Abstract
Background/Aim: Gastric outlet obstruction (GOO) is common in patients with advanced pancreatic cancer. Previous reports have documented the efficacy of gastrojejunostomy (GJ) bypass and endoscopic procedures. However, the optimal timing for GJ bypass remains unclear. This study aimed to determine the optimal timing for GJ bypass in pancreatic cancer.
Patients and Methods: This study included patients who underwent GJ bypass for GOO with pancreatic cancer at our institute between 2009 and 2023. Asymptomatic GOO was defined as gastroduodenal transit disturbance on imaging studies but without severe symptoms. Patients were divided into early and late groups based on the day of oral intake resumption.
Results: The study included 65 patients. The median time to resume postoperative oral intake was four days. According to the median value, 38 and 27 patients were assigned to the early and late recovery groups, respectively. Peritoneal dissemination was more common in the late-stage group (p=0.05). Asymptomatic GOO, laparoscopic surgery, and linear stapler use were significantly more common in the early recovery group. Multivariate analysis showed that symptomatic GOO was an independent variable for late recovery (odds ratio=12.7; p=0.0006; 95% confidence interval=2.7-95.1).
Conclusion: Early surgical intervention with GJ is recommended when asymptomatic GOO is suggested by imaging. This approach may facilitate earlier postoperative recovery and support continuity in multidisciplinary treatment strategies for pancreatic cancer.
- Gastrojejunostomy
- palliative surgery
- gastric outlet obstruction
- asymptomatic gastric outlet obstruction
- pancreatic cancer
Introduction
Malignancies of the pancreas, biliary tract, and gastroduodenum are often diagnosed in the advanced stages, or may grow without responding to chemotherapy or chemoradiation. Malignant gastric outlet obstruction (GOO) is a symptom caused by various tumors, including those of the pancreas, biliary tract, and gastroduodenum, and is defined as mechanical obstruction of the pylorus or duodenum secondary to compression or infiltration from these malignancies. Consequently, oral intake is often difficult or impossible to achieve (1). Patients with malignant GOO present with persistent nausea, vomiting, anorexia, abdominal pain, weight loss, malnutrition, and dehydration. GOO worsens the patient’s general condition, forcing the patient to discontinue systemic chemotherapy and preventing the improvement of the general condition.
The two main treatments for GOO are surgery and endoscopy. Previous reports have shown that the surgical approach has the advantage of longer relief of symptoms and less need for reintervention. In contrast, the endoscopic approach has shown that enteral stents using self-expanding metal stents rapidly improve enteral dysfunction with minimal invasion. However, enteral stents have also shown a high dysfunction rate owing to tumor growth and a higher rate of reintervention.
Recently, endoscopic ultrasound-guided gastroenterostomy (EUS-GE) has been reported as a novel technique for managing malignant GOO (2, 3). While previous studies have discussed the advantages and disadvantages of various treatment approaches for GOO (4), the optimal timing of intervention remains unclear. In this study, we retrospectively evaluated gastrojejunostomy (GJ) bypass for GOO due to pancreatic cancer (PC) and aimed to determine the optimal timing for GJ bypass.
Patients and Methods
Patients and study design. We searched a database maintained postoperatively at our institute between April 2009 and December 2023. Patients who underwent surgical GJ bypass for GOO due to PC were included in this study. This study was approved by the Institutional Review Board (R2022148).
Patients with PC and persistent nausea or vomiting due to gastric or duodenal strictures were defined as having symptomatic GOO. In contrast, patients with not persistent nausea or vomiting but with stenosis of the stomach or duodenum invasion or compression by the PC revealed by computed tomography (CT) or other imaging studies were diagnosed with asymptomatic GOO. Oral intake was graded using the GOO Scoring System (GOOSS) (0=no intake, 1=liquid only, 2=soft solids, 3=almost full diet, and 4=full diet) (5). Symptomatic GOO was defined as a GOOSS score of 0, or persistent nausea, vomiting, or anorexia not related to chemotherapy or other treatment. Clinical success was defined as the ability to tolerate at least a liquid diet after GJ bypass without symptoms due to GOO (GOOSS: ≥2) (6). To evaluate the optimal timing of GJ bypass, the patients were divided into two groups. Patients who were able to resume oral intake earlier after GJ bypass and who achieved clinical success were included in the early group. The remaining patients were assigned to the late recovery group. The threshold for classifying the patients into two groups was the median number of days that the patients were able to resume oral intake after GJ bypass. The exclusion criteria were as follows: (i) severe peritoneal dissemination or cancerous peritonitis, (ii) history of upper abdominal surgery, (iii) recurrence of PC after curative resection, and (iv) missing medical records.
