OBJECTIVES: Patients with chronic pancreatitis are managed by extracorporeal shock wave lithotripsy (ESWL) for pancreatic stones. Stone density on noncontrast computed tomography (NCCT) is used to predict stone composition and fragility, but no report investigating the pancreatic stone density has been documented. We aimed to assess the usefulness of pancreatic stone density on NCCT in predicting the ESWL outcome. METHODS: We evaluated 128 consecutive patients with pancreatic stones who underwent ESWL. Pancreatic stone density on NCCT was measured in Hounsfield units. Patients with complete stone removal were compared with those with incomplete stone removal. Patient characteristics; imaging findings, including stone density; and interventions were evaluated as potential predictors. The optimal cutoff value of variables to differentiate complete removal from incomplete removal was determined by receiver operating characteristic analysis. RESULTS: Complete stone removal was achieved in 66 patients (51.6%). When the density threshold was set at 820.5 Hounsfield units, complete stone removal was achieved in 52 patients (78.8%) with lower-density stones. In multivariate analysis, single stone (P = 0.007) and lower-density stone (P < 0.001) revealed significant association with complete stone removal. CONCLUSIONS: The measurement of pancreatic stone density before therapy can help predict therapeutic outcomes.
OBJECTIVES:Patients with chronic pancreatitis are managed by extracorporeal shock wave lithotripsy (ESWL) for pancreatic stones. Stone density on noncontrast computed tomography (NCCT) is used to predict stone composition and fragility, but no report investigating the pancreatic stone density has been documented. We aimed to assess the usefulness of pancreatic stone density on NCCT in predicting the ESWL outcome. METHODS: We evaluated 128 consecutive patients with pancreatic stones who underwent ESWL. Pancreatic stone density on NCCT was measured in Hounsfield units. Patients with complete stone removal were compared with those with incomplete stone removal. Patient characteristics; imaging findings, including stone density; and interventions were evaluated as potential predictors. The optimal cutoff value of variables to differentiate complete removal from incomplete removal was determined by receiver operating characteristic analysis. RESULTS: Complete stone removal was achieved in 66 patients (51.6%). When the density threshold was set at 820.5 Hounsfield units, complete stone removal was achieved in 52 patients (78.8%) with lower-density stones. In multivariate analysis, single stone (P = 0.007) and lower-density stone (P < 0.001) revealed significant association with complete stone removal. CONCLUSIONS: The measurement of pancreatic stone density before therapy can help predict therapeutic outcomes.