Jejunal interposition helps prevent reflux gastritis

BACKGROUND/AIMS: Jejunal interposition after distal gastrectomy is reported to prevent both duodenogastric reflux and rapid gastric emptying. However, comparing primary reconstruction with this procedure and Billroth-I in terms of clinical evaluation by the same surgeon is rare. In this study, the benefit of this procedure was retrospectively evaluated as compared to the Billroth-I method. METHODOLOGY: Of 30 patients with early gastric cancer located at the middle third of the stomach, 15 underwent distal gastrectomy with jejunal interposition and the other 15 underwent Billroth-I gastrectomy by the same surgeon. Isoperistaltic jejunal interposition measuring 10-12 cm was used. All the anastomoses without jejunojejunostomy were performed using auto-suture staplers. Assessment of postoperative symptoms and functions was performed one year after surgery. RESULTS: The mean operation time was significantly longer after jejunal interposition (p < 0.01). No serious complications occurred in either group, and the hospital stay after operation was also similar. There were no significant differences in terms of postoperative symptoms, food intake, and recovery of body weight. The incidence of bile regurgitation and reflux gastritis was very low or zero in the jejunal interposition group, which indicated differences (p < 0.05, p < 0.01, respectively). Reflux esophagitis was not found in jejunal interposition, but two patients after Billroth I showed grade B esophagitis. As regards gastric emptying, the retention capacity was very poor and there was no significant difference between the two groups. CONCLUSIONS: Jejunal interposition after distal gastrectomy was superior to the Billroth-I procedure in terms of reflux gastritis prevention. However, dumping syndrome and rapid gastric emptying were not prevented.     

Postoperative evaluation of the jejunal pouch reconstruction following proximal and distal gastrectomy for cancer

BACKGROUND/AIMS: To evaluate the efficacy of the jejunal pouch reconstruction following subtotal proximal and distal gastrectomy, a retrospective study examining the postoperative condition of patients who underwent different methods of reconstruction after gastrectomy for cancer was undertaken. METHODOLOGY: Various parameters indicative of postoperative function were evaluated at one year postoperatively, and two major groups were examined. The "proximal gastrectomy" group was composed of patients who underwent either 1) proximal gastrectomy with an interposed jejunal pouch (PG-pouch), 2) proximal gastrectomy with simple jejunal interposition (PG-inter), or 3) total gastrectomy with simple jejunal interposition (TG). The "distal gastrectomy" group was composed of patients who underwent either 4) distal gastrectomy with an interposed jejunal pouch (DG-pouch), 5) distal gastrectomy with simple jejunal interposition (DG-inter), or 6) distal gastrectomy with Billroth 1 reconstruction (B-1). RESULTS: Volume of meal intake was better preserved and the incidence of abdominal symptoms were less frequent in the PG-pouch and DG-pouch groups. In the PG-inter, DG-inter and DG-pouch groups, none of the patients experienced heartburn or had endoscopic findings consistent with reflux esophagitis, while 2 patients (20.0%) in the PG-pouch group complained of heartburn with evidence of reflux esophagitis on endoscopy. Increase in blood acetaminophen level was milder in both the PG-pouch and DG-pouch groups, signifying improved gastric emptying. CONCLUSIONS: The jejunal pouch interposition following proximal and distal gastrectomy seems to confer clinical benefit in terms of postoperative function, especially in the form of meal intake, abdominal symptoms, and gastric emptying. The side effect of an improved reservoir may be the incidence of reflux esophagitis seen in 2 patients in the PG-pouch group.     

A novel operative technique on proximal gastrectomy reconstructed by interposition of a jejunal J pouch with preservation of the vagal nerve and lower esophageal sphincter.

BACKGROUND/AIMS: This article describes the surgical techniques and postoperative status for proximal gastrectomy reconstructed by interposition of a jejunal J pouch with preservation of the vagal nerve and lower esophageal sphincter. METHODOLOGY: We have performed a new technique for reducing postgastrectomy sequelae such as reflux esophagitis, early dumping syndrome, and microgastria in early gastric cancer located in the proximal third of the stomach. The technique consists of proximal gastrectomy with preservation of the hepatic, pyloric, celiac branch of the vagal nerve, and abdominal esophagus (lower esophageal sphincter), and reconstruction by interposition of a jejunal J pouch. To reserve pyloric function, pyloroplasty can be omitted by preservation of the pyloric branch from the vagal nerve. To restore loss of reservoir function, the reconstruction is performed with an interposed jejunal J pouch. Sacrifice of the mesenteric arcades is kept to a minimum to preserve the autonomic nerve and blood flow in the mesentery. RESULTS: All of the patients who underwent this operation were able to eat an adequate amount of food at 6 months after surgery and they were satisfied with their postoperative status. And that, we have not experienced postgastrectomy disorders such the dumping syndrome and reflux esophagitis. CONCLUSIONS: Therefore, this method is useful for preventing the postoperative disorders in patients with early gastric cancer located in the proximal third of the stomach.     

