Abstract
Purpose
Systemic Inflammation Response Index (SIRI), based on peripheral neutrophil, monocyte, and lymphocyte counts, was recently developed and used as a marker to predict the survival of patients with malignant tumours. Cancer stem cells (CSCs) can contribute to gastric cancer progression and recurrence. It is not clear whether SIRI is associated with CSCs during gastric cancer development.
Methods
The SIRI was developed in a training cohort of 455 gastric cancer patients undergoing curative resection between 2007 and 2009, and validated in a validation cohort of 327 patients from 2010 to 2011. CD44 + CSCs were measured on tumour sections by immunohistochemical analysis.
Results
An optimal cut-off point for the SIRI of 0.82 divided the gastric cancer patients into a low SIRI group (SIRI < 0.82) and a high SIRI group (SIRI ≥ 0.82) in the training cohort. Compared with patients who had a SIRI < 0.82, patients who had a SIRI ≥ 0.82 had a shorter disease-free survival (DFS) (HR 2.529; 95% CI 1.922–3.326; p < 0.001) and shorter disease-special survival (DSS) (HR 2.692; 95% CI 2.022–3.585; p < 0.001) in the training cohort, comparable DFS and DSS findings were observed in the validation cohort, even for patients in pathological TNM stage of I subgroup. A SIRI ≥ 0.82 was significantly associated with older age, larger tumour, higher pathological TNM stage, lymphovascular invasion, and perineural invasion. Additionally, patients in the low SIRI group were prone to DFS and DSS benefits from postoperative adjuvant chemotherapy. Univariate and multivariate analyses revealed that SIRI was an independent predictor for DFS and DSS. Furthermore, gastric cancer patients with CD44 + CSCs scores had higher SIRI level (mean 1.198 vs. 0.835; p < 0.001). In patients with CD44 + CSCs, those with SIRI ≥ 0.82 had higher recurrence rates and shorter survival time than patients with SIRI < 0. 82.
Conclusions
SIRI was a useful prognostic indicator of poor outcomes in patients with gastric cancer and is a promising tool for gastric cancer treatment strategy decisions. The dismal outcomes in patients with high SIRI might be related to CSCs.
Electronic supplementary material
The online version of this article (doi:10.1007/s00432-017-2506-3) contains supplementary material, which is available to authorized users.
Keywords: Systemic Inflammation Response Index, Cancer stem cells, Gastric adenocarcinoma, Curative resection, Survival
Introduction
Gastric cancer is the fifth most common malignancy and the third leading cause of cancer death worldwide (Ferlay et al. 2015). Currently, surgery remains the main treatment for gastric cancer. However, the survival rate of gastric cancer patients is less than 30% (Siegel et al. 2016). Even after curative resection, approximately 35–70% patients have recurrence or metastasis within 5 years (Huang et al. 2016). To improve the control of post-surgical relapses, various regimens for adjuvant chemotherapy have been tested over the past several decades. For patients with advanced gastric cancer, a fluoropyrimidine-based chemotherapy forms the backbone of chemotherapy (Bang et al. 2010), and with the administration of a regimen of capecitabine plus platinum, patients with locally advanced gastric cancer can obtain improved overall survival (Noh et al. 2014). Additionally, patients with tumour recurrence or metastasis might have better outcomes when using chemotherapy combined with targeted treatments (Bang et al. 2010). However, approximately 50% of all gastric cancers do not respond to treatment, and only a few patients achieve disease stabilization or a partial response to treatment (Cunningham et al. 2008). Therefore, it is important to develop a prognostic tool that can identify patient subpopulations with a high risk of recurrence and metastasis and allow rational future therapies to be tailored for those patients.
The mechanisms for recurrence/metastasis of gastric cancer remain uncertain. Cancer stem cells (CSCs) may account for tumour development and progression and tumour recurrence and metastasis following therapy. CSCs are a subpopulation of tumor cells and featured by their capabilities of self-renewal and tumor initiation (Nguyen et al. 2012). Additionally, studies have demonstrated that purported CSCs showed resistance to chemotherapy (Takaishi et al. 2009; Yoon et al. 2014). Recent studies have demonstrated that CSCs may not be a fixed cell population and may exhibit plasticity regulated by tumor microenvironmental factors (Gupta et al. 2009; Magee et al. 2012). Studies have shown that tumor stromal cells, including BMFs (Zhu et al. 2016), MSC (Tsai et al. 2011), and endothelial cells (Hamerlik et al. 2012) induce CSC-like sphere formation of cancer cells, thereby promoting distant tumour metastasis (Magee et al. 2012; Zhang et al. 2016).
Cancer-related inflammation also plays an important role in cancer development and progression and immune and inflammatory cells are considered as essential components of the tumour microenvironment (Hanahan and Weinberg. 2011). The routine immune and inflammatory cells such as neutrophils, monocytes, and lymphocytes, present and detectable in the systemic circulation, may contribute to tumour cell invasion and metastasis (Diakos et al. 2014). Studies have shown that neutrophils can promote the development and progression of cancer by providing an adequate tumour microenvironment via secretion of cytokines and chemokines (Gregory and Houghton 2011). Monocytes, particularly, tumour-associated macrophages (TAMs), which are derived from circulating monocyte populations, have been reported to be a key player in the tumour microenvironment by encouraging tumour progression and metastasis (Galdiero et al. 2013b). Furthermore, several studies have indicated that factors secreted by modifying neutrophils and TAMs affect tumour stem cell stemness (Yu et al. 2017; Wan et al. 2014) and can thus influence on sensitivity to chemotherapy resistance (Takaishi et al. 2009; Yoon et al. 2014). Lymphocytes play a crucial role in cancer immune surveillance and defence by inducing cytotoxic cell death and inhibiting tumour cell proliferation and migration (Mantovani et al. 2008). Thereby, dictating the host immune response to malignancy (Li et al. 2017; Ock et al. 2017). Considering these factors, several inflammation and immune-based prognostic scores, such as neutrophil–lymphocyte ratio (NLR), and monocyte–lymphocyte ratio (MLR), have been developed to predict survival and recurrence in many types of cancer, including gastric cancer (Ock et al. 2017; Wang et al. 2016; Deng et al. 2015; Zhou et al. 2014a). However, an integrated indicator based on peripheral neutrophil, monocyte, and lymphocyte counts, which might be able to better reflect the balance of host inflammatory and immune status, has not yet been reported in gastric cancer. Moreover, the potential effects of peripheral neutrophil, monocyte, and lymphocyte counts on gastric cancer recurrence and metastasis have not been elaborated.
The Systemic Inflammation Response Index (SIRI), based on peripheral neutrophil (N), monocyte (M), and lymphocyte (L) counts, was recently developed and has proved to be a marker to predict the poor survival of patients with pancreatic cancer (Qi et al. 2016). However, the prognostic value of the preoperative SIRI in gastric cancer patients is still unclear. CD44 has been described as a cell surface marker of human gastric cancer CSCs (Takaishi et al. 2009; Zhang et al. 2016), CD44-positive gastric cancer cells exhibited stem cell properties of self-renewal, asymmetrical division and differentiation properties, and giving rise to the more or less differentiated cells composing the tumor mass (Takaishi et al. 2009). Besides, CD44-positive gastric cancer cells have been shown to be more resistant to treatments than the more differentiated cancer cells, and may therefore lead to cancer recurrence after treatment (Yoon et al. 2015). Therefore, CD44 + CSCs may act as the mastermind or root cause of gastric cancer metastasis and be crucial for gastric cancer metastatic potential, and CD44 + CSCs can be regulated by their stromal microenvironment. Therefore, we tested whether systemic inflammation may promote gastric cancer CD44 + CSCs function and affect patient survival.
