Background

Gastrointestinal cancers (GICs) are one of the leading causes of cancer-related mortality in the world, which majorly include cancers of colorectum and stomach (CRC and GC). In addition to growth and aging of global population, behavioral risk factors, such as smoking and dietary patterns, played an important role in the rising global burden of GIC [1]. Many treatment strategies, such as surgery, neoadjuvant chemotherapy, and adjuvant chemotherapy [2], are available for GIC patients. Moreover, early GIC patients can be cured by appropriate treatment, and the 5-year overall survival (OS) rate is 90%; however, for patients with advanced-stage GIC, the 5-year OC rate is still very low [3, 4].

Surgery, radiotherapy, and chemotherapy are routine treatments for GIC in most patients; they have the efficacy of eliminating and destroying primary tumors. However, these traditional treatments are often ineffective for advanced-stage cancers [5]. Cancer immunotherapy is a promising treatment method, which has made great progress in improving anti-tumor immunity [6]. Autoimmune therapy for malignant tumors is considered to be a feasible method, which mainly depends on the interference and inhibition effects of killer cells induced by the tumor, infiltrating lymphocytes and lymphokines, and CD3 monoclonal antibody [7, 8]. Dendritic cells (DCs) have the ability to present antigens, making them an attractive vehicle for the delivery and development of therapeutic tumor vaccines [9] . Cytokine-induced killer cells (CIK) is a heterogeneous effector CD3 + CD56 + NKT cell population that can be expanded from peripheral blood mononuclear cells (PBMC) in vitro [10,

Results

Literature search and study selection

A total of 466 studies were identified from PubMed, Cochrane library, and Embase. Then, three additional records were found by manually searching the reference lists of other studies. After deleting the duplications, 351 studies were selected. Then, 330 studies were discarded because of their irrelevance to the topic of interest. Of the remaining 21 studies, 3 were excluded for unavailability of data for statistics and four were excluded as they did not include controls. Besides, five studies did not focus on the advanced-stage GIC. Finally, a total of nine studies, including 1113 patients, met the inclusion and exclusion criteria in this meta-analysis11–19. The flow diagram of the search process is shown in Fig. 1.

Fig. 1
figure 1

Flow diagram of study identification

Full size image

Study characteristics

The key characteristics of all included studies are summarized in Table 1. All the studies involved patients with advanced GIC followed up for at least 24 months. Nine studies from 2006 to 2017 compared CIK/DC–CIK plus chemotherapy with chemotherapy alone for treating advanced GIC. These nine studies were assessed using the Cochrane Collaboration’s tool for the risk of bias. A graph and summary of the risk of bias are shown in Figs. 2 and 3. Four studies did not mention randomization, and three studies did not provide information of allocation concealment.

Table 1 Characteristics of the studies included in this meta-analysis
Full size table
Fig. 2
figure 2

Risk-of-bias assessments for the randomized trials included in the meta-analysis. a Risk-of-bias summary. b Risk-of-bias graph. Symbols: (+), Low risk of bias; (?), unclear risk of bias; and (−), high risk of bias

Full size image
Fig. 3
figure 3

Forest plot of the comparison of survival. a Overall survival (OS). b Progression-free survival

Full size image

Meta-analyses

Efficacy

Clinical responses were assessed in terms of the OS and PFS to evaluate the prognosis.

Among the nine trials, three reported 1-year OS rate, five reported 3-year OS rate, and four reported 5-year OS rate (Fig. 3a). As a slightly significant heterogeneity was detected, the fixed-effects model was used. Chemotherapy combined with CIK/DC–CIK immunotherapy showed a significant increase in 1-year OS (3 studies, 292 patients, RR = 1.43, 95% CI = 1.23–1.66, Pheterogeneity = 0.310, I2 = 14.6%), 3-year OS (5 studies, 727 patients, RR = 1.43, 95% CI = 1.25–1.64, Pheterogeneity = 0.179, I2 = 36.3%), and 5-year OS (4 studies, 580 patients, RR = 1.84, 95% CI = 1.41–2.40, Pheterogeneity = 0.654, I2 = 0%) compared with those of chemotherapy alone.

In terms of PFS, three studies presented relevant data of 1-year PFS, five reported 3-year PFS, and four reported 5-year PFS. As shown in Fig. 3b, chemotherapy combined with immunotherapy significantly prolonged 1-year PFS (3 studies, 292 patients, RR = 1.34, 95% CI = 1.13–1.59, Pheterogeneity = 0.972, I2 = 0%), 3-year PFS (5 studies, 727 patients, RR = 1.39, 95% CI = 1.20–1.62, Pheterogeneity = 0.664, I2 = 0%), and 5-year PFS (4 studies, 580 patients, RR = 1.99, 95% CI = 1.52–2.60, Pheterogeneity = 0.727, I2 = 0%) compared with chemotherapy alone.