The effect of preserving antral innervation and of a pyloroplasty on gastric emptying after vagotomy in man

A prospective randomized trial of 40 duodenal ulcer patients is reviewed. The patients had one of four operations (selective vagotomy, proximal gastric vagotomy, selective vagotomy plus pyloroplasty, or proximal gastric vagotomy plus pyloroplasty). The gastric emptying of a hypertonic fluid meal was assessed before and three to four months after operation. Selective vagotomy without a drainage procedure results in gastric retention and should no longer be considered as a method of treatment for duodenal ulcer. Proximal gastric vagotomy without a drainage procedure does not lead to gastric retention. Initial gastric emptying is more rapid after proximal gastric vagotomy but the final emptying time is the same as before operation and this operation alters the pattern of gastric emptying less than the other operations. Pyloroplasty added to either selective or proximal gastric vagotomy results in loss of the normal regulation of gastric emptying, very rapid initial gastric emptying, and a significant increase in the incidence of ;dumping'. It appears from these studies that ;dumping' is due to rapid gastric emptying and mainly due to the drainage procedure.     

Esophagogastrectomy without pyloroplasty

There is no consensus on the need for pyloroplasty after esophagectomy or proximal gastrectomy with an esophagogastrostomy and vagotomy. Arguments for routine pyloroplasty include prevention of postoperative delayed gastric emptying. Arguments against include prevention of postoperative dumping syndrome and bile reflux esophagitis. The purpose of this study was to assess clinical outcomes of patients undergoing esophagogastrectomy without routine pyloroplasty. All patients undergoing esophagogastrectomy or proximal gastrectomy with esophagogastrostomy from October 1996 to September 2002, inclusive were reviewed for age, gender, diagnosis, type of resection, pathology, short-term complications, long-term complications, remedial procedures performed, and postoperative gastric emptying scintigraphy. 58 patients were studied. Postoperative mortality was 6.9%, and anastomotic leak rate 12.1%. Eleven patients (19%) had symptomatic gastroparesis, two required pyloric balloon dilation and one a pyloroplasty. No patients complained of dumping symptoms; reflux requiring medical intervention occurred in seven (12.1%), and anastomotic stricture requiring dilation occurred in five (8.6%). Omitting a pyloroplasty does not lead to a high frequency of symptomatic delayed gastric emptying. Maintaining the pylorus may protect patients from dumping syndrome, and bile reflux esophagitis with its potential noxious effects on the remaining esophageal mucosa.     

Functional outcomes according to the size of the gastric remnant and type of reconstruction following open and laparoscopic proximal gastrectomy for gastric cancer

Background/Aims: We compared functional outcomes between different types of reconstruction following open or laparoscopic 1/2- or 2/3-proximal gastrectomy for gastric cancer. Methodology: Resection and reconstruction were performed by one of the following 6 methods, depending on the depth of cancer invasion and the date of the procedure relative to introduction of laparoscopic proximal gastrectomy: open proximal 2/3-gastrectomy with jejunal interposition (2/3PG-int, n=7), open proximal 1/2-gastrectomy with jejunal interposition (1/2PG-int, n=5), laparoscopic proximal 1/2-gastrectomy followed by double tract reconstructions with small (3cm) jejunogastrostomy (L1/2 PG-DT(S), n=19) and laparoscopic proximal 1/2-gastrectomy followed by double tract reconstructions with large (6cm) jejunogastrostomy (L1/2PG-DT(L), n=10). Open total gastrectomy with jejunal interposition (TG, n=12) and laparoscopic total gastrectomy with Roux-en-Y reconstruction (LTG, n=14) represented control procedures. Results: Comparison of postoperative/preoperative body weight ratios and food intake ratios revealed better preservation among patients with a larger remnant stomach and with easy flow of food into the remnant stomach (the 1/2PG-int and L1/2PG-DT(L) groups). Conclusions: Better functional outcomes were observed in patients with a large remnant stomach and with easy flow of food into the remnant stomach regardless of whether they underwent open or laparoscopic procedures.     

