Breast cancer organoids and their applications for precision cancer immunotherapy

One approach for native TME models is retaining and expanding endogenous immunocytes within organoids as a cohesive unit, which is a relatively simple immunotherapeutic approach. This co-culture model can be applied in the study of breast. Zumwalde et al. [47] successfully established organoids sourced from the mammary ductal epithelial cells of the human. Apart from specifying the intraepithelial lymphocyte compartment of healthy human breast tissues, they also pinpointed a T-lymphocyte subset, whose BC cell reactivity could be increased through pharmacological targeting. According to their findings, the leukocyte communities of breast organoids were unlike the peripheral blood counterparts, and the utilization of Vδ2 (+) T-cell reactivity to the Food and Drug Administration-approved bisphosphonates, as a novel immunotherapeutic strategy, could suppress BC growth. This is the first immunocyte-organoid co-culture model reported in the immunological study of BC.

The submerged Matrigel system is extensively used for culturing patient-derived carcinoma cells in a mixture of tissue culture medium and 3D matrix Matrigel. But stromal components are not retained by the regularly submerged Matrigel organoids. For the precision study of BC–TME and immunotherapy, ECM, stromal cells, and additional exogenous immunocytes are required to reconstruct BC–TME. Many studies have reported their novel models for co-culturing BCOs and immune cells. Hanley et al. [48] developed BCO-autologous stromal cell co-culture systems and revealed their dynamic molecular interactions. They further demonstrated that the infiltrative capacity and molecular phenotype of BC cells could be affected by the adjacent mammary cancer-associated fibroblasts (CAFs). Dhimolea et al. [49] co-cultured HR-positive BC cell line spheroids or PDOs in 3D ECM, alone or together with bone marrow stromal cells (BMSC), and highlighted the role of BMSC in affecting metastatic microenvironment and mediating hormone-independent tumor growth. These studies showed that adding additional immune cells and other stromal cells to BCOs cultured by submerged Matrigel culture is a feasible approach to reconstruct BC–TME. However, matrix components and immune cells are usually not fully preserved in tissue processing stage, and restoring BC–TME to its initial stage is a laborious and time-consuming process. This is also a common problem in other co-culture approaches, such as ALI culture and microfluidics 3D culture. Solving this problem is of great significance to the development of immune organoid models.

Through the ALI method for creating organoids, the tumors can grow as a cohesive unit, and en bloc preservation of carcinoma cells can be allowed, retaining their native stroma. Tumor organoids developed from minced fragments of primary tissues are added to a collagen gel that is inside an inner Transwell dish. Culture medium diffuses into the inner dish from the outer dish via the permeable Transwell. The ALI method is applied to expose the top collagen layer to the air, so the cells receive adequate oxygen [50, 51]. This type of organoid culture achieves the TME recapitulation by intrinsically retaining the fibroblasts and multiple parental tumor immunocytes, such as various immune cells (B cells, T cells, NK cells, and macrophages), without requiring artificial reconstruction, making them distinct from submerged Matrigel culture.

Presently, it has been observed that a PD-1 blocking antibody can initiate anti-tumor immune responses within ALI organoids from several types of carcinomas, in an approach that seems independent the PD-L1 expression status of tumors [52]. Notably, the ALI culture fails to remedy the short preservation deficiency of stromal myofibroblasts, such as SMA and vimentin, which would decline over a 6-week period in the organoid cultures [53]. It is challenging to investigate innovative approaches to ALI culture for breast organoids. The ALI-PDOs en bloc having endogenous immune stroma is less effective in breast organoids, because growing BCOs and maintaining TILs derived from BC are incredibly more challenging than the rest of the carcinoma types. Preserving the vascular system that carries immune cells may be helpful for ALI culture, but perfusion remains a huge challenge.

