Acute pancreatitis (AP) is a condition of high metabolic stress related to the release of inflammatory mediators and proteolytic enzymes. The severity of inflammatory response depends on the associated local and systemic complications (1,2). Cases with parenchymal or peripancreatic necrosis can lead to multiorgan failure, which increases mortality (1,2). Therefore, early and aggressive management of local and systemic complications is of paramount importance. Pancreatic or peripancreatic inflammation could be self-limiting or lead to pseudocyst or walled-off pancreatic necrosis. These complications are specifically a consequence of pancreatic duct (PD) disruption, which is contained by an inflammatory fibrous wall. Chronic pressure and secondary infection can result in decompression in proximate organs creating a spontaneous pancreatic fistula (PF) (3–7). A fistula is defined as an abnormal communication between 2 epithelial surfaces (8). Fistulae are most often internal, in that they drain into another viscera. Such internal PFs include pancreaticocolonic, gastric, duodenal, jejunal, ileal, pleural, and bronchial fistulae. PFs are also classified based on their number either as simple (i.e., a single PF) or complex (i.e., multiple PFs) (8).
Unlike postoperative PF which has a well-developed classification system and management protocol (9), there is a lack of consensus or guidelines on the classification and management of AP-associated PF. PFs clearly increase the morbidity of AP, producing serious consequences such as hemorrhage, prolonged sepsis, respiratory distress, prolonged length of stay, and a higher cost of hospitalization. Clarifying the risk factors, pathogenesis, and natural history of PF may help guide management and avoid these complications in an already compromised patient population (5,6,10). This review will cover all aspects of AP-associated PFs including the state of the art in managing this complication.
The incidence of AP-associated gastrointestinal fistulae (GIFs) varies between 4.0% and 41.0% and incidence is higher in patients who undergo open necrosectomies (41.0%–47.0%) which are complex and typically external (3,4,10). The most common location of spontaneous internal fistulae is transverse colon and duodenum (Figures 1–6) (10,11). In a pooled analysis of 97 patients from 43 reports, 15.4% of patients had colonic complications (necrosis, fistulae, and strictures), and of these, 27.0% of them had colonic fistulae with transverse colon being the most common location (83.0%) (12). The higher prevalence in more recent studies is likely because their samples are enriched with more severe AP and because of improved detection by advanced imaging.
Pleural effusions, mostly left-sided, are relatively common in AP (4.0%–6.0%). They may be transudative or exudative and are postulated to occur from migration of peripancreatic fluid through the diaphragm or through crural openings. True epithelialized pancreaticopleural fistulae (PPFs) are an infrequent complication (0.4%) of AP (13,14). PPFs have been most commonly reported in patients with chronic pancreatitis (CP) and usually appears during acute exacerbations of CP (15). Subclinical or chronic pancreatic inflammation, often secondary to alcohol ingestion, can lead to pancreatic ductal disruption and PPF (7). Because of non-GI presentations, a high degree of clinical suspicion is often required to diagnose PPF. The major reason for discrepancies in incidence of PF is the type and severity of AP patients included in these studies. Most studies have included transferred patients with severe AP (SAP), suggesting high baseline severity of AP (5,6,11).
RISKS AND ASSOCIATIONS
The major risk factors for internal PF include male sex, alcohol abuse, degree of severity of AP, and infected pancreatic necrosis (IPN) (3–6). A significant male preponderance has been reported in various PF studies, ranging from 60% to 100% (5,6,10,11,16,17). Alcohol use has been associated with development of both GIF and PPF in 70%–83% of cases in early series (3,7). However, more recent studies include more patients with gallstone etiology, and PFs are being observed in this group as well (5,6). In a study of 52 patients with GIFs, 29 had biliary disease and only 7 had alcohol as etiology (5). Similarly, gallstone disease was the cause of AP with fistula in 55.5% compared with 9.2% in another study by Jiang et al. (5,6). PF seems to be a consequence of every etiology of pancreatitis with some variability in the incidence.
