Effectiveness and safety of EUS-guided choledochoduodenostomy using lumen-apposing metal stents (LAMS): a systematic review and meta-analysis.

BACKGROUND: Endoscopic ultrasound-guided choledochoduodenostomy (CDD) is emerging as an alternative technique for biliary drainage in patients who fail conventional endoscopic retrograde cholangiopancreatography (ERCP). The lumen-apposing metal stents (LAMS) are being increasingly used for CDD. We performed a systematic review and meta-analysis to evaluate the effectiveness and safety of CDD using LAMS. METHODS: We performed a systematic search of multiple databases through May 2019 to identify studies on CDD using covered self-expanding metal stents. Pooled rates of technical success, clinical success, adverse events, and recurrent jaundice associated with CDD using LAMS were estimated. A subgroup analysis was performed based on use of LAMS with electrocautery-enhanced delivery system (EC-LAMS). RESULTS: Seven studies on CDD using LAMS (with 284 patients) were included in the meta-analysis. Pooled rates of technical and clinical success (per-protocol analysis) were 95.7% (95% CI 93.2-98.1) and 95.9% (95% CI 92.8-98.9), respectively. Pooled rate of post-procedure adverse events was 5.2% (95% CI 2.6-7.9). Pooled rate of recurrent jaundice was 8.7% (95% CI 4.5-12.8). On subgroup analysis of CDD using EC-LAMS (5 studies with 201 patients), the pooled rates of technical and clinical success (per-protocol analysis) were 93.8% (95% CI 90.4-97.1) and 95.9% (95% CI 91.9-99.9), respectively. Pooled rate of post-procedure adverse events was 5.6% (95% CI 1.7-9.5). Pooled rate of recurrent jaundice was 11.3% (95% CI 6.9-15.7). Heterogeneity (I2) was low to moderate in the analyses. CONCLUSION: CDD using LAMS/EC-LAMS is an effective and safe technique for biliary decompression in patients who failed ERCP. Further studies are needed to assess CDD using LAMS as primary treatment modality for biliary obstruction.

Outcomes in Pancreas Transplantation With Exocrine Drainage Through a Duodenoduodenostomy Versus Duodenojejunostomy.

Until recently, pancreas transplantation has mostly been performed with exocrine drainage via duodenojejunostomy (DJ). Since 2012, DJ was substituted with duodenoduodenostomy (DD) in our hospital, allowing endoscopic access for biopsies. This study assessed safety profiles with DD versus DJ procedures and clinical outcomes with the DD technique in pancreas transplantation. DD patients (n = 117; 62 simultaneous pancreas-kidney [SPKDD ] and 55 pancreas transplantation alone [PTADD ] with median follow-up 2.2 years) were compared with DJ patients (n = 179; 167 SPKDJ and 12 PTADJ ) transplanted in the period 1998-2012 (pre-DD era). Postoperative bleeding and pancreas graft vein thrombosis requiring relaparotomy occurred in 17% and 9% of DD patients versus 10% (p = 0.077) and 6% (p = 0.21) in DJ patients, respectively. Pancreas graft rejection rates were still higher in PTADD patients versus SPKDD patients (p = 0.003). Hazard ratio (HR) for graft loss was 2.25 (95% CI 1.00, 5.05; p = 0.049) in PTADD versus SPKDD recipients. In conclusion, compared with the DJ procedure, the DD procedure did not reduce postoperative surgical complications requiring relaparatomy or improve clinical outcomes after pancreas transplantation despite serial pancreatic biopsies for rejection surveillance. It remains to be seen whether better rejection monitoring in DD patients translates into improved long-term pancreas graft survival.

Late recurrent bleeding episodes from duodenojejunostomy after pancreas transplantation.

