Mural nodule of 10 mm or larger as predictor of malignancy for intraductal papillary mucinous neoplasm of the pancreas: Pathological and radiological evaluations.

BACKGROUND

We had previously reported that mural nodule (MN) ≥10 mm was optimal predictor of malignancy for intraductal papillary mucinous neoplasm (IPMN). However, little is known about its microscopic findings and imaging detectability.

METHODS

Medical records and resected specimens of consecutive patients with IPMNs harboring MN ≥ 10 mm were reviewed. Imaging detectability was determined on reports basis. Malignant IPMNs (noninvasive + invasive carcinomas) were microscopically classified according to localization of high-grade dysplasia (HGD) within MN.

RESULTS

Thirty-six patients were included. Imaging detectability of MN ≥ 10 mm in CT, MRI, US and EUS were 64%, 68%, 89%, and 97%, respectively. Thirty-three (92%) IPMNs were histologically diagnosed as malignant. Thirty percent of malignant IPMNs were classified into "diffuse HGD within MN", 40% into "focal HGD within MN", and 30% into "HGD outside MN", in which HGD was not located within MN but in low papillary epithelia around MN. Overall sensitivity of pancreatic juice cytology was calculated as 58%, and for "diffuse HGD within MN", "focal HGD within MN", and "HGD outside MN" as 80%, 62%, and 30%, respectively (p = 0.0237). Univariate-analysis showed localization of HGD within MN was associated with true positive cytology (OR = 5.33, p = 0.043).

CONCLUSIONS

Detectability of MN ≥ 10 mm is excellent in US and EUS. Although HGD is observed within MN in 70% of malignant IPMNs, HGD is located only in low papillary epithelia around MN in the remaining 30%, in which sensitivity of pancreatic juice cytology is shown to be inadequate.