IAGS 38~893-896,1990

Early Endoscopic Sphincterotomy in the Management of Acute Gallstone Pancreatitis in Elderl; Patients J

EfiahouShemesh, MD,*Abraham Czemiak, MD,f ShfomoSchneabaum, MD, f and Shlomo Nass,MD# Eighteen elderly patients with acute attacks of gallstone pancreatit is underwent early endoscopic sphincterotomy of the papilla of Vater. Eleven patients were considered to be at high risk for surgery due to chronic cardiorespiratory or renal problems. The outcome of these patients was compared with that of 20 consecutive elderly patients with gallstone pancreatitis treated at the same time by means other than endoscopic sphincterotomy. Endoscopic sphincterotomy resulted in an immediate clinical improvement in all patients, except in one patient who developed transient cholangitis; there was no mortality. In

contrast, there was one death ( 5 % )and 2Ooh morbidity in the controls. Mean hospitalization period was shorter in patients undergoing sphincterotomy (6 compared with 9.5 days), although the patients managed by sphincterotomy were initially more seriously ill than controls. Only two of the 2 2 high-risk patients underwent elective cholecystectomy; all others were well during a mean follow-up of 22 months. It is concluded that early endoscopic sphincterotomy is highly effective and safe in acute attacks of gallstone pancreatitis in elderly high-risk patients. J Am Geriatr Soc 38:893-896,1990

cute gallstone pancreatitis (AGP), caused by impaction of common bile duct stone at the papilla of Vater, occurs in about 5% of patients with cholelithiasis and is associated with a sigruficantmorbidity (42%) and mortality (1 1%), mainly in elderly patients.1-3 Early alleviation of the obstructed common bile duct is useful, both to hasten recovery and to prevent recurrent pancreatitis, which occursin up to 60% of patients not undergoing drainage of the bile ducts.'> Until recently, surgical decompression was the treatment of choice for AGP. However, urgent surgery in acutely ill patients is associated with sigruficantmorbidity and mortality. It has recently been shownthat endoscopicretrogradecholangiopancreatography (ERCP) with endoscopic sphincterotomy (ES) is safe and helpful in the diagnosis and treatment of AGP.6-10 However, the reported experience is still small, and the role of ERCP and ES in AGP, particularly in the seriously ill and aged, has not been fully defined. The present study examines the outcome and safety of early ES in the treatment of acutely ill older patients having gallstone pancreatitis.

PATIENTS AND METHODS

A

During 6 years, from 1982 through 1987,19 elderly patients with AGP were managed by early ERCP and ES as initial treatment (group A). The control group (group B) consisted of 20 consecutive patients, hospitalized during the same time period, who did not undergo ERCP and ES; they were treated conservatively at first, followed usually by cholecystectomy under elective timing and conditions. Acute gallstone pancreatitis was diagnosed on the basis of clinical presentation and serum amylase over 1,000 IU. There were 12 men and 27 women, with a mean age of 75.6 years (range, 65-92 years). The groups were not significantlydifferent in age or gender.

Clinical Data The patients' clinicaldata are summarized in Table 1. All patients underwent ultrasonographic investigationof the liver bfiary system and pancreas, and laboratory studies of liver biochemistry and serum amylase. Gallstones were detected in all patients (either by ultrasonography or by ERCP). None of the patients gave a history of alcoholism or had evidenceof hyperparathyroidism. Thirteen patients (five in group A, eight in group B) had undergone cholecystedomy FromtheDepartmentsof *Gastroenterology,tSwgery, and-diology, Sheba Medical Center, Sackler school of Medicine, Tel-Aviv previously. Thirteen patients (nine in group A, four in University,Tel-Aviv, Israel. group B) with an intact gallbladder containing stones Address correspondenceand reprint requests to Eliahou Shemesh, MD, Gastroenterology Department, Sheba Medical Center, Tel-Ha- and having serious chronic cardiorespiratory or renal problems were considered high-risk for surgery. Nine shomer 52621, Israel. 0 1990 by the American GeriutricsSociety

0002-8614/90/$3.50

894

IAGS-AUGUST 1990-VOL 38, NO. 8

SHEMESH ET AL

TABLE 1. CLINICAL DETAILS OF PATIENTS WITH GALLSTONE PANCREATITIS MANAGED BY EARLY ERCP AND ES (GROUP A) AND PATIENTS NOT MANAGED BY ERCP AND ES (GROUP B)

