Abstract
Background/Aim: The present study aimed to compare the utility of various inflammatory marker- and nutritional status-based prognostic factors, including many previous established prognostic factors, for predicting the prognosis of stage IV gastric cancer patients undergoing non-curative surgery. Patients and Methods: A total of 33 patients with stage IV gastric cancer who had undergone palliative gastrectomy and gastrojejunostomy were included in the study. Univariate and multivariate analyses were performed to evaluate the relationships between the mGPS, PNI, NLR, PLR, the CONUT, various clinicopathological factors and cancer-specific survival (CS). Results: Among patients who received non-curative surgery, univariate analysis of CS identified the following significant risk factors: chemotherapy, mGPS and NLR, and multivariate analysis revealed that the mGPS was independently associated with CS. Conclusion: The mGPS was a more useful prognostic factor than the PNI, NLR, PLR and CONUT in patients undergoing non-curative surgery for stage IV gastric cancer.
The aim of palliative surgery for stage IV gastric cancer is to relieve distressing symptoms, including cancer pain, tumor hemorrhage and inability to eat (1). However, it is controversial whether non-curative surgical treatment improves the prognosis of patients with incurable stage IV gastric cancer. The prognosis of cancer patients is determined by many factors, among which tumor- and patient-related factors are particularly important; their use as prognostic factors remains problematic due to the wide range of tumor- and patient-related factors and because the results are subject to bias. Thus, in order to establish the optimal medical treatment and determine the optimal timing of surgery it is crucial to identify effective markers of the immunological and nutritional status that can be used to predict the prognosis. Furthermore, it is necessary to find factors that can accurately predict the response to surgical treatment in patients with stage IV gastric cancer in order to identify the patients who are more likely to benefit from non-curative surgery.
In the past few decades, investigators have demonstrated that the presence of a systemic inflammatory response and the preoperative immunonutritional status are associated with a poor prognosis in patients with various types of cancer (2-5). The Glasgow Prognostic Score (GPS) (5-12), the Prognostic Nutritional Index (PNI) (13-15), the neutrophil lymphocyte ratio (NLR) (10, 16-21), the platelet lymphocyte ratio (PLR) (21-24) and the controlling nutritional status (CONUT) (25-27) have been reported to have prognostic value in patients with many types of cancer. Some studies have demonstrated that these inflammatory and nutritional factors may predict the prognosis of patients with advanced gastric cancer (10-12, 17-22). However, there have been no reports on the factors that are most useful for predicting the prognosis of patients with stage IV gastric cancer after non-curative surgery, including gastrectomy and gastrojejunostomy. Thus, in the present study, we investigated the usefulness of markers of inflammation and the nutritional status for predicting the prognosis of patients undergoing non-curative surgery for incurable stage IV gastric cancer.
Patients and Methods
A total of 271 gastric cancer patients underwent surgical resection or gastrojejunostomy in our Institution between January 2006 and December 2016. Among these patients, 49 patients with stage IV gastric cancer were enrolled in the present study. This study was approved by the ethics committee of our hospital (IRB No. 201704). All of the medical records were reviewed retrospectively. Patients who died within 30 days of surgery and who had other malignancies were excluded from the study. We excluded 16 patients for the following reasons: death within 30 days (n=2), concurrent malignancies (n=3) and incomplete clinical data (n=11). The remaining 33 cases included in this study had adequate clinical information and follow-up data. Non-curative surgery, including gastrectomy and gastrojejunostomy, was defined as described previously (12). The pathological diagnoses and classifications were made according to the 7th edition of the UICC TNM Classification of Malignant Tumors (28).