In this study, we performed a large-scale retrospective cohort study to evaluate the prognostic value of SIRI in localized or regional gastric cancer patients after curative resection and investigated the association of SIRI with CSCs.
Materials and methods
Patients and clinical specimens
This retrospective study recruited two independent cohorts of patients with gastric cancer from the Department of Gastroenterological Surgery of Harbin Medical University Cancer Hospital (Harbin, China). The training cohort that comprised of 455 patients who had undergone curative resection was obtained between 2007 and 2009. From 2010 to 2011, a validation cohort of patients with gastric cancer (n = 327) undergoing curative resection was recruited. The medical data and follow-up information of gastric cancer patients were identified from our prospective database. In this study, all patients were diagnosed with gastric adenocarcinoma and were not treated with neoadjuvant chemotherapy. The patients who were diagnosed with other gastric malignances such as lymphoma and gastrointestinal stromal tumours were excluded from the study. Additionally, remnant gastric cancers were excluded from this analysis. Exclusion criteria also included active infection or inflammatory disease within 1 month before blood test, and death within the 3 months after surgery. Patients who were diagnosed with advanced or early stage tumours with lymph node metastasis were candidates for postoperative adjuvant chemotherapy.
The tumour size was defined according to the longest diameters of the samples. The eighth edition of the AJCC TNM staging classification for carcinoma of the stomach was used for tumour staging. The Lauren classification was defined as intestinal type, diffuse type, and mixed type. The histological grade was classified as G1, G2, G3 adenocarcinoma, signet ring cell carcinoma, and mucinous adenocarcinoma. Lymphovascular invasion and perineural invasion were diagnosed by H&E-stained slides.
Blood samples were obtained for neutrophil, monocyte, and lymphocyte counts, which were measured in the laboratory 10 days before surgery. The NLR and MLR were calculated by dividing the absolute neutrophil (N) count by the absolute lymphocyte (L) count and the absolute monocyte (M) count by the absolute lymphocyte (L) count, respectively. The optimal NLR and MLR cut-off values for tumour recurrence were calculated by applying receiver operating characteristic (ROC) analysis in training cohort. The results of the ROC analysis revealed that the optimal cut-off points for the NLR and MLR were 2.10 and 0.22 in training cohort, respectively (Supplementary Fig. S 1a and S 1b). Subsequently, the NLR and MLR were stratified into NLR < 2.10 or NLR ≥ 2.10 and MLR < 0.22 or MLR ≥ 0.22, respectively, for all subsequent analysis. Patient tissues in the training cohort were obtained for immunohistochemical analysis to detect CD44 + CSCs. Patients were followed up carefully after surgery at 6–12 month intervals for the first 5 years after surgery and then annually. The postoperative follow-up was carried out by regular out-patient visits and telephone interviews. The last follow-up date was March 31, 2016. The overall follow-up rate was 87.00% (455/523) and 91.09% (327/359) in the training cohort and the validation cohort, respectively. Disease-free survival (DFS) was defined as the date of surgery to the date of identification of disease recurrence, either radiological or histological. Disease-specific survival (DSS) was calculated from the date of surgery to the date of death from gastric cancer. Patients who died of causes unrelated to the disease were censored at the last follow-up. This retrospective study was approved by the ethics committee of Harbin Medical University Cancer Hospital, China.
Expression of CD44 in cancer tissues
Immunohistochemical analysis of CD44 was performed using the biotin–streptavidin–peroxidase method, as previously described. In brief, paraffin-embedded and formalin-fixed tissues were cut into 4 μm sections and incubated on slides. The sections were dewaxed in xylene and graded alcohols, hydrated and washed in phosphate-buffered saline (PBS), immersed in 3% H2O2 for 30 min to block endogenous peroxidase, and washed three times with PBS for 5 min. The sections were then incubated for 10 min on a hot plate (95–99 °C), allowed to cool for 20 min, and incubated with 1% bovine serum albumin in PBS at room temperature for 30 min. The sections were incubated for 2 h in a moist chamber at 4 °C with a primary monoclonal anti-CD44 antibody (dilution 1:300, EPR1013Y Abcam). The sections were then incubated with secondary antibody (dilution 1:500, ab205718 Abcam) at 37 °C temperature for 20 min. Each section was performed with diaminobenzidine (DAB) reagent sets for 2 min, and was then counterstained with haematoxylin. Negative controls were treated identically without primary antibodies.
The immunoreactivity was assessed with the immunohistochemistry score (HSCORE) system as previously described, which evaluated the staining intensity and percentages of positive cells stained at a specific magnitude of intensity. The HSCORE was calculated using the equation HSCORE = ∑Pi (i) (I = 0, 1, 2, 3, Pi = 0–100%), in which i was the staining intensity (0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining) and Pi was the percentage of positive cells on a scale of 0–100%. Positive cell count was assessed in five randomly selected fields of vision with a microscope (400×). The expression levels of CD44 were classified as negative (HSCORE < 30) and positive (HSCORE ≥ 30). Cut-off values for CD44 (HSCORE < 30) were modified according to the results of previous studies (Wang et al. 2011).
The SIRI
The SIRI was defined as follows: SIRI = N × M/L, where N, M, and L are the pretreatment peripheral neutrophil, monocyte, and lymphocyte counts, respectively. The optimal SIRI cut-off values for tumour recurrence after curative resection in the training cohort were calculated using a ROC analysis. The Youden Index (sensitivity + specificity − 1) was used to select a threshold to estimate sensitivity and specificity. The results of the ROC analysis revealed an optimal SIRI cut-off point of 0.82 (Supplementary Fig. S1c). Consequently, the SIRI scores were classified as SIRI < 0.82 or SIRI ≥ 0.82 for all subsequent analyses.
Statistical analysis
SPSS 19.0 software (Version 19.0, Chicago, IL, USA) and Graphpad prism 5.0 were used for all statistical analyses. We determined the optimal discriminator value for NLR MLR SIRI using receiver operating characteristic (ROC) curve analysis. The ratio closest to the point with maximum sensitivity and specificity was selected as the cut-off value. Analyses of variance and Pearson’s Chi-square tests were used to assess any associations between variables. Clinical outcomes were calculated by Kaplan–Meier survival curves, and the groups were compared using the log-rank test. Stepwise multivariate Cox proportion analysis was performed. The level of significance permitting multivariate analysis inclusion and the statistical significance for all other tests used was set at p < 0.05.
Results
Patients’ characteristics
In total, 782 patients were enrolled in the study. The clinicopathological characteristics of patients in training cohort and validation cohort are shown in Table 1. In the training cohort, the mean DFS and DSS were 58.5 and 63.1 months, respectively. 232 of 407 patients received postoperative adjuvant chemotherapy. 223 (49.0%) patients had confirmed recurrence after curative resection, and 209 (45.9%) had died at last follow-up. The median follow-up duration was 77.53 months (range 3.03–111.73). The mean age of the patients was 57.6 years (range 29.0–89.0). The mean number of harvested lymph nodes after surgical resection was 24.8 (range 10–74).
Table 1.