Group A

Group B

No. of patients 19 20 Ma1e:female 6:13 6:14 77.1 (65-91) 74.3 (65-92) Mean age (yr)(range) Previous cholecystectomy 5 8 Acute cholecystitis 4 5 Serum biochemistry > 1,500 > 1,000 Amylase 1 - 15.4 (4.5) 1 - 8.0 (2.5) Bilirubin (mg90) (mean) N Transaminase (IU) N+NX2 Alkaline phosphatase N-NX3 N-NX2 11 4 High-risk patients Mean follow-up (mo) (range) 20 (7-70) 21 (7-71) ERCP, endoscopic retrograde cholangiopancreatography;ES, endoscopic sphincterotomy; N, normal value.

TABLE 2. RESULTS OF STUDY

Group A No. of patients Sphincterotomy attempts (% of success) Choledocholithiasis Cholelithiasis Mean no. of gallstones per patient (range) Common bile duct Gallbladder % of gallstones < 1 cm Pancreatic duct Irregularity Dilatation Hospitalization (d) No. of complications Mortality

Group B

19 19 (10090)

20 0

15 11

10 15

2.7 (1-5) 9 (5-20) 90

5 12 6 0 0

1.2 (1-2) 9 (2- 15) 65

9.5 1 1

Radiologic Findings Gallstones were found in all patients of both groups. Cholelithiasis was found in 11 patients (58%) of group A and 15 patients (75%) of group B. Choledocholithiasis was present in 76% of group A and 50% of group B. In both groups, the gallGroup A Endoscopic retrograde cholangiopancreastones were small and multiple. The mean number of tography was performed under light sedation using a gallbladder stones per host was nine in both groups small dose of Diazepam and Demerol; the procedure (range, 3-20). Sludge was found in the common bile took between 10 and 20 minutes. Olympus types JFlT duct of all patients in group A and in all patients underand JFlTlO fiberoptic endoscopes were used. The indigoing surgical exploration of the common bile duct. Pancations for early ERCP were either an acutely ill highcreatic duct abnormalities compatible with pancreatitis risk patient or a patient with unresolved pancreatitis were found in five patients of group A (two mild, three with protracted hyperamylasemia (of 4 days or more). moderate); there were no pancreatic pseudocysts. All Endoscopic retrograde cholangiopancreatography inother patients of group A had varying degrees of diffuse vestigations and ES were performed by a standard techdilatation of the pancreatic duct, with normal secondary nique." During ERCP, care was taken to inject the conbranches. trast medium (60% Sodium Diatrizonate and our standard contrast medium) into the pancreatic duct at a Endoscopic Appearance of the Papilla of Vater low pressure. All patients received antibiotics during The mucosa of the papilla was edematous and friable in and for at least 3 days after the procedure. Detection of all patients. In six patients the orifice of the papilla was stone(s)in the common bile duct was followed by diath- patulous, presumably indicating the recent passage of ermic ES and removal of stones with basket or balloon stones. Impacted gallstones were found in nine patients catheters. Five patients were investigated within 24 of group A: three patients had stones protruding hours of admission, three patients between 24 and 48 through the orifice of the papilla, and six had a promihours, four patients between 3 and 7 days, and seven nent papilla without protrusions of gallstones. Nine papatients between 5 and 7 days after admission. Eleven tients had duodenal diverticula adjacent to the papilla of patients were jaundiced (total bilirubin, 5 to 15 mg%) at Vater. the time of ERCP, and six patients were investigated soon after bilirubin returned to normal. Outcome of Early ERCP and ES Endoscopic retrograde cholangiopancreatography achieved early accuRESULTS rate diagnosis in all patients. Endoscopic sphincterotThe outcome and safety of ERCP and ES in patients omy was successful in all patients and resulted in an with AGP is presented in Table 2. Patients of group A immediate clinical improvement, including in the four were sicker than their controls based on Ranson's classi- patients with acute cholecystitis. Two patients later unfication.l 2 derwent elective cholecystectomy successfully. Surgery patients (four in group A, five in group B) had cholecystitis (with typical pain, tender mass, and ultrasonographic appearance) in addition to the symptoms of biliary or pancreatic obstruction.

