Integrating traditional Chinese medicine with pancreatic cancer organoids: Bibliometric insights and translational perspectives

Abstract

BACKGROUND

Pancreatic ductal adenocarcinoma (PDAC) remains a highly lethal malignancy with a 5-year overall survival of approximately 12%. While patient-derived organoids have emerged as a predictive three-dimensional preclinical platform, traditional Chinese medicine (TCM) and natural compounds have concurrently demonstrated significant multitarget anticancer effects against PDAC.

AIM

To systematically characterize the knowledge architecture of pancreatic cancer organoids and TCM through bibliometric analysis, and identify potential structural gaps at their intersection.

METHODS

A comprehensive bibliometric analysis was conducted on 1066 original research and review articles retrieved from the Web of Science Core Collection spanning from January 2001 to May 2026. VOSviewer and CiteSpace were utilized to map global collaboration networks, institutional authorship, keyword co-occurrence, and citation ecosystems.

RESULTS

The global literature on PDAC organoids has shown sustained exponential growth, encompassing contributions from 58 countries led by the United States, China, and Germany. Bibliographic coupling and burst detection identified drug screening and drug resistance as the most intensely active current research frontiers. However, TCM and natural compounds were entirely absent across all six analytical dimensions examined. No TCM-affiliated institutions, authors, journals, or relevant keywords appeared within the collaborative or thematic networks of the PDAC organoid literature.

CONCLUSION

Multidimensional absence reveals a low correlation between pancreatic cancer organoid technology and research in TCM. Utilizing organoid platforms to evaluate the multitarget pharmacology of TCM represents a underdeveloped translational research direction in precision oncology.

Key Words: Pancreatic ductal adenocarcinoma; Patient-derived organoids; Traditional Chinese medicine; Bibliometric analysis; Drug screening; Natural compounds; Precision oncology; Tumor microenvironment

Core Tip: Pancreatic cancer is difficult to treat. Recently, scientists have started using organoids as a tool for testing new cancer drugs. Natural compounds from traditional Chinese medicine show potential because they can attack cancer cells in multiple ways. We analyzed > 1000 scientific papers to determine if researchers are using these advanced three-dimensional models to test traditional medicines. There was little research combining these two promising fields. By revealing this gap, we aim to encourage the scientific community to use patient-specific tumor models to evaluate natural medicines. This integration could accelerate the discovery of new, personalized treatments for pancreatic cancer, bridging the gap between natural remedies and modern precision medicine.

INTRODUCTION

Pancreatic cancer has emerged as one of the most lethal human cancers. Pancreatic ductal adenocarcinoma (PDAC) accounts for approximately 90% of all pancreatic malignancies. Due to the deep anatomical location of the pancreas and the asymptomatic nature of the disease in its early stages, > 80% of patients present with locally advanced or metastatic disease at the time of diagnosis, at which point curative surgical resection is no longer feasible[1]. Consequently, the 5-year overall survival rate remains approximately 12%, and the 10-year survival rate is only a devastating 1%[2]. The systemic treatment of PDAC has evolved incrementally over the past two decades, with gemcitabine combined with nab-paclitaxel and modified FOLFIRINOX representing the current first-line chemotherapeutic standards for metastatic disease[3]. Despite these advances, median overall survival for patients with metastatic PDAC remains < 1 year, and therapeutic resistance ultimately prevails in virtually all cases[1].

To address current clinical predicaments, organoid technology has emerged as a revolutionary preclinical platform[4]. Organoids are three-dimensional (3D) microstructures derived from stem cells or patient tumor tissues[5]. They self-organize in vitro and faithfully recapitulate the complex structural and functional features of native organs[6]. Compared to traditional two-dimensional (2D) cell cultures and animal models, organoids offer distinct advantages. They preserve patient-specific genetic and phenotypic heterogeneity[7]. They also accurately replicate the highly complex in vivo tumor microenvironment (TME)[8], and these innovative platforms enable high-throughput drug screening. They simultaneously involve fewer ethical restrictions and require significantly shorter experimental cycles[9]. This technology has already achieved remarkable success in modeling other malignancies, including colorectal and breast cancers[10]. These achievements provide valuable paradigms for pancreatic cancer research. Consequently, organoids currently represent the most promising platform to overcome experimental bottlenecks and guide precision medicine[11].

Pancreatic cancer organoid research was established through pioneering work by the Tuveson and Clevers laboratories, who developed 3D murine and human PDAC models that closely recapitulate tumor biology in vivo[12]. Landmark studies by Boj et al[13] and Tiriac et al[14] demonstrated that patient-derived organoids faithfully preserve tumor histology, genomic architecture, and intratumoral heterogeneity, while enabling functional drug sensitivity profiling correlated with clinical outcomes[15]. Compared with conventional 2D cultures, PDAC organoids maintain 3D cytoarchitecture, cell-cell interactions, and phenotypic fidelity unavailable in monolayer systems[16,17]. They can be rapidly generated from small biopsy specimens, including endoscopic ultrasound-guided fine-needle aspiration/biopsy samples, with a mean establishment success rate of 89.7%[17,18]. Coculture with cancer-associated fibroblasts and immune cells further enables TME recapitulation[19]. Current applications encompass drug sensitivity testing, gemcitabine resistance elucidation, molecular subtyping, and biomarker discovery[20,21]. Ex vivo pharmacotyping achieves clinical response concordance with an area under curve value of 0.80-0.90[22], positioning PDAC organoids as the most physiologically relevant platform for preclinical pharmacological evaluation.

While organoids provide the physiologically relevant platform, traditional Chinese medicine (TCM) offers a rich source of multitarget compounds; a synergy that remains largely untapped. TCM has a long history of clinical application in oncology and has attracted increasing scientific interest as a complementary treatment strategy for PDAC[23,24]. Several TCM-derived compounds have demonstrated meaningful anticancer activity in preclinical studies. Emodin, isolated from Rheum palmatum, inhibits pancreatic cancer cell proliferation, suppresses metastasis via nuclear factor kappa B downregulation, and represses tumor angiogenesis through the transforming growth factor-β/Smad pathway[25,26]. Oxymatrine, extracted from Sophora flavescens, induces apoptosis in PANC-1 cells and exerts antiangiogenic effects by suppressing nuclear factor kappa B-mediated vascular endothelial growth factor signaling[27,28]. Matrine inhibits proliferation and induces apoptosis in BxPC-3 and PANC-1 cell lines both in vitro and in xenograft models[29], while triptolide reverses hypoxia-induced epithelial-to-mesenchymal transition and cancer stem cell features[30]. Oridonin and ursolic acid have demonstrated the ability to overcome gemcitabine resistance in PDAC models[31]. Beyond single compounds, Astragalus polysaccharide enhances the antitumor efficacy of targeted agents[32], and Chinese herbal injections combined with chemotherapy have improved performance status and reduce treatment-related adverse reactions in clinical trials[33]. TCM thus offers a multitarget pharmacological profile theoretically suited to the genetic heterogeneity and therapeutic resistance of PDAC. However, the existing evidence base relies predominantly on 2D cell line assays and xenograft models, which inadequately recapitulate the complex TME and patient-specific biology of PDAC. A more physiologically relevant experimental platform is therefore required to systematically evaluate the therapeutic potential of TCM compounds in this disease.