Clinical characteristics of patients in training cohort and validation cohort
| Variables | No. of patients (%) | p value | |
|---|---|---|---|
| Training cohort | Validation cohort | ||
| No. of patients | 455 | 327 | |
| Age | 0.635 | ||
| <70 | 368 (80.9) | 260 (79.5) | |
| ≥70 | 87 (19.1) | 67 (20.5) | |
| Sex | 0.689 | ||
| Male | 321 (70.5) | 235 (71.9) | |
| Female | 134 (29.5) | 92 (28.1) | |
| ASA status | 0.377 | ||
| <3 | 389 (85.5) | 272 (83.2) | |
| ≥3 | 66 (14.5) | 55 (16.8) | |
| Tumor site | 0.588 | ||
| Upper | 74 (16.3) | 58 (17.7) | |
| Middle | 122 (26.8) | 95 (29.1) | |
| Lower | 259 (56.9) | 174 (53.2) | |
| Tumor size | 0.521 | ||
| <5 | 236 (51.2) | 162 (49.5) | |
| ≥5 | 219 (48.1) | 165 (50.5) | |
| Pathological TNM stage | 0.095 | ||
| I | 96 (21.1) | 87 (26.6) | |
| II | 120 (26.4) | 92 (28.1) | |
| III | 239 (52.5) | 148 (45.3) | |
| Lauren classification | 0.188 | ||
| Intestinal | 282 (62.0) | 187 (57.2) | |
| Diffuse | 60 (13.2) | 58 (17.7) | |
| Mixed | 113 (24.8) | 82 (25.1) | |
| Histological grade | 0.599 | ||
| G1/G2 | 188 (41.3) | 129 (71.8) | |
| G3/signet ring cell/mucinous | 267 (58.7) | 198 (71.8) | |
| Lymphovascular invasion | 0.955 | ||
| No | 315 (69.2) | 227 (69.4) | |
| Yes | 140 (30.8) | 100 (30.6) | |
| Perineural invasion | 0.111 | ||
| No | 180 (39.6) | 148 (45.3) | |
| Yes | 275 (60.4) | 179 (54.7) | |
| NLR | 0.683 | ||
| <2.1 | 234 (51.4) | 173 (52.9) | |
| ≥2.1 | 221 (48.6) | 154 (47.1) | |
| MLR | 0.842 | ||
| <0.22 | 212 (46.6) | 150 (45.9) | |
| ≥0.22 | 243 (53.4) | 177 (54.1) | |
| SIRI | 0.580 | ||
| <0.82 | 233 (51.2) | 174 (53.2) | |
| ≥0.82 | 222 (48.8) | 153 (46.8) | |
| Fua | 0.804 | ||
| Yes | 232 (57.0) | 162 (56.1) | |
| No | 175 (43.0) | 127 (38.8) | |
ASA American Society of Anesthesiology, NLR neutrophil lymphocyte ratio, MLR monocyte–lymphocyte ratio, SIRI Systemic Inflammation Response Index
aFluoropyrimidine-based adjuvant chemotherapy, mostly including capecitabine plus platinum, capecitabine alone, or S1 (combined tegafur, gimeracil, and oteracil), in patients at advanced stage or early stage tumors with lymph node metastasis in this retrospective study
In the validation cohort, the mean DFS and DSS were 44.3 and 49.4 months, respectively. 162 of 289 patients received postoperative adjuvant chemotherapy. 150 (45.9%) patients had confirmed recurrence after curative resection, and 128 (37.6%) had died at the final follow-up. The median follow-up duration was 56.33 months (range 4.9–76.3). The mean number of harvested lymph nodes after surgical resection was 27.5 (range 10–79). The mean age of the patients was 57.6 years (range 29.0–86.0). The clinicopathological characteristics were similar between the two cohorts.
Association of the SIRI with clinicopathological variables
In the training cohort, there were 233 (51.21%) patients in the low SIRI group and 222 (48.79%) patients in the high SIRI group. We found that a high SIRI was significantly associated with older age (p = 0.001), larger tumour (p = 0.004), higher pathological TNM stage (p < 0.001), lymphovascular invasion (p = 0.001), and perineural invasion (p < 0.001). It was not associated with sex, tumour site, Lauren classification, or histological grade (Supplementary Table S1a).
The relationship between SIRI and clinicopathological variables in the validation cohort is shown in Supplementary Table S1b. There were 174 patients (53.21%) in the low SIRI group and 153 patients (46.79%) in the high SIRI group. A high SIRI was also more likely to have older age (p = 0.001), larger tumour (p = 0.030), higher pathological TNM stage (p < 0.001), lymphovascular invasion (p < 0.001), and perineural invasion (p < 0.001). It was also not associated with sex, tumour site, Lauren classification, or histological grade.
DFS analysis
The mean DFS were 83.274 months for patients with SIRI < 0.82 and 52.248 months for patients with SIRI ≥ 0.82 (p < 0.001) (Fig. 1a). The correlation between the SIRI and DFS was further confirmed in the independent validation cohort. We observed that patients with SIRI < 0.82 had a longer mean DFS (59.440 vs. 39.900, p < 0.001; Fig. 1c).
Fig. 1.
The prognostic significance of the SIRI in gastric adenocarcinoma patients undergoing resection in pathological TNM stages I, II, III and early stage subgroups of I in training cohort and validation cohort. Kaplan–Meier analysis of DFS and DSS for the SIRI in pathological TNM stages I, II, III in training cohort (a, b) and validation cohort (c, d). Kaplan–Meier analysis of DFS and DSS for the SIRI in pathological TNM stage I in training cohort (e, f) and validation cohort (g, h)
Univariate analysis of the training cohort indicated that the age, ASA, tumour site, tumour size, pathological TNM stage, histological grade, lymphovascular invasion, perineural invasion, NLR, MLR, SIRI, and postoperative adjuvant chemotherapy all had a statistically significant association with DFS, whereas sex and Lauren classification had no prognostic significance for DFS. A low SIRI was associated with prolonged DFS (HR 2.529; 95% CI 1.922–3.326; p < 0.001; Table 2). Univariate analyses of the validation cohorts confirmed that a low SIRI was associated with better DFS (HR 2.531; 95% CI 1.818–3.525; p < 0.001; Table 2). We also noted that the NLR and MLR correlated with DFS both in training cohort and in validation cohort (Table 2).
Table 2.
