NSCLC organoids are predominantly established from surgical tumor specimens (Ma J. et al., 2025). Typically, tissues are mechanically dissociated into small fragments and digested enzymatically using collagenase or neutral protease at 37 °C to yield single cells or small clusters, which are subsequently cultured to form NSCLC organoids (Kim and Park, 2025). To improve tumor cell purity and better recapitulate tumor heterogeneity, Li Z. et al. (2021) optimized the use of minimally invasive biopsy specimens or malignant pleural effusions, reducing contamination from normal pulmonary cells. The efficiency of tumor cell isolation depends critically on enzymatic digestion parameters, including enzyme type, concentration, digestion mode, and duration, which influence both cell yield and viability. To mitigate enzymatic damage to tumor cells, preserve tissue integrity, and streamline cell isolation, Tamura et al. (2018) and Takahashi et al. (2019) developed a method using mechanically minced tumor tissue for suspension culture. After stromal depletion, prolonged culture (3–6 months) in matrix-free medium yielded a biobank of 15 NSCLC organoids. Li Z. et al. (2021), Li Z. et al. (2020) observed that some NSCLC organoids exhibit limited proliferative capacity after serial passaging, regardless of enzymatic conditions. Alternative protocols (Li et al., 2018; Sachs et al., 2019) addressed this by employing mechanical dissociation during passaging, fragmenting organoids without enzymatic treatment, thereby improving long-term expansion. Further refinements in organoid handling have enhanced practicality. Kim et al. (2019) identified organoid diameter (100–150 μm) and storage temperature (4 °C) as critical factors for cryopreservation, achieving 70% post-thaw recovery via low-speed centrifugation and liquid nitrogen storage. Compared to 2D cultures, the 3D complexity of PDOs introduces additional challenges. To streamline culture conditions, Zhang et al. (2016) designed a superhydrophobic microwell array chip that isolates organoids using hydrophobic microstructures. This platform simplifies medium exchange, improves viability, and increases culture success rates by reducing technical variability.

Another critical challenge in the establishment of NSCLC organoids is the maintenance of genetic stability over prolonged culture periods (Li Z. et al., 2024). As organoids undergo long-term expansion, genetic drift that manifested as the accumulation of somatic mutations and chromosomal aberrations can occur, potentially altering the tumor’s genomic landscape (Nam et al., 2022). This genetic instability can profoundly impact drug response data by introducing mutations that confer resistance to certain therapies or enhance sensitivity to others (Broutier et al., 2017). For instance, mutations in the EGFR signaling pathway, such as the acquisition of the T790M mutation, could render organoids resistant to EGFR inhibitors, despite initial sensitivity (Lim et al., 2024). Similarly, mutations in KRAS, particularly G12C, could alter the tumor’s response to targeted therapies like EGFR-TKIs and MEK inhibitors by activating downstream MAPK and PI3K/AKT/mTOR pathways that promote tumor survival and drug resistance (Song et al., 2024; Ma X. et al., 2025). Genetic drift may also activate the Wnt/β-catenin pathway, which regulates tumor progression and stem cell maintenance, further contributing to therapy resistance (Yan et al., 2022). Additionally, ferroptosis, a regulated form of cell death suppressed by genetic alterations such as GPX4 loss (Liao et al., 2024), may be dysregulated in PDOs, particularly those exposed to chemotherapy, leading to enhanced resistance. These alterations, driven by genetic drift, underscore the potential for PDOs to evolve differently from the original tumor, complicating the predictive value of drug response assays. Therefore, long-term culture-induced genetic drift must be carefully considered when interpreting PDO-based drug sensitivity data.