In comparison with IPN patients without GIF, IPN patients with GIF had higher acute physiology and chronic health evaluation II score (11.8 vs 14.5, P < 0.01) and multiple organ dysfunction score (2.0 vs 3.0, P = 0.01) (6). Similarly, in a study of 244 patients with SAP, on multivariate logistic analysis, the risk of GIF was 3-fold higher in patients with IPN (odds ratio [OR (5,6,16)] 3.01; 95% confidence interval [CI] 1.69–15.02, P = 0.004). Other risks included high-modified computed tomography (CT) severity index (OR 1.18; 95% CI 1.09–2.55, P = 0.03) and possibly B-blood type (OR 1.006; 95% CI 0.74–3.56, P = 0.04) (5).
The clinical manifestations of internal PF vary based on the acuity of pancreatitis (acute vs chronic) and location of the fistula. The most serious and life-threatening fistulae–related complications include hemorrhage and sepsis (6). Nearly 60% of patients with pancreaticoenteric (colon, duodenum, stomach, jejunum, and ileum) fistulae present with GI bleeding (11,18). Hemorrhage results from inflammation and necrosis induced erosion of vessels in the bowel wall. In a study of 12 patients with GIFs, 7 (58.0%) presented with GI bleeding (11). Additional signs of GIFs include vomiting of necrotic, feculent, or purulent material, diarrhea with or without blood, or persistent sepsis (11,12,16). Most GIFs are diagnosed beyond 4–8 weeks after the onset of AP (4–6,12).
Dyspnea (65%–76%) is the most common presentation in patients with PPF, followed by cough (27.0%), chest pain (23.0%), abdominal pain (29%), and fever (19,20). Recurrent pleural effusions associated with PPF are common (15,19,20).
A delay in diagnosis of PFs can lead to increase in length of stay, intensive care unit stay, and even in mortality (6). Oral contrast-enhanced CT performed 72–96 hours after the onset of symptoms is considered the first-line test for diagnosing AP (21). Based on the CT findings, modified CT severity index can be calculated which is a measure of severity of pancreatic inflammation and necrosis and predicts SAP, IPN, and fistulae (22,23). CT findings of PFs include an edematous and indistinct gut wall (Figure 6), and the presence of air in the pancreatic bed can raise the suspicion of PF (11,21,22).
In addition to CT, more costly and invasive tests may selectively be used to enhance the detection of PFs. Although CT scan has a sensitivity of 63.0%, magnetic resonance cholangiopancreatography (MRCP) and endoscopic retrograde cholangiopancreatography (ERCP) have sensitivities of 83.0% and 46.0%–78.0%, respectively (24). Although ERCP is invasive and carries risk, it does allow for simultaneous therapeutic interventions (15). ERCP may fail to detect fistula if there is an obstructing stone or stricture in the main PD proximal to the fistula (7).
In the case of PPFs and pancreatic ascites, aspiration of pleural and/or peritoneal fluid for analysis can be useful. The presence of high amylase (>1,000 U/L) and high protein (>3.0 g/dL) in pleural fluid is pathognomic of PPF. However, it is important to rule out other causes of amylase- and protein-rich pleural fluid, such as tuberculosis, leukemia, lymphoma, and malignancy (13,15,25). A low pleural fluid protein does not exclude PPF, since patients with acute or subacute pancreatitis could have hypoproteinemia because of malnourishment. Clarifying the source of the ductal leak is likewise important in planning management of PPF. This is most commonly performed with ERCP (sensitivity100.0%) or MRCP with or without secretin (sensitivity 83.0%) (26).