A collagen gel mixture can be used to culture murine- or patient-derived organotypic tumor spheroids (MDOTS/PDOTS) in microfluidic 3D devices [54]. The patient-derived samples of tumor tissues were subjected to enzymatic and mechanical fragmentation for the MDOTS/PDOTS culture. The obtained samples were a nonuniform mixture of macroscopic tumor fragments, single cells, and spheroids [55]. Tumor spheroids are grown in a media-assisted 3D gel in the central zone of the media channels of the microfluidic 3D device, which runs parallel and is situated on either side of the central zone. Cultivation and assessment of MDOTS/PDOTS from syngeneic immunoreactive murine models and tumor samples of patients are achieved for 1–2 weeks, such as Merkel cell carcinoma and melanoma [56, 57]. As shown by flow cytometric profiling of immunocytes, apart from retaining tumor cells, the MDOTS and PDOTS also retain myeloid populations (tumor-associated macrophages, monocytes) and autologous lymphocytes (B and T cells). The results showed that these models can retain myeloid cell and autologous lymphoid communities and respond to immune checkpoint blockade (ICB) in 3D short-term microfluidic culture. Therefore, MDOTS/PDOTS profiling is an innovative platform for ICB assessment, where the clinically relevant specimens of patients and recognized murine models are utilized. Through this in vitro culture, a preclinical model of BC was also developed [58]. This model has certain value in screening the classic chemotherapeutic agents in real time and great potential to play a role in immune studies. Furthermore, Truong et al. [59] co-cultured BC cells and patient-derived fibroblasts in 3D tumor and stromal sites so that the TME spatial organization can be mimicked on a microfluidic chip. They studied the tumor-stroma interactions and further revealed that CAFs promoted invasion through the upregulation of glycoprotein nonmetastatic B in BC cells. These outcomes indicated the ability of this co-culture model to recapitulate patient-specific TME for exploring the tumor–stroma interactions in BC. Therefore, microfluidics 3D culture is a promising tool to co-culture organoids and immune cells for immunological research on BC. In terms of limitations, specialized equipment is required, and the immune components may decline over time in microfluidics 3D culture.

Organ-on-chip (OoC) is an emerging technology combining cell biology, microfabrication, and microfluidics. Epithelial organoid cultures can be integrated into organ-on-a-chip platforms to form the “organoid-on-a-chip” system, a more complex culture system for organoids [60, 61]. “ Organoid-on-a-chip” is a biochip system that combines the technological advantages of organoids and organ-on-chip and produces the same physiological and metabolic linkage reaction with multiple human organs. This model can overcome the disadvantages of organoids by making organoids more uniform and mimicking the bodily physical conditions, e.g., by providing culture media perfusion. Besides, with the aid of integrated sensors and actuators, microfluidic devices can be used to perform parameter assay as well as culture condition surveillance and control [62]. The multi-organoid-on-a-chip was also established [62, 63]. Critical parameters of the immune microenvironment and TME are recapitulated via the organoid-on-a-chip platform of tumors so that the synergistic and independent effects of various tumor progression components can be systematically comprehended [64, 65]. This co-culture model has progressively become an innovative and reliable tool for investigating how tumors evade immunity by affecting TME and how they resist immunotherapy.

In the field of BC research, organoid-on-a-chip has been used for surveilling the primary tumor responses to immunotherapies in patients. Zhang et al. [66] reported that MCF7 and MDA-MB-231 BC cells have been used to investigate whether on-chip testing of personalized immunotherapy was achievable by applying their mini-tumor chip. Herein, primary tumors were loaded on a chip after dissociation into single cells. On this basis, the primary BC cells on-chip were capable of responding to the anti-PD1 therapy or NK cell therapy at varying efficiencies. This on-chip reaction stresses the effectiveness of the chips in assessing the immunotherapy responses of patient tumors, with great potential to become the gold standard for preclinical screening of individualized therapy. However, the improvement of mechanical conditions and culture media for varying tissues is still tremendously challenging. Besides, whether current on-chip tumor models can fully simulate biological processes in vivo remains unclear, their application value should be investigated for individualized immunotherapy by clinical results in the future.

Cellular immunotherapy, also known as ACT, harnesses the killing power of immune cells to fight against cancer. The initial step of ACT immunotherapy is the isolation of immunocytes from either a patient (autologous cells) or a donor (allogeneic cells), which are subsequently genetically engineered, expanded, and activated ex vivo and eventually reinjected into the patients [75, 76]. Recently, ACT has made significant research and clinical advances in many types of cancer. The major ACT therapies include engineered T-cell receptor (TCR), NK cell, tumor-infiltrating lymphocyte (TIL) [77], and chimeric antigen receptor (CAR) T-cell therapies.

As mentioned above, Zumwalde et al. [47] successfully co-cultured Vδ2+ T lymphocytes with organoids derived from human breast mammary ductal epithelial cells and demonstrated the potential of utilizing Vδ2 (+) T cells to respond to bisphosphonate drugs as the novel immunotherapy approach to inhibit BC growth. This study first utilized organoid models for ACT therapy.