Despite these parallel advances, the intersection of TCM and PDAC organoid research remains virtually unexplored. Current reviews of organoid applications in digestive system cancers and PDAC precision medicine consistently omit natural compounds and herbal medicine as pharmacological candidates for evaluation in this platform[34-36]. Conversely, preclinical TCM studies continue to rely on 2D monolayer cultures and xenograft models; neither of which recapitulates the patient-specific tumor heterogeneity and microenvironmental complexity that organoids offer[37,38]. No systematic synthesis of this cross-disciplinary space currently exists. This disconnect is particularly striking, as drug screening and resistance research represent not only the cutting-edge areas where organoid platforms demonstrate their greatest translational value, but also the fields in which multitarget TCM compounds are likely to offer their most distinctive pharmacological advantages. The absence of a unified knowledge framework for this intersection represents both a critical gap in current pancreatic cancer research and an unrealized translational opportunity.

To characterize this gap systematically, we applied bibliometric analysis to the PDAC organoid literature. Bibliometrics offers distinct advantages over narrative reviews. It quantifies publication trends, maps collaboration networks, identifies knowledge foundations through co-citation analysis, and detects emerging research frontiers through keyword burst detection. Crucially, it can objectively measure the presence or absence of specific research communities within a broader field. These properties make it particularly well suited to evaluating the integration status of TCM within an established scientific domain.

Firstly, in our research, we describe the global research landscape of PDAC organoids; its growth trajectory, geographic and institutional structure, and knowledge architecture. Secondly, we established the current status of TCM and natural compound research within this landscape, quantifying its scope and identifying the dimensions of its absence. Thirdly, we provided prospective direction for cross-disciplinary integration, identifying the research frontiers most amenable to TCM-based pharmacological investigation. This study aimed to comprehensively summarize the current application status, research progress, and future prospects of pancreatic cancer organoids in the fields of TCM and natural compounds, thereby providing novel insights for achieving innovative breakthroughs in this area of research.

MATERIALS AND METHODS

All bibliometric data were sourced from the Web of Science Core Collection, with the search conducted on May 16, 2026. First, we used a topic-based search formula: TS = ((“pancreatic cancer“ OR “pancreatic carcinoma“ OR “pancreatic adenocarcinoma“ OR “pancreatic ductal adenocarcinoma“ OR “PDAC“ OR “pancreatic neoplasm*“ OR “pancreatic tumor*“ OR “pancreatic tumor*“ OR “exocrine pancreatic cancer“ OR “pancreatic malignancy“ OR “cancer of the pancreas“) AND (organoid* OR “patient-derived organoid*“ OR “PDO“ OR “PDOs“ OR “tumor organoid*“ OR “tumor organoid*“ OR “cancer organoid*“ OR “pancreatic organoid*“ OR “3D model*“ OR “3D culture*“ OR “3D tumor model*“ OR “three-dimensional model*“ OR “patient-derived explant*“ OR “mini tumor*“ OR “mini tumor*“ OR “ex vivo model*“ OR “tissue organoid*“) AND (“traditional Chinese medicine“ OR “Chinese medicine“ OR “Chinese herb*“ OR “herbal medicine“ OR TCM OR “natural product*“ OR phytochem*)). However, we found that this search formula yielded no results, so we deleted the part with TCM to continue our research. The final search formula was: TS = ((“pancreatic cancer“ OR “pancreatic carcinoma“ OR “pancreatic adenocarcinoma“ OR “pancreatic ductal adenocarcinoma“ OR “PDAC“ OR “pancreatic neoplasm*“ OR “pancreatic tumor*“ OR “pancreatic tumor*“ OR “exocrine pancreatic cancer“ OR “pancreatic malignancy“ OR “cancer of the pancreas“) AND (organoid* OR “patient-derived organoid*“ OR “PDO“ OR “PDOs“ OR “tumor organoid*“ OR “tumor organoid*“ OR “cancer organoid*“ OR “pancreatic organoid*“ OR “3D model*“ OR “3D culture*“ OR “3D tumor model*“ OR “three-dimensional model*“ OR “patient-derived explant*“ OR “mini tumor*“ OR “mini tumor*“ OR “ex vivo model*“ OR “tissue organoid*“)). The wildcard operator (*) was used to capture all morphological variants of the root terms. Publications were restricted to January 2001 to May 2026. To ensure analytical rigor, only original research articles and reviews were retained for further analysis. Document types considered peripheral to the scope of this study, including conference abstracts, editorial materials, letters, book chapters, corrections, news items, and retracted publications were systematically excluded. For each included record, the following fields were extracted: (1) Article title; (2) Author names; (3) Institutional affiliations; (4) Country of origin; (5) Source journal; (6) Publication year; and (7) Citation metrics. A total of 1066 eligible articles were ultimately selected and included in the study scope. Bibliometric mapping and visualization were performed using VOS viewer (version 1.6.20) and Cite Space (version 7.0.0) (Figure 1).

Figure 1 Prisma diagram for screening literature on pancreatic cancer organoids. WOSCC: Web of Science Core Collection.

RESULTS

Annual publication trends and overall research landscape

To characterize the temporal evolution of scholarly output in the field of pancreatic cancer organoids, an analysis of annual publication volume was conducted across the entire study period from 2001 to 2026. There has been a pronounced and sustained increase in the number of publications over the past two decades, conforming closely to an exponential growth model described by the equation y = 0.5181e^0.2539x (R² = 0.9257) (Figure 2). The high coefficient of determination indicated an excellent fit between the observed data and the exponential trajectory, confirming the reliability and robustness of this predictive trend.

Figure 2  Trend in the number of annual publications on pancreatic cancer organoid research.

During the early phase spanning from 2001 to 2015, annual publication output remained consistently low, with < 10 articles recorded per year. This protracted period of minimal activity reflected the nascent stage of organoid technology and the limited application of 3D culture systems to pancreatic cancer research at that time. Since 2016, the number of publications has increased annually, with the rate of growth exceeding that of previous periods. Research during this period established standardized operating procedures for patient-derived pancreatic tumor organoids and has driven the widespread adoption of this model system. The subsequent phase from 2019 onward was characterized by exponential acceleration in research activity. Annual publications increased from 50 in 2019 to 86 in 2020, surpassing 100 for the first time in 2021 (n = 117), and reaching successive peaks of 123 (2022), 165 (2024), and 177 (2025). As of the search date (May 16, 2026), publications for the current year had already reached 82; a number that, when extrapolated to a full calendar year, is consistent with a continued growth trend. However, the data for 2026 are partial, reflecting publications indexed up to the search date of May 16, 2026, and are therefore not directly comparable to full calendar-year counts (Figure 2).

The general bibliometric profile of this field – encompassing 1066 in clouded publications from 58 countries and regions, a cumulative citation counts of 42356 (average citations per item: 39.73), and an h-index of 92 – underscores the substantial academic impact and international scope of pancreatic cancer organoid research. The United States and Harvard University were identified as the most productive country and institution, attesting to the prominent role of North American research centers in shaping this field. These findings collectively demonstrate that pancreatic cancer organoid research has transitioned from an emerging niche into a rapidly expanding and globally recognized scientific discipline (Table 1).

