In this work we first developed a murine pancreatic tumor model using CD24(+) CD44(+) pancreatic cancer stem cells (CSC) supported by an electrospun scaffold. Unlike conventional models, the use of CSC and the scaffold, which were biologically and chemically defined, afforded scientists a reliable platform to evaluate novel chemotherapy regimens. CD24(+) CD44(+) CSC successfully initiated tumorigenesis in vitro on the scaffold without suffering apoptosis, evidencing the lack of cytotoxicity of scaffolding materials. Also, the scaffold contributed to the acceleration of in vivo tumorigenesis and increased the likelihood of tumor formation. Using this model, we set out to explore the effectiveness of irinotecan/gemcitabine (IRIN-GEM), a chemotherapy regimen, for pancreatic cancer. Our study showed that IRIN-GEM induced a tumor regression whereas gemcitabine alone could only arrest the tumor growth. Further study suggested that the superior performance of IRIN-GEM could be attributed to its capacity to demolish the CD24(+) CD44(+) CSC sub-population by inducing a large-scale apoptosis. The use of highly proliferative yet homogenous CD24(+) CD44(+) CSC along with a chemically defined scaffold accelerated the tumor formation and significantly reduced the variability associated with conventional murine models. Armed with this new model, we discovered that IRIN-GEM would be a promising chemotherapy candidate for patients with advanced pancreatic cancer.