Treatment of pancreatic cancer: what can we really predict today?

Managing pancreatic cancer remains a big challenge due to its worse course and prognosis. However, therapeutic options and multimodal strategies are increasing nowadays, including new agents, new regimens and chemoradiation. Recently, the FOLFIRINOX regimen has been reported to be more active than gemcitabine in selected metastatic patients. In this setting, it will be of utmost interest to guide our therapeutic choice not only on clinical and pathological findings, but also on specific biomarkers that will predict tumor behavior and patient outcome (prognostic markers), and benefit from specific agents or regimens (predictive markers). In the near future, we will have to build both our therapeutic interventions and our clinical research based on an accurate patients' clinical selection and on biomolecular markers. In this review, we aimed to highlight and discuss some of the recent results reported on biomarkers in pancreatic cancer that may predict, i.e., preferential metastatic diffusion after surgery, like CXCR4, or predict gemcitabine efficacy in an adjuvant setting as well as in advanced disease, like hENT1. An important effort for translational research in pancreatic cancer research is thus required to validate such markers, while some important questions concerning tissue availability and processing, methodology of analysis, and design of future prospective trials, need to be addressed.