Pancreaticopericardial fistula: a rare complication of chronic pancreatitis

Abstract

Pancreaticopericardial fistula (PPF) is an extremely rare clinical problem encountered in patients with chronic pancreatitis. The diagnosis should be suspected if a patient presents with pericardial effusion on a background of chronic pancreatitis. Significantly raised amylase in the pericardial fluid offers an important clue for the diagnosis. CT is the initial imaging modality to look for pancreatic and pericardial changes. The therapeutic options include medical, endoscopic or surgical interventions. Medical and endoscopic therapies are the preferred modes of treatment while surgery is reserved for those who fail these measures.

Background

We report a case of pancreaticopericardial fistula (PPF) in a patient with chronic pancreatitis, which was managed successfully using endoscopic treatment.

Case presentation

A 27-year-old man, a known case of ethanol-related chronic pancreatitis, presented with a 2-week history of abdominal pain and New York Heart Association class 2 dyspnoea for 1 week. He was a diagnosed case of chronic pancreatitis for 1 year. He had been investigated at another facility, where MR cholangiopancreatography (MRCP) had been performed 1 week prior to the current presentation, which showed atrophied pancreatic parenchyma, pancreatic duct (PD) stricture in the neck with upstream dilation and a tiny heterogenous intensity inflammation in the tail of the pancreas extending into the subphrenic region and around the oesophagus. On admission, physical examination revealed blood pressure of 130/90 mm  Hg, pulse 120/min, temperature 100°F and oxygen saturation 94% on room air. The patient had decreased air entry in the left infrascapular region. Cardiac examination revealed muffled heart sounds and a pericardial rub. His abdomen was soft with neither organomegaly, a palpable lump nor tenderness, and with normal bowel sounds. Initial laboratory work up revealed haemoglobin 11.1 g/dL, total leucocyte 38 900/mm³, platelet count 150 000/mm³, and normal liver and kidney function tests. Serum amylase levels were 155 mg/dL (normal up to 110 mg/dL) and lipase levels were 300 mg/dL (normal up to 300 mg/dl). EKG was normal and chest X-ray showed a water bottle shaped heart with mild left sided pleural effusion. The patient was given intravenous fluids along with supportive treatment after which he reported of worsening dyspnoea. A two-dimensional (2D) echocardiography (ECHO) was performed, which was suggestive of mild to moderate pericardial effusion with strands in the pericardium (figure 1). He underwent a CT of the abdomen and thorax, which showed presence of an atrophied pancreas with ductal dilation and irregularity involving side branches, with a neck stricture and calcifications. There was also a cystic collection in the tail of the pancreas extending superiorly into the left subphrenic region curving anterior to the oesophagus to enter the diaphragmatic hiatus. This was communicating into the right paracardiac region with rim enhancing pericardial inflammatory collection suggestive of PPF (figure 2). The largest component of this collection in the left subdiaphragmatic region measured 4.7×3.7×9 cm. There was basal left sided pleural effusion. The cardiac team was consulted, but pericardial drainage was deferred in view of the fibrinoexudative nature of the effusion and as the patient did not have cardiac tamponade. Pleural fluid analysis showed transudative effusion with normal amylase content. In view of the PPF and PD stricture, and the collection around the tail of the pancreas, endoscopic retrograde cholangio ancreatography (ERCP) was planned. The pancreatogram showed a distal cut-off at the tail and extravasation of contrast above the diaphragm into the pericardium (figure 3). A 5 French 10 cm plastic stent was placed in PD. Post-stenting, the patient's abdominal pain and breathing difficulty gradually settled over a period of 24 h. His 2D ECHO after 10 days showed interval resolution of the pericardial effusion. A repeat CT after 2 weeks showed decreased subdiaphragmatic collection and pericardial fluid. The patient was subsequently discharged and stent removed after 2 months, during which a repeat pancreatogram showed no leak (figure 4). The patient is on follow-up and doing well.

Figure 1.

Echocardiography image showing pericardial effusion.

Figure 2.

CT (coronal section) image showing fluid collection extending across diaphragm to right paracardiac region.

Figure 3.

Pancreatogram showing leak in the pancreatic tail region with extension across the diaphragm to pericardium.

Figure 4.

Pancreatogram (taken 2 months later) showing absence of leak in pancreatic tail region.

Outcome and follow-up

A repeat echocardiography postendotherapy showed resolution of the pericardial effusion, repeat CT 2 weeks later showed resolution of the subdiaphragmatic collection. Stent removal and subsequent pancreatogram performed after 2 months showed healing of the PPF.

