Gastric Stump Cancer: Demanding a More Suitable Denition and TNM Staging System

Background The incidence of gastric stump cancer (GSC) following resection of gastric cancer has increase. However, the denition between GSC and recurrent gastric cancer (RGC) is still being debated. This study was conducted to compare the clinicopathological characteristics and outcome for this two groups. Patients and Methods The investigators retrospectively reviewed patients in the Surveillance Epidemiology and End Results (SEER) databases from 1975 to 2016 to identify patients who underwent resection of gastric cancer subsequently experienced metachronous gastric cancer. GSC was dened as cancer occurring in remnant stomach of ≥ 10 years after gastrectomy, while RGC was dened as <10 years after gastrectomy. T-test, Pearson χ2-test and Kaplan-Meier estimator method was used in the present study.


Introduction
Gastric stump cancer (GSC) was de ned as "cancer occur in the remnant stomach after surgery for benign disease at least ve years and those after surgery for gastric cancer at least 10 years.'' According to the previous reported, the incidence of GSC range 1.0-3.0 % [1]. Most of previous GSC studies focused on benign disease after gastrectomy. However, gastrectomy for peptic ulcer has been greatly reduced in recent decades since drug therapy. It can expect that the number of patients with GSC after gastrectomy for benign diseases will decrease in the future [2]. On the other hand, partial gastrectomy has been widely accepted for early gastric cancer. And since the early detection and early treatment for gastric cancer, the number of patients with long-term survival after gastrectomy has increased. Therefore, some studies have reported that the number of patients with GSC after partial gastrectomy for gastric cancer will increase in the future [2][3][4].
However, to date, only a few studies have described the current status of GSC on the basis of long-term follow up data after surgery for gastric cancer and most of those studies were performed as single institutional and base on small number cases [5][6][7]. There were not studies clari ed whether there are differences in clinicopathological characteristics and prognosis between patients with GSC and patients with RGC. In addition, the de nition of GSC and RGC still controversy since the time limit was various.
This study aimed to reveal the clinicopathological characteristics, treatment and outcomes between GSC and RGC after gastrectomy for gastric cancer and further discuss whether it is necessary to limit 10 years later for the de nition of GSC following resection of gastric cancer.

Study Population
A retrospective cohort was identi ed by use of the Surveillance Epidemiology and End Results (SEER) databases, which is a population-based database that assembles data related to cancer patients in the United States. We obtained the study cases in 18 SEER registries from 1975 to 2016 (November 2018 submission) and identi ed patients who underwent gastrectomy for primary gastric cancer (ICD-O-3 site codes C34.0-C34.9) and subsequently experienced a second gastric cancer. The select gastrectomy codes included "site-speci c surgery" codes (20, 30, 40,50,60and 70) for cases before 1998 and "surgery primary site" codes (30-33, 40-42, 50-52, and 60-63) after 1998. The enrolled demographic variables include gender, age and race. The TMN stage were re-stage base on the eight edition AJCC stage. GSC was de ned as cancer occur in the remnant stomach after the rst primary cancer surgery at least 10 years despite whether the histology is the same or not. RGC was de ned as cancer occur in the remnant stomach after the rst primary cancer surgery within 10 years. Patients underwent total gastrectomy or nonsurgical were excluded in the study. The patient selection algorithm is described in Fig 1. Statistical analysis.
Statistical analysis was performed by SPSS (version22, IBM Corp). χ2-test was performed for categorical variables and Student's t-test was performed for continuous variables. Unadjusted overall survival (OS) and cancer-speci c survival (CSS) were analyzed using the Kaplan-Meier estimator method. A two-sided P-value of less than 0.05 was considered statistically signi cant.

Results
A total 167,747 patients who underwent gastric cancer, 1006 patients (0.6%) were second primary gastric cancer which meet the multiple primary criteria of SEER. Among these 1,006 patients, 93 (9.2%) met the GSC criteria and 282 (28.0%) met the RGC criteria ( Fig. 1).

Incidence trend
The incidence trend of RGC and GSC were shown in Figures 2 and 3. We can observe that both the RGC and GSC were increased in the following years. The increase was more obvious after 2007 in both groups.

Patients characteristic
The rst primary gastric cancer characteristics of two study patients are described in Table 1

Survival analysis
Both the median CSS and median OS between the GSC and RGC were not statistically signi cant (40.0 months vs. 25.0 months, p=0.032 and 20.0 months vs. 16.0 months, p=0.547, respectively) (Fig 4). To evaluated the effective of surgery treatment, we compared both RGC and GSC according to CSS and OS.
The results showed that surgery treatment or not, the CSS and OS have statistically signi cant (64.0 months vs. 10.0 months in median OS, p 0.001) (Fig. 5). Similarly, in GSC group, the median OS were 33 months by surgery treatment and 13 months in patients not underwent surgery (p=0.014) (Fig. 6).

