From spheroids to organoids: next-generation models for CAR-T cell therapy research in solid tumors

Abstract

Chimeric Antigen Receptor T-cell (CAR-T) therapy is a revolutionary immunotherapy involving the genetic modification of T cells to express chimeric receptors targeting specific tumor antigens. Over the past decade, CAR-T therapy has significantly advanced with the development of five generations of CAR-T cells, each introducing modifications to enhance T cell efficacy, persistence, and the ability to overcome immune evasion mechanisms. The manufacturing of CAR-T cells has also evolved, employing techniques such as viral vector transduction or CRISPR-based gene editing, lipid nanoparticle, or transposon mediated approaches, to optimize their function. However, the development of CAR-T therapy for solid tumors faces significant challenges, primarily due to the hostile tumor microenvironment (TME), which traditional two-dimensional (2D) culture systems fail to accurately replicate. This review explores the potential of three-dimensional (3D) culture models, including spheroids and organoids, as tools for studying CAR-T cells in the context of solid tumors. Unlike 2D models, 3D systems offer a more physiologically relevant environment, better mimicking the TME, tumor heterogeneity, and immune interactions which CAR-T cells must encounter. We examine the advantages and limitations of 2D versus 3D models and discuss four key methods for generating spheroids/organoids: direct cell aggregation, scaffold-based, microfluidic, organs-on-chip and bioprinting, and patient-derived organotypic tumor approaches. Moreover, we explore the use of murine models in preclinical CAR-T research, highlighting their role in studying the dynamics of CAR-T cell trafficking, efficacy, and off-target effects. While CAR-T therapy has shown impressive success in some hematological malignancies, there is still a critical need for improved models to study CAR-T efficacy against solid tumors, particularly in relation to the TME. 2D models remain a valuable tool but should be combined with 3D models and in vivo murine studies for more accurate clinical outcome predictions. As we advance toward preclinical and clinical applications, ongoing efforts to develop and refine 3D culture systems are essential for overcoming the unique challenges of CAR-T therapy in solid tumors.

Keywords: 3D culture; car-t; chimeric antigen receptor T cells; immunotherapy; organoid; solid tumor; spheroid; tumor microenvironment.

Figure 1

Schematic representation of the five distinct generations of CAR. The first-generation CAR construct consists of an extracellular antigen-recognition domain linked to the intracellular CD3ζ signaling domain. The second-generation contains a co-stimulatory domain (CD28 or 4-1BB) added alongside CD3ζ. The third-generation CAR incorporates more than one co-stimulatory domain. The fourth-generation secretes cytokines upon activation. The fifth-generation CAR incorporates a truncated cytoplasmic domain of the IL-2 receptor β-chain, coupled with a STAT3-binding motif between the co-stimulatory domain and the CD3ζ signaling domain to activate the JAK-STAT. This figure was adapted and generated on Biorender.

Figure 2

CAR-T manufacturing processes. The first method to generate CAR-T cells is the CRISPR-Cas9 genome editing followed by viral transduction or the direct adenoviral, retroviral or lentiviral transduction of the CAR sequence (I). The second method includes the CAR construct delivered using lipid nanoparticles (II). The third method involves mRNA electroporation transposon-mediated approaches to integrate the CAR sequence into T cells (III). This figure was generated on Biorender.

Figure 3

Comparison of 2D vs. 3D co-culture system. In a 2D co-culture system containing a monolayer of tumor cells and fibroblasts, CAR-T cells can interact directly and this could be observed in real-time while evaluating cytokines production, cytotoxicity and proliferation without external factors. However, this system does not represent the microenvironment and the migration of tumor cells or CAR-T cells cannot be followed according to gradient of oxygen or nutrients. In a 3D co-culture system, CAR-T migration, infiltration, cell-cell-interaction and tumor killing can be monitored between the several layers of cells with or without gradients of oxygen and nutrients. This figure was generated on Biorender.

Figure 4

Differences between monolayer 2D cultures, spheroids and organoids. Monolayer cultures consist of single cell type or multiple cell types while spheroids and organoids are composed of multiple cell types. Additionally, organoid structure requires at least epithelial or mesenchymal cells. The architecture complexity increases from monolayers (flat structure) to spheroids (tissue-like aggregates) and organoids (organ-like structures). Furthermore, organoids are composed of monolayers or spheroids composed of monolayers. This figure was generated on Biorender.

Figure 5

Schematic representation of direct spheroid formation. First method is the liquid overlay technique where cells aggregate together spontaneously to form a spheroid after being seeded. Second method is the hanging drop technique using a U-bottom well to facilitate natural aggregation. Third method is the same as the first one while cells are seeded by growth factors. Fourth method is the use of ultra-low or -low attachment plate with a specific coatting to avoid cell adherence to plastic. This figure was generated on Biorender.

Figure 6

Schematic representation of using matrix coupled with microfluidic, organ-on-chip and bioprinting. Microfluidic device is composed of different channels, one to culture the spheroid or organoid with CAR-T cells while other channels can bring oxygen, nutrients, chemotaxis molecules or growth factors or can mimick blood vessels. Channels can be coupled with filters to remove debris and dead cells. This figure was generated on Biorender.

Figure 7

Schematic representation of patient-derived organotypic tumor preparation. Solid tumors are extracted from human patient organ biopsies and cultured on a matrix. Extracted tumor can be sliced depending on their size and can be cultured as organoids or spheroids before injecting CAR-T cells. This figure was generated on Biorender.