Date of Award

Summer 8-15-2022

Author's School

McKelvey School of Engineering

Author's Department

Biomedical Engineering

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Brain cancer severely threatens human health due to its disruption of neurological function, poor prognosis, and substantial reduction in quality of life. Glioblastoma (GBM) is the most devastating brain cancer; not only is it the most common malignant primary tumor in adults, but also it has a median survival of 14 months with a 5-year survival rate of less than 5%. Despite advances in multidisciplinary treatment that includes surgical resection, radiation therapy, and chemotherapy, almost all patients experience tumor progression and nearly universal mortality within 2 years. However, advances in patient care have suggested that the accurate diagnosis of molecular subtypes is critical for individualized targeted treatment and improving survival outcome for brain cancer patients.Conventional diagnostic evaluation begins with neuroimaging and continues with surgical tissue biopsy to confirm the diagnosis and acquire the molecular profile of the tumor. Though tissue biopsy is the gold standard for molecular characterization, there are significant risks for patients because the procedure is invasive. Liquid biopsy is a minimally invasive approach that enables genetic profiling by detecting circulating tumor-derived biomarkers that were shed by tumors into the blood circulation. However, blood-based liquid biopsy is inherently limited by the blood-brain barrier (BBB) that hinders the release of molecular biomarkers, leading to a low detection sensitivity for GBM. The combination of focused ultrasound (FUS) with microbubbles is an established technique to disrupt the BBB noninvasively and transiently with high precision (on the order of millimeter). Though this has conventionally been used to deliver drugs from the bloodstream to the brain tissue of interest, it is hypothesized that this FUS-induced BBB disruption enables molecules to be released from the tissue into the blood circulation. Under this “two-way trafficking” hypothesis, FUS-enabled blood-based liquid biopsy (sonobiopsy) can release brain tumor-derived biomarkers into the blood circulation to improve the sensitivity for noninvasive molecular characterization of GBM. In this work, we evaluated the feasibility, safety, and efficacy of sonobiopsy in small and large animal models to provide a minimally invasive, spatiotemporally-controlled, and sensitive molecular characterization of brain diseases. First, we evaluated the impact of different sonobiopsy parameters on the extent of biomarker release and tissue damage in a mouse GBM model. The blood collection time after FUS sonication was an important factor to minimize the effect of clearance and maximize the level of biomarkers detected in the plasma. Importantly, careful optimization of key sonobiopsy parameters, e.g., FUS pressure, microbubble dose, and sonication volume, was necessary to increase the release of circulating biomarkers while minimizing the potential for tissue damage. With the optimized parameters, sonobiopsy significantly increased the plasma level of GBM-derived biomarkers and improved the detection sensitivity for two clinically relevant mutations. Second, sonobiopsy was performed in a non-tumor pig model to demonstrate the potential for clinical translation. A customized sonobiopsy device was developed to target a specific brain area and release brain-specific biomarkers into the blood circulation. Importantly, sonobiopsy significantly increased the plasma level of these biomarkers without causing detectable tissue damage. This large animal study demonstrated that sonobiopsy has the potential to be safely translated to humans. To further underscore the potential for clinical translation of sonobiopsy, a pig GBM model was developed to assess the feasibility of sonobiopsy to release GBM-derived biomarkers and improve the detection sensitivity for two clinically relevant mutations. We achieved localized BBB disruption and the plasma level of GBM biomarkers significantly increased shortly after FUS sonication in the large animal tumor model. Importantly, sonobiopsy improved the detection sensitivity for two mutations without causing off-target damage. This addressed the fundamental limitation—obtaining specimens with a sufficient abundance of circulating tumor biomarkers—for the minimally invasive, sensitive molecular characterization of GBM. Lastly, we evaluated the impact of sonobiopsy as a platform technology to aid in the diagnosis of other brain diseases. After performing sonobiopsy in a transgenic mouse model of tauopathy, there was a significant increase in the plasma levels of pathologic proteins and a key marker for neurodegeneration. This demonstrated the potential to use sonobiopsy for the noninvasive diagnosis of neurodegenerative disorders. In summary, this work provided evidence that supports the clinical translation of sonobiopsy as a minimally invasive, spatiotemporally-controlled, and sensitive molecular characterization of brain diseases. This enhanced capability could have an important impact throughout the continuum of patient care from brain disease diagnosis and treatment monitoring to recurrence detection. In addition, sonobiopsy could support the investigation of disease-specific molecular mechanisms and accelerate the development of targeted therapy.


English (en)


Hong Chen

Committee Members

Aadel A. Chaudhuri, H M. Gach, Eric C. Leuthardt, James D. Quirk,