Introduction
Gastric cancer (GC) is one of the leading digestive system malignancies. According to the data of GLOBOCAN in 2020, GC ranked the fifth in morbidity and the fourth in mortality among all malignancies worldwide [
1]. In recent years, there has been an increasing trend in the incidence of early gastric cancer (EGC) and proximal gastric cancer, and proximal gastrectomy (PG), which is one of the function-preserving surgery has gained extensive attention [
2,
3]. In the Japanese Gastric Cancer Treatment Guidelines 2021 (6th edition), PG is suggested for proximal gastric cancer with clinical stage of cT1N0 to preserve more than half of the distal stomach [
4].
PG is capable of maintaining the volume of the remnant stomach and preserving the physiologic functions, such as the secretion of the intrinsic factors. These advantages can decrease the incidence of postoperative body weight loss and anemia compared to total gastrectomy (TG) [
5]. However, due to the damage to the lower esophageal sphincter and the angle of His, the postoperative complications caused by PG, especially the reflux esophagitis (RE), may severely impair postoperative quality of life (QoL). Various functional digestive tract reconstruction methods after PG have been reported to reduce postoperative complications, such as double tract reconstruction, jejunum interposition and gastric tube reconstruction. But there is no consensus on the optimal reconstruction method after PG.
This systematic review and meta-analysis aim to comprehensively search comparative studies concerning reconstruction methods after PG and compare the postoperative short-term outcomes and long-term QoL.
Materials and methods
This systematic review and meta-analysis was conducted following Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statements [
6]. The protocol was registered on the PROSPERO website with the registration number CRD42022381357.
Literature search strategy
Two authors (Bailong Li and Yinkui Wang) independently searched the databases of PubMed, Embase, Web of Science, and Cochrane Library to identify relevant studies published before December 2022. The search terms “proximal gastric cancer”, “proximal gastrectomy”, “reconstruction”, “anastomosis” and Medical Subject Headings (MeSH) “surgical anastomosis” were used to form the search strategies. Moreover, the references cited in the included articles and previously published reviews were manually searched to identify additional relevant studies. If duplicated studies were published by the same authors with an accumulated number of patients or updated follow-up, only the latest article was included.
Inclusion and exclusion criteria
The inclusion criteria were: (1) Patients were pathologically diagnosed with GC or adenocarcinoma of the esophagogastric junction (AEG); (2) Patients underwent laparoscopic proximal gastrectomy, laparoscopic-assisted, or open proximal gastrectomy; (3) The reconstructive techniques included using linear stapler, circular stapler, and hand-sewn anastomosis. (4) Comparative studies concerning two or more digestive tract reconstruction methods after PG.
The exclusion criteria were: (1) Patients were pathologically diagnosed with other digestive tract tumors, such as gastrointestinal stromal tumors; (2) Single-arm study; (3) Case reports or all the cases added up less than 10; (4) Review articles, meeting abstracts and comments; (5) Studies that did not provide necessary data for statistical analysis.
Outcome definition
The postoperative short-term outcomes included postoperative complications, such as anastomotic leakage and anastomotic stricture. The surgical and postoperative recovery indicators were also collected, including intraoperative blood loss, operation time and postoperative hospital stay. The long-term QoL included postoperative symptoms such as reflux, dysphagia and distention and nutritional status. The nutritional status included body weight loss and hematological indexes, such as hemoglobin, albumin, and vitamin B12. The reflux esophagitis was confirmed by endoscopy 12 months after surgery and classified by Los Angeles classification [
7]. Los Angeles classification degree B or more severe degrees were recorded and analyzed. The degree of food residue was evaluated according to the RGB classification, and a grade ≥ 2 was recorded [
8].
Data extraction and quality assessment
Two authors (Bailong Li and Yinkui Wang) independently screened titles, abstracts, and full texts of studies. Disagreement between the two reviewers was resolved through a discussion with another author (Fei Shan).
