Prospective Comparison of EUS, ERCP, and EUS-Guided True Cut Biopsy (EUS-TCB) for Suspected Chronic Pancreatitis

Abstract

Background: Without histology, the diagnosis of chronic pancreatitis (CP) is made by clinical presentation and radiographic data. While EUS-FNA may be used to aid the diagnosis, the role of cytology for this indication remains debatable. Aims: 1) determine the utility and safety profile of a novel EUS-guided tru-cut biopsy (EUS-TCB) device (Quick-Core, Wilson-Cook Medical, Inc.) for the diagnosis of suspected chronic pancreatitis; 2) Compare EUS and ERCP to each other and individually to EUS-TCB for suspected CP. Methods: Inclusion criteria: Subjects between 18-60 with >3 months of upper abdominal pain and suspected CP (regardless of severity). Exclusion criteria: prior pancreatic or upper GI tract surgery, thrombocytopenia, anemia, coagulopathy, CAD, pancreatic cancer or cyst, or acute pancreatitis or pancreatic stent within the previous 3 months. Subjects with ≥3 EUS criteria for CP within the body and neck on radial EUS underwent attempted transgastric EUS-TCB followed within one week by ERCP. The endoscopist performing ERCP was blinded to the EUS results. All pathology was examined by one expert pathologist. The severity of CP was stratified by Cambridge classification for ERCP and by the number of ductal/parenchymal criteria present at EUS (normal: 0; equivocal: 1-2, mild: 3-4; moderate: 5-6; severe >6). Agreement between EUS and ERCP was evaluated by a weighted kappa statistic. Results: Of 29 subjects screened, 18 were enrolled (11 male, 18 Caucasian; mean age: 42), all of whom underwent EUS. Of these, 9 had ≥3 EUS criteria and 8 underwent attempted EUS-TCB. One subject with calcific pancreatitis could not be safely biopsied. EUS-TCB results: normal pancreas (n=5; mean biopsy length: 3.4 mm), nondiagnostic (n=2), device malfunction (n=1). No biopsies were diagnostic for CP. Complications from EUS-TCB: acute pancreatitis (n=1) and abdominal pain without pancreatitis (n=1), both of whom were discharged after a one day hospitalization. Five refused ERCP following EUS; the remaining 13 underwent ERCP without complications. Of the 5 with normal histology by EUS-TCB, ERCP was normal (n=1) or showed mild (n=3) or moderate CP (n=1). Agreement between EUS and diagnostic ERCP was moderate (kappa: 0.55, 95% CI: 0.24-0.86). Conclusion: In this preliminary study, EUS-TCB was relatively safe but did not demonstrate histologic CP with clinically suspected early disease. Agreement between EUS and ERCP was moderate. This study was supported by a grant from Wilson-Cook Medical, Inc.