Study population

A retrospective review of the institutional GIST database was conducted. The data of the patients with primary GIST located in the duodenum and proximal jejunum within 20 cm of Treitz’s ligament who underwent surgical resection between January 2008 and December 2023 at Peking University Cancer Hospital (PKUCH) were selected.

The inclusion criteria for patient selection were as follows: (1) between 18 and 80 years of age; (2) distal segmental duodenectomy; and (3) pathologically confirmed primary GIST located in the D3/D4 duodenum or jejunum within 20 cm of the Treitz’s ligament. The exclusion criteria were: (1) operations conducted in an emergency setting; (2) locally recurrent or metastatic GIST involving duodenum or proximal jejunum; (3) other duodenal resection methods beyond distal segmental duodenectomy, including distal gastrectomy with proximal duodenectomy, pancreaticoduodenectomy, wedge duodenal wall resection or endoscopic duodenal resection; (4) merely jejunal resection without duodenal resection for proximal jejunal GIST; and (5) history or presence of any severe, unstable, systemic disease.

Until the end of 2017, our standard practice for cases of GISTs involving the proximal jejunum or the D3/D4 duodenum that required distal segmental duodenectomies was to perform only an end-to-side duodenojejunostomy without a gastrojejunostomy. However, we frequently encountered DGE, leading to many patients having to receive long-term gastrointestinal decompression therapy and being unable to consume food, potentially delaying subsequent imatinib therapy. In early 2018, we began exploring a modified approach. This new strategy included the addition of a gastrojejunostomy and a Braun’s jejunojejunostomy. For this study, patients were divided into two groups: the non-bypass group (patients who had undergone surgery by the end of 2017 without a gastrojejunostomy and Braun’s jejunojejunostomy) and the bypass group (those who were treated thereafter with a gastrojejunostomy and a Braun’s jejunojejunostomy).

Surgical approach

The same surgical team carried out the following surgical procedure with the same main steps:

After a thorough exploration of the abdominal cavity, the Cattell–Braasch maneuver was employed to mobilize the entire small intestine and mesocolon to facilitate exposure of the distal duodenum.

The distal duodenum and proximal jejunum were mobilized and sectioned distal to the tumor with the mesentery. The tumor, along with the duodenum and jejunum, was then dissected and repositioned posterior to the superior mesenteric vessels on the right side.

In instances where pancreas involvement was detected, the decision would be made to proceed with pancreaticoduodenectomy, partial pancreatic resection, or distal pancreatectomy based on the extent of the involvement. For large tumors near the junction of D2 and D3, catheterization through the cystic duct after a cholecystectomy was used to identify and protect the Vater’s ampulla and to delineate a safe resection margin.

The duodenum was sectioned proximal to the tumor and distal to the Vater’s ampulla. Jejunal ascension was performed retrocolically, followed by an end-to-side anastomosis with the duodenal stump (Fig. 1a). For patients with the gastrojejunostomy procedure, an antecolic side-to-side gastrojejunostomy was stapled 40 cm distal to the duodenojejunal anastomosis along the greater curvature. A Braun’s antimesenteric jejunojejunostomy was then performed 15 cm proximal to the gastrojejunostomy for the afferent limb and 40 cm distal for the efferent limb (Fig. 1b).

Fig. 1

(a). Patients in the non-bypass group underwent end-to-side duodenojejunostomy. (b). Patients in the bypass group also underwent an antecolic side-to-side gastrojejunostomy and a Braun antimesenteric jejunojejunostomy

Postoperative care

The patients had routine nasogastric tubes (NGT) inserted during the operation. According to the institutional protocol, it is standard practice for patients to receive prophylactic nasogastric decompression after segmental duodenectomy and undergo routine upper gastrointestinal series (UGIS) between postoperative days (POD) 5 and 7 for the early detection of anastomotic leakage. Once the UGIS did not reveal any anastomotic fistula and the contrast agent transmitted rapidly into the distal jejunum, the NGT was removed immediately, and the patient was advised to resume eating.

Complication definition

Postoperative clinical data concerning any complications and deviation from the normal postoperative course was also collected. DGE and other incidents such as postoperative pancreatic fistula (POPF) were graded following the standards published by the International Study Group of Pancreatic Surgery (ISGPS) [16, 17].

It should be noted that the true incidence of grade A DGE could not be determined in the study as all the patients were allowed to remove the nasogastric tube after the UGIS was done 5–7 days postoperatively due to our institutional protocol. Considering that grade A DGE generally only causes minor disturbances during the resumption of solid food intake, it typically has minimal clinical impact and may cause only a slight deviation from the clinical pathway [16, 18]. We thus mainly compared the incidence grades B-C (moderate and severe) DGE between the non-bypass group and the bypass group. The other postoperative complications were graded by the Clavien–Dindo classification and considered “major” if they were Grade III or higher [19].

Data handling and statistical analysis

We collected and analyzed demographic information for variables including age, gender, and body mass index (BMI). Clinicopathologic variables such as symptoms, diabetes, smoking, preoperative hemoglobin (HGB), tumor size, tumor location, liver metastases, and American Society of Anesthesiologists (ASA) classifications were collected as well, and treatment variables such as neoadjuvant imatinib therapy, operation time, estimated blood loss, multi-visceral resection (MVR), intensive care unit (ICU) stays, NGT removal interval, NGT reinsertion, highest body temperature, and postoperative stays count were also incorporated into our study. It is important to note that according to the criteria of the Chinese Working Group on Obesity and thus our study, BMI is classified into the following three groups: low and normal weight (18.5–24), overweight (24 to < 28), and obesity [20].

Statistical analysis was performed with SPSS Statistics (IBM Corp. Released 2019. IBM SPSS Statistics for Windows, Version 26.0. Armonk, NY: IBM Corp.). Standard descriptive statistics were calculated for categoric data (i.e., frequency and percentage) and continuous data (i.e., median and range), as listed in Table 1. Independent sample t-tests, Chi-square tests, and nonparametric Mann-Whitney U tests were used to compare variables between the non-bypass group and the bypass group for significant differences.

Table 1 Clinicopathologic characteristics of patients

Each clinicopathological variable associated with DGE was analyzed using binary logistic regression models, with results presented as odds ratios (ORs) and 95% confidence intervals (95% CIs). Initially, each covariate was evaluated in a univariate model and retained if the p-value was less than 0.1. Subsequently, a stepwise selection method was employed to identify significant covariates. For our usage, a p-value of less than 0.05 was considered statistically significant for all analyses.

Continuous variables (age, hemoglobin level, tumor size, operation time, estimated blood loss, highest body temperature) and categorical variables (gender, symptoms, BMI, smoking status, tumor location, liver metastases, ASA classification, neoadjuvant imatinib therapy, multi-visceral resection, ICU admission, GIST risk classification, complication grade) were analyzed for associations with grade B-C delayed gastric emptying (DGE) using univariate logistic regression. Results are reported as ORs with 95% CIs. Variables with a p < 0.1 in univariate analysis were included in a multivariable logistic regression model to identify independent predictors of grade B-C DGE.

Independent sample t-tests assessed correlations between operative time and various factors, including age, gender, symptoms, BMI, smoking status, HGB levels, tumor size, risk classification, location, liver metastases, neoadjuvant imatinib therapy, ASA classification, MVR, and bypass. Variables with a p < 0.05 were included in multiple linear regression analysis. Tumor size and hemoglobin levels were categorized by mean values.