Expansion of epigenetic alterations in EFEMP1 promoter predicts malignant formation in pancreatobiliary intraductal papillary mucinous neoplasms

Extensive methylation of the EFEMP1 gene promoter can discriminate invasive from benign IPMNs with superior accuracy owing to their stepwise accumulation of tumor progression.

A sample of 65 IPMNs and 30 normal pancreatic tissues was analyzed. IPMNs were divided into the following three subsets according to pathological findings: 31 with low-grade dysplasia (low grade), 11 with high-grade dysplasia (high grade), and 23 with associated invasive carcinoma (invasive Ca). Mutations in the KRAS or GNAS genes were analyzed by Sanger sequencing, and methylation status of two discrete regions within the EFEMP1 promoter, namely region 1 and region 2, was analyzed by bisulfite sequencing and fluorescent high-sensitive assay for bisulfite DNA (Hi-SA). Expression status of EFEMP1 was investigated by immunohistochemistry (IHC).

Although limited understanding exists for the presence of specific genetic mutations and aberrantly methylated genes in pancreatobiliary intraductal papillary mucinous neoplasms (IPMNs), the fundamental understanding of the dynamics of methylation expansion across CpG dinucleotides in specific gene promoters during carcinogenesis remains unexplored. Expansion of DNA methylation in some gene promoter regions, such as EFEMP1, one of the fibulin family, with tumor progression has been reported in several malignancies. We hypothesized that DNA hypermethylation in EFEMP1 promoter would expand with the tumor grade of IPMN.

KRAS mutations were detected in 39, 55, and 70 % of low-grade, high-grade, and invasive Ca, respectively. GNAS mutations were observed in 32, 55, and 22 % of low-grade, high-grade, and invasive Ca, respectively. The methylation of individual regions (region 1 or 2) in the EFEMP1 promoter was observed in 84, 91, and 87 % of low-grade, high-grade, and invasive Ca, respectively. However, simultaneous methylation of both regions (extensive methylation) was exclusively detected in 35 % of invasive Ca (p = 0.001) and five of eight IPMNs (63 %) with extensive methylation, whereas 20 of 57 (35.1 %) tumors of unmethylation or partial methylation of the EFEMP1 promoter region showed weak staining EFEMP1 in extracellular matrix (p = 0.422). In addition, extensive EFEMP1 methylation was particularly present in malignant tumors without GNAS mutations and associated with disease-free survival of patients with IPMNs (p < 0.0001). Conclusions: Extensive methylation of the EFEMP1 gene promoter can discriminate invasive from benign IPMNs with superior accuracy owing to their stepwise accumulation of tumor progression.