A multidisciplinary approach is often required to manage internal PF. In this review, we have divided the management of PF into 3 main categories:
Medical management of PF
Medical therapy is the mainstay of treatment for all PF. The biggest factor is to manage the sepsis caused by undrained pancreatitis associated collections and necrosis. Based on the location of fistulous tract (upper GI, colonic, or bronchus), the success rate of medical management varies from 30.0% to 84.0% (6,13,15,19). The goal of medical therapy is to control the degree of inflammation and prevent infectious complications. Meticulous supportive care of necrotizing pancreatitis is of paramount importance, including aggressive fluid support with ringer lactate during the first 24 hours. In patients with IPN, intravenous antibiotics with efficacy on gut-derived bacteria and good penetration into the pancreas should be given. Patients with symptomatic or infected pancreatic fluid collections will benefit from well-timed drainage, and this may in turn decrease the risk of PF (5,16,27). Early initiation of EN helps to preserve gut mucosal integrity, inhibit bacterial translocation, and decrease systemic inflammation and risk of infection (5,27). In a study of 344 patients with SAP, those started on early EN (48–72 hours) had significantly lower risk of developing GIF (OR 0.26; 95% CI 0.18–0.73, P = 0.016) (5). Nasogastric feeding is not inferior to nasojejunal feeding, but clinical decision should be made based on the site of fistula (gastric or duodenal), presence of gastric outlet obstruction, and intolerance to gastric tube feeding (27,28). Total parenteral nutrition is reserved in patients with luminal obstruction, intolerance to tube feeding or unable to reach caloric goal with tube feeding (27).
Another important but debatable aspect in the management of internal PF is the role of somatostatin analogs such as octreotide (19,29–32). In a recent meta-analysis of 7 randomized clinical trials of 297 patients with fistulae of the GI tract, no difference in PF closure was observed with somatostatin analogs compared with other treatments (32). However, in patients with PPF, octreotide has been shown to reduce fistula output and decrease time to fistula closure, with success rates between 33.0% and 65.0% (19,33,34). Despite marginal success of somatostatin analogs for treating external fistulae, they are considered less likely to help with internal PF, with the exception of PPF where benefit has been observed (19,33).
In cases of failure of medical/conservative management for PF, endoscopic/interventional procedures are the next step. The decision to perform ERCP should be carefully considered after reviewing the cross-sectional imaging. If there is a completely disconnected duct, e.g., in the case of complete necrosis of the neck or body of the pancreas, then ERCP has a very low chance of success and may instead lead to superinfection of the associated pancreatic collection. However, if there if the gland integrity is preserved, ERCP may allow for identification and treatment of more limited main duct or side-branch injuries to ameliorate the leak or help close the fistulae. Ideally, a pancreatic stent should be placed to bridge the site of leaking. However, if this is not possible (e.g., a leak in the most distal tail), an endoscopic sphincterotomy and placement of a short transpapillary stent may decrease intraductal pressure. This could decrease inflammation, accelerate healing, and possibly can heal ductal discontinuity (15,16,35,36). In a series of 43 patients, ERCP and stenting for PPF showed a success rate of 100.0% (20). Endoscopic concomitant biliary and pancreatic stenting has been shown to be effective in closing pancreatobiliary fistula (37).
Other endoscopic techniques have been tried with varying success in patients with AP-associated PFs. Endoscopic placement of a nasopancreatic drain allows for intermittent suction and repeat pancreatograms to confirm fistula closure without need of repeat ERCP. Once the inflammation/infection is controlled, drain can be internalized by placement of stent in PD (38,39). Endoscopic ultrasound (EUS) or interventional radiology–guided rendezvous ERCP in another viable alternative in cases where conventional ERCP fails to gain access to the PD (40,41). Endoscopic clips were used in a patient to successfully close multiple colonic fistulae as a consequence of AP (42). Few case reports have also suggested using low-dose irradiation of the pancreas for the treatment of internal PF (43,44). But, more data are needed before clips, and radiation therapy could be used routinely as a treatment for PF.
