Identification of prognosis-related molecular subgroups based on DNA methylation in pancreatic cancer

Pancreatic cancer (PC) is one of the most lethal and aggressive cancer malignancies. The lethality of PC is associated with delayed diagnosis, presence of distant metastasis, and its easy relapse. It is known that clinical treatment decisions are still mainly based on the clinical stage and pathological grade, which are insufficient to determine an appropriate treatment. Considering the significant heterogeneity of PC biological characteristics, the current clinical classificatory pattern relying solely on classical clinicopathological features identification needs to be urgently improved. In this study, we conducted in-depth analyses to establish prognosis-related molecular subgroups based on DNA methylation signature.

DNA methylation, RNA sequencing, somatic mutation, copy number variation, and clinicopathological data of PC patients were obtained from The Cancer Genome Atlas (TCGA) dataset. A total of 178 PC samples were used to develop distinct molecular subgroups based on the 4227 prognosis-related CpG sites. By using consensus clustering analysis, four prognosis-related molecular subgroups were identified based on DNA methylation. The molecular characteristics and clinical features analyses based on the subgroups offered novel insights into the development of PC. Furthermore, we built a risk score model based on the expression data of five CpG sites to predict the prognosis of PC patients by using Lasso regression. Finally, the risk score model and other independent prognostic clinicopathological information were integrative utilised to construct a nomogram model.

Novel prognosis-related molecular subgroups based on the DNA methylation signature were established. The specific five CpG sites model for PC prognostic prediction and the derived nomogram model are effective and intuitive tools. Moreover, the construction of molecular subgroups based on the DNA methylation data is an innovative complement to the traditional classification of PC and may contribute to precision medicine development, therapeutic efficacy prediction, and clinical decision guidance.