Table 1 General bibliometric characteristics of the included publications on pancreatic cancer organoids (2001-2026).

Indicator	Value
Total number of included studies	1066
Research article	901
Review	165
Time span	2001-2026
Number of countries/regions involved	58
Sum of times cited	42356
Average citations per Item	39.73
H-index	92
Countries with the most publications	United States
Institutions with the most publications	Harvard University

Analysis of country- and institutional-level collaboration networks

As illustrated in the co-authorship network map, the contributing countries formed several distinct collaborative clusters. The United States occupied the most central node, with pronounced links to Germany, England, Italy, and the Netherlands, collectively constituting the primary backbone of international co-authorship in this field. China, despite its substantial publication volume, was positioned at the network periphery with comparatively sparse intercluster connections, suggesting a predominantly domestic collaborative orientation (Figure 3A). The overlay visualization encodes average publication year on a blue-to-yellow gradient (2021-2023), revealing an important temporal dimension. The United States, Germany, and most Western European countries appear in cooler hues, indicating earlier and more historically established engagement with the field. China, by contrast, appears in a distinctly warm yellow tone, identifying it as a late but rapidly accelerating contributor (Figure 3B).

Figure 3 Visualization of bibliometrics at the country level. A: The co-authorship network map of countries; B: The co-authorship overlay map of countries; C: Top 15 countries by number of publications.

The United States led in publication volume with 404 documents, followed by China (218) and Germany (166), with all remaining top-15 countries contributing < 100 publications each (Figure 3C). The comprehensive indicator table adds further nuance: While the United States dominated across all dimensions – total citations (23 422), average citations per document (57.98), collaboration links (32), and total link strength (299) – the Netherlands (average citations = 168.31), Sweden (127.83), and Switzerland (110.25) demonstrated disproportionately high citation efficiency relative to their output volumes. China’s average citation rate (22.17) and total link strength (112) lagged behind countries of comparable publication scale, suggesting that its contributions, while growing rapidly in quantity, are still consolidating in terms of international influence and collaborative integration (Table 2).

Table 2 Top 15 countries based on comprehensive indicators.

Rank	Country	Documents	Citations	Average citations	Norm citations	Links	Total link strength
1	United States	404	23422	57.98	492.40	32	299
2	China	218	4833	22.17	220.65	21	112
3	Germany	166	5891	35.49	160.94	27	185
4	Japan	90	3064	34.04	81.79	17	45
5	Italy	80	4813	60.16	95.28	24	99
6	Netherlands	65	10940	168.31	119.01	18	80
7	England	64	4444	69.44	92.54	23	97
8	South Korea	56	1244	22.21	50.07	7	18
9	France	55	2137	38.85	49.66	16	49
10	Spain	53	3532	66.64	57.99	19	84
11	Canada	36	2670	74.17	46.16	15	48
12	Sweden	24	3068	127.83	44.80	16	41
13	Australia	22	808	36.73	25.21	15	38
14	Switzerland	20	2205	110.25	26.09	14	33
15	Denmark	18	454	25.22	19.05	12	26

At the institutional level, the co-authorship network revealed multiple geographically coherent research clusters. Harvard Medical School occupied the most central position as the largest node, with the highest total link strength (121), reflecting its role as the primary institutional hub. A prominent North American cluster encompassed Harvard Medical School, Cold Spring Harbor Laboratory, Johns Hopkins University, Dana Farber Cancer Institute, and Memorial Sloan Kettering Cancer Center, collectively representing the core of high-output, high-impact research. A Chinese institutional cluster was similarly identifiable at the network periphery, formed by Shanghai Jiao Tong University, Sun Yat-sen University, Fudan University, and associated institutions (Figure 4A).

Figure 4 Visualization of bibliometrics at the institution level. A: The co-authorship network map of institutions; B: Top 15 institutions by number of publications.

Harvard Medical School ranked first in publication output (n = 52), while Cold Spring Harbor Laboratory, despite ranking second in documents (n = 36), recorded the highest citation impact of any institution (total citations = 8015; average citations = 222.64), reflecting the seminal nature of its contributions to pancreatic cancer genomics and organoid biology (Figure 4B and Table 3). Johns Hopkins University ranked third (n = 32, average citations = 124.66), completing the leading triad of North American institutions. Among the top 15, four Chinese institutions were represented – Shanghai Jiao Tong University (rank 9), Sun Yat-sen University (rank 10), the Chinese Academy of Sciences (rank 11), and Fudan University (rank 13) – all of which are comprehensive biomedical research universities.

Table 3 Top 15 institutions based on comprehensive indicators.

Rank	Institutions	Documents	Citations	Average citations	Total link strength
1	Harvard Med Sch	52	2786	53.58	121
2	Cold Spring Harbor Lab	36	8015	222.64	95
3	Johns Hopkins Univ	32	3989	124.66	74
4	German Canc Res Ctr	27	962	35.63	57
5	Tech Univ Munich	26	1142	43.92	54
6	Univ Nebraska Med Ctr	26	2325	89.42	48
7	Dana Farber Canc Inst	25	1716	68.64	88
8	Mem Sloan Kettering Canc Ctr	25	3996	159.84	74
9	Shanghai Jiao Tong Univ	25	782	31.28	31
10	Sun Yat Sen Univ	23	791	34.39	17
11	Chinese Acad Sci	22	715	32.50	40
12	Univ Verona	22	2647	120.32	34
13	Fudan Univ	21	596	28.38	32
14	Aix Marseille Univ	20	508	25.40	17
15	Univ Calif San Diego	20	780	39.00	48

Critically, no TCM-affiliated institution appeared in the network, indicating that China’s contribution to this field is driven exclusively by conventional oncological and biomedical centers. This structural absence highlights both the existing gap and the unrealized potential for integrating TCM-oriented research institutions into the pancreatic cancer organoid landscape.

Analysis of author collaboration networks and core author identification

The co-authorship network identified several geographically and institutionally coherent research clusters among authors meeting the minimum threshold of three publications. Tuveson, David A occupied the most central position in the network, serving as the primary hub connecting multiple subclusters, while Batra, Surinder K and Kleger, Alexander led two distinct peripheral clusters, reflecting independent but parallel research trajectories[39]. All identifiable core authors were affiliated with conventional biomedical or oncological institutions in North America and Europe. No author with a TCM-affiliated institutional background appeared within the network, further corroborating the institutional-level findings described above (Figure 5A).

Figure 5 Visualization of bibliometrics at the author level. A: The co-authorship network map of authors; B: Top 15 authors by number of publications.

Application of Price’s Law to the full author pool (n = 7723) yielded a minimum publication threshold of m ≥ 0.749√24 = 3.7, rounded up to 4. A total of 260 authors met this criterion, representing only 3.37% of all contributors, yet collectively accounting for 918 of the 1066 included publications (86.12%). This pronounced concentration of output among a small author elite is consistent with the highly skewed knowledge production patterns typical of rapidly emerging research fields.

