Applications and challenges of patient-derived organoids in hepatobiliary and pancreatic cancers

Abstract

Hepatobiliary and pancreatic (HBP) cancers are among the most aggressive malignancies, with recurrence and metastasis driven by tumor heterogeneity and drug resistance, presenting considerable challenges to effective treatment. Currently, personalized and accurate treatment prediction models for these cancers are lacking. Patient-derived organoids (PDOs) tumor are three-dimensional in vitro models created from the tumor tissues of individual patients. Recent reports and our cultivation data indicate that the success rate of cultivating organoids for HBP cancers consistently exceeds 70%. The predictive accuracy of these tumor organoids has been shown to surpass 90%. However, PDOs still face notable limitations, especially in simulating the tumor microenvironment, including tumor angiogenesis and the surrounding cellular context, which require further refinement. While co-culture techniques and microfluidic platforms have been developed to mimic multi-cellular environments and functional vascular perfusion, they remain insufficient in accurately recapitulating the complexities of the in vivo environment. Additionally, PDOs are needed to fully assess their potential in predicting the efficacy of multi-drug combination therapies. This review provides an overview of the applications, challenges, and prospects for organoid models in the study of HBP cancer.

Key Words: Patient-derived tumor organoids; Hepatobiliary and pancreatic cancers; Tumor microenvironment; Co-culture; Drug prediction

Core Tip: Hepatobiliary and pancreatic (HBP) cancers are highly aggressive, with recurrence and metastasis driven by tumor heterogeneity and drug resistance, posing significant treatment challenges. Current personalized and accurate prediction models are lacking. Patient-derived organoids (PDOs) tumor, three-dimensional in vitro models from patient tumor tissues, show over 70% cultivation success and over 90% predictive accuracy. However, PDOs face limitations in simulating the tumor microenvironment despite advances in co-culture and microfluidic techniques. Additionally, PDOs' potential in predicting multi-drug therapy efficacy requires further assessment. This review outlines the applications and challenges of organoid models in HBP cancer research.

INTRODUCTION

Hepatobiliary and pancreatic (HBP) cancers are highly aggressive, with some of the highest incidence and mortality rates worldwide[1]. According to the latest cancer statistics, the 5-year relative survival rates for pancreatic cancer and liver cancer (including intrahepatic bile duct cancer) between 2013 and 2019 were 13% and 22%, respectively, making them the lowest and second-lowest survival rates among all cancer types[2]. Typically, HBP cancers progress rapidly and are associated with a high risk of recurrence and metastasis, owing to their high heterogeneity and resistance to therapy, creating significant challenges for treatment[3-5]. Although emerging therapies, such as immune checkpoint inhibitors and targeted therapies, have led to some improvements in survival outcomes, only a small number of patients currently benefit from them[6]. Especially in cholangiocarcinoma and pancreatic cancer, conventional chemotherapy remains the mainstream treatment option, highlighting the need to further optimize treatment strategies[7,8]. Current approaches to predicting tumor response to treatment are still quite limited. One of the more popular methods involves using NGS to identify tumor-related gene mutations and molecular features, which can help guide treatment decisions for certain patients. However, given that the underlying mechanisms and therapeutic targets for HBP cancers are still being explored, the high cost and limited availability of these methods pose significant barriers to clinical implementation[9].

Patient-derived organoids (PDOs) tumor are three-dimensional (3D) cell culture models derived from patient tissue, preserving the genomic and pathological characteristics of the original tumor while closely preserving its biological features and tissue structure in vivo. These models are promising means of informing clinical treatment decisions[10,11]. Organoids were first described in 2009[12], and PDOs were successfully established in 2011[13]; now, PDOs are widely applied in cancer research. PDOs typically require only a small number of cells initially and can proliferate rapidly under specific conditions, greatly reducing both culture time and costs[14]. Furthermore, PDOs maintain genetic stability and demonstrate strong modelling accuracy and reliability, even during long-term culture[15]. These organoids have proven invaluable in investigating disease mechanisms and conducting high-throughput drug screening as approaches towards advancing precision medicine[16]. While the average success rate for generating cancer organoids exceeds 70%[17], success rates vary significantly across different types of PDOs. For instance, the success rate for generating colon cancer organoids can reach up to 90%[18], whereas it remains below 20% for prostate cancer[19]. Although the different tumor culture rates are related to their invasiveness and degree of differentiation, considering that most current culture techniques are derived from colon cancer, there is still a need to improve corresponding culture methods for other tumors to meet the demands of future clinical applications[13].

