Comparison of Current EUS Criteria to the Proposed “Rosemont Criteria” in Patients with An Established Diagnosis of Chronic Pancreatitis

Abstract

Introduction: EUS has been proposed as the most accurate test to diagnose CP, and ≥4 criteria are accepted as diagnostic. This assigns equal importance to all endoscopic changes. The Rosemont criteria were proposed to be a more reliable grading system for CP. Aim: Compare the previous EUS criteria and the newly proposed Rosemont criteria in patients with established CP. Methods: The endoscopy database was searched from 2006 to 2008 for patients with an established diagnosis of CP who had a EUS. The current EUS criteria were: parenchymal lobularity, hyperechoic foci, hyperechoic strands, cysts, main pancreas duct (MPD) dilatation, MPR irregularity, hyperechoic duct walls, visible side branches and calcifications. More than 4 of these criteria were considered diagnostic for CP. The Rosemont criteria were: Major A: hyperechoic foci with shadowing or MPD calculi; Major B: lobularity with honeycombing; Minor: cysts, dilated duct >3.5 mm, irregular PD contour, dilated side branches > 1 mm, hyperechoic duct wall, strands, non-shadowing hyperechoic foci, lobularity with non-contiguous lobules and where categorized as most consistent, suggestive, or indeterminate for chronic pancreatitis. The diagnosis of CP was established by combining clinical criteria (exocrine and/or endocrine insufficiency), with imaging (CT and/or MRI in all), ERCP (Cambridge III or IV in 6), and histology (pancreatectomy 1, tail resection 1, Beger resection 2, Whipple resection 3). Results: We evaluated 23 patients (M=17, F=6) with a mean age of 51.3 years. The etiology of CP was idiopathic (11), alcohol (10), hypertriglyceridemia (2). Seventeen patients had ≥ 4 current EUS criteria, and 18/23 fulfilled major Rosemont criteria for CP. The sensitivity of the current EUS criteria was 79%, and 82% using Rosemont criteria. PPV was 74% and 78%, respectively. In patients with histologically confirmed CP (n=7) the sensitivity of EUS was 88% for current and 100% for Rosemont criteria, respectively. The PPV was 86% for current criteria and 100% for Rosemont criteria. Conclusions: The diagnosis of CP is difficult to establish and the actual performance of diagnostic tests is speculative because tissue is usually not obtained. In an attempt to normalize EUS criteria for CP the Rosemont criteria establish more restrictive parameters for the diagnosis of CP. In a group of patients with established CP the current EUS criteria and the newly proposed Rosemont criteria were equally efficacious at establishing the diagnosis of CP. Overall sensitivity increased for both when compared to the gold standard of histology. Performance of the criteria in patients with less severe disease needs to be studied.