Despite major advancements in medical and nonsurgical interventions, a significant percentage of patients with PF require eventual surgical interventions. Surgery is aimed at managing the necrosis including debridement with external drainage. The crucial question in the management of PF is the timing of surgery because early pancreatic resection might not be possible in acute phase of fistula formation (16). Patients with necrotic bowel, massive bleeding from fistulous tract, or complex fistulas might need emergent/urgent surgical intervention (4,11). Most of the upper GI and PPF fistulas close spontaneously after sphincter pressure and obstruction are relieved by sphincterotomy and perhaps drainage or debridement of the collection (6,20,45). On the other hand, 50–70% of colonic fistulas require surgical interventions including bowel resection with ileostomy or colostomy (6,11,12). In patients with PF and disconnected PD syndrome, specific surgical approach depends on the character and size of pancreatic tail. A pancreatic tail <6 cm, especially associated with splenic vein thrombosis, will lead to distal pancreatectomy/splenectomy. A larger, nonatrophic pancreatic tail can be considered for Roux-en-Y pancreatiojejunostomy (7,16,20,46). A duct to mucosa anastomosis supersedes any consideration of anastomosis to a fistulous tract composed inherently of fibrous tissue (Figures 1–6).
Based on above discussion, we suggest diagnostic and management strategies for internal PFs as shown in Figure 7.
Spontaneous PF is a rare but challenging complication of AP, especially in patients with IPN. Male sex, alcohol abuse, patients with severe AP, and infected necrosis are at risk of developing internal PF. Early and aggressive fluid therapy for AP, antibiotics in cases of IPN, timely drainage of necrotic collection, and enteral nutrition can prevent PF formation. GI bleeding (60%) is the most common manifestation of pancreaticoenteric fistulae. A high index of suspicion is required in patients with pleuropancreatic fistula as the majority (65%–76%) usually present with respiratory symptoms and GI symptoms are presenting complaint in only 30% of patients. High amylase (>1,000 U/dL) and protein (>3 g/dL) in pleural fluid is pathognomic of PPF. CT scan is the initial test of choice because of rapid availability, but MRCP and ERCP have better sensitivity, and ERCP also offers therapeutic opportunities. PD should be assessed with ERCP/EUS or MRCP for leak or disruption because this could affect management strategies. However, ERCP with PD manipulation should be avoided in patients with complete ductal disruption in the setting of necrosis of neck or body of the pancreas. Certain complications of PF including hemorrhage, perforation, and sepsis could be life-threatening and need urgent and aggressive therapy. In nonurgent or chronic cases, management of internal PF involves control of sepsis, decreasing fistula output with somatostatin analogs (in PPF), and decreasing intraductal pressure with ERCP or EUS/interventional radiology–guided interventions. More than 60.0% of the internal PF close with medical and nonsurgical interventions. Colonic fistulae, medical refractory-PF, or with disconnected PD syndrome often require surgical interventions. In conclusion, internal PF is a complication of AP with IPN often with disruption or disconnection of PD. A multidisciplinary approach including internist, gastroenterologist, nutritionist, interventional radiology, and surgical team specialized in pancreatobiliary pathologies is required for the successful management of these fistulae.
CONFLICTS OF INTEREST
Guarantor of the article: Prabhleen Chahal, MD.
Specific author contributions: A.S., M.A., R.G., M.W., T.S., and P.C.: study concept. A.S., M.A., and R.G.: literature search. All authors: manuscript writing. M.W., T.S., and P.C.: critical revisions and supervision.
Financial support: None to report.
Potential competing interests: None to report.
1. Buter A, Imrie CW, Carter CR, et al. Dynamic nature of early organ dysfunction determines outcome in acute pancreatitis. Br J Surg 2002;89:298–302.
2. Petrov MS, Shanbhag S, Chakraborty M, et al. Organ failure and infection of pancreatic necrosis as determinants of mortality in patients with acute pancreatitis. Gastroenterology 2010;139:813–20.
3. Doberneck RC. Intestinal fistula complicating necrotizing pancreatitis. Am J Surg 1989;158:581–3; discussion 583–4.
4. Gupta R, Kulkarni A, Babu R, et al. Complications of percutaneous drainage in step-up approach for management of pancreatic necrosis: Experience of 10 years from a tertiary care center. J Gastrointest Surg 2020;24:598–609.
5. Hua Z, Su Y, Huang X, et al. Analysis of risk factors related to gastrointestinal fistula in patients with severe acute pancreatitis: A retrospective study of 344 cases in a single Chinese center. BMC Gastroenterol 2017;17:29.
