Gastric cancer was the fifth-most common cancer and the fourth most frequent cause of cancer-related mortality worldwide in 2020 (1). The standard treatment for locally advanced gastric cancer is gastrectomy and postoperative adjuvant chemotherapy.

The 5-year survival rate is reported to be 96.7% for early-stage gastric cancer, and 51.9% for advanced gastric cancer (2,3). The postoperative recurrence rate is 50%, and the 5-year survival rate after recurrence is <5%. Advanced gastric cancer still has an unfavorable prognosis. To improve the long-term prognosis of advanced gastric cancer, the risk of recurrence must be predicted, and more aggressive treatment must be given to those at high risk of recurrence.

Since Virchow first proposed the relationship between inflammation and cancer in 1863 (4), it has become clear that systemic inflammation plays an important role in the etiology and progression of cancer (5). Previous studies have shown that the lymphocyte-to-monocyte ratio (LMR), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR) and preoperative serum pre-albumin, could be potentially used as predictors of the prognosis of cancer (5-7). However, these indicators only assess either the inflammatory or nutritional status.

Recently, the systemic inflammation score (SIS) has been reported to have an effect on the postoperative prognosis of colorectal and esophageal cancer (8,9). SIS includes systemic inflammation and the nutritional status. Thus, it is a superior marker than any other single inflammatory or nutritional indicator. However, there are limited studies evaluating the clinical impact of the SIS in gastric cancer.

This study therefore aimed to clarify whether preoperative SIS affects the short-term and long-term oncological outcomes in gastric cancer patients.

Patients. Patients were retrospectively selected from the medical records of consecutive patients diagnosed with primary gastric cancer who underwent gastrectomy at Yokohama City University from 2000 to 2015. The inclusion criteria were as follows: 1) histologically proven adenocarcinoma, according to the 15th edition of the Japanese Classification of Gastric Carcinoma (JCGC) published by the Japanese Gastric Cancer Association (JCGC) (10), 2) complete (R0) resection of gastric cancer with lymphadenectomy (D1 or more) as a primary treatment, 3) stage I-III disease diagnosed pathologically based on the 15^th edition of the JCGC, and 4) a laboratory blood analysis, including serum albumin (Alb) and the white blood cell (WBC) count, in order to determine the SIS within one week before surgery. Patients who were diagnosed with remnant gastric cancer, who had received preoperative chemotherapy, and who had synchronous or multiple cancers, or whose medical records were incomplete or inaccurate were excluded.

Surgical procedure and adjuvant treatment. All patients underwent distal or total gastrectomy with lymph node dissection. D1+nodal dissection was performed for clinical stage IA disease, while D2 dissection was performed for clinical stage ≥IB. Patients with a pathological diagnosis of II or III received one year of postoperative adjuvant chemotherapy. As a rule, patients with pathological stage II disease were treated with S-1 monotherapy, and patients with pathological stage III disease were treated with S-1 in combination with docetaxel or capecitabine plus oxaliplatin.

SIS definition. SIS was calculated by the perioperative Alb level and LMR. According to previous studies, a score of 0 was defined as Alb ≥4.0 g/dl and LMR ≥4.44; a score of 1 was defined as either Alb <4.0 g/dl or LMR <4.44; and a score of 2 was defined as Alb <4.0 g/dl and LMR <4.44. The SIS was evaluated within seven days prior to surgery (11).

Evaluations and statistical analyses. The significance of differences between the SIS and clinicopathological parameters was determined by the χ² test. Overall survival (OS) and recurrence-free survival (RFS) curves were calculated by the Kaplan–Meier method. Univariate and multivariate survival analyses were analyzed using a Cox proportional hazards model. p-Values <0.05 were considered to be statistically significant.

Patients. A total of 258 patients were assessed in this study. Based on the 1-,3- and 5-year OS rates, a cutoff value of 2 for SIS was used for the preset study (Table I). Of the 258 total patients, 171 (66.3%) were classified in the Low-SIS group, and 87 (33.7%) in the High-SIS group.

On comparing background characteristics between the Low- and High-SIS groups, the findings were quite similar. The proportions of males (69.6% vs. 73.6%, p=0.506), individuals with an alcohol habit (68.4% vs. 66.7%, p=0.775), individuals with a smoking habit (65.5% vs. 60.9%, p=0.469), incidence of hypertension (38.6% vs. 40.2%, p=0.799), and morbidity of chronic obstructive pulmonary disease (33.3% vs. 44.8%, p=0.071) were equivalence in the Low- and High-SIS groups, while the median age (66.4 years vs. 72.9 years, p<0.001), morbidity of diabetes mellitus (15.2% vs. 25.3%, p=0.049) and median body mass index were considerably higher in the High-SIS group (22.9 vs. 21.9, p=0.014).

