If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The value of spleen-preserving lymphadenectomy in total gastrectomy for gastric and esophagogastric junctional adenocarcinomas: A long-term retrospective propensity score match study from a high-volume institution in China
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Department of Gastrointestinal Surgery and Laboratory of Digestive Surgery, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
Reprint requests: Jian-Kun Hu, MD, PhD, FRCS, Department of Gastrointestinal Surgery, and Laboratory of Gastric Cancer, West China Hospital, Sichuan University, No. 37 Guo Xue Xiang Street, Chengdu 610041, Sichuan Province, People’s Republic of China.
Department of Gastrointestinal Surgery and Laboratory of Gastric Cancer, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, and Collaborative Innovation Center for Biotherapy, Chengdu, China
The benefit of removing the splenic lymph nodes in patients with proximal gastric cancer has been controversial. The purpose of our study was to investigate the importance of performing a splenic hilar lymph node dissection without splenectomy in patients undergoing total gastrectomy for gastric cancer.
Methods
From January 2006 to December 2015, we retrospectively reviewed patients who underwent a curative total gastrectomy for gastric cancer. Propensity score matching was used to balance any potential discrepancy of the other covariates between patients with and without splenic hilar lymph node dissection. Survival analysis, Cox univariate and multivariate analysis, and subgroups analysis were conducted to determine the value of splenic hilar lymph node dissection. After matching, 2 nomograms among patients with and without splenic hilar lymph node dissection were established respectively, the C-index, calibration curve and decision curve analysis were used to further evaluate the value of splenic hilar lymph node dissection.
Results
The rate of metastatic splenic hilar lymph nodes in the 274 patients undergoing splenic hilar lymph node dissection was 16.4% (45/274). Patients undergoing splenic hilar lymph node dissection had better survival outcomes than those not undergoing splenic hilar lymph node dissection before (P = .003) and after (P = .003) propensity score matching. Cox multivariate analysis also confirmed that splenic hilar lymph node dissection was an independent prognostic factor both before (hazard ratio 1.284, 95% confidence interval 1.042–1.583, P = .019) and after (hazard ratio 1.480, 95% confidence interval 1.156–1.894, P = .002) propensity score matching. Subgroup analysis indicted that splenic hilar lymph node dissection offered better survival outcomes for esophagogastric junctional adenocarcinoma (P < .001, P for interaction = .018). After propensity score matching, the nomogram of patients with splenic hilar lymph node dissection (C-index 0.735, 95% confidence interval 0.695–0.774) also indicated a statistically significant advantage compared with that without splenic hilar lymph node dissection (C-index 0.708, 95% confidence interval 0.668–0.748, P < .001).
Conclusion
Our study suggests that spleen-preserving splenic hilar lymph node dissection should be an essential procedure among patients undergoing total gastrectomy.
Introduction
Gastric cancer (GC) remains one of the most common causes of cancer-related death worldwide.
Total gastrectomy (TG) with a concomitant D2 lymphadenectomy has been recommended for advanced esophagogastric junctional adenocarcinoma (EGJA), upper third GC, middle third GC, and lower third GC that invades the middle and/or upper third of stomach or tumors involving all of the stomach.
Gastric cancer surgery: morbidity and mortality results from a prospective randomized controlled trial comparing D2 and extended para-aortic lymphadenectomy–Japan Clinical Oncology Group study 9501.
Whether a No. 10 lymph node dissection (No. 10 LND) should be included in a D2 lymphadenectomy had been always in dispute. Several previous reports have suggested that patients undergoing a No. 10 LND might obtain considerable survival benefits,
Survival benefits from splenic hilar lymph node dissection by splenectomy in gastric cancer patients: relative comparison of the benefits in subgroups of patients.
Splenic hilar lymph node metastasis independently predicts poor survival for patients with gastric cancers in the upper and/or the middle third of the stomach.
