Using genomic technologies and mouse modeling including patient-derived xenografts (PDX) and genetically engineered mouse models (GEMM), we are discovering and validating novel biomarkers and therapeutic targets of pediatric cancers
We developed both the concepts and applications of clinically relevant biomarkers that could predict chemoresistance and metastatic potential of osteosarcoma at the time of initial diagnosis before the initiation of treatment. This set the stage for developing personalized predictive medicine. This work has been continuously funded by a series of NCI grants that spanned more than 11 years with the current TARGET (Therapeutically Applicable Research to Generate Effective Therapies) project being the third study that is approaching its conclusion. Based on the largest cohort of osteosarcoma cases ever assembled, we show that osteosarcoma is a genomic instability disease which has important therapeutic implications.
Together with a large team of investigators, we established the premise that pediatric brain tumors including germ cell tumors could be molecularly classified into prognostically relevant subgroups using genomic profiling. In addition, potential novel targets could also be identified for many of these subgroups that could be exploited for personalized therapy. We initially started with gene expression profiling by microarrays and later incorporated next generation sequencing as well as epigenetic profiling and has begun applying these genomic profiling platforms in molecular classification in the clinical setting.
Together with Melissa Bondy and Sanjay Shete from MDACC, we put together an international consortium of 13 centers, Gliogene, to study the genetic predisposition of human glioma, using initially the genetic linkage approach to study familial cases and later expanding to genome-wide association approach to include all sporadic cases. Since then this consortium has collected and studied more than 500 glioma cases from over 100 families by linkage analysis and also more than 4500 cases of sporadic gliomas by GWAS which identified a germline variant in the POT1 gene in the familial cases and at least 13 risk loci for glioma from GWAS.
Together with Xiao-Nan Li from TCH/BCM, we have successfully established and characterized the first patient-derived orthotopic xenograft (PDOX) models from various pediatric brain tumors including medulloblastoma, glioma, ependymoma, PNET, and germ cell tumors. These xenografts have been shared with multiple investigators in the field and are now used routinely in the NCI Pediatric Preclinical Testing Consortium (PPTC) for drug testing.