Is proximal gastrectomy indicated for locally advanced cancer in the upper third of the stomach?

Abstract Aim To treat upper third gastric cancer, proximal gastrectomy (PG), a function‐preserving procedure, is recommended for early lesions when at least half the distal stomach can be preserved, while total gastrectomy (TG) is standard for locally advanced lesions. Oncological feasibility, when applying PG for such lesions, remains unknown. Methods We reviewed patients undergoing TG for clinical (c) T2–T4 upper third gastric cancer between 2006 and 2015. Preoperative tumor locations were further classified into the cardia, fornix, and gastric body based on endoscopic findings. The metastatic rate and therapeutic value index for lymph node (LN) dissection were determined, and characteristics of patients with distal LN (No. 4d, 5, and 6) metastasis (DLNM) were reviewed. In addition, patients with pathological tumor invasion to the middle third (M) region were investigated. Results We studied 167 patients. There were 8 (4.8%) with DLNM and 41 (24.6%) with pathological tumor invasion to the M region. As to regional stations, therapeutic indices for LN dissection at stations No. 4d, 5, 6, and 12a were zero or extremely low. No DLNM was detected in cT2 lesions or cT3/T4 lesions located within the cardia and/or the fornix. In addition, none of the lesions located within the cardia and/or the fornix by preoperative endoscopy extended to the M region in the pathological specimen. Conclusions For upper third gastric cancer, PG without No. 12a dissection might be acceptable for cT2–T4 lesions located within the cardia and/or the fornix when considering the risk of DLNM and cancer‐positivity in the distal stump.


| INTRODUC TI ON
Worldwide, gastric cancer is among the most life-threatening malignant neoplasms. 1,2 The incidence of gastric cancer in the upper third of the stomach has recently been rising in both Western and Asian countries. [3][4][5] As a therapeutic strategy for upper third gastric cancer indicated for surgical treatment, proximal gastrectomy (PG), a function-preserving procedure, is advocated for lesions diagnosed at an early stage when more than half of the distal stomach can be preserved. 6 In contrast, total gastrectomy (TG) is now the standard procedure for locally advanced lesions in the upper third of the stomach. 6 Comparing TG and PG for early gastric cancer, PG is considered to be more advantageous in mitigating body weight loss, maintaining nutritional status, and not causing deterioration of quality of life postoperatively. [7][8][9] Therefore, provided that oncological safety is assured, PG may also be the preferred surgical treatment for locally advanced gastric cancer in the upper third region. As to esophagogastric junctional (EGJ) cancer, PG can be selected even for an advanced tumor if the primary lesion is less than 4 cm in size, based on the oncological safety of lymph node (LN) metastasis, as stated in the Japanese gastric cancer treatment guidelines. 6 However, there are oncological concerns when applying PG for advanced upper third gastric malignancies other than EGJ tumors.
That is, an optimal extent of LN dissection in performing PG for advanced cancer has not yet been established, with only a few reports focusing on this issue, 10,11 nor has the relationship between actual locations of primary lesions and LN metastasis been investigated in sufficient detail. Moreover, it is essential to ensure an adequate distal tumor margin and sufficient volume of the remnant stomach after gastric dissection.
Here we evaluated pathological metastasis involving the aforementioned regional LNs and the distal tumor margin in patients undergoing TG for clinically advanced gastric cancer in the upper third of the stomach. The present results are anticipated to contribute to determining the criteria for applying PG to advanced lesions.

| Assessment of clinical staging
For clinical T factor diagnosis, findings obtained by endoscopy and the features noted on computed tomography (CT) by an experienced radiologist were reviewed, and the depth of tumor invasion of the wall was finally determined at the gastric cancer team conference including surgeons, endoscopists, and chemotherapists. Regional LNs with a long-axis diameter of 10 mm or more on CT were diagnosed as metastatic nodes and their station numbers were also examined.
Clinical stages were determined according to the 14th edition of the Japanese Classification of Gastric Carcinoma. 13

| Pathological metastasis and therapeutic value index for LNs at each station
As to the regional LNs to be dissected, the rate of LN metastasis

| Evaluation of tumor location based on preoperative endoscopy and pathological specimen
We assessed tumor location preoperatively based on the endoscopic findings. In particular, the locations of the distal tumor border were further divided into three regions, ie, the cardia, the fornix, and the gastric body ( Figure 1). The location of the cardia was defined as being within 2 cm of the EGJ in the stomach. Representative endoscopic photographs of tumors included in this study are presented in

| S TATIS TIC AL ANALYS IS
The patient background characteristics, surgical details, and pathological findings were collected from our database and information contained in electronic medical records. The relationships between clinical characteristics and pathological findings, including LN metastasis and tumor location, were investigated. All continuous variables are expressed as median values. Statistical analyses were conducted using the Mann-Whitney U-test and the chi-squared test. A P-value less than .05 was considered to indicate a statistically significant difference. All statistical analyses were performed with JMP Pro 13 (SAS Institute Japan, Japan) for windows.

| Clinicopathological characteristics
Patient clinical characteristics are shown in Table 1. In total, 167 patients were included in this study. As to the location of the distal tumor border according to preoperative endoscopy, 12 patients  Table 3 shows the LN metastasis rate, 5-y OS, and the therapeutic value index for dissection of each LN station. In patients with cT2 lesions, the metastatic rate and the therapeutic index of LN were both zero at stations No. 4d, 5, and 6. On the other hand, the LN metastasis rate was zero only at station No. 6 of the three stations examined patients with cT3/T4 tumors. Among the regional stations examined, however, the therapeutic indices of LN dissection at stations No. 4d and 5 were extremely low, at 1.0 and 1.4, respectively.

| Rate of LN metastasis and therapeutic value index of LN dissection for each station
The therapeutic index for dissection of LNs at station No. 12a was also zero in all patients.

| D ISCUSS I ON
We evaluated the pathological status of regional LNs, the therapeutic index for each nodal station, and the pathological tumor location in patients undergoing TG for cT2-4 upper third gastric cancer, to TA B L E 3 Metastatic ratio and therapeutic value index for dissection of LNs at each station   Insufficient volume of the remnant stomach after PG is reportedly associated with deterioration of postoperative quality of life and skeletal muscle loss. 22,23 In practice, the aim is generally to preserve more than 2/3 of the preoperative gastric volume in performing PG for early lesions. 7,24 Even when applying PG to advanced tumors, it is apparently essential that at least half of the distal stomach, as recommended in the guidelines, be preserved. 6 Large tumors and pathologically advanced T stage are reportedly risk factors for Third, functions of the remnant stomach such as peristalsis and retention after PG for locally advanced cancer were not evaluated. A multicenter study with a large sample size is required to clarify these issues and overcome the limitations of this study.
In conclusion, for locally advanced gastric cancer in the upper third of the stomach, PG without No. 12a dissection might be acceptable for lesions located within the cardia and/or the fornix, given that neither DLNM nor a distal cancer-positive margin was detected in these cases.

D I SCLOS U R E
The protocol for this research project was approved by the Souya Nunobe https://orcid.org/0000-0003-3012-5632