Should splenic hilar lymph nodes be dissected for Siewert type II and III esophagogastric junction carcinoma based on tumor diameter? A retrospective database analysis.
; however, these randomized controlled trials all included a splenectomy to accomplish the No. 10 LND. The recent treatment guideline of the 15th Japanese Gastric Cancer Association (JGCA) did not recommend splenectomy for removal of the No. 10 lymph node (No. 10 LN) in D2 lymphadenectomy during TG.
however, the long-term survival outcomes of No. 10 LND and which subgroup of patients might truly benefit from No. 10 LND was still unknown.
The purpose of this study was to assess the value of a spleen-preserving, splenic hilar lymphadenectomy for patients undergoing TG and to investigate the potential benefit for subgroups analysis of patients with GC involving specific regions of the stomach and the esophagogastric junction by stratified analysis.
Methods
Ethical standards
This study was based on data collected from the database of the Surgical Gastric Cancer Patient Registry of West China Hospital (WCH-SGCPR) under registration number WCHSGCPR-2018-09. The establishment of this database was approved by the Research Ethics Committee of West China Hospital. With respect to the GC database, medical records were anonymized and deidentified before analysis and informed consent of individual patients was waived because of the retrospective nature of the analysis.
Study population
We reviewed retrospectively 837 patients who underwent radical TG for GC and EGJA in the Gastrointestinal Surgery Department West China Hospital from January 2006 to December 2015. The inclusion criteria were as follows: (1) histologically confirmed gastric adenocarcinoma confirmed by biopsy; (2) patients who underwent radical resection (pathologically confirmed R0 resections); (3) patients who underwent TG; (4) patients in whom the number of harvested lymph nodes was ≥15; and (5) patients without any preoperative oncologic treatment. The exclusion criteria included (1) other gastric neoplasms; (2) gastric remnant cancer; (3) patients who underwent a distal or proximal gastrectomy alone without a D1+/D2/D2+ lymphadenectomy; (4) patients with peritoneal dissemination and/or other distant organ metastasis; (5) patients who underwent R1/R2 resections; (6) patients in whom the number of harvested lymph nodes was <15; and (7) patients with any preoperative oncologic treatment.
Definition of No. 10 LND
Our study was a retrospective review of 837 patients who underwent radical TG for GC and EGJA. All the patients underwent a TG plus D1+/D2/D2+ lymphadenectomy according to the JGCA guideline. Since No. 10 LND including a splenectomy had been controversial for patients with early-stage GC or without greater curvature invasion, we decided to detect the value of spleen-preserving No. 10 LND. In our study, harvested lymph nodes were examined by experienced pathologists, and patients were divided according to the final pathologic results into those with a No. 10 LND (No. 10 LND+ group) and those who did not undergo a No. 10 LND (No. 10 LND– group). If the pathologic reports indicated that only adipose or fibrous connective tissue was detected in No. 10 region, these patients were also divided into the No. 10 LND– group.
Operative procedures for TG
Before operation, all patients in our study routinely had a chest and abdominal computed tomography (CT) imaging to evaluate the clinical stage of the tumor. The potential of metastasis in the No.10 LN for GC and the inferior mediastinal LNs for EGJA was routinely evaluated by CT; when the lymph nodes had CT enhancement and the smallest diameter of these LNs was ≥1.0 cm, they were deemed clinically involved.