Survival of the Low- and High-SIS groups. The 3- and 5-year OS rates were 83.6% and 78.9%, respectively, in the Low-SIS group and 61.5% and 55.6% in the High-SIS group, which amounted to a statistically significant difference (Figure 1) (p=0.001). Each clinicopathological factor was classified as shown in Table II and evaluated for its prognostic value. The univariate analysis of OS showed pathologic T factor and SIS to be statistically significant prognostic factors. The SIS was set for the final multivariate analysis model [hazard ratio (HR)=1.784, 95% confidence interval (CI)=1.039-3.063, p<0.05]. The respective 3- and 5-year RFS rates were 45.3% and 77.4% in the Low-SIS group and 21.4% and 48.6% in the High-SIS group, showing a statistically significant difference (Figure 2) (p=0.001). Each clinicopathologic factor was categorized and its prognostic value analyzed as shown in Table III. Univariate analysis of RFS showed that pathologic factors and SIS were significant prognostic factors. The SIS was therefore also selected for the final multivariate analysis model (HR=1.710, 95%CI=1.027-2.847, p<0.05). When comparing the site of recurrence, there were significant differences in the rates of hematological recurrence and local recurrence between the High- and Low-SIS groups (Table IV).

Clinical course of adjuvant treatment and postoperative complications between the Low- and High-SIS groups. When comparing the patients who required adjuvant treatment after surgery between the Low- and High-SIS groups, 32.2% (55/171) of the patients were eligible for adjuvant treatment in the Low-SIS group, while 59.8% (52/87) were eligible in the High-SIS group. The difference was statistically significant (p<0.001). On the other hand, when comparing the rate of adjuvant treatment after surgery between the Low- and High-SIS groups, 74.1% (14/40) of patients in the Low-SIS group and 49.0% (25/51) of the patients in the High-SIS group received adjuvant treatment. The difference was statistically significant (p=0.08). The postoperative surgical complication rate was 37.4% (64/171) in the Low-SIS group and 39% (34/87) in the High-SIS group (p=0.796). Regarding the details of postoperative complications, the incidence rates in the Low-SIS vs. High-SIS groups were as follows: postoperative pneumonia, 7% vs. 3.4%; (p=0.247); anastomotic leakage, 7% vs. 11.5% (p=0.224); abdominal abscess, 2.9% vs. 3.4% (p=0.818); and anastomotic stenosis, 1.2% vs. 0% (p=0.672). There were no significant differences.

The objective of this study was to identify the clinical value of the preoperative SIS as a prognostic factor for curative resection of gastric cancer. The major findings were that the preoperative SIS was a significant prognostic factor for gastric cancer patients undergoing curative resection. In addition, the Low-SIS group showed significantly longer OS and RFS than the High-SIS group. Therefore, out results suggested that preoperative SIS may be a useful prognostic factor for patients with resectable gastric cancer.

The HR of SIS for gastric cancer in this study was 1.784 (95% CI=1.039-3.063, p=0.036). Prior studies have shown similar results. For example, Hara et al. evaluated the preoperative clinical impact of SIS in 138 gastric cancer patients undergoing curative treatment (12). They divided patients into an SIS-low group (n=45) and an SIS-high group (n=93) according to the preoperative SIS. The results showed that the 5-year OS rate was 86.6% in the SIS-low group and 66.6% in the SIS-High group (p=0.004). The 5-year RFS rate was 79.1% in the SIS-Low group, whereas it was 58.0% in the SIS-High group (p=0.007). The HR of the SIS was 2.143 (95%CI=1.126-4.078, p=0.020). Moreover, Lin et al. found the importance of preoperative SIS in patients undergoing radical resection of gastric cancer (13). They classified SIS into three groups, score 0 (n=781), score 1 (n=614), and score 2 (n=391). High SIS was associated with significantly higher local recurrence, peritoneal metastasis, and distant metastasis compared to low SIS (all p<0.001). Multivariate analysis also showed that SIS was associated with OS (SIS=1: HR=1.250, p=0.038; SIS=2: HR=1.728, p<0.001) and RFS (SIS=1: HR=1.248, p=0.055; SIS=2: HR=1.648, p<0.001). In addition, Inagaki et al. demonstrated the SIS accurately predicted the prognosis after the radical resection of gastric cancer (14). They categorized their study population into preoperative SIS 0 (n=955), SIS 1 (n=584), and SIS 2 (n=225) groups. They had assigned a cutoff value of 3.4 for the LMR. OS and RFS of patients in preoperative SIS 0, 1, and 2 groups became shorter with higher SIS (p<0.0001), identifying SIS 1 and 2 as independent prognostic factors (HR=1.35, 95%CI=1.06-1.272, p=0.0125 and HR=1.63, 95%CI=1.21-2.19, p=0.0012). Considering these findings, SIS is potentially useful as a prognostic predictor for resectable gastric cancer.

