Should Splenic Hilar Lymph Nodes be Dissected for Siewert Type II and III Esophagogastric Junction Carcinoma Based on Tumor Diameter?

Abstract The aim of the study is to identify the value of a spleen-preserving No. 10 lymphadenectomy (SPL) for Siewert type II/III adenocarcinoma of the esophagogastric junction (AEG). From January 2007 to June 2014, 694 patients undergoing radical total gastrectomy for Siewert type II/III AEG were analyzed. Oncologic outcomes were compared between SPL and no SPL (No. 10D+ and No. 10D–) groups. The incidence of No. 10 lymph node metastasis (LNM) was 12.3%. No significant differences in the incidence of No. 10 LNM were found between Siewert type II AEG with tumor diameters of <4 cm and ≥4 cm (P = 0.071). However, Siewert type III AEG with a tumor diameter ≥4 cm showed a significantly higher frequency of No. 10 LNM compared with a tumor diameter <4 cm (P < 0.001). The No. 10D+ group had superior 3-year overall survival (OS) and disease-free survival (DFS) rates compared with the No. 10D− group (P = 0.030 and P = 0.005, respectively). For patients with Siewert type II and type III AEG with a tumor diameter <4 cm, the 3-year OS and DFS rates were similar between the 2 groups. However, the No. 10D+ group had better 3-year OS (66.6% vs 51.1%, P = 0.019) and DFS (63.2% vs 45.9%, P = 0.007) rates for Siewert type III AEG with a tumor diameter ≥4 cm. A multivariate Cox regression showed that SPL was an independent prognostic factor in Siewert type III AEG with a tumor diameter ≥4 cm. SPL may improve the prognosis of Siewert type III AEG with a tumor diameter ≥4 cm, whereas SPL may be omitted without decreasing survival in patients with Siewert type II or type III AEG with a tumor diameter <4 cm.


INTRODUCTION
T he incidence of adenocarcinoma of the esophagogastric junction (AEG) is increasing. 1,2 Siewert and Stein 3 classified AEG into 3 subgroups according to the location of the tumor's epicenter. Perigastric regional nodal metastasis comprises the majority of Siewert type II/III AEG cases, and the incidence of No. 10 lymph node metastasis (LNM) is reported to be 4.8% to 15.0%. [4][5][6][7] Biological behavior of AEG differs among types and particularly with regard to the incidence of No. 10 LNM. 8,9 Therefore, the value of a No. 10 lymphadenectomy for Siewert type II and III AEG is fairly controversial. 9,10 Several recent studies have indicated that tumor diameter not only is prognostic in gastric cancer but also correlates with LNM. 11,12 The Japanese Gastric Cancer Association and the Japan Esophageal Society assessed the status of LNM in AEG with a tumor diameter <4 cm and developed a flow diagram to identify the extent of the lymphadenectomy. They did not recommend a No. 10 lymphadenectomy in the 4th edition of Japanese gastric cancer treatment guidelines. 13 However, whether a No. 10 lymphadenectomy should be performed for AEG with a tumor diameter !4 cm was not addressed. Therefore, this study sought to investigate the survival benefits of a spleen-preserving No. 10 lymphadenectomy (SPL) for Siewert type II and III AEG with tumor diameters of <4 cm and !4 cm.

Patients
From January 2007 to June 2014, a prospectively maintained database identified 694 patients with Siewert type II/III AEG who underwent a transabdominal radical total gastrectomy. Based on whether they underwent a spleen-preserving No. 10 lymphadenectomy (SPL), patients were divided into No. 10Dþ (n ¼ 293) and No. 10D-(n ¼ 401) groups. The necessity of a No. 10 lymphadenectomy was evaluated according to the preoperative examination and intraoperative exploration by the same group of surgeons, all of whom had significant experience in laparoscopic radical gastrectomy. A No. 10 lymphadenectomy was also performed by the same group of surgeons. The inclusion criteria were as follows: histological confirmation of Siewert type II or III AEG; pathological confirmation of stage T1 to T4a; no evidence of distant metastasis; a completely transabdominal approach; and curative R0. The exclusion criteria included preoperative radiotherapy or chemotherapy, combined major organ resection, presence of a tumor invading >3 cm into the esophagus, or incomplete pathological data. Preoperative imaging studies, including computed tomography (CT) scanning, ultrasonography (US) of the abdomen and endoscopic US, were routinely performed following an endoscopic examination. The severity of complications was classified according to the Clavien-Dindo grade, and major complications were at least grade III. 14 Staging was determined according to the 7th edition of the International Union Against Cancer (UICC) TNM classification. 15 Adjuvant chemotherapy with 5-fluorouracil (5-FU)-based regimens (generally 5-FU with cisplatin) was recommended for every patient with positive nodal disease or advanced cancer. Written informed consent was obtained from all patients prior to surgery. This study was approved by the institutional review board of Fujian Medical University Union Hospital.

