Patient-derived xenograft models for oncology drug discovery

The success of targeted therapies for cancer patients rests on three major components: the right target(s), the right drug and drug combination, and the right patient population. Although much progress has been made in understanding the mechanism of disease and in refi ning pharmaceutical properties of therapeutic agents, the attrition rates between target discovery and drug marketing approval have been high, especially in oncology. One of the main reasons underlying this undesirable statistics is believed to be the lack of predictive power of the model systems used in the preclinical setting. Several strategies have been employed with the aim of improving the predictive value of the preclinical studies, such as incorporating genomic profi ling and molecular segmentation into model selection, and enhancing the development and application of patient-derived xenograft models even during early stage of drug discovery. This brief review will summarize some of the recent concept and practice in incorporating patient-derived models into all stages of drug discovery process, from target to clinical development