2025 Theses Doctoral

Technologies for Patient-Specific Drug Screening, Monitoring and Treatment Towards Precision Medicine

Kennedy, Darragh Gerard

Precision medicine has traditionally focused on personalized cancer therapeutics, however patient-specific tools are needed to optimize diagnosis and treatment options for a variety of conditions beyond cancer. In this dissertation, we develop patient-specific technologies that can improve patient outcomes across diagnosis, drug screening and treatment.

Glioblastoma multiforme (GBM) is among the most deadly forms of malignant brain tumor and is characterized by profound intra-tumoral heterogeneity. While chemotherapy can be effective against proliferating GBM cells, we need patient-derived models that recapitulate the full diversity of GBM phenotypes. In Aim 1, we developed a microfluidic device that can be used to culture or screen drugs on thick, intact GBM slices while preserving GBM microenvironmental cues. The platform enables a single cell-specific drug readout and preserves native drug responses to a better degree than standard culture platforms.

More than 30 million Americans have diabetes and are at higher risk of developing peripheral artery disease (PAD). Advanced PAD can manifest as chronic limb-threatening ischemia (CLTI), which affects approximately 2 million Americans over the age of 40 and is associated with higher risk of limb loss. Many of these patients do not have viable options for revascularization using traditional surgical or endovascular methods, which carry their own risks of adverse events. As such, there is a clear need for therapeutics that can rapidly restore blood perfusion in end-stage CLTI patients. In Aim 2, we used iPSC-derived endothelial cells to generate pre-vascularized organoids as an injectable therapeutic for revascularization. We demonstrated growth of capillary-like structures when cultured in vitro and showed that the secretome of organoids increased endothelial cell migration and tube formation, key hallmarks of angiogenesis, compared to single-cell equivalent suspensions. Finally, we showed that organoids increased vascularization in the chick chorioallantoic membrane (CAM) assay.

Chronic wounds are characterized by dysregulated healing and affect millions globally. The incidence of chronic wounds is expected to increase due to the increasing number of individuals with diabetes, obesity and cancer. Current methods to study wound healing progression are limited, expensive, and require bulky equipment. In Aim 3, we demonstrated the utility of a non-invasive ultrasound probe to accurately provide information regarding the progression of complex wounds. We used a handheld, commercially available and inexpensive ultrasound probe to take images of fast- and slow-healing chronic wounds on diabetic mice and were able to discern differences in wound healing between groups by analyzing these images.

Taken together, the technologies in this dissertation present a range of tools for optimizing healthcare at various steps of the patient’s journey, from diagnostics to drug-screening through to therapy.