In response to a number of late, repetitive bleeding episodes from the site of the enteric anastomosis, we herein analyze the clinical courses and etiologies of 379 consecutively performed pancreas transplants between January 2000 and December 2016. Duodenojejunostomies for enteric drainage were performed at the upper jejunum in a side to side, double layer fashion. Five patients (1.3%) developed recurrent late hemorrhagic episodes originating from the graft duodenal anastomosis. Bleeding from the anastomotic site was associated with hematochezia, hemodynamic instability and decrease in serum hemoglobin. Mean onset was 6.4(±2.8) years after transplantation. Bleeding was recurrent (mean 5.2 ± 2.6) and required 9(±2.5) interventions. Hypervascularization, mucosal vulnerability, and bleeding at the site of the enteric anastomosis could be identified in all cases. In four patients, the enteric pancreas anastomosis was resected and a new duodenojejunostomy was performed. No pancreas graft loss occurred due to bleeding. In two patients, hepatic cirrhosis and portal hypertension were identified, one patient had a liver fibrosis as putative cause for the repetitive bleeding episodes. Late anastomotic hemorrhage is a rare but severe complication following pancreas transplantation. The treatment is challenging and includes endoscopy, interventional radiology, and surgery. Hepatic conditions with an increased portal pressure may be the underlying cause.

Intracorporeal delta-shaped gastroduodenostomy in reduced-port robotic distal subtotal gastrectomy: technical aspects and short-term outcomes.

BACKGROUND: Gastroduodenostomy is preferred as a method of reconstruction following distal subtotal gastrectomy. However, in initial reports on reduced-port gastrectomy, gastroduodenostomy has rarely been performed therein because of technical difficulties. The present study describes a novel intracorporeal gastroduodenostomy technique applicable during reduced-port robotic distal subtotal gastrectomy. METHODS: Data were retrospectively reviewed for cases of reduced-port (three-port) robotic distal subtotal gastrectomy with intracorporeal delta-shaped gastroduodenostomy performed from February 2016 to December 2016. The reduced-port approach used a Single-Site™ port via a 25-mm infraumbilical incision and two additional ports. We performed intracorporeal gastroduodenostomy using a 45-mm robotic or laparoscopic endolinear stapler. All staplers were inserted via a port on the left lower abdomen. RESULTS: In our initial experience with intracorporeal gastroduodenostomy, 28 consecutive patients underwent successful surgery with the technique without needing to convert to open, laparoscopic, or conventional five-port robotic surgery. Mean operation time was 201.1 min (110-282 min), and no major complications, including anastomosis-related problems, were recorded. CONCLUSIONS: Intracorporeal delta-shaped gastroduodenostomy was safely and feasibly applied during reduced-port robotic gastrectomy with acceptable operative outcomes and no major complications. Intracorporeal gastroduodenostomy should be considered during reduced-port distal subtotal gastrectomy.

Feeding duodenostomy decreases the incidence of mechanical obstruction after radical esophageal cancer surgery.

BACKGROUND: Nutritional support influences the outcome of gastroenterological surgery, and enteral nutrition effectively mitigates postoperative complications in highly invasive surgery such as resection of esophageal cancer. However, feeding via jejunostomy can cause complications including mechanical obstruction, which could be life threatening. From 2009, we began enteral feeding via duodenostomy to reduce the likelihood of complications. In this study, we compared duodenostomy with the conventional jejunostomy feeding, mainly looking at the catheter-related complications. METHODS: The database records of 378 patients with esophageal cancer who underwent radical esophagectomy with retrosternal or posterior mediastinal gastric tube reconstruction in our department from January 1998 to December 2012 were examined. Of the 378 patients, 111 underwent feeding via duodenostomy (FD) and 267 underwent feeding via jejunostomy (FJ), and their records were reviewed for the following catheter-related complications: site infection, dislodgement, peritonitis, and mechanical obstruction. RESULTS: Mechanical obstruction occurred in 12 patients in the FJ group but none in the FD group (4.5 % vs. 0 %, P = 0.023). Of the 12 cases, 7 (58.3 %) required surgery of which 2 had bowel resection due to strangulated mechanical obstruction. Catheter site infection was seen in 14 cases in the FJ group, of which 2 (14.2 %) had peritonitis following catheter dislocation, while only one case of site infection was seen in the FD group (5.2 % vs. 0.9 %, P = 0.078). CONCLUSIONS: Feeding via duodenectomy could be the procedure of choice since neither mechanical obstruction nor relaparotomy was seen during enteral feeding through this technique.