Jejunal pouch interposition with an antiperistaltic conduit as a pyloric ring substitute after standard distal gastrectomy: a comparison with the use of an isoperistaltic conduit

BACKGROUND/AIMS: We performed jejunal pouch interposition with a short antiperistaltic conduit as a pylorus substitute after gastrectomy for gastric cancer and compared the outcome with an isoperistaltic conduit. METHODOLOGY: After a standard distal gastrectomy and lymph node dissection, a 15-cm-long pouch was formed using 3 linear staples (Endo-GIA) and interposed between the residual stomach and duodenum. The distal jejunal limb was made into a 3-cm-long isoperistaltic conduit in the isoperistaltic group (n = 17), and the proximal jejunal limb was made into a 3-cm-long antiperistaltic conduit in the antiperistaltic group (n = 8). Postoperatively, the patients were interviewed periodically to document any complaints. A dual-phase, dual-isotope radionuclide gastro-pouch-emptying study was performed 1 and 6 months after surgery. RESULTS: None of the patients developed postoperative complications and showed discomforts of dumping, stasis or reflux esophagitis. The dietary volume and body weight of patients gradually increased in both groups after 6 months. The combined radioisotope retention rate for the pouch and residual stomach was 31% for liquid food and 35% for solid food in the isoperistaltic group after 120 min, and 41% and 57%, respectively, in the antiperistaltic group. The pattern and emptying rate for solid food in the antiperistaltic group were more similar to those in healthy individuals than in the isoperistaltic group. CONCLUSIONS: The gastro-pouch-emptying test in the antiperistaltic group demonstrated acceptable emptying for a pyloric ring substitute. A reasonably good quality of life has been obtained for patients having an antiperistaltic jejunal conduit.     

Results of follow-up endoscopy in patients who underwent proximal gastrectomy with jejunal interposition for gastric cancer

BACKGROUND/AIMS: The present study evaluates the findings of long-term follow-up endoscopy in patients who underwent proximal gastrectomy with jejunal interposition for gastric cancer. METHODOLOGY: A total of 55 patients (45 males, 10 females; 32 to 79 years; mean, 55.9 years), who underwent proximal gastrectomy with jejunal interposition, were enrolled in the present study. We reviewed the findings of follow-up endoscopy of all patients with particular reference to the development of esophagitis, jejunitis, jejunal ulcer and secondary tumors. RESULTS: We found reflux esophagitis in 6 patients (10.9%) between 12 and 35 months with an average of 22 months after surgery. Jejunitis was discovered in 5 patients (9.0%) between 6 and 96 months with an average of 29 months after surgery. Jejunal ulcer was revealed in 6 patients (10.9%) between 6 and 75 months with an average of 37 months after surgery. Tumors of the remnant stomach, early gastric cancer and gastric adenoma, were identified in 2 patients (3.6%) at 24 months and 69 months, respectively. CONCLUSIONS: Jejunal interposition combined with proximal gastrectomy does not always prevent complications related to regurgitation of gastric content, and may not be a suitable treatment in view of postoperative endoscopic surveillance. Further studies are required to identify an appropriate surgical approach to proximal gastrectomy for gastric cancer.     

Postoperative evaluation of pylorus-preserving gastrectomy for early gastric cancer

BACKGROUND/AIMS: There have been many reports proposing some advantages of pylorus-preserving gastrectomy for gastric ulcer compared to the conventional distal gastrectomy. However, it is not clear whether similar results will be obtained from the patients with early gastric cancer. METHODOLOGY: Of 50 patients with early gastric cancer, 25 underwent pylorus-preserving gastrectomy under strict criteria and the other 25 underwent distal gastrectomy with Billroth I anastomosis by the same surgeon. The subjects were then interviewed and examined periodically to assess symptoms, food intake, body weight and serum nutritional parameters. Endoscopy and a radioisotope gastric emptying test was performed 1 year after the operation. RESULTS: Many of the patients with pylorus-preserving gastrectomy complained of gastric fullness after meals, resulting in poor food intake; a significant between-group difference was found up to 1 year after the operation. A low incidence of reflux gastritis and slow gastric emptying were confirmed in the patients after pylorus-preserving gastrectomy. CONCLUSIONS: Pylorus-preserving gastrectomy has advantages over distal gastrectomy in terms of the avoidance of dumping syndrome and protection against duodeno-gastric reflux. However, more time was necessary for improved gastric fullness or food intake. Pylorus-preserving gastrectomy should be applied in younger patients with early gastric cancer expecting long survival.     

