Research progress of plant-derived chemical compounds for overcoming pancreatic cancer drug resistance.

INTRODUCTION
Pancreatic cancer is a highly lethal malignancy that has become the seventh leading cause of cancer-related deaths globally. It is expected that by 2030, pancreatic cancer will become the second leading cause of cancer-related deaths, only behind lung cancer[1]. North America, Europe, and Australia are regions with a high incidence of pancreatic cancer. Over the past 20 years, the number of diagnoses of pancreatic cancer has increased from 196000 cases in 1990 to 441000 cases in 2017, representing a twofold increase[2], because of the insidious nature of pancreatic cancer, more than 80% of diagnoses are made at an advanced stage, resulting in a five-year survival rate of only 11%[3]. It is widely recognized that smoking, alcohol abuse, diabetes, obesity, aging, family history, and genetics are high-risk factors influencing the incidence of pancreatic cancer[4]. In countries with a higher development index and an aging population (where the elderly population aged 65 and above exceeds 7%), there are generally more accurate medical examination tools, leading to a higher absolute number of cases of pancreatic cancer. In contrast, developing countries, especially in Africa, face challenges in introducing advanced imaging and accurate pathological diagnostic methods. This makes high-quality diagnosis of pancreatic cancer a luxury, and as a result, some global data on pancreatic cancer cases may be missing[5]. Although the absolute number of cases is high, patients in developed regions with well-established treatment systems tend to have better overall survival (OS) than patients in economically underdeveloped regions. Currently, the mainstream treatment involves R0 resection surgery combined with postoperative chemotherapy, with gemcitabine + nab-paclitaxel (AG) and fluorouracil, leucovorin, irinotecan, and oxaliplatin (FOLFIRINOX) regimens being the primary chemotherapy options. However, due to the highly resistant nature of pancreatic cancer, the benefits of chemotherapy regimens are suboptimal. As a result, scientists have made significant efforts to identify various resistance mechanisms in pancreatic cancer. On the basis of these mechanisms, new therapeutic approaches, such as neoadjuvant therapy, targeted immunotherapy, and cancer vaccines, have been explored. Many scientists believe that plant-derived compounds offer advantages, including the reduction of anticancer drug toxicity, having fewer side effects, and being cost-effective, making them a promising addition to cancer treatment regimens. Numerous studies have confirmed that medicinal plants and their active compounds help control diseases and inhibit carcinogenesis. Furthermore, active compounds from natural products or herbs have shown substantial effects in enhancing the efficacy of anticancer drugs while reducing their toxicity. Novel drug combinations consisting of traditional chemotherapy regimens and natural products are expected to achieve better outcomes than conventional chemotherapy drugs, while reducing adverse reactions[6].

CURRENT TREATMENTS FOR PANCREATIC CANCER AND THEIR LIMITATIONS
The mainstream treatment regimen for pancreatic cancer remains surgery followed by postoperative chemotherapy. Due to the insidious nature of pancreatic cancer, only a small proportion of patients are eligible for surgical treatment. The median survival time is approximately 10-12 months for patients who undergo diagnostic evaluation and receive treatment, whereas untreated patients typically survive for only 5-6 months. The standard first-line treatment regimens are mFOLFIRINOX [a combination of 5-fluorouracil (5-FU), leucovorin, irinotecan, and oxaliplatin] or AG[7-10]. No other combination has demonstrated both superior survival benefits and reduced treatment toxicity compared with these two regimens. A recent study demonstrated that the modified version of the mFOLFIRINOX regimen, when used as adjuvant therapy, yields significantly longer survival compared to gemcitabine alone. The median disease-free survival in the modified mFOLFIRINOX group was significantly higher than in the gemcitabine group (21.6 months vs 12.8 months). However, the modified mFOLFIRINOX regimen is associated with greater toxicity, despite improvements compared to the traditional FOLFIRINOX regimen[11]. As a result, gemcitabine remains the most commonly used chemotherapy drug for pancreatic cancer in clinical practice. Unfortunately, approximately 80% of patients relapse and die within a year[12].
The disappointing outcomes are primarily attributed to factors such as inadequate lymph node dissection during surgery, undetected micro-metastases, chemotherapy drug resistance, and poor efficacy of postoperative adjuvant therapies, which result in suboptimal clinical outcomes. Given the low rate of complete R0 resection during surgery and the high likelihood of recurrence within a year after surgery, chemotherapy is generally considered the mainstream treatment for pancreatic cancer. Drug resistance remains a major factor influencing the effectiveness of chemotherapy regimens, with gemcitabine, the first-line chemotherapy drug for pancreatic cancer, being particularly affected. The specific mechanisms of resistance to gemcitabine are still a subject of ongoing debate.

CHEMORESISTANCE IN PANCREATIC CANCER
This review summarizes recent advances in understanding the mechanisms of drug resistance, including the following aspects: Drug transport disorders, key intracellular enzymes, epithelial-mesenchymal transition (EMT), the tumor microenvironment (TME), abnormal expression of resistance proteins, and the regulation of potential target genes and signaling pathways by non-coding RNAs (ncRNAs). Based on literature review, it is undeniable that pancreatic cancer cells exhibit higher resistance to gemcitabine compared to other chemotherapy drugs.

Drug transport disorders
Chemotherapy drugs require nucleoside transporters (NTs) to pass through the plasma membrane and exert their effects[13]. Some studies have demonstrated that reducing human equilibrative nucleoside transporter 1 (hENT1) expression results in decreased cellular uptake of gemcitabine[14-18]. Patients with low expression of NTs have significantly shorter median survival times[19-24], it is inferred that a lack of hENT1 can lead to significant gemcitabine resistance[18]. Bhutia et al[25] showed that pharmacological inhibition of human concentrative nucleoside transporter 1 (hCNT1) degradation modestly increased the surface expression of hCNT1 in MIA PaCa-2 cells, and the cellular transport of gemcitabine. Constitutive expression of hCNT1 reduced the clonogenic survival of MIA PaCa-2 cells and significantly enhanced gemcitabine transport and chemoresistance[25].

The role of intracellular enzymes
Once chemotherapy drugs enter the cell via NTs, two nucleotidases play essential roles in deoxycytidine kinase (dCK) and ribonucleotide reductase (RR). dCK is the primary rate-limiting enzyme for the metabolism and activation of chemotherapy drugs within the cell[26]. Knockdown of dCK can induce drug resistance in pancreatic cancer[27]. Moreover, overexpression of dCK can restore drug sensitivity in resistant strains[28,29]. RR, another rate-limiting enzyme in the DNA synthesis pathway, primarily converts ribonucleotides into deoxyribonucleotide triphosphates, which are crucial for DNA replication and repair. RR consists of two subunits, M1 and M2[30]. The inhibition of RR by gemcitabine diphosphate (dFdCDP) activates gemcitabine[31,32]. Recent meta-analyses have shown that patients with low expression of RR subunit 1 (RRM1) have better OS and disease-free survival than those with high RRM1 expression[33-36]. This suggests that RRM1 expression is an indicator of poor survival in pancreatic cancer patients undergoing chemotherapy. Wang et al[37] and Nakahira et al[38] have experimentally demonstrated that overexpression of RRM1 and RRM2 proteins in pancreatic cancer cell lines can induce stable genetic resistance. Hence, both dCK and RR are key factors influencing drug resistance.

