Adoptive cell therapy (ACT) involves the transfer of antitumor lymphocytes into a tumor-bearing host. It is a potent and feasible immunotherapy for certain advanced or relapsed malignancies, although it requires significant front-end “personalization” for each patient [63]. ACT was initially developed to generate lymphokine-activated killer (LAK) cells, which could directly lyse tumor cells [64, 65]. Then, strategies to isolate and expand tumor antigen-specific T cells were developed. Specifically, tumor-infiltrating lymphocytes (TILs) were isolated from resected tumors and expanded ex vivo by coculturing them with patient tumors and the IL-2 cytokine. TILs in combination with IL-2 had about a 50 % objective tumor response in patients with metastatic melanoma [66–69]. Moreover, TILs expanded from EGCs may provide a new and promising approach for patients with metastatic esophageal and gastric cancers [38].

The authors conducted a phase I/II trial for esophageal SCC with adoptive cell therapy [70]. Peripheral blood mononuclear cells were stimulated in vitro with autologous tumor cells. T cells were directly injected into primary tumors, metastatic lymph nodes, pleural spaces, or ascites in combination with IL-2. The objective tumor responses were achieved in half of the patients. Four of 11 patients (36 %) had confirmed complete or partial response. Furthermore, one patient with recurrent esophageal SCC had a partial response to the therapy [71].

Adoptive cell therapy has also had some success in patients with gastric cancer. Expanded T cells, called cytokine-induced killer (CIK) cells, were found to have appreciable antitumor activity against human gastric cancer [72]. At an effector to target cell ratio of 30:1, CIK cells were able to destroy 58 % of MKN74 human gastric cancer cells, suggesting that CIK cells can be developed for ACT. CIK cells combined with chemotherapy in postoperative stage III–IV gastric cancer patients significantly improved overall survival time and disease-free survival time compared to conventional chemotherapy alone [73]. In other nonrandomized or randomized trials, patients treated with chemotherapy combined with CIK cells had increased survival rates compared to those who received chemotherapy alone [74, 75]. In addition to CIK cells, ex vivo expanded human NK cells can acquire cytolytic activity against gastric tumor cells [76]. Currently, there is no FDA-approved ACT protocol for the treatment of cancer; however, the recent explosion of data regarding ACT should usher these novel strategies into daily clinical practice. To bolster this transition from the bench to the clinic, future trials need to address the barriers raised by Tregs, the use of engineered culture systems, and the genetic modification of T cells [77–80]. Moreover, clinical data concerning the efficacy of ACT in EGCs are insufficient and additional trials are required.

DCs are antigen-presenting cells that most effectively activate the adoptive immune response. Antigen presentation by DCs is critical for the induction of antitumor T-cell immunity. Gastric cancer patients with high levels of infiltrated DCs had a lower frequency of lymphatic invasion and had increased 5-year survival rates. Therefore, DC-based vaccinations could provide a novel immunotherapeutic approach for advanced gastrointestinal cancer patients [81, 82].

Several clinical studies have investigated DC-based vaccinations in patients with esophageal and gastric cancers (Table 9.2). In a clinical study of 12 patients with advanced gastrointestinal carcinoma (6 stomach, 3 esophagus, and 3 colon), Sadanaga et al. reported that ex vivo generated autologous DCs pulsed with MAGE-3 peptide were an effective and safe antitumor vaccine [83]. In this study, patients were immunized every 3 weeks for 3 months without experiencing toxic side effects. Peptide-specific CTL responses were detected in four of eight patients. Tumor markers decreased in seven patients, and tumors regressed (evidenced by imaging studies) in three patients, suggesting that DCs are safe and promising components for vaccine development. Kono et al. published a phase I vaccination trial in nine gastric cancer patients using DC pulsed with immunodominant HLA-A2-restricted HER2/neu (p369) peptides [84]. There were no adverse effects noted in the immunized patients. HER2/neu peptide-specific immune responses were detected in six of nine immunized patients (67 %), and peptide-specific hypersensitivity responses occurred in three of nine patients (33 %). One of the patients underwent PR response concurrent with a decrease in tumor markers, and another patient demonstrated SD for a period of 3 months.

Homma et al. generated a vaccine with fused autologous DCs and tumor cells (DC/tumor-fusion vaccine) [85]. The study consisted of 22 patients with advanced cancer, including 3 with gastric cancer. One gastric cancer patient had significantly elevated levels of serum antinuclear antibodies following treatment, which might have resulted from the immune response induced by the vaccine. Malignant ascitic effusion eventually was resolved in this patient, and their serum levels of tumor markers decreased. Fujiwara et al. performed a pilot study involving the intratumoral administration of ¹¹¹In-labeled DC in combination with chemotherapy (Adriamycin, cisplatin, and 5-FU) before surgical treatment in five esophageal cancer patients [86]. No adverse effects directly related to the intratumoral DC administration were observed. None of the antibodies against the 28 tumor antigens were upregulated. Moreover, enhancement of NY-ESO-1-specific cellular immune response was not observed. According to scintigraphic images obtained after treating each patient, DCs remained at the injection sites and did not drain in lymph nodes, suggesting that intratumoral DC administration does not elicit an optimal clinical response.

The field of cancer immunotherapy has significantly progressed ever since Boon and his colleagues made the observation in 1991 that a tumor-associated antigen (TAA) can be targeted by cytotoxic T lymphocytes [87–90]. Since then, technical advances have facilitated the identification of many TAAs and peptide epitopes that can be targeted for cancer immunotherapy [91]. For example, esophageal and gastric cancers express a variety of TAAs as potential targets for immunotherapies, and several clinical trials involving these TAAs have had promising results (Table 9.3) [92–99, 100, 101].