A critical challenge in establishing NSCLC organoids from surgical specimens is the overgrowth of normal cells, which reduces tumor purity (Lee et al., 2021). To address this, Kumamot et al. (2008) found that nutlin-3a supplementation selectively eliminates non-tumor cells, promoting the expansion of TP53-mutant organoids. However, prolonged use of such agents may induce genomic instability and alter tumor-specific characteristics (Kucab et al., 2017). Alternative approaches, such as embedding dissociated tumor cells in matrigel with lung epithelial-optimized media, have proven more effective in suppressing normal cell growth while preserving cancer genomic integrity (Kim et al., 2019; Shi et al., 2020). Recent advances focus on leveraging organoids for personalized therapy (Han et al., 2022; Kim SK. et al., 2021). For instance, Banda et al. (2020) cultured organoids in erlotinib-containing matrigel, monitoring genomic evolution and resistance mechanisms. Their findings recapitulated acquired resistance in NSCLC, providing insights for therapeutic optimization. Beyond primary tumor-derived models, genetically engineered normal organoids offer a controlled system to study carcinogenesis. Naranjo et al. (2022) induced oncogenic mutations in normal lung organoids, triggering malignant transformation and tumor formation upon murine transplantation. Similarly, Dost et al. (2020) generated KRAS-mutant organoids, which exhibited diminished epithelial biomarker expression, validating their use in early-stage cancer modeling. While 3D matrigel-based systems (Shi et al., 2020) facilitate tumor-specific studies, their lack of native stromal and immune components restricts microenvironmental recapitulation (Yuki et al., 2020). Consequently, co-culture models integrating tumor organoids and immune cells are emerging as a vital tool to investigate tumor-immune crosstalk, potentially advancing NSCLC immunotherapy research (Dijkstra et al., 2018; Nakamura et al., 2019).

PDOs demonstrate significant potential in predicting patient-specific drug responses and elucidating resistance mechanisms. Li YF. et al. (2020) evaluated five natural compounds in PDOs, revealing distinct response patterns: berberine exhibited selective efficacy in PDOs but not in conventional cell lines, while betaine induced resistance in both models. Notably, PDOs also enable real-time monitoring of acquired resistance. Halofuginone (HF) was shown to suppress proliferation, induce cell cycle arrest, and restore cisplatin sensitivity in resistant PDOs, confirming its dual role as a chemosensitizer and pathway inhibitor (Li H. et al., 2021). Further mechanistic studies identified recombinant DCLK1 as a key regulator of cancer stemness via Wnt/β-catenin signaling, driving EGFR-TKI resistance. Yan et al. (2022) demonstrated that DCLK1 inhibition restored TKI sensitivity and synergized with EGFR inhibitors in TKI-resistant PDOs. These findings highlight the role of DCLK1 in promoting stemness and its potential as a therapeutic target in overcoming resistance. Additionally, Weng et al. (2023) developed a novel topoisomerase I inhibitor with potent activity against refractory EGFR-mutant organoids, including those with ex19del/T790M/C797S mutations, and demonstrated synergistic effects with PARP/ATR inhibitors and immunotherapy. These findings underscore PDOs’ utility in identifying drug-tolerant persister (DTP) cells and developing strategies to overcome therapeutic resistance (Mikubo et al., 2021) (Figure 1).

Resistance mechanisms associated with cancer stem cells are critical for sustained tumor growth and survival. In particular, drug-tolerant persister (DTP) cells evade chemotherapies and targeted agents (Li et al., 2023a). MAPK inhibitors induce PINK1-mediated mitophagy in DTP cells, promoting their survival and proliferation (Li et al., 2023b). When combined with chloroquine, an autophagy inhibitor, MAPK inhibitors regain their cytotoxic effects, indicating that targeting mitophagy could offer a strategy to overcome resistance in DTP populations (Li et al., 2023b; Wang B. et al., 2023). Metabolic alterations also play a significant role in resistance to cancer therapies (Lin et al., 2023). One such example is ferroptosis, a form of regulated necrosis. Ferroptosis suppression by β5-integrin has been linked to chemotherapy resistance in PDOs. SRC or ceramidase inhibition reactivated ferroptosis in chemotherapy-resistant PDOs, thereby reversing drug resistance and providing a potential therapeutic avenue for targeting metabolic escape mechanisms in resistant NSCLC tumors (Beekhof et al., 2023; Luk et al., 2024). Despite these advancements, tumor organoids are not yet clinically approved for guiding therapy. Only when organoids exhibit drug responses consistent with matched patients can they be reliably used in personalized medicine (Xiang et al., 2024). Additionally, tumor heterogeneity complicates the evaluation of organoid predictive value. Thus, while promising, organoid-based drug screening remains a substantial challenge requiring further optimization (Table 1).

