RNA-seq based mapping strategies to uncover heterogeneity in survival among Pancreatic Ductal Adenocarcinoma (PDAC) patients

Pancreatic ductal adenocarcinoma (PDAC) is categorized as the seventh leading cause of cancer mortality in the world. Little is known about predictive markers for long-term survival. In this work, we performed a series of transcriptome computational analyses to better understand patient heterogeneity between longterm (LT) and short-term (ST) survivors. Using a discovery cohort of 19 PDAC patients from CHU-Liège (Belgium), we first identified differentially expressed genes between LT/ST. The 216 predicted genes could be linked to multiple metabolic and cell cycle related pathways. Second, we performed unsupervised system biology approaches to obtain gene modules for our PDAC samples. In particular, important modules obtained via weighted gene co-expression network analysis (WGCNA) showed significant correlation with clinical features, including overall survival, tumour size, and tumour invasion. Third, we created individual-level perturbation profiles (PEEP) and found that both group-level and individual-level approaches indicated a change in secondary metabolic pathways and FoxO signalling pathways when comparing ST with LT patients. In addition, individual-level gene expression make-ups seemed to suggest a larger heterogeneity among long-term survivors compared to short-term survivors. In conclusion, despite the small sample size, but using testing strategies for small samples whenever possible, we have shown how the combination of multi-level information can give important clues towards PDAC prognosis and patient follow-up in personalized medicine.