Comment
The de nition of gastric stump cancer remains controversial. It was rst described and de ned in 1922 as ''cancer in the remnant stomach after surgery for benign disease at least 5 years'' [8]. Subsequently, studies suggested that cancer appeared in remnant stomach after surgery for gastric cancer at least 10 years also classi ed as GSC in 1982 [9]. However, the Japanese Gastric Cancer Association suggested that GSC refers to all cancers detected from the remnant stomach after partial gastrectomy, irrespective of the primary disease and interval [10]. In China, it still suggested that the time interval of GSC must at least 10 years after surgery for rst gastric cancer [11] . The time interval of GSC was varied in previous studies [1]. Since the rarely morbidity and long-time follow up, the clinicopathology, treatment and the de nition of GSC following resection of gastric cancer lacks of a large sample study.
Previous studies often focus on compare the clinicopathology of GSC between gastric cancer and benign disease or compare the prognosis of GSC and primary gastric cancer (PGC). Most studies have reported that the prognosis between GSC and PGC have no signi cant difference [6,12,13] and revealed that the 5year survival rates were 41.7 % vs. 45.2 % [1].
With increasing number of partial gastrectomy for gastric cancer and the number of patients with longterm survival after gastrectomy, the number of patients with GSC following resection of gastric cancer has increased [2,14]. The present study also approved this trend. This raises two practical questions: 1) Should the interval of GSC following the rst resection for malignant disease be at least 10 years? 2) How to treatment base on TMN stage after the normal anatomic structure has been destroyed?
In order to answer the rst question, we compared the clinicopathology between GSC and RGC, male and younger patients may more likely to develop GSC and there was no statistically signi cant on others variable. This difference may be explained that younger and early gastric cancer patients may more likely to have long-term survival more than 10 years. On the other hand, the prognosis was not signi cant difference between GSC and RGC. Thus, we suggest that the de nition of GSC should not be limited to 10 years after surgery. However, different to the traditional recurrence of gastric cancer, in the present study, RGC were classi ed as multiple primary gastric cancer by SEER database, which was de ned as the following rules: 1) Tumors diagnosed more than one year apart, and 2) Tumors with ICD-O-3 histology codes that are different at the rst (x xxx), second (x xxx) or third (x xxx) number are multiple primaries. 3) These rules are NOT used for tumor(s) described as metastases. Based on the present study outcome and this de nition, we suggest the de nition of GSC may be based on the following rules: 1 cancer occur in the remnant stomach after surgery for benign disease at least ve years and those after surgery for gastric cancer at least 10 years. 2) Or tumors with difference histology between the primary cancer and second cancer after surgery within 10 years. Similarly, many previous study de ned GSC as cancer that occurs within 10 years following the initial gastrectomy for gastric cancer and is thought to have developed independently of the initial lesion [2,15]. With the development of molecular markers, the analysis of molecular genetic characteristics might be expected to improve the speci city of the de nition of GSC.
In order to answer the second question, we evaluated the effective of surgery treatment for GSC and RGC. The result showed that surgery still suit for GSC and RGC since the satisfactory prognosis compare to those no underwent surgery. Previous studies considered that although previous gastric resection tends to render surgery for GSC more challenging, due to adhesions and anatomical changes, R0 resection remains as the best treatment for GSC [14][15][16][17]. However, due to the anatomical changes for rst gastric cancer, the patterns of lymph drainage from lesions in the remnant stomach were also changed. Studies have shown that in most patients with previous cancer, lymph drainage from the remnant stomach was newly developed to the lower mediastinum, to the jejunal mesentery, or to the duodenal side through the anastomotic line made at the initial surgery. The reason was probably due to the surgical ablation of preexisting lymphatics and lymph nodes in the peri-gastric areas at the initial cancer surgery [15,18,19].
This situation reveals that the N stage of TNM staging system based on primary gastric cancer is unsuitable for GSC patients. Base on this conclusion, some studies recommended develop a new and more suitable staging systems for GSC [12,20]. However, a consensus has not been reach on the new N stage. More studies are needed to explore a suitable N stage and appropriate treatment base the new stage. For the surgery treatment, R0 resection should comprise a total gastrectomy and radical lymph node resection, especially the jejunal mesentery of the anastomosis [14].
Although the present study has revealed some signi cant clinical information, there were some limitations. The SEER database did not provide information on the type of resection (R0, R1 or R2), lymph node dissected pattern, or pattern of recurrence. Furthermore, for the classi cation of multiple primary gastric cancer, a second primary versus a recurrence can be complex. There may be some misclassi cation cases on the study. In addition, the missing variable on the present study may cause bias.

Conclusions
The number of patients of GSC following resection of gastric cancer has increased. The de nition of GSC may not be limited to cancer occurrence in the remnant stomach after surgery for benign disease for at least ve years, and those after surgery for gastric cancer for at least 10 years. Patients who had tumors with difference histology between the primary cancer and second cancer after surgery within 10 years may classi ed as GSC. R0 resection may be the rst choice for GSC, but lower mediastinum lymph and jejunal mesentery of the anastomosis resection must be taken into consideration due to the high incidence of lymph node metastasis. The data that support the ndings of this study are available from the corresponding author upon reasonable request.

Funding information
This work was supported by The National Natural Science Foundation of China (81272637). The Sanming Project of Medicine in Shenzhen (SZSM201612022 to ZL and YW).

Con icts of interest
The authors have declared no con icts of interest.

Ethics approval
The study was approved by the Ethics Committee of the Seventh A liated Hospital of Sun Yat-sen University.

Consent to participate
All of the data was obtained from the SEER database thus informed consent was not required in the study.

Consent for publication
Not applicable.