Gemcitabine remains a first-class chemotherapeutic drug for pancreatic cancer. However, due to the rapid development of gemcitabine resistance in pancreatic cancer, gemcitabine alone or in combination with other anti-cancer drugs only showed limited effect in the clinic. It is extremely challenging to effectively and efficiently determine the optimal drug regimens. Thus, identification of appropriate prediction biomarkers is critical for the rational design of gemcitabine-based therapeutic options. Herein, a pancreatic cancer stem cell (PCSC) model exhibiting chemoresistance to gemcitabine was used to test the activity of clinical cancer drugs in the presence or absence of gemcitabine. As determined by combinatorial treatment, several types of drugs resensitized gemcitabine-resistant PCSCs to gemcitabine, with sorafenib (EGFR inhibitor)/gemcitabine and sunitinib (TBK1 inhibitors)/gemcitabine drug combinations being the most preferred treatments for PCSCs. Following the validation of the PCSC model by an antibody array test of 15-gene expression of stemness biomarkers, NANOG showed markedly different expression in PCSCs compared to the parental cells. From comprehensive analysis of stem cell index versus combination index, a stemness-related correlation model was successfully constructed to demonstrate the correlation between NANOG expression and synergism. Cancer cell stemness was ascertained to be highly relevant to NANOG overexpression that can be abrogated by synergized gemcitabine-drug combinations. Therefore, NANOG works as a therapeutic biomarker for predicating efficient combinatorial treatment of gemcitabine in pancreatic cancer.
Read full abstract