6. Jiang W, Tong Z, Yang D, et al. Gastrointestinal fistulas in acute pancreatitis with infected pancreatic or peripancreatic necrosis: A 4-year single-center experience. Medicine (Baltimore) 2016;95:e3318.
7. Kaman L, Behera A, Singh R, et al. Internal pancreatic fistulas with pancreatic ascites and pancreatic pleural effusions: Recognition and management. ANZ J Surg 2001;71:221–5.
8. Fielding GA, McLatchie GR, Wilson C, et al. Acute pancreatitis and pancreatic fistula formation. Br J Surg 1989;76:1126–8.
9. Bassi C, Marchegiani G, Dervenis C, et al. The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after. Surgery 2017;161:584–91.
10. Tsiotos GG, Smith CD, Sarr MG. Incidence and management of pancreatic and enteric fistulas after surgical management of severe necrotizing pancreatitis. Arch Surg 1995;130:48–52.
11. Kochhar R, Jain K, Gupta V, et al. Fistulization in the GI tract in acute pancreatitis. Gastrointest Endosc 2012;75:436–40.
12. Mohamed SR, Siriwardena AK. Understanding the colonic complications of pancreatitis. Pancreatology 2008;8:153–8.
13. Machado NO. Pancreaticopleural fistula: Revisited. Diagn Ther Endosc 2012;2012:815476.
14. Wronski M, Slodkowski M, Cebulski W, et al. Optimizing management of pancreaticopleural fistulas. World J Gastroenterol 2011;17:4696–703.
15. Sasturkar SV, Gupta S, Thapar S, et al. Endoscopic management of pleural effusion caused by a pancreatic pleural fistula. J Postgrad Med 2020;66:206–8.
16. Kaiser J, Hackert T, Büchler MW. Management of pancreatic fistula in acute pancreatitis. In: The Pancreas: An Integrated Textbook of Basic Science, Medicine, and Surgery, 3rd edn. Wiley: Hoboken, NJ, 2009, pp 316–22.
17. Sikora SS, Khare R, Srikanth G, et al. External pancreatic fistula as a sequel to management of acute severe necrotizing pancreatitis. Dig Surg 2005;22:446–51; discussion 452.
18. Berne TV, Edmondson HA. Colonic fistulization due to pancreatitis. Am J Surg 1966;111:359–63.
19. Aswani Y, Hira P. Pancreaticopleural fistula: A review. JOP 2015;16:90–4.
20. Kord Valeshabad A, Acostamadiedo J, Xiao L, et al. Pancreaticopleural fistula: A review of imaging diagnosis and early endoscopic intervention. Case Rep Gastrointest Med 2018;2018:7589451.
21. De Waele JJ, Delrue L, Hoste EA, et al. Extrapancreatic inflammation on abdominal computed tomography as an early predictor of disease severity in acute pancreatitis: Evaluation of a new scoring system. Pancreas 2007;34:185–90.
22. Bollen TL, Singh VK, Maurer R, et al. A comparative evaluation of radiologic and clinical scoring systems in the early prediction of severity in acute pancreatitis. Am J Gastroenterol 2012;107:612–9.
23. Mortele KJ, Wiesner W, Intriere L, et al. A modified CT severity index for evaluating acute pancreatitis: Improved correlation with patient outcome. AJR Am J Roentgenol 2004;183:1261–5.
24. van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J Med 2010;362:1491–502.
25. Villena V, Perez V, Pozo F, et al. Amylase levels in pleural effusions: A consecutive unselected series of 841 patients. Chest 2002;121:470–4.
26. Timmerhuis HC, van Dijk SM, Verdonk RC, et al. Various modalities accurate in diagnosing a disrupted or disconnected pancreatic duct in acute pancreatitis: A systematic review. Dig Dis Sci 2021;66(5):1415–24.
27. Baron TH, DiMaio CJ, Wang AY, et al. American Gastroenterological Association clinical practice update: Management of pancreatic necrosis. Gastroenterology 2020;158:67–75.e1.