Univariate analyses of the SIRI for the prediction of DFS and DSS in patients with gastric adenocarcinoma in training cohort and validation cohort
| Variables | DFS | DSS | ||
|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | |
| Training cohort | ||||
| Age (≥70 vs. <70) | 1.710 (1.265–2.312) | <0.001 | 1.831 (1.347–2.488) | <0.001 |
| Sex (female vs. male) | 0.981 (0.733–1.313) | 0.895 | 0.945 (0.698–1.279) | 0.713 |
| ASA status (≥3 vs. <3) | 2.041 (1.474–2.826) | <0.001 | 2.236 (1.610–3.105) | <0.001 |
| Tumor site | 0.003 | 0.001 | ||
| Upper | Ref | Ref | ||
| Middle | 0.702 (0.480–1.026) | 0.658 (0.446–0.969) | ||
| Lower | 0.557 (0.396–0.782) | 0.524 (0.370–0.741) | ||
| Tumor size (≥5 vs. <5) | 2.024 (1.547–2.647) | <0.001 | 2.066 (1.565–2.729) | <0.001 |
| TNM stage | <0.001 | <0.001 | ||
| I | Ref | Ref | ||
| II | 2.133 (1.192–3.818) | 2.414 (1.280–4.552) | ||
| III | 7.243 (4.330–12.116) | 8.219 (4.663–14.488) | ||
| Lauren classification | 0.157 | 0.182 | ||
| Intestinal | Ref | Ref | ||
| Diffuse | 1.435 (0.988–2.085) | 1.377 (0.943–2.011) | ||
| Mixed | 1.123 (0.823–1.533) | 0.938 (0.671–1.310) | ||
| Histological grade (G3/signet ring cell/mucinous vs. G1/G2) | 1.357 (1.034–1.782) | 0.027 | 1.274 (0.963–1.686) | 0.089 |
| Lymphovascular invasion (Yes vs. No) | 1.884 (1.441–2.463) | <0.001 | 1.938 (1.471–2.552) | <0.001 |
| Perineural invasion (Yes vs. No) | 2.157 (1.606–2.897) | <0.001 | 2.237 (1.643–3.045) | <0.001 |
| NLR (≥2.1 vs. <2.1) | 2.080 (1.588–2.724) | <0.001 | 2.264 (1.708–.3.001) | <0.001 |
| MLR (≥0.22 vs. <0.22) | 1.835 (1.397–2.410) | <0.001 | 1.963 (1.476–2.609) | <0.001 |
| SIRI (≥0.82 vs. <0.82) | 2.529 (1.922–3.326) | <0.001 | 2.692 (2.022–3.585) | <0.001 |
| Fua (No vs. Yes) | 1.532 (1.174–2.000) | 0.002 | 1.744 (1.325–2.294) | <0.001 |
| Validation Cohort | ||||
| Age (≥70 vs. <70) | 1.598 (1.109–2.302) | 0.011 | 1.699 (1.151–2.506) | 0.007 |
| Sex (Female vs. Male) | 1.280 (0.906–1.807) | 0.160 | 1.199 (0.823–1.748) | 0.344 |
| ASA status (≥3 vs. <3) | 2.390 (1.661–3.441) | <0.001 | 2.505 (1.707–3.676) | <0.001 |
| Tumor site | <0.001 | 0.001 | ||
| Upper | Ref | Ref | ||
| Middle | 0.976 (0.637–1.496) | 1.052 (0.662–1.672) | ||
| Lower | 0.515 (0.340–0.781) | 0.534 (0.339–0.841) | ||
| Tumor size (≥5 vs. <5) | 1.967 (1.412–2.741) | <0.001 | 2.239 (1.551–3.233) | <0.001 |
| TNM stage | <0.001 | <0.001 | ||
| I | Ref | Ref | ||
| II | 2.901 (1.565–5.380) | 4.287 (2.056–8.940) | ||
| III | 6.902 (3.937–12.099) | 8.415 (4.227–16.753) | ||
| Lauren classification | 0.237 | 0.184 | ||
| Intestinal | Ref | Ref | ||
| Diffuse | 1.327 (0.871–2.020) | 1.367 (0.867–2.154) | ||
| Mixed | 1.313 (0.900–1.915) | 1.394 (0.931–2.087) | ||
| Histological grade (G3/signet ring cell/mucinous vs. G1/G2) | 1.343 (0.960–1.881) | 0.084 | 1.522 (1.050–2.207) | 0.026 |
| Lymphovascular invasion (Yes vs. No) | 2.486 (1.799–3.437) | <0.001 | 2.378 (1.676–3.374) | <0.001 |
| Perineural invasion (Yes vs. No) | 2.605 (1.835–3.698) | <0.001 | 2.825 (1.931–4.171) | <0.001 |
| NLR (≥2.1 vs. <2.1) | 1.906 (1.377–2.637) | <0.001 | 1.988 (1.396–2.830) | <0.001 |
| MLR (≥0.22 vs. <0.22) | 1.868 (1.335–2.612) | <0.001 | 1.882 (1.306–2.711) | 0.001 |
| SIRI (≥0.82 vs. <0.82) | 2.531 (1.818–3.525) | <0.001 | 2.573 (1.792–3.696) | <0.001 |
| Fua (no vs. yes) | 1.607 (1.159–2.230) | 0.004 | 1.980 (1.391–2.819) | <0.001 |
DFS disease-free survival, DSS disease-special survival, HR hazard ratio, CI confidence interval, ASA American Society of Anesthesiology, NLR neutrophil lymphocyte ratio, MLR monocyte lymphocyte ratio, SIRI Systemic Inflammation Response Index
a Fluoropyrimidine-based adjuvant chemotherapy, mostly including capecitabine plus platinum, capecitabine alone, or S1 (combined tegafur, gimeracil, and oteracil), in patients at advanced stage or early stage tumors with lymph node metastasis in this retrospective study. Cox proportional hazard-rate model in patients at advanced stage or early stage tumors with lymph node metastasis in this retrospective study
The multivariate analysis of the training cohort suggested, after adjusting for other variables, pretreatment SIRI was independently associated with DFS (HR 1.753; 95% CI 1.187–2.588; p = 0.005; Table 3). Multivariate analyses of the validation cohort also confirmed that the SIRI was associated with DFS (HR 2.050; 95% CI 1.297–3.241; p = 0.002; Table 3). Whereas, the prognostic significance of NLR and MLR could not be verified in the multivariate analysis (Table 3).
Table 3.
Multivariate analyses of the SIRI for the prediction of DFS and DSS in patients with gastric adenocarcinoma in training cohort and validation cohort
| Variables | DFS | DSS | ||
|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | |
| Training cohort | ||||
| Age (≥70 vs. <70) | 1.183 (0.812–1.725) | 0.381 | 1.095 (0.745–1.610) | 0.644 |
| ASA status (≥3 vs. <3) | 1.491 (1.000–2.224) | 0.050 | 1.551 (1.031–2.333) | 0.035 |
| Tumor site | 0.293 | 0.577 | ||
| Upper | Ref | Ref | ||
| Middle | 0.936 (0.634–1.384) | 0.883 (0.594–1.314) | ||
| Lower | 0.764 (0.529–1.103) | 0.818 (0.562–1.191) | ||
| Tumor size (≥5 vs. <5) | 0.919 (0.682–1.238) | 0.579 | 0.976 (0.718–1.326) | 0.875 |
| TNM stage | <0.001 | <0.001 | ||
| I | Ref | Ref | ||
| II | 1.402 (0.688–2.857) | 1.498 (0.710–3.159) | ||
| III | 4.619 (2.370–9.005) | 4.739 (2.356–9.532) | ||
| Histological grade (G3/signet ring cell/mucinous vs. G1/G2) | 1.455 (1.095–1.932) | 0.010 | NA | |
| Lymphovascular invasion (Yes vs. No) | 1.237 (0.922–1.660) | 0.156 | 1.281 (0.950–1.728) | 0.104 |
| Perineural invasion (Yes vs. No) | 1.162 (0.842–1.604) | 0.362 | 1.229 (0.881–1.714) | 0.225 |
| NLR (≥2.1 vs. <2.1) | 1.091 (0.777–1.532) | 0.616 | 1.166 (0.821–1.656) | 0.391 |
| MLR (≥0.22 vs. <0.22) | 0.974 (0.692–1.372) | 0.882 | 0.945 (0.663–1.347) | 0.755 |
| SIRI (≥0.82 vs. <0.82) | 1.753 (1.187–2.588) | 0.005 | 1.804 (1.208–2.695) | 0.004 |
| Fua (No vs. Yes) | 1.404 (1.051–1.875) | 0.022 | 1.606 (1.191–2.167) | 0.002 |
| Validation Cohort | ||||
| Age (≥70 vs. <70) | 0.909 (0.590–1.400) | 0.664 | 1.020 (0.647–1.609) | 0.931 |
| ASA status (≥3 vs. <3) | 1.947 (1.301–2.915) | 0.001 | 1.995 (1.311–3.035) | 0.001 |
| Tumor site | 0.212 | 0.073 | ||
| Upper | Ref | Ref | ||
| Middle | 1.129 (0.717–1.777) | 1.295 (0.791–2.122) | ||
| Lower | 0.783 (0.500–1.226) | 0.763 (0.466–1.249) | ||
| Tumor size (≥5 vs. <5) | 1.112 (0.766–1.615) | 0.576 | 1.081 (0.718–1.627) | 0.710 |
| TNM stage | <0.001 | <0.001 | ||
| I | Ref | Ref | ||
| II | 2.522 (1.135–5.601) | 2.707 (1.146–6.393) | ||
| III | 5.166 (2.383–11.201) | 4.530 (1.951–10.522) | ||
| Histological grade (G3/signet ring cell/mucinous vs. G1/G2) | NA | Ref 1.687 (1.117–2.547) | 0.013 | |
| Lymphovascular invasion (Yes vs. No) | 1.386 (0.938–2.048) | 0.101 | 1.264 (0.835–1.914) | 0.268 |
| Perineural invasion (Yes vs. No) | 1.196 (0.765–1.868) | 0.433 | 1.190 (0.732–1.934) | 0.483 |
| NLR (≥2.1 vs. <2.1) | 0.752 (0.481–1.177) | 0.212 | 0.833 (0.523–1.326) | 0.441 |
| MLR (≥0.22 vs. <0.22) | 1.255 (0.834–1.888) | 0.275 | 1.090 (0.699–1.698) | 0.704 |
| SIRI (≥0.82 vs. <0.82) | 2.050 (1.297–3.241) | 0.002 | 2.325 (1.411–3.831) | 0.001 |
| Fua (no vs. yes) | 1.933 (1.352–2.765) | <0.001 | 2.447 (1.656–3.617) | <0.001 |
DFS disease-free survival, DSS disease-special survival, HR hazard ratio, CI confidence interval, ASA American Society of Anesthesiology, NLR neutrophil lymphocyte ratio, MLR monocyte lymphocyte ratio, SIRI Systemic Inflammation Response Index
aFluoropyrimidine-based adjuvant chemotherapy, mostly including capecitabine plus platinum, capecitabine alone, or S1 (combined tegafur, gimeracil, and oteracil), in patients at advanced stage or early stage tumors with lymph node metastasis in this retrospective study. Cox proportional hazard-rate model in patients at advanced stage or early stage tumors with lymph node metastasis in this retrospective study
DSS analysis
The mean DSS was 87.861 months for patients with SIRI < 0.82 and 58.094 months for patients with SIRI ≥ 0.82 (p < 0.001) (Fig. 1b). The correlation between the SIRI and DSS was further confirmed in the independent validation cohort. We observed that patients with SIRI < 0.82 had a better mean DSS (63.543 vs. 48.459; p < 0.001; Fig. 1d).