Recent studies into HBP cancer have utilized PDOs to identify tumor-associated molecular markers[20], assess metabolic alterations[21], and perform high-throughput drug screening[22]. Current organoid models, however, fail to accurately replicate key aspects of the tumor microenvironment, such as blood perfusion and immune cell infiltration. This limitation significantly reduces the precision and reliability of drug sensitivity assays[16,23]. To address these limitations, co-culture technologies and microfluidic platforms have been developed. Moreover, due to the highly malignant nature of HBP cancer, combination therapy is standard practice, highlighting the need for further investigation into the sensitivity of multi-drug regimens[7,8,24]. Our team has successfully established 180 PDOs following the process outlined in the flowchart, including 83 liver cancer samples, 61 biliary tract cancer samples, and 36 pancreatic cancer samples (Figure 1). The average success rate for organoid formation was 77.65%, and drug sensitivity tests were performed on 7 conventional drug regimens[25,26]. Additionally, we also improved the success rate of PDO culture by utilizing metabolites (such as lactate) and performed tumor–macrophage co-culture to further simulate the tumor microenvironment[27-29].

Figure 1 Flow chart of patient-derived hepatobiliary and pancreatic cancer organoid culture and its applications in our team. The organoid culture process involves dissociating fresh tumor tissue into small pieces, followed by digestion, dissolution, filtration, centrifugation, and resuspension. The cell is then embedded in Matrigel and cultured in a medium enriched with relevant factors in an incubator. PDO: Patient-derived organoid. Created by BioRender (BioRender.com).

This review aims to summarize the current applications of HBP tumor organoids in clinical treatment, analyze existing challenges, and discuss future research directions. Specifically, based on the team's previous research, it analyzes the factors contributing to the current variations in culture success rates, uncovers the limitations of microenvironment simulation and combination drug therapy prediction, and provides preliminary insights into potential solutions, such as co-culture models and microfluidic chips.

APPLICATION OF HBP TUMOR ORGANOIDS

Liver cancer organoids

The success rates for developing liver cancer organoid cultures are reported to be from 26% to 75.6% (Table 1)[11,14,22,25,30-44], with half of the studies reporting rates below 50%. In contrast to other gastrointestinal cancers, liver cancer organoids exhibit lower success rates, which might be due to the absence of epithelial stem cell characteristics in hepatocellular carcinoma (HCC): This is known to impact the proliferation of HCC cells in organoid culture systems[22]. Furthermore, the success rate of organoid culture is linked with the degree of differentiation and the proliferation rate of the tumor itself. Some studies have indicated that only moderately-to-poorly differentiated liver cancers with a Ki-67 index greater than 5 are likely to successfully form organoids[30,31]. In addition, tumor heterogeneity and the specific sampling region also play a role in influencing culture outcomes[11]. There is also evidence reporting a 26% success rate for liver cancer biopsy cultures, which is lower than that for surgical resection samples, possibly due to the insufficient amount of fresh tissue available, which limits the number of initial cells for culture[45]. Altogether, while liver cancer organoid models have proven to be accurate in predicting tumor behavior[17,22,31], further validation of the accuracy of the prediction is crucial. Additionally, there is a need to further increase the number of cohorts to better guide clinical drug treatment based on organoid drug sensitivity results. Additionally, HCC patients show low sensitivity to chemotherapy, with the combination of immunotherapy and anti-angiogenic drugs frequently being the preferred treatment option[24,46]. Given the known unique features of the immune microenvironment and the high vascularization of liver cancer, liver tumor organoids and their co-culture with immune cells, in addition to their vascularization, are considered crucial for accurately predicting therapeutic responses[17]. Compared to traditional PDO models, co-culture models incorporating immune cells and cancer-associated fibroblasts demonstrate comparable effectiveness in assessing responses to chemotherapy or targeted anti-cancer therapies[32]. Moreover, these models have exhibited a higher predictive potential for responses to anti-programmed death ligand-1 compounds[32].

