Appropriate indications of initial endoscopic ultrasound evaluation for detecting mural nodules in branch duct intraductal papillary mucinous neoplasms of the pancreas.

OBJECTIVE: The presence of mural nodules is a strong predictor of malignancy in patients with branch-duct intraductal papillary mucinous neoplasms (BD-IPMNs). Endoscopic ultrasound (EUS) is the most sensitive imaging for the detection of mural nodules. The aim was to evaluate whether initial EUS evaluation is necessary to detect mural nodules in all patients with BD-IPMNs.

MATERIAL AND METHODS: We reviewed retrospectively the medical records of all 104 patients with BD-IPMNs diagnosed by EUS from January 2008 to December 2011. Of the 104 patients, 13 (12.5%) had mural nodules (MN-positive group) and 91 (87.5%) did not (MN-negative group).

RESULTS: Patients in the MN-positive group were significantly older (69 years vs. 61 years, p = 0.013), had larger-sized cysts on EUS (24.1 ± 8.3 mm vs. 16.2 ± 8.0 mm, p = 0.001), a higher rate of main pancreatic duct dilatation (30.8% vs. 3.3%, p = 0.004), and higher serum CA 19-9 concentrations (67.8 ± 109.4 U/mL vs. 17.5 ± 27.0 U/mL, p = 0.012), than patients in the MN-negative group. Cyst sizes measured by EUS, CT, and MRCP did not differ significantly. Multivariate logistic regression analysis showed that cysts ≥16 mm in size on CT (odds ratio [OR], 9.84; 95% confidential interval [CI], 1.08-89.93; p = 0.043) and main pancreatic duct dilatation (OR, 12.65; 95% CI, 1.62-98.99; p = 0.016) were independent predictors of mural nodules.

CONCLUSIONS: Patients with BD-IPMNs, cyst size ≥16 mm, and main pancreatic duct dilatation should be further evaluated by EUS for the presence of mural nodules.