Univariate analysis of the training cohort indicated that the age, ASA, tumour site, tumour size, pathological TNM stage, lymphovascular invasion, perineural invasion, NLR, MLR, SIRI, and postoperative adjuvant chemotherapy all had a statistically significant association with DSS, whereas sex, histological grade, and Lauren classification had no prognostic significance for DSS. A low SIRI was associated with prolonged DSS (HR 2.692; 95% CI 2.022–3.585; p < 0.001; Table 2). Univariate analyses of the validation cohorts confirmed that a low SIRI was associated with better DSS (HR 2.573; 95% CI 1.792–3.696; p < 0.001; Table 2). We also noted that, the NLR and MLR correlated with DSS both in training cohort and in validation cohort (Table 2).
In the multivariate analysis we found that pretreatment SIRI of the training cohort was independently associated with DSS (HR 1.804; 95% CI 1.208–2.695; p = 0.004; Table 3). Multivariate analyses of the validation cohort confirmed that the SIRI was associated with DSS (HR 2.325; 95% CI 1.411–3.831; p = 0.001) after adjustment for other characteristics (Table 3). Whereas, the prognostic significance of NLR and MLR could not be verified in the multivariate analysis (Table 3).
The prognostic significance of SIRI in patients in TNM stage
To investigate the prognostic efficiency of the SIRI, the SIRI was reanalysed according to the pathological TNM stage. We stratified patients into TNM I as early-stage diseases and TNM II + III as advanced-stage diseases. In the training cohort, we found that patients with high SIRI subgroups have shorter DFS and DSS in advanced-stage diseases (Supplementary Fig. S2a and S2b, p < 0.001 and p < 0.001, respectively). Even in early-stage diseases, worse DFS and DSS were found to be associated with high SIRI subgroups (Fig. 1e, f; HR 3.015; 95% CI 1.122–8.101; p = 0.021, HR 3.649; 95% CI 1.193–11.161; p = 0.015, respectively).
In the validation cohort, worse DFS and DSS in advanced-stage diseases were further confirmed in high SIRI subgroups (Supplementary Fig. S2c and S2d, p < 0.001 and p < 0.001, respectively). In early-stage diseases, shorter DFS was found to be associated with high SIRI subgroup (Fig. 1g, HR 3.764; 95% CI 1.260–11.240; p = 0.011). A tendency of worse DSS was found to be associated with high SIRI subgroup, in spite of no statistical significance (Fig. 1h, HR 3.604; 95% CI 0.898–14.472; p = 0.053).
Correlation between SIRI and chemotherapy
Precious data suggest that system immune contributes significantly to chemotherapy efficacy. Thus, we evaluated the benefit of 5-Fu-based treatment according to the level of SIRI scores in patients who received postoperative adjuvant chemotherapy. A total of 407 and 289 patients at advanced or early-stage tumors with lymph node metastasis were candidates for postoperative adjuvant chemotherapy in training cohort and validation cohort, respectively. In the training cohort, patients who received postoperative chemotherapy had better DFS and DSS (Supplementary Fig. S3a and S3b, p = 0.002 and p < 0.001, respectively). Similar results were obtained from the training cohort (Supplementary Fig. S3c and S3d, p = 0.004 and p < 0.001, respectively).
In addition, a low SIRI was associated with reduced risk of DFS and DSS in patients who received postoperative adjuvant chemotherapy (Fig. 2a, b, p = 0.001 and p < 0.001, respectively). A low SIRI was also associated with reduced risk of DFS and DSS in the validation cohort (Fig. 2c, d, p = 0.001 and p < 0.001, respectively). In contrast, a high SIRI was not associated with a better DFS and DSS in patients with postoperative adjuvant chemotherapy (Fig. 2e, f, HR 1.174; 95% CI 0.840–1.640; p = 0.347, HR 1.284; 95% CI 0.914–1.802; p = 0.148, respectively). The result of validation cohort further conformed that patients with high SIRI did not have improved DFS and DSS with postoperative adjuvant chemotherapy (Fig. 2g, h, HR 1.262; 95% CI 0.836–1.905; p = 0.267, HR 1.443; 95% CI 0.935–2.225; p < 0.095, respectively).
Fig. 2.
The correlation between the SIRI and postoperative adjuvant chemotherapy in training cohort and validation cohort. Kaplan–Meier analysis of DFS and DSS for postoperative adjuvant chemotherapy in low SIRI in training cohort (a, b) and validation cohort (c, d). Kaplan–Meier analysis of DFS and DSS for postoperative adjuvant chemotherapy in high SIRI in training cohort (e, f) and validation cohort (g, h)
Together, these results indicate that the SIRI influences the efficacy on the administration of 5-Fu-based chemotherapy, patients with low SIRI scores might benefit more from postoperative adjuvant chemotherapy.
Correlation between SIRI and CD44 + CSCs
The correlation between SIRI level and CD44 + CSCs scores was further evaluated. Scatter plot analyses revealed a significant positive correlation between the SIRI and CD44 + CSCs scores (Fig. 3a; R 2 = 0.121; p < 0.001). The rate of CD44 + CSCs were significantly higher in the SIRI ≥ 0.82 group (Fig. 3b; 68.47 vs. 43.78%; p < 0.001).
Fig. 3.