EMT
EMT is a phase of tumor cell phenotype transformation that facilitates the acquisition of a more invasive mesenchymal phenotype, often accompanied by morphological changes in cancer cells and alterations in genomic and proteomic levels[39]. The role of mesenchymal transcription factors in chemotherapy resistance has garnered increasing attention in recent years[40]. EMT is mediated by various key genes and signaling pathways. E-cadherin is strongly negatively correlated with zinc-finger E-box binding homeobox 1 (Zeb-1) and is closely related to resistance. Zeb-1 and other EMT regulators are implicated in maintaining drug resistance in human pancreatic cancer cells[41]. Silencing Zeb-1 upregulates E-cadherin expression and restores drug sensitivity by increasing the expression of epithelial markers such as epithelial-V-like antigen 1 and myelin and lymphocyte protein 2[42]. Signaling pathways, such as Notch, tumor necrosis factor α, transforming growth factor β, and hypoxia-inducible factor-1α (HIF-1α) pathways, are involved in EMT induction in pancreatic cancer cells[43]. Silencing Zeb-1 reverses EMT, restores the expression of epithelial marker genes, and enhances sensitivity to therapeutic agents, suggesting that Zeb-1 and other EMT regulators are involved in maintaining drug resistance in human pancreatic cancer cells.

TME
The TME encompasses blood vessels, immune cells, fibroblasts, signaling molecules, and the extracellular matrix[44]. Tumors can influence the microenvironment by releasing extracellular signals, promoting tumor angiogenesis, and inducing peripheral immune tolerance[45]. The microenvironment, in turn, can affect the efficacy of gemcitabine in treating cancer cells. The pancreatic cancer TME comprises pancreatic stellate cells (PSCs), fibroblasts, inflammatory cells, endothelial cells, and nerve cells[46,47]. These components contribute to drug resistance in pancreatic cancer[48-50]. Co-culture experiments confirmed that PSCs can enhance chemotherapy resistance to gemcitabine through the Notch signaling pathway by increasing Hes1 expression[51]. Previous studies have demonstrated that Notch signaling inhibitors or silencing Hes1 expression can reverse chemoresistance induced by PSCs[49]. Liu et al[50] discovered that periostin is overexpressed only in the stroma of pancreatic ductal adenocarcinoma (PDAC) and PSCs, suggesting its role in conferring resistance to gemcitabine. Zhang et al[52] demonstrated that PSCs activate the focal adhesion kinase-protein kinase B (AKT) and extracellular signal-regulated kinase 1/2 pathways and secrete interleukin-6 through the stromal cell-derived factor-1α/C-X-C receptor 4 axis, thereby promoting gemcitabine resistance. Emerging evidence indicates that aberrant activation of Hedgehog signaling is associated with tumor malignancy and drug resistance[53-55]. In vitro experiments have demonstrated that Hedgehog signaling promotes cancer progression by regulating cancer cell proliferation, malignancy, metastasis, and the expansion of cancer stem cells (CSCs)[56-59]. Hedgehog activation increases the stroma in the pancreatic cancer microenvironment, and its inhibition can reduce α-smooth muscle actin levels. Reduction of the stroma component can increase mean vascular density, leading to improved blood perfusion and drug delivery[60]. Blocking Hedgehog signaling with cyclopamine has been shown to reduce PDAC proliferation and reverse drug resistance[61]. Further exploration in mouse models suggests that inhibition of Hedgehog signaling leads to transient increases in tumor vasculature and gemcitabine concentration, resulting in temporary disease stabilization[60]. Resveratrol reprograms tumor-associated macrophages by inhibiting the signal transducer and activator of transcription 3 signaling pathway, thereby reducing the release of pro-tumorigenic factors and enhancing the penetration of chemotherapeutic agents[62]. Quercetin has been shown to promote the migration of dendritic cells to lymph nodes via the Toll-like receptor 4/receptor for advanced glycation end-products (RAGE) pathway, which subsequently initiates T cell responses and enhances the efficacy of chemotherapeutic drugs[63]. A considerable number of phytochemicals have a significant effect on remodeling the TME, which should not be underestimated.

Abnormal expression of resistance proteins

Abnormal expression of transport proteins: Members of the ATP-binding cassette (ABC) transporter superfamily catalyze the translocation of substrates across cellular membranes and are widely expressed in all human tissues[64,65]. Multiple transporters are involved in resistance and disease pathways related to transporter dysfunction. Among them, ABCB1 is a multi-specific multidrug efflux pump involved in cellular detoxification, and cancer cells can exploit it to develop multidrug resistance (MDR) in response to chemotherapy. ABCG1 transports cholesterol and sphingolipids, playing a role in the reversal of cholesterol transport from peripheral tissues[66,67]. ABCG2 is a highly physiologically important multidrug efflux pump. Together with ABCB1, ABCG2 forms a pair of drug-resistant transmembrane transporters that act as gatekeepers in the blood-brain barrier. ABCG2 also transports various drugs and influence their pharmacokinetics. It plays a role in the transport of sulfated steroids and urate, and is reported to cause MDR in various cancer cells[68-70].

High expression of P-glycoprotein: P-glycoprotein (P-gp), encoded by the MDR1 gene, is a 1280-amino acid single-chain protein that acts as an efflux transporter, influencing the absorption, distribution, and elimination of various chemotherapy compounds. The exact mechanism of P-gp resistance remains unclear. It is generally believed that P-gp utilizes ATP hydrolysis to actively efflux drugs from cells, thereby reducing the intracellular drug concentration below the effective killing threshold. Some researchers have reported that P-gp expression is significantly higher in pancreatic cancer cells than in normal pancreatic tissue, which may contribute to the low efficacy of chemotherapy and the development of resistance[71].

Elevated expression of lung resistance-related protein: Lung resistance-related protein (LRP), initially known as LRP, was first identified by Scheper in 1993 in a non-small cell lung cancer cell line (SW-1573) that lacked P-gp expression. The LRP gene is located on chromosome 16, near the genes encoding MDR-associated proteins and protein kinase C-beta, and may mediate resistance through a transport mechanism[72]. LRP exhibits tissue-specific expression patterns and is widely detected in normal human tissues and tumor cells. The exact mechanism by which LRP contributes to resistance remains unclear. Some researchers believe that LRP may cause MDR by: (1) Preventing drugs from entering the cell nucleus through nuclear pores, or if drugs do enter, pumping them out before they can exert their effect; and (2) Facilitating the entry of drugs into vesicles in the cytoplasm, from where they are eventually excreted by exocytosis[72,73]. Other researchers suggest that LRP may not directly contribute to tumor MDR but works synergistically with multidrug resistance-associated protein in mediating tumor resistance[74]. Therefore, we believe that the increased expression of LRP may be associated with resistance to pancreatic cancer.

