Tu1680 MAnagement of Patients With Pancreatic Cysts Using Integrated Molecular Pathology

Abstract

surgical resection had a diagnosis of malignancy. Conclusion: 1) Patients with pancreas mass lesions who have EUS FNA cytology results suspicious for malignancy have a 90% chance of subsequent malignancy confirmed. 2) Patients with pancreas mass lesions who have atypical EUS FNA cytology results have a 68% chance of subsequent malignancy confirmed. Furthermore, repeat EUS FNA is performed more often in patients with atypical cytology. 4) The percentages of patients who ultimately underwent a surgical operation were similar between the two groups. 5) Further research is needed to know the true risk of malignancy in indeterminate cytology groups. Tu1678 External Retrospective Validation of Needle-Based Confocal LASER Endomicroscopy (nCLE) Criteria for Pancreatic Cysts David L. Carr-Locke, Manoop S. Bhutani, Oleh Haluszka, Frank G. Gress, Kevin E. Woods, Bertrand Napoleon* Beth Israel, New York, NY; MD Anderson, Houston, TX; Temple University, Philadelphia, PA; Columbia University, New York, NY; Emory University, Atlanta, GA; Hopital Prive Jean Mermoz, Lyon, France Introduction: nCLE is an imaging technique, enabling microscopic observation of solid organs, in vivo and in real-time, during an EUS-FNA procedure. This technique could potentially provide useful information to establish a differential diagnosis between mucinous and non-mucinous neoplasms. A recent prospective multicenter French study, CONTACT (Needle-based confocal laser endomicroscopy (nCLE) for the diagnosis of pancreatic cysts: validation of the described criteria, Napoleon et al., UEGW 2014, OP 282), assessed the diagnostic performance of nCLE for the diagnosis of pancreatic cysts. In this study, the nCLE specificity for serous cystadenoma (SCA), mucinous cystic neoplasm (MCN), and pseudocyst were 100%, 95%, and 96% respectively. This US retrospective study aims at externally validating those results through a retrospective evaluation of the nCLE performance and the interobserver agreement for the characterization of pancreatic cysts. Materials and Methods: Five independent gastroenterologists conducted a blinded, individual, retrospective review of 31 nCLE sequences after a short training session on image interpretation delivered by an nCLE expert. Training consisted of a review of current nCLE criteria for each cyst type with corresponding sample sequences. The sequences then reviewed consisted of 13 serous cystadenomas, 6 mucinous cystadenomas, 5 IPMNs, and 7 pseudocysts. Their performance was compared to that of CONTACT reviewers. Results: Performance results are reported in table 1. Performance values obtained correlated with CONTACT results, especially for those sequences evaluated with “high confidence”. Interobserver agreement were kappa Z 0.71, 0.65, and 0.9 for SCA, MCN, and pseudocysts respectively. Conclusions: This study demonstrates that after brief training on nCLE image interpretation criteria for the characterization of pancreatic lesions, a high level of performance can be achieved for the characterization of pancreatic lesions. This suggests that accurate interpretation of nCLE images for pancreatic cysts can be learned rapidly, with a high level of confidence in 72.6% of the cases.