Tuveson, David A led in publication volume (n = 24) and also recorded high citation impact (total citations = 4556; average citations = 189.83) (Figure 5B). However, a notable divergence between productivity and citation influence was observed: Clevers, Hans, despite ranking seventh in publication output (n = 14), accumulated the highest total citations (6732) and the highest average citations per document (480.86) among all top authors, identifying him as the single most influential contributor to the foundational knowledge base of this field (Table 4).

Table 4 Top 15 authors based on comprehensive indicators.

Author	Documents	Citations	Average citations	Norm citations	Links	Total link strength
Tuveson, David A	24	4556	189.83	53.37	15	52
Kleger, Alexander	18	390	21.67	17.28	15	58
Corbo, Vincenzo	16	2306	144.13	29.26	17	48
Tiriac, Herve	15	3102	206.80	30.15	14	39
Batra, Surinder K	15	483	32.20	19.53	10	61
Seufferlein, Thomas	15	329	21.93	13.89	8	49
Clevers, Hans	14	6732	480.86	47.79	13	19
Ponnusamy, Moorthy P	13	462	35.54	18.37	10	61
Reichert, Maximilian	11	319	29.00	6.86	2	6
Hwang, Chang-il	10	2683	268.30	26.05	18	37
Rauth, Sanchita	10	246	24.60	15.12	10	55
Park, Youngkyu	9	3670	407.78	37.46	12	33
Muthuswamy, Senthil K	9	1097	121.89	18.57	4	20
Gout, Johann	9	185	20.56	8.34	8	34
Huang, Ling	8	1088	136.00	16.75	4	18

Analysis of journal distribution and source characteristics

A total of 365 journals published the 1066 included articles. Cancers ranked first in publication output (n = 56, 5.25%), followed by Scientific Reports (n = 26) and International Journal of Molecular Sciences (n = 25) (Figure 6A). The top 20 journals collectively accounted for 372 publications, representing 34.90% of the total corpus, indicating a moderate degree of publication concentration.

Figure 6 Analysis of top 20 journals defined by different fields. A: Top 20 journals by number of publications; B: Top 20 journals by subject area; C: The Bradford curve of all included journals.

The comprehensive indicator table reveals that 15 of the top 20 journals are classified as Q1, with impact factors ranging from 2.6 (PLoS One) to 33.3 (Cancer Discovery). High-impact journals including Gastroenterology (IF = 25.9), Gut (IF = 26.2), Cancer Research (IF = 16.6), and Nature Communications (IF = 15.7) featured prominently, with the top 20 journals collectively accounting for 72.85% of total citations, reflecting a strong concentration of academic influence among a small number of high-quality outlets (Table 5).

Table 5 Top 20 journals based on publication titles comprehensive indicators.

Journal	Record count	% of 1066	IF	JCR
Cancers	56	5.25	4.4	Q2
Scientific Reports	26	2.44	3.9	Q1
International Journal of Molecular Sciences	25	2.35	4.9	Q1
Journal of Experimental Clinical Cancer Research	25	2.35	12.8	Q1
Clinical Cancer Research	24	2.25	10.2	Q1
Gastroenterology	22	2.06	25.9	Q1
Cancer Research	21	1.97	16.6	Q1
Nature Communications	21	1.97	15.7	Q1
Cancer Letters	20	1.88	10.1	Q1
Frontiers in Cell and Developmental Biology	19	1.78	4.3	Q2
Frontiers in Oncology	15	1.401	3.3	Q2
Advanced Science	13	1.22	14.1	Q1
Gut	13	1.22	26.2	Q1
Molecular Cancer Therapeutics	12	1.13	5.5	Q1
Advanced Healthcare Materials	11	1.03	9.6	Q1
Plos One	11	1.03	2.6	Q2
Cell Reports	10	0.94	6.9	Q1
Oncogene	10	0.94	7.3	Q1
Biomaterials	9	0.84	12.9	Q1
Cancer Discovery	9	0.84	33.3	Q1

Subject area analysis of the top 20 journals identified oncology as the dominant discipline, contributing 192 publications (51.6%), followed by multidisciplinary journals (n = 71, 19.1%), cell and molecular biology (n = 54, 14.5%), gastroenterology (n = 35, 9.4%), and biomaterials and bioengineering (n = 20, 5.4%). No journals affiliated with TCM appeared among the top 20 source titles, corroborating the pattern of TCM-oriented research absence identified at the country, institutional, and author levels (Figure 6B).

Bradford’s Law of Scattering was applied to delineate the core publication outlets in this field. With 1066 publications distributed across 365 journals, the corpus was divided into three zones of approximately equal output (355 documents each). Zone 1 (core zone) comprised 19 journals, zone 2 (related zone) encompassed 71 journals, and zone 3 (peripheral zone) contained the remaining 275 journals, yielding a ratio of approximately 1:3.74:14.47, with a consistent Bradford multiplier of n = 3.8, indicating good conformity with Bradford's theoretical prediction (Figure 6C). This conformity confirms that the field has developed a stable and identifiable core publication structure, wherein a small number of specialized journals serve as the primary vehicles for knowledge dissemination – a pattern characteristic of a maturing research domain.

Analysis of keyword co-occurrence network and research theme evolution

To characterize the thematic structure of pancreatic cancer organoid research, key word co-occurrence analysis was performed using Cite Space (v7.0.0) based exclusively on author-assigned keywords (DE field). The resulting network comprised 261 nodes and 646 links (density = 0.019), with 209 nodes (80%) forming the largest connected component, indicating a well-integrated knowledge structure (Figure 7). Pancreatic cancer emerged as the most prominent node, recording the highest frequency (n = 279) and centrality (0.76) among all keywords, consistent with its role as the foundational conceptual anchor of the field. Other high-frequency terms included PDAC (n = 102, centrality = 0.21), TME (n = 47), precision medicine (n = 39), personalized medicine (n = 32), and patient-derived organoids (n = 28), collectively delineating the principal research dimensions of the corpus (Table 6). No keywords related to TCM, herbal compounds, or natural products appeared among the top 25 terms, extending the pattern of TCM research absence from the country, institutional, and author levels to the thematic level as well.

Figure 7  Keyword co-occurrence network.

Table 6 Top 25 high-frequency keywords.

Keywords	Frequency	Centrality
Pancreatic cancer	279	0.76
Pancreatic ductal adenocarcinoma	102	0.21
Tumor microenvironment	47	0.13
Precision medicine	39	0.21
Personalized medicine	32	0.2
Nude mice	30	0.39
Patient-derived organoids	28	0.04
Drug screening	19	0.16
Salmonella typhimurium a1-r	18	0.13
3D culture	15	0.19
Pancreatic ductal adenocarcinoma (PDAC)	13	0.08
Patient-derived orthotopic xenograft	12	0.07
Cancer stem cells	11	0.1
Recombinant methioninase	9	0
Drug resistance	8	0
3D cell culture	7	0.02
Extracellular matrix	7	0.05
Precision oncology	7	0.01
Methionine dependence	6	0.01
Targeted therapy	6	0.05
Combination therapy	6	0.05
Patient-derived organoid	6	0
Three-dimensional culture	5	0.16
Kras mutation	5	0
Stem cells	4	0

PDAC: Pancreatic ductal adenocarcinoma; 3D: Three-dimensional.

