Introduction

Since the pioneering work of Sato et al. [1] in constructing colorectal cancer organoid models, there has been significant progress in the development of tumor organoids. The establishment and advancement of tumor organoids have provided a novel approach for creating more physiologically relevant human tumor models. In comparison to commonly used cancer models like cancer cell lines and primary patient-derived tumor xenografts (PDTXs), tumor organoids offer distinct advantages. The advantages have resulted in their utilization across diverse domains, encompassing drug screening, genome editing, and oncogene identification [2].

Chimeric Antigen Receptor T-Cell Immunotherapy (CAR-T therapy) has garnered significant interest as a burgeoning treatment for tumors, demonstrating promising outcomes in the management of specific hematological tumors such as B-cell acute lymphoblastic leukemia (B-ALL), B-cell non-Hodgkin lymphoma (B-NHL), and multiple myeloma (MM) [3,4,5]. The scientific community has shown considerable attention towards the efficacy of CAR-T therapy in addressing hematological tumors, perceiving it as a potential remedy for solid tumors. Nevertheless, the effectiveness of this therapy is constrained by factors such as tumor heterogeneity, limited transport and infiltration of CAR-T cells into tumor tissues, and the presence of immunosuppressive microenvironments within the tumor [6]. These crucial concerns emphasise the necessity of preclinical research models for CAR-T therapy.

Tumor organoids can preserve primary tumour tissue characteristics more completely, allowing for a realistic simulation of the interaction between tumour and CAR-T cells in vitro. Researchers have co-cultured CAR-T cells with tumor organoids to validate the anti-tumour effects of new target CAR-T cells, modified CAR-T cells, and CAR-T combination therapies.

However, there remains a significant amount of unexplored research potential within the currently utilized co-culture models of tumor organoids and CAR-T cells. This review summarizes a total of 10 research papers on the use of tumor organoids for CAR-T therapy from March 2019 to June 2023. At the same time, based on the new advances in organoid technologies from September 2014 to June 2023, optimization suggestions were proposed for this co-culture model from five perspectives: preserving or reconstructing the immune microenvironment, systematization, vascularization, standardized culture procedures, and expanding the tumor organoid resource library. All cited articles are sourced from the Pubmed database. We hope that this review will offer a fresh perspective to assist CAR-T therapeutics researchers in the successful implementation of this co-culture model in both fundamental and translational CAR-T therapies.

A basic model for co-culture of tumor organoids with CAR-T cells

As of June 2020, there are currently over 500 ongoing clinical trials investigating CAR-T therapy for various tumor types, with numerous additional trials in the developmental phase. The development of a practical, cost-effective, and realistic in vitro model for early CAR-T research is of utmost importance as it allows for the evaluation of the anti-tumor activity of CAR-T cells, a detailed exploration of the mechanism of action of CAR-T cells, and an enhancement of the structural design of CAR-T cells. Furthermore, there is a growing clinical demand for personalized medicine models that can predict the efficacy of CAR-T treatments and identify viable combinations of CAR-T therapy strategies. Co-culture models of tumor organoids have become a prominent choice owing to their remarkable resemblance to the original tumor in terms of histopathological morphology, genomic alterations, and expression profiles of specific markers.

Submerged Matrigel culture is a classic approach for organoid culture and is applicable to tumor organoids [7]. This method involves dissociating tumor tissue into a dispersed suspension of tumor cells using enzymatic or physical techniques. These cells are then embedded in a gel and placed in a culture medium to facilitate the growth of tumor organoids. The culture medium not only contains essential components for organoid growth but also includes pathway inhibitors and/or growth factors. And the additives and culture conditions being adapted based on the tumor organoids type [8,9,10]. Common additions comprise Wnt3a, R-spondin, epidermal growth factor (EGF), and Noggin—a bone morphogenesis (BMP) inhibitor which together promote stem cell growth, differentiation, and self-renewal [11] (Fig. 1)

Fig. 1
figure 1

Preparation process of tumor organoids and CAR-T co-culture basic model (by Figdraw.)