In our study the indications for TG were as follows: (1) tumor located in and/or invading the middle or upper third of the stomach (including the esophagogastric junction) with advanced stage; (2) tumor located in the distal third of the stomach and intraoperative frozen section confirmed metastasis in the No. 4sa and/or No. 2 lymph nodes, and/or tumor invading the middle third of stomach; (3) a positive resection margin confirmed by frozen section after distal or proximal gastrectomy; and (4) bulky lymph nodes detected in the upper region of stomach, whatever the tumor location. Intraoperative frozen section was used routinely to ensure tumor-free margins. Esophagogastric continuity was restored with a Roux-en-Y esophagojejunostomy: The length of the Roux-en-Y was 45 cm; the esophagojejunal anastomosis was performed using a mechanical stapler with a 21 or 25 mm circumference and external seromuscular reinforcement with 3-0 silk sutures. The duodenal stump was closed using a mechanical stapler with external reinforcement with interrupted 3-0 silk sutures. One drain was left routinely left along the left side of the esophagojejunal anastomosis. Combined resections (including gallbladder, spleen, left lobe of liver, small intestine, transverse colon, and pancreatic body and tail) were performed selectively only when tumor invaded the adjacent structures or to ensure an en bloc dissection of bulky metastatic LNs. At the end of every operation, the surgeons would routinely identify and classify all lymph nodes groups according to the Japanese classification.
Total gastrectomy plus D2 lymphadenectomy was performed routinely for patients with advanced lesions; D1/D1+ lymphadenectomy was performed selectively among patients with early clinical stage GCs (involvement of only the mucosa and submucosa). A spleen-preserving No. 10 LND was performed routinely as follows: the lymphatic tissue at the splenic hilum was dissected with care to preserve the spleen and splenic vessels. Splenectomy was not performed unless the tumor invaded the spleen, there were bulky lymph nodes metastasis in splenic hilum, or there was a splenic injury. All the operations were performed by experienced surgeons specialized in GC surgery.
Clinicopathologic materials
The clinicopathologic data reviewed included demographic parameters, maximal diameter of the tumor (cm), macroscopic types, pT stage, pN stage, pM stage (pTNM), vascular invasion, neural invasion, and tumor differentiation (well, moderate, poorly differentiated, and signet-ring cell type). Operation-related parameters included various types of resection, number of LNs harvested, concomitant organ resection, and short-term operative outcomes. The clinicopathologic features were classified according to the JGCA (third English version).
The major postoperative complications were defined as those requiring reoperation or other interventions. We used the Clavien-Dindo classification of postoperative complications.
Operative mortality was any death, regardless of reason, occurring before discharge or within 30 days postoperatively whether the patient was in or out of the hospital.
In this study, overall survival (OS) was the primary endpoint and was calculated from date of TG to the time of death from whatever cause to the latest follow-up. All the patients were followed-up periodically by either outpatient visits, telephone interviews, network tools, and letters. The follow-up interval was every 3 to 6 months during the first 2 postgastrectomy years, every 6 to 12 months during the subsequent 3 years, and annually thereafter until death.
Patients lost to long-term follow-up were also recorded; the main causes of loss to follow-up were their inability to return for outpatients visit and problems with their contact information precluding our ability to contact them. As of October 2019, among the 837 patients, 776 (92.7%) had complete follow-up. The median follow-up was 91.2 months.
Propensity score matching
Because patients were not allocated randomly to the No. 10 LND+ or No. 10 LND– groups, treatment selection bias probably did exist in our study. Therefore we used propensity score matching (PSM) to minimize the selection bias and balance some covariates that might be associated with survival outcomes.
We performed PSM on factors with a different distribution between 2 groups that might be associated with survival outcomes, such as tumor size, anatomic locations, lymphadenectomy, combined resection, pT stage, pTNM stage, and the number of harvested LNs. In light of the sample size and the distribution of patients in 2 groups, the No. 10 LND+ and No. 10 LND–groups were paired 1:1 based on these propensity scores using nearest neighbor matching method.
After repeatedly matching, we confirmed the caliper size as 0.1 to minimize bias and intensify the conclusions of our study. PSM was performed by packages of “Match It” (http://CRAN.R-project.org/package=MatchIt) in R (Version 3.6.1; R Foundation for Statistical Computing, Vienna, Austria).
