (ImmPACT® DAB; Vector Laboratories, Burlingame, CA) and AP substrate (ImmPACT® Vector® Red; Vector Laboratories), which were visible as brown and red, respectively. The counterstain was then performed using Mayer’s hematoxylin solution (FUJIFILM).
Images of 3 fields (×200), which were randomly selected in a blinded manner, were acquired using a Nikon Eclipse Ts2 microscope (Nikon Instech, Tokyo, Japan). The separation of stains was achieved using the color deconvolution plug in of ImageJ software 1.47 (National Institutes of Health, USA) . Using the cell counter plug in of ImageJ software, the number of stained cells was determined for each image. The frequency of PD-1+ macrophages was represented by the ratio of the number of PD-1+ CD68+ cells to that of CD68+ cells.
Immunofluorescence staining for PD-1 and CD68 was performed as previously described . Four µm thick paraffin-embedded sections were incubated with primary antibodies, which were the same antibodies used in immunohistochemistry, overnight at 4°C. The slides were then incubated with Goat Anti-mouse IgG H＆L (Alexa Fluor® 488) and Goat Anti-rabbit IgG H＆L (Alexa Fluor® 647) (Abcam plc., Cambridge, UK; 1:500 dilution) for 30 min at room temperature. After washing with PBS, slides were mounted with ProLong Gold antifade reagent with 4,6-diamidino-2-phenylindole (Thermo Fisher Scientific) and examined using a Nikon Eclipse Ts2 microscope (Nikon Instech).
The differences between the frequency of PD1+ macrophages in GC tissue and that in non-cancerous gastric tissue were compared by paired t-test. The differences of clinicopathologic characteristics between two groups were compared by Mann-Whitney U test. Receiver operating characteristic (ROC) analysis was used to determine the Youden index. The frequency of PD1+ macrophages with the Youden index was used as an optimal cutoff value. Survival rates were calculated using the Kaplan-Meier method and their differences were determined using the log-rank test. Cox’s proportional hazards model was used to perform univariate analyses. Cox’s proportional hazards model and a stepwise procedure were used for multivariate analyses. A value of P < 0.05 was considered statistically significant. SPSS statistics version 24 (SPSS Inc., Chicago, IL, USA) and GraphPad Prism version 6 (GraphPad Software, Inc., La Jolla, CA, USA) software were used for all statistical analyses.