The correlation between the SIRI levels and CSCs in training cohort. a Relationship between the SIRI levels and CSC scores; b CD44 + CSCs rates of patients in the SIRI groups; Kaplan–Meier analysis of DFS (c) and DSS (d) for the SIRI in patients with CD44 + CSCs
In light of the close relationship between CD44 + CSCs and the SIRI, we further explored the prognostic significance of the SIRI in subgroups of patients presenting with CD44 + CSCs. In patients with detectable CD44 + CSCs, patients with an SIRI ≥ 0.82 had higher recurrence rates (68.42 vs. 53.92%; p = 0.019) and a shorter DFS (mean 48.417 vs. 62.875 months; p = 0.010; Fig. 3c) compared to patients with SIRI < 0.82. In terms of DSS, we found that the DSS rates were lower in the SIRI ≥ 0.82 group than in the SIRI < 0.82 group (50.00 vs. 66.45%; p = 0.009) (mean 54.661 vs. 69.444 months; p = 0.005; Fig. 3d).
Discussion
Systemic immune and inflammatory cells such as neutrophils, monocytes, and lymphocytes have prognostic value in many types of cancer, including gastric cancer (Li et al. 2017; Eo et al. 2015; Chen et al. 2016), as do the neutrophil/lymphocyte ratio (NLR) and the monocyte/lymphocyte ratio (MLR) (Ock et al. 2017; Wang et al. 2016; Deng et al. 2015; Zhou et al. 2014a). Recently, a novel Systemic Inflammation Response Index (SIRI) based on neutrophil, monocyte, and lymphocyte counts was developed and shown to be an independent predictor of recurrence and survival for pancreatic cancer patients after surgery (Qi et al. 2016). Its prediction ability was shown to be greater than that of the NLR and MLR and was related to chemotherapy resistance, higher serum inflammatory cytokine/chemokine levels and shorter outcomes. In our study, we confirmed that this preoperative SIRI can be used to predict the survival of patients with gastric adenocarcinoma after curative resection. We found that a high preoperative SIRI indicated a worse prognosis and was associated with poor clinicopathological prognostic factors and chemotherapy resistance. Meanwhile, SIRI calculations, which are based on standard laboratory measurements of total neutrophil, monocyte and lymphocyte counts, are routinely performed in the clinical setting. Thus, there is a potential for the SIRI to be used as a marker for tumour recurrence and chemotherapy resistance, which might provide a powerful test enabling accurate and early decision-making to tailor the most effective therapy according to characteristics of individual tumours.
Cancer-related inflammation is recognized as the seventh hallmark of cancer (Hanahan and Weinberg 2011) and there is increasing evidence that local immune responses and systemic inflammation contribute to the development and progression of malignancies and are associated with the survival of patients with cancer. Systemic inflammation consists of circulating immune cells, circulating cytokines, small inflammatory proteins, and circulating immune cells (Diakos et al. 2014). These mediators and circulating immune and inflammation cells, present and detectable in the systemic circulation, mark the presence and progression of cancer (Gabrilovich et al. 2012). The tumour-derived secretome can condition the bone marrow to increase myelopoiesis. In the context of inflammation, neutrophil and monocyte might be released from the bone marrow to the peripheral blood (Gabrilovich et al. 2012; Shi and Pamer 2011) and have profound consequences on tumour progression. The prognostic and predictive usefulness of circulating neutrophils is apparent as an independent measure or as part of the NLR in many types of cancer, including gastric cancer. Studies have consistently demonstrated decreased survival in patients who present with a high peripheral absolute neutrophil count or NLR (Li et al. 2017; Ock et al. 2017; Wang et al. 2016; Deng et al. 2015). Similarly, high numbers of circulating monocytes have been associated with increased tumour progression and poorer survival in patients with gastric cancers (Deng et al. 2015; Zhou et al. 2014a; Eo et al. 2015). In this study, we have demonstrated that this novel SIRI can predict the survival of patients with gastric cancer who received curative resection, even for patients in pathological TNM stage of I subgroup.
Additionally, we observed a correlation between SIRI and older age, larger tumour, higher pathological TNM stage, lymphovascular invasion, and perineural invasion. Tumor size and cancer stage suggest tumour burden, whereas cancer stage, lymphovascular invasion, and perineural invasion indicate tumour invasion. Larger tumours, high cancer stage, and poor clinicopathological feathers indicate a large tumour burden and a high degree of invasion. Therefore, the SIRI scores may reflect overall tumor burden and may be correlated with clinical outcomes. Therefore, our results may indicate that patients with SIRI scores ≥0.82 would have a worse outcome after curative resection compared with those with SIRI scores <0.82. This finding allows clinicians to potentially improve treatment outcomes by identifying candidates for aggressive therapy.
Local immune and system inflammation response encompasses tumour-derived and host-derived cytokines, small inflammatory protein mediators, and infiltrating immune and inflammation cells. Immune and inflammation cellular infiltrates are present in most tumours. The components of such infiltrates include TAN-related cell types, TAMs-related cell types, mast cells, and T cells. Of these cell types, neutrophils and macrophages are essential components of the inflammatory microenvironment of tumours (Grivennikov and Karin 2010). The recruitment of neutrophils into tumours could be attracted by different chemokines as neutrophil chemoattractants, such as CXCL5 in hepatocellular carcinoma (Zhou et al. 2012), CXCL5 hepatic cholangiocarcinoma (Zhou et al. 2014b), and CXCL6 in gastrointestinal tumours (Gijsbers et al. 2005). The TANs located in human tumours generally have an N2 phenotype driven by TGF-beta (Fridlender et al. 2009) and have important roles on promoting malignant transformation, angiogenesis, and tumour progression and have provided initial clues about the immunoregulatory role in cancer (Galdiero et al. 2013a). The infiltrated monocyte could be controlled by different chemokines as monocyte chemoattractants, such as CCL2, CCL5, CCL7, CCL8 and CXCL1 produced by stromal or tumoral cells (Galdiero et al. 2013a). Likewise, The TAMs located in human tumours generally have an M2 phenotype from signals that are derived from regulatory T cells present in tumours or from the tumour cells themselves (including M-CSF, IL-10 and TGF-β) (Gabrilovich et al. 2012). The TAMs mediate their effects via promoting angiogenesis and lymphangiogenesis, promoting invasion and metastasis, remodelling tissues, and altering responses to chemotherapeutic agents and the presence of these cells also suppress anti-tumoural adaptive immunity protection from chemotherapy-induced apoptosis (Gabrilovich et al. 2012; Galdiero et al. 2013a). More specifically, proinflammatory, tumour-derived soluble factors (e.g., IL-1b, IL-6) and cytokines released by activated T cells (e.g., IL-4, IL-10) initiate immunosuppressive pathways and promote myeloid-derived suppressor cell differentiation into immunosuppressive macrophages (Diakos et al. 2014; Wan et al. 2014; Mantovani et al. 2008).
In the current study, we also evaluated the correlation between CSCs and outcomes in patients who received curative resection. Most importantly, we identified a good association between the SIRI and CD44 + CSCs and patients who had high SIRI scores exhibited a high CD44 + CSCs rate, suggesting that the SIRI can reflect the status of tumour invasion. Several results indicated that neutrophil-secreted factors affect tumour stem cell stemness by targeting oncogenes such as c-myc, p63 (Yu et al. 2017), and stat3 (Jiang et al. 2017). Macrophage-secreted factors such as IL-6 affect tumour stem cell stemness through targeting oncogene stat3 (Wan et al. 2014). Studies have demonstrated that purported CSCs are more resistant to chemotherapy than non-CSCs (Takaishi et al. 2009; Yoon et al. 2014) and stem cells expressing CD44 were considered as tumour stem cells including gastric cancer (Takaishi et al. 2009; Zhang et al. 2016). These finding suggested that the chemotherapy resistance of gastric cancer cells may be a result of CD44 + CSCs which could be the result of a series of events initiated by the inflammatory cytokines released by TANs and TAMs.