Table 1 Types of established patient-derived hepatobiliary and pancreatic cancer organoids.

Tumor type	Sample source	Success rate (sample size)	Drug testing	Time	Quality grade	Ref.
HCC, cholangiocarcinoma, combined hepatocellular cholangiocarcinoma	Res	25.0% (3/12), 75.0% (3/4), 100.0% (2/2)	29 anti-cancer compounds	2017	Low	Broutier et al[31]
HCC, ICC	Bx	33.0% (8/24), 60.0% (3/5)	Sorafenib	2018	Low	Nuciforo et al[30]
HCC, ICC	Res	40.9% (27/66)	Common clinical regimens	2024	Medium	Rao et al[22]
HCC, ICC	Res	75.6% (399/528)	Common clinical regimens	2024	High	Yang et al[11]
HCC, ICC	Res	> 60% (64/N/A)	301 compounds	2024	High	Walz et al[47]
HCC	Res or Bx	54.5% (12/22)	Common clinical regimens	2023	Medium	Zou et al[32]
ICC, GBC, PDAC	Res	50% (3/6), 20% (1/5), 50% (1/2)	339 anti-cancer compounds	2019	Low	Saito et al[35]
GBC	Res	12.2% (5/41)	29 anti-cancer compounds	2022	Low	Yuan et al[36]
ICC	Res or Bx	69.5% (16/23)	Gemcitabine and cisplatin	2023	Medium	Lee et al[33]
ECC, GBC, ICC	Res, Res, Res or Bx	81.3% (13/16), 40.0% (4/10), 82.6% (38/46)	Common clinical regimens	2023	Medium	Ren et al[34]
GBC, ECC	Res	85.7% (6/7)	Common clinical regimens	2021	Low	Wang et al[25]
PDAC	Res	80.0% (8/10)	N/A	2015	Low	Boj et al[37]
PDAC	Res or Bx	73% (101/138)	Fluorouracil	2018	High	Tiriac et al[38]
PC, ECC	Res or Bx	62.6% (52/83)	Common clinical regimens	2019	High	Driehuis et al[43]
IPMN	Res	81.3% (13/16)	N/A	2021	Medium	Beato et al[39]
PDAC	Res or Bx	37%-60% (36/N/A)	Common clinical regimens	2024	Medium	Kim et al[40]
PC	Bx	48.1% (13/27)	Common clinical regimens	2023	Medium	Choi et al[42]
PC	Bx	93.3% (14/15)	Common clinical regimens	2024	Medium	Yang et al[14]
Cholangiocarcinoma, GBC, perihilar cholangiocarcinoma, PDAC, IPMN	Res	60.0% (12/20), 100.0% (6/6), 100.0% (2/2), 40.0% (8/20), 33.3% (2/6)	Integrin-linked kinase inhibitor	2021	Medium	Shiihara et al[41]

Studies with fewer than 10 successful cultures are classified as having low prevalence, those with 10–50 successful cultures are considered to have moderate prevalence, and studies with more than 50 successful cultures are considered to have high prevalence. Bx: Biopsy; ECC: Extrahepatic cholangiocarcinoma; GBC: Gallbladder cancer; HCC: Hepatocellular carcinoma; ICC: Intrahepatic cholangiocarcinoma; IPMN: Intraductal papillary mucinous neoplasm; N/A: Not available; PC: Pancreatic cancer; PDAC: Pancreatic ductal adenocarcinoma; Res: Resection.

