New EUS Criteria of Chronic Pancreatitis: Is EUS Over-Diagnosing Chronic Pancreatitis?

Abstract

Background and Aim: The diagnosis of chronic pancreatitis (CP) at an early stage is still a great challenge. The gold standard of CP has been either endoscopic retrograde pancreatography (ERP) and/or function tests. ERP, however, only can evaluate pancreatic ductal system, not parenchyma. Recently, endoscopic ultrasonography (EUS) is considered useful modality for the diagnosis of the CP because of its ability to detect pancreatic parenchymal abnormalities. The findings of EUS cannot be confirmed by other modalities. It is the advantage, whereas it's just the weakness of EUS. Among patients with CP, EUS may reveal abnormal parenchyma in the form of hyperechoic changes, such as hyperechoic foci and strands. However, since these findings sometimes identified in patient with normal pancreatic duct on ERP, it will be confused whether it may diagnose early CP or not. To prevent the over-diagnosis, we tried taking ‘fine-reticular pattern (FRP)’ in EUS criteria of CP as normal pancreatic parenchymal pattern. The aim of this study was to investigate the implication of hyperechoic abnormalities and FRP on EUS in normal ERP. Patients and Methods: One hundred ten patients who were undergone both EUS and ERP within 2 months were enrolled, excluding in patient with pancreatic cancer, IPMN and lower biliary ductal cancer. CP was defined using the Cambridge classification of ERP findings. On EUS images, we evaluated FRP as normal pancreatic parenchyma, and hyperechoic foci/strands as hyperechoic abnormalities. Then we gave 1 point to each criterion of EUS criteria and -1 point when FRP was seen, and regarded 2 more points as CP in ERP normal patients. Results: Normal pancreatic duct on ERP was identified in 86 patients. FRP was seen in 76 patients who had normal ERP (p < 0.01). On the other, in 43 patients who had no EUS abnormalities, all patients were identified FRP. Moreover, in 22 patients who had no FRP on EUS, all patients were identified parenchymal abnormalities. These results strongly suggested FRP indicated normal pancreatic parenchyma. On the basis of above findings, EUS abnormalities on normal ERP were investigated. Hyperechoic abnormalities on EUS were seen in 46 patients with normal ERP. In these 46 patients, FRP on EUS was seen in 36 patients. 57 patients were diagnosed as normal in conventional EUS criteria and 75 were as normal in criteria added FRP. The latter criteria was closer to ERP (p = 0.001). Conclusion: Hyperechoic EUS abnormalities were not necessarily mean that presence of early chronic pancreatitis. New EUS criteria of CP which is added a FRP criterion may be useful for over-diagnosis especially at an early stage of CP.