Statistical methods
The normality tests of the data were performed routinely before statistical analysis by parametric tests. Quantitative data were expressed as mean ± SD. Two-tailed χ2 tests or Fisher exact tests were performed for the categorical variables. Overall survival (OS) was calculated according to Kaplan-Meier method and compared by the log-rank test. The survival curves were constructed by the packages of “survival,” “survminer,” and “ggplot2” (http://CRAN.R-project.org/package=survival) in R. The logistic regression was carried by the package of “stats,” “DAAG,” and “regplot” (http://CRAN.R-project.org/package=DAAG) in R.
The Cox proportional hazard regression model with conditional backward stepwise analysis was conducted to perform univariate and multivariate survival analysis, and the nomogram was developed by the package of Regression Modeling Strategies (http://CRAN.R-project.org/package=rms) in R. Decision curve analysis (DCA) was performed to determinate the clinical usefulness of a new nomogram.
The novel biomarker-based ABC (age, biomarkers, clinical history)-bleeding risk score for patients with atrial fibrillation: a derivation and validation study.
DCA was performed by the package of “rmda” (http://CRAN.R-project.org/package=rmda) in R. A stratified analysis was also performed to detect which subgroup of patients might obtain a survival benefit from the No. 10 LND, and their interactions with OS were also tested in our study. The results of stratified analyses were presented by a forest plot that was constructed using the packages “grid,” “magrittr,” “checkmate,” and “forestplot” (http://CRAN.R-project.org/package=forestplot) in R. All these analyses were performed in R (Version 3.6.1; http://www.Rproject.org/). A two-tailed P < .05 was considered with statistically significant.
Results
Comparison of clinicopathologic findings before and after PSM
There were 274 patients who underwent No. 10 LND among 837 patients undergoing TG for GC and EGJA. Comparison of baseline materials indicated that tumor size (P = .002), number of harvested LNs (P < .001), distribution of anatomic locations in the stomach (P = .015), lymphadenectomy (P < .001), combined resection (P = .009), and pT4 (P = .003) were different between 2 groups. The previously mentioned factors might bring different survival outcomes; therefore the covariates for PSM included tumor size, anatomic locations, lymphadenectomy, combined resection, number of harvested LNs, pT stage, and pTNM stage (P = .076). After 1:1 matching, there were 237 cases left in each group. After PSM, the clinicopathologic features between 2 groups were balanced before and after PSM (Table I).
Table IDetails of clinicopathologic features before and after propensity score matching
No. 10 lymph nodes were not detected in final pathologic reports in the No. 10 LND– group; a D2/D2+ lymphadenectomy was performed as a standard procedure.
† No. 10 lymph nodes were not detected in final pathologic reports in the No. 10 LND– group; a D2/D2+ lymphadenectomy was performed as a standard procedure.
‡ Combined resection of organs included gallbladder, spleen, left lobe of liver, transverse colon, pancreatic body and tail, and small intestine.
§ TNM stage was according to seventh American Joint Committee on Cancer staging system for gastric cancer.
Operative parameters and postoperative complications
The numbers of harvested LNs in the No. 10 LND+ group were greater than that of the No. 10 LND– (41.2 ± 16.0 vs 33.8 ± 11.3; P < .001) (Table I). The operative duration was also greater in No. 10 LND+ group (254 ± 43 vs 243 ± 53 minutes; P = .002). There were no differences in postoperative days, blood loss, and proportion of postoperative complications between 2 groups (18.6% [51/274] in the No. 10 LND+ group and 17.2% [97/563] in No. 10 LND– group; P = .692). There was also no difference for each grade of Clavien-Dindo classification for postoperative complications. Similarly, 30-day operative mortality also did not differ between 2 groups (1.1% [4/274] in No. 10 LND+ and 0.9% (5/563) in No. 10 LND– group; P = .721; Table II).
