Outcome of pancreatic ascites in patients with tropical calcific pancreatitis managed using a uniform treatment protocol

Transcription

1 Indian J Gastroenterol 2009(May June):28(3): CASE SERIES Outcome of pancreatic ascites in patients with tropical calcific pancreatitis managed using a uniform treatment protocol Prakash Kurumboor Deepak Varma Mahendra Rajan Naduthottam Palanisami Kamlesh Roshin Paulose Ramesh Ganesh Narayanan Mathew Philip Abstract Pancreatic ascites or internal pancreatic fistula is a known complication of chronic pancreatitis. This condition is associated with considerable morbidity and mortality. The management approach of pancreatic ascites in tropical calcific pancreatitis is infrequently reported owing to the low incidence of this condition. Between December 2005 and June 2007, 11 patients with pancreatic ascites with tropical calcific pancreatitis (male:female 7:4, mean age 29.5 [14.2] years) were treated. A retrospective analysis of patients who underwent endotherapy and surgery for this condition based on an institutional protocol was performed. The end point was resolution of pancreatic ascites and relief of symptoms. All patients had pancreatic ascites, and one patient also had pancreatic pleural effusion. Endoscopic transpapillary stenting was possible in nine patients (81 ). Identification of site of leak and placement of an endoscopic stent across the PD disruption was possible in five (45 ) patients. All these patients had relief of ascites. Mean number of endotherapy sessions required before control of ascites was 1.8. Among the remaining four (36.6 ) patients who had ERCP, placement of stent across the leak was unsuccessful; however stenting helped stabilize the general condition and nutritional status. These four patients and two patients who failed ERP underwent lateral pancreatojejunostomy surgery. Morbidity was observed in three patients who underwent surgery and P. Kurumboor D. Varma M. Rajan N. P. Kamlesh R. Paulose R. G. Narayanan M. Philip PVS Memorial Hospital, Kaloor, Kochi , Kerala, India P. Kurumboor ( ) drkprakash@vsnl.com Received: 4 November 2008 / Revised: 10 January 2009 / Accepted: 7 February 2009 Indian Society of Gastroenterology 2009 one patient died due to sepsis and hemorrhage. All patients who had surgical drainage had complete relief of ascites and symptoms. In patients with pancreatic ascites in tropical calcific pancreatitis endotherapy and transpapillary stenting helps in resolution of ascites in nearly half of the patients. In the remaining patients preliminary conservative management followed by surgical pancreatic ductal drainage provides good relief of symptoms. Keywords Diabetes mellitus Pancreatic endotherapy Steatorrhea Introduction Pancreatic ascites or internal pancreatic fistulae and pancreatic pleural effusion are uncommon complications of chronic pancreatitis. Pancreatic ascites can occur due to the rupture of a pseudocyst or disruption of a main pancreatic duct (MPD) during the natural course of chronic pancreatitis. Pancreatic ascites was first reported in 1953 when Smith described two cases of ascites associated with chronic pancreatitis. 1 Because of the low incidence of pancreatic ascites the exact incidence of this clinical condition is not well known. There are no data of pancreatic ascites on tropical calcific pancreatitis (TCP). Traditionally conservative approach comprising bowel rest, parenteral nutrition and somatostatin analogues have been tried with resolution of ascites in of cases. 2 7 This approach has also been associated with significant mortality up to 17 2 in an earlier study. However, with improved imaging techniques and with better understanding of the pathophysiology of the disease, interventional approach is currently favored for patients who do not improve following an initial conservative trial. 7 9 Planning of therapy endoscopic or surgical requires accurate localization of the disruption and delineation of morphology of the duct. This is currently done based on endoscopic retrograde pancreatography (ERP) findings, 7,9 though magnetic cholangiogram findings may show up breach in MPD. 8 This paper aims at analyzing the management of pancreatic ascites in patients with TCP based on preliminary ERP findings.

