INTRODUCTION: Recurrent unilateral pleural effusion in a patient with chronic pancreatitis should heighten suspicion of a pancreatico-pleural fistula. The predominance of pulmonary findings with minimal clues to pancreas as the initiator of the effusion make this diagnosis challenging.
CASE PRESENTATION: A 54-year old male with medical history of chronic alcoholic pancreatitis presented to his primary physician’s office with increasing shortness of breath and severe weakness. Since he had decreased breath sounds on left side, he was sent for a chest radiograph, which confirmed a large left pleural effusion. Pleural fluid analysis confirmed yellow exudative effusion, but no evidence of infection or malignancy. Pleural fluid analysis showed pH 8.0, protein 2.0 g/dL, LDH 8603 U/L, but no amylase or lipase levels were drawn. Despite a video-assisted thoracoscopic surgical decortication and two tube thoracostomy procedures within next nine months, he developed a loculated left effusion. He was admitted to outside hospital with acute pancreatitis, and soon developed multi-organ failure requiring ventilation, ionotropic support and dialysis. CT chest confirmed large left sided effusion with patchy airspace consolidation. Intravenous antibiotics were followed by chest tube insertion and drainage. He improved clinically with supportive measures, and was weaned off the ventilator. In the next few days, his effusion became larger and he developed a hydro-pneumothorax as a complication of multiple thoracic interventions. On transfer to our facility, he had decreased left sided breath sounds on auscultation. CBC and electrolytes were in normal range but amylase (243 U/L) was elevated. Malnutrition was confirmed with albumin (1.9 g/dL) and pre-albumin (9.0 mg/dL). Thorax tomography confirmed large loculated hydro-pneumothorax. No esophageal perforation was seen. Ultrasound guided thoracocentesis yielded grayish pleural fluid with elevated amylase (19,987 U/L) and lipase (16,125 U/L). All cultures remained negative. Pleural fluid analysis was significant for pH 7.25, triglyceride 41 mg/dL, protein - 1.6 g/dL, LDH - 3818 IU/L, cholesterol - 9 mg/dL, chylomicrons- negative, and Cytology negative. A distinct fluid collection intimately related to the tail of pancreas with a fistulous tract piercing the diaphragm was seen on MRI abdomen with MRCP. Initial non-surgical treatment included endoscopic retrograde cholangiopancreatography with pancreatic stent and octreotide. Total parenteral nutrition was initiated. Three months later, conservative measures failed to close the fistula. The fistulous tract was surgically removed with distal pancreatectomy.
DISCUSSION: This case emphasizes the importance of performing additional testing on pleural fluid if routine laboratory data fail to yield a diagnosis. Patients with chronic alcohol use may have minimal abdominal complaints except mild pain and malnutrition.
CONCLUSIONS: A high index of suspicion is required, in those with a large (usually left-sided) pleural effusion and history of alcohol abuse or chronic pancreatitis, with addition of amylase and lipase levels to pleural fluid analysis. Other causes of high amylase in pleural fluid include acute pancreatitis (small, transient, usually left sided), esophageal perforation, and certain tumors.
Reference #1 Rockey DC, Cello JP. Pancreaticopleural fistula. Report of 7 patients and review of the literature. Medicine (Baltimore). 1990;69:332-344
Reference #2 Materne R, Vranckx P, Pauls C, et al. Pancreaticopleural fistula: diagnosis with magnetic resonance pancreatography. Chest. 2000;117:912-914
Reference #3 Poddar U, Kochhar R, Singh A, et al. Pancreatcopleural fistula: successful treatment with octreotide. Indian J Gastroenterol. 1995;14:145-146 Burgess NA, Moore HE, Williams JO, Lewis MH. A review of pancreatico-pleural fistula in pancreatitis and its management. HPB Surg. 1992;5:79-86 Saeed ZA, Ramirez FC, Hepps KS. Endoscopic stent placement for internal and external pancreatic fistulas. Gastroenterology. 1993;105:1213-1217
DISCLOSURE: The following authors have nothing to disclose: Vipul Kumar, Shashank Jain, Anthony Donato
No Product/Research Disclosure Information