Introduction

Lymph node status is an important indicator of survival in patients with gastric adenocarcinoma (GA) (1). Despite the fact that malignant cells are limited to mucosa and submucosa regardless of lymph node status in early gastric cancer, the predisposing factors for lymph node metastasis are still investigated for providing better outcome and survival rate (2). Since early GA has good prognosis, routine screening is not performed in most countries, therefore, GA is usually diagnosed at advanced stage. Surgical team is mostly unaware of the exact lymph node status in GA patients preoperatively despite imaging studies. Multiple factors, such as extent of node dissection and experience of the pathologist may affect the pathological lymph node status in patients with GA (1). Generally, dissection of 16 or more regional nodes is recommended to determine the correct nodal status, but this is not a strict surgical criterion (1). Some studies reported that D2 dissection, compared to D1 dissection, and dissecting more than 15 lymph nodes were favorable for optimal surgical outcome (3,4). Retrieval of less than 15 lymph nodes is considered an inadequate surgery and lymph node ratio is supplemental for nodal staging (3).

In this study; we aimed to evaluate the effect of different nodal systems on survival in patients with GA who underwent D2 lymph node dissection with more than 15 lymph nodes harvested.

Methods

The study was approved by Clinical Research Ethics Committee of Mersin University (no: 2016/185). Informed consent was not obtained from the patients due to the retrospective nature of the study. This retrospective study evaluated GA patients who underwent curative gastrectomy in our tertiary care center from January 2001 to January 2015. Only the resectable tumors arising from the gastric mucosa and tumors that arise from the first cranial 5 cm of gastric mucosa without crossing the esophagogastric junction were included. Patients with metastatic tumors, patients who underwent palliative interventions, patients having gastric cancer other than adenocarcinoma and patients with dissection of fewer than 16 lymph nodes were excluded from the study. One hundred and sixty-one patients who fulfilled the inclusion criteria were enrolled. Fifteen patients were also excluded from the study due to death within 30 days after surgery (5). The remaining 146 patients were evaluated in detail for survival. No patients had known metastatic disease, ascites and neoadjuvant chemoradiotherapy. All surgeries were performed by the same surgical team. Adjuvant chemotherapy indication (combination of 5-fluorouracil with cisplatin, docetaxel or none) was decided by the multidisciplinary council (department of general surgery, medical oncology, gastroenterology, pathology and radiology) for each patient.

Patients’ data were collected from the database. The patients were evaluated with regard to age, gender, site and size of the tumor, extent of gastrectomy, complications of the surgery, stage and grade of the tumor, nodal staging according to different nodal evaluation systems, and presence of adjuvant chemotherapy. Effects of these variables on survival were investigated.

All the patients underwent endoscopy to determine the location of the tumor and pathological samples were collected preoperatively; either computed tomography scan or magnetic resonance imaging were mainly performed to determine the stage after the pathological diagnosis. Primary tumor site was defined as “distal 1/3”, “middle 1/3”, “proximal 1/3” and “linitis plastic”. Extent of the surgery was decided according to tumor location; subtotal gastrectomy was performed with safety margin, especially for distal tumors. Total gastrectomy was mostly performed in all proximal-middle location tumors and linitis ones. Addition of “splenectomy” to the procedure was performed due to extent of surgery or iatrogenic laceration. The tumors were divided according to their grade and type into 3 categories: well-differentiated, moderately-differentiated and poorly differentiated/signet-ring cell type’. Status of lymphovascular invasion (LVI) and perineural invasion (PNI) were also evaluated. T stage was defined/redefined (6). For nodal evaluation, different lymph node staging systems were used. Total number of harvested lymph nodes and metastatic ones were calculated separately. N stage was defined according to the 8th edition of the Tumor, node and metastasis classification (6). Secondly, the number of metastatic lymph nodes and the total number of harvested lymph nodes were calculated and the ratio of metastatic to total retrieved nodes (MLR) was categorized as “MLR 0 (0%)”, “MLR 1 (1-5%)”, “MLR 2 (6-10%)”, “MLR 3 (11-20%)”, “MLR 4 (21-30%)” and “MLR 5 (>30%)” (7). Finally, percentage of node positivity (PNP) was defined as “PNP 0 (0%)”, “PNP 1 (1-20%)”, “PNP 2 (21-50%)” and “PNP 3 (>50%)” (3). Stage (1a, 1b, 2a, 2b, 3a, 3b and 3c) was also evaluated for survival analysis. Mortality information was confirmed via the Central Civil Registration System.

The patients were mainly evaluated by physical examination, chest X-ray (semi-annually), abdominal computed tomography (semi-annually) and upper endoscopy (annually) in the control visits. Follow-up visits were scheduled at three-month intervals for the first two years, six-month intervals for the third to fifth postoperative years, and once a year there after.