Biliary tract cancer organoids

Based on recent studies, the success rate of culturing biliary tract cancer organoids has increased to 69.5%-74.4%[33,34], which is a significant improvement compared to the rates of 12.2%-36.4% reported in earlier studies[35,36]. The challenge in culturing biliary tract cancer organoids is primarily attributed to the small size of surgical resection samples. In addition, these resection samples often contain a high proportion of non-cancerous stromal cells, such as fibroblasts, which reduce the relative number of tumor cells, leading to contamination of the organoid culture. The success of culturing organoids can be enhanced further by microscopically isolating and extracting tumor organoids from the contaminating non-tumor cells[35]. In terms of culture sources, cholangiocarcinoma organoids are primarily cultured from solid tissue samples. Unlike bladder cancer PDOs, which are cultured from urine, no studies have reported culturing cholangiocarcinoma organoids from liquid biopsy sources, such as bile[47]. Regarding organoid function, cholangiocarcinoma organoids have shown a 92.3% accuracy in predicting drug sensitivity to clinical treatments[34]. Moreover, organoid morphology can be used to classify cholangiocarcinoma subtypes, helping to guide treatment decisions based on the gene expression profiles of the organoids[33].

Pancreatic cancer organoids

The success rate for establishing organoid culture with pancreatic cancer tissue ranges from 37% to 93% (Table 1)[11,14,22,25,30-44], with approximately 50% of studies reporting success rates of around 80%[37-39]. It has been found that tumor size, whether from surgical resection or biopsy, is a limiting factor for successful pancreatic cancer organoid culture[40]. Additionally, specimens obtained after neoadjuvant chemotherapy are particularly challenging to grow in organoid culture, often resulting in little-to-no growth or very slow growth[41]. Furthermore, pancreatic cancer organoids have been successfully cultured from ascitic fluid and other liquid biopsy sources, with a success rate of 48.1%. Drug sensitivity testing has shown that clinical responses observed in patient-derived samples can be replicated[42]. However, the limited availability of clinical data on treatment regimens and patient responses has hindered the predictive accuracy of pancreatic organoid models[43]. Despite this, existing data still support the potential of organoids in guiding pancreatic cancer treatment. Models based on drug sensitivity results can accurately predict responses in treatment-naïve patients, with prediction accuracies of 91.1% for first-line and 80.0% for second-line therapies. Based on this study, patients receiving treatments guided by predicted tumor sensitivity have demonstrated significantly longer progression-free survival (PFS) compared to those treated with regimens predicted to be resistant[48]. Additionally, the transcriptomic features found from drug sensitivity assays have been shown to reflect treatment responses in pancreatic cancer patients[38,40,43]. For example, patients with gemcitabine-sensitive characteristics demonstrated significantly prolonged PFS under gemcitabine treatment, while gemcitabine sensitivity features did not predict PFS or overall survival (OS) improvement in untreated cohorts, indicating that these features are drug-dependent[38].

The success rates of HBP cancer organoid cultures reported in the literature vary significantly[19]. This study summarizes key findings in the culture of HBP tumor organoids and drug sensitivity testing (Table 1)[11,14,22,25,30-44]. Based on the number of successfully cultured tumor organoids, we have provided a simple rating for the current research on HBP tumor organoids. Studies with fewer than 10 successful cultures are classified as having low prevalence, those with 10–50 successful cultures are considered to have moderate prevalence, and studies with more than 50 successful cultures are considered to have high prevalence. The culture success rates of HBP tumor organoids are influenced by factors such as specimen source (biopsy, surgical resection, post-chemotherapy), specimen properties (size, differentiation, invasiveness), sample size, and culture conditions[33]. Furthermore, the definition of cultural success varies. Typically, successful organoid culture is defined by the manifestation of appropriate organoid morphology, pathological features, and genomic consistency with the parental tissue. However, some studies define success as stable and continuous passaging for over a year[35]. Different studies also use varying methods to calculate success rates. Some calculate success based on the number of cultured tissues, while others use the number of patients as the denominator. When multiple samples are taken from the same patient, the calculation of culture success rates may differ[30]. However, continuous advancements in culture techniques have markedly improved the overall success rates of organoids compared to earlier studies[44]. Furthermore, although the efficacy of using organoid-based drug sensitivity molecular features and direct drug sensitivity results to guide therapeutic strategies has been partially validated, higher quality evidence is still required to further confirm the clinical reliability of these approaches[11,34,48]. As organoid culture methodologies continue to expand, this issue is expected to be addressed progressively.