These results parallel the well-established association between cancer and host immune and inflammation environments, they lend support through clinical evidence. A better understanding of the role of neutrophils, monocytes, and lymphocytes in cancer will help elucidate the association between cancer and immunity and inflammation. Furthermore, patients with gastric cancer who have a high SIRI might benefit from anti-inflammatory and immuno-therapy after surgery with agents such as Alox5 inhibitors including Zileuton (Zil) (Coffelt et al. 2015), Trabectedin, a new cytotoxic drug to tumour-associated macrophages and circulating monocytes via activation of caspase-8-dependent apoptosis (Barat et al. 2017) and target therapy such as a gamma-secretase inhibitor targeting CD44 + CSCs (Germano et al. 2013).
There are a few limitations of this study. First, this study is a retrospective study in nature, possible selection bias, detection bias, and performance of analysis bias might be confounded. Second, the peripheral blood findings were not compared to the findings of peritumoural and intratumoural inflammation in the primary tumour tissue. Nevertheless, the peripheral blood results provide a novel horizon to understand theories of SIRI in carcinogenesis. Finally, there was some heterogeneity in the treatment used for patients after surgical resection that led to a different clinical prognosis. CSCs are not a fixed cell population, CD44 + CSCs are thought to define human cancer CSCs, including gastric cancer (Takaishi et al. 2009; Zhang et al. 2016). The measurement of CD44 expression in gastric cancer tissue by immunohistochemical analysis might fail to capture some CSC subsets, potentially leading to an underestimation of the CSC levels. Thus, using a marker-independent enrichment method and multi-marker approach for CSC detection will improve the sensitivity of CSC detection and will be helpful for clarifying the true relationship between the SIRI and CSC levels.
Therefore, the SIRI may serve as an “immunologic signature” in gastric cancer and could potentially be used to predict response to chemotherapy, which would allow clinicians to tailor future therapies for those patients who would benefit most from chemotherapy.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
We thank Xiuwen Lan, Hongyu Gao, and Zhiguo Li for their excellent technical assistance. We thank Li Chen, Wenpeng Wang, Shubin Song, and Yimin Wang for data collection and analysis. We thank Chunfeng Li and Hongfeng Zhang for fruitfull help.
Funding
This study was funded by 20122307110022.
Compliance with ethical standards
Conflict of interest
The author(s) declared no potential conflicts of interest.
Ethical approval
This study was approved by the Ethics Committee of Harbin Medical University Cancer Hospital. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.
Informed consent
Informed consent was obtained from all individual participants included in the study.
References
- Bang YJ, Van Cutsem E, Feyereislova A, Chung HC, Shen L, Sawaki A, Lordick F, Ohtsu A, Omuro Y, Satoh T, Aprile G, Kulikov E, Hill J, Lehle M, Rüschoff J, Kang YK, ToGA Trial Investigators (2010) Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial. Lancet 376:687–697 [DOI] [PubMed] [Google Scholar]
- Barat S, Chen X, Cuong Bui K, Bozko P, Götze J, Christgen M, Krech T, Malek NP, Plentz RR (2017) Gamma-secretase inhibitor IX (GSI) impairs concomitant activation of notch and Wnt-beta-catenin pathways in CD44+ gastric cancer stem cells. Stem Cells Transl Med 6:819–829 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen XF, Qian J, Pei D, Zhou C, Røe OD, Zhu F, He SH, Qian YY, Zhou Y, Xu J, Xu J, Li X, Ping GQ, Liu YQ, Wang P, Guo RH, Shu YQ (2016) Prognostic value of perioperative leukocyte count in resectable gastric cancer. World J Gastroenterol 22:2818–2827 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Coffelt SB, Kersten K, Doornebal CW, Weiden J, Vrijland K, Hau CS, Verstegen NJ, Ciampricotti M, Hawinkels LJ, Jonkers J, de Visser KE (2015) IL-17-producing γδ T cells and neutrophils conspire to promote breast cancer metastasis. Nature 522:345–348 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cunningham D, Starling N, Rao S, Iveson T, Nicolson M, Coxon F, Middleton G, Daniel F, Oates J, Norman AR, Upper Gastrointestinal Clinical Studies Group of the National Cancer Research Institute of the United Kingdom (2008) Capecitabine and oxaliplatin for advanced esophagogastric cancer. N Engl J Med 358:36–46 [DOI] [PubMed] [Google Scholar]
- Deng Q, He B, Liu X, Yue J, Ying H, Pan Y, Sun H, Chen J, Wang F, Gao T, Zhang L, Wang S (2015) Prognostic value of pre-operative inflammatory response biomarkers in gastric cancer patients and the construction of a predictive model. J Transl Med 13:66 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Diakos CI, Charles KA, McMillan DC, Clarke SJ (2014) Cancer-related inflammation and treatment effectiveness. Lancet Oncol 15:e493–e503 [DOI] [PubMed] [Google Scholar]
- Eo WK, Jeong DW, Chang HJ, Won KY, Choi SI, Kim SH, Chun SW, Oh YL, Lee TH, Kim YO, Kim KH, Ji YI, Kim A, Kim HY (2015) Absolute monocyte and lymphocyte count prognostic score for patients with gastric cancer. World J Gastroenterol 21:2668–2676 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, Forman D, Bray F (2015) Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 136:E359–E386 [DOI] [PubMed] [Google Scholar]
- Fridlender ZG, Sun J, Kim S, Kapoor V, Cheng G, Ling L, Worthen GS, Albelda SM (2009) Polarization of tumor-associated neutrophil phenotype by TGF-beta: “N1” versus “N2” TAN. Cancer Cell 16:183–194 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gabrilovich DI, Ostrand-Rosenberg S, Bronte V (2012) Coordinated regulation of myeloid cells by tumours. Nat Rev Immunol 12:253–268 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Galdiero MR, Bonavita E, Barajon I, Garlanda C, Mantovani A, Jaillon S (2013a) Tumor associated macrophages and neutrophils in cancer. Immunobiology 218:1402–1410 [DOI] [PubMed] [Google Scholar]
- Galdiero MR, Garlanda C, Jaillon S, Marone G, Mantovani A (2013b) Tumor associated macrophages and neutrophils in tumor progression. J Cell Physiol 228:1404–1412 [DOI] [PubMed] [Google Scholar]
- Germano G, Frapolli R, Belgiovine C, Anselmo A, Pesce S, Liguori M, Erba E, Uboldi S, Zucchetti M, Pasqualini F, Nebuloni M, van Rooijen N, Mortarini R, Beltrame L, Marchini S, Fuso Nerini I, Sanfilippo R, Casali PG, Pilotti S, Galmarini CM, Anichini A, Mantovani A, D’Incalci M, Allavena P (2013) Role of macrophage targeting in the antitumour activity of trabectedin. Cancer Cell 23:249–262 [DOI] [PubMed] [Google Scholar]
- Gijsbers K, Gouwy M, Struyf S, Wuyts A, Proost P, Opdenakker G, Penninckx F, Ectors N, Geboes K, Van Damme J (2005) GCP-2/CXCL6 synergizes with other endothelial cell-derived chemokines in neutrophil mobilization and is associated with angiogenesis in gastrointestinal tumors. Exp Cell Res 303:331–342 [DOI] [PubMed] [Google Scholar]