Prospective randomized clinical trial of a change in gastric emptying and nutritional status after a pylorus-preserving pancreaticoduodenectomy: comparison between an antecolic and a vertical retrocolic duodenojejunostomy.

BACKGROUND: Although an antecolic duodenojejunostomy was reported to reduce post-operative delayed gastric emptying (DGE) compared with a retrocolic duodenojejunostomy after a pylorus-preserving pancreaticoduodenectomy (PPPD), the long-term effects of these procedures have rarely been studied. The aim of this prospective, randomized, clinical trial was to investigate the influence of the reconstruction route on post-operative gastric emptying and nutrition. METHODS: Reconstruction was performed in 116 patients with an antecolic duodenojejunostomy (A group, n = 58) or a vertical retrocolic duodenojejunostomy (VR group, n = 58). Post-operative complications, including DGE, gastric emptying variables assessed by (13) C-acetate breath test and nutrition, were compared between the two groups for 1 year post-operatively. RESULTS: The incidence of DGE was not significantly different between the procedures (A group: 12.1%; VR group: 20.7%, P = 0.316). At post-operative month 1, gastric emptying was prolonged in the VR versus the A group but not significantly so. At post-operative month 6, gastric emptying was accelerated significantly in the A versus the VR group. Post-operative weight recovery was significantly better in the VR versus the A group at post-operative month 12 (percentage of pre-operative weight, A group: 93.8 ± 1.2%; VR group: 98.5 ± 1.3%, P = 0.015). CONCLUSIONS: A vertical retrocolic duodenojejunostomy was an acceptable procedure for the lower incidence of DGE and may contribute to better weight gain affected by moderate gastric emptying.

Endoscopic ultrasound-guided choledoco-duodenostomy as an alternative to percutaneous trans-hepatic cholangiography.

BACKGROUND: Endoscopic ultrasonography (EUS)-guided choledochoduodenostomy (CDS) is an alternative to percutaneous transhepatic cholangiography (PTC) drainage in patients with an obstructed biliary system where conventional endoscopic retrograde biliary drainage (ERBD) has been unsuccessful. METHODS: Five EUS-CDS procedures were reviewed to assess whether successful decompression was achieved and maintained. RESULTS: There was technical success in each instance with no immediate complications. There was a significant fall in the median bilirubin of 164 mmol/l. The median follow-up was 44 days. In one patient the stent migrated with no adverse outcome. CONCLUSION: EUS-CDS is a viable alternative to PTC with fewer complications and comparable success rates. EUS-CDS may offer a future route for novel therapeutic advances.

Evaluation of risk factors for morbidity and mortality after pylorectomy and gastroduodenostomy in dogs.

OBJECTIVES: To (1) identify and describe the type and frequency of postoperative complications after pylorectomy and gastroduodenostomy in dogs and (2) identify preoperative and intraoperative risk factors, including the presence of neoplasia, prognostic for patient mortality after surgery. STUDY DESIGN: Case series. ANIMALS: Dogs (n=24) treated by pylorectomy and gastroduodenostomy. METHODS: Medical records (2000-2007) for 2 teaching hospitals of dogs treated that had pylorectomy and gastroduodenostomy were reviewed. Pre-, intra-, and postoperative data were obtained from the medical record. RESULTS: Of the 24 dogs, 75% survived 14 days, but 10 (41%) died by 3 months. Overall median survival time (MST) was 578 days. On log-rank univariate analysis, preoperative weight loss (P=.001) and malignant neoplasia (P=.01) were associated with decreased survival time. Dogs with malignant neoplasia had a MST of 33 days. Common postoperative morbidity included hypoalbuminemia (62.5%) and anemia (58.3%). CONCLUSIONS: Pylorectomy with gastroduodenostomy has a good short-term outcome but long-term survival time is poor in dogs with malignant neoplasia. CLINICAL RELEVANCE: Overall, most dogs treated with pylorectomy and gastroduodenostomy survived the postoperative period; however, preoperative weight loss and malignant neoplasia were associated with decreased survival time. Because dogs with malignant neoplasia have markedly shortened survival times, pertinent preoperative, diagnostics steps should be exhausted to identify underlying neoplasia.

