A prospective comparison of the yield of EUS in primary vs. recurrent idiopathic acute pancreatitis

Abstract

Background: It is uncertain whether EUS should be performed after a single episode of idiopathic pancreatitis vs. recurrent episodes or if clinical factors can predict positive EUS findings.

Methods: Consecutive patients with a single episode of idiopathic pancreatitis or with recurrent episodes underwent EUS (with analysis of bile for bilirubinate and cholesterol crystals, when possible). The diagnostic yield was compared for patients with a single episode of idiopathic pancreatitis and recurrent episodes (stratified by cholecystectomy status). Predictors of positive EUS findings were sought. EUS was considered "positive" if it identified any possible cause of pancreatitis other than chronic pancreatitis.

Results: A total of 370 patients were studied (246 no-cholecystectomy group [134 single episode of idiopathic pancreatitis, 112 recurrent episodes] and 124 post-cholecystectomy group [67 single episode of idiopathic pancreatitis, 57 recurrent episodes]). Overall, EUS yielded a positive finding in 29.2%. For patients in the no-cholecystectomy group, positive EUS findings were not significantly more frequent in those with a single episode of idiopathic pancreatitis vs. those with recurrent episodes (31.3% vs. 32.1%; p = 0.89). In the post-cholecystectomy group, the yield was not significantly different for single episode of idiopathic pancreatitis (29.9%) vs. recurrent episodes (17.5%) ( p = 0.15). Chronic pancreatitis was the only abnormality identified in 30.9% of patients in the no-cholecystectomy group vs. 26.6% of those in the post-cholecystectomy group ( p = 0.24). It was the most common abnormality found in all 4 subgroups (range 16.4%-42.0%) and was approximately twice as frequent in patients with recurrent episodes vs. a single episode of idiopathic pancreatitis (no-cholecystectomy: 42.0% vs. 21.6%, p = 0.0008; post-cholecystectomy: 38.6% vs. 16.4%, p = 0.008). Analysis of bile revealed crystals in 38/80 (47.5%) patients in whom it could be performed. Patients with positive EUS findings tended to be older.

Conclusions: In patients with idiopathic pancreatitis, the yield of EUS is not significantly different after an initial attack or after recurrent attacks. Therefore, it is reasonable to perform EUS after an initial attack of idiopathic acute pancreatitis, especially in older patients.