Chapter 1: Tissue Engineering Models for Cancer Pathology

Despite recent advances in treatment, cancer remains a leading cause of death in developed countries.¹  Whereas molecular therapies²  and immunotherapies³  offer great promise, our understanding of cancer biology and its translation into viable therapies is still emerging. Most candidate cancer therapeutics begin their journey in petri dish studies (the current gold standard) using cultivated cancer cell lines. Successful drugs then advance to animal studies where they are further validated and tested for safety and efficacy. Ultimately, a select few compounds advance to the patient testing stage, where only ∼37% of compounds perform as in animal studies⁴  and the chance of overall success is only ∼25–50%.⁵ 

There are several factors that limit successful therapy development. Cell lines employed in petri dish studies may bear little genetic and phenotypic resemblance to patient tumors.⁶  Additionally, cell culture models typically contain a single cell type that does not capture the rich heterogeneity of tumor tissue. Petri dishes often do not include cues from other cell types, the immune system, mechanical and topographical effects of the local tissue anatomy, or support vascularization critical to tumor development. Thus, many features of cancer biology are lost in petri dish models. Animal models have greater success in mimicking human cancer pathology. However, apart from ethical concerns, animal–animal variation, and the costs of animal husbandry, animal models have their own limitations. These models often employ artificially implanted xenograft tumors,⁷  rather than genetic models that promote spontaneous tumor development with greater homology to human pathology. Because these models usually employ immunocompromised mice, they fail to reproduce cancer-specific immune responses.⁸  Animals also have a much shorter lifespan than humans, which can limit the ability to probe the biology of complex disease processes, such as metastasis. Models that successfully bridge the gap between petri dish and animal studies would greatly improve our understanding of cancer biology, potentially leading to improved therapies at reduced cost.

Any such models must have the ability to recapitulate features of the tumor microenvironment (TME). The concept of the TME originates from the “Soil and Seed” hypothesis developed by Stephen Paget in the 1890s.⁹  In this theory, it is suggested that both the “seed”, i.e., the cancer cell, and the “soil” in which it grows (i.e., the surrounding tissue) contribute to metastasis and cancer progression. Unfortunately, this theory lay mostly dormant until the 1970–1980s when researchers began to explore the roles of the immune system and vasculature in tumorigenesis and cancer progression.¹⁰  Researchers discovered that cancer is influenced by a vast host of cues extending beyond the cancer cells themselves. This “soil” includes not just the physiological tissue in which cancer cells find themselves, but also the immune cells, normal cells, and rich chemical signaling between them. Physical and topographical anatomy was found to be important. The concept of tumor detection via palpitation can be directly related to observed tissue stiffening in regions of a tumor.¹¹  These mechanical changes have been shown to form part of a feedback loop in which cell signaling is altered. Similarly, cancer cells have been shown to recognize and exhibit preferences toward certain topographical structures. For example, in the brain, glioblastoma (brain cancer) cells migrate preferentially on fibrous structures and flat sheets, known as Scherer's structures because of their identification by Hans-Joachim Scherer in 1938.¹²  The TME also includes other anatomical structures and their corollary effects, such as the tumor vasculature and its associated angiogenesis, leaky vasculature, and increased interstitial fluid pressure. Growth of the tumor generates yet other changes to its local environment, including compressive solid stress, hypoxia, and a necrotic core. These macroscale effects induce changes in chemical signaling that initiate genetic and epigenetic changes and an altered cell phenotype that can impact tumorigenesis, tumor cell invasion and proliferation, metastasis, and resistance to therapy.

The TME is now increasingly recognized as an important contributor to cancer progression and treatment. Particularly, understanding the feedback loop between TME and cancer cells may be critical to therapy development. However, two-dimensional (2D) tissue culture studies do not capture the rich chemical, topographical, and mechanical characteristics of the tumor niche. In contrast, animal models present many TME cues, but offer little control of the local microenvironment. Genetic engineering may be employed to investigate specific signaling mechanisms active in cancer.¹³  Specific cell or tumor types can be grafted into host animals to recreate the desired cell–cell interactions.⁸  However, it is difficult to isolate specific variables that may be interrelated, such as cell–cell signaling, angiogenesis, and increased interstitial fluid pressure, which may all co-exist in the same animal model. Similarly, it may be challenging to decouple macroscale anatomical variables, such as angiogenesis and hypoxia induced from tumor growth. There is a unique and important niche for TME models that go beyond 2D petri dish studies, yet provide the engineering control that is difficult to provide in animal models.