A hybrid endo-laparoscopic therapy for common bile duct stenosis of a choledocho-duodenostomy after a Roux-en-Y gastric bypass.

The essential growth of the number of Roux-en-Y gastric bypass procedures will obviously be accompanied by an increase of cases of common bile duct lithiasis. It seems evident that a close cooperation between surgeon and endoscopist will be needed on a routine basis. A laparoscopic-assisted transgastric ERCP is a well-documented approach to investigate the pancreatico-biliary tree in patients where the duodenum has been bypassed as in Roux-en-Y gastric bypass. In this case we present the possibility of assisting the endoscopist not only by providing access to the gastric remnant but also by helping with laparoscopic instruments during duodenoscopy. A formally obese woman who had benefited from a RYGB developed recurrent jaundice despite a precedent common bile duct exploration and choledocho-duodenostomy. A laparoscopic-assisted transgastric endoscopy revealed an obstructed choledocho-duodenostomy caused by accretions around a migrated clip. The obstructing clip could be extracted by laparoscopic instruments under endoscopic control.

Robotic choledochoduodenostomy for benign distal common bile duct stricture: how we do it.

Benign bile duct stricture poses a significant challenge for gastroenterologists and general surgeons due to the inherent nature of the disease, difficulty in sustaining long-term solutions and fear of pitfalls in performing biliary tract operations. Operative management with an open biliary bypass is mainly reserved for patients who have failed multiple attempts of endoscopic and percutaneous treatments. However, recent advances in minimally invasive technology, notably in the form of the robotics, have provided a new approach to tackling biliary disease. In this technical report, we describe our standardized method of robotic choledochoduodenostomy in a 59-year-old woman with history of Roux-en-Y gastric bypass who presents with benign distal common bile duct stricture following failure of non-operative management. Key steps in this approach involved adequate duodenal Kocherization, robotic portal dissection and creation of a side-to-side choledochoduodenal anastomosis. The operative time was 200 min with no intraoperative complications and estimated blood loss was less than 50 mL. No abdominal drains were placed. The patient was discharged home on postoperative day 1 tolerating regular diet and able to resume her usual activities within 1 week of her operation. A video is attached to this report.

Pancreas Transplant with Duodeno-Duodenostomy and Caval Drainage Using a Diamond Patch Graft: A Single-Center Experience.

BACKGROUND The surgical technique used in pancreas transplant is essential for patient safety and graft survival, and problems exist with conventional strategies. When enteric exocrine drainage is performed, there is no method of immunologic monitoring other than direct graft pancreas biopsy. The most common cause of early graft failure is graft thrombosis, and adequate preventive and treatment strategies are unclear. To overcome these disadvantages, we suggest a modified surgical technique. MATERIAL AND METHODS Eleven patients underwent pancreas transplant with our modified technique. The modified surgical techniques are as follows: 1) graft duodenum was anastomosed with recipient duodenum to enable endoscopic immunological monitoring, and 2) the inferior vena cava was chosen for vascular anastomosis and a diamond-shaped patch was applied to prevent graft thrombosis. RESULTS No patient mortality or graft failure occurred. One case of partial thrombosis of the graft portal vein occurred, which did not affect graft condition, and resolved after heparin treatment. All patients were cured from diabetes mellitus. There were no cases of pancreatic rejection, but 2 cases of graft duodenal rejection occurred, which were adequately treated with steroid therapy. CONCLUSIONS This modified surgical technique for pancreas transplant represents a feasible method for preventing thrombosis and allows for direct graft monitoring through endoscopy.

Current assessment of choledochoduodenostomy: 130 consecutive series.

