Surgery alone or combined with chemo- and/or radiation therapy remains the primary treatment for gastric cancer (GC) to date and immunotherapeutic tools such as monoclonal antibodies are only slowly being implemented. This is partly due to the fact that the immune microenvironment in GC during chemoradiation and other treatment modalities is still poorly understood. 7 gastric cancer (GC) cell lines were tested for their response to chemoradiation using 5-FU in combination with X-ray irradiation. We conducted flow cytometric analysis to determine the cells' ability to undergo immunogenic cell death (ICD) and their expression of the two immunosuppressive proteins programmed death-ligand 1 (PD-L1) and galectin-9 (Gal-9). We evaluated the overall immunogenicity of two cell lines (MKN7, MKN74) in co-culture experiments with human monocyte-derived dendritic cells (Mo-DCs). Chemoradiation induces distinct responses in different GC cell lines. We observe ICD in vitro in all tested GC cell lines in the form of calreticulin (CRT) translocation to the plasma membrane. As a resistance mechanism, these cells also upregulated Gal-9 and PD-L1. Mo-DC maturation experiments showed that GCs provoked the maturation of Mo-DCs after chemoradiation in vitro. The addition of α-PD-L1 blocking antibody further enhanced the immunogenicity of these cells while improving DC viability. Blocking Tim-3, as the main receptor for Gal-9, had no such effect. Our findings suggest that the benefits of chemoradiation can substantially depend on tumor subtype and these benefits can be offset by induced immune evasion in GC. Combination treatment using checkpoint inhibitors could potentially lead to enhanced immune responses and yield better patient outcomes.

Although basic, translational and clinical research suggest a possibility of synergistic effect of radiation-induced immunogenic cell death with immune checkpoint inhibitors, the effectiveness of concurrent therapy with radiotherapy and immunotherapy is not fully established. Phase I/II, open single arm, prospective clinical trial was conducted and eligible patients were unresectable advanced or recurrent gastric cancer patients (n = 40) who developed progression after primary and secondary chemotherapy with multiple metastasis assessable in imaging (one lesion must be ≥2cm). Radiotherapy of total 22.5 Gy/5 fractions/5 days was given to the largest or symptomatic lesion and nivolumab was administered day 15-22 at 3 mg/kg or 240mg/body every 2 weeks to a total of 6 administrations. The primary endpoint is disease control rate of non-irradiated lesions as an abscopal effect. The secondary endpoints are MST, safety and proportion of local control rate for irradiated lesion. As an ancillary analysis, immunologic parameters including high-dimensional MHC multimer analysis and TCR repertoire analysis, and ctDNA analysis are designed. Total 41 patients were enrolled and 40 patients were evaluated as full analysis set, since 1 patient was judged as an unmatched case to inclusion criteria. The clinical evaluation as an abscopal effect were CR=1 (2.5%), PR=5 (12.5%), SD=3 (7.5%), PD=15 (37.5%) and NE=16 (40%). The disease control rate and the response rate for non-irradiated target lesions was 22.5% and 15%, respectively. The MST was 230 days (157-330, 95%CI) and 1-year survival rate was 28.6%. Any adverse event more than Grade 3 was observed in 16 cases in 41 patients (39%). The most frequent AE more than grade 3 were anemia (19%) and appetite loss (12%). The local control for irradiated lesions were seen in CR=5 (12.5%), PR=6 (15%), SD=5 (12.5%), PD=4 (10%) and NE=20 (50%). The ancillary analysis for immunological monitoring is under investigation. The combination of nivolumab with radiotherapy demonstrated promising anti-tumor activity with marked prolonged survival time in gastric cancer. No new safety issues were detected in the combination.

Circulating tumor cells (CTCs) are tumor cells that are released into the bloodstream from the primary tumor or metastatic lesions, and have the potential to act as precursors of metastases in several types of cancer. Analysis of CTCs are non-invasive and real-time monitoring of cancer patients, thus now used as one of the liquid biopsy-based prognostic markers. In esophageal carcinoma (EC), a meta-analysis indicated that CTC-positive patients were associated with poor progression-free survival (PFS) and overall survival (OS). Japanese guideline recommends neoadjuvant chemotherapy (NAC) with 5FU/CDDP(FP) followed by curative surgery for stage Ⅱor Ⅲ esophageal squamous cell carcinoma (ESCC) and definitive chemoradiotherapy with FP for stageⅣa ESCC. We analyzed CTCs in EC patients to investigate whether NAC or chemotherapy (CT) have enough power to eliminate CTCs, or alter their phenotypes or not. Peripheral blood samples from 31 EC patients before treatment and 25 patients of them after treatment were analyzed for the presence of CTCs with microfluidic chips. CTCs are defined as Pan-CK (+), DAPI (+) and CD45 (-), and classified into epithelial-CTCs and mesenchymal-CTCs by Vimentin expression by staining and scanning systems. Before treatment, CTCs were detected in 26 of 31 patients (84%) regardless of their clinical stage. Comparing 25 patients before and after treatment in each patient, their average number of CTCs increased from 9.92 to 11.12 in 4ml blood samples. There was no correlation between number of CTCs and clinical stage or clinical/pathological response or radiation therapy. The average ratio of epithelial/mesenchymal CTCs has changed only slightly from 0.45 to 0.47. The number and phenotype of CTCs did not affect the response of NAC or CT. Our quantitative and qualitative analysis for CTCs shows that NAC and CT with FP does not have enough power to eliminate CTCs and alter their phenotypes.

AT-rich interactive domain 1A (ARID1A) is a tumor suppressor gene and frequently mutated in gastric cancer (GC). Although ARID1A mutations are not a druggable target for conventional treatments, novel therapeutic strategies based on a synthetic lethal approach are effective for ARID1A-deficient cancers. EZH2, the histone methyltransferase and a member of the polycomb repressive complex 2 (PRC2), catalyzes the trimethylation of lysine 27 on histone 3 and showed a synthetic lethality in ARID1A-mutated ovarian cancer, but its role in GC has not been investigated yet. The selective sensitivity of the EZH2 inhibitors for ARID1A-deficient GC cells was evaluated using cell viability and colony formation assays enrolling three kinds of EZH2 inhibitors. The expression of PI3K/AKT signaling genes was investigated using TCGA's cBioPortal database to determine whether the homeostasis between ARID1A and EZH2 is related to cell proliferation and survival via the PI3K/AKT signaling pathway. We also evaluated the expression of EZH2, PD-L1, and PD-L2, MSI status, and EBV infection were investigated in ARID1A negative GC using patient samples. EZH2 inhibitors decreased the viability of ARID1A-deficient cells in a dose-dependent manner and demonstrated the selective sensitivity to ARID1A-deficient cells in vitro experiment system. A bioinformatics approach revealed that the PI3K/AKT signaling tended to be activated in ARID1A-deficient GC enhancing cell viability and, furthermore, down-regulation of EZH2 in ARID1A-deficient GC was related to normalization of PI3K/AKT signaling pathway. Clinicopathological characteristics of ARID1A negative GC with ARID1A, EZH2, PD-L1, and PD-L2 expressions, MSI, and EBV status were summarized. The present findings provide a rationale for the selective sensitivity of EZH2 inhibitors against ARID1A-deficient GC and suggest the potential efficacy of targeted therapy using EZH2 inhibitors in this patient population.