Biomaterials research affords a pathway to TME models that close the gap between 2D petri dish and animal studies. Biomaterials research has arisen through the twin pillars of materials that promote wound repair (i.e., tissue engineering and regenerative medicine) and materials that evoke disease pathology enabling ex vivo study. Biomaterials have advanced significantly from their earliest modern implementations that focused primarily on form and function over biomimicry.¹⁴  Early use of synthetic materials, such as titanium rods for bone repair or poly lactic-co-glycolic acid (PLGA) scaffolds, has given way to use of natural materials, such as hydroxyapatite and collagen. These advances in biomaterials science have improved understanding of human biology and disease pathology. For example, the observation that cells spread preferentially on stiffer substrates has led to the field of cancer mechanobiology.¹⁵  Thus, the concept of using biomaterials to explore TME is a natural outcome of systematic development of this field. Although growth of biomaterials-based TME models has exploded in recent years (Figure 1.1), TME model development has its earliest roots in studies dating back decades.

The concept of matrix engineering was introduced in 1971, with a goal of applying this approach for tissue regeneration applications.^16,17  In the late 1980s and early 1990s, the concept of tissue engineering began to gain traction (Figure 1.2). Specifically, the term “tissue engineering” first originated in 1987 at a National Science Foundation bioengineering panel meeting, and subsequently, a definition was proposed at the first tissue engineering meeting in 1988.^18,19  A primary strategy in tissue engineering involves combining cells with biomaterial scaffolds in which they are cultured to create a tissue engineered construct for implantation in vivo. Examples demonstrating this concept included degradable polymeric scaffolds for cell transplantation in the early 1990s;²⁰  as well as the famous ear grown on the back of a mouse reported in 1997.²¹  Around the same time, pioneering work by Prof. Bissell in 1992 demonstrated the ability of three-dimensional (3D) culture environments comprised of basement membrane to distinguish normal versus malignant breast epithelial cells²²  (Figure 1.2). Importantly, the malignant phenotype observed in this system could be reversed by inhibition with antibodies that neutralized matrix binding structures,²³  demonstrating the impact of the microenvironment. This set the stage for the development of TME models, including those incorporating non-malignant stromal cells present in the TME^24,25  (Figure 1.2). Other natural materials including 3D Collagen-I matrices were also beginning to be utilized in this time frame to study cell–matrix interactions in breast cancer.²⁶  The fundamental idea of a tissue engineered construct has been applied to create physiologically relevant in vitro model systems to study the TME and various aspects of cancer progression.

Single cell patterning techniques were reported as early as the late 1960s.²⁷  However, the ability to micropattern, wherein cells could be deposited in regions defined by specific shapes, was not fully achieved until the mid-1990s,^28–30  and, around the same time, dynamic imaging technologies were being used to observe cell–matrix interactions in vitro³¹  (Figure 1.2). Approximately five years later, the ability to pattern cells and proteins using microfluidic technologies was demonstrated.^32,33  Collectively, these advances paved the way for tumor-on-a-chip platforms and thus have been instrumental in studying the TME. Examples of tumor-on-a-chip platforms appeared as early as 2010³⁴  and, now, such platforms exist for a variety of cancers. Advances in increasing platform complexity, such as the incorporation of multiple cell types present in the TME, including immune cells, are expected in the coming years.

