Shu-Wen Qin ,Yue Zhang ,Jia-Wei Fan ,Ying Wang ,Xiao-Qing Shi ,Li-Fen Mao,*

1Nursing School of Soochow University,Suzhou,China.2Department of Nursing,The First Affiliated Hospital of Soochow University,Suzhou,China.

Abstract The current study aimed to summarize valid and up‑to‑date high‑quality evidence to compare the impact of early oral feeding (EOF) and delay oral feeding (DOF) among patients undergoing gastrectomy.A comprehensive literature review of the PubMed,Cochrane library,Medline,Embase,the CNKI,Wangfang Library and CBM databases was conducted from the earliest data of each to May 2021.Studies were selected to compare the EOF protocol versus DOF protocol undergoing gastrectomy.The primary outcome assessed was anastomotic leaks.Secondary outcomes included time to passage of flatus,postoperative ileus and postoperative length of hospital stay(LOS).Analysis was performed with Review Manager 5.3 software.Sisteen studies,comprising of 3477 patients were included in the analysis.There was no significant difference in risk of anastomotic leak (odds ratio [OR] 0.73;95% confidence interval [CI] 0.45 to 1.20; P=0.22; I2=0%; P=0.47) and postoperative ileus (OR 1.61;95% CI 0.93 to 2.77; P=0.09;I2=0%, P=0.85) between the EOF and DOF groups.Patients had a reduction in time to passage of flatus (WMD -0.66d;95% CI -0.86 to -0.46; P <0.001; I2=94%; P <0.001),and postoperative LOS (WMD -1.66d,95% CI -1.97 to -1.35, P <0.001; I2=65%).EOF as compared with DOF is not associated with an increase in clinically relevant complications,Moreover,EOF is associated with shorter LOS and early the recovery of gastrointestinal function.

Keywords:feeding;gastrectomy;Meta‑analysis;nutrition;systematic review

Introduction

In China,gastric cancer (GC) is the second most common cancer and the second in cancer‑related deaths [1].Surgical resection is still the primary curative treatment for GC.Gastrectomy is one of the most complicated surgical procedures in the department of gastrointestinal surgery accompanying clinically significant postoperative stress and symptoms.Available studies have suggested that gastrectomy‑related morbidity and mortality range from 10 to 46% and from 0 to 13%,respectively [2,3].The nutritional status for patients undergoing major abdominal surgery is a key factor determining the patients’postoperative outcome.Usually,oral feeding is delayed after gastrectomy,until the passage of flatus due to the traditional point that the anastomosis stressed by food before gastrointestinal function will increase the risk of anastomotic leakage and intestinal obstruction[4,5].

Henrik Kehlet [6] proposed the concept of enhanced recovery after surgery (ERAS),it can really modulate surgical stress response,reduce postoperative complications,accelerate functional recovery,shorten the length of hospital stay,for which the concept quickly gained popularity among surgeons globally.Early oral feeding (EOF) as the most important contents of ERAS,was introduced for upper gastrointestinal surgery.Clinical randomized controlled trials (RCTs)[7-9] has already indicated the advantages of EOF after colorectal cancer surgical resection,such as shorten postoperative length of stay(LOS) and decreased postoperative morbidity and mortality compared with traditional time of oral feeding.A meta‑analysis [10] about open upper gastrointestinal surgery showed that EOF could reduce the length of hospital stay and bowel recovery time without increasing postoperative complications.However,the safety and feasibility of EOF after gastrectomy is still controversial,and high‑quality research evidence supporting this practice is scarce [10].

The purpose of the current meta‑analysis is to provide a comprehensive,updated summary of relevant high‑quality evidence comparing the effect of EOF and DOF in adults undergoing gastrectomy.Historically,surgeons tended to believe it was required that gastrointestinal decompression with a nasogastric tube and a fasting period of 2-5 days after gastrointestinal surgery was performed to prevent anastomotic leakage [5].Therefore,in this study,the primary outcome measure was anastomotic leakage.Secondary outcomes measures were time to passage of flatus and postoperative LOS.