JAGS-AUGUST 1990-VOL 38, NO. 8

was presumably shortened due to the preoperative ES because the surgeon did not have to explore the common bile duct. In our experience, early ES was technically simpler than ES in patients without AGP, due to a more patulous orifice of the papilla and easier incision through the friable edematous mucosa at the papilla of Vater. Among patients of group B, three patients underwent early surgical treatment, 15 underwent elective surgery later, and two patients were treated conservatively throughout. Two patients of group B were readmitted for recurrent AGP within 1 and 2 months after discharge (before the planned elective cholecystectomy). All other patients of group B underwent elective cholecystectomieswithin 3 to 6 months after discharge. Mean hospital stay for patients of groups A and B was 6 and 9.5 days, respectively. None of the patients of group A or B was readmitted for biliary symptoms within the 12 months after discharge. Complications and Mortality One patient of group A developed cholangitis, which improved with conservative treatment. There was no mortality related to ERCP or ES. Two patients in group B developed cardiopulmonary complications; one of them died, the other recovered. Four patients of group B (20%)developed cholangitiS. High-Risk Patients Nine patients of group A and four patients of group B had severe medical illnesses. All of them had gallstones both in the common bile duct and gallbladder. One of such patients in group A and all four patients of group B had acute cholecystitis. All nine patients of group A made a rapid and uneventful recovery after ES. One of the four high-risk patients of group B died of cardiopulmonary and renal failure. Two of the nine patients of group A and three of the four patients of group B underwent elective cholecystectomy. During a mean follow up of 22 months (for both groups), there were no biliary or pancreatic symptoms in all the other high-risk patients. DISCUSSION Acute gallstone pancreatitis is defined as an attack of pancreatitis in patients with gallstones, with no other detectable etiology.' Acute gallstone pancreatitis results from an impaction of common bile duct stones at the papilla of Vater. More than a century ago, Claude Bernard induced experimental pancreatitis by the injection of bile into the pancreatic duct of a dogi3 Opie, in 1901, showed than an impacted gallstone at the papilla of Vater in patients with a common channel could cause pancreatitis in humans." Acosta and Ledesma formed gallstones more frequently in the stools of patients with acute AGP than in patients with cholelithiasis without pancreatitis (85Y0 vs. 8%).3 These three observations form the basis for the theory that pancreatitis is the

SPHINCTEROTOMYIN GALLSTONE PANCREATITIS 895

result of gallstone migration with a transient obstruction of the papilla of Vater; reflux of infected bile into the obstructed pancreatic duct presumably activates pancreatic enzymes in the pancreatic duct, resulting in pancreatic damage and edema.15 Severe symptoms occur in about 25% of patients.16 The severity of symptoms is directly correlated with the degree of obstruction at the papilla. An early decompression of the biliary system should and does hasten recovery from AGP. However, surgery in severely ill patients is associatedwith sigruficantmortality and morbidity. Until recently, ERCP was felt to be contraindicated in patients with pancreatitis because this test by itself may cause pancreatitis. However, in recent years it has been shown that ERCP and ES, when used cautiously, is helpful in patients with AGP (Table 3).6-10 Yet the utility and safety of early ERCP and ES are still controversial. In the present study, ES was safely performed during the course of acute pancreatitis (sometimes with cholecystitis) and resulted in immediate improvement in all patients. Endoscopic retrograde cholangiopancreatographywas performed by an experienced endoscopist, and the only technical precaution was to inject less contrast medium than usual into the pancreatic duct. Despite the fact that ERCP and ES were performed in more seriously ill patients than in the control group, faster recovery was achieved in these patients, with fewer complicationsand no mortality. There was some difference in the two groups of patients; bile duct dilatation was present in all group A patients but only in 60% of group B patients, and liver function tests were more abnormal in group A patients. Although this may apparently favor group A for any decompression procedure, we would emphasize that the difference in outcomes still existed when we excluded the 40% of patients without dilated bile ducts in group B. In our experience, ERCP and ES, even as an emergency procedure, has proved safe and effective, providing a rapid accurate diagnosis and more rapid clinical improvement than in the controls. Endoscopic sphincterotomy was successfuleven in septicpatients, and was well tolerated in high-risk patients. In contrast with previous suggestions to dilute contrast material to prevent irritation in AGP, we found a 60% concentration of contrast material as harmless. Therefore, considering our favorable experience with early AGP and ES, the amount and pressure of contrast injection and volume contrast concentration seem the most important risk factors. Endoscopic sphincterotomy was technically easier for inpatients with AGP compared with controls due to the more patulous opening of the papilla after stone passage and to the softness of tissue caused by the mucosal edema of the papilla of Vater. In such patients, ES is a very short procedure. The early performance of ERCP is mainly beneficial

