Utilizing patient derived xenograft mouse models to improve liquid biopsies for glioblastoma multiforme
Glioblastoma multiforme (GBM) is an aggressive brain cancer affecting 12,000 new patients in the U.S. annually. There are no effective biomarkers for monitoring treatment response during chemotherapy in patients with GBM. In several cancer types, circulating tumor DNA (ctDNA) in plasma has emerged as a potential biomarker for tumor genotyping, treatment monitoring, and early tumor detection. However, in patients with GBM, ctDNA is difficult to detect in the blood, most likely because ctDNA released by tumor cells does not cross the blood-brain barrier (BBB), an interface between the central nervous system and blood. To validate this hypothesis and potentially overcome this challenge, we wanted to evaluate whether transient disruption of BBB will increase the amount of ctDNA in blood. However, this approach cannot be tested directly in patients until it is demonstrated pre-clinically. Therefore, we developed a ctDNA assay for GBM Patient Derived Xenograft (PDX) mouse models. This multiplexed droplet digital PCR assay is comprised of 4 human-specific targets (FAM labeled) and 4 mouse-specific targets (TET labeled). Using this assay, we can accurately differentiate human DNA from mouse DNA, with high specificity. Any human DNA detectable in plasma of a PDX mouse model can be considered tumor-derived ctDNA. In preliminary results, we successfully detected ctDNA in plasma of mice with subcutaneous tumor implants. In contrast, ctDNA levels in mice with intra-cranial implants were very low. In ongoing work, we are further optimizing this assay and establishing limits of detection. Once this assay is finalized, we will evaluate potential approaches to improve ctDNA detection in model systems and in patients with GBM.