**1. Introduction**

Historically, the objective of tissue engineering and regenerative medicine (TERM) has been to apply the principles of engineering and life sciences to the development of biological substitutes that restore, maintain, or improve the function of a tissue or whole organ [1]. While this objective remains intact, the focus in the field has been extended to the implementation of engineered tissues that will never be implanted into patients, but will transform the way we study human tissue physiology in vitro [2–5]. Each tissue and organ is unique and has well defined functions, anatomy and cellular, molecular and soluble components. In vivo, individual cells are harbored in specific niches where they integrate many external cues (including those that arise from extracellular matrix (ECM), mechanical stimulation and soluble signals from adjacent and distant cells) to generate a basal phenotype and respond to perturbations in their environment. The development of 3D platforms with well-defined architectures resembling native cellular environments has contributed to significant advances, among other tissues, in liver or heart modeling [6–8]. The integration of three dimensionality, multi-cellular interactions, patient-specific polymorphisms, fine control of chemical parameters (pH, oxygen level, biochemical gradients) and ECM composition are the main assets of this engineered tissues [4,9,10].

#### **2. Modeling Solid Tumors in Vitro**

Cancer is a heterogeneous dynamic disease in which the associated stroma plays a critical role as a pro-tumorigenic environment, drug desensitization inductor and drug penetration barrier [11]. 3D engineered cancer models have been used to overcome major issues of conventional 2D planar cultures and animal models. The average success rate for candidate drugs in translating from animal models to clinical cancer trials is less than 8% [12]. Biological di fferences among humans and animals limit their ability to mimic complex processes such as carcinogenesis and tumor physiology, progression and metastasis. Mice are the most frequently used animal models. Crucial genetic, molecular, immunologic and cellular di fferences between mice and humans prevent them from serving as e ffective models [9]. Significant progress has been made, such as humanizing mice by transplanting human cells or obtaining patient-derived tumor xenografts (so called PDTX or avatar mouse). Nevertheless, such models are still challenging and expensive to adopt for routine use. Furthermore, fundamental di fferences in telomerase regulation between rodents and humans [13] have raised questions regarding the reliability of transgenic and inducible mouse cancer models, and discrepancies between certain rodent and human cytokines generate uncertainty for mouse models [9,14]. 2D planar cultures' lack of architecture, cell-cell and cell-ECM interactions, and the exposure of cells to high-sti ffness substrates like culture plates could a ffect cell behavior in terms of gene expression profile and drug sensitivity. For example, the PI3K–AKT–mTOR pathway is a central regulator of cell growth, proliferation, survival, metabolism and aging. Riedl et al. reported significant di fferences in mTOR activity and crosstalk between AKT-mTOR-S6K and the MAPK pathway in spheroids vs. planar cultures of colorectal cancer Caco-2 cells, including alterations in the responses in treatments with inhibitors of AKT, mTOR and S6K axis or of the MAPK (ERK) axis, which are ongoing pharmacological targets [15]. Moreover, the role of specific ECM signaling in regulating gene expression and cell fate has been largely validated as a pivotal agen<sup>t</sup> in cancer progression and drug resistance. The attachment of tumor cells to the ECM may trigger cell adhesion-mediated drug resistance (CAM-DR). Several receptors such as integrins and their ligands, including fibronectin (FN) or hyaluronic acid (HA), are involved in this process. The interaction between α4β1 integrin on tumor cells and FN induces progressive drug resistance in chronic lymphocytic leukemia (CLL) and acute myeloid leukemia (AML). Also, β1 integrin-mediated PI3K activation overrides treatment-induced cell cycle arrest and apoptosis in various solid tumors [16]. As the specific crosstalk between a given cancer and its stroma varies for each cancer type and perhaps for each patient, in vitro models that better reflect the in vivo human environments and their heterogeneity may provide more accurate indications of patient outcome [17,18].

TERM has been used to explore several approaches for modeling solid tumors (Table 1). Sca ffold-free models such as spheroids and organoids have achieved grea<sup>t</sup> in vitro results [15,19]. Organoids, cell aggregates deriving from one or several stem cells able to self-organize and phenocopy essential aspects of the organs they derive from, are of grea<sup>t</sup> interest from the point of view of drug testing, as they are easily compatible with high throughput screening technologies (HTS). Genetic modification of organoids allows disease modeling and organoids can be grown from patient tumor tissues (tumoroids) and recapitulate better native tumors arising in superior models for patient-specific drug testing [20]. Other approaches incorporate polymeric substrates with tunable composition, sti ffness or functionality into the equation, as in sca ffold or hydrogel-based models [21,22]. Including bioreactors and perfused microfluidic chambers gives strict control of oxygen, temperature, pH or nutrients and precise spatiotemporal control over gradient formation [23]. Advances in bioprinting techniques endow tumor-on-a-chip models with specific properties such as anisotropy or complex physiological architecture [22,23]. More recent approaches integrate the above-mentioned features in realistic systems that can even include cancerous vascularized tissues embedded in chemico-physically defined environments with ECM and healthy neighboring cells under dynamic perfused conditions.


**Table 1.** Summary of different TERM approaches for modeling solid tumors.

human colonic microvascular endothelial cells.

#### *Int. J. Mol. Sci.* **2020**, *21*, 5747