全胃切除后单腔空肠间置代胃术疗效分析

目的探讨全胃切除术后采用40cm长度的顺行性间置空肠代胃术重建消化道的临床效果.方法对39例行间置空肠代胃术患者随访一年,观察其代胃的储存量、排空时间、反流情况、进食量、体能恢复、临床症状、体重及血浆营养参数(血红蛋白、血浆总蛋白).结果术后二周时一次吞服钡剂量为120ml(100～150),代胃排空时间平均40分钟(20～90),未见钡剂反流入食道,无间置肠段梗阻,术后一年时进食米量2～3两/餐,进餐次数为3～4次/日,无倾倒综合症,均可以从事轻～中度的体力活动.入院时与术后一年的血浆营养参数与体重比较有明显差异.(P＜0.01)结论采用40cm长度的顺行性间置空肠代胃术操作简便,临床效果良好.     

Gastric emptying and sieving of solid food and pancreatic and biliary secretions after solid meals in patients with nonresective ulcer surgery

This study was undertaken to compare with previously published findings in normal subjects and subjects after truncal vagotomy and antrectomy the effects of nonresective ulcer surgery on (a) gastric emptying, grinding, and sieving of solid food and on (b) pancreatic and biliary secretions. Six subjects with proximal gastric vagotomy and 7 subjects with truncal vagotomy with pyloroplasty were studied using a previously validated indicator perfusion system with its aspiration port placed in the proximal jejunum. All subjects were given a meal of 30 g of 99mTc-liver, 60 g of beefsteak, and 100 ml of H2O. In conjunction with a gamma-camera to measure total gastric emptying of 99mTc-liver, this method allowed us to estimate the fraction of 99mTc-liver emptied from the stomach as particles of less than 1-mm diameter; in addition, we were able to measure jejunal concentrations and outputs of bile salts and pancreatic enzymes. In subjects with proximal gastric vagotomy, all parameters studied were indistinguishable from normal. Subjects with truncal vagotomy and pyloroplasty behaved similarly to subjects with vagotomy and antrectomy, showing (a) early precipitous emptying of food, (b) heterogeneous distribution of half-emptying times, (c) near normal concentration of biliary and pancreatic secretions, (d) markedly reduced jejunal flow rates, and (e) a reduction in postcibal trypsin secretion. In contrast to subjects after truncal vagotomy and antrectomy, however, the majority of subjects with vagotomy and pyloroplasty did not show a persistent defect in grinding and sieving of solid food.     

Gastric Emptying and Dumping after Proximal Gastric Vagotomy

One hundred dumping provocation tests were performed on patients who had had proximal gastric vagotomy without drainage in the pre-and postoperative period. Other than monitoring the symptoms of dumping on test, various other objective criteria, Le., plasma volume fall, gastric emptying, and blood sugar changes were also measured simultaneously. Whether dumping was assessed clinically or on test, it seemed to be equally prevalent after proximal gastric vagotomy without drainage compared to vagotomy with drainage or partial gastrectomy. Although the pattern of gastric emptying after proximal gastric vagotomy alone was not significantly different from that before operation in those patients who had dumping symptoms on test, the half-life of the meal was significantly shorter and the rates of initial emptying were much faster. The plasma volume fall was also significantly greater and blood sugar was rise steeper in these patients. The objective measurements have not only helped in determining the incidence of dumping after proximal gastric vagotomy but have also proved to be of value in categorizing patients with doubtful symptoms and have given some insight into the etiopathogenesis of this syndrome.     