The study of TME has significantly benefitted from advances in tissue engineering and regenerative medicine. For example, the ability to precisely engineer matrices using natural and synthetic polymers or their combination, rather than employing available natural matrices, has been applied since the early 2000s to study cancer cell–matrix interactions in vitro³⁵  (Figure 1.2). Accordingly, imaging tools have evolved to dynamically image interactions of tumor cells with those matrices, which have since been applied to understand how tumor cells interact with their surrounding matrix.³⁶  Technological advances in creating complex matrices for tissue engineering using bioprinting have also been applied for TME models since early 2010³⁷  (Figure 1.2). Such techniques provide many advantages for TME studies, including the potential for high throughput, which can be subsequently employed for various applications such as drug screening. Finally, the need to increase TME model complexity resulted in organoid models beginning in the early 2010s^38,39  (Figure 1.2), and examples of such models are already reported in the literature for several cancer types. Such models closely recapitulate several features observed in tumors in vivo.

This book will explore the application of bioengineering strategies, including scaffold design and synthesis, chemical signaling and delivery, and co-culture, microfluidics, and organ-on-a-chip tools to study the TME.

Chapter 2, by Witz and Israely introduces the concept of the TME, including the reciprocal nature of interactions between the normal and tumor cells, the extracellular matrix (ECM) surrounding them, and the chemical and mechanical signaling present. The chapter also discusses types of cells residing in the TME, and how TME can be leveraged for new therapies.

Chapter 3, by Nain and Sharma, discusses fibrous ECM components and their importance in the TME. This chapter introduces methods to mimic fibrous ECM that are expanded upon in later chapters, including micropatterning (Chapter 7), photolithography (Chapter 7), electrospinning (Chapter 8), and hydrogel scaffolds (also discussed in Chapters 4–5). The chapter ends by discussing cell responses to fibrous ECM components, including cell–ECM mechanics and cell adhesion and migration.

In Chapter 4, Mierke discusses mechanical features of the TME, including how cell types in the TME contribute to these features that subsequently influence tumor progression. Various in vitro biomaterial-based strategies employed to mimic the mechanical features of tumors, such as hydrogels, are discussed. In addition, the utilization of tumor mechanics concepts for developing better therapeutic approaches is highlighted.

In Chapter 5, Pradhan and colleagues discuss how scaffold chemistry can be employed to study the TME. The composition as well as modification of both naturally-derived scaffolds and synthetic scaffolds (e.g., hydrogels) to recapitulate the TME are discussed. In addition, recent advances enabling spatiotemporal modulation of biomaterial properties to study the dynamically evolving TME is presented.

In Chapter 6, Curvello and Loessner discuss approaches for mimicking multi-cellular features of the TME. Strategies for deriving biomaterial-based culture platforms incorporating various cell types present in the TME are presented. The ability of such platforms to mimic key features of the TME including vascularization, metabolic and inflammatory profiles are highlighted in this chapter.

Chapter 7, by Nelson et al., discusses methods to pattern cells within or on scaffolds. The chapter reviews cell types and cell–cell and cell–ECM interactions prevalent in the TME. Techniques for patterning 2D scaffolds, including self-assembled monolayers and microcontact printing are discussed. Then, methods for 3D cell culture and patterning, including methods for cell spheroid formation, are introduced. The chapter concludes with applications of lithography to cell patterning.

Chapter 8, by Zhao, focuses on electrospinning methods of scaffold generation. The chapter reviews the history and principles of electrospinning, materials employed in fibrous scaffold generation, and fiber geometries and orientations. The chapter concludes with a discussion of electrospun fibrous scaffold use in cancer research, including studies of the TME and cancer biology, detection, and treatment.

In Chapter 9, Scott and Guelcher discuss techniques for TME scaffold fabrication using additive manufacturing. The advantages and disadvantages, along with design considerations, and biomaterials employed for various techniques such as stereolithography, fused deposition modeling, selective laser sintering, 3D printing, and bioprinting are discussed. In addition, bioprinting techniques including inkjet, extrusion, stereolithography, and laser-assisted printing and their utility in constructing TME models are highlighted.

Chapter 10, contributed by Song et al., describes applications of microfluidics to TME biomaterial models. This chapter begins with a discussion of flow and gradient based features in the TME, such as O₂ and interstitial fluid pressure gradients. Methods to integrate microfluidic devices with TME scaffold materials and to establish specific gradient types are provided. The chapter concludes by detailing methods that use flow to orient specific ECM features to more closely mimic microarchitectures found in the TME.