Blood samples were obtained within a week before surgery to measure the white blood cell, neutrophil and lymphocyte counts, and the C-reactive protein (CRP), albumin, total cholesterol, carcinoma embryonic antigen (CEA) and carbohydrate antigen (CA) 19-9 levels. Patients were evaluated using the following values: the modified GPS (mGPS) (patients with both an elevated CRP level [>0.5 mg/dl] and hypoalbuminemia [<3.5 g/dl] were assigned an mGPS of 2; patients with one of these blood chemistry abnormalities were assigned a score of 1; and those with no abnormalities were assigned a score of 0) (29), the PNI (10×serum albumin level [g/dl] + 0.005 × peripheral lymphocyte count [/mm3]) (30), the NLR (neutrophils/lymphocytes [mm3/mm3]), the PLR (platelet/lymphocytes [mm3/mm3]) and the CONUT (a score composed of the serum albumin concentration, the total peripheral lymphocyte count, and the total cholesterol concentration) (31).
We established effective cutoff levels for the different indexes in which a large amount of difference was observed between two groups. For example, the mGPS was tested at 0-1 and 2, or 0 and 1-2; the PNI was tested at set cutoff levels of 40, 45, 50; the NLR was tested at set cutoff levels of 2.0, 2.5, 3.0, 3.5, 4.0; and the PLR was tested at set cutoff levels of 150, 200, 250; the CONUT was tested at cutoff levels of ≤3 vs. ≥4, or ≤4 vs. ≥5, or ≤5 vs. ≥6. Then, we divided each group based on the preoperative mGPS (0–1 vs. 2), PNI (>40 vs. ≤40), NLR (<2.5 vs. ≥2.5), PLR (≤200 and >200), CONUT (≤4 vs. ≥5).
The potential prognostic factors for advanced gastric cancer were as follows: age (<70 years vs. ≥70 years); sex (male vs. female); body mass index (<22 vs. ≥22); Eastern Cooperative Oncology Group performance status (0-1 vs. ≥2); American Society of Anesthesiologists (ASA) physical status (0-1 vs. ≥2); histological subtype (well and moderately differentiated adenocarcinoma vs. poorly differentiated and undifferentiated adenocarcinoma); tumor depth (≤T3 vs. T4); number of metastatic sites (≤1 vs. ≥2), metastatic sites (such as the liver, lung, and peritoneum); peritoneal cytology (positive vs. negative); CEA level (<5 vs. ≥5); CA19–9 level (≤37 vs. >37); mGPS (0-1 vs. 2); PNI (>40 vs. ≤40); NLR (<2.5 vs. ≥2.5); PLR (≤200 and >200); CONUT (≤4 vs. ≥5); surgical treatment (gastrectomy vs. gastrojejunostomy); chemotherapy (performed vs. not performed); and the presence or absence of postoperative complications (defined as Grade >II according to the Clavien-Dindo classification (32)). The associations between the mGPS, PNI, NLR, PLR, CONUT, the clinicopathological parameters and overall survival (OS) were assessed.
Intergroup comparisons were performed using the Mann-Whitney U-test for continuous and ordinal variables, and the chi-squared test and Fisher's exact test for categorical variables. The OS rates were calculated using the Kaplan–Meier method, and the differences in survival rates were compared using the log-rank test. OS was defined as the time from the first day of surgical treatment until death. Both univariate and multivariate analyses were performed using Cox's proportional hazards regression to assess the potential prognostic factors. p-Values of <0.05 were considered to indicate statistical significance in all of the analyses. All of the statistical analyses were performed using the IBM SPSS Statistics for Windows software program (Version 22.0, IBM Corporation, Armonk, NY, USA).
Results
Patient characteristics. Table I summarizes the background information of the patients with each prognostic factor. The rates of liver and peritoneal metastasis were significantly lower, whereas the incidence of gastrectomy in the PNI >40 group was significantly higher than that in the PNI ≤40 group. All patients in the PNI >40 group underwent gastrectomy. The rate of peritoneal metastasis in the NLR ≥2.5 group was significantly higher than that in the NLR <2.5 group and the number of metastatic sites was significantly increased in the NLR ≥2.5 and PLR >200 groups. The incidence of chemotherapy in the mGPS (0-1) and NLR <2.5 groups was significantly higher than that in the mGPS (2) and NLR ≥2.5 groups, respectively. There were no significant differences in the clinicopathological factors of the CONUT ≤4 and CONUT ≥5 groups.