Full size image

It should be emphasised that the application of the Submerged Matrigel culture approach for tumor organoids is not fixed and should be customized to suit the unique attributes of the tumor organoids. Jacob et al. [12] generated glioblastomas organoids by directly culturing micro-dissected tumor fragments, as opposed to dispersing the tumor tissues into cell suspensions. Tumour fragments were manipulated directly to ensure the survival of sensitive cells and reproduced the hypoxic gradient, whilst preserving part of the mesenchymal component in tumor organoids. This facilitated the study of tumour cell-mesenchymal cell interactions. Fujiii et al. [10] discovered that genetic mutations in tumour cells had a tendency to impact the nature of the additive. The majority of colorectal cancer (CRC) organoids and all adenoma organoids were capable of autonomous growth without the need for exogenous R-spondin/Wnt3a. This can be attributed to the existence of mutations in one of the pivotal protein genes involved in the Wnt signalling pathway, which sustains the activation of said pathway. For example, variation-induced activation of TCF7L2, CTNNB1, APC. Hence, it is imperative to establish the requirement for growth factor/pathway inhibitors based on the genetic characteristics of the tumor organoids. Incorrect application of growth factor/pathway grafts may result in induced clonal selection of the tumor organoids, while the incorporation of incorrect components may complicate the interpretation of drug therapy [12]. The instructions/literature should be read in detail to determine the appropriate culture procedure and additives for the use of Submerged Matrigel culture for tumor organoids.

How can the anti-tumour activity of CAR-T cells be accurately analyzed in a co-culture model? Three studies offer valuable insights into this inquiry. Zou et al. [13] utilised a Caspase3/7 green fluorescent probe to label tumor organoid cells undergoing apoptosis, and also employed flow cytometry to quantitatively analyse CAR-T cells with elevated killing virulence expressing CD107a, IFN-γ. Schnalzger et al. [14] performed lentiviral transduction of luciferase/GFP into tumor organoids. They quantified the anti-tumour activity of CAR-T cells based on the remaining fluorescence intensity, which decreased as tumor organoid cells died. Additionally, Yu et al. [15] labelled CAR-T cells with CD8 and granzyme B, observing differences between the two markers in experimental and control groups. LDH, IL-2, TNF-α, and IFN-γ in the cell matrix were quantified. In brief, the evaluation of the anti-tumour activity of CAR-T cells comprises three perspectives: apoptosis of tumor organoids, killing activity of CAR-T cells, and the content of relevant cytokines in the cell matrix.

Current application of tumor organoids and CAR-T cell co-culture models

Validation of the anti-tumour effect of newly targeted CAR-T cells

CAR-T therapy has demonstrated promising outcomes in specific haematological malignancies, but its efficacy in solid tumours is considerably limited. Despite the identification of several potential targets, it is crucial to investigate novel targets in order to advance the field [16]. Therefore, there is an urgent need for a preclinical model that can faithfully replicate the unique surface markers found on human cells and can be readily constructed to assess the therapeutic efficacy of CAR-T on new targets. In this regard, tumor organoids offer a viable solution.

Yu [15] and Jacob [12] independently developed organoid models for bladder cancer and glioblastoma, and proved the degree of preservation of tumor organoids on biological characteristics of primary tumor tissue. Subsequently, they conducted co-culturing experiments by introducing MUC1-CAR-T cells and EGFR VIII-CAR-T cells to their respective tumor organoids, thereby verified the anti-tumorigenic effect of these two innovative CAR-T cell targets. Tumors overexpressing MET are generally insensitive to small molecule targeted drugs therapy. In light of this, Chiriaco et al. [17] devised two MET-CAR constructs with distinct structures specifically for MET-overexpressing tumours, and assess the anti-tumour effect of these two constructs by utilising different tumour organoid models that overexpress MET. Both types of MET-CAR-T cells can overcome resistance to small molecule targeted drugs against MET, and their anti-tumor activity is correlated with MET expression levels. Li et al. [18] co-cultured NSCLC organoids expressing B7-H3 with B7-H3-CAR-T cells to verify their anti-tumor activity prior to brain metastasis. Then, CCR2b was expressed on the CAR-T cell surface. Finally, it was confirmed in the patient-derived tumor xenograft (PDTX) model that its binding with CCR2 on the surface of tumor cells can promote CCR2b-B7-H3-CAR-T cells to penetrate the blood–brain barrier.