Low expression of topoisomerase II: Topoisomerases (Topo) are enzymes that correct DNA supercoiling by cleaving and rejoining the phosphodiester bonds in one or both strands of DNA. Topo II acts on both strands of DNA and usually requires ATP as an energy cofactor. Recent studies have shown that changes in Topo II activity and function are involved in mediating MDR or alternative mechanisms of resistance[75]. Research has shown that the expression of Topo II in pancreatic cancer cells decreases, which may be linked to the development of drug resistance[76].

Upregulated expression of glutathione S-transferases: Glutathione is the most important low-molecular-weight antioxidant synthesized in cells. Glutathione S-transferases (GSTs) catalyze the initial step in the glutathione conjugation reaction and are key enzymes in this process, existing in various forms and primarily localized in the cytoplasm. GSTs are overexpressed in multiple types of tumors, including pancreatic, ovarian, non-small cell lung, breast, colon, and lymphoma[77]. Studies have shown that GSTs protect cells from damage caused by exposure to both endogenous and exogenous electrophilic substances, exhibiting mutagenic and antitumor properties. They are highly expressed in tumor tissues and are associated with tumor resistance. In-depth research on GSTs lays the foundation for their application in cell protection and antitumor therapy in Table 1[64-92].

ncRNAs
ncRNAs are a class of RNAs that lack protein-coding potential and were once considered non-essential components of the genome. However, increasingly diverse ncRNAs, including microRNAs (miRNAs), long ncRNAs (lncRNAs), and circular RNAs (circRNAs)[93], regulate tumor progression through chromatin remodeling, interactions with cytoplasmic mRNA, and the formation of competitive endogenous RNA networks[94,95]. In addition to their roles in cell proliferation, invasion, migration, and apoptosis, ncRNAs also coordinate metabolic reprogramming, such as glycolysis, mitochondrial function, glutamine, and lipid metabolism[94], thereby driving drug resistance in various types of cancer, including prostate cancer and breast cancer[96,97]. Therefore, ncRNAs have increasingly become recognized as important therapeutic targets and potential biomarkers in cancer treatment[93,97]. A growing body of evidence suggests that ncRNAs play a key role in regulating the chemosensitivity of pancreatic cancer to gemcitabine. Different ncRNAs can affect drug uptake, tumor cell cycle, apoptosis, and EMT by regulating the expression of potential target genes and related signaling pathways, thereby participating in pancreatic cancer resistance[98]. The following section summarizes relevant studies conducted over the past five years.

miRNA and chemoresistance in pancreatic cancer: Recent studies have shown that miRNA families play a significant role in the chemotherapy resistance process of pancreatic cancer. Overexpression of miR-210 enhances drug resistance by antagonizing gemcitabine-induced cell cycle arrest[99]. Elevated expression of miR-30a-3p suppresses intercellular communication by downregulating connexin 43, thereby hindering the intercellular transfer of cytotoxic agents such as gemcitabine through gap junctions and ultimately promoting drug resistance[100]. miR-146a-5p targets the 3’-untranslated region of tumor necrosis factor receptor-associated factor 6 and also downregulates the tumor necrosis factor receptor-associated factor 6/nuclear factor kappa B (NF-κB)-p65/P-gp axis, which collectively suppresses pancreatic cancer cell proliferation and promotes resistance[101]. Functional studies have shown that gemcitabine causes pancreatic cancer cells to downregulate miR-21 expression and upregulate FasL expression. After gemcitabine treatment, the increase in FasL expression induces apoptosis, while ectopic expression of miR-21 partially protects cancer cells from gemcitabine-induced apoptosis, confirming that FasL is a direct target of miR-21. Therefore, elevated serum levels of miR-21 have been proposed as a potential predictor of chemotherapy sensitivity in advanced pancreatic cancer[102]. Inhibition of miR-221/222 expression has been shown to arrest the cell cycle, promote apoptosis, and preserve gemcitabine sensitivity during treatment[103]. miR-126 targets vascular endothelial growth factor-A, influencing tumor angiogenesis and progression. Moreover, miR-126 plays a crucial role in cancer cell invasion and metastasis by regulating genes involved in cell adhesion and migration. In addition, dysregulation of miR-126 contributes to chemoresistance, affecting cancer cell survival and therapeutic response[104].
Inhibition of miR-497 significantly accelerates gemcitabine resistance, migration, and invasion in pancreatic CSCs. Overexpression of NFκB1 has been shown to promote gemcitabine-treated pancreatic CSC viability, migration, and invasion. However, miR-497 overexpression partially suppresses this process[105]. Ye et al[106] found that miR-7, by targeting polyadenosine-diphosphate-ribose polymerase 1/NF-κB signaling, acts as an important regulator of cellular senescence. miR-7 expression is decreased in gemcitabine-resistant cell lines, and its restoration makes pancreatic cancer cells sensitive to gemcitabine, reversing resistance and cellular senescence. miR-136-5p negatively regulates ZNF32, thereby reducing the proliferation and migration abilities of pancreatic cancer cells (PANC-1 and CFPAC-1), and reversing gemcitabine resistance[107]. The tumor suppressor miRNA-3662 inhibits chemotherapy resistance and aerobic glycolysis in PDAC cells via a negative feedback loop with HIF-1α[108].
In addition to gemcitabine, miRNAs have been reported to reverse resistance to other chemotherapy agent used in pancreatic cancer treatment. miR-499a-5p targets phosphatase and tensin homolog and activates the phosphatidylinositol 3-kinase (PI3K)/Akt pathway, promoting proliferation, migration, and 5-FU resistance in pancreatic cancer cells. Moreover, miR-499a-5p has been shown to affect the expression of MDR-related genes, including ABC subfamily B member 1 (P-gp), ABC subfamily C member 1 (MRP1), and ABC subfamily G member 2 (BCRP)[109]. The tumor suppressor miR-1291-5p enhances pancreatic cancer cell sensitivity to arginine deprivation (PEGylated arginase) and chemotherapy (cisplatin) by targeting argininosuccinate synthase 1 and glucose transporter 1, which regulate protein degradation and glycolysis[110]. Meijer et al[111] showed that pancreatic cancer patients who respond well to FOLFIRINOX exhibit lower levels of miR-181a-5p in both cancer tissue and plasma samples. They proposed that plasma miR-181a-5p could be a specific biomarker for monitoring FOLFIRINOX response, and that decreases in both plasma miR-181a-5p and carbohydrate antigen 19-9 levels are associated with better prognosis after FOLFIRINOX treatment[111]. A research team investigating the sensitizing effects of miRNA on FOLFIRINOX demonstrated, through high-throughput validation, that compared to control PDAC cells, miR1307 knockout (miR1307KO) cells accumulated more chemotherapy-induced cell death and DNA damage. In contrast, re-expression of miR1307 in miR1307KO cells rescued these effects[112] (Table 2).

