The application of patient-derived organoid in the research of lung cancer

Lung cancer is the most common cancer and the leading cause of cancer-related death worldwide. However, mechanisms of its progression remained unclear and new treatments against this disease are rapidly emerging. As a novel preclinical model, patient-derived organoid (PDO) can also be established from the patient’s tumor tissue and cultured in the laboratory, which preserves the key biological characteristics of the original tumor. Compared to the patient-derived xenograft (PDX) model of lung cancer, the culture success rate is improved, and the time and cost of model establishment are largely reduced. PDO is also expected to provide a more individual model to predict the efficacy of anti-cancer treatment in vitro. This paper summarizes the current application of PDO in the translational research of lung cancer.


Introduction
Lung cancer is the most common cancer worldwide. It is also the leading cause of cancer related death (1.8 million deaths each year) [1]. Lung cancer development is closely associated with genetic alterations such as epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, and kirsten rats arcomaviral oncogene homolog (KRAS) mutation [2], and its progression and metastasis are also associated with microenvironment changes. Although new drugs that targeting the key molecular events may impede or even reverse this process, their effects remain hardly predictable and all develop drug

PDO
reproduces the biological characteristics of lung cancer in vitro

PDO as a preclinical model for lung cancer
Lung cancer PDO has some advantages over other preclinical models, comparing to traditional cancer cell lines and the PDX model (Table 1).
Cell lines remains the most used in vitro model in cancer research. However, lung cancer cell lines also have some significant limitations. Firstly, it does not reflect individual tumor characteristics real-time and the success rate of establishment from primary lung cancer is less than 5% [14]. Secondly, random genetic drift caused by long-term culturing and repetitive passaging produces artifacts cannot reflect the genetic background and individual epigenetic differences of lung cancer [15]. In addition, cancer cells cultured in three dimensions are more close to physiological state in vivo differentiating them from those monolayer cultured cell lines [16]. Furthermore, experimental therapies using established cell lines for preclinical testing has a high failure rate in phase III clinical trials, which indicates its inherited application defects in tumor research and drug development [17,18].
Lung cancer PDX mice model, as compared to tumor cell lines, retains more genomic and phenotypic characteristics of the primary tumor. Drug testing results based on PDX models reproduce the clinical outcomes, thus it is considered a potential preclinical model for translational research and personalized drug screening [19]. Transplantation of human leukocytes and purified CD34 (+) hematopoietic stem cells into immunodeficient mice may mimics the functional human immune system, allowing PDX to be used in vivo to assess the therapeutic response of lung cancer cells to Pembrolizumab and Nivolumab [20]. However, to establish a successful PDX model is time consuming and costly, thus significantly limiting its application in large-scale drug screening and personalized drug guidance [21]. Although PDX provides complete tumor environment for cancer cell growth, transplanted human stromal cells have tendency to be replaced by murine counterparts. Species-specific cytokines may cause drastic changes in drug responses [19,22]. Besides, genetic alterations and chromosome abrasions are also inevitable in PDX passaging [21].
PDO summarizes histopathology, gene expression profile, and treatment sensitivity of the primary tumor [23]. PDOs derived from different lung cancer patients show different morphology under hematoxylin-eosin (HE) staining but maintain similar morphology with the primary tumor tissues. HE staining and immunohistochemical is also used to identify the purity of tumor cells in lung cancer PDO [5,[23][24][25][26]. According to the whole-exon sequencing, wholegenome sequencing and RNA-seq of lung cancer PDO, Kim and their colleagues reported that short-term cultured lung organoids have retained 92.7 and 77% of the driver mutations in the primary tissue, respectively [4,27]. Notably, some PDOs harbored additional driver mutations that were not detected in the match tissue may related to cross-cell contamination, limitations of genetic tests or low frequencies in the original tumors. Compared to short-term cultured PDO, long term passage (passage > 10) PDO maintain the overall mutation spectrum and copy number variation detected in the original tissues, although 80% of long-term cultured PDO have increased mutation numbers, suggesting the sub-clonal expansion [4,5,23,25,27].

