Determining the Origins and Vulnerabilities of Pediatric Treatment-Induced Gliomas

ABSTRACT Pediatric treatment-induced high-grade glioma (HGG), also known as radiation-induced glioma (RIG), is an incurable secondary cancerous brain tumor that affects children who have previously received cranial radiotherapy (RT), most commonly for leukemia or a separate, primary brain tumor (the two most common

childhood cancers). They affect up to 4% of children who have received cranial RT and account for up to 10% of pediatric brain tumor deaths. They have no known effective treatment, and research on these uniformly fatal tumors has been unacceptably limited. We recently broadly characterized these tumors’ molecular features in

a large patient sample cohort. We found that while RIG and primary pediatric HGG share histological similarities, their genetic alterations and gene expression profiles are different. RNA-seq analysis of treatment- induced HGG shows they form two expression subgroups (A and B) of approximately equal incidence.

Findings in both groups suggest the presence of germline DNA repair defects that may place some patients at increased risk of developing these tumors after receiving RT. We have characterized these tumors’ cells of origin through single-cell RNA-Seq. We have also developed a world-first set of patient-derived cell culture and

orthotopic xenograft (PDX) models of these tumors. We found that RIG is vulnerable to many clinically relevant therapies, including DNA disruptors and MEK inhibitors. For this proposal, we hypothesize that germline susceptibility to DNA disruption from RT contributes to oncogenesis of these tumors, but this origin in DNA

disruption in turn renders them susceptible to DNA disrupting treatment like RT and specific chemotherapies, so that combining DNA disrupting treatment with targeting of tumor-specific molecular dependencies will improve outcomes. Our first aim is to discover and validate germline risk factors for RIG through DNA

sequencing of the germlines of patients and controls and then testing putative genetic variants conferring risk through an in vitro stem cell model. Second, we will inhibit predicted pathways of tumor growth in three types of RIG cells of origin to establish pathways key to tumor growth and treatment resistance. We will supplement the

predicted pathways with screens based on orthogonal functional genomic techniques in vitro and in vivo. Lastly, we will use our patient-derived models of these tumors to test combinations of effective chemotherapies and RT in cell culture and take the most successful treatment plans for validation in our PDX models. We will

also conduct orthogonal assessment of treatment efficacy through our newly opened RIG patient registry. Through these aims, we will lead research on these tumors from current limited biological understanding to advanced understanding of tumor origins with a pathway to prevention, with implications for other secondary

cancers as well. Critically, we will also understand rational, combination treatment susceptibilities. We will be able to translate this knowledge to a RIG-specific clinical trial to finally improve outcomes for these tragic patients currently dying of brain tumors caused by treatment that helped cure their primary cancer.