Data of studies were extracted onto Excel spreadsheets by two authors (Bailong Li and Baocong Li). The following data were extracted: (1) Study characteristics, including the first author, country, study period, year of publication, study design, and the number of patients in each group; (2) Patient characteristics, including age, gender, tumor location, surgical approach, the reconstruction methods, the reconstruction techniques and the size of the remnant stomach; (3) Short-term outcomes, including the number of the overall postoperative complications, the number of specific complications; the intraoperative indexes, such as intraoperative blood loss, operation time, and postoperative hospital stay; (4) Long-term QoL, including the number of patients with postoperative long-term symptoms, the body weight loss, and hematological indexes 12 months after surgery.
Two authors (Bailong Li and Baocong Li) independently assessed the quality of included studies. The disagreement between the two authors was resolved by discussion. Newcastle-Ottawa quality assessment scale (NOS) was used to evaluate the risk of bias in nonrandomized studies [
9]. Considering 8 items associated with selection, comparability and exposure, and the total score was calculated by summing the values. The total score was 9. Cochrane Handbook for Systematic Reviews of Interventions was used to assess the risk of bias in randomized controlled trials (RCTs) [
10]. The risk of bias consists of selection bias, performance bias, detection bias, attrition bias, reporting bias, and other bias. Each term was judged as “low risk”, “high risk” or “unclear risk”.
Statistical analysis
Review Manager (RevMan) (Version 5.4, The Cochrane Collaboration, 2020) was used to perform the statistical analysis. Continuous variables were pooled using the mean difference (MD) with a 95% confidence interval (95% CI), and discontinuous variables were pooled with an odds ratio (OR) with 95% CI. Heterogeneity was tested with Cochran’ Q test and quantified with the I
2 inconsistency test. Heterogeneity was graded as low (I
2 < 25%), moderate (I
2 = 25–50%), and high (I
2 > 50%). If I
2>50%, a random effects model was adopted, otherwise, a fixed effects model was used. Sensitivity analysis was performed to identify the sources of heterogeneity by removing one study in meta-analysis at a time. Subgroup analysis was conducted to adjust for the heterogeneity if appropriate. Publication bias was explored graphically with funnel plots to detect asymmetry and outliers. In some studies, mean value and standard deviation were estimated from the median, range, and sample size according to Hozo’s report [
11]. A
p-value<0.05 is considered statistically significant.
Discussion
In previous studies, PG could maintain comparable oncological radicalness to TG in EGC [
39]. Moreover, in the KLASS 05 randomized clinical trial, patients who received PG showed better nutritional status and physical and social functions than patients received TG [
40]. However, postoperative complications such as RE and anastomotic stricture may severely impair the long-term QoL after PG. Therefore, the effectiveness of anti-reflux, the incidence of postoperative complications and the QoL of various reconstruction methods after PG need to be compared urgently. By comprehensively searching the literature, a total of 27 comparative studies were identified and included in this systematic review, including 4 RCTs and 23 nonrandomized studies. The present study is an updated meta-analysis with a larger sample size to evaluate the short-term outcomes and long-term QoL of various reconstruction methods. To our best knowledge, this is the first systematic review and meta-analysis involving patients’ long-term QoL and nutritional status after PG. In addition, this is also the first meta-analysis that comprehensively included comparative studies concerning the reconstruction methods after PG.
This systematic review divided reconstruction methods into esophagogastric anastomosis (EG group) or esophagojejunal anastomosis (EJ group). We found that EJ group had advantages in decreasing the incidence of RE and stricture compared to the EG group. The previous studies showed that esophagogastric anastomotic sites were narrower and more likely to develop strictures than esophagojejunal anastomoses, which was consistent with our results [
12]. However, due to the technique complexity, the EJ group required more operation time and presented more intraoperative blood loss and longer postoperative hospital stay than the EG group. Additionally, fewer early postoperative complications were observed in EG group. These results suggested that esophagogastric anastomosis and esophagojejunal anastomosis each had its advantages. Concerning the postoperative nutritional status, we found that the EG group had advantages in maintaining weight than the EJ group. The possible reason was that all the food passed through the stomach and duodenum passage in the EG group, and the nutrient substance could be absorbed fully. But there was no difference in hemoglobin and albumin 12 months after surgery between the two groups.