Regulation of chemoresistance by lncRNA family members in pancreatic cancer: Studies have shown that certain members of the lncRNA family can regulate the process of chemotherapy resistance in pancreatic cancer. For example, lncRNA SLC7A11-AS1 is highly expressed in gemcitabine-resistant pancreatic cancer cell lines. It stabilizes nuclear factor erythroid-2-related factor 2 (Nrf2), a key regulator of antioxidant defense, through interaction with β-transducing repeat-containing protein 1. Nrf2, in turn, promotes chemoresistance by reducing the levels of reactive oxygen species in pancreatic cancer cells[113].
Similarly, lncRNA HOXA transcript at the distal tip (HOTTIP) is significantly upregulated in pancreatic cancer tissues and cell lines compared with non-cancerous pancreatic tissue and the non-tumor pancreatic cell line HPDE6. Functional inhibition of HOTTIP induces cell cycle arrest and suppresses cell invasion by suppressing EMT. Moreover, HOTTIP silencing enhances the antitumor effects of gemcitabine both in vitro and in vivo. These functions were partially mediated through the activation of HOXA13[114].
Plasmacytoma variant translocation 1 (PVT1), first discovered in 1986, has been shown to possess oncogenic potential in various malignancies[115]. However, studies on the PVT1 gene in pancreatic cancer only emerged recently[116], including research on the negative regulation of gemcitabine sensitivity in pancreatic cancer cells by PVT1. A susceptibility allele, rs1561927, was identified in PVT1[117]. Studies using quantitative reverse-transcription polymerase chain reaction to measure PVT1 expression in pancreatic tissues have revealed that PVT1 is expressed at significantly higher levels in pancreatic cancer tissues compared to non-cancerous tissues, and its expression is positively correlated with poor patient survival[118] (Table 3).

Role of circRNAs in chemoresistance in pancreatic cancer: Research on circRNAs in pancreatic cancer chemoresistance is relatively limited. Liu et al[119] found that circHIPK3 expression was elevated in both gemcitabine-resistant pancreatic cancer tumor tissues and gemcitabine-resistant pancreatic cancer cells. CircHIPK3 promotes gemcitabine resistance in pancreatic cancer cells by regulating RASSF1 through miR-330-5p targeting. Exosomal circZNF91, induced by hypoxia, was transferred to normoxic pancreatic cancer cells, where it competitively binds with miR-23b-3p, thus preventing miR-23b-3p from inhibiting the expression of the deacetylase sirtuin1. Upregulation of sirtuin1 enhances the deacetylation-dependent stability of HIF-1α protein, promoting glycolysis and conferring gemcitabine resistance-a finding further validated in mouse models[120]. Hong et al[121] also discovered that circZNF91 promotes pancreatic cancer growth and metastasis by sequestering miR-519 and targeting secreted modular calcium-binding protein 2, which increases the resistance of pancreatic cancer cells to gemcitabine. In another study, circ-MTHFD1 L was shown to promote DNA damage repair by upregulating the expression of RPN6, thereby enhancing gemcitabine resistance in pancreatic cancer cells. Remarkably, inhibition of circ-MTHFD1 L in combination with olaparib was able to reverse gemcitabine resistance in pancreatic cancer[122]. CircFARP1 plays a crucial role in CAF-induced gemcitabine resistance through activation of the caveolin 1/miR-660-3p/Leukemia inhibitory factor axis. Suppression of circFARP1 significantly inhibited PDAC growth and gemcitabine resistance in patient-derived xenograft (PDX) models[123]. CircLMTK2, an isoform of circZNF91, is overexpressed in gemcitabine-resistant pancreatic cancer cells and tissues. Knockdown of circLMTK2 promoted miR-485-5p sponge-mediated regulation of p21-regulated kinase 1, reducing gemcitabine resistance in pancreatic cancer cells[124] (Table 4 and Figure 1).

EXPERIMENTAL MODELS OF DRUG RESISTANCE IN PANCREATIC CANCER
Here, we summarize the major experimental models used to investigate drug resistance in pancreatic cancer, based on extensive research and previous literature. These models can be broadly divided into four types: In vitro, in vivo, ex vivo, and specialized mechanistic models. Below, we provide a brief overview of these models, comparing their advantages, limitations, and applicability (Figure 2).

In vitro models

Two-dimensional cell line models: The classic experimental model for studying drug resistance. Examples of gemcitabine-resistant cell lines include PANC-1/gemcitabine[119], MIA PaCa-2/gemcitabine[125], and multi-drug-resistant cell lines such as Capan-1/P-gp (overexpressing P-gp)[126]. Researchers can use this model to select suitable cell lines based on the drugs identified in their experiments or conduct high-throughput screening of drugs in the pre-experiment phase. This is a bidirectional screening approach where methods such as CCK-8 or MTT assays can be used to observe changes in IC50 values. Researchers also explore the mechanisms of reversing drug resistance. The significant advantage of this model is its ability to study drug efficacy and mechanisms; however, it cannot replicate the TME, which is a notable limitation.

Three-dimensional organoid models: Three-dimensional (3D) organoid models are classified into patient-derived organoids (PDOs) and cell line-derived organoids. These models have gained considerable attention because they aim to retain tumor heterogeneity and stem cell properties, making them more representative of in vivo drug resistance phenotypes. PDO-based 3D organoid models are particularly useful for screening patient-specific drug combinations and predicting therapeutic responses[127]. Derived directly from patients’ tumor tissues, PDOs faithfully reproduce patient-specific molecular and phenotypic characteristics, enabling individualized testing of drug combinations and prediction of clinical therapeutic responses[128]. For instance, Tiriac et al[129] established a living biobank of pancreatic cancer organoids and demonstrated that their in-vitro drug-sensitivity profiles correlated strongly with patient outcomes in vivo, underscoring their potential as predictive diagnostic tools. Similarly, Huang et al[130] reported that organoid-based chemosensitivity assays successfully guided gemcitabine-based combination therapy, achieving improved clinical responses in PDAC patients. Moreover, recent advances in co-culture organoid models, which incorporate PSCs, immune components, or endothelial cells, have allowed researchers to more accurately recapitulate tumor-stroma interactions that drive chemoresistance[131]. Such integrated 3D systems are increasingly used to explore novel treatment strategies aimed at overcoming resistance and optimizing individualized therapy[132]. Collectively, 3D organoid models serve as a vital bridge between preclinical research and clinical application, offering a robust and adaptable platform for precision medicine in pancreatic cancer.

Co-culture systems models: Co-culture models simulate the TME by co-culturing selected pancreatic cancer cell lines with PSCs[125] or macrophages (tumor-associated macrophages)[133]. This model is used to study matrix-mediated drug resistance mechanisms, such as how PSCs secrete interleukin-6 to induce gemcitabine resistance in tumor cells.

In vivo models

Xenograft models: Among the various experimental systems, xenograft models play a crucial role in in vivo studies of pancreatic cancer drug resistance. Drug-resistant cell lines, such as AsPC-1/gemcitabine, are commonly implanted into immunodeficient mice to assess the reversal of chemoresistance in vivo, typically through parameters such as tumor volume reduction and improved survival rates[134]. However, a major limitation of xenograft models is the lack of an immune microenvironment.