Lung cancer PDO in the research of oncogenesis
Metabolic reprogramming is one of the major steps of oncogenesis [28]. Glutamine synthetase (GS) is overexpressed in cancer and promotes cancer cell growth through glutamine anabolic metabolism [29]. Knockdown of GS in lung cancer PDO can inhibit organoid growth. Depletion of GS also restores the sensitivity of PDO to the microtubule drug Paclitaxel [30]. Using lung cancer PDO, Chen et al. and their colleagues showed reactive oxygen species (ROS) play pivotal roles in epithelial-mesenchymal transition and cell invasion and migration. Fangchinoline, a small molecule drug, revise this malignant progression by reducing cytoplasmic ROS and inhibit the Akt-mTOR signaling  [31]. It was also found that NF-κB and MYC were overexpressed in CD133 (+) CD44 (+) lung cancer PDO, and treatments targeting these signaling pathways may be is a possible treatment for the patients [32]. High invasiveness of CD133 (+) colonies in lung cancer PDO are also associated with activation of AXL, TGFβ, and JAK1, proofed by effective treatment of AXL inhibitor TP-0903 and JAK inhibitor Ruxolitinib [33].
Besides genomic research, lung cancer PDO can be also used to explore the epigenetic changes of lung cancer. Sca-1 (+) CD24 (+) double-positive lung adenocarcinoma (LUAD) cells are a subtype with high tendency of proliferation and invasion. Using LUAD PDO, Rowbotham et al. found that this aggressive phenotype is controlled by H3K9 methyltransferases G9a, a suppressor of tumor-propagating cell phenotype in lung cancer cells. For specific individuals with this subtype, demethylase inhibitors rather than methylase inhibitors contribute to the treatment of advanced lung cancer [34].

The diversity of sample sources for lung cancer PDO
Various research groups have reported the successful establishment of lung cancer PDO from different sources ( Table 2). Compared to PDX and cancer cell line, lung cancer research using PDO is at a climbing stage [35]. Besides surgically resected tumor, PDO can also be cultured from malignant pleural effusion or biopsy tissue. Although the success rate remains to be improved (the current success rate is approximately from 28 to 83%) [4,9,11,36,37].

Co-culture PDO model in the research of tumor microenvironment
Changes in tumor microenvironment (TME) affect cancer growth, migration, and invasion, thus are associated with patient prognosis and treatment response [38]. TME includes extracellular matrix (ECM) and stromal cells. There are as many as 52 subtypes of stromal cells in lung cancer, and the subtypes have different proportions among different patients [39]. As mentioned above, cancer cell lines cannot reflect the actual phenotype of the primitive tumor. Although PDX contains a tumor microenvironment, it is difficult to observe and intervene its microenvironment in cancer research. Establishing cell-matrix and cell-cell interactions in tumor PDO may help to understand carcinogenic mechanisms and drug development [40].
Co-culture by supplementing the medium with stromal cells, including endothelial cells, immune cells, and fibroblasts, allows in vitro observation and quantitative study of cell-cell interactions [13,[41][42][43] (Fig. 1). The specific CD133 (+) stem cell colony presenting in LUAD PDO has a higher proliferation rate due to its excess expression of the Wnt signaling pathway promoter PORCN. Extracellular vesicles secreted by co-cultured peripheral fibroblasts further promote PORCN expression, creating a friendly microenvironment for tumor cell proliferation [44]. Additionally, co-culture of LUSC PDO with carcinoma-associated fibroblast (CAF), which involve tumor progression by actively interaction with other cell types in the tumor microenvironment, enhanced PDO formation [45].
Besides the natural ECM, the emergence of extracellular matrix removal culture has brought a breakthrough for organoid co-culture model. New materials such as composite hydrogels allow the co-culture of PDO and stromal cells, and have steerable biochemical signals and independent mechanical properties changes, making the growth, development, and morphology of PDO more controllable [46]. Using PEG-fibrin hydrogel as a scaffold for PDO, Del Bufalo and their colleagues found that lung cancer cells cocultured with fibroblast MRC5 grow faster, and the combination of MRC5 and endothelial cells HUVECS in the co-culture system further enhanced tumor cell proliferation and invasion. These results highlight the potential of the coculture system in the research of TME [47].

