Abstract 2035: A 3-dimensional tumoroid model made up of lung cancer cells, fibroblasts, and endothelial cells is predictive of drug activity in animal models

Abstract

Abstract Lung cancer is a very heterogeneous type of cancer. Among various subtypes, non-small-cell lung cancer (NSCLC) comprises about 80 % of all cases, and the KRAS mutant subtype, has a generally poor prognosis and response to chemotherapy or targeted therapeutics. There is an urgent unmet medical need for development of effective treatments. Anticancer drugs are often selected based on studies performed on adherent cultures of tumor cells where the cells grow in two dimensional monolayers. However, tumors grow in 3 dimensions and are not composed exclusively of malignant cells, but of a mixture of cancer cells and stromal cells embedded in an extracellular matrix. These interactions involve a complex mechanical and biochemical interplay that is missed when cells are cultured in classical monolayer-based models. NSCLC 3-D heterotypic models that preserve epithelial and stroma cell interactions in a well-defined 3-D microenvironment should more accurately identify drug candidates for lung cancer treatment. We developed a 3-D heterotypic in vitro lung tumor model that recapitulates the human tumor microenvironment and is useful for selection of novel anticancer drugs. This in vitro 3-D KRAS mutant NSCLC model includes H358 epithelial cancer cells, stromal cell types such as lung cancer-associated fibroblasts (CAFs), and microvascular endothelial cells (HMVECs) cultured on a relevant cellular substrate component such as collagen type I. Each cell type was fluorescently labeled and the cultures grew to form what we named a “tumoroid” structure. The tumoroid is a palpable 2 mm3, spheric, dense structure that resembles the architecture of a human lung tumor and has features such as a capillary-like vasculature structure and areas of epithelial/stromal interaction as confirmed by microscopy and H&E staining. This model was used to test agents impacting known pathways in the tumor and stromal microenvironment to determine their effect on tumoroid viability and apoptosis. Test compounds included a DR5 agonistic monoclonal antibody, CNTO 95 (pan αv integrin monoclonal antibody), Cetuximab (EGFR monoclonal antibody) and Bevacizumab (VEGF monoclonal antibody). Cetuximab, CNTO 95, and DR5 significantly inhibited H358 tumoroid viability, and Bevacizumab disrupted the tumoroid capillary-like structures. Conversely, Cetuximab and CNTO 95 did not inhibit H358 cell viability when grown in 2-D cultures and Bevacizumab did not have an impact on H358 cells grown on plastic. However, previous publications have shown that Cetuximab and Bevacizumab inhibit H358 tumor growth in animal models. The results suggest that the in vitro tumoroid model is more predictive of the activity seen in animal models than 2-D cell culture techniques. The tumoroid model may therefore be useful for anticipating the effectiveness of new drugs in vivo and could reduce the need for evaluating numerous compounds in animal models. Citation Format: Luciana F. Macedo, Elizabeth Kaiser, Bradley Heidrich, Barbara Bushey, Catherine Ferrante, Deborah Marshall. A 3-dimensional tumoroid model made up of lung cancer cells, fibroblasts, and endothelial cells is predictive of drug activity in animal models. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2035. doi:10.1158/1538-7445.AM2014-2035