MS 26.04 PDx Model

Abstract

The tumor microenvironment differs significantly from the controlled environment of in vitro cell culture. Whereas cell lines are typically grown in plastic vessels containing a pH buffered, nutrient rich liquid medium with 10% fetal bovine serum and incubated at 37 ºC and 5% CO2, cancer cells in tumors experience different oxygen tension, growth substrate, nutrient, waste, and growth factor gradients, pH, and the presence of stromal support cells and infiltrating immune cells. The process of establishing a new cell line requires adaptation to this environment, which results in loss of genetic heterogeneity as well as irreversible epigenetic reprogramming that is maintained even when cultured cancer cells are re-introduced to an in vivo setting. Patient-derived xenografts (PDX) are direct human tumor xenografts established and maintained exclusively in mouse hosts. While these models can never perfectly recapitulate an autochthonous human tumor, they are increasingly used as tools in cancer research due to their utility in modeling therapeutic response with higher fidelity than either cell lines or cell line xenografts. A growing interest in defining the role of the tumor microenvironment and in testing the efficacy of immunotherapy in vivo has driven advances in PDX model development. The tumor microenvironment consists of stromal cell components including blood vessels, fibroblasts, and infiltrating immune cells as well as extracellular matrix, growth factors, and nutrients. In PDX, components of the tumor microenvironment are either provided by the mouse host or are excreted by the tumor cells themselves. However, paracrine effects between tumor and stroma may not be entirely replicated in a mouse host due to a lack of cross-species cytokine reactivity. The absence of compatible stroma may bias PDX engraftment toward tumors that are less dependent on paracrine factors or which are more adept at recruiting mouse stromal support cells through enhanced expression of mouse reactive factors. It remains a significant challenge to accurately assess the mechanistic activity of therapeutic approaches designed to inhibit these interactions in the absence of human tumor stroma. Anatomical context can also have significant impact on PDX tumor biology. PDX can be established as subcutaneous flank tumors or at any of a variety of orthotopic sites. Lung PDX are amenable to orthotopic growth in the lung, and can thus be used to model lung cancer growth and metastasis within its normal anatomical context. Orthotopic tumors can have vastly altered metastatic potential and organ preference as well as differential response to anti-cancer therapeutics in comparison to subcutaneous flank tumors. One of the primary limitations of PDX as a model for cancer is the need to use immunocompromised mouse hosts. The degree to which the immune system can be modeled with PDX is dependent on the mouse host chosen and the type of human immune cells used to reconstitute the human immune cell component. PDX can be established in a great variety of different strains of immunocompromised mice including athymic nude mice, severe combined immune deficiency (SCID) and non-obese-diabetic (NOD)-SCID strains, Rag null strains, and profoundly immunocompromised strains in which IL2-Rγc has been disrupted (NSG, NOG, BRG). Each of these strains differs with respect to the type and function of hematopoietic cells. For example, athymic nude mice have intact natural killer (NK) cells, whereas NSG mice do not; NSG mice therefore develop primary tumors and metastases at a much faster rate. Humanized mice are immunocompromised mice that have partially reconstituted human immune components for the purposes of modeling the behavior of the human immune system in a cancer context. NSG, NOG, and BRG mice have been engrafted with isolated peripheral blood mononuclear cells (PBMC) or tumor infiltrating lymphocytes (TIL) primarily to study mechanisms of lymphocyte recruitment; however, a major limitation of this approach is the rapid onset of graft versus host disease in the mouse host. Alternatively, different strains of mice can be engrafted with human CD34+ hematopoietic stem cells (HSC), which results in mouse hosts with fully human lymphocytes, monocytes, dendritic cells and in some cases NK cells. A variety of mouse strains engineered to express human cytokines such as IL-3 and GM-CSF have been developed to promote improved functional human immune system components. PDX models of lung cancer are growing in complexity, variety, and sophistication. These in vivo cancer models will be an integral component in a suite of tools for studying many aspects of lung cancer biology in a research environment. Recent advances in the humanization of mouse hosts promises to expand the possibilities of studying cancer immunology and immunotherapy of human tumors in an experimental setting in vivo.