2 Outcome of pancreatic ascites in tropical calcific pancreatitis 103 Methods Between December 2005 and June 2007, 180 patients with TCP presented to our unit. Diagnosis of TCP was made in patients with recurrent abdominal pain, steatorrhea, diabetes mellitus, presence of large intraductal calculi and/or changes of chronic pancreatitis, and no history of alcohol abuse. 10,11 Out of these patients, 11 (6.1 ) had pancreatic ascites; one patient had associated pancreatic pleural effusion also. Their mean age was 29.5 (14.2) years, and male to female ratio was 7:4. Patients presented clinically with abdominal pain, abdominal distention and generalized weakness. Investigations and assessment All patients underwent preliminary investigations, ultrasonography, and ascitic fluid assay. Diagnosis of pancreatic ascites was made if the ascitic fluid amylase is more than 1000 IU/ml. 5,6 A CT scan was done to delineate the morphology of pancreatic parenchyma, calculi, associated pseudocyst, size of the pancreatic duct and its course. Further investigations and planning of treatment was based on the treatment protocol followed in the department (Fig. 1). Planning of therapy Patients with poor nutritional status were started on with nil per orally, nasojejeunal feeds ( kcal/day), octreotide injections 100 mcg tid subcutaneously and symptomatic measures. Endosonography followed by ERCP was used in planning of the therapeutic strategy. On ERCP, site of the leak and presence of strictures if any were identified. Further, a pancreatic sphincterotomy and placement of stent across the site of leak/stricture was performed. In those who had successful endoscopic stenting i.e., stent placed across the site of duct disruption, conservative treatment was continued with periodic assessment of the clinical status. A complete clinical response was taken as relief of symptoms, ascites and no leak on ERP. Partial relief was defined as persistence of amylase-rich ascites on ultrasonography despite clinical improvement and persistence of duct leak on follow up ERP. In those patients who had no clinical relief of ascites and continued symptoms, surgical therapy was planned. Similarly, patients who failed endotherapy (failure to canulate/stent MPD) underwent surgery following stabilization of the general condition. During surgery, identification of the duct disruption, complete ductotomy, stone clearance and lateral pancreatojejunostomy (LPJ) was performed. Results The patient details and relevant laboratory data are given in Table 1. The mean ascitic fluid amylase value was 11,387 (range ,560) IU/L. CT scan demonstrated dilated PD and stones in four patients, calculi in all patients and cysts in relation to pancreas in five patients. CT scan also showed disruption of pancreatic duct in three patients which correlated with ERP findings. In five patients duct disruption was partial and in the remaining distal duct could not be demonstrated due to persistent leak of dye due to the presence of complete stricture. The details of ERP findings, outcome of therapy and follow up are detailed in Table 2. Endotherapy group: Endotherapy could be successfully done in nine patients Five patients had successful outcome (relief of ascites and abdominal pain) after endotherapy. Three patients had leak from the body region and two patients had leak of dye from the cyst in the region of pancreatic tail. In these patients duct distal to the pancreatic leak Fig. 1 Therapeutic algorithm followed in patients with TCP and pancreatic ascites Indian J Gastroenterol 2009(May June):28(3): Springer

3 104 Kurumboor, et al. Table 1 Demographic features of patients with pancreatic ascites Parameters Mean (SD) Range Age (years) 29.5 (14.2) Male: Female 7:4 Duration of symptoms (months) Symptom (n [ ) Abdominal pain 11 (100 ) Abdominal distension 5 (45.5 ) Diabetes mellitus 4 (36.4 ) Investigations Hemoglobin (g/dl) 10.8 (1.4) Serum proteins (g/dl) 5.6 (0.5) Serum albumin (g/dl) 2.9 (0.7) Ascitic fluid amylase (IU/L) (11381) Table 2 Imaging and treatment details, and outcome of the cohort of patients with pancreatic ascites Nutritional Follow up Sex/Age USG CT Endoscopic pancreatography findings and treatment details therapy Surgery (months) Stent ERP Response Site of leak Stricture Stent across leak session ERP M/15 HC, Cyst tail Tail cyst No Yes Yes 1 Good Yes No 9 M/19 Cyst body, caculi Body cyst No Yes Yes 1 Good No No 18 M/42 Dil PD, Atrophy, Body Yes Yes Yes 3 Good Yes No 14* calculi M/12 Dil PD, Mul. Body No Yes Yes 2 Good Yes No 15* Calculi F/24 HC, Cyst tail Tail No Yes Yes 1 Good No No 10 F/44 Dil PD, Atrophy, Body Yes Yes No 3 Partial Yes LPJ 15 calculi M/50 Mul calculi, Cysts Tail cyst Yes No No 0 Nil Yes LPJ 14 body tail M/44 Dil PD, Mul. Neck Yes Yes No 4 Partial No LPJ 6 Calculi M/38 HC, Tail mass, Tail No Yes No 2 Partial Yes DP-LPJ 16 cyst F/21 HC, Tail mass, Tail Yes Yes No 3 Partial Yes DP-LPJ Died cyst F/19 Mul calculi, Cysts Neck Yes No No 0 Nil Yes CJ 10 head, body Abbreviations: Dil Dilated, Mul Multiple, HC head calculi, PD pancreatic duct, LPJ lateral pancreatojejunostomy, DP-LPJ distal pancreatectomy LPJ *on follow up endotherapy could be well seen and probably had partial pancreatic duct disruption. In all these patients stent could be placed across the demonstrated site of