Efficacy

Efficacy was assessed in terms of CR, PR, and ORR. The analysis result is shown in Fig. 4; however, the RR of CR (3 studies, 207 patients, RR = 1.89, 95% CI = 0.18–20.20), PR (3 studies, 207 patients, RR = 1.86, 95% CI = 0.93–3.71, Pheterogeneity = 0.773, I2 = 0%), and ORR (3 studies, 207 patients, RR = 1.51, 95% CI = 0.97–2.37, Pheterogeneity = 0.417, I2 = 0%) did not infer significant difference between the two groups. Three studies evaluated the effect of CIK/DC–CIK immunotherapy on the QOL of patients with advanced GIC. Significantly improved QOL was found in the CIK/DC–CIK immunotherapy group compared with the chemotherapy-alone group (3 studies, 245 patients, WMD = 16.09, 95% CI = 1.66–30.52, Pheterogeneity < 0.001, I2 = 98.8%) (Figure S1).

Fig. 4
figure 4

Forest plot of the comparison of complete response (CR), partial response (PR), and overall response rate (ORR)

Full size image

The safety of CIK/DC–CIK therapy in the treatment of advanced GIC was evaluated in this meta-analysis. As shown in Figure S2, no significant difference was observed in terms of thrombocytopenia (2 studies, 283 patients, RR = 1.04, 95% CI = 0.93–1.16, Pheterogeneity = 0.497, I2 = 0%), nausea, vomiting (3 studies, 329 patients, RR = 0.67, 95% CI = 0.35–1.28, Pheterogeneity = 0.003, I2 = 82.9%), abnormal liver function (2 studies, 301 patients, RR = 1.08, 95% CI = 0.95–1.23, Pheterogeneity = 0.373, I2 = 0%), neutropenia (2 studies, 150 cases, RR = 0.55, 95% CI = 0.31–1.00, Pheterogeneity = 0.955, I2 = 0%), and myelosuppression (2 studies, 193 patients, RR = 0.73, 95% CI = 0.48–1.11, Pheterogeneity = 0.088, I2 = 65.7%) between the two groups.

Subgroup analysis

This study probed the detailed results in subgroup analyses stratified by cancer type (GC or CRC). All subgroup results were quite consistent with the overall results. The results are summarized in Table 2.

Table 2 Subgroup analysis based on cancer type
Full size table

Sensitivity analysis

Sensitivity analyses were performed to assess the stability of the results by sequentially removing each study. The removal of any single study did not change the overall statistical results, indicating that the results of this study were statistically robust (Fig. 5).

Fig. 5
figure 5

Sensitivity analysis examining the influence of individual study on pooled results. a Overall survival. b Progression-free survival. c Partial response. d Overall response rate

Full size image

Publication bias

The outcome of OS, with the largest number of included studies, was chosen to test the publication bias. Visual examination of the funnel plot (Fig. 6a) revealed a considerable degree of asymmetry. In addition, publication bias was statistically significant by Egger’s test or Begg’s test (Begg’s test, P = 0.034; Egger’s test, P = 0.016). Therefore, a sensitivity analysis was conducted using the trim-and-fill method (Fig. 6b) [21]. After imputing six unpublished studies, the trim and fill sensitivity analysis did not change the general result (RR = 0.255, 95% CI = 0.176–0.333, P < 0.01).

Fig. 6
figure 6

Funnel plot for publication bias. Each point represents a separate study for the indicated association. a Funnel plot of OS. b Trim-and-fill plot of OS

Full size image

Discussion

GIC is still a fatal threat to human health due to metastasis, with recurrence inducing refractory advanced tumor stage and poor prognosis. Yan et al. [22] proved that the recurrence rate of patients with GC ranged from 40 to 65% due to distant metastasis and local recurrence. Adoptive cellular immunotherapy can be used in combination with standard therapy as adjuvant and palliative treatment after operation to improve the survival rate and QOL of patients with GIC. The CIK cells have shown the best efficacy in this treatment. Compared with lymphokine-activated killer cells, CIK cells are more readily available and show higher tumor-specific cytotoxic activity [23,24,25,26,27]. To date, several clinical trials have used chemotherapy plus DC–CIK immunotherapy to treat advanced GIC. However, in these trials, the clinical protocol applied was not standard, blurring the evaluation of treatment effects. In this study, a large number of comprehensive trials were investigated to achieve higher statistical reliability. This meta-analysis showed that chemotherapy combined with CIK/DC–CIK immunotherapy improved the OS, PFS, and QOL without causing serious adverse events.