PDXs models: PDX models are derived from the patient’s own tumor, preserving the genetic and pathological characteristics of the primary tumor. These models are ideal for studying individualized drug resistance and play an irreplaceable role in testing targeted drugs (e.g., the reversal effect of polyadenosine-diphosphate-ribose polymerase inhibitors in BRCA-mutant PDX models)[135]. Essentially, 3D organoid models are derived from PDX models.

Genetically engineered mouse models: Genetically engineered mouse models enable the specific knockout or editing of genes to develop pancreatic cancer, thereby simulating the natural progression of drug resistance[136]. This model overcomes the limitations of xenograft models by incorporating an immune microenvironment; however, it incurs high financial and time costs, necessitating careful consideration by researchers (Figure 3).

Ex vivo models

Ex vivo models involve the excision of fresh tumor tissue from patients through surgical resection, which is then cultured in tissue slices. Drug sensitivity testing can be performed within a short period (72 hours)[137]. This model retains the original tumor’s cellular composition and spatial structure while maximizing the utilization of existing tissue. It can be widely used in clinical settings for short-term drug sensitivity tests. However, it is essential to note that tissue slices cannot be passaged, making long-term studies or mechanistic investigations unsuitable for this model.

Specialized drug resistance mechanism models
In models where HIF-1α is known to mediate pancreatic cancer resistance, researchers have developed ways to simulate hypoxic environments to study how HIF-1α contributes to drug resistance. This is achieved by using hypoxic incubators (1% O2) to replicate the tumor’s hypoxic conditions[138]. Another specialized model focuses on exosome-mediated drug resistance. In this model, exosomes collected from drug-resistant cells are used to treat sensitive cells, thereby inducing cross-resistance and providing insights into the transfer of resistance mechanisms[139] (Table 5).

REPRESENTATIVE PHYTOCHEMICALS AND THEIR RESEARCH ADVANCES

Polyphenols
Dietary polyphenols, also referred to as phenolic compounds represent one of largest and most widely distributed groups of natural products in the plant kingdom. To date, over 8000 phenolic structures have been identified, including more than 4000 flavonoids[140-142]. The diversity and widespread distribution of polyphenols in plants increase their frequency of use. They can be classified into phenolic acids, flavonoids, polyphenolamides, and non-flavonoid polyphenols[143]. Representative substances include curcumin, resveratrol, and epigallocatechin gallate (EGCG).

Curcumin: Curcumin, a well-characterized polyphenol compound extracted from the turmeric plant, Curcuma longa. It was first isolated in 1815 and is also known as bisdemethoxycurcumin. Curcumin exhibits a broad spectrum of pharmacological activities, including anti-inflammatory, antioxidant, and anti-tumor properties, along with the ability to enhance radio- and chemotherapy sensitivity and protection of liver and kidney functions[144]. NF-κB is considered one of the major targets of curcumin’s activity, and much research has focused on this aspect[145]. Curcumin has poor bioavailability[146]. Clinical experiences using curcumin and the anti-metabolite gemcitabine in the treatment of advanced pancreatic cancer show that fewer than 10% of patients exhibit objective responses, with little impact on survival. However, the safety of this combination was demonstrated in phase I/II studies, with minimal toxic side effects[147,148]. The potent anti-proliferative activity of curcumin, which interacts with several intracellular signaling pathways, may enhance the antitumor effect of gemcitabine[149]. In addition to gemcitabine, curcumin has been reported to enhance the efficacy of other cytotoxic drugs, including cisplatin, oxaliplatin, and 5-FU[150,151]. While not all of these studies were focused on pancreatic cancer, they remain highly instructive.

Resveratrol: Resveratrol is a naturally occurring polyphenol synthesized by various plant species in response to damage, ultraviolet irradiation, and fungal attacks[152]. Among the more than 70 plant that contain resveratrol, red grape skins contain the highest amounts, with approximately 50 μg to 100 μg of resveratrol per gram of wet weight[153]. Interestingly, resveratrol maintains low biological toxicity to normal pancreatic cells[154,155]. This creates a foundation for its use in combination with chemotherapy drugs. Resveratrol sensitizes pancreatic cancer cells to various chemotherapy drugs, including gemcitabine, which has potential implications for cancer treatment. In an in situ human pancreatic cancer model, resveratrol significantly suppressed tumor growth (P < 0.001), and co-administration with gemcitabine further potentiated this inhibitory effect (P < 0.001). The combination of gemcitabine and resveratrol significantly downregulated markers of proliferation (Ki-67) and microvascular density (CD31) in tumor tissues (P < 0.001 vs control; P < 0.01 vs gemcitabine alone). Compared to the vehicle control, resveratrol suppressed NF-κB activation and reduced the expression of cyclin D1, cyclooxygenase-2, intercellular adhesion molecule-1, matrix metallopeptidase-9, and survivin. Overall, resveratrol can enhance the effect of gemcitabine by inhibiting markers of proliferation, invasion, angiogenesis, and metastasis[156]. The chemoresensitizing effect of resveratrol on tumor cells appears to be mediated through its regulation of various cell signaling molecules, including drug transporters, NF-κB, and signal transducer and activator of transcription 3[62,156,157].

EGCG: EGCG is the most abundant catechin in green tea. It is worth noting that both curcumin and EGCG are among the few plant-derived compounds approved for clinical trials. EGCG, when used alone, reduces the migration, invasion, and growth of pancreatic cancer cells. Moreover, when combined with gemcitabine, it enhances the therapeutic effects. EGCG and gemcitabine work synergistically by modulating EMT markers and inhibiting the Akt pathway, thereby suppressing pancreatic cancer cell growth, migration, and invasion[158]. A recent study investigated the effects and mechanisms of EGCG in combination with chemotherapeutic agents on pancreatic cancer cell growth, revealing that its primary mechanism involves the regulation of glycolysis. The results indicated that this specific catechin inhibits the growth of pancreatic cancer cells, and this inhibition is further enhanced under glucose-deprived conditions[159]. EGCG, when combined with bleomycin (an anticancer chemotherapy drug), exhibited anti-proliferative effects by inducing apoptosis in pancreatic cancer cells in vitro. EGCG also sensitized pancreatic cancer cells to various chemotherapy drugs, thereby overcoming drug resistance - a rare and valuable finding[160].

Flavonoids
Flavonoids are essentially a subclass of polyphenolic compounds. After an extensive literature review, the editors have categorized them as a distinct group due to the large number of studies and detailed classifications. More than 10000 flavonoid compounds are known[161], which can be further divided into subgroups. These include flavanols, flavanones, flavonols, flavones, isoflavones, and anthocyanins[162]. The representative compounds selected in this review are quercetin and apigenin.