This study found that SIS was associated with OS and RFS in gastric cancer patients who underwent radical resection. There are some possible explanations for our findings. First, the SIS status was related to the introduction of adjuvant treatment after surgery. In the present study, the proportion of patients who required adjuvant chemotherapy was higher in the High-SIS group than that in the Low-SIS group. However, the introduction rate in the High-SIS group was lower than that in the Low-SIS. A similar result was observed in the previous study by Sato et al., which revealed that SIS is a simple predictor of the incidence of postoperative complications and survival in pT2-4 gastric cancer patients (15). They categorized their patients into preoperative SIS 0 (n=78), SIS 1 (n=73), and SIS 2 (n=36) groups. The incidence of advanced cancer in the SIS 0, 1, and 2 groups was 78.2%, 79.5%, and 91.7%, respectively. Postoperative adjuvant chemotherapy for advanced cancer was administered to 73.8%, 55.2%, and 33.3% of the patients in the SIS 0, 1, and 2 groups, respectively, with patients with high SIS values showing lower rates. Taken together, the SIS status may have a clinical impact on the adjuvant chemotherapy course. Second, the SIS status was related to postoperative surgical complications. In the present study, the High-SIS group tended to have a higher rate of elderly patients, obese patients, and patients with diabetes mellitus, who are also known to have a higher risk of postoperative complications. In this study, although the complication rate did differ to a statistically significant extent according to the SIS, all patients for whom adjuvant chemotherapy could not be introduced had postoperative complications; this included 16 patients in the High-SIS group and 9 in the Low-SIS group, which amounted to a significant difference. A similar result was observed in a previous study: Sato et al. evaluated the clinical significance of preoperative SIS on short- and long-term prognosis in pT2-4 gastric cancer patients who underwent radical resection (15). They categorized their patients into preoperative SIS 0 (n=78), SIS 1 (n=73), and SIS 2 (n=36) groups. They clarified that there was a stepwise increase in the prevalence of postoperative complications in proportion to the SIS (p=0.043). In another study, Shoka et al. analyzed a multi-institutional dataset with the aim of identifying a predictor of post-gastrectomy pneumonia (16). They categorized their study population into preoperative SIS 0 (n=565), SIS 1 (n=586), and SIS 2 (n=264) groups. The prevalence of postoperative pneumonia gradually increased in parallel with the SIS, as did the overall complication rate (SIS 0, 20.4%; SIS 1, 20.8%, and SIS 2, 31.1%; p=0.001). Taken together, the SIS status was demonstrated to be correlated with the incidence of postoperative complications.

This study is accompanied by several limitations. First, it was a single-site study. Second, because it is retrospective, it may be subject to selection bias. Third, the mechanism of this study is not clarified. The interrelationships among nutritional status, immune status, and the dynamics of cytokines and related proteins need to be further evaluated. Despite these limitations, our study demonstrates that the preoperative SIS is a simple prognostic predictor for gastric cancer.

In conclusion, our study showed that the preoperative SIS could predict tumor recurrence and survival of gastric cancer patients and can help predict the long-term outcomes and determine the indications for postoperative adjuvant chemotherapy.

The Authors declare no conflicts of interest in association with the present study.

MJ and TA made substantial contributions to the concept and design. MJ, TA, MF, TI, KK, KK2 (Keisuke Kazama), SS, HT, NY and YR made substantial contributions to the acquisition of data and their analysis and interpretation. MJ, TA, HT, NY and YR were involved in drafting the article or revising it critically for important intellectual content. MJ and TA gave their final approval of the version to be published.

This work was supported by JSPS KAKENHI Grant Number 21K08688.