Co-culture and immunotherapy evaluation
Inhibitors of immune checkpoint pathways such as the programmed cell death protein-1/programmed death ligand-1 (PD-1/PD-L1) pathway have shown exciting therapeutic effects in lung cancer recently [48,49]. However, their effects in individual patient remain largely unpredictable [50]. PDO, along with co-culture system, is assessed as an in vitro and patient-based platform to observe T cell-mediated tumor recognition to help us understand the key factors that determine successful anti-tumor immune responses and to screen suitable therapeutic schedule [13] (Fig. 1).
Recent reports point to the potential feasibility of lung cancer PDO co-culture in immuno-oncology research. The first method is called the reductive simulation method. Researchers hope to generate tumor-reactive T cells by coculturing matched patients' peripheral blood lymphocytes (PBL) with tumor cells. The success rate of the tumor-reactive CD8 (+) T cell population is between 33 and 50% hence the co-culture system is a potential platform for evaluating the interaction between tumor cells and T cells in lung cancer PDO [13,51]. In another co-culture model of peripheral blood mononuclear cell (PBMC) and non-small-cell lung the original tumor tissue. In contrast, PD-1 inhibitor causes significant tumor cell lysis in the co-culture system of PBMC and lung cancer PDO, indicating the importance of co-culturing technique for immunotherapy evaluation [10]. Moreover, to increase the immunotherapy testing speed of in vitro platform, Ding et al. developed an automated microfluidic droplet platform that can rapidly generate considerable amount of lung cancer PDO and reliably evaluate the cancer (NSCLC) PDO, MEK-targeted drugs and immune checkpoint inhibitors have synergistic anti-tumor effects by increasing T cell reactivity [12]. xCELLigence is a noninvasive, label-free, real-time cell impedance monitoring technique that evaluates the efficacy of immune checkpoint inhibitors in vitro by assessing cytolysis [52]. Using xCEL-Ligence as a evaluation tool, PD-1 inhibitor Nivolumab or Pembrolizumab alone has no significant effect in lung cancer PDO because of the loss of immune cells retained from  [54]. In addition, through the co-culture system, PDOs are also suitable for evaluating the efficacy of chimeric antigen receptor engineered T (CAR-T) cells. CAR-T cells exhibit antitumor activity in LUAD PDO by targeting B7-H3 [55]. We expect an improved and more stable PDO model to demonstrate the feasibility of providing efficacy prediction for precision immunotherapy in future clinical trials.
Besides the reductive simulation method, another holistic approach is to culture PDO based on an air-liquid interface (ALI), which preserves the original immune components of the tumor, rather than adding additional blood cells to produce endogenous and homologous tumor infiltrating lymphocytes (TILs). This organoid model derived from NSCLC patients preserves the functional and original tumor microenvironment and successfully simulates the immune checkpoint blockade to recover the anti-tumor activity of activated TILs, and the only drawback is that this kind of TIL cannot remain in the medium for more than 60 days Fig. 1 Co-culture system of lung cancer PDO. Co-cultured with fibroblasts or endothelial cells, lung cancer PDO is allowed to research cell interactions in TME. Co-culture system is also suitable for immuno-therapy evaluation. Endogenous TILs or exogenous TILs can be generated successfully with the co-culture of original immune stroma and PBL/PBMC correspondingly subsequent PDX experiments and phase II clinical studies with 53.5% of ORR [37].
Lung cancer PDO can also be used for drug development on neo-cancer targets and large-scale drug screening. As a preclinical model, Lung cancer PDO verifies the anticancer activity of MFF (D) 8-11 peptide mimic, and CKD9 inhibitors SNS032, LY2857785, AZD4573 [66,67]. Different from traditional organoid culture systems [26,68], the microfluidic platform developed by Jung and their colleagues and the InSMAR chip designed by Hu and their colleagues can generate PDO with uniform size and load various synthetic or natural anticancer drugs combinations continuously. It is proved that those devices accurately summarize the response of different tumor subtypes and predict the patient's drug resistance which is highly consistent with PDX and clinical treatment data [9,69]. Their work reduced the drug susceptibility testing time in PDO to one week. Fast, high-throughput, and accurate, these features improve the feasibility of PDO as a new preclinical model in individualized medicine.

Challenges and limitations of lung cancer PDO
Establishing pure tumor organoids and maintaining a stable success rate remain significant challenges for PDO application and transformation [11]. Lung cancer organoids culture medium mimics colorectal cancer organoids culture medium, and mismatch of growth factor additives often leads to overgrowth of normal airway organoid (AO) in lung cancer PDO [36]. Although many research groups are actively adjusting the formula to improve the success rate, such as using minimum basic medium or improved medium M26 [23,27], it is still essential to establish the optimal medium for lung cancer PDO with variable tumor subtypes [70]. B27 supplements, N-acetylcysteine, nicotinamide, SB202190, FGF, and other supplements provide the necessary survival protection signals for cell growth and contribute to organoid formation, but the specific concentrations still need to be adjusted in experiment [70]. Y-27,632, WNT/R-spondin, and EGF may be detrimental to the growth of LUAD PDO [36,70]. Nutlin-3a and Palbociclib can be used to screen lung cancer cells with TP53 or RB1 mutations from normal organoids [13,36,70]. This may contribute to the separation of tumor organoids harbored specific mutation but is not suitable for all samples in PDO purification because lung cancer driver mutations are diverse, and wild-type mutations in normal lung cells [71]. It was also reported that the culture conditions of NSCLC PDO and SCLC PDO were different. R-spondin1 and Wnt3a were important factors for the long-term culture of SCLC tumor

