Background: Pancreatic head cancer and pancreatic body/tail cancer are considered to have different clinical presentations and to have altered outcomes.
Methods: Ninety cases of pancreatic adenocarcinoma (PDAC) from our institution were used as a discovery set and 166 cases of PDAC from the TCGA cohort were used as a validation set. According to the anatomical location, the cases of PDAC were divided into the pancreatic head cancer group and the pancreatic body/tail cancer group. Firstly, the different biological functions of the two groups were assessed by ssGSEA. Meanwhile, ESTIMATE and CIBERSORT were conducted to estimate immune infiltration. Then, a novel anatomical site-related risk score (SRS) model was constructed by LASSO and Cox regression. Survival and time-dependent ROC analysis was used to prove the predictive ability of our model in two cohorts. Subsequently, an integrated survival decision tree and a scoring nomogram were constructed to improve prognostic stratification and predictive accuracy for individual patients. In addition, gseaGO and gseaKEGG pathway analyses were performed on genes in the key module by the R package.
Results: Overall survival and the objective response rate (ORR) of patients with pancreatic body/tail cancer were markedly superior to those with pancreatic head cancer. In addition, distinct immune characteristics and gene patterns were observed between the two groups. Then, we screened 5 biomarkers related to the prognosis of pancreatic cancer and constructed a more powerful novel SRS model to predict prognosis.
Conclusions: Our research shed some light on the revelation of gene patterns, immune and mutational landscape characterizations, and their relationships in different PDAC locations.
Keywords: immunologic and hallmark gene sets; pancreatic adenocarcinoma; pancreatic body/tail cancers; pancreatic head cancer; prognosis.
Copyright © 2022 Ge, Ma, Mao, Xu, Zhang, Li, Wang, Yao, Yue, Jiao, Wang, Zhuo, Han, Hu, Zhang, Cui and Wang.