• Cancer prevention
• pancreatic cancer

Pancreatic cancer is a frequent and devastating cause of morbidity and mortality in many parts of the world with a 5 year survival of only 5%.1 Unfortunately, these dismal statistics have not improved appreciably over the last 30 years despite an explosive growth in our understanding of the disease's pathogenesis. Indeed, a recent model of human pancreatic cancer provides important histological and molecular insights,2 but for the most part, these insights await translation into improved markers for risk assessment, new agents for prevention and therapy, as well as intermediate endpoints that could prove useful in guiding intervention trials.

Despite these translational limitations, our growing knowledge of human cancer development has stimulated the development of many new, genetically driven rodent models that promise to provide novel mechanistic insights into cancer development and progression, and/or serve as improved preclinical systems in which to evaluate promising preventive or therapeutic agents. Of course, the relevance of these models to human applications depends on several factors.3 4 Fundamentally, the most promising new animal models should: (1) employ genetic manipulations that mimic key molecular determinants of human disease—optimally, in nature, magnitude, timing, and frequency; (2) reflect histopathological and phenotypic features of human disease that evolve over time; and (3) offer reproducible results across time and various laboratories. Additional desirable features of somewhat lesser importance include: (1) ready public access, (2) efficiency of use (eg, regarding the frequency and timing of meaningful endpoints), (3) modest cost, (4) breeding efficiencies (eg, for generation and maintenance of the colony), and (5) an intact immune system. When an animal model is used to …