Introduction Cholelithiasis usually can be managed successfully by endoscopic sphincterotomy. Choledochoduodenostomy (CDD) is one of the surgical treatment options but its acceptance remains debated because of the risk of reflux cholangitis and sump syndrome. The aim of this study was to assess the current features and outcomes of patient undergoing CDD. Patients and methods We retrospectively analysed the surgical results of consecutive 130 patients treated by CDD between 1991 and 2013 and excluded five cases with a malignant disorder. Indications for surgery included endoscopic management where stones were difficult or failed to pass and primary common bile duct stones with choledochal dilatation. Incidences of reflux cholangitis, stone recurrence, pancreatitis or sump syndrome were investigated and the data between end-to-side and side-to-side CDD were compared. Results Reflux cholangitis and stone recurrence was 1.6% (2/125) and 0% (0/125) of cases by CDD. There is no therapeutic-related pancreatitis in CDD. Sump syndrome was not also observed in side-to-side CDD. Conclusions This study is a first comparative study between end-to-side and side-to-side CDD. The surgical outcomes for CDD treatment of choledocholithiasis were acceptable. The incidence of reflux cholangitis, stone recurrence, pancreatitis and sump syndrome was very low.

Stapled anastomosis versus hand-sewn anastomosis of gastro/duodenojejunostomy in pancreaticoduodenectomy: A systematic review and meta-analysis.

BACKGROUND: Controversy exists regarding the best anastomotic method for pancreaticoduodenectomy (PD). We aimed to evaluate the perioperative outcomes of PD with stapled anastomosis (SA) versus hand-sewn anastomosis (HA) of gastrojejunostomy or duodenojejunostomy. METHODS: We conducted a systematic search of electronic information sources, including MEDLINE; EMBASE; CINAHL; the Cochrane Central Register of Controlled Trials (CENTRAL); the World Health Organization International Clinical Trials Registry; ClinicalTrials.gov; ISRCTN Register, and bibliographic reference lists. We applied a combination of free text and controlled vocabulary search adapted to thesaurus headings, search operators and limits in each of the above databases. Delayed gastric emptying (DGE), postoperative pancreatic fistula (POPF), anastomotic bleeding, anastomotic leak, intra-abdominal abscess and mortality were defined as the outcome parameters. Combined overall effect sizes were calculated using fixed-effect or random-effects models. RESULTS: We identified 1 randomised controlled trial (RCT) and 5 observational studies reporting a total of 890 patients who underwent PD with SA (n = 300) or conventional HA (n = 590). Our analysis demonstrated that SA significantly reduced postoperative DGE (OR: 0.37, 95% CI 0.25-0.54, P < 0.00001) but significantly increased anastomotic bleeding (OR: 13.4, 95% CI 2.96-57.41, P = 0.0007) compared to HA. No significant difference was found in POPF (OR: 0.83, 95% CI 0.56-1.21, P = 0.33); anastomotic leak (OR: 0.50, 95% CI 0.09-3.79, P = 0.58); intra-abdominal abscess (OR: 1.39, 95% CI 0.71-2.70, P = 0.34); or mortality (RD: -0.01, 95% CI 0.03-0.02, P = 0.65) between two groups. CONCLUSIONS: Our analysis demonstrated that compared to conventional HA, SA may be associated with lower incidence of DGE after PD without increasing the risk of clinically significant POPF, anastomotic leak or mortality. However, it is associated with higher rate of anastomotic bleeding which mandates careful and precise haemostasis of the stapled line. Considering the current limited evidence, no definitive conclusion can be drawn. Future research is required.

Bilateral Dilated Nonreactive Pupils in a Neonate After Surgery.

Fixed and dilated pupils are disturbing when encountered during a physical examination in the pediatric intensive care unit, particularly when sedation or neuromuscular blockade confounds the neurologic examination. Rocuronium, a nondepolarizing neuromuscular drug, does not cross the blood-brain barrier and is not considered a causative agent for fixed mydriasis. We report a case of bilateral fixed and dilated pupils in a 1-week-old low-birth-weight neonate, which we contend was secondary to centrally mediated neuromuscular blockade.

Delayed Gastric Emptying After Stapled Versus Hand-Sewn Anastomosis of Duodenojejunostomy in Pylorus-Preserving Pancreaticoduodenectomy: a Randomized Controlled trial.

BACKGROUND: A retrospective analysis indicated that the incidence of delayed gastric emptying (DGE) was less after using a circular stapler (CS) for duodenojejunostomy than that after hand-sewn (HS) anastomosis in pylorus-preserving pancreaticoduodenectomy (PpPD). This randomized clinical trial compared the incidence of DGE postoperative after CS duodenojejunostomy with that of conventional HS anastomosis in PpPD. METHODS: We randomly assigned 101 patients (age 20-80) undergoing PpPD to receive CS duodenojejunostomy (group CS, n = 50) or HS duodenojejunostomy (group HS, n = 51) in two Japanese cancer center hospitals between 2011 and 2013. The patients were stratified by institution and size of the main pancreatic duct (<3 or ≥3 mm). The primary endpoint was the incidence of grade B or C DGE according to the international definition with a non-inferiority margin of 5 %. This trial is registered with University hospital Medical Information Network (UMIN) Center: UMIN000005463. RESULTS: Per-protocol analysis of data on 95 patients showed that grade B or C DGE was found in 4 (8.9 %) of 45 patients who underwent CS anastomosis and in 8 (16 %) of 50 patients who underwent HS anastomosis (P = 0.015). There were no differences in the overall incidence of DGE (P = 0.98), passage of the contrast medium through the anastomosis (P = 0.55), or hospital stays (P = 0.22). CONCLUSIONS: CS duodenojejunostomy is not inferior to HS anastomosis with respect to the incidence of clinically significant DGE, justifying its use as treatment option.

Ten-year follow-up of a prospective, randomized trial of selective proximal vagotomy with ulcer excision and partial gastrectomy with gastroduodenostomy for treating corporeal gastric ulcer.

Between 1975 and 1980, 30 patients with type I corporeal gastric ulcer were randomly allocated to undergo selective proximal vagotomy with ulcer excision or partial gastrectomy with gastroduodenostomy. Sixteen patients underwent selective proximal vagotomy (1 was excluded from the follow-up since microscopic examination of the excised ulcer revealed an early gastric cancer) and 14 underwent partial gastrectomy. No significant differences in the clinical results were found 3 years after surgery. During a median follow-up of 10 years, ulcer recurred in 3 patients after selective proximal vagotomy and in 2 after partial gastrectomy. One patient in each group had recurrent ulcer without symptoms and received no treatment. Two selective proximal vagotomy patients and three partial gastrectomy patients had epigastric pain with or without ulcer. One patient with selective proximal vagotomy underwent a second operation because of epigastric pain and recurrent ulcer. Bowel habits remained unchanged in all but one patient in each group, and mild or moderate dumping was recorded for two patients in each group. Very good or good results (modified Visick scale) were recorded for 11 of 15 patients after selective proximal vagotomy and for 10 of 14 patients after partial gastrectomy. Except for one patient in each group who had moderate dumping, patients classified as Visick III or IV had no symptoms during treatment with antacids or H2-blockers, or had asymptomatic ulcers and needed no treatment. Selective proximal vagotomy reduced the median acid response to insulin hypoglycemia and to pentagastrin by 100% and 80%, respectively, for at least 3 to 5 years, and partial gastrectomy reduced the median acid response to pentagastrin by 97%. In our opinion, selective proximal vagotomy with ulcer excision is an alternative to partial gastrectomy for surgically treating type I gastric ulcer.

Superior results with combined kidney-pancreas transplants.

From February of 1987 to February of 1991 the authors performed 23 pancreas transplants for Type I diabetes mellitus. Eight of the pancreas transplants were in patients who had a previous kidney transplant, 14 were simultaneous kidney and pancreas transplants, and 1 was in a pre-uremic diabetic. Two patients have been retransplanted after losing first grafts. All pancreata were retrieved from heart-beating cadaver donors. Pancreata were transplanted into the iliac fossa of the recipient using the iliac artery and vein as arterial inflow and venous outflow, respectively. Drainage of the pancreatic ductal system was accomplished by anastomosing either a patch or segment of duodenum surrounding the ampulla of Vater to the urinary bladder. All pancreata functioned initially with no patient requiring insulin 6 hours after surgery. Two grafts were lost early due to thrombosis of the venous drainage of the transplant; 4 grafts were lost to acute rejection; 3 were lost to chronic rejection; and 1 patient died with a functioning pancreas. One-year graft survival for all pancreatic grafts is 62 per cent. One-year patient survival is 96 per cent. One-year pancreatic graft and patient survival for the 14 combined kidney-pancreas transplants is 88 per cent and 100 per cent, respectively. Two kidneys transplanted with pancreata also were lost to acute rejection. Pancreas transplantation has proven to be a viable treatment alternative for selected patients with Type I diabetes mellitus. Long-term results are best when pancreas transplantation is done in combination with renal transplantation.

Comparision of modified and conventional delta-shaped gastroduodenostomy in totally laparoscopic surgery.

AIM: To evaluate the safety and feasibility of a modified delta-shaped gastroduodenostomy (DSG) in totally laparoscopic distal gastrectomy (TLDG). METHODS: We performed a case-control study enrolling 63 patients with distal gastric cancer (GC) undergoing TLDG with a DSG from January 2013 to June 2013. Twenty-two patients underwent a conventional DSG (Con-Group), whereas the other 41 patients underwent a modified version of the DSG (Mod-Group). The modified procedure required only the instruments of the surgeon and assistant to complete the involution of the common stab incision and to completely resect the duodenal cutting edge, resulting in an anastomosis with an inverted T-shaped appearance. The clinicopathological characteristics, surgical outcomes, anastomosis time and complications of the two groups were retrospectively analyzed using a prospectively maintained comprehensive database. RESULTS: DSG procedures were successfully completed in all of the patients with histologically complete (R0) resections, and none of these patients required conversion to open surgery. The clinicopathological characteristics of the two groups were similar. There were no significant differences between the groups in the operative time, intraoperative blood loss, extension of the lymph node (LN) dissection and number of dissected LNs (150.8 ± 21.6 min vs 143.4 ± 23.4 min, P = 0.225 for the operative time; 26.8 ± 11.3 min vs 30.6 ± 14.8 mL, P = 0.157 for the intraoperative blood loss; 4/18 vs 3/38, P = 0.375 for the extension of the LN dissection; and 43.9 ± 13.4 vs 39.5 ± 11.5 per case, P = 0.151 for the number of dissected LNs). The anastomosis time, however, was significantly shorter in the Mod-Group than in the Con-Group (13.9 ± 2.8 min vs 23.9 ± 5.6 min, P = 0.000). The postoperative outcomes, including the times to out-of-bed activities, first flatus, resumption of soft diet and postoperative hospital stay, as well as the anastomosis size, did not differ significantly (1.9 ± 0.6 d vs 2.3 ± 1.5 d, P = 0.228 for the time to out-of-bed activities; 3.2 ± 0.9 d vs 3.5 ± 1.3 d, P = 0.295 for the first flatus time; 7.5 ± 0.8 d vs 8.1 ± 4.3 d, P = 0.489 for the resumption of a soft diet time; 14.3 ± 10.6 d vs 11.5 ± 4.9 d, P = 0.148 for the postoperative hospital stay; and 30.5 ± 3.6 mm vs 30.1 ± 4.0 mm, P = 0.730 for the anastomosis size). One patient with minor anastomotic leakage in the Con-Group was managed conservatively; no other patients experienced any complications around the anastomosis. The operative complication rates were similar in the Con- and Mod-Groups (9.1% vs 7.3%, P = 1.000). CONCLUSION: The modified DSG, an alternative reconstruction in TLDG for GC, is technically safe and feasible, with a simpler process that reduces the anastomosis time.