Discussion

Chronic pancreatitis can have associated pseudocysts that are mostly located within the pancreas. Ductal disruption may cause the collections of enzyme rich secretions reaching the pleura, mediastinum and pericardium. PPF is a type of internal fistula, in which pancreatic secretions drain directly into the pericardial cavity. Mediastinal extension of pseudocysts along with pericardial effusion is reported in the literature.¹ In this case report, we describe a most unusual complication of chronic pancreatitis in the form of PPF, which was successfully managed by endotherapy.

A pseudocyst or acute fluid collection is a collection of pancreatic secretions, blood and cellular debris, due to disruption of the PD. Chronic pancreatitis leakage into the serosal cavity causing ascites and pleural effusions occurs in 3–5% of patients.¹ The proposed mechanisms can be indirect pseudocyst communication or directly by duct disruption in the form of fistulae.

Although there have been reports of mediastinal extension of pseudocysts, to the best of our knowledge, very few cases of PPFs have been reported in the medical literature.^1–4 Earlier, it was believed that pancreatic enzymes induce chemical pericarditis and pleurisy. If the communication occurs anteriorly it may cause pancreatic ascites whereas retroperitoneal duct disruption may cause mediastinal extension. In the mediastinum, it may present either as a pseudocyst or pancreaticopleural fistula, or PPF. For the management of PPF, pericardial drainage has been tried previously with the need for pericardial and pancreatic drains.⁵ Surgical management for those not benefiting after 2–4 weeks of conservative and endoscopic therapy is also described. PPFs may not respond to PD stenting and may portend a poor prognosis, hence a rapid and accurate diagnosis of this life-threatening complication is crucial.

The clinical features are often variable, and depend on the location and size of the communication. Patients may present with dyspnoea, chest pain, syncope or cardiogenic shock. Abdominal symptoms may or may not be present. Our patient presented with dyspnoea and abdominal pain.

Clinically, pericardial effusion with a high fluid amylase level in a patient with pancreatitis suggests a diagnosis of PPF. A direct demonstration of this fistulous communication is difficult; however, multiple diagnostic modalities may be used to establish the diagnosis. CT may often be suggestive, especially if a pseudocyst or peripancreatic fluid collection is observed. MRCP delineates the PD and may identify the fistulous tract in a good number of patients.

ERCP is an important modality in PPF as it provides both diagnostic as well as therapeutic options.^{6 7} In our patient, clinical suspicion of PPF and demonstration of tract on CT imaging, established the diagnosis.

Though there are no established standard guidelines for the management of PPF, based on case reports and small observational studies, the initial management may include nutritional support and octreotide to inhibit pancreatic secretions. This management may suffice for small leaks but the most common recommended therapy is endoscopic PD stenting.^8–10 Optimising nutritional status and limiting oral intake may be effective with small leaks. Octreotide, a potent inhibitor of pancreatic secretion, has also been used in some small studies, with efficacy. However, endoscopic stenting by ERCP is the most effective treatment.^{9 10} Location of the duct disruption could be the most important predictor of who would benefit from stenting for PD leak. Successful endotherapy depends on passing the guidewire across the leak and deploying a stent of appropriate length.¹¹ If the leak persists after 2–4 weeks of conservative and endoscopic therapy, surgery may be required depending on the site of the leak. In a case series presented by Kozarek et al,¹² of 18 patients having PD leak who underwent pancreatic endotherapy, seven required surgery due to recurrent collections or ongoing pain. In another retrospective study of eight patients with PPF who underwent PD stenting, only a single case responded to endotherapy.¹³ In a case series, covered metallic pancreatic stent placement was used to manage PD leaks.¹⁴ Another case report of PPF complicating acute pancreatitis has been published, the case was initially managed with octreotide and intrapericardial triamcinolone, but the pericardial effusion recurred. Finally, ERCP with PD stenting to bridge the leak was successful.¹⁵ There is a case report of PPF complicating chronic pancreatitis, where an elective surgical procedure such as panceaticojejunostomy was carried out with successful outcome.²

In conclusion, we present a rare case of CP with PPF that was managed with ERCP and stenting.

Learning points.

• Pancreaticopericardial fistula is a rare complication of chronic pancreatitis.

• Early diagnosis is necessary to prevent formation of fibrinous adhesions in the pericardium.

• Successful endotherapy can result in resolution of pancreaticopericardial fistulae.