Quercetin: Quercetin is a flavonol[163]. Data support that commercially available quercetin can be administered orally at a daily dose of 1 g, with an absorption rate of up to 60%, thereby maintaining a sufficient blood concentration. This absorption efficiency is higher than that of EGCG and curcumin, both of which are considered safe and easily absorbed by the human body[164]. Quercetin has been shown to increase the sensitivity of pancreatic cancer cells to chemotherapy drugs, including BET inhibitors, doxorubicin, gemcitabine, sulforaphane (SFN), anthracyclines, and tumor necrosis factor-related apoptosis-inducing ligand[165]. Additionally, many signaling molecules are involved in oxidative and inflammatory pathways mediated by Toll-like receptor 4 during the development of chemotherapy resistance. One study investigated the anti-cancer effects and mechanisms of quercetin in gemcitabine-resistant cancer cells. In this study, BxPC-3, PANC-1, HepG2, and Huh-7 cell lines were tested. The cell proliferation trend showed that quercetin had a significant cytotoxic effect on resistant cell lines (including HepG2 and PANC-1). Compared with gemcitabine alone, combination therapy with quercetin resulted in greater anticancer benefits[166]. Another researcher focused on the RAGE, a major protein involved in drug resistance. It was found that silencing RAGE led to the downregulation of the classic PI3K/AKT/mechanistic target of rapamycin pathway, thus killing resistant cells. Quercetin showed a significant effect similar to si-RAGE silencing, effectively reducing RAGE expression and promoting gemcitabine sensitivity in MIA Paca-2 GEMR pancreatic cancer cells, suggesting an unexpected response when quercetin and gemcitabine were combined[63].

Apigenin: Apigenin possesses potent pharmacological activity and exhibits low biological toxicity[167]. However, its bioavailability is relatively low[168], at only 5%-10%[169]. Therefore, there is a need for research into more effective absorption derivatives and models. Recently, a study indirectly demonstrated that apigenin can overcome chemotherapy resistance in pancreatic cancer by inducing apoptosis in human pancreatic cancer cells through inhibiting the glycogen synthase kinase-3β/NF-κB signaling cascade[170]. However, the sensitizing effect of apigenin on chemotherapy drugs is limited and does not show a concentration-dependent pattern. A cell proliferation assay was performed using different concentrations (0-50 μM) of several chemotherapy drugs, combined with pre-treatment with apigenin at various time points (0, 6, 24, and 42 hours). Pre-treatment with 13 μM apigenin for 24 hours followed by gemcitabine a 36-hour pretreatment was found to be optimal for inhibiting cell proliferation, resulting in 59%-73% growth inhibition[171].

Terpenoids
Since 1995, 162 terpenoid compounds have been discovered, and they are classified into five categories based on their structures: Monoterpenes, sesquiterpenes, diterpenes, triterpenes, and tetraterpenes. Terpenoid compounds are known for their anticancer, hepatoprotective, anti-inflammatory, antidiabetic, and neuroprotective properties[172]. A substantial of original research studies and reviews articles have highlighted the broad-spectrum anticancer activities and therapeutic potential of terpenoids[173]. The representative compounds selected for this review, which have been extensively studied, include β-elemene, paclitaxel (a semisynthetic derivative of taxol), and artemisinin.

β-elemene: β-elemene has been summarized by researchers as being effective in treating tumors across multiple systems (Integumentary System, Nervous System, Respiratory System, Immune System, Digestive System, Reproductive System, Endocrine System, Urinary System) in combination with various chemotherapy drugs (cisplatin, gefitinib, aldesleukin, oxaliplatin, doxorubicin, etc.) to enhance their effectiveness[174]. Xu’s study[175] found that β-elemene (30 μmol/L) significantly reduced the expression of P-gp. P-gp overexpression can cause MDR, and the reversal of drug resistance by β-elemene is likely related to this effect. Although there is a lack of studies specifically on β-elemene in pancreatic cancer, based on its sensitizing effects on chemotherapy drugs in many other types of tumors, we can reasonably speculate that β-elemene may have the potential to reverse drug resistance in pancreatic cancer.

Paclitaxel: Paclitaxel is a very special terpenoid compound that plays a key role in pancreatic cancer chemotherapy regimens. It is also one of the chemotherapy drugs that strengthens our belief in the great potential of plant-derived compounds in pancreatic cancer treatment. Paclitaxel is commonly used in combined with carboplatin or cisplatin as a first-line chemotherapeutic agent, and its most clinically accepted form is nab-paclitaxel (albumin-bound paclitaxel)[176]. This subtype has milder biological toxicity compared to solvent-based paclitaxel formulations. The phase III MPACT trial in metastatic pancreatic cancer patients showed that, compared to gemcitabine monotherapy, nab-paclitaxel combined with gemcitabine had superior efficacy, including the primary endpoint of OS (median OS: 8.7 months vs 6.6 months; P < 0.001)[177].

Artemisinin: Artemisinin was first isolated and for the treatment of malaria in China in 1972. It is a highly effective and cost-efficient drug for treating malaria and is classified as a sesquiterpene compound[178]. Subsequent research has demonstrated that artemisinin derivatives exhibit significant anticancer activity in various cancer cell lines and animal tumor models. Multiple clinical trials have confirmed its potential as an anticancer agent[179]. Furthermore, dihydroartemisinin enhances T-cell proliferation and immune activity, thereby promoting an anti-cancer immune response[180]. Importantly, dihydroartemisinin increased the sensitivity of BxPC-3 and PANC-1 cell lines to gemcitabine and tumor necrosis factor-related apoptosis-inducing ligand therapies, both in vitro and in vivo[181,182]. In a PDAC xenograft model, dihydroartemisinin significantly enhanced the tumor-suppressive effect of gemcitabine by approximately 25%[181].

Alkaloids
Alkaloids, derived from plants, are iconic specialized metabolites with numerous biological activities. Plants utilize these naturally occurring nitrogen-containing compounds as a defense mechanism to cope with various challenging environmental factors. These organic compounds have been used in traditional and modern medicine to treat various diseases. The diverse functional groups attached to their core structures are responsible for their various biological properties[183]. Alkaloids are generally categorized into six main categories: Piperidine alkaloids, isoquinoline alkaloids, indole alkaloids, terpenoid alkaloids, steroid alkaloids, and other alkaloids[184]. The representative compounds selected for this review are berberine, vinblastine, and camptothecin.

Berberine: Berberine is a naturally occurring plant-derived polyphenol found in several plants/herbs, including turmeric, goldenseal, Chinese goldthread, anemone, and Oregon grape[185]. As a widely used plant extract in clinical practice, a meta-analysis has shown that berberine significantly affects blood glucose levels, insulin resistance, inflammatory markers, colorectal adenomas, and Helicobacter pylori infection[186]. However, its poor solubility in water may hinder its effective absorption in the small intestine. Despite this, substantial progress has been made in improving its bioavailability[187]. Studies have demonstrated that berberine exhibits effective anti-inflammatory[188,189], antioxidant[189,190], antidiabetic[190-192], and anticancer properties[193-195]. In combination with lovastatin, berberine treatment of Panc02 cells resulted in a twofold increase in the percentage of cells in the G1 phase, accompanied by moderate increases in the number of G1 phase cells. Interestingly, although pretreatment with mevalonate pathway products (i.e., cholesterol synthesis) weakened the anticancer effect of lovastatin, it did not reduce the anticancer efficacy of berberine, suggesting that berberine enhanced the anticancer efficacy of lovastatin. In the Panc02 xenograft model in C57BL/6 mice, when treated with berberine (oral, 100 mg/kg/day) and lovastatin (intraperitoneal injection, 30 mg/kg/day), the tumor volume showed a slight reduction, with lovastatin alone showing a more pronounced effect. Still, it was significantly reduced with the combination therapy[196,197].

