Assessment of Drug Effects in Cancer Xenograft Models with Multiple Tumor Outcomes

• Activity Number: 646
• Type: Contributed
• Date/Time: Thursday, August 7, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Elande Baro*+
• Companies: University of Maryland Baltimore County
• Keywords: missing data ; multivariate

Abstract

The tumor xenograft model is an animal model that is often used in cancer drug development for demonstrating efficacy of a treatment regimen. In xenografts, severe combined immunodefficient mice are grafted with human cancer cells and then treated with different regimens of anticancer agents and tumor growth is measured over a period of time. One mouse is usually grafted one tumor. However, it is very important to investigate the efficacy of a treatment for multiple tumors grafted in one mouse. For instance, in breast cancer, the occurrence of multiple tumors affecting one or both breasts at initial diagnosis is possible. Non-informative missing may occur when a mouse is sacrificed while informative missing may occur when a tumor volume becomes unmeasurable. We present a quasi F-test for comparison of two treatments effects and a growth curve model for modeling the dose response relationship that can be used for analysis of cancer xenografts models of mice with multiple tumors, in the presence of missing data. Simulation results as well as a real data application are provided.