leak. Though ascites improved after the first endoscopic stenting, two of them are on endotherapy and periodic stent exchange for recurrent pain due to ductal obstruction and calculi. Resolution of ascites was noted after mean 36.2 (6.4) days. In the remaining four patients, full length of the pancreatic duct could not be demonstrated due to leakage of dye from the disruption with probable complete duct disruption/stricture. In these patients negotiation of stent across leak was not feasible despite multiple attempts (Table 2) and stent was left close to the site of leak. These patients also required more sessions due to migration of stent. Two patients had resolution of ascites and two had partial relief in ascites persistence of amylase-rich ascites on ultrasonography. Though these patients had clinical improvement (partial resolution of ascites), they subsequently required surgery due to persistent duct leak and persistent pain. Three patients developed infected ascites following ERP as evidenced

4 Outcome of pancreatic ascites in tropical calcific pancreatitis 105 by fever, leucocytosis and raised counts in the ascitic fluid and two patients had stent migration. Surgery group: Four patients who had partial relief of symptoms after endotherapy and two patients who had failed stenting underwent LPJ. In those who had partial relief, two patients had relief of ascites; however, they required surgery due to persistent pain and leak of dye from the PD on follow up ERP. These two patients had multiple calculi and stricture body-neck region. One of these patients had portal hypertension also. The remaining two patients had only partial relief of ascites after ERP and stenting. In one of these patients, during surgery stent was seen coming out of site of PD disruption instead of going beyond the stricture. Out of the two patients who failed stenting one had multiple pseudocysts with dye leaking from the head cyst. This patient underwent cystojejunostomy alone due to dense inflammatory adhesions and difficulty in exposing pancreas. She remains symptom-free now. Two patients who had cyst and inflammatory mass in tail of pancreas required distal pancreatectomy and LPJ. In both these patients stent could not be advanced beyond the site of leak and total duct disruption of pancreatic duct in the tail region. One of these patients had massive secondary hemorrhage in the second postoperative week requiring re-exploration; she however died due to septicemia in the third postoperative week. Follow up: All patients are followed up for a median of 11.5 months. All patients who had surgery have relief of ascites and symptoms. All patients who had endotherapy had relief of ascites; however two patients are on regular stent exchange once in three months due to recurrent pain due to chronic pancreatitis. Discussion Tropical calcific pancreatitis is a disease form of chronic pancreatitis in young non-alcoholic individuals characterized by large intraductal calculi, ductal strictures and parenchymal changes. 10,11 The disease is characterized with recurrent pain, diabetes, and/or steatorrhea and has susceptibility for cancer. 11,12 There are no data to suggest the incidence of pancreatic ascites or pleural effusion in patients with tropical pancreatitis. Though the exact incidence of pancreatic ascites is not well known in chronic pancreatitis, it has been reported that ruptured pseudocyst exists in approximately 80 of cases, a disruption of the pancreatic duct exists in 10, and an obscure leakage site exists in the remaining Management of these patients are challenging due to the compromised nutritional state, generalized weakness, diabetes mellitus and disease related factors like multiple strictures, large stones and inflammatory mass. There are many reports of successful endoscopic treatment of pancreatic ascites with chronic pancreatitis. 7,9,14,15 Generally, in a partly disrupted pancreatic duct, a stent which bridges across the disruption is associated with successful outcome. 9,15 Indian J Gastroenterol 2009(May June):28(3): Conversely, a total duct disruption and stricture of pancreatic duct are poor predictors of successful outcome. 7,9 A similar outcome was observed in the current study. We followed an algorithm predominantly based on ERP findings and success of initial endoscopic stenting. All patients who had successful stenting across the site of disruption had control of ascites and symptoms. This was achieved with a mean of 1.8 sessions. Two patients have had their stent removed and two of them are on repeated stent exchange as they had recurrent pain after stent removal, probably due to calculi in the head duct though they had no ascites. We feel that even when the endoscopic stent placement across the disruption was unsuccessful, placement of stent or nasopancreatic drain up to the level of disruption along with nasojejeunal feed and supportive therapy is helpful in stabilizing patient s clinical condition. This was observed in four patients who had preliminary endotherapy and nasojejunal feeds and eventually required surgery. Success of endoscopic treatment was based on three factors in this study: ability to pass the stent across the site of disruption of the duct; absence of strictures and stones, and ability to traverse the stricture. Patients who had only partial relief of symptoms and those with failed endotherapy required surgery. The aims of the surgery in patients with chronic pancreatitis are: wide drainage of pancreatic duct (specific repair of the leaking site is often unnecessary) often with a LPJ, removal of pancreatic stones and ductal strictures, and drainage of the cysts and external drainage of abscess. Surgery for pancreatic ascites is often difficult due to the inflammatory process in the peripancreatic tissue, mesentery and due to the presence of pseudocysts and abscess. 