Table IIOperative parameters and Clavien-Dindo classification of postoperative complications for patients in our study
Operative parameters and classification of complications
The overall rate of No. 10 LN metastasis was 16.4% (45/274); however, when the tumor involved the greater curvature, the rate of No. 10 LN metastasis was greater (36.8% [28/76]; P < .001). The presence of No. 10 LN metastasis indicated a worse prognosis when compared with those without metastasis (median survival time 23.8 vs 88.4 months; hazard ratio [HR] = 2.992; P < .001; Supplemental Digital Contents Fig S1a). The logistic regression analysis confirmed that tumor size (odds ratio [OR] = 1.370, P < .001), involvement of the greater curvature (OR = 5.562, P = .002), combined resection (OR = 6.027, P = .002), and M stage (OR = 7.206, P = .002) were independent factors of No. 10 LN metastasis (Supplemental Digital Contents Table S1). The nomogram of multivariate logistic model for No.10 LN metastasis was also depicted in Supplemental Digital Contents Fig S1b.
Survival analysis of No. 10 LND
Before PSM, the median OS of patients in the No. 10 LND+ group was greater than that of No. 10 LND– group (66.3 vs 37.0 months; P = .003), and the estimated 5-year OS was also greater (51.7% vs 36.8%; P = .003) (Fig 1, A). After PSM, the group undergoing a No. 10 LND still had better median OS and 5-year OS (65.3 vs 35.3 months and 51.2% vs 35.6%; P = .003; Fig 1, A).
Fig 1(A) The overall survival curves of No.10. LN dissection in total gastric cancer patients before propensity score match (P = .003, HR = 1.360). (B) Survival curves of No. 10 LN dissection in EGJA patients before propensity score match (P < 0.001, HR = 1.816). (C) For Siewert type II before propensity score match (P = .002, HR = 1.839). (D) For Siewert type III before propensity score match (P = .032, HR = 1.882). (E) The overall survival curves of No. 10 LN dissection in total gastric cancer patients after propensity score match (P = .003, HR = 1.426). (F) Survival curves of No. 10 LN dissection in EGJA patients after propensity score match (P = .005, HR = 1.688). (G) For Siewert type II after propensity score match (P = .039, HR = 1.596). (H) For Siewert type III after propensity score match (P = .043, HR = 1.915).
Before PSM, there were 129 EGJA patients in No. 10 LND+ group and 255 in No. 10 LND– group and the estimated 5-year OS of the EGJA No. 10 LND+ group was greater than that of the No. 10 LN D– group (57.1% vs 38.3%; P < .001; Fig 1, B). Stratified analysis of the No. 10 LND+ versus the No. 10 LND+ groups according to Siewert types for EGJA indicated an OS for Siewert type II of 60.9% vs 35.4% (P = .002; Fig 1, C) and for Siewert type III of 48.7% vs 28.2% (P = .032; Fig 1, D). After PSM, there were 112 EGJA patients in both groups and a No. 10 LND still provided a 5-year OS benefit to EGJA patients (56.4% vs 34.7%; P = .005; Fig 1, B). The stratified analysis after PSM according to Siewert types also indicated a better OS for Siewert type II (59.8% vs 38.9%; P = .039; Fig 1, C) and for Siewert type III (49.4% vs 23.9%; P = .043; Fig 1, D).
Subgroups analysis according to clinical findings
We performed subgroups analysis stratified by clinicopathologic findings. The results also indicated that a No. 10 LND could bring a survival benefit for patients in TNM stage II (P = .024), pN+ (P = .016), and pM0 (P = .004); with adjuvant chemotherapy (P = .009) and without signet cells (P = .014); in tumors with poor differentiation (P = .005); in Bormann types 0 to II (P = .004); in tumors without greater curvature invasion (P = .001); and in tumors located along the lesser curvature (P = .008); and at the esophagogastric junction (P < .001). Stratified analysis indicated most subgroups could obtain potential survival benefit from No. 10 LND; these results are depicted with a forest plot (Fig 2).
Fig 2Forest plot for subgroup analysis: a statistically significant interaction was found in EGJA patients, No. 10 LN dissection indicated significantly better survival than without No. 10 LN dissection among EGJA patients.