This study confirmed the safety of CIK/DC–CIK immunotherapy for advanced GIC patients, and the adverse events caused were tolerated by all patients. No significant difference was observed in terms of common adverse events, such as thrombocytopenia (RR = 1.04, 95% CI = 0.93–1.16, Pheterogeneity = 0.497, I2 = 0%), nausea, vomiting (RR = 0.67, 95% CI = 0.35–1.28, Pheterogeneity = 0.003, I2 = 82.9%), abnormal liver function (RR = 1.08, 95% CI = 0.95–1.23, Pheterogeneity = 0.373, I2 = 0%), neutropenia (RR = 0.55, 95% CI = 0.31–1.00, Pheterogeneity = 0.955, I2 = 0%), and myelosuppression (RR = 0.73, 95% CI = 0.48–1.11, Pheterogeneity = 0.088, I2 = 65.7%) in the CIK/DC–CIK immunotherapy group compared with the chemotherapy-alone group. The CIK/DC–CIK immunotherapy enhanced the efficiency of conventional therapy in treating advanced GIC. Compared with the conventional therapy group, the 1-year OS, 3-year OS, 5-year OS, 1-year PFS, 3-year PFS, and 5-year PFS of patients in the combination therapy group improved remarkably. Moreover, the combination therapy improved the QOL (WMD = 16.09, 95% CI = 1.66–30.52, Pheterogeneity < 0.001, I2 = 98.8%) of the patients by relieving pain, decreasing fatigue and insomnia, and improving appetite.

Chemotherapy is thought to damage the immune system. Immunotherapy is a type of cancer treatment that helps the body’s own immune system fight cancer, many clinical trials of immunotherapy have been conducted in multiple centers. DCs are an important immunotherapeutic cell type. Ishigami et al. [28] demonstrated that patients with GC having a high level of DC cell infiltration were less likely to have lymph node metastasis and a significantly increased 5-year survival rate. In addition, CIK cells exhibit strong cytotoxicity against a variety of tumor cell lines as well as newly isolated tumor samples (e.g. liver cancer [29], lung cancer [30], glioma [31], and GC [32]). Compared with other immune cells, CIK cells proliferate rapidly and exhibit strong antitumor activity and a broad spectrum of targeted tumors [14, 33]. Therefore, CIK/DC–CIK cell-based immunotherapy is a promising treatment for a number of cancers. Zhao et al. [34] showed that “GP regimen combined with DC–CIK immunotherapy could reduce postoperative recurrence and prolong survival in patients with non-small-cell lung cancer.” Ma et al. [35] demonstrated that CIK cell therapy had significant advantages in prolonging the median OS rate, PFS, DCR, ORR, and QOL of patients with hepatocellular carcinoma. Other studies of gastric cancer [13], advanced renal cancer [36], and metastatic nasopharyngeal carcinoma [37] demonstrated that DC–CIK cells could improve the prognosis of patients. CIK cells have become a promising immunotherapy method in tumor therapy due to their easy availability and strong antitumor activity, which is of great significance for tumor prognosis [38].

Heterogeneity is a problem with most meta-analyses. In the present meta-analysis, heterogeneity was found in overall analyses; thus, the random-effects model was used. Sensitivity analysis was performed to assess the stability of the results by sequentially removing each study. The removal of any single study did not change the overall statistical results, indicating that the results of this study were statistically robust. Furthermore, this study probed the detailed results in subgroup analyses stratified by cancer type (GC or CRC). Based on the data collected, this study suggested that the cancer type at least partly contributed to the between-study heterogeneity.

This meta-analysis had some limitations that might have affected the interpretation of results. First, the efficacy of CIK/DC–CIK immunotherapy was affected by many factors, such as injection mode, tumor stage, and metastasis cycle. Further, a detailed analysis needs to be carried out on the basis of complete literature, standardization of treatment options, and limitation of patient participation criteria. Second, some data on adverse events could only be used in two trials, which might have led to bias. Third, publishing bias existed in this study. Although the trim-and-fill method was used in this study to confirm the results, some negative data that were possibly omitted might have affected the results. Finally, the heterogeneity could not be completely eliminated in the analysis.

Conclusion

In summary, this meta-analysis confirmed that the combination of CIK/DC–CIK immunotherapy and chemotherapy was safe and applicable for patients with advanced GIC, and hence might serve a feasible choice to prolong survival and improve QOL. Therefore, CIK/DC–CIK immunotherapy is an effective therapy for advanced GIC treatment.