CAR-T-cell therapy refers to one type of ACT immunotherapies. CAR-T cells are equipped with specific antibodies to recognize antigens in autologous tumor cells and further induce cytotoxic effects, achieving remarkable successes in the treatment of hematologic malignancies recently. CAR-T-cell therapy also exerts a role in the treatment of BC but is not as effective as expected [78]. The main reason is probably that solid tumors usually face multiple barriers to ACT, such as immunosuppressive TME, antigen specificity, and toxicities. BCOs may be efficient platforms for showing TME and assess the tumor-specific cytotoxicity of T cells. Wallstabe et al. [67] once established standardized and scalable BCOs from MDA-MB-231 with architectural and phenotypical features of TNBC. Using these 3D tumor models, they investigated the antitumor function of CAR-T cells and obtained proof of concept for their safety and efficacy before the clinical application. They further demonstrated potent antitumor effects of receptor tyrosine kinase-like orphan receptor 1-specific CAR-T cells.

Adoptive transfer of bispecific antibody-armed activated T cells (BATs) exhibited promising antitumor activity in the clinical trials of solid tumors. Thakur et al. [68] hypothesized that the release of BAT-induced tumor-targeting effectors (TITE) might play the role of a potent antitumor and immune-activating immunotherapy. The TITE exhibited potent cytotoxic activity against multiple BC cells in a 3D tumorsphere model. They believed that TITE could offer a clinically controllable cell-free platform to target various tumor types containing BC, regardless of the mutation-prone and heterogeneous nature of the tumor cells.

The efficacy of NK cell immunotherapy in BC was also evaluated by a microfluidic platform [69]. This model contains a 3D BC spheroid in a 3D ECM and two flanking lumens lined with endothelial cells, replicating pivotal structures and components within the immune response. It was discovered that NK cells could detect tumor spheres farther and faster than antibodies. Once inside the spheroid, NK cells can also destroy tumor cells completely, both at the spheroid periphery and the innermost layers. Besides, Yang et al. [79] showed that mesothelin-targeted CAR-NK cells derived from induced pluripotent stem cells had a certain efficacy in killing TNBC cells in several preclinical models, including in vitro organoid models. Parikh et al. [80] established PDOs from common epithelial cancers, including BC, and demonstrated their utility as an effective tool for selecting TCRs and TIL in ACT.

All the obtained findings showed the application of the model for detecting novel therapeutic approaches of ACT to increase immunotherapy against solid tumors.

Antibody-based immunotherapy is a leading type of cancer immunotherapy that particularly and directly restricts cancer cell survival, activates the immune system to eradicate cancer cells, or delivers cytotoxic compounds [81]. Recently, tumor organoids have been used as preclinical models to investigate the efficacy of antibody-based checkpoint blockade immunotherapy. The immune–tumor organoids are also used for detecting new strategies for an antibody-based combination treatment of cancers.

Organoid models have been used to research the value of antibody-based therapeutics in TNBC. Dees et al. [70] reported that the treatment of trophoblast cell-surface antigen 2 (Trop2) and trophoblast cell-surface antigen 2 (CEACAM5) expressing 3D-TNBC spheroids with CD3 × Trop2 or CD3 × CEACAM5 bispecific antibodies in combination with human peripheral blood mononuclear cells significantly hindered the TNBC cell growth. Besides, the addition of an antagonistic anti-PD-1 monoclonal antibody to this model further increased cell death in 3D TNBC spheroids. These findings indicated that combining T-cell-redirecting bispecific antibodies with immune checkpoint inhibitors (ICIs) provides a practical approach to improving antitumor efficacy and surmounting the immunosuppressive TME in TNBC. Ning et al. [82] reviewed representative cutting-edge antibody-based therapeutics in TNBC in clinical use and trials and suggested that antibody-based therapeutics hold great promise in TNBC. For a better clinical application of antibody-based therapeutics in TNBC, effective preclinical verification using reliable models is particularly crucial. Considering the advantages of immune–tumor organoids in reconstructing the TME, BC organoids can be an ideal model for future studies on antibody-based therapeutics in BC.

ICB therapy has exhibited promising results in various cancers. In recent years, extensive efforts have been made to develop effective immunotherapy to enhance clinical outcomes in BC, particularly for TNBC. Growing evidence suggests that BC is markedly heterogeneous concerning immune infiltration and the TME, and that lymphocyte infiltration into tumors is related to a better prognosis and better clinical responses to chemotherapy [83, 84]. However, genetic heterogeneity, the lack of actionable targets, and immune evasion lead to limited clinical response rates to ICB therapy. In metastatic BC, lasting responses occur in only approximately 5% of patients and are mostly limited to TNBC [85]. When ICB is used in combination with chemotherapy in patients with PD-L1-positive tumors, response rates may increase. But most BC patients do not benefit from ICBs. To further study the response of BC to ICB, better immune models in vitro on antitumor immunity are urgently needed.