Before PSM, the Cox univariate and multivariate analysis indicated that No. 10 LND (HR = 1.284, 95% confidence interval [CI] 1.042–1.584; P = .019), tumor size (HR = 1.060, 95% CI 1.023–1.098; P = .001), differentiation grade (HR = 1.347, 95% CI 1.068–1.700; P = .012), cross-sectional anatomic location (P = .002), T stage (P = .004), N stage (P < .001), and M stage (HR = 1.967, 95% CI 1.418–2.730; P < .001) were all independent prognostic factors for GC in our study (Table III). After PSM, the Cox multivariate analysis also indicated that a No. 10 LND (HR = 1.480, 95% CI 1.156–1.89; P = .002), tumor size (HR = 1.058, 95% CI 1.010–1.109; P = .018), cross-sectional anatomic locations (P = .003), N stage (P < .001), and M stage (HR = 1.617, 95% CI 1.052–2.484; P = .028) remained independent prognostic factors for GC (Supplemental Digital Contents Table S2).
Table IIIUnivariate and multivariate survival analysis in this study by Cox proportion hazard model before PSM
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
3.177
0.967–10.43
.057
∗ Others include upper third gastric cancer, middle third of stomach, distal gastric cancer, and tumor invading the entire stomach. / indicate that factors in front of the slash (/) are treated as the reference of the factors behind the slash.
Based on the independent prognostic factors confirmed by Cox multivariate analysis after PSM, the predictive nomograms of patients with No. 10 LND and patients without No. 10 LND were established respectively (Fig 3, A and 3, D). The C-index of the No. 10 LND+ nomogram (0.735, 95% CI 0.695–0.774) was greater than that of the No. 10 LND– (0.708, 95% CI 0.68–0.748; P < .001). The calibration curves of the No. 10 LND+ nomogram also had better consistency with ideal predictive curves (Fig 3, B and 3, E).
Fig 3(A) Nomogram of patients with No. 10 LN dissection after propensity score match. (B) Calibration curves for nomograms of patients with No. 10 LN dissection after propensity score match. (C) Decision curves for patients with No. 10 LN dissection after propensity score match. (D) Nomogram of patients without No. 10 LN dissection after propensity score match. (E) Calibration curves for nomograms of patients without No. 10 LN dissection after propensity score match. (F) Decision curves for patients without No. 10 LN dissection after propensity score match.
We also performed DCA on these 2 nomograms to calculate the net benefit of a No. 10 LND quantitatively. DCA revealed that the threshold probability of a No. 10 LND+ nomogram was nearly 11%, which was less than 31% in No. 10 LND– nomogram. Accordingly, the optional range of probability that patients could derive a net survival benefit of a No. 10 LND+ was statistically significantly greater than No. 10 LND– (0.11–0.84 vs. 0.31–0.89). DCA indicated that No. 10 LND in TG could predict nearly 15% more patients would obtain a net survival benefit (Fig 3, C and 3, F).
Discussion
In recent decades, TG plus standard D2 lymphadenectomy is gradually becoming a mainstay for surgical treatment of GC with the increased incidence of EGJA and upper third gastric cancer.
Gastric cancer surgery: morbidity and mortality results from a prospective randomized controlled trial comparing D2 and extended para-aortic lymphadenectomy–Japan Clinical Oncology Group study 9501.
The updated 15th edition of the treatment guidelines of the JGCA excluded a No. 10 LND as performed concomitantly with a splenectomy from the domain of a D2 lymphadenectomy in TG based on the result of the JCOG-0110 trial.
But the JCOG-0110 only addressed the value of splenectomy in TG for proximal gastric carcinoma without greater curvature invasion, whereas the value of a spleen-preserving No. 10 LND was not further investigated.
