LDHA is a prognostic biomarker on the immune response in pancreatic adenocarcinoma and associated with m6A modification.

PURPOSE

N6-methyladenosine (mA) is tightly associated with the progression of pancreatic ductal adenocarcinoma (PDAC). Another prominent feature of PDAC is metabolic reprogramming, which provides sufficient nutrients to support rapid cell growth via the tumor microenvironment. However, the underlying influences of mA-associated metabolic genes on the PDAC microenvironment remain poorly understood. Therefore, we sought to construct a survival prediction model using mA-related genes to clarify the molecular characteristics of PDAC.

METHODS

In the present study, mA-related metabolic genes were obtained from The Cancer Genome Atlas (TCGA) pancreatic adenocarcinoma dataset and subjected to coexpression analysis. Consensus clustering recognized two distinct subgroups with different immune cell infiltration patterns according to the expression of mA-associated metabolic genes. Multivariate Cox regression analyses and least absolute shrinkage and selection operator (LASSO) analysis were adopted to create an mA-related metabolism model. A nomogram including clinical features and the risk score based on the expression of mA-related metabolism regulators was constructed.

RESULTS

A four-gene signature comprising ATP8B2, GMPS, LDHA and SDR39U1 was built to predict the overall survival (OS) of PDAC patients. This signature also robustly predicted survival in two independent validation cohorts from the International Cancer Genome Consortium (ICGC) and ArrayExpress (E-MTAB-6134). The four-gene signature divided patients into high- and low-risk groups with distinct OS values as verified by the log-rank test. Among the four genes, LDHA was upregulated in both PDAC tissues and cell lines.

CONCLUSIONS

Collectively, we analyzed the immune microenvironment, predicted drug sensitivity and assessed the implications of the mutation landscape based on the crosstalk between mA regulators and metabolic rewiring, and we also constructed a novel signature based on mA-associated metabolic genes to predict PDAC prognosis.