28. Singh N, Sharma B, Sharma M, et al. Evaluation of early enteral feeding through nasogastric and nasojejunal tube in severe acute pancreatitis: A noninferiority randomized controlled trial. Pancreas 2012;41:153–9.
29. Martineau P, Shwed JA, Denis R. Is octreotide a new hope for enterocutaneous and external pancreatic fistulas closure? Am J Surg 1996;172:386–95.
30. Rana SS, Sharma R, Kang M, et al. Natural course of low output external pancreatic fistula in patients with disconnected pancreatic duct syndrome following acute necrotising pancreatitis. Pancreatology 2020;20:177–81.
31. Schorn S, Vogel T, Demir IE, et al. Do somatostatin-analogues have the same impact on postoperative morbidity and pancreatic fistula in patients after pancreaticoduodenectomy and distal pancreatectomy? A systematic review with meta-analysis of randomized-controlled trials. Pancreatology 2020;20:1770–8.
32. Gans SL, van Westreenen HL, Kiewiet JJ, et al. Systematic review and meta-analysis of somatostatin analogues for the treatment of pancreatic fistula. Br J Surg 2012;99:754–60.
33. Ali T, Srinivasan N, Le V, et al. Pancreaticopleural fistula. Pancreas 2009;38:e26–31.
34. Oh YS, Edmundowicz SA, Jonnalagadda SS, et al. Pancreaticopleural fistula: Report of two cases and review of the literature. Dig Dis Sci 2006;51:1–6.
35. Boerma D, Rauws EA, van Gulik TM, et al. Endoscopic stent placement for pancreaticocutaneous fistula after surgical drainage of the pancreas. Br J Surg 2000;87:1506–9.
36. Lau ST, Simchuk EJ, Kozarek RA, et al. A pancreatic ductal leak should be sought to direct treatment in patients with acute pancreatitis. Am J Surg 2001;181:411–5.
37. Carrere C, Heyries L, Barthet M, et al. Biliopancreatic fistulas complicating pancreatic pseudocysts: A report of three cases demonstrated by endoscopic retrograde cholangiopancreatography. Endoscopy 2001;33:91–4.
38. Miyachi A, Kikuyama M, Matsubayashi Y, et al. Successful treatment of pancreaticopleural fistula by nasopancreatic drainage and endoscopic removal of pancreatic duct calculi: A case report. Gastrointest Endosc 2004;59:454–7.
39. Schoefl R, Haefner M, Pongratz S, et al. Endoscopic treatment of fistulas and abscesses in pancreatitis: Three case reports. Endoscopy 1996;28:776–9.
40. Irani S, Gluck M, Ross A, et al. Resolving external pancreatic fistulas in patients with disconnected pancreatic duct syndrome: Using rendezvous techniques to avoid surgery (with video). Gastrointest Endosc 2012;76:586–93 e1-3.
41. Mallery S, Matlock J, Freeman ML. EUS-guided rendezvous drainage of obstructed biliary and pancreatic ducts: Report of 6 cases. Gastrointest Endosc 2004;59:100–7.
42. Hwang SO, Lee TH, Park JW, et al. Endoscopic management of multiple colonic fistulae secondary to acute pancreatitis (with video). Gastrointest Endosc 2010;71:395–7; discussion 397.
43. Kavin H, Sobel JD, Dembo AJ. Pancreatic ascites treated by irradiation of pancreas. Br Med J 1971;2:503–4.
44. Morton RE, Deluca R, Reisman TN, et al. Pancreatic ascites: Successful treatment with pancreatic radiation. Am J Dig Dis 1976;21:333–6.
45. Halttunen J, Weckman L, Kemppainen E, et al. The endoscopic management of pancreatic fistulas. Surg Endosc 2005;19:559–62.
46. Jang JW, Kim MH, Oh D, et al. Factors and outcomes associated with pancreatic duct disruption in patients with acute necrotizing pancreatitis. Pancreatology 2016;16:958–65.