Therefore, we performed this retrospective analysis to determine the importance of a spleen-preserving No. 10 LND. Based on our results, we found that No. 10 LND brought a considerable OS benefit for patients who underwent TG with a spleen-preserving No. 10 LND.
Several previous reports stated that the metastatic rate of No.10 LNs in patients with middle or upper third gastric cancers was 8.8% to 20.9%, which was a bit less than our results.
Considering the metastatic rate, the poor prognosis of patients with metastases to the No.10 LNs, and the lack of difference in complications with and without spleen-preserving No. 10 LND, we maintain that a No. 10 LND should be an essential procedure for improving the survival outcomes in patients with appropriate risk factors for the presence or suspicion of metastatic involvement. Previous studies also indicated that the metastatic rate of the splenic hilar nodes exceeded 10%, which was similar to the rate of positive perigastric nodes, such as No. 4sa and No. 4sb, and they suggested that the priority of a No. 10 LND was greater than that of Nos. 11p, 11d, and 8a LNa.
These findings might be interpreted to indicate that the route of No.10 LN metastasis was similar to that of the No. 4sa and No. 4sb LNs, which followed the greater curvature lymphatics to splenic hilum. Our study also confirmed that incision of the greater curvature was an independent risk factor for No. 10 LN metastasis because lymph nodal pathways in this location would be more at risk for drainage to the splenic hilar region along left gastroepiploic, short gastric, posterior gastric, and splenic vessels. In our study, the metastatic rate of LN 4sa was 6.9% and 4sb was 16.1%, suggesting that the left gastroepiploic vessels were one of the main routes of metastasis to the splenic hilum, which further supported the suggestion that a No. 10 LND should be performed as a standard component of a D2 lymphadenectomy for advanced gastric cancer, especially for tumors of the greater curvature.
also supported the therapeutic value of a No. 10 LND because this nodal group was close to the perigastric lymph nodes like such as Nos. 3, 1, 7, and 2 LNs. The metastatic rate of the No. 10 LN in the JCOG-0110 trial was only 2.4%, which was significantly less than many other reports; consequently, the JCOG-0110 trial could not detect the importance of a spleen-preserving No. 10 LND.
In our study, possibly because of the relatively small sample size of greater curvatures, a No. 10 LND did not confer a survival benefit; however, the obviously greater metastatic rate of these patients should be given more attention by clinical surgeons. In the past a No. 10 LND would be performed by its inclusion with the splenectomy, which was later considered to be too aggressive, whereas the therapeutic value of spleen-preserving No. 10 LND should not be overlooked.
In contrast, the development of more advanced operative techniques and alternative devices for dissection/hemostasis allow a safer spleen-preserving No. 10 LND, for instance, the amplifying effect of a laparoscopic or robotic approach allows better visualization of the anatomic layers and satisfactory short-term results.
Furthermore, 3-dimensional imaging technology as applied to CT images has made the preoperative evaluation of the splenic vascular anatomy even better, thereby enhancing the safety of a No.11 LN and No. 10 LND and allowing a spleen-preserving LN dissection.
The minimally invasive operations and advanced preoperative evaluation have made a No. 10 LND better tolerated and more feasible without unduly prolonging the operative duration.
For EGJA patients, the need for a No. 10 lymphadenectomy was always controversial. Most previous studies have rejected the therapeutic value of No. 10 LND in EGJA considering the low metastatic rate and the index of estimated benefit from lymph node dissection.
The optimal extent of lymph node dissection for adenocarcinoma of the esophagogastric junction differs between Siewert type II and Siewert type III patients.
The treatment guidelines of the JGCA (Version 4) also stipulated that a No. 10 LND should be excluded from a conventional D2 lymphadenectomy for EGJAs for tumors with a diameter ≤4 cm.
Contrary to previous studies, our study found a substantially greater incidence of No. 10 LN metastasis (16.3%) and confirmed a considerable survival benefit of a No. 10 LND in patients with EGJA. The peculiar anatomic location and biologic properties of EGJA might imply a special lymphatic flow pathway. Indeed, Siewert type III carcinomas were similar to other advanced middle and upper third GCs, which might suggest a greater incidence of No. 10 LN metastasis.
