PDX Clinical Trial Design in Anti-Cancer Research

Animal models are useful tools for understanding cancer biology and genetics and serve as an essential platform for the preclinical development of anticancer therapeutics. In this context, cancer-bearing patient-derived xenograft (PDX) models, also called cancer avatars, have successfully replaced the traditional cell linederived models in recent years. PDX-based studies are now widely used for preclinical testing of novel treatments as well as tailoring personalized medicine. For anti-cancer research, however, the use of PDX models propagated from a unique patient does not fully represent the true therapeutic efficacy and toxicity of a drug. That is why many studies in this format later failed to show efficacy and safety in human clinical trials. Hence, the concepts of PDX clinical trials and co-clinical trials have gained importance and prospered in recent years. A PDX clinical trial implies investigation on a set of PDXs originated from multiple patients prior to an early phase human trial, whereas a co-clinical trial refers to drug response assays, in parallel and simultaneously with a human clinical trial, on a set of PDX models established from the same clinical trial participants. A carefully designed PDX- /co- clinical trial requires a meticulous calculation of the sample size, enrollment of pathologically and molecularly diverse patients, and selection of suitable endpoints and outcome measures. With a special focus on PDX clinical trial design in anti-cancer research, this chapter specifically addresses how to develop cancer-bearing PDX models, what to consider in characterizing them, how to track their fidelity to the parental tumor, how to estimate the number of animals included in a PDX trial, how to achieve greater power in the translation of final outcomes, what are the minimum endpoints to be considered, and what measures are preferred for evaluating the response to therapeutic interventions.

Keywords: Cancer Avatars, Clinical Trials, Malignant Neoplasms, Patient- Derived Xenograft Models, Xenograft Model Antitumor Assays.