Esophagogastrostomy reconstruction after limited proximal gastrectomy

BACKGROUND/AIMS: Recent advances in diagnostic techniques have led to the detection of an increasing number of early gastric cancers in the upper third of the stomach. The objective of this study was to determine the most appropriate surgical treatment for these cancers. METHODOLOGY: The clinicopathologic characteristics of 35 patients with early gastric cancer in the upper third of the stomach who underwent three different types of gastrectomies were reviewed retrospectively from hospital records between January 1992 and August 1999. RESULTS: Patients undergoing limited proximal gastrectomy with esophagogastrostomy reconstruction had shorter operation times and less blood loss than those for patients undergoing total gastrectomy or proximal gastrectomy with jejunal interposition. No lymph node metastasis was identified in any of these patients. Heartburn due to reflux esophagitis was seen in a few patients of each group, but they were successfully treated by antacids. The extreme reduction in food intake volume was more frequently experienced in patients with total gastrectomy than those with both proximal gastrectomies. When mortality due to other disease was excluded, all patients survived without recurrence. CONCLUSIONS: A limited proximal gastrectomy with esophagogastrostomy reconstruction decreased surgical risk and realized preservation of maximal function.     

A randomized clinical trial of pouch reconstruction after total gastrectomy for cancer: which is the better technique, Roux-en-Y or interposition?

BACKGROUND/AIMS: To determine the optimum pouch reconstruction after total gastrectomy, we conducted a randomized trial to compare the usefulness between PR and PI in terms of quality of life. METHODOLOGY: Thirty patients younger than 70 years of age were selected and randomly classified into the following 2 groups: pouch and Roux-en-Y (PR; n = 15) and pouch and interposition (PI; n = 15). In each subject, the postoperative symptoms, food intake in a single meal, body weight, serum nutritional parameters, endoscopy, emptying time of the gastric substitute, and gallstone formation were evaluated. RESULTS: There were no significant differences in terms of the postprandial symptoms, food intake, body weight, and serum nutritional parameters until 2 years postoperative. Endoscopy showed a tendency of food stasis in the PR group, although this was not significant. The gastric emptying test in the PR group indicated that the retention capacity was slightly superior to that in the PI group. CONCLUSIONS: The impact of the duodenal passage on symptoms and nutrition could not be ascertained.     

Evaluation of reconstruction after proximal gastrectomy: prospective comparative study of jejunal interposition and jejunal pouch interposition

BACKGROUND/AIMS: To prevent various distresses after proximal gastrectomy, reconstruction by interposed jejunal pouch has been advocated as an organ-preserving surgical strategy to ensure favorable quality of life for the patients. METHODOLOGY: Proximal gastrectomy was performed in 9 patients with gastric cancer in the upper third of the stomach. Four patients were randomly selected for reconstruction by jejunal pouch interposition (JPI group), while 5 had reconstruction by jejunal interposition (JI group). The patients who underwent JPI and JI were followed up to evaluate resumption of normal diet, change in body weight, and clinical symptoms. RESULTS: The JPI group showed a significant dietary advantage. Three months after surgery, JPI patients could eat more than 80% of the volume of their preoperative meals, whereas JI patients ate less than 50%. The percentage of postoperative body weight loss was higher in the JI group than in the JPI group because the volume of the remnant stomach was more adequate in the latter. Moreover, it was easier to enter the remnant stomach and duodenum for endoscopic fiberscopy in the JPI group for the treatment of hepato-biliary pancreatic disease. CONCLUSIONS: JPI is an effective method for preservation of gastric function after proximal gastrectomy.     

Serious complications after a proximal gastrectomy with a jejunal pouch interposition for gastric cancer

Proximal gastrectomy with jejunal interposition is a common surgical method in Japan, because the procedure has been shown to give a better post-operative quality of life. Some complications are associated with it. However, esophageal candidiasis and linear marginal ulcer along the gastrojejunal anastomosis after the surgical method has never previously been reported. We herein report a case of a patient who developed serious complications after proximal gastrectomy with jejunal interposition. A 68-year-old man underwent proximal gastrectomy with a jejunal pouch interposition for reconstruction for type 1 gastric cancer. Twenty-three months after the procedure, he complained of dysphagia and epigastric pain. Esophagogastroduodenoscopy showed esophageal candidiasis. The patient improved symptomatically following antifungal medication with fluconazole. Eleven months later, the patient developed severe pneumonia. In subsequent days, a melena episode occurred. Esophagogastroduodenoscopy revealed a linear marginal ulcer along three-fourths of the gastrojejunal anastomosis. The ulcer was drug resistant. The patient died of respiratory failure. Jejunal pouch interposition after a proximal gastrectomy can be associated with significant complications. Further studies are required to identify the best condition of the procedure.     

