Purpose: Pancreatic adenocarcinoma is the fourth leading cause for cancer-related mortality with a survival rate of less than 5%. Late diagnosis and lack of effective chemotherapeutic regimen contribute to these grim survival statistics. Relapse of any tumor is largely attributed to the presence of tumor-initiating cells (TIC) or cancer stem cells (CSC). These cells are considered as hurdles to cancer therapy as no known chemotherapeutic compound is reported to target them. Thus, there is an urgent need to develop a TICtargeted therapy for pancreatic cancer. Experimental Design: Weisolated CD133+ cells from a spontaneous pancreatic ductal adenocarcinoma mouse model and studied both surface expression, molecular markers of pancreatic TICs. We also studied tumor initiation properties by implanting low numbers of CD133+ cells in immune competent mice. Effect of Minnelide, a drug currently under phase I clinical trial, was studied on the tumors derived from the CD133+ cells. Results: Our study showed for the first time that CD133+ population demonstrated all the molecular markers for pancreatic TIC. These cells initiated tumors in immunocompetent mouse models and showed increased expression of prosurvival and proinvasive proteins compared to the CD133-non-TIC population. Our study further showed that Minnelide was very efficient in downregulating both CD133-and CD133+ population in the tumors, resulting in a 60% decrease in tumor volume compared with the untreated ones. Conclusion: As Minnelide is currently under phase I clinical trial, its evaluation in reducing tumor burden by decreasing TIC as well as non-TIC population suggests its potential as an effective therapy. © 2014 AACR.