Esophagogastrostomy is a conventional method with technical simplicity advantages. The qualitative analysis showed that the incidence of anastomotic leakage of esophagogastrostomy was the lowest among all reconstruction methods, and the incidence of early postoperative complications was lower than DTR, JI and JPI. Moreover, the operation time was shorter and intraoperative blood loss was lesser. However, over one fourth of patients developed RE after esophagogastrostomy, which was highest in all reconstruction methods. And it showed a higher incidence rate of anastomotic stricture than DTR. Some previous studies showed that the incidence of RE could be reduced by modified operation, such as side overlap esophagogastrostomy and SPADE operation [
41,
42]. However, the effectiveness of these modified operations remains to be confirmed by studies with larger sample size.
Many studies demonstrated that jejunal interposition was associated with a lower risk of RE. Our study showed that JI could significantly decrease the incidence of RE to 7.1% compared to EG. Despite the favorable anti-reflux effectiveness, the procedure was more technically complex, requiring three anastomoses. Compared to EG, the operation time was longer, the blood loss was more, and the postoperative hospital stay was also longer. However, no significant difference in postoperative leakage and early postoperative complications was found between JI and EG, which was consistent with the previous reports [
43]. It might attribute to the skillful operation of surgeons and anastomosis performed by stapler under direct vision [
21,
22]. In the present study, JPI had no notable advantages in RE, anastomotic stricture, and postoperative complications compared to JI. The operation time was longer in JPI. Additionally, by endoscopic examination, Nakamura et al. reported that residual food was revealed in 92% of patients after JPI, which might decrease the long-term QoL [
25].
In terms of DTR, our results showed an excellent effectiveness in anti-reflux, with the incidence of RE decreasing to 7.6%. Compared to esophagogastrostomy, DTR showed lower rates of anastomotic stricture. In terms of surgical safety, although DTR needed three anastomoses and longer operation time, there was no significant difference in postoperative complications between esophagogastrostomy and DTR. These results suggested that DTR was a safe technique after PG. Compared to JI, DTR adds a side-to-side anastomosis between remnant stomach and jejunum, effectively avoiding emptying dysfunction. In addition, we found that the residual food was least in DTR among all reconstruction methods. Regarding postoperative nutrition, we found no difference in hemoglobin, albumin, vitamin B12, and iron between esophagogastrostomy and DTR. Two studies previously reported the long-term QoL of DTR [
15,
19]. Ji reported that DTR had apparent advantages over esophagogastrostomy based on EORTC QLQ-C30 and EORTC QLQ-STO22 questionnaires [
19]. To sum up, DTR is a safe technique with excellent anti-reflux effectiveness and long-term QoL.
In the present study, GT showed a good effectiveness in reducing RE. In qualitative analysis, only 4.5% of patients presented RE which was the lowest in all methods. However, the anastomotic stricture occurred in 14.5% patients who underwent GT, which was highest. Moreover, only a few studies with small sample sizes investigated this reconstruction method, and the effectiveness remains to be confirmed.
In the previous study, Shoji et al. reported a 4.2% of incidence rate of RE after the DFT [
44]. A meta-analysis also demonstrated that DFT could lower the rates of complications compared to DTR, EG, JI and JPI [
43]. However, all four studies included in the meta-analysis were single-arm studies. By comprehensive searching, only one comparative study investigating DFT was included in the present study.
There were several limitations in the present study. First, most of the studies included were retrospective and nonrandomized studies. This certainly attenuated the evidence level. Second, in terms of postoperative nutritional status, some original data could not be obtained. And the questionnaires for evaluating long-term QoL were quite inconsistent between studies, so some long-term indexes could not be compared. Third, all the studies included in the present study were reported by Asian authors. The results need further confirmation in other countries. Fourth, the reconstruction techniques might be related to the incidence of postoperative complications. However, no studies included in the meta-analysis provided the comparative data in terms of the reconstruction techniques.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.