CURRENT APPLICATION BOTTLENECKS OF HBP TUMOR ORGANOIDS

Tumor microenvironment simulation

In the HBP cancer microenvironment, the interactions between cells, including immune cells and stromal cells[49], play a crucial role in tumor initiation and progression[50,51]. Additionally, the spatial distribution differences of these cells within tissues also contribute to tumor heterogeneity. Despite originating from the same primary tumor, tissues from different tumor regions exhibit varying proportions of cellular components, leading to the formation of heterogeneous organoids[10,52]. When tumor PDOs are initially cultured, they retain components of the surrounding microenvironment. However, over time, only the tumor cell-related components persist[53]. Additionally, organoids lack a functional vascular system[16,54], which plays a crucial role in tumor growth and metastasis by supplying essential nutrients and providing pathways for tumor cells to disseminate to distant sites[55]. Furthermore, endothelial cells contribute not only to vascular formation but also to cancer organoid proliferation by secreting angiogenic factors[56], which promote immune suppression, metabolic alterations, therapy resistance, and tumor metastasis[57].

A relevant strategy to simulate the tumor microenvironment is co-culturing PDOs with tumor-associated cells (Table 2)[27,32,51,58-73]. The cell types currently utilized in co-cultures of HBP tumor organoids include macrophages, lymphocytes, cancer-associated fibroblasts, natural killer cells, peripheral blood mononuclear cells, and endothelial cells. Studies have shown that co-culturing with tumor-associated macrophages (TAMs) leads to increased expression of genes related to proliferation and cell survival in pancreatic ductal adenocarcinoma (PDAC) organoids. This process is thought to promote organoid growth through pathways such as RAS protein signaling[58]. Moreover, co-culture successfully established the diversity of TAMs, including subtypes that regulate tumor angiogenesis and growth, thereby recapitulating the complexity of the PDAC tumor microenvironment[58]. Compared to organoids cultured alone, bile duct organoids co-cultured with TAMs exhibited reduced sensitivity to gemcitabine and cisplatin. Clinical efficacy assessments indicate that the co-culture model provides a more accurate prediction of drug responses than single-culture organoids[27]. Similarly, in pancreatic cancer co-cultures, macrophages were observed to impart tumor stemness and confer chemoresistance to drugs such as gemcitabine[59].

Table 2 Co-culture model of patient-derived organoids for hepatobiliary and pancreatic cancer.