4,16 However, good results of ductal drainage and LPJ in patients with pancreatic ascites have been reported. 2,4,6,17 Duct drainage has been proven to be sufficient in patients with chronic pancreatitis with pseudocysts. 18 In patients with pancreatic ascites, identification of pancreatic duct and site of leak is often difficult due to inflammation and hence cystojejunostomy to the leaking cyst is an effective method in treatment. 17 In patients with TCP, large duct stones and strictures are common and inadequate duct drainage is associated with recurrent pain. 12 Similar strategy was adopted in the current study and all except one patient had undergone LPJ which takes care of pancreatic ascites was well as duct stones and strictures. Selection of patients for surgery should be carefully done. Endotherapy along with enteral feeds prior to the surgery helps to control the ascites and improve nutritional status of these patients. Contrary to one previous study observation of shortened small bowel mesentery and technical difficulty in establishing anastomosis, 16 all patients in this study had undergone LPJ. All patients had symptomatic relief of ascites and pains on a median follow up of 11.5 months. In conclusion, an algorithmic approach based on ERP findings and planning of therapy provides excellent relief in symptoms. Endoscopic transpapillary stenting across 1 Springer

5 106 Kurumboor, et al. the site of leak results in resolution of ascites in nearly half of the patients. Patients in whom endotherapy is not successful probably a group of patients with ductal strictures and stones surgical duct drainage (LPJ) provides relief of symptoms. References 1. Smith EB. Hemorrhagic ascites and hemothorax associated with benign pancreatic disease. Arch Surg 1953;67: Lipsett PA, Cameron JL Internal pancreatic fistula. Am J Surg 1992;163: Pottmeyer EW III, Frey CF, Matsuno S. Pancreaticopleural fistulas. Arch Surg 1987;122: da Cunha JE, Machado M, Bacchella T, et al. Surgical treatment of pancreatic ascites and pancreatic pleural effusions. Hepatogastroenterology 1995;42: Cameron JL, Keiffer RS, Anderson WJ, Zuidema GD. Internal pancreatic fistulas: pancreatic ascites and pleural effusions. Ann Surg 1976;184: Kaman L, Behera A, Singh R, Katariya RN. Internal pancreatic fistulas with pancreatic ascites and pancreatic pleural effusions: recognition and management. ANZ J Surg 2001;71: Chebli JM, Gaburri PD, de Souza AF, et al. Internal pancreatic fistulas: proposal of a management algorithm based on a case series analysis. J Clin Gastroenterol 2004;38: Fulcher AS, Capps GW, Turner MA. Thoracopancreatic fistula: clinical and imaging findings. J Comput Assist Tomogr 1999;23: Varadarajulu S, Noone TC, Tutuian R, Hawes RH, Cotton PB. Predictors of outcome in pancreatic duct disruption managed by endoscopic transpapillary stent placement. Gastrointest Endosc 2005;61: Balakrishnan V, Nair P, Radhakrishnan L, Narayanan VA. Tropical pancreatitis a distinct entity, or merely a type of chronic pancreatitis? Indian J Gastroenterol 2006;25: Barman KK, Premalatha G, Mohan V. Tropical chronic pancreatitis. Postgrad Med J 2003;79: Ramesh H, Augustine P. Surgery in tropical pancreatitis: analysis of risk factors. Br J Surg 1992;79: Fernandez-Cruz L, Margarona E, Llovera J, Lopez-Boado MA, Saenz H. Pancreatic ascites. Hepatogastroenterology 1993;40: Bhasin D, Rana SS, Siyad I, et al. Endoscopic transpapillary nasopancreatic drainage alone to treat pancreatic ascites and pleural effusion. J Gastroenterol Hepatol 2006;21: Telford JJ, Farrell JJ, Saltzman JR, et al. Pancreatic stent placement for duct disruption. Gastrointest Endosc 2002;56: Selvakumar E, Vimalraj V, Rajendran S, et al. Pancreaticogastrostomy for pancreatic ascites Hepatogastroenterology 2007;54: Dhar P, Tomey S, Jain P, Azfar M, Sachdev A, Chaudhary A. Internal pancreatic fistulae with serous effusions in chronic pancreatitis. ANZ J Surg 1996;66: Nealon WH, Walser E. Duct drainage alone is sufficient in the operative management of pancreatic pseudocyst in patients with chronic pancreatitis. Ann Surg 2003;237: Indian Journal of Gastroenterology J Mitra Memorial Award The Indian Journal of Gastroenterology bestows this award for the best original scientific contribution published in the Journal during the year This award carries a prize of Rs , and will be given to the department(s) submitting the selected paper. The paper will be selected by a scientific committee appointed by the Editor, from among all the Original Articles published in the Journal during the year. In the event of a tie, the award will be distributed equally. Terms for eligibilty will apply. The award has been made possible by a generous endowment from M/s J Mitra and Co Ltd, New Delhi.