the Harwood Academic Publishers imprint, part of The Gordon and Breach Publishing Group. Printed in Malaysia.

Bleeding from portal hypertensive gastropathy (PHG) has been estimated to account for upto 30% of all upper gastrointestinal haemorrhage in patients with cirrhosis and portal hypertension. Although portal hypertension seems to be an essential prerequisite, the precise mechanisms responsible for the development of PHG are unknown. The aim of this study was to examine the role of injection sclerotherapy of oesophageal varices in the development of PHG. Gastric emptying was studied using a radionuclide test meal with the emptying characteristics of a slow liquid in 57 patients with cirrhosis and/or portal hypertension (median age 53 yrs), of whom 34 had received injection sclerotherapy for their oesophageal varices and 20 normal healthy volunteers (median age 42 yrs). As vagal damage is associated with more rapid emptying of liquids, despite hold up of solids, this technique might be expected to demonstrate such damage if gastric emptying was accelerated. The results indicated that there was no difference in the rate of gastric emptying between normal healthy volunteers and portal hypertensive patients. However, patients who had received injection sclerotherapy emptied their stomachs faster than those who had not (p<0.05). Furthermore, the speed of gastric emptying correlated directly with the number of injections (r=0.41; p=0.02) and the volume of sclerosant injected (r=0.39; p=0.03). These observations suggest that injection sclerotherapy for oesophageal varices results in disturbances of gastric emptying that may contribute to the pathogenesis of portal hypertensive gastropathy.


INTRODUCTION
Gallstones and biliary sludge are the most com- mon etiology of acute pancreatitis [1][2][3].Chole- cystectomy is mandatory to avoid recurrence, whereas the timing of surgery is still controver- sial [4].Laparoscopic cholecystectomy is now the standard procedure in the management of gallbladder stones, but experience in instances of acute pancreatitis is limited [5][6][7].Moreover, in a multicenter study, most surgeons regarded acute pancreatitis as a contraindication for laparoscopic cholecystectomy [8], while others advocate early surgery, performed shortly after resolution of acute pancreatitis, pointing out the increased technical difficulty of the laparoscopic method [3].The wide acceptance of endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic sphincterotomy for urgent de- compression of the biliary system [9][10][11][12][13], en- ables the surgeon to plan an elective (interval) *Corresponding author.Tel.: 972-3-5028607, Fax: 972-9-9586544.laparoscopic cholecystectomy for the residual gallstones.
With this background we have evaluated our experience with acute pancreatitis patients predicted to have mild disease, treated conserva- tively initially and having interval laparoscopic cholecystectomy performed within 8-12 weeks from the attack.

PATIENTS AND METHODS
All patients presenting to our department be- tween January 1995 and June 1997 with acute biliary pancreatitis were evaluated for participa- tion in the study.Overall, about 40-50 patients with acute pancreatitis are hospitalized yearly in the department.Only patients that signed an informed consent were included in the study.
Acute pancreatitis was defined as acute abdo- minal pain with elevated serum and/or urine levels of amylase (serum levels > 700 IU/1, nor- mal range 70-220IU/1, urine levels >1500IU/1, normal < 1000 IU/1).Abdominal ultrasonography was performed on admission to determine the biliary etiology and confirm the diagnosis.Patients were stratified according to the severity of the disease by Ranson's criteria and included in the study only if there was evidence of gallbladder disease, no other etiologic cause of pancreatitis, with Ranson's score of 3 or less, (mild pancreatitis) and they did not' have previ- ous abdominal surgery (except appendectomy).Treatment was according to a standard protocol with no oral intake, nasogastric drainage, intra- venous fluids and oxygen mask if required.No drugs were prescribed except for non-narcotic analgesia.
Endoscopic retrograde cholangiopancreato- graphy (ERCP) was performed in all patients with acute pancreatitis on the first available endoscopy list.Endoscopic sphincterotomy was performed when the common bile duct was dilated with calculi or sludge demonstrated on ultrasonographic examination.If the initial ERCP failed it was repeated on an elective basis prior to surgery.Clinical improvement was defined as a reduction of abdominal pain and tenderness, normalization of laboratory values and neutral fluid balance.Oral diet was reintroduced gradually, the patients were discharged for an elective laparoscopic cholecystectomy 8-12 weeks later.
Surgery was performed by two experienced laparoscopists (P.S. and O.C.) using a three- cannula technique.The cystic duct and cystic artery were separately ligated with metal clips and the gallbladder was removed through the supraumbilical incision, which was dilated if necessary.The degree of difficulty of the pro- cedure was determined by the presence of adhesions to the gallbladder area, difficulty of dissection in the Calot's triangle, intraoperative bleeding, and the need for a drain.Following surgery, patients were managed with mobiliza- tion and reintroduction of diet as soon as tolerated.

