Background: Pancreatic ductal adenocarcinoma (PDAC) patients with diabetes mellitus (DM) have poor overall survival. Underlying mechanisms have not been fully clarified. This presents an opportunity for precision-oncology for which we systematically analysed publicly-available PDAC transcriptome data.
Methods: PDAC TCGA RNASeq data were used. Analyses were restricted to only 'high purity' and 'head' as anatomical site. Patients were characterised by: (1) Gene expression classification, and (2) Weighted gene correlation network analysis (WGCNA) to identify co-expression patterns of genes. Newly identified gene signature subclasses of pancreatic head PDAC were associated with clinical and functional characteristics of patients.
Results: Consensus clustering identified two patient subclasses within PDAC involving pancreatic head. WGCNA identified 11 distinct networks of gene expression patterns across two sub-classes. Class 1 patients demonstrated a significant upregulation of Module 5 and Module 6 gene expression compared to Class 2. Class 1 predominantly expressed the acinar, ductal and islet cell gene signatures. There were significantly less patients with DM in Class 1 subclass compared to Class 2 (p < 0.037). Patients with DM had significant downregulation of pathways involved in cellular metabolism, hormone secretion and paucity of islet cell markers with no reduced survival compared with non-diabetics.
Conclusions: A significant proportion of patients with PDAC of pancreatic head and DM exhibit downregulation of pathways involved in cellular metabolism, hormone secretion and signalling accompanied by a paucity of islet expression. Investigating the relationship between DM and gene expression profiles in patients with PDAC presents opportunities to improve overall survival in diabetics with PDAC.
Keywords: general surgery; genetics; glucagon; hepatopancreaticobiliary surgery; insulin; pancreas; transcriptome.
© 2021 Royal Australasian College of Surgeons.