Surgical Procedures
All patients were treated with a radical total gastrectomy. In terms of the lymphadenectomy, a D2 lymphadenectomy including SPL was performed on patients in the No. 10Dþ group. [16][17][18] In the No. 10D-group, the extent of the lymphadenectomy was the same as in the No. 10Dþ group but without the SPL. Specifically, after entering the retropancreatic space from the superior border of the pancreas, the left gastroepiploic vessels were sufficiently exposed at the tail of the pancreas and the lower pole of the spleen. These vessels were sufficiently vascularized and dissected. However, the remaining short gastric vessels were divided at their roots directly to the left cardiac region. The splenic vessels were not vascularized, and the lymph nodes in this region were not dissected. An intraoperative frozen section analysis was performed during every operation, and the adequacy of the extent of esophageal resection was confirmed.

Measurement of Tumor Diameter
The tumor diameter was measured according to the Japanese Classification of Gastric Carcinoma. 19 Briefly, the resected stomach was opened along the greater or lesser curvature to expose the entire mucosa. Then the opened stomach was examined macroscopically with the mucosal side up, and the diameter and thickness of each tumor were recorded. The longest diameter was used in the present study.

Follow-Up
Patients were followed up every 3 months for 2 years and then every 6 months from 3 to 5 years. Most routine follow-up consisted of laboratory tests, chest radiography, abdominopelvic ultrasonography, or computed tomography. Endoscopic examination was performed every 12 months. Overall survival (OS) was calculated from the day of surgery until death or until the final follow-up date of June 2015, whichever occurred first. Recurrence was diagnosed based on radiological or histological signs of disease. Disease-free survival (DFS) was calculated from the day of surgery to the day of recurrence or death.

Statistical Analysis
The chi-square test or Fisher's exact test was used to compare categorical variables, and Student's t tests were used to evaluate continuous variables. Multivariate analysis that used binary logistic regression models was performed to further evaluate factors found to be significantly prognostic on univariate analysis. Cumulative survival rates were compared using the Kaplan-Meier method and the log-rank test. All statistical analyses were performed using SPSS v. 18.0 for Windows (SPSS Inc., Chicago, IL). Values of P < 0.05 were considered to be statistically significant.

Patient Characteristics
The general clinicopathologic characteristics are summarized in Table 1. There were 293 patients in the No. 10Dþ group     type II AEG and 15.4% in patients with type III AEG (P ¼ 0.046). A significantly higher rate of No. 10 LNM was observed in patients with a tumor diameter !4 cm than with those with a tumor diameter <4 cm (17.8% vs 1.0%, P < 0.001). A stratified analysis demonstrated that there were no significant differences in No. 10 LNM for Siewert type II AEG with tumor diameters of <4 cm and !4 cm (P ¼ 0.071). However, Siewert type III AEG with a tumor diameter !4 cm showed a significantly higher rate of No. 10 LNM compared with a tumor diameter <4 cm (21.6% vs 0.0%, P < 0.001) (Figure 1). For patients with No. 10 LNM, the percentages of N1-3 were 11.1%, 11.1%, and 77.8%, respectively, and they were 21.0%, 17.1%, and 28.0% for patients without No. 10 LNM, respectively. N stage was significantly more advanced for patients with compared to without No. 10 LNM (P < 0.001).

Oncologic Outcomes
The median follow-up period was 47 (range 11-100) months. The No. 10Dþ group showed superior 3-year OS and DFS rates compared to the No. 10D-group (P ¼ 0.030 and P ¼ 0.005, respectively) ( Table 2 and Figure 2A and B) . For patients with Siewert type II AEG, the 3-year OS and DFS rates were similar between the 2 groups (Table 2). However, the No. 10Dþ group showed significantly longer 3-year OS (72.9% vs 55.1%, P ¼ 0.002) and DFS (69.8% vs 50.4%, P < 0.001) in patients with Siewert type III AEG (Table 2, Figure 3A and B).

Stratified Analysis of Survival According to Tumor Diameter
For Siewert type II AEG with different tumor diameters (<4 cm and !4 cm), the 3-year OS and DFS rates did not differ significantly between the No. 10Dþ and No. 10D-groups ( Table 2). In patients with Siewert type III AEG and a tumor diameter <4 cm, the differences in the 3-year OS and DFS rates between the 2 groups were not statistically significant (Table 2). However, the No. 10Dþ group had better 3-year OS (66.6% vs 51.1%, P ¼ 0.019) and DFS (63.2% vs 45.9%, P ¼ 0.007) rates for patients with Siewert type III AEG and a tumor diameter !4 cm ( Table 2 and Figure 4A and B).