Statistical Analysis

Data were summarized as mean ± standard deviation, numbers (n), percent (%), minimum and maximum. Categorical variables were evaluated via a chi-square test and likelihood ratio. Student’s t-test and the Mann-Whitney U test were used for continuous variables. We used Kaplan-Meier curves and log-rank test to estimate 1-year, 3-year and 5-year survival rates and Cox regression to investigate the univariate and multivariate effects of different parameters on overall survival time. Significant cut-off values for the total number of harvested lymph nodes and metastatic lymph nodes were obtained by calculating the ROC curve. Statistical 13.3 Software was used for statistical analysis. A p value of less than 0.05 was considered statistically significant.

Results

The operative mortality rate was 9.3% (15/161). The mean follow-up period for the 146 patients was 24 (12-102) months. Gender and age had no statistically significant effect on survival (Table 1). Parameters affecting survival are documented in detail in Table 1 and Table 2. Twenty-one (14.4%) patients had complication with prolonged hospital-stay due to mainly major wound infection, leakage, pulmonary disease and cardiac problems. Splenectomy was added to gastrectomy in 39 (26.7%) patients. Twenty-three (15.8%) patients had signet-cell histology. Signet-cell histology and poorly differentiated tumors were found in 68 (46.6%) patients. Both LVI and PNI positivity were seen in 103 (70.5%) patients. 108 (74%) patients had adjuvant chemotherapy. Most of the patients had advanced disease: T3-4 (71.9%), N2-3 (70.6%) and stage 3 (65.1%). The mean survival time was found to be 818.80±692.42 (66-3065) days. 1-year, 3-year and 5-year survival rates were 65.1%, 39.7% and 30.8%, respectively.

Overall Survival

Gender, age, differentiation of tumor and LVI and PNI statuses were not found to have a statistically significant effect on overall survival. In univariate analysis, it was observed that the total number of harvested lymph nodes and metastatic lymph nodes, length of hospital stay, tumor location, surgery type, preserving spleen, T staging, N staging, PNP, MLR, stage and adjuvant chemotherapy had a statistically significant effect on overall survival (Table 1). In multivariate analysis, having distal tumor, having chemotherapy, higher number of total harvested lymph nodes and lower N stage were found to have a statistically significant effect on overall survival (Table 2).

Discussion

In our study, we found that gender, age and differentiation of tumor were not found to have a statistically significant effect on survival (Table 1). In multivariate analysis, having distal tumor location, having adjuvant chemotherapy, higher number of the total harvested lymph nodes and lower N status were found to have a statistically significant effect on survival (Table 2).

Gastric cancer is the fifth most common malignancy and third-leading cause of cancer-related death worldwide (8). Surgery is the main treatment modality (8,9). Besides a successful surgery, some clinicopathologic factors, such as lymph node status and presence of advanced tumor were also found to take a role in long-term prognosis for GA (10). Lymph node status is one of the most important predictor of survival in GA, whereas a cut-off point for the total number of harvested lymph node during surgical dissection has not been found to estimate the postoperative prognosis in previous studies (1,8,9). Chen et al. (9) reported that the number of lymph nodes harvested was mainly dependent on the surgeon’s technique and pathologist’s experience; the main problem was staging of GA according to the number of harvested lymph nodes. Dissection of more than 15 lymph nodes is recommended to prevent the false down-staging (9). We included patients with at least 16 lymph nodes harvested in this study. We found that harvesting more than 33 lymph nodes had statistically significant effect on overall survival (Table 3).

Survival

When survival rates of GA were investigated in our study, we found 1-, 3- and 5-year survival rates of 65.1%, 39.7% and 30.8%, respectively. Kim et al. (10) declared that majority of deaths (70.5%) occurred within the first 2 years after gastrectomy and also declared that 1-, 3- and 5-year survival rates were 75.1, 48.1 and 37, respectively, similar with our results (10). Reported results demonstrated that there was a great variation for 5-year survival of GA, ranging from 5% to 90% depending on the extent of the disease, tumor and patient-related factors (10-18).

Tumor Location and Treatment

Patients with distal tumor location and subtotal gastrectomy had better prognosis in our study. Only five patients with distally located tumors had total gastrectomy to have negative margin, most of patients with distal tumor (94%) had subtotal gastrectomy. Distal tumor location and/or subtotal gastrectomy are still controversial prognostic factors for GA. Distal tumor location and resection with subtotal gastrectomy were similarly demonstrated to be better prognostic factors with our study results (1,9,19). In multivariate analysis, prognostic effect of having distal tumor location was more prominent than subtotal gastrectomy (p=0.018 vs 0.306). On the contrary, some studies claimed that organ preserving surgery had an unfavorable effect (20,21). Additional splenectomy during total/subtotal gastrectomy has been shown to have no effect on prognosis of GA (22,23). In this study, it was found that spleen-preserving surgery was associated with improved survival (Table 1). Poor prognosis due to adding splenectomy to the surgery may be associated with increased morbidity. However, preserving spleen had no prognostic effect on survival in multivariate analysis (Table 2). According to our study results, routine splenectomy was not necessary during gastrectomy unless the spleen is primarily affected by the tumor. The effect of chemotherapy on the GA prognosis remains controversial. Some studies reported that adjuvant chemotherapy had no effect on survival, whereas a recent literature review presented favorable effects of additional adjuvant chemotherapy (1,7,24,25). According to our study results, additional adjuvant chemotherapy seems to be a favorable prognostic factor for survival. Furthermore, chemotherapy was found to be a favorable prognostic factor in multivariate analysis.

