Preclinical Models and Imaging Modalities of the Tumor Microenvironment in Metastasis

Editorial

This collection portrays systems beyond the traditional cancer research experimental methods for the study of the tumor microenvironment, an essential element for tumor growth and metastasis. Such systems are tridimensional, involve high-depth and high-resolution imaging (often over time, thus requiring complex imaging analysis tools), and use patient derived tissue and/or employ synthetic or de-cellularized scaffolds. 

For diseases with unmet clinical needs, such as pancreatic, colorectal, and brain cancer, in which the translatability of mouse syngeneic or xenograft models is low, the establishment and expansion of biobanks is essential. Bürtin et al.¹ provide a guide into how they accomplish this, including the expansion of donor material into patient derived xenografts (PDXs), thus establishing a living biobank that closely recapitulates patients’ histology and disease courses. Through constant protocol optimization and close collaboration and communication with surgical staff, the authors report outgrowth rates of 63% for colorectal PDXs and 48% for pancreatic PDXs. This progress falls short of being useful as a test-treatment proxy for individual patients in immediate need of treatment options, solely due to the long period of time required for model establishment. It is nonetheless useful to better understand tumor biology and help future patients with similar tumor characteristics. Additionally, Tew et al.² describe the establishment and use of central nervous system (CNS) metastatic PDXs, highlighting the importance of the route of tissue inoculation for successful outgrowth and study purposes.

Alternatively to PDX models, patient derived tumor slice ex vivo cultures provide a potentially quicker and more economical system, with a native matrix and architecture—both multicellular and heterotypic—which can be used to test drug responses. Braun et al.³ present their organotypic slice cultures (OTSCs) for pancreatic primary tumors and metastases. They convey a method from xenograft establishment and monitoring, including: manufacturing of a device to support live imaging of the ex vivo slice; tumor slice collection; slice culture and ex vivo growth; and imaging and image analysis. In another example, Ciraku et al.⁴ provide a comprehensive method to study brain metastases and their treatment through the use of ex vivo brain metastatic tumor slices.

Tumor adjacent tissue and extracellular matrix are also key to fully understand metastasis. Jelinek et al.⁵ provide a PDX to study cancer associated fibroblasts’ impact on tumor growth, using patient derived dermal fibroblasts to co-inject with human melanoma cell lines. Mayorca-Guiliani et al.⁶ describe a method to decellularize mice heart and lungs and generate an intact scaffold, which may be used to seed cells, and more importantly, to understand the location/topology and extracellular matrix (ECM) composition of these organs. The lungs are particularly a very common site of metastasis for several tumor types. This method may also be used to generate other ECM scaffolds. Collagen is one of the major components of the ECM, and topological changes in fibrillar collagen have been associated with disease progression. Liu et al.⁷ provide an open-source image analysis tool to quantify fibrillar collagen organization from high-dimensional, high-resolution confocal tomography images. Part of the relevance of the ECM is its adaptability to external factors, such as pressure. The ECM adopts different biophysical properties in response to these external factors that can impact tumor growth. Using collagen and fibrinogen gels, Calitz et al.⁸ introduce a biomimetic model for liver cancer that mimics its ECM tunability in a 3D environment. The authors create experimental conditions that recapitulate the biophysical properties of the fibrotic and cirrhotic liver, as well as the hepatocellular carcinoma (HCC) to study tumor-stroma interactions in the HCC. The created biomimetic models (covering a wide range of stiffness) exhibit more representative treatment responses when compared to 2D cultures.

Other heterotypic approaches to mimic metastasis include organoid cultures, organ on chips, and/or microfluidic co-culture systems, which are evolving into complex systems to accommodate the co-culture of immune and cancer cells with very different culture condition needs. Chakrabarti et al.⁹ present an organoid system that includes T and myeloid cells, as well as human-derived autologous gastric cancer cells in an organoid/immune cell co-culture which serves as a preclinical model to predict targeted therapy efficacy. De Ninno et al.¹⁰ describe a similar concept using microfluidic co-culture systems compatible with dynamic and multiparametric measurements. The manufacturing, seeding, imaging, and analysis of cancer cells and the tissue microenvironment (TME) of a novel 3D microfluidic blood brain niche (µmBBN) is described by Oliver et al.¹¹. The authors show the use of artificial intelligence to identify the phenotypic differences of cancer cells that are capable of surpassing the blood brain barrier, and assign those cells an objective brain metastatic potential. This can be used to answer basic and translational questions about metastasis, therapeutic efficacy, and the role of the TME in both.

Collectively, the authors contributing to this collection provide the TME and metastasis field a range of innovative, ready to use methods to answer key questions that 2D monotypic in vitro cultures cannot address. Ultimately, the wide implementation of these methods will improve our understanding of tumor ecosystems, and therefore help in developing TME targeted therapies against metastatic disease, a condition that remains a challenge to treat and with a high impact on patients’ survival and quality of life.