A Study of Hydroxychloroquine in Combination With Gemcitabine/Abraxane in Pancreatic Cancer

Recent strategies have focused on improving the efficacy of gemcitabine either by improving the method of delivery, or by combining gemcitabine with other non-cross resistant agents. A sequence of Phase III combination studies of gemcitabine in combination (with oxaliplatin, and with the targeted therapies bevacizumab and cetuximab) have been negative, though based on strikingly positive Phase II data generated in cancer centers. Several studies suggest that taxanes are active in pancreatic cancer, but a randomized trial of gemcitabine with taxanes has not been preformed, probably on the basis that the differences in Phase II were insufficiently persuasive. The development of a novel taxane conjugate with albumin, abraxane, with established activity in breast cancer, prompted a Phase II trial of gemcitabine/abraxane by Von Hoff (6). Phase I/II data were highly promising, with response rates of the order of 40%, with tolerable toxicity, and a one-year survival of about 48%. A phase III trial of gemcitabine versus gemcitabine/abraxane is in progress, and based on these promising data has served as the control chemotherapy for previous SU2C trials. The development of a more intensive, but toxic regimen (FOLFIRINOX) in no way diminishes the enthusiasm for this chemotherapy backbone, given the activity in Phase II trials that appears comparable (7). Given the promise of this regimen, and the possibility of making a substantial improvement in outcome with additional targeted interventions, we propose to continue to use this regimen in the current study. Of particular interest in extending these studies to pancreatic cancer is the finding that autophagy inhibition is particularly deleterious to cell lines bearing a mutant Kras protein. Additional studies as part of the SU2C pancreatic cancer project reveal that an autophagy program is activated in the presence of mutant Kras, and thus prompts the testing of this strategy in a setting in which Kras is commonly (about 85%) mutated (SU2C, unpublished data).