Optimise existing CAR-T cells

The identification of novel targets broadens the potential applicability of CAR-T therapy, and on this basis, it is also meaningful to optimise existing CAR-T cells in multiple ways to enhance their killing effect. The presence of tumor organoids also provides a platform for evaluating the anti-tumour effect of CAR-T after optimisation.

The optimization of CAR-T cells can be undertaken from various perspectives, often involving the modification of the CAR structure to enhance the cytotoxicity of CAR-T cells. Thokala et al. [19] replaced the single stranded fragment variable region (scFv) of CD19-CAR-T cells with monoclonal antibody (mAb) 806, which can target various EGFR mutants. In order to evaluate the anti-tumor efficacy of improved CAR-T cells, researchers co-cultured them with GBM organoid containing multiple EGFR mutations. The improved CAR-T cells successfully targeted and eliminated multiple tumor copies, significantly reducing the possibility of esca** tumors through antigen loss. Wang et al. [20] constructed a modified CAR-T cell targeting glypican-3 (GPC3) in hepatocellular carcinoma (HCC).This modification involved substituting the cd8α-derived hinge region in the conventional CAR structure with a 4-1bb-derived hinge region containing 11 cysteine residues. By co-culturing with HCC organoids, it was observed that modified CAR-T cells showed stronger effectiveness in inhibiting tumor growth.

Zou et al. [13] attempted to optimize CAR-T therapy by screening a subset of CAR-T cells with strong anti-tumor activity. Researchers divided HBVs CAR-T cells into two groups based on whether express CD39. Subsequently, co-culturing these populations with HBV + HCC organoid models revealed that CD39 + HBVs CAR-T cells exhibited more significant apoptosis induction in HCC organoids. Therefore, CD39 can serve as an indicator to distinguish the subgroups of CAR-T cells with stronger activity in HBV + HCC.

Qiao et al. [44] developed microorganoids spheroids (MOSs) using emulsion microfluidics. Additionally, an automated MOS seeding, processing and imaging system has also been developed. Based on this efficient therapeutic analysis platform, I believe that the aforementioned prospects will soon be achievable. Due to the presence of TME, which leads to a poor response of CAR-T therapy to solid tumours, our knowledge of the mechanisms by which TME affects CAR-T cells is currently limited. However, tumor organoids co-culture models with preserved TME offer a realistic platform for further in-depth research in this area. Inadequate local infiltration of CAR-T cells following infusion is a frequent issue encountered by patients undergoing CAR-T cell therapy. It is closely related to the presence of TME, but there is also a correlation between the physiological process of CAR-T cells drilling out of blood vessels. The phenomenon of vascular mimicry has been confirmed in various malignant tumors [45], so what impact will this phenomenon have on the infiltration of CAR-T cells? This presents an interesting area for exploration. The construction of a co-culture model for vascularization provides strong support for the study of the interaction between CAR-T cells and tumor blood vessels.

Finally, the utilisation of tumor organoids co-culture models in scientific research in oncology treatments is not limited to CAR-T therapies. Schnalzger et al. [14] utilised colorectal cancer organoids to carry out killing toxicity assessments of CAR-NK. Recently, CD19-CAR-T cells have been used for the treatment of systemic lupus erythematosus, achieving encouraging therapeutic effects [46, 47]. At the same time, some research teams have successfully used organ chip technology to construct and simulate organoid models of autoimmune diseases [48, 49]. The good news from these two aspects has given us great inspiration—let's imagine that with the development of autoimmune disease organoid models, the autoimmune disease organoids and CAR-T cell co-culture models will become a preclinical model with great potential, just like the tumor organoids and CAR-T cell co culture models, widely used in preclinical research, pre infusion efficacy evaluation, and other aspects, which will benefit a large patient population. In brief, co-culture models are expected to see extensive use in future scientific research on adoptive cell therapy (ACT). The potential for growth in co-culture models remains high, and researchers are encouraged to conduct thorough explorations of such opportunities.