Methodology

Search Strategy

A comprehensive literature review of the PubMed,Cochrane library,Medline and Embase,the CNKI,Wangfang Library and CBM databases were conducted from the earliest data of each to May 2021.All studies evaluating the effect of EOF versus DOF protocols in patients undergoing open or laparoscopic gastrectomy were identified using search criteria.The search string used the following keywords and was modified for each:(‘‘gastrectomy’’ [MeSH] OR gastric cancer OR gastric neoplasm OR gastric carcinoma OR gastric tumor OR stomach cancer OR stomach neoplasms OR gastric surgery OR stomach carcinoma) AND (‘‘oral feeding’’ [MeSH] OR intake) (See Supplement Table 1).There were no language restrictions.This systematic review and meta‑analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta‑Analyses (PRISMA) statement[11].

Selection criteria

All eligible studies conformed to the following criteria:(1) age≥18 years,(2) Participants were gastrectomy patients;(3) studies should reporte at least one of the following index:anastomotic,time to passage of flatus,postoperative LOS;(4) randomized controlled trials and cohort trials comparing EOF and DOF protocols were included;(5)studies were completed in May 2021 with a completed dataset.In this study,early feeding and late or traditional feeding all were defined by the authors of the studies.

Data Extraction

Two authors (SQ,YZ) independently extracted relevant data from eligible studies according to predefined criteria.Extracted data dimensions included study characteristics (country,date,study design,sample size),baseline population characteristics (age,type of surgery,sex),intervention,comparison and outcomes of interest (anastomotic leak,time to passage of flatus,postoperative LOS).

Risk of Bias:Assessment of Individual Studies

Cochrane Collaboration’s tool [12] and the Newcastle‑Ottawa Scale[13] were used to assess risk of bias of the RCT studies and the individual nonrandomized studies,respectively.Two authors (JF,YW)evaluated the quality of all articles independently.Any disagreements were discussed and solved by consensus.For RCT studies,low risk bias of Cochrane Collaboration’s tool at least five domains and less than four domains corresponded to high and low quality,the remaining studies were of moderate quality.For cohort studies,the total scores of the NOS were 9 and 0～3,4～6,7～9 was considered as low,moderate and high quality,respectively.

Statistical Analysis

We summarized the baseline characteristics of the included patients,chi‑square test andttest were conducted to compare differences between groups.The results for dichotomous outcomes were reported as odds ratio (OR) with 95% confidence interval (CI) and as weighted mean difference (WMD) for continuous data.TheI2statistic was used to assess the statistical heterogeneity.The fixed effect model was applied for meta‑analysis when the heterogeneity was low (I2≤50%,P＞0.01).When the heterogeneity was high (I2≥50%,P＜0.01),the source of heterogeneity would need to be identifyed,then we run a subgroup analysis according to the source of clinical heterogeneity,the random effect model was applied if clinical heterogeneity was not existed,only descriptive analysis was used once the source of heterogeneity could not be determined.the meta‑analysis was conducted by Review Manager Version 5.3 software (Cochrane Collaboration,Oxford,England;http://www.cochrane.org/software/revman.htm) [14].

Results

Identification of studies

A total of 3477 records were retrieved from Cochrane library (430 studies),PubMed (361studies),Medline (1900 studies),Embase (668 studies),the CNKI (33 studies),the Wanfang Library (69 studies) and the CBM (16 studies),of which 933 were duplicates.The selection process and search results were summarized in a flowchart (Figure 1).No limitation in terms of the types of clinical trial design were adopted.

Characteristics of studies

Sixteen studies were included [15-30].A total of 3226 patients undergoing gastrectomy were randomized to either EOF protocol (n=1609) or DOF protocol (n=1617).Table 1 summarized the characteristics of the sixteen included quantitative studies.Among the included sixteen studies,six were RCT studies[20,22,24,27,28,30],ten were cohort studies [15-19,21,23,25,26,29].Ten studies were performed in China [15,17,19-22,24,25,27],two in South Korea[18,26],two in Korea [28,29],one in Japan [30] and one in Italy[23].In the RCTs,the risk of bias was either low or unclear using the Cochrane Collaboration’s tool (Table 2).The non‑RCTs were assessed with the Newcastle‑Ottawa Scale for risk of bias resulting in 0 to 4 stars per category,indicating high to low bias (Table 2).All studies had low to moderate quality.

Interventions and Cointerventions

As inclusion criteria of this systematic literature review,patients in intervention group started oral feeding were earlier than them in control group for every study (Table 1).No notable differences were observed between the two groups regarding the baseline patient characteristics(Table 3).