Surgical procedures. Laparoscopic surgery is usually performed for GJ bypass. Patients who underwent upper gastrointestinal tract surgery were selected for open GJ. In cases of severe intra-abdominal adhesions or PC invasion, continuing laparoscopic surgery was difficult, and the patient underwent open surgery.
Laparoscopic GJ bypass. Patients were placed in a supine position with their legs apart. Five trocars were used in the study. The first 12 mm trocar for laparoscopy was inserted through a small umbilical incision. Carbon dioxide pneumoperitoneum was established, and the remaining four trocars were placed in the upper abdomen. After all the trocars were in place, the abdominal cavity was carefully observed to detect peritoneal dissemination or liver metastasis. Peritoneal lavage cytology was performed.
After dissection of the greater omentum, the posterior wall of the stomach was exposed, and no direct invasion of the PC was confirmed. A small hole was created on the mesenteric side of the jejunum. A small hole was also made in the posterior wall of the gastric antrum, close to the greater curvature. The limb of the 60 mm linear stapler was inserted into the jejunum. The jejunum was elevated via the antecolic route, the other limb was inserted into the stomach, and GJ was performed. GJ was routinely performed on the posterior wall of the greater curvature of the stomach. The common hole was closed using absorbable sutures. Jejunojejunal side-to-side anastomosis was performed with a 45 mm linear stapler using a procedure similar to that used for GJ. In patients with confirmed PC invasion of the posterior stomach wall, GJ was performed on the anterior wall of the stomach. In patients who underwent hand suturing, the procedure of anastomosis and selection of the anastomotic position were the same as those using linear staplers.
Data collection. We evaluated the clinical features, intraoperative and postoperative findings, and treatment after GJ. Blood examinations were performed, including preoperative nutritional status, total protein, albumin, cholinesterase, total cholesterol (T-Cho), and lymphocyte count. The Controlling Nutritional Status (CONUT) score, which is calculated using serum albumin, the number of lymphocytes, T-Cho, and the Prognostic Nutritional Index (PNI), which is calculated as follows: 10 × serum albumin (g/dl) + 0.005 × lymphocyte count (per mm3), was used to evaluate the preoperative nutritional status (7, 8). PC resectability was diagnosed based on the National Comprehensive Cancer Network Guidelines Version 2.2021 (9). Postoperative complications were defined as grades III-V according to the Clavien–Dindo classification (10).
Statistical analysis. Differences in the nominal dates between the two groups were examined using the chi-square test. Differences in quantitative variables were evaluated using the Mann–Whitney U-test. Logistic regression analysis was used for multivariate analysis. Statistical significance was set at p<0.05. All statistical analyses were performed using JMP software (version 15.0; SAS Institute, Cary, NC, USA).