The optimal extent of lymph node dissection for adenocarcinoma of the esophagogastric junction differs between Siewert type II and Siewert type III patients.
As we know, the esophagogastric junction and the cardia are not totally covered by visceral peritoneum, which might allow EGJAs to more readily infiltrate the serosa and lead to lymphatic metastases spreading along with No. 2 and No. 4sa LNs to the splenic hilum,
Comparison on clinicopathological features and prognosis between esophagogastric junctional adenocarcinoma (Siewert II/III types) and distal gastric adenocarcinoma: retrospective cohort study, a single institution, high volume experience in China.
which would explain the greater incidence of No. 10 LN metastasis in EGJA. In addition, many collateral lymphatic pathways might exist or develop among the pericardial, left inferior phrenic, and superior splenic pole vessels.
This study has the potential limitations inherent to any retrospective study, including selective bias, detective bias, and statistical bias of such a study format. Second, our research was based on a database from a single center that lacked any internal or external validation of the accuracy of the data extraction. Third, although the clinical findings of patients with and without a No. 10 LND did not appear to be different after our PSM, there were potentially important differences on the mean number of harvested LNs between the 2 groups. Finally, there were only 68 patients with GC invading the greater curvature who underwent a No. 10 LND in our study; although we performed a survival analysis and a stratified analysis for this group and we found a potential trend of better prognosis with a No. 10 LND, the statistical analysis was not robust.
In conclusion, a spleen-preserving splenic hilar lymph node dissection should be recommend in patients undergoing a TG, especially for adenocarcinomas of the esophagogastric junction.
Conflict of interest/Disclosure
The authors have no conflicts of interest or financial ties to disclose.
Acknowledgments
All the authors thanks the funding sources which was supported by grant from (1) Sichuan Province Youth Science & Technology Innovative Research Team, No.2015TD0009; (2) 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University, No. ZY2017304; (3) Fund for Fostering Academic and Technical Leaders of Sichuan Province, No. [2016] 183-19, No. [2017] 919; and (4) Sichuan Province Cadre Health Care Research Project (No.2017–114). The authors thank the substantial work of Volunteer Team of Gastric Cancer Surgery (VOLTGA) based on the multidisciplinary team (MDT) of gastrointestinal tumors, West China Hospital, Sichuan University, China, for the establishment of gastric cancer database and the continual follow-up. The authors thank Danil Galiullin for assistance with English grammar in our manuscript.
Gastric cancer surgery: morbidity and mortality results from a prospective randomized controlled trial comparing D2 and extended para-aortic lymphadenectomy–Japan Clinical Oncology Group study 9501.
Survival benefits from splenic hilar lymph node dissection by splenectomy in gastric cancer patients: relative comparison of the benefits in subgroups of patients.
Splenic hilar lymph node metastasis independently predicts poor survival for patients with gastric cancers in the upper and/or the middle third of the stomach.
Should splenic hilar lymph nodes be dissected for Siewert type II and III esophagogastric junction carcinoma based on tumor diameter? A retrospective database analysis.
The novel biomarker-based ABC (age, biomarkers, clinical history)-bleeding risk score for patients with atrial fibrillation: a derivation and validation study.
The optimal extent of lymph node dissection for adenocarcinoma of the esophagogastric junction differs between Siewert type II and Siewert type III patients.
Comparison on clinicopathological features and prognosis between esophagogastric junctional adenocarcinoma (Siewert II/III types) and distal gastric adenocarcinoma: retrospective cohort study, a single institution, high volume experience in China.
30526-2&id=gr1.jpg){kind=link}
30526-2&id=gr2.jpg){kind=link}
30526-2&id=gr3.jpg){kind=link}
30526-2&id=figs1.jpg){kind=link}