A total of 261 out of 1261 keywords were selected for further analysis. Cluster analysis yielded 13 distinct research clusters with excellent modularity (Q = 0.844) and silhouette scores (S = 0.9384), confirming the robustness and internal coherence of the identified thematic groupings. The most substantively relevant clusters included #0 pancreatic cancer (core disease cluster), #2 tumor microenvironment, #6 patient-derived xenograft model, #7 overcoming drug resistance, and #8 nab-paclitaxel resistance, together reflecting the dominant research directions of tumor biology, model development, and therapeutic response (Figure 8A). The timeline view further illustrated that these clusters maintained sustained research activity across the study period, with drug-resistance-related clusters exhibiting broad temporal spans, suggesting persistent and unresolved scientific attention (Figure 8B).

Figure 8 Keyword co-occurrence network analysis. A: Keyword clustering; B: Keyword timeline; C: Keyword timezone.

The time zone visualization revealed a clear sequential evolution in the thematic focus of the field. Pancreatic cancer and basic oncological descriptors constituted the foundational vocabulary of early publications (approximately 2002-2012). Patient-derived animal models – represented by nude mice and patient-derived orthotopic xenograft – dominated the intermediate period (approximately 2013-2018). The subsequent phase (2018-2022) was characterized by the emergence of patient-derived organoids, 3D culture, and personalized medicine, marking the transition toward ex vivo patient-specific modeling. Most recently, drug screening and drug resistance appeared as active and growing nodes in the rightmost temporal zone (2022-2026), indicating a shift toward translational and pharmacological applications (Figure 8C).

This trajectory was corroborated by burst detection analysis. Among the 22 key words with the strongest citation bursts, drug screening exhibited the most intense and currently active burst (strength = 4.38, 2024-2026), followed by drug resistance (strength = 3.14, 2024-2026) and PDAC (strength = 2.57, 2024-2026). The burst of patient-derived organoid during 2023-2024 (strength = 2.50) reflected the recent standardization and dissemination of organoid methodology, while the concurrent burst of drug screening directly identified pharmacological application, including evaluation of novel compounds, as the foremost current research frontier (Figure 9). The absence of any TCM-related or phytochemical-related burst keywords substantiates the conclusion that natural compound research has yet to emerge as a recognized thematic direction within this domain.

Figure 9  Top 22 keywords with the strongest citation bursts.

Co-citation analysis of references and journals

Reference co-citation analysis was conducted to map the intellectual base underlying pancreatic cancer organoid research. The resulting network comprised 435 nodes and 1490 links (density = 0.0158), with 159 nodes (36%) forming the largest connected component – a pattern reflective of the still-diversifying knowledge structure of the field (Figure 10A). Cluster analysis yielded four thematically coherent groups with exceptional quality metrics (Q = 0.859, S = 0.9747), confirming the robustness of the identified knowledge clusters. The four clusters – #1 cell-derived organoid, #2 subtype-specific difference, #3 human pancreatic cancer organoid, and #4 pancreatic cancer precision medicine – delineate a sequential intellectual trajectory: From the establishment of organoid technology, through molecular subtyping of pancreatic cancer, to precision medicine applications (Figure 10B). This progression reveals that the knowledge base of the field is structurally organized around the translational pipeline from model development to clinical implementation.

Figure 10  Co-citation analysis of references. A: Co-citation network of references; B: Co-citation cluster of references.

Tiriac et al[14] (2018, Cancer Discovery, co-citation count = 141) and Boj et al[13] (2015, Cell, n = 124) ranked as the two most frequently co-cited references, consistent with their foundational roles in establishing patient-derived pancreatic cancer organoid biobanks and demonstrating their utility for drug sensitivity profiling (Table 7)[13,14,40-57]. Seino et al[52] (2018, Cell Stem Cell, n = 83), Raghavan et al[53] (2021, Cell, n = 80), and Driehuis et al[54] (2019, PNAS, n = 80) consolidated the organoid methodological cluster. Öhlund et al[55] (2017, Journal of Experimental Medicine) recorded the highest betweenness centrality in the network (0.14) despite ranking seventh in co-citation frequency, identifying it as a critical bridge reference connecting the TME and organoid construction clusters.

Table 7 Top 20 co-citation of references based on comprehensive indicators.

Author	Year	Journal	Count of citations	Centrality
Tiriac et al[14]	2018	Cancer Discov	141	0.02
Boj et al[13]	2015	Cell	124	0.02
Seino et al[52]	2018	Cell Stem Cell	83	0.06
Raghavan et al[53]	2021	Cell	80	0.03
Driehuis et al[54]	2019	P Natl Acad SCI USA	80	0.04
Bailey et al[40]	2016	Nature	70	0.01
Öhlund et al[55]	2017	J Exp Med	61	0.14
Chan-Seng-Yue et al[41]	2020	Nat Genet	55	0.06
Halbrook et al[42]	2023	Cell	54	0
Tsai et al[43]	2018	Bmc Cancer	53	0.01
Huang et al[44]	2015	Nat Med	51	0.08
Hiroshima et al[45]	2014	Oncotarget	49	0.01
Vlachogiannis et al[56]	2018	Science	45	0.01
Cancer Genome Atlas Research Network[46]	2017	Cancer Cell	42	0
Siegel et al[47]	2021	Ca-Cancer J Clin	41	0
Murakami et al[48]	2016	Oncotarget	41	0.01
Metildi et al[49]	2014	J Surg Oncol	40	0.01
Drost et al[50]	2018	Nat Rev Cancer	40	0
Neal et al[57]	2018	Cell	40	0.01
Baker et al[51]	2016	Trends Cancer	39	0

Journal co-citation analysis identified 234 source journals forming a network of 1367 Links (density = 0.0501), with 162 journals (69%) constituting the largest connected component, reflecting a more integrated knowledge dissemination structure relative to the reference-level analysis (Figure 11A). Cluster analysis produced 12 thematic groupings (Q = 0.7795, S = 0.9237), including clusters directly relevant to organoid research (#9 patient-derived organoid, #11 pancreatic exocrine organoid) alongside tumor biology (#5 pancreatic tumorigenesis) and preclinical evaluation (#4 preclinical evaluation) clusters, collectively confirming that the knowledge base of this field spans basic oncology, model technology, and translational research (Figure 11B). Cell (n = 747) and Cancer Research (n = 735) ranked as the most frequently co-cited journals, followed by PNAS (n = 679), Nature (n = 659), and Clinical Cancer Research (n = 624) (Table 8). Science recorded the highest betweenness centrality (0.40) despite ranking ninth in citation frequency, indicating its disproportionate role as a cross-disciplinary bridge within the citation network. All 20 most frequently co-cited journals belong to oncology, molecular biology, or multidisciplinary science categories.

Figure 11  Co-citation analysis of journals. A: Co-citation network of journals; B: Co-citation cluster of journals.

Table 8 Top 20 co-citation of journals based on comprehensive indicators.