- Gregory AD, Houghton AM (2011) Tumor-associated neutrophils: new targets for cancer therapy. Cancer Res 71:2411–2416 [DOI] [PubMed] [Google Scholar]
- Grivennikov SI, Karin M (2010) Inflammation and oncogenesis: a vicious connection. Curr Opin Genet Dev 20:65–71 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gupta PB, Chaffer CL, Weinberg RA (2009) Cancer stem cells: mirage or reality? Nat Med 15:1010–1012 [DOI] [PubMed] [Google Scholar]
- Hamerlik P, Lathia JD, Rasmussen R, Wu Q, Bartkova J, Lee M, Moudry P, Bartek J Jr, Fischer W, Lukas J, Rich JN, Bartek J (2012) Autocrine VEGF-VEGFR2-Neuropilin-1 signaling promotes glioma stem-like cell viability and tumor growth. J Exp Med 209:507–520 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144:646–674 [DOI] [PubMed] [Google Scholar]
- Huang L, Liu S, Lei Y, Wang K, Xu M, Chen Y, Liu B, Chen Y, Fu Q, Zhang P, Qin K, Cai Y, Fu S, Ge S, Yuan X (2016) Systemic immune-inflammation index, thymidine phosphorylase and survival of localized gastric cancer patients after curative resection. Oncotarget 7:44185–44193 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jiang YX, Yang SW, Li PA, Luo X, Li ZY, Hao YX, Yu PW (2017) The promotion of the transformation of quiescent gastric cancer stem cells by IL-17 and the underlying mechanisms. Oncogene 36:1256–1264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li TJ, Jiang YM, Hu YF, Huang L, Yu J, Zhao LY, Deng HJ, Mou TY, Liu H, Yang Y, Zhang Q, Li GX (2017) Interleukin-17-producing neutrophils link inflammation stimuli to disease progression by promoting angiogenesis in gastric cancer. Clin Cancer Res 23:1575–1585 [DOI] [PubMed] [Google Scholar]
- Magee JA, Piskounova E, Morrison SJ (2012) Cancer stem cells: impact, heterogeneity, and uncertainty. Cancer Cell 21:283–296 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mantovani A, Allavena P, Sica A, Balkwill F (2008) Cancer-related inflammation. Nature 454:436–444 [DOI] [PubMed] [Google Scholar]
- Nguyen LV, Vanner R, Dirks P, Eaves CJ (2012) Cancer stem cells: an evolving concept. Nat Rev Cancer 12:133–143 [DOI] [PubMed] [Google Scholar]
- Noh SH, Park SR, Yang HK, Chung HC, Chung IJ, Kim SW, Kim HH, Choi JH, Kim HK, Yu W, Lee JI, Shin DB, Ji J, Chen JS, Lim Y, Ha S, Bang YJ, CLASSIC trial investigators (2014) Adjuvant capecitabine plus oxaliplatin for gastric cancer after D2 gastrectomy (CLASSIC): 5-year follow-up of an open-label, randomised phase 3 trial. Lancet Oncol 15:1389–1396 [DOI] [PubMed] [Google Scholar]
- Ock CY, Nam AR, Lee J, Bang JH, Lee KH, Han SW, Kim TY, Im SA, Kim TY, Bang YJ, Oh DY (2017) Prognostic implication of antitumor immunity measured by the neutrophil–lymphocyte ratio and serum cytokines and angiogenic factors in gastric cancer. Gastric Cancer 20:254–462 [DOI] [PubMed] [Google Scholar]
- Qi Q, Zhuang L, Shen Y, Geng Y, Yu S, Chen H, Liu L, Meng Z, Wang P, Chen Z (2016) A novel systemic inflammation response index (SIRI) for predicting the survival of patients with pancreatic cancer after chemotherapy. Cancer 122:2158–2167 [DOI] [PubMed] [Google Scholar]
- Shi C, Pamer EG (2011) Monocyte recruitment during infection and inflammation. Nat Rev Immunol 11:762–774 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Siegel RL, Miller KD, Jemal A (2016) Cancer statistics, 2016. CA Cancer J Clin 66:7–30 [DOI] [PubMed] [Google Scholar]
- Takaishi S, Okumura T, Tu S, Wang SS, Shibata W, Vigneshwaran R, Gordon SA, Shimada Y, Wang TC (2009) Identification of gastric cancer stem cells using the cell surface marker CD44. Stem Cells 27:1006–1020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tsai KS, Yang SH, Lei YP, Tsai CC, Chen HW, Hsu CY, Chen LL, Wang HW, Miller SA, Chiou SH, Hung MC, Hung SC (2011) Mesenchymal stem cells promote formation of colorectal tumors in mice. Gastroenterology 141:1046–1056 [DOI] [PubMed] [Google Scholar]
- Wan S, Zhao E, Kryczek I, Vatan L, Sadovskaya A, Ludema G, Simeone DM, Zou W, Welling TH (2014) Tumor-associated macrophages produce interleukin 6 and signal via STAT3 to promote expansion of human hepatocellular carcinoma stem cells. Gastroenterology 147:1393–1404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang T, Ong CW, Shi J, Srivastava S, Yan B, Cheng CL, Yong WP, Chan SL, Yeoh KG, Iacopetta B, Salto-Tellez M (2011) Sequential expression of putative stem cell markers in gastric carcinogenesis. Br J Cancer 105:658–665 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang SC, Chou JF, Strong VE, Brennan MF, Capanu M, Coit DG (2016) Pretreatment neutrophil to lymphocyte ratio independently predicts disease-specific survival in resectable gastroesophageal junction and gastric adenocarcinoma. Ann Surg 263:292–297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yoon C, Park DJ, Schmidt B, Thomas NJ, Lee HJ, Kim TS, Janjigian YY, Cohen DJ, Yoon SS (2014) CD44 expression denotes a subpopulation of gastric cancer cells in which Hedgehog signaling promotes chemotherapy resistance. Clin Cancer Res 20:3974–3988 [DOI] [PMC free article] [PubMed] [Google Scholar] [Retracted]
- Yu PF, Huang Y, Han YY, Lin LY, Sun WH, Rabson AB, Wang Y, Shi YF (2017) TNFα-activated mesenchymal stromal cells promote breast cancer metastasis by recruiting CXCR2+ neutrophils. Oncogene 36:4824–4890 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang X, Hua R, Wang X, Huang M, Gan L, Wu Z, Zhang J, Wang H, Cheng Y, Li J, Guo W (2016) Identification of stem-like cells and clinical significance of candidate stem cell markers in gastric cancer. Oncotarget 7:9815–9831 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhou SL, Dai Z, Zhou ZJ, Wang XY, Yang GH, Wang Z, Huang XW, Fan J, Zhou J (2012) Overexpression of CXCL5 mediates neutrophil infiltration and indicates poor prognosis for hepatocellular carcinoma. Hepatology 56:2242–2254 [DOI] [PubMed] [Google Scholar]
- Zhou X, Du Y, Xu J, Huang Z, Qiu T, Wang X, Qian J, Zhu W, Liu P (2014a) The preoperative lymphocyte to monocyte ratio predicts clinical outcomes in patients with stage II/III gastric cancer. Tumour Biol 35:11659–11666 [DOI] [PubMed] [Google Scholar]
- Zhou SL, Dai Z, Zhou ZJ, Chen Q, Wang Z, Xiao YS, Hu ZQ, Huang XY, Yang GH, Shi YH, Qiu SJ, Fan J, Zhou J (2014b) CXCL5 contributes to tumor metastasis and recurrence of intra hepatic cholangiocarcinoma by recruiting in filtrative intratumoral neutrophils. Carcinogenesis 35:597–605 [DOI] [PubMed] [Google Scholar]
- Zhu L, Cheng X, Shi J, Lin J, Chen G, Jin H, Liu AB, Pyo H, Ye J, Zhu Y, Wang H, Chen H, Fang J, Cai L, Wang TC, Yang CS, Tu SP (2016) Crosstalk between bone marrow-derived myofibroblasts and gastric cancer cells regulates cancer stemness and promotes tumorigenesis. Oncogene 35:5388–5399 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.