Characterization of prostate tumor spheroid growth and response to treatment using dynamic optical coherence tomography

Abstract

Prostate cancer, the most prevalent cancer in North American men, is often treated with radiation and in the case of advanced disease, with the first-line chemotherapeutic docetaxel (DTX). The use of in vitro cell culture as a model for in vivo patient tumors has been instrumental in understanding the cancer biology underpinning tumor development, progression, and treatment. Conventionally, in vitro cells were cultured as 2D monolayers on hard, flat surfaces. However, it has become well understood in the last half-century that the behavior of in vivo tumors is better replicated by small 3D aggregates of in vitro cancer cells called tumor spheroids. While cells proliferate along the spheroid periphery, the center of the spheroid succumbs to starvation and waste build up. In between forms an intermediate layer of hypoxic and non-proliferative quiescent cells: two characteristics associated with treatment resistance. Despite the physiological pertinence of spheroid culture, its 3D nature challenges conventional biological methods. Moreover, cell culture geometry influences cell behavior not only during treatment, but in post-treatment recovery and throughout measurement as well. The clonogenic assay is a gold standard method for quantifying cell survival following treatment; however, it requires spheroids to be disaggregated and cultured in 2D for colony formation, which alters the cellular response. Proliferation assays quantify cellular activity inside intact spheroids, but similarly lack spatial resolution. Although fluorescence microscopy (FM) enables spatially resolved spheroid evaluation, it remains largely qualitative rather than quantitative. Proliferation assays and FM also require the addition of exogenous agents that are invasive to the sample and struggle to penetrate large spheroids. Alternatively, optical coherence tomography (OCT) enables non-invasive, high-speed, volumetric imaging of biological tissues with cellular resolution. Analysis of temporal OCT intensity fluctuations generates dynamic OCT (dOCT) images that can provide a quantitative and spatially resolved measurement of cellular activity throughout the spheroid. Nevertheless, investigating treatment response in spheroids is known to be tedious and time consuming, and some questions of interest cannot be accessed experimentally with adequate accuracy. As such, mathematical in silico models of spheroids have become increasingly prevalent. The heterogeneity inherent to in vivo tumors and in vitro spheroids is recapitulated by discrete in silico methods like agent-based modeling (ABM). These methods simulate each cancer cell as an individual “agent” that responds to neighboring cells, nutrients, drugs, and other components of its local environment. In particular, the cellular Potts model (CPM) is a form of ABM with phenomenological utility, given adequate validation to experimental observations. As such, it can describe, visualize, interpolate, and potentially even extrapolate experimental data to guide future experiments. This dissertation investigated and validated emerging and under-used in vitro tools for characterizing prostate tumor spheroid growth and response to treatment with radiation and DTX. First, I created a semi-automated masking process to isolate spheroids from volumetric OCT images for high resolution morphological analysis. For analysis of dynamic motion, I generated dOCT images using a frequency banding method that had only been previously subject to qualitative evaluation. Then, I developed a technique to quantify cellular activity in volumetric dOCT images of masked spheroids. Visual and quantitative comparison of live and formaldehyde-fixed spheroids imaged with dOCT and FM confirmed that dynamic motion measured with the dOCT method was associated with cellular activity. To mitigate influences unrelated to cellular activity, the average measurement of formaldehyde-fixed spheroids was subtracted from quantitative dOCT measurement as background. Post-treatment recovery of prostate tumor spheroids exposed to DTX was investigated with repeated measurement via Alamar Blue (AB) proliferation assay and longitudinal observation of clonogenic assay colony formation. Excellent agreement was observed between the quantitative dOCT method and AB proliferation assay over two weeks of longitudinal spheroid growth. Volumetric morphological analysis supported the measured cellular activity trends. However, fixation-subtraction could not be performed in a spatially sensitive manner and the dOCT images failed to resolve the longitudinal formation of a necrotic spheroid core that was observed via FM. Nonetheless, dOCT images and quantification of spheroids post-radiation demonstrated good agreement with FM and AB, respectively. Spheroids treated with radiation and DTX demonstrated better survival compared to monolayer culture. Monolayer culture treated with low dose DTX demonstrated higher post-treatment cellular activity and faster colony formation, but clonogenic survival remained lower than the untreated control. This effect was also observed in spheroids, albeit to a lesser extent. Experimental observations were probed with novel in silico CPMs of prostate tumor spheroid treatment. This thesis serves as a step towards validating emerging and under-used in vitro tools for spheroid evaluation in well-studied conventional cancer treatments. Once validated, these in vitro tools are particularly well suited for discovery and testing of novel targeted cancer treatments since spheroid protein and gene expressions are more physiologically representative than monolayer culture.