Pathologic Features

Signet-cell histology and poorly-differentiated tumors were found in 68 (46.6%) patients. The results of the studies concerning the relationship between grade and survival of patients with GA are challenging. Chen et al. (9) and Kim et al. (10) reported that poorly-differentiated histology and signet-cell type were poor prognostic factors. Biondi et al. (1) reported lower survival rate in grade 3 (poorly differentiated) and linitis plastica groups (five-year survival rate 58.6 and 42.5%, respectively). On the contrary, we found no statistically significant relationship between tumor grade-histology and survival time, similar with some studies (7,19,21). When we evaluated in detail, patients with poorly-differentiated tumors and well-differentiated tumors had similar node-negativity rate (16.2% and 17.9%, respectively). However, patients with worse tumor histology had advanced N status (p=0.031), 31 of 78 (39.7%) patients with well or moderately differentiated tumors had N3 status, 40 of 68 (58.8%) patients with signet-cell histology and poorly differentiated tumors had N3 status. Taghavi et al. (26) declared that when adjusted for stage, signet ring cell did not portend a worse prognosis. The extensiveness of the tumor with lymph node metastasis may be a more important prognostic factor rather than tumor histology and grade according to our study results. LVI-PNI positivity was associated with survival in some studies (10,23,27). On the contrary, LVI-PNI positivity was not found to be associated with poor prognosis in our study.

Stage

In this study, lower stage was found to be associated with better survival (Figure 1). Stage and status of T and N were valuable parameters for survival (Table 1). Tumor detection in an early stage has been shown to be a very important prognostic factor (28). We also found the favorable prognostic effect of early stage on nodal status. The results of other studies demonstrated that patients with early stage tumors had statistically significantly higher survival rates, similar with our study results (7,23).

Nodal Status

In our study, we found that higher total number of harvested lymph nodes and lower number of metastatic lymph nodes had favorable effects on survival. Harvesting more than 33 lymph nodes provide beneficial effects on survival (Table 3). Some studies also reported that MLR was a determinant factor for survival (29). We included patients with more than 15 lymph nodes to prevent inaccurate surgery criticism or down-staging effect. In multivariate analyses, among the nodal parameters, higher total number of harvested lymph nodes and lower N stage were found to be favorable prognostic factors for survival (Table 2).

When we evaluated the effect of MLR and PNP on survival, we observed that higher MLR and higher PNP were statistically significantly worse prognostic factors for survival. Lee et al. (7) similarly reported that five-year survival decreased significantly with increasing MLR (7). Siewert et al. (30) declared significant deterioration of the prognosis on survival for stage II GA when more than 20% of the removed lymph nodes were metastatic. Most of the patients (63.2%) had MLR “0” in the study by Lee et al. (7), however, fewer patients (17.8%) had MLR “0” in our study. The reason for the different results may be related with different population features. One of the main differences between the two studies is no harvested lymph node cut-off in the study of Lee et al. (7) but >15 nodes was an inclusion criterion in our study (7). Shen et al. (8) reported that most patients (85%) had advanced tumors and harvesting more than 30 lymph nodes was associated with better survival for T3, T4, node-positive and stage 3-4 patients. In the present study, most of our patients (71.9% for T3-T4, 65.1% for stage 3) had also advanced tumors. Different cut-off values/systems may be used for lymph node ratio other than MLR. Harvesting more than 33 lymph nodes was associated with better survival in this study (Table 3). PNP level was also correlated with estimated results. Similar categorization values were found to be associated with reasonable predictive value in some studies (3). As well as adequate surgery, pathological evaluation is also important for lymph node ratio and survival (1,9).

Study Limitations

This study had some limitations. Firstly, this is a retrospective study. Secondly, chemotherapy regimen was not unique and was not detailed. Small sample size was another limitation of the study. Further large-scale randomized controlled studies are needed. On the other hand, the study was based on new staging systems; the study population was homogenous, had at least 16 harvested lymph nodes and mostly had advanced-stage tumors.

Conclusion

When screening is not performed routinely, GA is usually diagnosed at advanced stage. Having distal tumor location, having adjuvant chemotherapy, higher total number of harvested lymph nodes and lower N status were found to be more important than grade-histological type of the tumor in survival rates in patients having advanced tumors.

Authorship Contributions

Surgical and Medical Practices: M.B., R.Ç., A.Ö., H.C., M.M.D., T.K. Concept: M.B., R.Ç., A.Ö., M.M.D. Design: M.B., R.Ç., A.Ö., H.C. Data Collection or Processing: M.B., A.Ö., H.C., B.T. Analysis or Interpretition: M.B., B.T. Literature Search: M.B. Writing: M.B.

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.