While patient-derived organoids (PDOs) are invaluable in drug screening, their utility in immunotherapy testing remains limited due to the absence of key immune components that are critical for assessing immune-related responses (Jeong and Kang, 2023; Wang M. et al., 2025). This limitation restricts the ability of PDOs to accurately model the immune microenvironment of tumors, which plays a significant role in the efficacy of immunotherapies, such as immune checkpoint inhibitors (Teijeira et al., 2022). To address this limitation, recent studies have established advanced co-culture platforms that incorporate immune components into PDOs, thereby reconstructing a more immunocompetent tumor microenvironment (Magré et al., 2023; Dao et al., 2022). Immune cells such as peripheral blood mononuclear cells (PBMCs), tumor-infiltrating lymphocytes (TILs), dendritic cells, macrophages, and even chimeric antigen receptor T cells (CAR-T cells) have been successfully integrated into NSCLC PDO systems. These immune cells can be derived either autologously from patient or from allogeneic sources (Sun et al., 2022; Li P. et al., 2024). Co-culture techniques vary, including embedding immune cells into Matrigel with PDOs, overlaying immune cells on organoid surfaces, and using organoid-on-a-chip models to preserve spatial architecture and facilitate direct immune-tumor interaction (Dao et al., 2022; Fang et al., 2023). To overcome this challenge, recent studies have developed co-culture systems that integrate immune cells with PDOs, thus allowing for a more accurate representation of the tumor-immune interaction. These tumor–immune organoid platforms have been used to evaluate immune checkpoint inhibitors, such as PD-1/PD-L1 and CTLA-4 antibodies, providing insights into their efficacy in a more physiologically relevant context (Amato et al., 2025; Ou et al., 2022). For example, Hélène et al. co-cultured NSCLC PDOs with IFN-γ, which led to enhanced immune response assessments and a better understanding of the tumor’s immune evasion mechanisms (Lê et al., 2024). Furthermore, PDO-based immune checkpoint evaluation systems have also been introduced, where immune cells are incorporated into PDO cultures to assess how tumors interact with immune cells and respond to checkpoint inhibition (Yao et al., 2025).

In addition to immune components, the incorporation of other stromal cell types, particularly endothelial cells, has gained attention for reconstructing a more comprehensive tumor microenvironment within PDO systems (Madan et al., 2025; Podaza et al., 2022). Endothelial cells play a pivotal role in angiogenesis, tumor perfusion, and immune cell trafficking, all of which influence drug delivery and response (Xu et al., 2024; Zhao et al., 2021). Advanced organoid-on-a-chip platforms and microfluidic co-culture systems now allow the inclusion of human endothelial cells alongside PDOs, enabling the formation of rudimentary vascular networks within the extracellular matrix (Zhao et al., 2021; Lim et al., 2021). These vascularized organoid models better replicate in vivo gradients of oxygen and nutrients, facilitate dynamic cell–cell interactions, and improve drug penetration modeling. For example, co-culturing NSCLC PDOs with human umbilical vein endothelial cells (HUVECs) under perfusion flow conditions has been shown to support neovascular structure formation and recapitulate tumor angiogenic responses to VEGF inhibitors (Ma et al., 2022b). Additionally, incorporation of pericyte-like stromal cells further enhances the maturation and stability of these networks (Wang X. et al., 2023; Phan et al., 2017).

Notably, studies utilizing TIL-PDO co-cultures have demonstrated that cytotoxic CD8⁺ T cells infiltrate the organoid structure and mediate cell killing in response to PD-1 blockade, thereby validating the translational potential of such systems (Dijkstra et al., 2018; Magré et al., 2023). Similarly, ALI-PDO platforms preserve endogenous immune infiltrates within the PDO structure, offering a superior representation of in situ immunity (Neal et al., 2018; Thorel et al., 2024). Furthermore, time-lapse imaging and cytokine release assays (IFN-γ, Granzyme B) have been used to functionally assess immune activation and killing capacity (Yuki et al., 2020; Gu et al., 2024). To further enhance the physiological relevance of PDOs and enable dynamic modeling of tumor microenvironments, microfluidic platforms will be an important component of next-generation organoid systems (Szewczyk et al., 2024; Sontheimer-Phelps et al., 2019). Notably, microfluidic devices allow precise spatial and temporal control over cell–cell and cell–matrix interactions (Bhatia and Ingber, 2014; DePalma et al., 2025), which are essential for simulating immune infiltration, angiogenesis, and metastatic dissemination within PDO models (Wang J. et al., 2025). These innovative approaches hold the potential to significantly improve the predictive accuracy of PDOs for immunotherapy responses and guide the development of personalized immunotherapy strategies.