896 SHEMESH ET AL

IAGS-AUGUST 1990-VOL 38, NO. 8

TABLE 3. REPORTS OF RESULTS OF ERCP AND ES IN AGP IN LITERATURE Authods)

n

Year

% of Success

Safrany and Cotton6 SPUY' Rosseland and Solhaug Classen and Phillip' Neoptolmos et all0 Shemesh et alt Total

11 10 44 58 13 19 155

1981 1981 1984 1984 1986 1990

100 100 100 100 100 94 100

Complications 0 1*

0 0 0 1$ 1(0.6%)

Urgent Surgery

Mortality

0 0 0 0 0 0 0

0 0

0 0 0 0 0

ERCP, endoscopic retrograde cholangiopancreatography;ES, endoscopic sphincterotomy; AGP, acute gallstone pancreatitis. * Perforation. t Present series. 4 Cholangitis.

for elderly high-risk patients because hazardous surgery is saved in most of these patients. Surgery in such patients should be considered only when there is cholecystitis, and therefore most high-risk patients with AGP will not need surgery after ERCP and ES. A small minority will develop acute cholecystitis later and will need cholecystectomy. In the present series, ERCP and ES provided a safe and effective alternative to surgery in high-risk patients having gallstonesboth in the gallbladder and in the common bile duct. Nevertheless, because AGP frequently improves spontaneously, more experience should be gathered before the exact indications for ERCP and ES alongside surgery should be defined. We do not suggest that ERCP should be performed in all patients with AGP. In our view, ERCP and ES should be performed as an early procedure mainly in patients with two conditions: in seriously ill, high-risk elderly patients with AGP, and in patients with slowly resolving pancreatitis. Endoscopic retrograde cholangiopancreatography should be performed as an elective procedure when the diagnosis of AGP is suspected but not proved. In low-risk patients who will need surgery, ES helps by stabilizing the patient and shortening surgery by avoiding common bile duct exploration. In postcholecystectomy patients, ES is the definitive treatment.

REFERENCES 1. Pellegrini C A Biliary panmatitis, in Way LW, Pellegrini CA

(eds):Surgery of the Gallbladder and Bile Ducts. Philadelphia,

WB Saunders, 1987, p 315

2. Bell AM, ORourke MGE Gallstone panmatitis. Med J Aust 144:572, 1986 3. Acosta JM, Ledesma CL Gallstone migration as a cause of acute pancreatitis. N Engl J Med 290:484,1974 4. Dixon JA, Hillam J: Surgical treatment of biliary tract disease associated with acute panmatitis. Am J Surg 120:371,1970 5. Stone HH, Fabian TC, Dunlop WE: Gallstone panmatitis. biliary tract pathology in relation to time of operation. Ann Surg 194:304, 1981 6. Safrany L, Cotton PB A preliminary report: urgent duodenoscopic sphincterotomy for acute galistone panckatitis. Surgery 89:424, 1981 7. Spuy S V D Endoscopic sphincterotomy in the management of gallstone pancreatitis. Endoscopy 13:25, 1981 8. Rosseland AR, Solhaug J H Early or delayed endoscopic papillotomy (EPT) in gallstone pancreatitis. Ann Surg 199:165,1984 9. Classen M, Phillip J: Endoscopic retrograde cholangiopanma-

tography (ERCP) and endoscopic therapy in pancreatic disease. Clin Gastroenterol 13:819, 1984 10. Neoptolmos JP, London N, Slater ND, et al: A prospective study of ERCP and endoscopic sphincterotomy in the diagnosis and treatment of gallstone acute panmatitis. Arch Surg 121:697, 1986 11. Classen M, Demling L Endoskopische sphinktertomie der pa12. 13. 14. 15. 16.

pilla Vateri und steinextraktion aus dem ductus choledochus. Dtsch Med Wochenschr 99:496,1974 Ranson JHC: Acute pancreatitis: surgical management, in Go VLW, Gardner JD, Brooks FB, et al: (eds):The exwine pancreas, New York, Raven Press, 1986, p 503 Bernard C: Le corse de physiologie experimental, Paris, JB Bailiere, 1856 Opie E L The relationship of cholelithiasisto disease of the pancreas and fat necrosis. Am J Med Surg 12:27,1901 Ofstad E: Formation and destruction of plasma kinins during experimental acute hemorrhagic panmatitis in dogs. Scand J Gastroenterol Suppl5:1, 1970 Frey FF: Gallstone panmatitis. Surg Clin North Am 61:923, 1981