Tumor cells	Co-culture cells	Ratio (co-culture cells: Tumor cells)	Culture time (day)	Pattern	Ref.
HCC	T cells	10:1	1	Direct 3D co-culture	Liu et al[62]
HCC, intrahepatic cholangiocarcinoma	CAFs	2:1	10-14	Transwell culture	Liu et al[51]
HCC	PBMC, MSC	30:1:10	7	Chip	Zou et al[32]
HCC	T cells	10:1	1	Direct 3D co-culture	Zou et al[63]
Cholangiocarcinoma	T cells	25–50:1	7	Direct 3D co-culture	Zhou et al[64]
Extrahepatic cholangiocarcinoma	TAMs	1:5	14	Transwell culture	Guo et al[27]
PDAC	MSC, T cells	N/A	5	Direct 3D co-culture	Tai et al[65]
PDAC	CAFs	2:1	5	Transwell culture	Tao et al[60]
PDAC	TAMs	1:1	6	ALI culture	Tabe et al[58]
PDAC	MSC, endothelial cells	60:21:30	7	ALI culture	Takeuchi et al[61]
PDAC	CAFs	1:1	10	Transwell culture	Sheng et al[66]
PDAC	T cells	10:1	3	Hydrogel chip	Lahusen et al[67]
PC	TAMs	3:1	N/A	Transwell culture	Jiang et al[59]
PDAC	NK cells	2-5:1	3	Transwell culture	Beelen et al[68]
PC	CAFs	1:1	3	Transwell culture	Schuth et al[69]
PDAC	CAFs	20:1	10	Transwell culture	Shinkawa et al[70]
PDAC	T cells	1:1	14	Direct 3D co-culture	Meng et al[71]
PDAC	PBMC, NK cells	N/A	7–14	Direct 3D co-culture	Marcon et al[72]
PDAC	CAFs, T cells	N/A	3	Direct 3D co-culture	Tsai et al[73]

ALI: Air-liquid interface; CAFs: Cancer-associated fibroblasts; HCC: Hepatocellular carcinoma; MSC: Mesenchymal stem cells; N/A: Not available; NK: Natural killer; PBMC: Peripheral blood mononuclear cells; PC: Pancreatic cancer; PDAC: Pancreatic ductal adenocarcinoma; TAMs: Tumor-associated macrophages; 3D: Three-dimensional.

Currently, most studies are limited to co-cultures involving a single cell type[60], with research on multi-cellular co-cultures being scarce[61,74,75]. Co-culture systems are primarily divided into two formats: (1) Direct 3D co-culture; and (2) Indirect co-culture, which includes Transwell chambers, air–liquid interface, and microfluidic chips. Organoid-on-chip models are 3D, micro-engineered systems that incorporate multi-cellular layers, tissue interfaces, and continuous perfusion chambers for human microvascular networks. These systems optimize nutrient and oxygen delivery while facilitating waste removal[76]. In addition to optimizing co-culture conditions for tumor organoids and associated cells, organoid-on-chip models can also enable dynamic drug delivery. A microfluidic chip designed by Zou et al[32], which co-cultures HCC organoids with macrophages, has been shown to increase the success rate of HCC organoid culture while reducing the amount of culture medium required and lowering the time and costs associated with high-throughput drug screening. Furthermore, it demonstrated more accurate predictions of drug responses. Vascularized organoids are commonly generated in vitro by co-culturing tumor organoids with endothelial cells using microfluidic platforms[16], and these models are applied to study tumor metastasis and targeted therapies[54]. The study also demonstrated that co-culturing HCC organoids with endothelial cells resulted in the upregulation of angiogenic signals, such as monocyte chemoattractant protein-1, thereby influencing tumor progression[74].

Challenges in clinical prediction using tumor organoids

Patients with HBP cancer experience high rates of recurrence and metastasis, coupled with poor sensitivity to chemotherapy agents[77]. Despite progress in the development of targeted therapies and immunotherapies, which have shown promise in both preclinical and clinical studies, only a small proportion of patients experience effective responses[78]. Unfortunately, tumor heterogeneity and individual genetic variations contribute to considerable variability in drug sensitivity[79]. Existing evidence suggests that using organoids to test drug sensitivity offers valuable insights for single-agent treatment regimens. Advanced techniques, including genomics-based strategies, can further identify drug sensitivity traits, enabling stratified predictions for patients[11]. However, there is a lack of large-scale cohort studies directly validating these strategies[34,48]. This limitation is largely because HBP tumor organoids are often derived from surgically excised specimens, complicating the clinical evaluation of drug responses[43]. Therefore, improving the culture rates of late-stage biopsy specimens or establishing standardized methods for assessing drug sensitivity following post-surgical adjuvant therapy is critical for advancing organoid-based applications.