Table 1 Characteristics of the included studies

Table 2 Quality assessment of included studies

Table 3 Baseline characteristics of included patients

Table 3 Baseline characteristics of included patients (Continued)

Primary Outcome

Anastomotic leakage.Nine of the studies (2364 patients) [16,18-20,23,25,26,30] reported anastomotic leak.All data in the included studies were synthesized in forest plots,and there was no significant difference between the EOF and DOF groups with low heterogeneity(OR 0.73;95%CI 0.45 to 1.20;P=0.22;I2=0%;P=0.47)(Figure 2).This result indicated association between EOF/DOF and anastomotic leakage after gastrectomy.

Secondary Outcomes.

1.Time to passage of flatus

Fifteen studies (3123 patients) [15-26,28-30] were enrolled in the meta‑analysis of time to passage of flatus.Since the statistical heterogeneity among studies was high,random effect model was used(I2=94%;P<0.001).The results indicated time to passage of flatus in the EOF group was significantly shortened (WMD -0.66d;95% CI-0.86 to -0.46;P<0.001).Subgroup analysis based on different surgical approach showed the consequent conclusion:the patients in the laparoscopic subgroup [16,17,20-22,25] (WMD ‑0.98d;95% CI-1.44 to -0.51;P<0.001),the patients in the open subgroup [24,27-29] (WMD -0.58d;95% CI -0.91 to -0.24;P<0.001),the patients in the ‘Not Reported’ subgroup [15,18,19,23,26] (MD-0.38d;95% CI -0.63 to-0.14;P<0.001) (Figure 3)

2.Postoperative ileus

Ten of the studies (2462 patients) [15,16,18-23,26,30] reported postoperative ileus.The pooled results showed,compared with DOF,no significant difference in the risk of postoperative ileus in EOF group with low heterogeneity (OR 1.61;95% CI 0.93 to 2.77;P=0.09;I2=0%,P=0.85) (Figure 4).

3.Postoperative LOS

Eleven of the studies (2100 patients) [15,16,18‑21,25,26,28,30]reported postoperative LOS.The high statistical heterogeneity among studies determined the random effects model was used (I2=65%;P<0.001).The meta‑analysis revealed that the mean hospital stay was significantly shorter for EOF group than DOF group (WMD -1.66d,95% CI -1.97 to -1.35,P<0.001).Subgroup analysis was performed according to different surgical approach and the results showed same conclusion:the patients in the laparoscopic subgroup [16,20-22,25](WMD -1.72d,95% CI -2.14 to -1.29,P<0.001),the patients in the open subgroup [24,27,28] (WMD -1.93d,95% CI -2.32 to -1.54,P<0.001),the patients in the ‘Not Reported’ subgroup [15,19,26](WMD ‑1.47d,95% CI-2.29 to-0.65,P<0.001) (Figure 5).

Figure 1 Preferred Reporting Items for Systematic reviews and Meta-Analysis(PRISMA) flow diagram

Figure 2 Meta-analyses of anastomotic leakage between EOF group and DOF group.

Figure 3 Meta-analyses of time to passage of flatus between EOF group and DOF group.

Figure 4 Meta-analyses of postoperative ileus between EOF group and DOF group.

Figure 5 Meta-analyses of postoperative LOS between EOF group and DOF group.

Discussion

Gastric cancer is the fourth largest malignant tumor in the world,surgical resection is the only radical treatment GC,but it is also associated with extensive morbidity and mortality.EOF,as an important component of the ERAS program,has been proposed for the management of gastrointestinal and colorectal surgery [31].It is not only beneficial to maintain the integrity of intestinal mucosal barrier and promote the recovery of gastrointestinal function,but also can reduce the incidence of complications,shorten the LOS and reduce hospitalization costs.At present,the advantages of ERAS in surgery,especially in the field of tumor surgery,have been confirmed by many high‑quality research evidence.However,for patients undergoing gastrectomy,it is still getting a lot of attention and controversy about whether EOF could promote postoperative rehabilitation due to the concern of postoperative anastomotic leakage and other complications.High‑quality evidence focusing on the safety,feasibility and short‑term clinical outcomes of EOF after GC surgery is still scarce.

In this systematic review and meta‑analysis,the researchers compared the effect of EOF with traditional oral feeding after gastrectomy in terms of anastomotic leak,time to passage of flatus,postoperative ileus,postoperative ileus LOS.In total,sixteen studies,including six RCTs,met the predetermined inclusion criteria.