Marginal ulcer on the jejunum after proximal gastrectomy by jejunal interposition

Marginal ulcer after proximal gastrectomy has never been previously reported, despite that this procedure preserves the fundic gland area of the stomach, which secretes gastric acid. In this report, we describe a patient who developed a marginal ulcer on the oral side of the gastrojejunal anastomosis after proximal gastrectomy by jejunal interposition. This case serves as a reminder that gastric acid secretion of the remnant stomach must be carefully monitored after proximal gastrectomy in gastric cancer surgery.     

Gastrointestinal motility and gastric emptying after Billroth II gastrectomy in dogs

This study was undertaken to compare the effects of subtotal Billroth II gastrectomy on gastric emptying and gastrointestinal motility with previously published results in intact dogs and in dogs with subtotal Roux-Y gastrectomy. Extraluminal strain gauge transducers were used to study gastrointestinal motility after Billroth II gastrectomy in four conscious dogs. Gastric emptying was measured radiographically. In Billroth II dogs gastric emptying of low-viscosity meals was biphasic with an initial rapid emptying. The addition of nutrients to low-viscosity meals delayed gastric emptying accompanied with reduction in gastric and jejunal motility. Similar to that in Roux-Y dogs, gastric emptying of noncaloric medium-viscosity meals was delayed because of segmenting motor patterns of the jejunal loops, in contrast to the propulsive jejunal motor pattern in intact dogs. Nutrients added to medium-viscosity meals did not change the jejunal motor pattern; gastric emptying was delayed compared with intact dogs. Results show that meal viscosity and jejunal motor pattern influence gastric emptying after Billroth II gastrectomy.This study was supported by the Deutsche Forschungsgemeinschaft, grant Eh 64/3-2.     

Suture line recurrence in the jejunal pouch after curative proximal gastrectomy for gastric cancer: report of two cases

We report two patients with suture line recurrence in the jejunal pouch after curative proximal gastrectomy for gastric cancer. The first patient was a 60-year-old asymptomatic woman with gastric cancer (T2N0M0) after curative proximal gastrectomy with jejunal pouch interposition. She had to undergo a second resection for suture line recurrence in the jejunal pouch 12 months later. On examination of the resected specimen, histological examination revealed a moderately differentiated adenocarcinoma, which was similar to that of the primary tumor. The second patient was a 74-year-old man who was also diagnosed as having locoregional recurrences in the jejunal pouch after a curative proximal gastrectomy with an S-shaped pouch for gastric cancer (T2N0M0). Histological examination of the resected specimen revealed moderately differentiated adenocarcinoma, which had a similar histopathology to that of the primary tumor. During the first procedure, the jejunal pouch was formed using several disposable devices and the end-to-side esophagojejunostomy was performed with another circular stapler to avoid contamination through surgical instruments. Exfoliated cancer cells that may have detached from the primary tumor during the surgical procedures could have contributed to local recurrence along the longitudinal suture line of the pouch.     

Partial separating gastrojejunostomy for incurable cancer of the stomach or pancreas

BACKGROUND/AIMS: Advanced stomach or pancreas cancer with antral obstruction has been treated by gastrojejunostomy. The delayed return of gastric emptying, however, frequently occurs. The Devine exclusion procedure has been reported to be the better bypass operation in terms of oral intake, but it needs a drainage tube. In cases where the lesser curvature is invaded, this operation should be avoided. A method of gastroenterostomy, which is safe and shows good outcomes concerning oral intake, is desired. METHODOLOGY: Among 15 patients with advanced stomach or pancreas cancer, 8 received conventional gastrojejunostomy (CG Group), 3 Devine exclusive gastrectomy with a drainage tube (DE Group) and 4 partial separating gastrojejunostomy (PG Group). The partial separating gastrojejunostomy was performed as follows. The stomach was partially partitioned using GIA from the side of the greater curvature. The posterior side of the proximal stomach was anastomosed with the proximal jejunum using a circular stapler instrument. RESULTS: All patients in the DE and SG Groups could eat regular or semi-regular meals. The bleeding from tumor in the DE Group was less than that in the SG and CG Groups. CONCLUSIONS: In cases where the lesser curvature is invaded by tumor or lymph node metastasis, partial separating gastrojejunostomy would be recommended as a substitute for the Devine procedure.