The comparison of the survival rates. Thirty-one patients died and two were censored at the last date of follow-up. The median OS of all cases was 193 days. The OS rates of the mGPS (0-1) (p=0.0001), PNI >40 (p=0.047), and NLR <2.5 (p=0.013) groups were significantly better than those of their counterparts (Figure 1a, b and c).
Analysis of prognostic factors for OS. To obtain the adjusted hazards ratio for survival, a multivariate Cox proportional hazards model was fitted to the data after the backward elimination of candidate variables. Table II shows the relationships between the clinicopathological factors and OS among patients with incurable stage IV gastric cancer who underwent non-curative surgical treatment. The univariate analyses revealed significant differences in chemotherapy (p=0.001), mGPS (p=0.0001) and NLR (p=0.012) while a multivariate analysis of the clinicopathological factors that showed significant differences in the univariate analyses revealed that mGPS (HR=0.371, 95%CI=0.181-0.76, p=0.007) was the only independent prognostic factor that was significantly associated with OS (Table II).
Discussion
Serum albumin concentration is well known to be a reliable indicator of the nutritional status and state of systemic inflammation (33). A low serum albumin level is considered to be associated with various cancer survival outcomes. On the other hand, serum CRP elevation, which indicates the presence of systemic inflammation, has been shown to be an independent prognostic factor in gastric cancer (34). The GPS combines these two variables (the serum albumin level and the CRP level). Thus, the GPS is considered useful as it reflects the degree of malnutrition and inflammation, which are influenced by the development of cancer. Several studies have reported that the GPS is a useful prognostic factor in gastric cancer (8-12). Nozoe et al. (8) reported that a high GPS was independently associated with a worse prognosis in gastric cancer patients undergoing curative resection. Kubota et al. (9) reported that the GPS was a significant predictor of short- and long-term survival in patients with stage I–III gastric cancer who were undergoing curative surgery. We also previously reported that the mGPS was a significant prognostic marker in patients undergoing palliative surgery for stage IV gastric cancer (12).
Wang et al. (11) revealed that the GPS was superior to the NLR and PLR and that it was associated with OS in stage III gastric cancer patients undergoing potentially curative resection. Liu et al. (6) compared the prognostic value of different nutrition- and inflammation-based markers using an ROC analysis. They revealed that the C-reactive protein/albumin ratio was the best predictor of the survival of gastric cancer patients after curative resection. These results suggested that the inflammatory marker, which consisted of serum albumin and CRP, is better prognostic factor than the other inflammation and nutritional makers in patients with gastric cancer undergoing curative resection. However, there have been no reports of factors that predicted better survival in stage IV gastric cancer patients undergoing non-curative surgery. The results of the present study suggest that the mGPS may be a more useful prognostic factor than the PNI, NLR, PLR and CONUT for such patients.
The PNI was initially designed by Buzby et al. (35) to assess the immunological and nutritional aspects of patients who underwent surgical treatment for diseases of the digestive tract. Onodera et al. (30) proposed the modified PNI, which was calculated using the two values: the serum albumin concentration and the lymphocyte count in the peripheral blood. Recently, the PNI has been widely used to assess the preoperative condition and predict the surgical risk (including anastomotic deficiency) in patients undergoing surgery for gastrointestinal malignancies. Migita et al. (13) revealed that the overall and relapse-free survival rates in a PNI-low group were significantly lower than those in a PNI-high group among patients with stage I and stage III gastric cancer. A meta-analysis revealed that the PNI was significantly associated with poor OS in patients with stage I, II and III gastric cancer, but not in patients with stage IV gastric cancer (14). These results suggest that a low preoperative PNI has little impact on survival in patients with stage IV gastric cancer. In the present study, although the OS rate of PNI >40 group was significantly better than the PNI ≤40 group, the PNI was not an independent prognostic indicator after non-curative surgery in patients with stage IV gastric cancer from a result of multivariable analysis.