In Chapter 11, Soker and co-workers discuss modeling of the TME in tumor organoids. Techniques used to fabricate tumor spheroids as well as organoids, along with their advantages and disadvantages are highlighted. In addition, the utility of tumor organoid models to study the TME in various types of cancer are presented.

Chapter 12, by Walsh and colleagues, discuss advances in imaging technologies crucial for evaluating cell behaviors in these complex models. The chapter begins by detailing the challenges in complex scaffold imaging. Then, sample preparation methods to overcome some of those limitations are presented. The chapter discusses several imaging methodologies, including optical coherence tomography (OCT), confocal fluorescence microscopy, light sheet microscopy, multiphoton imaging, and magnetic resonance imaging (MRI). The chapter concludes with a description of methods for image analysis.

In Chapter 13, Raghavan et al. discuss the use of TME models in the field of immuno-oncology. The chapter begins by discussing immune cell types prevalent in the TME, and then introduces the concepts behind current and proposed immunotherapy treatments. The chapter concludes by discussing 3D models used in immuno-oncology research, including scaffold-free systems, hydrogels, organoid models, decellularized scaffolds, microfluidic models, and scaffolds generated by additive manufacturing.

In Chapter 14, Sempertegui and Fischbach discuss tissue engineered models of metastasis, with a focus on bone metastasis. The bone metastatic cascade is described highlighting the importance of the ECM as well as cellular interactions. Biomaterial-based strategies incorporating these components to mimic key features of the bone tissue to study tumor cell dissemination, dormancy, and metastatic outgrowth are described along with recent advancements in model development. Tissue engineered models provide a valuable tool to study not only bone metastasis, but also metastasis to other organs including the liver, lung, and brain.

In Chapter 15, McGuigan and co-workers discuss the application of various tissue engineered cancer models for drug screening. The various steps involved in utilizing such in vitro models for anticancer drug screening is discussed, including selection of the tumor type, setting up the culture model by identifying the key components of the TME, model validation, model scale up, followed by high throughput screening. Finally, the challenges involved with each of these steps are highlighted along with future directions.

The development of new therapies depends on the success of pre-clinical research models that can reduce animal requirements and de-risk the clinical trials process. It has become increasingly clear that 2D petri dish models alone are not sufficient to meet these needs. Building on the rich histories of tissue engineering, regenerative medicine, cell patterning, and 3D imaging, biomaterials scaffolds that mimic tumor features offer a path forward. Although these materials are rooted in the application of simple, natural hydrogels, they have matured into systems capable of capturing the complexities of the TME. TME models can be manufactured using natural or synthetic materials, whose chemical and mechanical properties can be engineered to match organ specific TME features. Scaffolds can be manufactured with complex geometries and seeded with a variety of cell types in designated orientations. These models can be integrated with flow via microfluidics, and with each other to generate organ-on-a-chip models. Advances in imaging technologies allow researchers to characterize cell–cell and cell–ECM interactions in situ, painting an increasingly detailed portrait of the interactions in the tumor milieu. These models are already enabling new insights into immuno-oncology, metastasis, and high throughput drug screening. As the capabilities of biomaterial TME scaffolds continue to improve, it is easy to envision a future in which such materials could be used in personalized medicine approaches to precisely tailor treatments to a given patient based on their own cells and tissue characteristics. Similarly, improved TME models will accelerate the cancer biology and drug development pipelines, enabling new therapies to reach the market faster and with lower development costs. Thus, biomaterial TME models are rapidly becoming part of the pre-clinical cancer research portfolio, taking their place alongside 2D petri dish and animal models.

S. R. acknowledges research support from the National Science Foundation (CBET 1749837), the American Cancer Society (RSG-21-032-01-CSM), METAvivor, and the Breast Cancer Research Foundation of Alabama. The authors thank Prof. Sanjay Kumar (University of California, Berkeley, CA, USA) for his input on the timeline presented in Figure 1.2.