RESULTS
During the study period, 19 patients (4 males and 15 females, age range 20-79 years, mean 59 years) with mild acute biliary pancreatitis qualified for the study.
Abdominal ultrasonography revealed a thick- ened gallbladder wall (more than 3 mm) in 14 patients.A dilated common duct (over 6 mm) and suspected choledocholithiasis were found in two patients.These two patients underwent ERCP with endoscopic sphincterotomy and clear- ance of the stones and sludge from the com- mon duct.
ERCP was successful in all patients, initially in 14 patients with acute pancreatitis and was repeated prior to surgery in the remaining 5 patients.
There were no complications related to the ERCP.
Laparoscopic surgery was performed after an interval of 8-12 weeks to allow the acute inflammation to settle.Table I presents the operative difficulty for laparoscopic cholecystectomy.Two patients underwent conversion to open sur- gery due to severe adhesions that prevented a satisfactory exposure of the anatomy in the Calot's triangle (one patient), and bleeding dur- ing the dissection of the cystic duct and artery (one patient).Another four patients of this group had marked adhesions, difficult dissections in the Calot's triangle and prolonged laparoscopic cholecystectomies.In one instance a drain was placed in the area of the cystic artery remn- ant due to unsatisfactory identification.The postoperative course was uneventful with a hospitalization stay of 2-5 days.

DISCUSSION
Acute biliary pancreatitis, is among the most commonly encountered complications of gallbladder stones and sludge.The majority of patients suffer mild attacks that respond promptly to medical treatment.Cholecystectomy is indicated to prevent recurrence.Laparoscopic cholecystectomy is now the procedure of choice, and unless absolute contraindications exist (uncorrectable Coagulopathy or concurrent diseases requiring laparotomy) all cholecystec- tomies are performed laparoscopically [14].Biliary pancreatitis in the acute phase, has been considered a relative contraindication to the laparoscopic approach as it increases the like- lihood of conversion, the operative difficulty and operating time [3,4,15,16].Moreover, ERCP with ES has become an important and very successful adjunctive in the management of acute biliary pancreatitis, offering the surgeon the possibility to postpone the intervention for several weeks and allow the acute inflamma- tion to settle.Therefore, even in spite of the fact that approximately one-quarter of patients have symptomatic recurrence within 6 weeks if not operated, and the rate increases with time [17- 19], the old concept of interval laparoscopic cholecystectomy has regained interest.Indeed, during the 8-12 weeks of delay there were no recurrent attacks in the patients of our group, since adequate drainage was demonstrated on ERCP, or created by the endoscopic sphincter- otomy.However, the results of our study, although based on small but carefully selected patient series, suggest that delayed (interval) laparoscopic cholecystectomy does not offer any benefit compared to reports of early interven- tions [3,4].The conversion rate was relatively high -10% for acute pancreatitis group, com- pared to our conversion rate for elective laparoscopic cholecystectomies -2.6% (8 from 303 laparoscopic cholecystectomies performed dur- ing the same period of the study).Moreover, dense, vascular adhesions to the gallbladder, a thick-walled gallbladder and difficult dissection in the Calot's triangle causing bleeding and prolonging operating time significantly were recorded in 6 patients of the acute pancreatitis group (including the conversion cases).We do not share the experience described of a dilated cystic duct that did not fit the standard ligating clip and required externally tied Roeder slip- knot ligation [3].The dilated cystic duct was encountered in the majority of patients operated early after the acute inflammation, therefore postponing surgery for 8-12 weeks, and ERCP with or without ES to ensure biliary drainage might explain the non-dilated cystic ducts in our series.The role of intra-operative cholangiography is controversial.In the presence of non- dilated bile ducts on ultrasonography, a normal ERCP or ES and with the knowledge that ERCP is feasible, we do not feel intra-operative cholangiography is essential.
In conclusion, our results suggest that the benefits obtained by interval laparoscopic cho- lecystectomy after acute pancreatitis are insuffi- cient to justify the risk of a further episode of acute pancreatitis, although early ERCP and ES are associated with a high success rate and low morbidity.In accordance with our results we recommend surgery during the same hospitali- zation, as soon as the signs and symptoms of acute pancreatitis have settled.