Clinical research and drug screening based on lung cancer PDO
In recent years, the use of PDO for individualized drug selection is under clinical investigation [5]. The application of PDO in a large-scale prospective clinical cohort helps to deeply integrate molecular biological characteristics and treatment response in cancer patients, reduce the time of clinic-laboratory-clinic cycle, and establish a medical platform for precision oncology [56]. Thirteen lung cancer PDO-related clinical trials have been registered in the Clinical Trials (Table 3).
PDO is an important pre-clinical model for determining mechanisms of resistance. Detection of BRAF V600E, KRAS G12D, KRAS G12V, and PIK3CA H1047R resistance-associated mutations in long-term cultured Erlotinibresistant lung cancer PDOs revealed that tumor cells often harbor multiple associated mutations, which means that partial Erlotinib resistant patients require combination therapy to address tumor resistance [57]. The addition of DCLK1 inhibitor DCLK1-IN-1 to the third-generation EGFR-TKI inhibitor Osimertinib resistant PDO downregulates the Wnt/β-catenin signaling pathway, restores tumor sensitivity to Osimertinib [58].
Cisplatin, as the first-line drug for SCLC and NSCLC chemotherapy, inevitably leads to drug resistance during treatment. The mechanism of chemotherapy resistance is complex, and verifying the hypothesis in the new model is necessary [59]. Several compounds were tested in cisplatinresistant lung cancer PDO to restore chemosensitivity, such as YPN-005, which antagonizes CDK7, Solamargine, which blocks the hedgehog signaling pathway, and Halofuginone, which inhibits PI3K/AKT and MAPK signaling pathways [60][61][62].
PDO is also an important pre-clinical model for predicting the efficacy of targeted therapies and can be used for prospective adjuvant targeted therapy experiments to reduce treatment costs and improve success rates [63]. In preliminary studies, clinical outcomes were consistent with drug response in PDO models [64]. LUAD PDO carrying EGFR and BRAF mutations successfully captures the clinical response of the tumor to Dabrafenib/Trametinib combination therapy [4]. The accuracy of PDO in predicting response of patients with ERBB2 exon 20 insertion and RET fusion treated with Poziotinib and Pralsetinib combination is 75% [4]. The high agreement between the PDO drug trial and clinical results promotes Amivantamab as an effective treatment option for NSCLC patients with EGFR Exon20ins [65]. Besides, PDO cultured from a patient with HER2-A775_G776YVMA insertional mutation summarize the patient's response to Pyrotinib, as evidenced by by suspension culture or separating tumor tissue into single cells in microfluidic systems for PDO formation, but we cannot ignore the importance of extracellular matrix in tumor progression [53,72]. To establish a reliable organoid model, genomic sequencing combined with morphological observation and marker staining or in vivo tumorigenesis organoids [25]. Another strategy for tumor purifying in lung cancer PDO is to remove non-tumor components as much as possible before inoculation. Using biopsy or metastatic tumor tissue to create organoids that lack normal lung epithelium helps to avoid the overgrowth of AO [4,36]. There are also reports about the removal non-tumor components     VAF may indicate that lung cancer PDO consist of a heterogeneous population [27], but whether organoids reflect the heterogeneity within tumors needs to be further explored.