Results
Sixty-nine patients underwent GJ for GOO and PC. Four patients were excluded for the following reasons: one had multiple small bowel obstructions due to peritoneal dissemination and three had recurrence after pancreatoduodenectomy (Figure 1). The remaining 65 patients were included in this study. Patient backgrounds are shown in Table I. The most common tumor location was the pancreatic head (69.2%). Fifty-five patients were diagnosed with unresectable PC, however, two patients were diagnosed with resectable PC and underwent radical resection after GJ bypass and subsequent neoadjuvant chemotherapy. Of 42 patients diagnosed with unresectable metastatic (UR-M) pancreatic cancer, 25 had confirmed peritoneal dissemination. Among all patients, 36 (55.4%) were diagnosed as positive by peritoneal lavage cytology. Sixty patients (92.3%) had duodenal outlet obstruction. Twenty patients (33.3%) had stenosis of the first or second portion of the duodenum, and 40 patients (66.7%) had stenosis of the third or fourth portion of the duodenum. Forty-two patients (64.6%) were diagnosed with symptomatic GOO and treated conservatively for 17 (1-102) days before undergoing GJ bypass. The operative findings are summarized in Table II. Thirty-seven patients underwent laparoscopic GJ bypass, and five underwent choledochojejunostomy at the same time. Forty-two patients (64.6%) underwent GJ with a linear stapler. The patients resumed oral intake on postoperative day (POD) 4 (2-13). The clinical success rate of the resumption of oral intake was 85.7%. Twenty-two patients were selected for best supportive care after GJ bypass. The remaining 41 patients received chemotherapy on POD 19 (6-69). Chemotherapy was resumed on POD 20 (4-113) (Table III).
Patient flowchart. Three patients were diagnosed with gastric outlet obstruction (GOO) after curative resection of pancreatic cancer, and one patient was diagnosed with multiple bowel obstructions due to peritoneal dissemination.
Characteristics of patients.
Operative findings.
Postoperative findings.
To evaluate the optimal timing of GJ bypass for GOO with PC, patients were classified into two groups based on the median POD value at which they started oral intake. Patients who started oral intake within four days after surgery were classified into the early recovery group, and those who started oral intake after POD5 were classified into the late recovery group. The early recovery group included 37 patients, whereas the late recovery group included 26 patients. There were no significant differences in patient backgrounds (Table IV). Peritoneal dissemination was more common in the late recovery group than in the early recovery group (p=0.05). Twenty-four patients in the late recovery group were diagnosed with symptomatic GOO (88.9%), 18 in the early recovery group were diagnosed with symptomatic GOO (48.6%), and significantly more patients were diagnosed in the late recovery group (p=0.003). Regarding intraoperative features, the early recovery group had significantly more laparoscopic surgeries, less blood loss, and more anastomoses using a linear stapler (Table V). However, there were no significant differences in the diameter of the anastomosis or anastomotic location in the stomach between the two groups (Table VI). Twelve patients (32.4%) in the early recovery group and 10 (38.4%) in the late recovery group chose the best supportive care after GJ bypass (p=0.62). Two of 25 patients in the early recovery group who received chemotherapy after GJ bypass underwent curative pancreatectomy.
Comparison of patients’ background.
Comparison of operative and postoperative findings.
Multivariable analyses of the factors that late postoperative oral intake.
In the early recovery group, patients experienced the first postoperative flatus significantly earlier, resumed water intake and oral feeding earlier, and had a significantly shorter postoperative hospital stay. The clinical success rate of oral intake was significantly higher in the early recovery group (35 patients, 94.5%) than in the late recovery group (seven patients, 26.9%). However, there were no significant differences in the duration of chemotherapy interruption (days) or resumption of chemotherapy after GJ bypass (POD) between the two groups (Table V).
Variables that were determined preoperatively to have a p-value less than 0.05 were included in a subsequent multivariate analysis to identify the risk factors for late postoperative recovery. In multivariate analysis, only symptomatic GOO was selected as an independent factor for late recovery [odds ratio=12.7; p=0.0006; 95% confidence interval (CI)=2.7-95.1] (Table IV).
Discussion
GOO is a common clinical condition resulting from locally advanced malignant tumors adjacent to the stomach or duodenum, such as gastric, duodenal, or PCs. GOO often develops in highly advanced PC, and its treatment is palliative. Palliative therapies for GOO are divided into two main categories: endoscopic and surgical. Self-expanding metal stents (SEMS) have been selected for endoscopic procedures since the late 1990s (11, 12). Recently, EUS-GE with the placement of a lumen-apposing metal stent has become an increasingly adopted approach (2, 6, 13-17). GJ with laparotomy or a laparoscopic approach is often used in surgical procedures. Previous studies, including systematic reviews, have often focused on procedures that are more effective for GOO (1, 4, 6, 18-21). Compared with surgical GJ bypass, SEMS tended to improve symptoms of GOO faster and have shorter life due to tumor growth. GJ bypass was performed in patients with a life expectancy of ≥2 months because of a longer term patency could be expected (17). However, EUS-GE was reported to have equivalent clinical success and reintervention rates but significantly fewer adverse events than surgical GJ bypass.