Journal	Count of citations	Centrality
Cell	747	0.08
Cancer Res	735	0.26
P Natl Acad SCI USA	679	0.05
Nature	659	0.06
Clin Cancer Res	624	0.06
Cancer Cell	546	0.11
Nat Commun	546	0.00
Cancer Discov	504	0.01
Science	487	0.40
Gastroenterology	485	0.02
Plos One	483	0.02
SCI Rep-UK	475	0.01
Nat Med	474	0.13
Nat Rev Cancer	465	0.30
New Engl J Med	404	0.04
Oncotarget	400	0.00
Cancers	380	0.05
Ca-Cancer J Clin	373	0.00
Oncogene	355	0.04
Brit J Cancer	340	0.01

No journal affiliated with TCM appeared anywhere within the co-citation network. This demonstrates that the two fields currently operate within entirely separate citation ecosystems with no cross-referencing of knowledge between them.

Bibliographic coupling analysis and identification of research fronts

To identify the current research fronts in pancreatic cancer organoid research, bibliographic coupling analysis was performed using VOSviewer (v1.6.20) on publications from 2021 to 2026, based on the principle that documents sharing a greater number of references are more likely to address related scientific questions. The resulting coupling network revealed four distinct research clusters; each representing an active thematic front within the contemporary literature (Figure 12A).

Figure 12  Document-coupled analysis. A: Document-coupled network; B: Document-coupled overlay sequences.

Three dominant research themes were identified. Cluster 1, the largest cluster comprising nine of the top 20 coupled documents, centered on the intersection of TME modeling and organoid technology. Cluster 2 encompassed six documents focused on mechanisms of drug resistance in pancreatic cancer. Cluster 3 addressed broader molecular mechanisms of pancreatic cancer. A smaller, more focused, cluster 6 was specifically representing pancreatic cancer organoid methodology. Analysis of total link strength – a more sensitive indicator of the centrality of a document within the research front than citation count alone – identified Schuth et al[58] (2022, total link strength = 643), Grossman et al[59] (2022, 614), and Suarez-Martinez et al[6] (2022, 569) as the most strongly coupled papers, suggesting that these works serve as knowledge hubs around which current research activity converges, regardless of their absolute citation rankings (Table 9)[6,58-76].

Table 9 Top 20 highly coupled documents based on comprehensive indicators.

Ref.	Journal	Title	Year	Citations	Total link strength	Cluster	Research theme
Tiwari et al[68]	Journal of Biomedical Science	Tumor Microenvironment: Barrier or Opportunity towards Effective Cancer Therapy	2022	290	427	1	Tumor microenvironment; Organoids
Chen et al[71]	Cancer Cell	Oncogenic Collagen i Homotrimers from Cancer Cells Bind to α3β1 Integrin and Impact Tumor Microbiome and Immunity to Promote Pancreatic Cancer	2022	216	188	2	Mechanisms of resistance in pancreatic cancer
Mahadevan et al[60]	Cancer Cell	KRAS G12D Inhibition Reprograms the Microenvironment of Early and Advanced Pancreatic Cancer to Promote Fas-mediated Killing by CD8+ T Cells	2023	182	432	2	Mechanisms of resistance in pancreatic cancer
Dilly et al[72]	Cancer Discovery	Mechanisms of Resistance to Oncogenic KRAS Inhibition in Pancreatic Cancer	2024	177	506	2	Mechanisms of resistance in pancreatic cancer
Mucciolo et al[62]	Cancer Cell	EGFR-activated Myofibroblasts Promote Metastasis of Pancreatic Cancer	2024	156	476	1	Tumor microenvironment; Organoids
Schuth et al[58]	Journal of Experimental & Clinical Cancer Research	Patient-specific Modeling of Stroma-mediated Chemoresistance of Pancreatic Cancer Using A Three-dimensional Organoid-fibroblast Co-culture System	2022	131	643	6	Pancreatic cancer organoids
Magré et al[73]	Journal for Immunotherapy of Cancer	Emerging Organoid-immune Co-culture Models for Cancer Research: From Oncoimmunology to Personalized Immunotherapies	2023	131	362	1	Tumor microenvironment; Organoids
Goodwin et al[74]	Cancer Research	Combination Therapies with CDK4/6 Inhibitors to Treat KRAS-mutant Pancreatic Cancer	2023	125	195	2	Mechanisms of resistance in pancreatic cancer
Proietto et al[65]	Frontiers in Oncology	Tumor Heterogeneity: Preclinical Models, Emerging Technologies, and Future Applications	2023	121	438	1	Tumor microenvironment; organoids
Haque et al[75]	Microsystems & Nanoengineering	Patient-derived Pancreatic Cancer-on-a-chip Recapitulates the Tumor Microenvironment	2022	113	260	1	Tumor microenvironment; organoids
Su et al[67]	Science Translational Medicine	Targeting SRC Reactivates Pyroptosis to Reverse Chemoresistance in Lung and Pancreatic Cancer Models	2023	113	14	1	Tumor microenvironment; organoids
Müller et al[63]	Cell Death and Differentiation	Elevated FSP1 Protects KRAS-mutated Cells from Ferroptosis during Tumor Initiation	2023	108	206	2	Mechanisms of resistance in pancreatic cancer
Schwörer et al[66]	Cancer Research	Hypoxia Potentiates the Inflammatory Fibroblast Phenotype Promoted by Pancreatic Cancer Cell-derived Cytokines	2023	107	318	1	Tumor microenvironment; organoids
Grossman et al[59]	Clinical Cancer Research	Organoid Sensitivity Correlates with Therapeutic Response in Patients with Pancreatic Cancer	2022	104	614	3	Molecular mechanisms of pancreatic cancer
Matthews et al[61]	Plos Computational Biology	Organoid: A Versatile Deep learning Platform for Tracking and Analysis of Single-organoid Dynamics	2022	102	45	1	Tumor microenvironment; organoids
Braxton et al[76]	Nature	3D Genomic Mapping Reveals Multifocality of Human Pancreatic Precancers	2024	97	111	3	Molecular mechanisms of pancreatic cancer
Suarez-Martinez et al[6]	Cell and Bioscience	3D and Organoid Culture in Research: Physiology, Hereditary Genetic Diseases and Cancer	2022	96	569	1	Tumor microenvironment; organoids
Wehrli et al[70]	Clinical Cancer Research	Mesothelin CAR T Cells Secreting Anti-fap/anti-CD3 Molecules Efficiently Target Pancreatic Adenocarcinoma and Its Stroma	2024	94	392	1	Tumor microenvironment; organoids
Nelson et al[64]	Nature Communications	USP25 Promotes Pathological HIF-1-driven Metabolic Reprogramming and Is A Potential Therapeutic Target in Pancreatic Cancer	2022	94	387	3	Molecular mechanisms of pancreatic cancer
Wang et al[69]	Cancer Research	Mitochondrial Calcium Uniporter Drives Metastasis and Confers A Targetable Cystine Dependency in Pancreatic Cancer	2022	90	109	2	Mechanisms of resistance in pancreatic cancer

The overlay visualization, encoding publication year on a blue-to-yellow gradient (2022-2025), confirmed that the identified research fronts remain actively evolving. While the majority of high-coupling nodes correspond to 2022 publications – reflecting their greater opportunity to accumulate shared citations – a distinct set of recent nodes including Loveless (2025), Ely (2025), Dilly (2024), and Braxton (2024) appear in warmer hues, indicating sustained and ongoing growth across all four thematic clusters into the most recent years of the study period (Figure 12B).