Univariate and Multivariate Survival Analyses and Clinicopathologic Characteristics of Patients With Siewert Type III AEG and a Tumor Diameter > -4 cm
The clinicopathologic characteristics of patients with Siewert type III AEG and a tumor diameter !4 cm were comparable between the No. 10Dþ and No. 10D-groups (Table 3). Factors potentially related to OS and DFS in the univariate analysis were No. 10 lymph node dissection (LND), age, tumor grade, T stage, and N stage ( Table 3). The multivariate model identified No. 10 LND, age, T stage, and N stage as independent predictors of OS and DFS (Table 4).

DISCUSSION
Surgical management is the primary treatment of AEG. Perigastric regional nodal metastasis comprises the majority of AEG cases, and the transabdominal approach is recommended for Siewert type II/III tumors invading <3 cm into the esophagus according to the results of the JCOG 9502 trial. 20 6,21 showed that the rate of No. 10 LNM in Siewert type II AEG was much lower than in type III AEG. In this study, the incidence of No. 10 LNM in Siewert type II AEG was 7.6%, which was also much lower than the rate of 15.4% in type III AEG (P ¼ 0.046). It has been reported that tumor diameter in proximal gastric cancer correlates with No. 10 LNM. 22 Shin et al 23 indicated that the splenic hilar metastasis group in proximal gastric cancer had a higher proportion of tumor diameters >5 cm (82.9%). Fang et al 5 found no incidence of No. 10 LNM in Siewert type II AEG with a tumor diameter <4 cm. The 4th edition of the Japanese gastric cancer treatment guidelines, published in 2014, also revealed a low incidence of No. 10 LNM in AEG with a tumor diameter <4 cm and discouraged the use of a No. 10 lymphadenectomy in the treatment of these tumors. 13 Therefore, we evaluated the frequency of No. 10 LNM in Siewert type II and III AEG with tumor diameters of <4 cm and !4 cm in this study. Of the 36 patients with No. 10 LNM, only 1 had a tumor <4 cm in diameter, and the rest had tumors !4 cm in diameter. For Siewert type III AEG, the No. 10 LNM rate was significantly higher for tumors with a diameter !4 cm compared with those with a diameter <4 cm. This is because the anatomical position and biological behavior of Siewert type III AEG with a larger tumor diameter are quite similar to those of advanced proximal gastric cancer, which has a higher rate of No. 10 LNM. 24 Tumor stage was more advanced and had a higher incidence of No. 10 LNM for a tumor diameter !4 cm because it is very difficult to perform a No. 10 lymphadenectomy but easy to damage the vessels and tissues in the splenic hilar area, which can induce more major complications for patients with a tumor diameter !4 cm.
Although there is a certain incidence of No. 10 LNM in Siewert type II and III AEG, the survival benefit of a No. 10 lymphadenectomy for Siewert type II and III AEG patients with a total gastrectomy is uncertain and controversial. A systematic review of spleen and pancreas preservation in an extended lymphadenectomy for advanced gastric cancer was presented by Brar et al. 25 They concluded that the preservation of the spleen and pancreas during an extended lymphadenectomy would not decrease the OS. Goto et al 9 evaluated 132 patients with Siewert type II/III AEG and found that a No. 10 lymphadenectomy did not increase the OS for Siewert type II/III AEG. Yang et al 26 indicated that the 5-year survival rates of patients with Siewert type II and III AEG were 51.6% and 57.2% in the No. 10Dþ group and 28.5% and 39.5% in the No. 10D-group, respectively. Although the difference in 5-year survival rates between the 2 groups did not reach statistical significance, the differences were obvious. Therefore, Yang et al recommended a No. 10 lymphadenectomy for Siewert type II/III AEG during a total gastrectomy. To estimate the therapeutic value of a No. 10 lymph node dissection, Hosokawa et al 27 calculated the index of estimated benefit from lymph-node dissection 28   10 LNM and a more advanced N stage. Therefore, a No. 10 lymphadenectomy would significantly improve the prognosis for patients with Siewert type III AEG and a tumor diameter !4 cm. Furthermore, our study revealed that a No. 10 lymphadenectomy was an independent prognostic factor in Siewert type III AEG with a tumor diameter !4 cm. To our knowledge, the present study is the first and largest to evaluate the oncologic outcomes of SPL for Siewert type II and III AEG based on tumor diameter. However, our study did have some limitations, including the use of data from a single center and its retrospective nature, which introduces a possible selection bias. The preoperative examination was unable to accurately assess the status of LNM in the splenic hilar. Therefore, a No. 10 lymphadenectomy should have been performed for some patients with No. 10 LNM, and the results of the metastatic ratio and survival differences may be biased. In addition, data to assess how survival could be attributed to a No. 10 LND versus chemotherapy were not available in our center. Therefore, in our future research, a further randomized prospective study is required to confirm our results, particularly for Siewert type II AEG, to provide evidence for the standardized treatment of AEG.