Precision oncology aims to personalize cancer treatment by targeting molecular alterations specific to individual tumors. This approach offers new therapeutic avenues for rare, resistant, or rapidly progressing cancers of children, adolescents, and young adults (CAYA). Despite progress in genomic and proteomic profiling, translating molecular findings into effective therapies for childhood tumors remains a significant challenge. The ability to assess pre-clinical efficacy by modeling patient-tumor-specific response could provide timely and meaningful information to guide treatment selection. This dissertation introduces the chick chorioallantoic membrane (CAM) model as a valuable alternative to conventional mouse models for generating patient-derived xenografts (PDXs). The primary advantage of the CAM-PDX model is its ability to create xenografts rapidly (in 3-7 days) using minimal tumor tissue. This rapid turnaround is crucial for testing therapies in time-sensitive situations involving a realistic in vivo environment. I established a dedicated CAM-PDX facility, where I optimized protocols specifically for engrafting pediatric tumor specimens. Through several case studies, I demonstrated the utility of this approach. In one, proteomics identified high SHMT2 levels in a progressive SETTLE tumor from an adolescent. Using the CAM-PDX model and in vitro assays, I validated the SHMT2 inhibitor sertraline as a potential therapy, confirming the tumor's reliance on the serine-glycine pathway. I also used the CAM-PDX model to evaluate mTOR inhibitors for a SPEN tumor and JAK inhibitors for a T-lymphoma, demonstrating how this model can test molecularly targeted drugs. Additionally, I generated xenografts from flash-frozen tumor samples, which expands the model’s utility to include archival clinical tissues that were not viably cryopreserved at point of collection. The viability and integrity of all xenografts were verified through growth patterns, biomarker imaging, and histological analyses. Overall, this work underscores that the integration of actionable targets with functional validation in CAM-PDX could potentially contribute to the development of personalized treatment strategies for difficult-to-treat pediatric cancers, offering a practical alternative to conventional in vivo models for time- and resource-sensitive research.
Tariq Ahmad Bhat (Thu,) studied this question.