Critically, no publication addressing TCM, herbal medicine, or natural compound intervention appeared among the top coupled documents, extending the pattern of TCM research absence to the research front level. The current research frontier of pancreatic cancer organoid science is thus defined exclusively by TME modeling, drug resistance mechanisms, and molecular oncology.

DISCUSSION

It is important to clarify the epistemological boundaries of this study. The bibliometric analysis presented above generates descriptive evidence regarding what exists and does not exist within the indexed literature. We identified a structural absence of TCM across all six analytical dimensions. However, bibliometric methods cannot, by themselves, determine why this absence exists, nor can they establish whether filling this gap would necessarily lead to translational benefits. The following subsections are therefore organized into two distinct categories: (1) Evidence-based observations derived directly from the bibliometric data; and (2) Hypothesis-driven perspectives on potential future directions, which are offered as reasoned speculation rather than empirically supported conclusions.

The bibliometric landscape mapped in this study reveals a mature, rapidly expanding field of pancreatic cancer organoid but it also highlights the deficiency of TCM. This finding is not merely a bibliometric artifact; it reflects a genuine structural gap between two scientifically productive domains.

This study used bibliometric analysis as its primary methodological framework. This approach offers several well-recognized strengths. It provides an objective, quantitative overview of a research field based on the entire indexed literature, free from the selection biases that affect narrative reviews. By applying multiple complementary analytical dimensions – including co-authorship networks, keyword co-occurrence, co-citation analysis, and bibliographic coupling – it can simultaneously map the sociological structure, intellectual base, thematic evolution, and current research frontiers of a field. Crucially for this study, bibliometrics can detect not only what is present in a field but also what is structurally absent. The multidimensional, convergent absence of TCM identified here across six analytical layers provides a level of evidential rigor that anecdotal observation cannot achieve.

However, several limitations must be acknowledged. First, the analysis was restricted to the Web of Science Core Collection, which has well-known coverage gaps for Chinese-language literature and TCM-specific journals. It is possible that relevant TCM-PDAC organoid research published in Chinese-language or nonindexed journals was not captured. However, this limitation is itself informative: The absence of such work from major indexed databases reflects its absence from international scientific discourse, which is the relevant domain for global research integration. Second, bibliometric methods capture publication-level metadata but cannot assess the quality, reproducibility, or clinical relevance of individual studies. Keyword co-occurrence analysis depends on author-assigned terms, which may be inconsistent, and citation counts may reflect recency, controversy, or methodological influence rather than scientific merit. Finally, bibliometrics cannot establish causality. The absence of TCM in the PDAC organoid field does not itself explain the reasons for this absence, which likely reflect a combination of disciplinary boundaries, funding structures, and publication language barriers rather than an inherent incompatibility between the two fields. Mechanistic and experimental studies will be required to translate the directional insights provided by this bibliometric mapping into actionable scientific knowledge.

As shown in Figures 13 and 14, PDAC organoids have demonstrated strong fidelity to primary tumor biology, preserving genomic, epigenomic, and transcriptomic features of parental tumors, and have been validated as platforms for drug response prediction and precision medicine[77,78]. Within this context, natural compounds and TCM-derived molecules represent a compelling but largely untapped category of candidate agents. The most direct evidence of potential comes from berberine, an isoquinoline alkaloid with well-documented anti-inflammatory and antiproliferative properties. Okuno et al[79] demonstrated that berberine overcomes gemcitabine resistance in PDAC through modulation of the Rap1/PI3K/Akt signaling axis. Critically, these findings were validated in patient-derived PDAC organoids, confirming the utility of the organoid platform for evaluating natural compounds in a clinically relevant context. Similarly, jorunnamycin A, a bioactive compound derived from a marine sponge, demonstrated potent anti-PDAC activity across cell lines, spheroids, and patient-derived organoids harboring diverse KRAS mutations, with synergistic effects when combined with oxaliplatin[80]. These studies demonstrate that patient-derived PDAC organoids are viable, and arguably superior platforms for evaluating natural compound efficacy and identifying mechanisms of chemosensitization. Given that PDAC is driven in > 90% of cases by KRAS mutations and is defined by a dense[81], immunosuppressive TME[82,83], the multitarget pharmacology of TCM formulations is particularly well matched to the biological complexity of the disease[84]. Natural compound screening in PDAC organoids therefore represents a high-priority and under-developed research direction.

Figure 13  Traditional Chinese medicine-pancreatic ductal adenocarcinoma organoid integration framework. PDO: Patient-derived organoid; EUS-FNA/FNB: Endoscopic ultrasound-fine-needle aspiration/fine-needle biopsy; TME: Tumor microenvironment; 3D: Three-dimensional; CAFs: Cancer-associated fibroblasts; NK: Natural killer; TCM: Traditional Chinese medicine; PDAC: Pancreatic ductal adenocarcinoma.

Figure 14  Summary diagram of the target sites of traditional Chinese medicine compounds in pancreatic ductal adenocarcinoma-related signaling pathways. PDAC: Pancreatic ductal adenocarcinoma; TME: Tumor microenvironment; VEGF: Vascular endothelial growth factor; TGF-β: Transforming growth factor-beta; 3D: Three-dimensional; BRAF: B-Raf proto-oncogene, serine/threonine kinase; MEK: Mitogen-activated protein kinase kinase; ERK: Extracellular regulated protein kinases; PI3K: Phosphatidylinositol 3-kinase; Akt: Protein kinase B; NF-κB: Nuclear factor kappa B; 2D: Two-dimensional; EMT: Epithelial-mesenchymal transition; STAT3: Signal transducer and activator of transcription 3; PDOs: Patient-derived organoids; TCMs: Traditional Chinese medicine.

The integration of TCM with PDAC organoid technology is not merely technically feasible; it carries distinct theoretical advantages grounded in TCM philosophy and the biological characteristics of PDAC. TCM is characterized by its multicomponent, multitarget, and multipathway mode of action, which stands in contrast to the single-target approach of conventional western pharmacology[85,86]. This systemic pharmacology is uniquely suited to cancers like PDAC, where therapeutic failure is driven by the simultaneous activation of redundant oncogenic pathways, stromal resistance, and immune evasion[81-83]. Organoid platforms, with their capacity to model tumor heterogeneity and microenvironmental complexity in three dimensions, are therefore better-suited than 2D monocultures to capture the pleiotropic effects of TCM interventions[14,54].

The TCM concept of syndrome differentiation and treatment (Bian Zheng Lun Zhi) – individualizing therapy according to patient-specific patterns of disease – aligns conceptually with the precision medicine paradigm that organoid technology is designed to enable[87-89]. Patient-derived organoids preserve interpatient heterogeneity and can serve as personalized avatars for predicting individual drug responses[54,56,90]. This creates an unprecedented opportunity to empirically test syndrome-based TCM prescriptions at the individual patient level, providing mechanistic and pharmacological evidence for a therapeutic philosophy that has historically resisted western validation[88].