However, to the best of our knowledge, there are no reports on the optimal timing of therapeutic interventions for GOO. In our study, the clinical success rate of GJ bypass for GOO in patients with PC was 85.7%. In the early recovery group, clinical success rate was 94.5%. Patients who were able to resume postoperative oral intake also had a higher clinical success rate for GJ bypass. Therefore, evaluating the factors that allow earlier resumption of postoperative oral intake is important to validate the optimal timing of GJ bypass for GOO with PC. The patients were classified into early and late recovery groups. There was a trend toward more cases of peritoneal dissemination in the late recovery group; however, there was no significant difference between the two groups. This suggests that peritoneal dissemination is not a reason for avoiding GJ bypass for GOO in PC. In cases of peritoneal dissemination, the indications for GJ bypass should be comprehensively judged based on peritoneal dissemination, PC progression, and general condition. The GJ bypass procedure showed no significant difference in operative time between laparoscopic and open surgery, with significantly less blood loss during laparoscopic surgery (data not shown). Laparoscopic surgery is the first choice for GJ bypass.
Patients with a PNI ≤40 have contraindications for intestinal anastomosis (8). No postoperative leakage was observed in any of the patients. In GOO patients with PC, owing to the continuation of chemotherapy, pancreatic exocrine insufficiency, decreased oral intake, and other factors, their nutritional status was considered inadequate. Our study suggests that surgery should be performed as early as possible if the patient’s general condition permits without adherence to preoperative nutritional management, using PNI and CONUT as indicators.
In this study, univariate analysis showed significant differences in symptomatic GOO, laparoscopic surgery, and GJ bypass with a linear stapler. However, there were no significant differences in peritoneal lavage cytology or peritoneal dissemination. These facts suggest that the presence of peritoneal lavage cytology or peritoneal dissemination alone is not a reason to avoid GJ bypass in GOO with PC. In the multivariate analysis, only preoperative symptomatic GOO was identified as a factor that delayed resumption of oral intake after GJ bypass. Preoperative CT revealed duodenal edema and a dilated stomach or duodenum in almost all patients; however, no findings that could predict the development of symptomatic GOO were found in this study.
Of the patients who underwent GJ bypass, chemotherapy was administered to 63%, and conversion surgery was performed in two patients in the early recovery group. Thus, GJ bypass for GOO in PC was not only a palliative treatment in the terminal state, but an essential component of multidisciplinary treatment. Multidisciplinary treatment is important for improving the prognosis of PC (22-27). Therefore, continuous treatment is necessary. In the present study, there were no significant differences in the duration of chemotherapy interruption or time to chemotherapy resumption after GJ bypass between the early and late recovery groups. This may be because the patients’ backgrounds were inconsistent.
Study limitations. First, it was a retrospective study with a small sample size. Second, the lack of comparison with endoscopic procedures limits evaluation of the optimal procedure for GOO in PC. Third, patients with severe cancerous peritonitis with massive ascites effusion or severe peritoneal dissemination were excluded. This may have resulted in selection bias.
Conclusion
Asymptomatic GOO is an independent factor that delays the resumption of oral intake after GJ bypass. Early surgical intervention is recommended if GOO is suggested on CT or other imaging studies. Moreover, GJ bypass is not only a palliative treatment for PC but also an aggressive option as part of the multidisciplinary treatment for advanced PC.
Footnotes
Authors’ Contributions
K.H. and I.H. contributed to data collection. K.H. analyzed the data and wrote the manuscript, the figure and the tables. T.F. supervised all activities. All Authors have read and approved the published version of the manuscript.
Conflicts of Interest
The Authors confirm there are no conflicts of interest in relation to this study.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received April 8, 2025.
- Revision received April 16, 2025.
- Accepted April 17, 2025.
- Copyright © 2025 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).