INVITED COMMENTARY ON
Schachter P, Peleg T, Cohen, O: "Interval laparo- scopic cholecystectomy in the management of acute biliary pancreatitis" Dr. Ronnie Tung-Ping Poon and Prof. Sheung-Tat Fan

Department of Surgery
The University of Hong Kong Medical Centre Queen Mary Hospital Hong Kong

COMMENTARY
The timing of cholecystectomy after acute gallstone pancreatitis is an important but un- resolved issue.A few studies in the open cholecystectomy era, including one prospective randomized trial, have shown that early chole- cystectomy within the same admission after resolution of pancreatitis was preferable as it avoided the risk of recurrent pancreatitis, while the operative morbidity and mortality were comparable to interval cholecystectomy [1][2][3][4].However, the management of gallstone pancrea- titis has undergone major changes over the past decade.First, early endoscopic retrograde cho- langiopancreatography (ERCP) and papillotomy have become widely accepted in the manage- ment of gallstone pancreatitis.Second, laparoscopic cholecystectomy has emerged as the standard operation for gallstone disease, even for patients presenting with pancreatitis.Hence there is a need to re-evaluate the optimal timing for cholecystectomy in the laparoscopic era.
The authors of this study attempted to address this issue by evaluating their results of interval laparoscopic cholecystectomy for gallstone pan- creatitis.A few previous studies of early laparoscopic cholecystectomy after pancreatitis have reported increased technical difficulties due to associated inflammatory changes, and hence a high conversion rate of 12-24% [5][6][7][8].The authors of the current study have found that- interval cholecystectomy 8-12 weeks after re- solution of pancreatitis was also associated with significant technical difficulties due to adhe- sions, and the conversion rate was 10% among 19 patients.This conversion rate was relatively high compared with 2.6% in their elective laparoscopic cholecystectomies, presumably for uncomplicated gallstone disease, and the authors concluded that interval cholecystectomy was not advantageous.The validity of the data was somewhat weakened by the small sample size.Their conclusion that early laparoscopic cholecystectomy in the same admission should be recommended could not be fully justified without any results of early operation in their study, although they did draw.a comparison with previous reports of early laparoscopic cho- lecystectomy after pancreatitis.Nevertheless, this study is worthy of inclusion in the literature, as little data on the results of interval laparoscopic cholecystectomy for gallstone pancreatitis were available.
The risk of recurrent pancreatitis or other complications of gallstone disease is the main argument against interval cholecystectomy.The authors of this study have quoted from the literature a 25% risk of symptomatic currence within six weeks, but the figure was derived from studies performed when early ERCP has not yet become a practice in the management of gallstone pancreatitis.The risk of recurrent pan- creatitis would presumably be reduced by early endoscopic removal of common bile stones.In fact, the authors found no recurrent attacks of pancreatitis in their patients.A previous study of 48 patients with gallstone pancreatitis who had undergone early ERCP followed by elective laparoscopic cholecystectomy in our institution also demonstrated no interim attacks of pancreatitis [9].These results suggested that the risk of recurrent pancreatitis was minimal with early ERCP and papillotomy.
The authors of this study have to be com- mended for their effort to resolve the question of the optimal timing for laparoscopic cholecystectomy after gallstone pancreatitis, but the controversy could not be settled without a prospective randomized study.A randomized trial of early versus interval laparoscopic chole- cystectomy is under way in our department, and it is hoped that a more definite answer could be provided after the conclusion of the trial.