According to the data includes in the present study,no significant difference of incidence rate of anastomotic leakage was observed between the EOF and DOF groups with low heterogeneity (I2=0%).The traditional view believed that postoperative fasting and placement of the nasogastric tube,as an necessary treatment,can really bring down the pressure in the digestive tract,reduce the anastomotic edema and provide sufficient time for anastomotic healing.However,our meta‑analysis result seem to be different.Gagner et.al [32] had argued that the occurrence of anastomotic leakage was mainly related to anastomotic ischemia,edema or improper anastomosis,otherwise,the healing of anastomosis mainly depends on the moderate tension of intestinal tissue on both sides of the anastomosis and good blood supply.Therefore,we can infer that the occurrence of anastomotic leakage after gastrectomy has little relationship with EOF,which is more related to the insufficient intraoperative anastomotic suture or insufficient postoperative anastomotic blood supply.At the same time,laparoscopic radical gastrectomy had obvious advantages in delicate operation,which can really reduce the occurrence of anastomotic leakage,in this meta‑analysis,over one third of patients received laparoscopic gastrectomy [33‑35].Moreover,a previous study [36] had been examined that EOF can provide immunological and nutritional benefits with better protein kinetics and preservation of the immune system than DOF.EOF can really accelerate wound healing and increase anastomotic strength [37].Animal studies has already confirmed that starvation after intestinal anastomosis resulted in poor quality of healing,and EOF can increase wound healing and strength in anastomoses[38‑40].Furthermore,for gastrointestinal anastomosis,the excretion amount of gastric fluid in the residual stomach and the possible reflux flow of intestinal fluid through the gastrointestinal anastomosis should not be less than 1000ml per day,in contrast,100‑200ml daily transoral fluid diet in the early postoperative period does not significantly increase anastomotic pressure,so there is no need to worry too much about the effect of EOF on anastomotic healing.Anyway,on the basis of mastering anastomotic techniques,oral feeding can promote gastrointestinal contraction,increase gastrointestinal blood perfusion,prevent food retention and secondary intestinal obstruction,avoid high pressure in gastrointestinal lumen,and then reduce the incidence of anastomotic leakage.

According to our meta‑analysis,EOF group showed shorter time to passage of flatus compared with DOF group,and there had no significant difference of postoperative ileus between EOF groups and DOF groups.It’s the key determinant of duration post‑gastrectomy that the ability eat sufficient amount of meal without adverse gastrointestinal symptoms.Early oral nutrition may enhance early adaptation to postoperative oral intake after gastrectomy.Intestinal function usually begins recovering 6-12 hours after surgery,therefore,enteral nutrition initiated after 24h of the surgery does not cause abdominal distention,instead,food stimulation can speed up gastrointestinal recovery while providing better nutritional support[41].Animal studies also indicated that EOF could accelerate the recovery of peristalsis and strengthen the immune response [39].In addition,the fear and anxiety caused by postoperative fasting and gastrointestinal decompression could be alleviated without the placement of nasogastric tube and EOF,and the psychological and gastrointestinal stress response of patients could also be reduced,which could make the patients feel more comfortable and conducive to accelerating the recovery of gastrointestinal function [16].A study has indicated that excessive intravenous fluid after surgery may result in delayed recovery of gastrointestinal function,cardiopulmonary dysfunction,would healing,and impaired tissue oxygenation [42].Restricted intravenous fluid administration is another advantage of early oral nutrition,and surgical outcomes in the EOF group,could possibly be related to this [18].

Compared with DOF,patients in EOF groups recovered gastrointestinal function more quickly,and they need intravenous fluids,which can really reduce their postoperative LOS,reduce hospitalization costs and avoid the waste of medical resources.

Limitation of study

The present analysis still exists several limitations.Firstly,the meta‑analysis lack of sufficient studies investigating EOF.Due to this,retrospective trials,which usually spanned numerous years and lacked precise data and may exist inevitably increasing bias,were also included in this analysis.Then major confounding factors that may have influenced the study results included the type of gastric operations,minor variations of the definition of EOF.Secondly,there are differences in the definition of some important indicators,such as the timing of oral nutrition initiation,consistency of oral intake and quantity of oral feeding.These differences may affect the rigor of the results.For future directions,researchers can conduct more rigorous RCT studies to demonstrate the advantages of EOF.

Conclusion

In general,based on the available evidence,there are no significant differences in anastomotic leaks and postoperative ileus between EOF group and DOF group,DOF protocol appears to have no protective effect after adult gastrectomy compared to allowing oral feeding early after surgery.Also,allowing early postoperative feeding is associated with faster recovery of gastrointestinal function and shorter LOS.