Vinblastine: Vinblastine and vincristine are representative members of the vinca alkaloid family and rank among the most significant plant-derived natural products contributing to cancer chemotherapy[198,199]. Both vinblastine and vincristine are effective clinical drugs that were introduced into clinical practice more than 50 years ago, used in combination to treat Hodgkin’s disease, testicular cancer, ovarian cancer, breast cancer, head and neck cancer, and non-Hodgkin lymphoma (vinblastine), or as curative treatment for pediatric lymphocytic leukemia and Hodgkin’s disease (vincristine). Four vinca alkaloid are used clinically, including vinblastine, vincristine, vindesine[200], and vinorelbine[201]. As a plant-derived compound directly used in cancer treatment, vinblastine, like paclitaxel, inevitably leads to drug resistance, which significantly hampers the treatment efficiency of vinca alkaloids. Consequently, many studies focus on restoring pancreatic cancer’s sensitivity to vinblastine. For example, inhibiting βIVb-tubulin in pancreatic cancer cells can increase sensitivity to vinca alkaloids (vincristine, vinorelbine, and vinblastine)[202]. Although research in this area is limited, we believe that vinblastine still has substantial potential to reverse drug resistance in pancreatic cancer.

Camptothecin: Camptothecin is derived from the Chinese tree Camptotheca acuminata, with mainland China being the only known source. It was originally isolated in 1966 as a potent anti-tumor agent[203]. Despite its good anticancer properties, camptothecin has extremely low bioavailability, which remains a critical limitation. As of June 2025, many researchers are working to improve its bioavailability and have made significant progress[204]. Topo-I serves as a key molecular target in the discovery and design of anticancer drugs. All clinically evaluated Topo-I inhibitors are camptothecin derivatives, including irinotecan (camptosar)[205] and topotecan (hycampitin)[206], which are Food and Drug Administration-approved for clinical use. Irinotecan, a component of the FOLFIRINOX regimen for pancreatic cancer, was shown in a global, multicenter phase III clinical trial (NAPOLI-1), which demonstrated improved survival in patients with metastatic pancreatic ductal adenocarcinoma who had previously received gemcitabine treatment, with a manageable safety profile[207]. Studies have found that phosphorylated aspirin (MDC-22) enhances the biological inhibition of irinotecan[208]. While much research focuses on the resistance issues related to the AG regimen for pancreatic cancer, only a few studies address FOLFIRINOX resistance. As an effective pancreatic cancer therapeutic compound, many studies aim to enhance the effects of camptothecin derivatives. It is worth considering whether camptothecin derivatives can be used as a variable in combination with existing drug treatment systems, which could be an interesting research direction for overcoming drug resistance.

Organosulfur compounds
Organosulfur compounds are found in foods like garlic, onions, broccoli, cabbage, and other vegetables from these plant families (Amaryllidaceae, Brassicaceae)[209,210]. Due to their ability to inhibit matrix metalloproteinases, suppress carcinogen-activating enzymes, and induce apoptosis, they have been investigated as potential antitumor agents and chemoresensitizers in combination therapies for colorectal cancer (CRC)[211-213]. Representative compounds include allicin and SFN.

Allicin: Garlic is widely used in cooking globally, and its active compound, allicin, is a potent organosulfur compound derived from garlic. Allicin is formed when the enzyme alliinase acts on another organic sulfur compound, alliin[214]. Studies have summarized that allicin has anticancer, antioxidant, anti-inflammatory, antifungal, and antibacterial properties[215]. Allicin is a lipophilic organosulfur compound and is unstable once synthesized, rapidly breaking down into a series of lipophilic organosulfur compounds, including diallyl disulfide, diallyl trisulfide, ajoene, allyl methyl trisulfide, diallyl sulfide, and others[216]. Among them, diallyl trisulfide has shown anticancer activity against pancreatic cancer. These compounds exert antitumor effects primarily by inhibiting proliferation, inducing apoptosis, suppressing migration and metastasis, and enhancing the chemotherapy sensitivity of tumor cells[217]. One study found that allicin affected a 3D drug-resistant model of pancreatic cancer[218]. Allicin has been found to reverse drug resistance in gastric cancer[219], hepatocellular carcinoma[220], CRC, and lung cancer[221] against 5-FU. Current research on allicin and its derivatives mainly focuses on gastric and CRC. As a safe and effective plant-derived compound, allicin warrants re-evaluated for its potential therapeutic effects against pancreatic cancer.

SFN: SFN is a natural, plant-derived compound found in cruciferous vegetables, such as broccoli and cauliflower, and has been widely studied for its neuroprotective and anticancer properties[222]. SFN demonstrates cytotoxic effects on pancreatic cancer, and a review of current experimental results suggests that SFN can regulate NF-κB-related target genes[223], induce cell cycle arrest, and increase reactive oxygen species levels to promote apoptosis[224]. Importantly, we found that combining SFN with gemcitabine in tumor xenografts overexpressing miR30a-3p reduced tumor volume and increased gemcitabine sensitivity[100]. This finding aligns with the ncRNA-mediated drug resistance mechanisms discussed earlier in this article. Despite these excellent preclinical results, no clinical trials have been conducted, and there is a need to fill this experimental gap.

Polysaccharides
Natural polysaccharides are large molecules with widespread biological activities, found in almost all plants. Due to the various methods of polysaccharide modification, numerous derivatives are available. Modern pharmacological research indicates that polysaccharides and their derivatives exhibit a variety of biological activities in vitro, including antioxidant, antitumor, immunomodulatory, antiviral, antibacterial, and anticoagulant functions[225-227]. Importantly, polysaccharides have been shown to have the ability to overcome drug resistance by targeting epigenetic regulators and components of the TME[228,229]. Depending on their biological source, polysaccharides can be classified into fungal polysaccharides, plant polysaccharides (originating from higher plants), animal polysaccharides, and bacterial polysaccharides. Representative substances include Astragalus polysaccharides (APS) and lentinan (LNT).

