Investigator initiated clinical and translational research program in childhood sarcoma

  • Beginning with an individual investigator award from the Cancer Prevention and Research Institute of Texas (RP150164), Dr. Leavey developed a long-lasting collaboration with Dr. Ovidiu Daescu, Department Chair of Computer Science at UT Dallas.
    • Together they have leveraged advances in artificial intelligence and machine learning with advances in bioimaging to study therapy-induced tumor necrosis in osteosarcoma, the most common childhood bone cancer. They demonstrated the ability to quantify necrosis in the histology of resected tumors and then, using creative image registration techniques, trained a model to quantify necrosis in pre-operative MRI and generated 3-D images of necrosis.
    • Their ongoing work now focuses on the creation of an osteosarcoma-specific computer-assisted diagnostic tool for nodule identification by chest CT.
  • Continuing the focus on advanced analytics, Dr. Leavey teamed with Dr. Guanghua Xiao, Professor in the O’Donnell School of Public Health, to interrogate rhabdomyosarcoma histology.
    • Together they were the first to describe that machine learning can separate histology types in rhabdomyosarcoma and, more importantly, that in a homogenous pathological cohort, the potential ability of machine learning to separate risk groups.
  • In continuing to focus his translational work on childhood sarcoma, Dr. Leavey paired with Dr. Ralph DeBerardinis to secure an R21 to study metabolic vulnerabilities in pediatric fusion positive sarcoma.
    • Together they demonstrated the feasibility of delivering intra-operative 13C-glucose infusions, allowing metabolic flux and metabolomic profile analysis for extra-cranial tumors in children.