Comparison of endoscopic ultrasonography and multidetector computed tomography for detecting and staging pancreatic cancer

Background: Accurate preoperative detection and staging of pancreatic cancer may identify patients with locoregional disease that is amenable to surgical resection. Objective: To compare endoscopic ultrasonography and multidetector computed tomography (CT) for the detection, staging, and resectability of known or suspected locoregional pancreatic cancer. Design: Prospective, observational, cohort study. Setting: Single, tertiary referral hospital in Indianapolis, Indiana. Patients: 120 participants with known or suspected locoregional pancreatic cancer. Interventions: Endoscopic ultrasonography followed by multidetector CT was performed in all patients. Patients with known or suspected pancreatic cancer deemed potentially resectable by 1 or both tests were considered for surgery. Measurements: Detection, staging, and resectability of pancreatic cancer. Surgically resected pancreatic cancer with negative microscopic histologic margins was considered resectable. Results: Of 120 patients enrolled, 104 (87%) underwent endoscopic ultrasonography and CT. Of the 80 patients with pancreatic cancer, 27 (34%) were managed nonoperatively, and 53 (66%) treated surgically had resectable (n = 25) or unresectable (n = 28) cancer. For the 80 patients with cancer, the sensitivity of endoscopic ultrasonography (98% [95% CI, 91% to 100%]) for detecting a pancreatic mass was greater than that of CT (86% [CI, 77% to 93%]; P = 0.012). For the 53 surgical patients, endoscopic ultrasonography was superior to CT for tumor staging accuracy (67% vs. 41%; P < 0.001) but equivalent for nodal staging accuracy (44% vs. 47%; P > 0.2). Of the 25 resectable pancreatic tumors in patients recommended for surgery, endoscopic ultrasonography and CT correctly identified 88% and 92%, respectively, as resectable. Of the 28 unresectable pancreatic tumors in patients recommended for surgery, endoscopic ultrasonography and CT correctly identified 68% and 64%, respectively, as unresectable. Limitations: Radiologists who read the scans and endosonographers were not blinded to previous radiographic information. Because of the modest sample size, CIs of the sensitivity estimates were sometimes wide. Conclusion: Compared with multidetector CT, endoscopic ultrasonography is superior for tumor detection and staging but similar for nodal staging and resectability of preoperatively suspected nonmetastatic pancreatic cancer.