A study previously showed that PDOs could functionally recapitulate the PD-1/PD-L1-dependent immune checkpoint system, thereby allowing the in vitro modeling of intrinsic and syngeneic immune cell responses [86]. Shelkey et al. [71] proposed a novel immune-enhanced tumor organoid (iTO) system to explore factors affecting the response to ICB therapy. They successfully showed the response to ICB therapy using the 4T1 TNBC murine cell line and matched splenocytes. Furthermore, the administration of bacterium derived from species found in the immunomodulatory host microbiome could increase the ICB-induced apoptosis of tumor cells and decrease the levels of the immune cell receptor. These results showed an approach to isolate individual factors that altered the response to ICB and streamlined the study of the effects of the microbiome on ICB efficacy. On the basis of these results, we can conclude that iTOs are robust platforms that help assess the efficacy of cancer immunotherapy, discover immune-oncology resistance mechanisms, and identify new combination therapies for ICB therapy. Zhou et al. [72] reported a high-throughput immune-drug screening approach on the basis of the functional interaction of mouse or BCOs and tumor-specific cytotoxic T cells. On this basis, they identified that the epigenetic inhibitors GSK-LSD1, CUDC-101, and BML-210 exhibited antitumor ability in orthotopic mammary tumors in mice. The epigenetic inhibitors increased antigen presentation mediated by the major histocompatibility complex class I on BC cell. Besides, BML-210 made breast tumors more susceptible to the PD-1 inhibitor. Thus, organoids can be used in ICB therapy and its combination therapy for BC. However, the material and composition of devices used in organoid cultures may affect the outcomes of these immunotherapies, including ICB therapy [87].

Oncolytic viruses are a class of viruses that selectively infect and damage cancer tissues without damaging normal tissues. Their use in cancers is gradually becoming a promising treatment approach [88]. However, oncolytic viruses for treating BC have not been successfully developed yet, because the traditional models adopted in previous studies failed to mimic the complex TME of human cancers sufficiently [18]. This has markedly hindered the approval of oncolytic viruses for BC treatment.

Nonetheless, organoid models show some advantages to observe the effect of oncolytic viruses in BC. Carter et al. [73] developed stable organoid models derived from BC tissues and observed the greatest oncolytic effects of oncolytic viruses that were engineered to denote a suicide gene (MeV-SCD and GLV-1h94) in the presence of the prodrug 5-fluorocytosine. They suggested that the organoid model provided a promising in vitro approach to benefit the testing and further engineering of virotherapeutic vectors in vivo. Similarly, Behrens et al. [74] also reconstructed BCO models to test the effectiveness of the oncolytic Urabe mumps virus in TNBC. They used the original oncolytic Urabe MuV clinical trial virus stock (MuV-U-Japan) that potently killed several established human BC cell lines in vitro, significantly extended the survival of nude mice with human TNBC MDA-MB-231 tumor xenografts in vivo, and exhibited significant killing activity against BC PDX cell lines grown as 3D organoids containing PDXs from patients resistant to anthracycline- and taxane-based chemotherapy. Moreover, the present study reveals the suitability of the MuV-UC virus for translation to modern clinical trials for treating patients with TNBC. However, immune cells were not included in the BC organoid models described above. More efforts are needed to incorporate immune cells into the organoid culture to improve BCOs. Once this model is successfully established, it allows a more accurate and realistic representation of the TME surrounding the tumor and the immune response of patients to oncolytic virotherapy, revealing its efficiency. It will also enable the evaluation of oncolytic viruses that have been specifically engineered to induce an immune response against tumor cells. We look forward to using such tumor-immune co-culture models to understand the potential of oncolytic virotherapy in BC treatment.

Now, we see the advantages of organoids in immunotherapy for BC. But, some important limitations of these co-culture models still exist. First, BCOs are usually derived from biopsies, which may lower the success rate of organoid modeling, and biopsy tissues often represent only a small fraction of tumor characteristics, potentially underestimating the overall complexity of BC tumors. Second, a cell culture model will miss or underestimate the genetic heterogeneity of BC tumors, especially considering BC can contain small “islands” with, e.g., cluster amplifications of Her2/neu, which are therapeutically relevant. Third, the BC–TME is complex and contains many cell types that we need to control; thus, the BC–TME usually cannot be fully replicated, even with exogenous immune components. Besides, the culture conditions of such organoids are specific, requiring many growth factors, but these growth factors may also affect other cells that are co-cultured with BCOs. A major focus of future studies should be the optimization of co-culture conditions further, considering factors such as cell growth factors and different immune cell components in the BC–TME.