Conclusion and perspective
Lung cancer PDO can reflect the genetic background of tumors better and is easier to culture than PDX, with lower cost and shorter culture period, showing significant advantages and potential in basic and clinical research for lung cancer. It is not only possible to establish PDO from surgical resection specimens but also feasible to establish PDO from malignant pleural effusion or infinitesimal biopsy specimens. The establishment of an optimized culture medium and co-culture system reflects a more realistic TME, which helps us understand the internal mechanism of tumor development and drug resistance. At the same time, large-scale organoid library construction or biobanking improves the use of drug efficacy tests and the credibility of drug screening results, which is expected to provide guidance for individualized medical treatment in the future. Lung cancer PDO is still facing some problems. Firstly, the success rate of PDO establishment is unstable. The repeatability of organoids between groups has become a experiments is essential for early identification in lung cancer PDO and clearance of AO [11]. There was no significant difference in the success rate of PDO culture from different lung cancer subtypes [27]. However, the success rate of PDO culture in early lung cancer may be lower than that in advanced lung cancer [12]. Due to different success rates and low repeatability in different research groups [4,5,9,36,57], we need a standardized methodology. And it would be beneficial to create a database containing the PDO/AO culture system data and histological genetic information [4,70].
Genomic drifting caused by invitro culturing, which is also affected by differently additives in the medium, is also a major limitation of the PDO models. Presently, organoids are cultured in modified Dulbecco's modified Eagle's medium, known as DMEM/F12 medium, with naturally extracted animal ECM matrix as a scaffold. The most widely used ECM is Matrigel extracted from mouse sarcoma [73]. But this natural Matrigel contains more than 1800 proteins, the unknown signals may have selective pressure on tumor growth [74]. This shaping from culture medium may cause specific bias of cell state and affect cell function [75]. Therefore, homogeneity between the PDO models and the primary tumor should be validated by sequencing before further application of this model. In addition, a widely distributed  cancer PDO, and the development prospect will be extensive [79] (Fig. 2).
1) Single-cell RNA-sequencing (scRNA-seq) is one of the most accurate tools to verify whether PDO truly generalizes the characteristics and heterogeneity of primary tumors. The scRNA-seq allows analysis of tumor cell genomes, transcriptomes, and epigenomes at the singlecell level, tracking the lineage of cells, which helps with the development of precision oncology [80]. The upgrade from RNA-seq to scRNA-seq is expected to match the unique clinical and biological characteristics of patients with the best treatment combination, thereby maximizing clinical benefits. Spatial transcriptomics is a rapidly developing method which allows preserving crucial spatial context of regulatory processes. Integration of spatial and single-cell transcriptomic data provide a comprehensive cell resolution spatial map for limitation of research and application. The referential steps for lung cancer PDO culture have been released, and the standardization protocol needs to be promoted [76,77]. Secondly, the purity of PDO depends on the sampling, digestion, and culture medium. Further optimization of culture methods and medium components is needed to confirm the applicability for various types of lung cancer [78]. In addition, despite the limited co-culture duration in primary stroma cells and lack of prospective validation for immunotherapy assessment, the co-culture technology of lung cancer PDO should be developed to reconstruct TME in vitro, which will help PDO predict clinical outcomes more accurately and promote the application of organoids in the field of immunotherapy [54].
With the rapid development of biomedical engineering, the advanced technology that has been applied to other tumor organoids can be introduced into the culture of lung Data availability Not applicable.

Declarations
Ethical approval Not applicable.

Competing interests
The authors have no competing interests to declare that are relevant to the content of this article.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/. morphogenesis research and tumor tissue architecture analysis [81,82]. 2) Genetic engineering enables organoids to study the role of interested gene mutations in cancer and develop corresponding anticancer therapies. Dox-induced HER2overexpressing conditioned cells were generated from iPSC cells using the CRISPR/Cas9 genome editing system, and corresponding organoids with typical AAH characteristics were successfully cultured. CRISPR/ Cas9 allows researchers to build specific organoids to explore vital evolutionary processes in early-stage lung cancer [24,83]. Interestingly, the induction of mature lung organoids from iPSC requires the co-culture of human fetal fibroblasts to support the differentiation of stem cells into alveoli [24]. Besides, CRISPR-HOT, developed by Artegiani and their colleagues, enables efficient and visual gene editing in organoids [84]. Rare and specific subsets are often observed in lung cancer [85]. Genome-Wide CRISPR/Cas9 screen for rare and interesting genes in organoids will significantly promote the development of tumor research [86]. 3) Advanced engineering methods are complementing the traditional organoid culture model. Mechanically adjustable biomaterials such as hydrogels and fibrous scaffolds are being developed to replace traditional ECM. At the same time, the use of microporous arrays, 3D bioprinting, and microfluidics in organoids realizes high-throughput drug testing, immunotherapy screening, and vascularization in organoids [87]. Lung cancer PDO-related microchips and microfluidic platforms have been developed to enable large-scale drug testing within a week [9,69]. The researchers also developed an automatic IF multiplexing for FFPE sections from lung cancer PDO to detect the expression of multiple critical biomarkers on a single slide [88]. With machine learning, SigMaps can integrate and generate molecular interactions of specific tumor-associated proteins in lung cancer and perform large-scale validation of predicted proteins in PDO [89].
In conclusion, as an emerging preclinical model, lung cancer PDO has excellent potential and broad prospects in the research and treatment of lung cancer. Although there are still some challenges, the rapid development of culture technology will bring greater possibilities for the application of lung cancer PDO in clinical treatment guidelines.