It is generally known that lymphopenia is a surrogate of impaired cell-mediated immunity, whereas neutrophilia is a response to systemic inflammation. The NLR, which is calculated as the neutrophilia count divided by the lymphocyte count, has been suggested as a marker of the immune response to systemic inflammation in patients with various malignancies. It is thought that a high NLR reflects increased systemic inflammation in the host and that it is associated with a poor prognostic outcome. A meta-analysis to investigate the association between the NLR and the prognosis of gastric cancer revealed that a high preoperative NLR is associated with poor survival in patients with gastric cancer (16). Several studies have demonstrated that the NLR is a prognostic and predictive biomarker in patients with advanced gastric cancer (17-21). Yamanaka et al. (17) revealed that the NLR is an independent prognostic factor in advanced gastric cancer patients undergoing palliative chemotherapy. Jung et al. (18) revealed that an elevated pre-operative NLR predicted poor OS following resection in patients with stage III–IV gastric cancer. Jin et al. (19) revealed that the NLR before chemotherapy and surgery were independent prognostic factors for progression-free survival in patients with stage III–IV gastric cancer undergoing neoadjuvant chemotherapy. However, to the best of our knowledge, the prognostic significance of the NLR in stage IV gastric cancer patients undergoing palliative surgery has rarely been studied. In the present study, although the NLR was a useful prognostic marker, it was not an independent prognostic factor for such patients.
The PLR is a prognostic marker in several types of malignancies (24). The relationship between the PLR and a poor prognosis may be explained through inflammatory processes caused by cancer cells. Systemic inflammation results in the release of various immunological mediators such as interleukin (IL)-1 and IL-6, which accelerate megakaryocyte proliferation leading to thrombocytosis (36). Thrombocytosis is considered to be a negative prognostic marker in gastric cancer (37). Meanwhile, lymphocytes play an important role in immune-surveillance for cancer, and prevent tumor growth (38). Lymphocytopenia has been reported to be associated with a poorer prognosis in patients with gastrointestinal malignancies, including gastric cancer (39). Thus, thrombocytosis and lymphocytopenia are both correlated with the degree of systemic inflammation in the host and are associated with the prognosis of tumor growth. However, in the present study, the PLR was not a useful prognostic marker in stage IV gastric cancer patients who underwent non-curative surgery.
The CONUT, which is calculated by the serum albumin concentration, the total peripheral lymphocyte count, and the total cholesterol concentration, was developed as a screening tool to allow the early detection of a poor nutritional status (31). In contrast to the GPS, PNI, NLR and PLR, the total cholesterol concentration is included in the calculation of the CONUT. Cholesterol is an essential component of the cell membrane that is involved in many of the biochemical pathways that are potentially correlated with the initiation and progression of cancer and the immune response (26). Several studies have suggested that a low serum total cholesterol level is associated with mortality in gastrointestinal duct cancer (40, 41). To date, a few studies have reported the CONUT as an independent prognostic factor in colorectal cancer (25) and esophageal cancer (26). However, there have been no reports on the relationship between the CONUT and the prognostic outcome in gastric cancer. The present study is the first report to evaluate the relationship between the CONUT and the survival of patients with stage IV gastric cancer. Unfortunately, there was no significant difference in the survival of stage IV gastric cancer patients with a CONUT of <4 and those with a CONUT of ≥5.
The present study is associated with several limitations. First, it was a retrospective study that was conducted at a single institution and included a small number of patients. Furthermore, a number of patients had to be excluded due to insufficient data. Thus, well-designed prospective studies that include a large number of advanced gastric cancer patients are required. Moreover, further investigations are necessary to determine whether the cut-off mGPS, PNI, NLR, PLR and CONUT value were correct.
In conclusion, the mGPS and NLR were useful prognostic factors that predicted postoperative survival in patients undergoing non-curative surgical treatment for stage IV gastric cancer. Moreover, the preoperative mGPS was the most useful prognostic factor for such patients.
Footnotes
Conflicts of Interest
The Authors have no conflicts of interest to declare.
- Received June 11, 2017.
- Revision received June 28, 2017.
- Accepted July 29, 2017.
- Copyright© 2017, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved