This is a prospective study, in the form of consecutive case series, the protocol of which was approved by the ethical committee of Assiut Faculty of Medicine. Included in the study are all patients presenting to the authors with a diagnosis of benign biliary stricture for which surgical correction is indicated in the period from October 2008 till February 2011. For those patients other therapeutic options as percutaneous transhepatic cholangiography (PTC) or endoscopic retrograde cholangiopancreatography (ERCP) dilatation or stenting were either not applicable or not successful.
For all patients, full medical history, clinical examination, laboratory investigations in the form of liver function tests (LFT), prothrombin time and concentration (PTT), complete blood count (CBC) and kidney function tests (KFT) were performed. Imaging studies were also carried out in the form of abdominal ultrasonography (US), computerized tomography scan (CT) of the abdomen and magnetic resonance cholangiography (MRC) if indicated. ERCP and/or PTC were performed, whenever applicable, whether for diagnosis or therapeutic trial. Thorough preoperative medical fitness assessment was done before surgery.
Surgical technique: under general intubation anesthesia after administration of prophylactic antibiotic in the form of third generation cephalosporin, a generous right subcostal incision was performed and could be extended on demand upward to the xyphoid process and/or to the left subcostal area. Thorough dissection and adesiolysis was performed to reach the CBD and prepare the unaffected proximal part for anastomosis (Figure ). The Roux jejunal loop was prepared with lengthening of the mesentery and then passed retrocolic to reach the porta hepatis (Figure ). Afterwards, BEG was constructed depending on the type; I, II or III. In all types, the HJ is end to side anastomosis using interrupted sutures of polyglactin 910 of 4–0 size. The level of HJ is dependent on Bismuth classification [10
] of the level of the stricture and also on the diameter of CBD and the operative circumstances. If the level was at the carena (Bismuth type III), the HJ could be performed, using the Hepp-Couinaud technique [23
], with widening of the stoma by extension through the wall of the left hepatic duct. When the two ducts constituted separate openings (Bismuth type IV), the septum in between, if applicable, could be sutured and cut to transform both ducts into single stoma. The rest of the anastomoses, including the enterogastrostomy (EG), are constructed according to the type of BEG. All JG are anastomosed to the anterior wall of the stomach as near as possible to the pyloric orifice. In BEG type I (Figure ), a side loop of the Roux jejunum is anastomosed at its apex to the stomach with another side to side enteroenterostomy as a conduit of bile away from the stomach. In BEG type II (Figure ), the Roux jejunum is transected about 10 to 15
cm from HJ without interference with its mesentery. The distal end is anastomosed to the stomach, while the proximal end is anastomosed to the side of the distal loop as a conduit of bile. In BEG type III (Figure ), HJ is constructed leaving the distal end of the Roux jejunum long enough to be anastomosed to the side of the stomach. All the enterogastrostomies and enteroenterostomies were in the form of single-layer continuous sutures of polyglactin 910 of 3–0 size. An intraperitoneal drain was left in the hepatorenal pouch before closing the incision.
Preparation of the proximal CBD stoma.
a) Preparation of the Roux jejunal loop after lengthening of the mesentery. b) Passing the loop retrocolic.
BEG type I.a) and b) Diagrams of the construction. c) Enterogastrostomy [EG]. d) Enteroenterostomy [EE].
BEG type II.a) and b) Diagrams of the construction. c) Enterogastrostomy [EG]. d) Enteroenterostomy [EE].
BEG type III.a) and b) Diagrams of the construction. c) Enterogastrostomy [EG].
The choice of the type of BEG was an evolving process. We began with the BEG type I. The endoscopist faced difficulty with this technique due to the presence of more than one lumen at the site of gastroenterostomy making the decision and technique of entery more complex. Consequently, we modified the reconstruction to the BEG type II which appeared to be relatively easier for the endoscopist. However, some difficulty was still present for the endoscopic approach due to the side to end enteroenterostomy needed for this type. Moreover, BEG type II was more complex to be performed surgically. Consequently the technique was further modified to BEG type III which was not only the simplest to perform surgically but also the easiest for the endoscopist to reach the anastomosis due to the presence of only one lumen at the site of the enterogastrostomy. This may explain the sequence of cases in relation to the type of BEG, the increased number of BEG type III and the concentration of the cases with this type toward the later period of the study. Consequently, toward the end of the study, we decided to perform all cases using BEG type III.
Postoperatively, antibiotics in the form of third generation cephalosporin were administered twice daily for at least five days. The patients were closely followed up for any bile leak, improvement in the LFT, restoration of gastrointestinal motility and oral feeding. The patients were discharged from the hospitals once they were well mobilized, oral feeding was restored, and LFT were relatively improved.
Follow up visits, after hospital discharge, were scheduled weekly for the first four weeks then monthly for three months then every six months or whenever there were abnormal symptoms. In each visit, the patients underwent clinical assessment specially manifestation of cholangitis and laboratory assessment specially serum bilirubin and alkaline phosphatase levels. In the fourth postoperative week, endoscopic assessment of the BEG shunt with evaluation of the HJ was scheduled for all patients.
Endoscopic technique: within the 4th postoperative week, gastroenteroscopy was performed using end-view gastroduodenoscope (pentax 3440, Tokyo, Japan). The aim was to assess the feasibility to access the HJ stoma and perform cholangiography. With the patient in the left lateral position, monitored anesthesia care (MAC) sedation with propofol was used with an initiation dose of 100–150 mcg/kg/min for a period of 3–5 minutes and a maintenance dose of 25–75 mcg/kg/min which was adjusted to clinical response. The endoscope was introduced through the esophagus to the stomach where we assessed the amount of bile in the stomach on a scale of 0 to 2 where 0 meant no bile, 1 meant minimal amount of bile staining the gastric mucosa, and 2 meant large amount of bile accumulating and needed to be sucked. The gastroenterostomy stoma was assessed as regard to its site, diameter, as well as the difficulties we faced to pass through it. We passed through the gastroenterostomy either directly with the scope or over a guiding catheter. Thereafter, when we reached the HJ stoma, we inject a diluted dye into the cannulated bile ducts using an ERCP catheter to obtain cholangiography. Time to reach the gastroenterostomy as well as time to reach the HJ was reported for each case. Failure of endoscopic access was defined as failure to reach the HJ and perform cholangiography. For failed cases, the cause of failure and the type of BEG performed were reported. After the end of each procedure, the endoscopist was asked to score the difficulty of it on a scale from 1 to 5 where 1 is the easiest and 5 is the most difficult. It is worth mentioning that all endoscopies were performed with the same endoscopist (HE) who has more than fifteen years of experience in gastrointestinal endoscopy including ERCP and therapeutic endoscopies.
During the follow up, if the patient suffered cholangitis that required medical treatment, imaging studies were performed including abdominal US and MRC to exclude the presence of HJ stricture. If stricture did exist, endoscopic diagnosis and therapeutic trial was performed aiming at either balloon dilatation and/or stent placement through the strictured anastomosis. According to Terblanche grading of clinical outcome of HJ (Table ) [24
], the patients were assessed and categorized.
Terblanche clinical grading of long term clinical results of hepaticojejunostomy
All data were prospectively collected for assessment and for comparing the results of the three different types of BEG.