TCM herbal compounds frequently exhibit low toxicity profiles and favorable safety records accumulated through centuries of clinical use. This is clinically significant in PDAC, where conventional first-line regimens (FOLFIRINOX, gemcitabine/nab-paclitaxel) are associated with severe systemic toxicity[91,92]. TCM-derived compounds used as sensitizers or adjuvants – as demonstrated by berberine in combination with gemcitabine – may offer pathways to improved therapeutic ratios[79]. In the organoid context, triptolide and honokiol, both derived from TCM sources, were notably found to lack the multicellular drug resistance observed with doxorubicin in liver organoid-like tumor tissue, suggesting that they circumvent 3D-architecture-mediated resistance mechanisms[93]. These properties make TCM compounds particularly attractive for organoid-based combinatorial screening.

While TCM-organoid integration in PDAC remains undeveloped, several fields have already established proof-of-concept, providing a roadmap for future work. In gastrointestinal inflammation, the classic TCM formula Gegen Qinlian decoction was evaluated in intestinal organoids subjected to RSL3-induced ferroptosis. Gegen Qinlian decoction reversed organoid morphological damage, restored epithelial barrier integrity, and modulated ferroptosis-associated proteins (glutathione peroxidase 4, SLC7A11, and ACSL4), recapitulating its in vivo therapeutic effects in DSS-induced colitis[94]. This study demonstrates that organoids can serve as mechanistic interrogation platforms for complex TCM formulations, bridging the gap between empirical clinical efficacy and molecular mechanism.

In neuroscience, a cortical organoid-on-a-chip platform with physiological hypoxia was used to investigate the neuroactive effects of tanshinone IIA, a TCM compound derived from Salvia miltiorrhiza. Tanshinone IIA was found to promote neural differentiation and maturation, suggesting therapeutic potential for neurodevelopmental disorders; a finding only possible in a system replicating the hypoxic prenatal microenvironment[95]. In rheumatoid arthritis, synovial organoids have been proposed as ideal platforms for testing TCM-based interventions that modulate synovial hyperplasia and inflammatory signaling, given their capacity to recapitulate the pathological microenvironment of the joint[96]. For coronavirus disease 2019, organoid and organ-on-a-chip models have been specifically reviewed as platforms to elucidate the mechanisms of TCM intervention and to enable rational screening of active TCM components; a framework equally applicable to PDAC[97]. In respiratory disease, lung organoids in a 96-well format were used to screen natural compounds including quercetin for senolytic activity, demonstrating that organoid platforms can support high-throughput TCM compound evaluation in chronic disease contexts[98].

Across these disease models, a convergent conclusion emerges: Organoid platforms resolve the key limitations of 2D cell lines in TCM research – particularly the absence of cell heterogeneity, 3D architecture, and physiologically relevant cell-cell and cell-matrix interactions[6] – while enabling mechanistic dissection of multitarget TCM activity at the molecular level[86,99,100].

Despite clear conceptual and preliminary empirical support, significant challenges must be addressed before TCM-PDAC organoid research can mature into a clinically impactful discipline[101]. The most fundamental challenge is the complexity of TCM formulations. Multiherb decoctions contain hundreds of bioactive components with overlapping, synergistic, and potentially antagonistic interactions. Disentangling which components are pharmacologically active and which targets they engage – in a 3D organoid system that itself possesses multilayered biological complexity – requires integrating network pharmacology, multiomics profiling, and advanced computational modeling[85,102].

A second major challenge is the poor physicochemical properties of many natural compounds. The experience with curcumin is illustrative: Despite potent programmed death-ligand 1 downregulation in monolayer culture, its limited penetration into dense 3D tumor architecture significantly compromises its efficacy in organoid and spheroid models[103]. The extracellular matrix of PDAC organoids presents particularly formidable physical barriers, including dense fibronectin networks, pH gradients, and extreme cell density[83,104]. Nanoparticle-based and liposomal delivery systems may improve compound penetration, but formulation optimization must be carefully balanced against potential interference with biological activity[103]. Encouragingly, hollow mesoporous silica nanoparticles loaded with natural product extracts (e.g., Mangifera indica seed kernel extract combined with doxorubicin) have demonstrated enhanced efficacy in 3D hepatocellular carcinoma tumor models, suggesting a viable engineering strategy[105].

Third, the standardization of PDAC organoid culture conditions – including extracellular matrix composition, growth factor cocktails, and passage protocols – remains inconsistent across laboratories[106]. This reproducibility challenge is compounded in TCM research by batch-to-batch variation in herbal preparations, making interstudy comparisons difficult[107]. Regulatory and quality standardization frameworks for both organoid models and TCM preparations are essential prerequisites for clinical translation[108].

Finally, while natural products have been tested for direct cytotoxicity in organoids, the immunomodulatory dimension of TCM action remains entirely unexplored in organoid systems. Hong and Lyu[109] identified that no published studies have investigated natural-product-mediated immunomodulation in organoid-immune coculture platforms; a critical gap given that immune evasion is a central driver of PDAC progression. Tumor-immune organoid coculture systems incorporating autologous immune cells (T cells, natural killer cells, and macrophages) represent the next-generation platform required to evaluate the immunopharmacological dimension of TCM in PDAC[57,107-111]. Arenobufagin, for example, was shown to modulate heat shock protein 90β/STAT3/programmed death-ligand 1 signaling and enhance T cell enrichment in patient-derived colorectal cancer organoids cocultured with immune cells[112], demonstrating the feasibility and power of this approach and its direct relevance to PDAC.

Looking forward, the convergence of patient-derived PDAC organoid technology, network-pharmacology-guided TCM target identification, AI-assisted compound screening, and multiomics mechanistic profiling creates an enabling infrastructure for systematic TCM–PDAC organoid research. Disease-derived organoid models that preserve pathological hallmarks – as demonstrated for fibrotic liver disease using CCl₄-injured tissue-derived organoids that successfully identified Sargassum japonica as an antifibrotic candidate[113] – provide a template for constructing PDAC-specific organoid platforms that model stroma-mediated resistance. Pilot studies applying this integrated pipeline to well-characterized TCM compounds (berberine, triptolide, ginsenoside Rg3, and Astragalus polysaccharide) against PDAC organoids with defined molecular subtypes would constitute a scientifically high-value and technically feasible near-term research agenda.

CONCLUSION

Through systematic bibliometric analysis of 1066 publications spanning 2001 to 2026, this study mapped the global research landscape of pancreatic cancer organoids. The field has undergone rapid and sustained expansion, with the United States, China, and Germany emerging as the leading contributors. Its intellectual base is well consolidated around organoid technology development, TME biology, and molecular subtyping, while drug screening and drug resistance have emerged as the most intensely active current research frontiers. However, our analysis revealed a deficiency of TCM across all six analytical dimensions examined: (1) Country collaboration networks; (2) Institutional co-authorship; (3) Author affiliations; (4) Source journals; (5) Keyword co-occurrence; and (6) Citation ecosystems. This multidimensional, convergent absence indicates not merely a quantitative gap in publication output, but a low relevance between two fields. This situation reveals that research into the modernization of TCM is lagging behind in the field of model innovation, while also highlighting that pancreatic cancer organoid models hold considerable potential for screening the active components of TCM and evaluating the bioequivalence of compound formulations. Consequently, the technical characteristics of pancreatic cancer organoids, their compatibility with TCM theory, and existing applications warrant further investigation, with a view to promoting the in-depth development of this interdisciplinary field.