Cancer treatment selection is hindered by a lack of personalized data, relying instead on empirical evidence to justify treatment selection. For multiple myeloma (MM), a hematologic cancer, treatment selection is complicated by the occurrence of drug resistance, and patients have been observed to respond in an unpredictable fashion. To address these challenges, there is a need for more personalized treatment selection methods. Since the 1950’s, tools and assays for testing individual patient samples with a variety of drug treatments, called chemosensitivity and resistance assays (CSRAs), have been under development. There are currently no CSRAs approved for clinical use due to a lack of data supporting improved clinical outcomes compared to current empirical methods. A lack of improved clinical outcomes has been linked to limitations associated with the tools and methods used to perform these assays, as they require higher quantities of patient sample than some patients can provide, and they do not sufficiently mimic in vivo conditions.

To address these limitations, my objective was to produce several individual advances for a platform previously developed to study MM. Using microfabrication and device material characterization, poly(dimethylsiloxane) (PDMS) drug absorption and its impact on cytotoxicity assays was studied. Diffusion modelling and analysis was performed alongside physical experiments to confirm near-uniform soluble factor diffusion across microsystems. Cell population heterogeneity measures were used to study the effects of varying cell population to assess the feasibility of reducing cell seeding populations. Finally, the importance of accurately mimicking in vivo drug treatment concentrations was assessed by comparing two approaches for analyzing cytotoxicity assay data. The combination of these advancements positions the platform for further application in clinical studies, and to further our basic understanding of MM, with a specific focus on disease progression and drug resistance.