Background:Currently, the role of EUS FNA in pancreatic neuroendocrine tumors (pNETs) is primarily for tumor detection and tissue diagnosis. However, few data exist regarding the comparative features of CT to EUS FNA.Aims: In patients undergoing evaluation of a suspected pNET, to compare EUS FNA to pancreatic protocol CT in terms of: 1.) tumor detection, and 2.) disease extent as it potentially impacts the surgical approach and extent of resection.Methods: From a prospectively maintained EUS database, all patients undergoing pNET staging FNA from 01/01/99 to 10/01/08 were identified. Clinical, radiologic, EUS, cytologic, and surgical data were abstracted and analyzed.Results: EUS FNA was performed to confirm and stage 108 patients (56.8 ±13.9 years, 52.8% female) with suspected pNETs, n=96 (88.9%) of which had a CT within 30 days of the EUS exam. CT identified a pNET in only 47 (49%) patients. In the remaining patients, CT was interpreted as normal in 15 (15.6%), inconclusive in 14 (14.6%), non-specific pancreatic cystic disease in 11 (11.5%), and misinterpreted the findings as adenocarcinoma in 9 (9.4%) patients, respectively. EUS more often identified multiple lesions, (p=0.008) and EUS and CT findings were similar in terms of lesion size, the presence of a cystic component and in the determination of locally advanced or metastatic disease. EUS detected multiple lesions in 19/108 (17.6%) patients among whom CT detected only one lesion. Among this group, the additional lesion(s) identified by EUS were located remote from the lesion seen with CT in 12% patients, thereby potentially altering the surgical approach or extent of resection. In an additional 15 patients, in whom EUS FNA identified a pNET, the CT was negative with no pathology detected. Among this group, the EUS FNA identified lesions were ≤2cm in 14 (93.3%), were located within the body/tail in 9 (60%), had multiple lesions in 4 (26.7%), a cystic component in 1 (6.7%) and all 15 (100%) had locally confined disease. Thus, collectively in 27 of the 108 (25%) patients the EUS FNA findings potentially altered patient management.Conclusion: Our data suggest helical CT identifies fewer than 50% of patients who have a pNET that is detected with EUS and confirmed with FNA. A subgroup of patients may be misdiagnosed by CT as having a ductal carcinoma. Additionally, the findings of EUS FNA may alter the clinical care of approximately 25% of patients through the identification of otherwise undetected lesions by CT and/or through the visualization of multiple remote lesions, thereby potentially altering the surgical approach or extent of resection.