Endoscopic ultrasound guided fine needle aspiration in the diagnosis of chronic pancreatitis (CP)

Differentiation of malignant disease from chronic pancreatitis (CP) in patients with abnormal pancreatic findings on CT or ERCP is often difficult. This study was designed to evaluate the role of endoscopic ultrasound (EUS) guided fine needle aspiration (FNA) in a group of patients with abnormalities of the pancreas on radiographic imaging studies. Methods: The EUS Database was used to identify patients with suspected pancreatic cancer who had EUS guided FNA with cytology suggestive of CP. The cytologic diagnosis of CP was defined as the presence of reactive acinar cells, fibrosis, and chronic inflammation . All patients had ERCP and/or CT prior to EUS with FNA. Final determination of benign vs malignant disease was made by surgical pathology and/or long-term clinical follow-up for clinical status. Results: Twenty-eight patients were identified; the mean age was 59 years (range: 27-84); 15 (54%) were male. Twenty-five patients had CT scan, 23 had ERCP, and 20 (71%) had both CT and ERCP. Twenty-four of the 25 patients with CT scan (96%) had an abnormality in the pancreas: 15 enlarged head of pancreas, 6 pancreatic mass, and 3 "double duct sign" (DDS, or simultaneous stenosis of the adjacent pancreatic duct (PD) and common bile duct). Nineteen of the 23 ERCPs were abnormal: 11 (48%) DOS, 4 side branch changes, 2 isolated PD strictures, 1 pancreas divisum, and 1 pancreatic cyst. Overall, 12/28 (43%) had a DOS on CT or ERCP, suggesting pancreatic cancer. The FNA cytology on all 28 was consistent with CP. There were no cases of malignant cytology in this group. Clinical or surgical follow-up was available on all 28 patients. Seven (25%) went to surgery; four of these had pancreatic adenocarcinoma and 3 had chronic pancreatitis. Of the 21 patients who did not go to surgery, the mean follow-up was 16 months (range: 7 - 24 months); all are alive and without evidence of malignancy. Sensitivity,specificity, positive and negative predictive (PPV & NPV) values of the DOS to detect pancreatic cancer were calculated. These were 75%, 63%, 25% and 94%, respectively. The PPV of EUS with FNA to detect CP was 82%. Conclusions: (1) The PPV of EUS with FNA for the identification of patients with CP is 82%. (2) The NPV of the DOS for the detection of pancreatic cancer in this population is 94%.

Duke Scholars

Author Jane Elizabeth Onken Medicine, Gastroenterology