Given the aggressive nature of HBP tumors, monotherapy is typically ineffective, and combination therapies are the standard treatment approach[7,24,80]. Identifying optimal drug combinations that benefit patients is a key therapeutic strategy. However, organoids currently cannot predict sensitivity to combination therapies directly[81]. Encouragingly, by integrating algorithms with PDO monotherapy drug sensitivity results and treatment regimens in practice, data models have been developed, demonstrating that better therapeutic outcomes are associated with higher matching between combination drug regimens and sensitive drug phenotype[82]. For example, in the treatment of PDAC, patients who received well-matched neoadjuvant and adjuvant chemotherapy exhibited significant improvements in both recurrence-free survival and OS. When the match was poor, the median recurrence-free survival was 8.5 months, whereas with a good match, it increased to 15.9 months[82]. Furthermore, in patients who have previously undergone treatment, prior therapies may interfere with organoid-based drug sensitivity predictions, leading to reduced accuracy[48].

FUTURE DIRECTIONS

The clinical application of HBP cancer PDOs depends on the maturation of organoid culture technologies. Optimizing culture medium components and refining culture techniques are essential strategies aimed at improving organoid culture efficiency[83,84]. To further enhance the clinical utility of PDOs and reduce reliance on tissue samples will be a key focus for the translational application of HBP tumor organoids, particularly enabling PDOs derived from biopsy samples for large-scale drug sensitivity testing. In support of this, there are studies that report the successful culture of tumor organoids from ascitic fluid and pleural effusions[42,85]. In the future, HBP tumor organoids may be derived from liquid biopsy samples, such as peripheral blood and bile, thus expanding the clinical diagnostic capabilities and applications of organoids[86]. Although studies also indicate that the cell count in ascitic or pleural fluid is relatively low, with tumor cells not being detected in approximately 46% of the fluid samples[42].

With the rapid advancement of medical biotechnology, the development of new biomaterial platforms is expected to synergize with microfluidic chips and organoid culture technologies. The integration of organoids with biotechnological advances will improve culture efficiency, reduce costs, and provide technical support for high-throughput drug screening[87,88]. Establishing standardized and automated systems for sampling, culturing, and testing will not only mitigate the impact of HBP tumor heterogeneity but also minimize operational errors, laying the foundation for large-scale applications of tumor organoids[32,35,89]. Moreover, novel imaging techniques, such as real-time imaging and droplet assays, will provide more accurate observations of tumor cell–immune cell interactions, enhancing the precision of tumor microenvironment simulations within organoids[90]. In addition, the introduction of next-generation sequencing will allow deeper exploration of the genomic landscape of tumors. By combining drug sensitivity data derived from these technologies with clinical data, these technologies will aid in identifying molecular signatures and therapeutic targets associated with drug sensitivity, ultimately guiding personalized treatment strategies[91-93]. Even more promising is the potential for artificial intelligence to further integrate the aforementioned materials, imaging, and sequencing techniques. The non-invasive, automated organoid evaluation process developed with artificial intelligence could enhance large-scale organoid analysis capabilities and significantly improve the accuracy of clinical drug predictions[94,95]. The breakthroughs in these technologies will advance HBP tumor organoid research to new heights, fostering significant progress in clinical diagnosis and treatment.

CONCLUSION

HBP cancer organoids have emerged as crucial tools for high-throughput drug screening and precision medicine. The success rate of organoid culture currently varies significantly, and the cultivation techniques need further improvement. Additionally, standardized protocols for cultivation still need to be established. While the drug prediction capabilities of organoids have been preliminarily validated, further high-quality evidence is required to strengthen their application, especially in predicting multi-drug combinations. The tumor microenvironment and vascularization are features that are challenging to model and, therefore, remain major limitations in improving the accuracy of organoid-based predictions. Co-culture models and microfluidic chip technologies show promise as potential solutions, although they are still in the early stages of development. In the future, with continuous integration of genomics, engineering, and information technologies, organoids are expected to become a powerful tool in clinical diagnostics and therapy.