Molecular signatures of mRNAs and miRNAs as prognostic biomarkers in pancreatobiliary and intestinal types of periampullary adenocarcinomas

•Periampullary carcinomas are distinct at level of histology rather than of origin.•Molecular findings stress the importance of stratifying tumors according to subtype.•Prognostic mRNAs and miRNAs are linked to the two histological subtypes.•mRNAs and miRNAs expression of subtypes are linked to specific enriched pathways.•Ten miRNA families are differentially expressed in the two subtypes.

Periampullary adenocarcinomas include four anatomical sites of origin (the pancreatic duct, bile duct, ampulla and duodenum) and most of them fall into two histological subgroups (pancreatobiliary and intestinal). Determining the exact origin of the tumor is sometimes difficult, due to overlapping histopathological characteristics. The prognosis depends on the histological subtype, as well as on the anatomical site of origin, the former being the more important. The molecular basis for these differences in prognosis is poorly understood. Whole-genome analyses were used to investigate the association between molecular tumor profiles, pathogenesis and prognosis. A total of 85 periampullary adenocarcinomas were characterized by mRNA and miRNA expressions profiling. Molecular profiles of the tumors from the different anatomical sites of origin as well as of the different histological subtypes were compared. Differentially expressed mRNAs and miRNAs between the two histopathological subtypes were linked to specific molecular pathways. Six miRNA families were downregulated and four were upregulated in the pancreatobiliary type as compared to the intestinal type (P < 0.05). miRNAs and mRNAs associated with improved overall and recurrence free survival for the two histopathological subtypes were identified. For the pancreatobiliary type the genes ATM, PTEN, RB1 and the miRNAs miR-592 and miR-497, and for the intestinal type the genes PDPK1, PIK3R2, G6PC and the miRNAs miR-127-3p, miR-377* were linked to enriched pathways and identified as prognostic markers. The molecular signatures identified may in the future guide the clinicians in the therapeutic decision making to an individualized treatment, if confirmed in other larger datasets.