Characterizing the Cytotoxic Response of Three-Dimensional Microtissues using Multi-Assay Outputs and Optical Coherence TomographydissertationsZein-Sabatto, AhbidcreatorHoffman-Kim, DianeReaderLee, JonghwanAdvisorMorgan, JeffreyReaderDarling, EricReaderToussaint, KimaniReaderBrown University. Biology and Medicine: Biomedical Engineeringsponsor2025xxv, 142 p.born digitalThesis (Ph. D.)--Brown University, 2025thesesThe implementation of functional precision medicine to cancer therapy involves directly examining the efficacy of treatments on patient tumors. This approach has the potential to match patients with the most effective treatment; however, challenges that preclude its widespread implementation include the loss of clinical relevance using traditional in vitro tumor culture platforms and the technical limitations associated with currently available viability assays for drug screening. The development of 3D in vitro cultures addresses the first challenge by recapitulating the tumor microenvironment and other physiological factors necessary to accurately model the patient tumor drug response. Nevertheless, there is a critical need for high-throughput screening techniques that assess cell viability to enable functional precision medicine. In this research, we validated and characterized the use of optical coherence tomography (OCT) viability imaging to noninvasively analyze changes in cell viability using four distinct biological mechanisms associated with cytotoxicity. First, we optimized the dose-response of HepG2 liver cancer, MCF7 breast cancer, and primary neonatal cortical rat microtissues using seven compounds that induced metabolic inhibition, membrane poration, DNA crosslinking, and microtubule overstabilization. Doses were initially optimized using the CellTiter-Glo 3D, mammalian Live/Dead, Caspase-Glo 3/7 3D, and Click-iT EdU proliferation assays. Lethal concentrations (LCs) that elicited 25, 50, and 75 percent cell death were determined for the compounds using the “best-fit” assay-treatment pair. Second, we collected multi-assay data by testing all assay-treatment combinations using the predetermined LCs for all tissue culture types. Linear mixed effects modeling and principal component analysis revealed the presence of multifaceted cellular injuries which highlight the need for multimodal approaches to measure cell viability. We also derived a novel metric, termed multi-mechanism lethal concentration 50, which simultaneously incorporated data from multiple assays into a single-value output relating to the potency of a compound. Finally, we presented a framework for analyzing and modeling cell viability using OCT imaging. OCT data was collected for each tissue culture type using the seven previously used compounds after 24 hours of treatment. A total of 77 metrics were extracted from the data covering changes in geometry, intensity, decorrelation, scattering attenuation, and diffusive motions. The collected data was preprocessed, filtered, and modeled using structural equation modeling to identify the relationship between cell viability, assays, and OCT metrics. The proposed framework served as a proof-of-concept to identify which OCT metrics were most attributed to changes in cell viability. Ultimately, the framework provided can be used in conjunction with other analytical approaches including machine or deep learning algorithms to generate an OCT based cell viability prediction model. The successful implementation of OCT viability imaging may help enable functional precision medicine applications in cancer therapy, toxicology, and novel drug development by overcoming current limitations in high-throughput screening techniques.CancerOptical coherence tomographyCell ViabilityMultivariate analysisFunctional Precision Medicine3D in vitro modelsEnglishRPB20250707