APS:
Astragalus is a leguminous plant known as “Huangqi” in China. According to the Chinese Pharmacopoeia, the medicinal part of Astragalus is its dried root[230]. The main components of Astragalus include APS, flavonoids, saponins, and alkaloids. APS is one of the important active natural components derived from Astragalus[231]. Numerous pharmacological studies have demonstrated that APS possess diverse biological activities, such as the regulation of blood glucose and lipids, anticancer and anti-aging effects, and immunomodulatory functions[232]. The mainstream view is that APS exerts its antitumor effects by enhancing immunity, inducing tumor cell apoptosis, and inhibiting tumor cell proliferation and metastasis[233]. Recent retrospective studies have found that APS combined with gemcitabine and tegafur-otastat potassium capsules (S-1) significantly improves short-term efficacy and long-term survival in pancreatic cancer patients, and alleviates chemotherapy-induced immunosuppression and adverse reactions (P < 0.05)[234]. In 2018, Wu et al[235] first demonstrated that apatinib showed potential inhibitory effects on pancreatic cancer cells (ASPC-1 and PANC-1), and APS enhanced the antitumor effects of apatinib by further downregulating the phosphorylation of AKT, extracellular signal-regulated kinase, and matrix metallopeptidase-9. These excellent clinical and preclinical results confirm that APS can restore pancreatic cancer sensitivity to chemotherapy drugs. In the latest clinical guidelines for the integrated treatment of advanced pancreatic cancer using Chinese and Western medicine in China, APS have been combined with the AG regimen. However, no additional double-blind or prospective clinical studies, so whether this drug can reverse pancreatic cancer resistance remains unclear.

LNT: LNT is a key active extract from shiitake mushrooms. Numerous pharmacological and clinical studies have demonstrated that LNT has antioxidant, immunomodulatory, anticancer, blood sugar-lowering, and blood lipid-lowering biological activities[236-238]. LNT is a macromolecule with β-1,3-D-glucan, and its unique molecular structure is closely related to its pharmacological activity, with the β-glycosidic bond in glucan being the key structure for its antitumor function[239]. In Japan and China, LNT has been incorporated into cancer adjuvant therapy systems[237,240]. Sun et al[241] demonstrated that LNT, when combined with gemcitabine chemotherapy, significantly suppressed the proliferation of UBC cells-an effect confirmed by both in vivo and in vitro[242] experiments. There is a reported case in Japan of gastric and pancreatic cancer that showed a significant response to combination chemotherapy with S-1, paclitaxel, and LNT[243]. Although there are related literature reports, due to the age of the studies, their authenticity is questioned; therefore, the editors have chosen not to summarize this part of the content. Given the enhanced chemotherapeutic effect of LNT when combined with gemcitabine, it is conceivable that its combination with the AG regimen in pancreatic cancer could be highly promising, considering its stable source, high yield, and the unique advantages of fungal-derived compounds (Figures 4 and 5, Table 6).

LIMITATIONS AND CHALLENGES
Most plant-derived compounds, being naturally occurring substances, show varying degrees of efficacy against pancreatic cancer and in reversing its drug resistance. However, many unresolved issues remain. Below, we summarize these problems into five key points: Pharmacokinetic defects, complexity of the TME diminishing drug efficacy, the double-edged sword effect of multi-target action, and challenges related to standardization and heterogeneity.

Pharmacokinetic defects
A major issue with the natural compounds mentioned above is their poor bioavailability. This limitation can be primarily attributed to low oral bioavailability and inadequate systemic concentration. Despite numerous studies confirming their surprising effects in reversing pancreatic cancer resistance, the compounds’ effectiveness will be meaningless if their oral bioavailability or blood concentration cannot attain the efficacy observed in vitro experiments. Efforts to address this issue through high-doses administration often fail to enhance absorption efficiency and may lead to additional toxicity. Previous studies have shown that poor bioavailability is associated with limited absorption, rapid metabolism, and systemic elimination[146,244]. Numerous researchers have investigated synthetic derivatives and materials-based delivery systems to overcome these limitations and enhance bioavailability, thereby improving drug delivery to tumor sites and achieving encouraging outcomes[187,244-246].

Complexity of the TME diminishing drug efficacy
The TME in pancreatic cancer is characterized by dense fibrotic stroma (with stroma accounting for > 70%) and immune-suppressive cell infiltration, which is not typically replicated in conventional in vitro cell line experiments. To simulate the TME, 3D organoid models are more appropriate. The weakening of drug efficacy is largely attributed to the physical barriers of the TME, immune evasion mechanisms, and high expression of MDR proteins. These factors contribute to the resistance of pancreatic cancer to chemotherapy. Recent research has shown that nanoparticle delivery systems can address this issue[247]. Through the application of advanced material technologies, these systems can sustain blood drug concentrations and achieve more precise and targeted delivery[248].

The double-edged sword effect of multi-target action
Plant-derived compounds can reverse resistance through multiple pathways[149]. For example, curcumin can enhance the efficacy of gemcitabine in combination by inhibiting NF-κB[249], and overcome pancreatic cancer chemoresistance through its actions on nuclear factor of activated T cells and Nrf2[250]. However, this multi-target approach is not always desirable. Multiple pathways can lead to unclear mechanisms and uncontrolled therapeutic effects, posing unpredictable risks. Achieving precision treatment remains a key hurdle for plant-derived compounds to become mainstream drugs.

Standardization and heterogeneity challenges
The issue of standardization is particularly evident in polysaccharide-based compounds. Due to the uniqueness of polysaccharides, their chemical structures cannot always be easily depicted, which results in inefficiency when quantifying active substances and large batch-to-batch variability[251]. Chemical structural variations and the interference of metabolic products are other challenges. Many bioactive substances undergo metabolic transformation in the human digestive organs, particularly in the liver and small intestine. For example, curcumin’s double bonds are reduced to dihydrocurcumin, tetrahydrocurcumin, hexahydrocurcumin, and octahydrocurcumin by reductases in enterocytes and hepatocytes[252]. Resveratrol is reduced to dihydroresveratrol conjugates and other derivatives, and the extensive metabolism in the intestines and liver results in an oral bioavailability of less than 1%. Even with increasing doses and repeated administration, the situation doesn’t improve significantly[253]. Quercetin and kaempferol are rapidly metabolized in the liver and circulate as methyl, glucuronide, and sulfate conjugates[254]. Delivering plant-derived compounds to tumor sites in their native structures, as verified in vitro, remains a pressing issue.

PERSPECTIVES AND FUTURE RESEARCH DIRECTIONS
Overcoming pancreatic cancer drug resistance using plant-derived compounds must address three major contradictions: Balancing multi-target specificity, TME penetration and immune remodeling synergy, and standardizing chemical heterogeneity. Future research should focus on integrating nanotechnology, biomarker stratification, and immune-metabolic reprogramming strategies to drive precision translation. To enhance research efficiency and facilitate the translation of in vitro and in vivo findings into clinical applications, selecting compounds that are it is advisable to prioritize compounds already approved by the Food and Drug Administration and National Medical Products Administration to ensure smoother clinical translation. Emerging delivery systems can be further optimized through nanotechnology-based approaches, including targeted liposomes and exosomes. Such systems may be activated by specific proteins in the cellular microenvironment or through biotransformation mediated by intestinal microbiota and hepatic enzymes, thereby enhancing bioavailability and overall delivery efficiency. At the same time, the rapid development of artificial intelligence technologies has brought new opportunities to researchers. For instance, platforms such as DeepPhytoScreen allow the construction of deep learning models using either public or user-provided datasets. These models facilitate the virtual screening of chemical probes or drug candidates against specific targets of interest. Through DeepScreening, users can conveniently build deep learning models and generate target-focused databases. The classification and regression models thus established can be employed for virtual screening the generated databases or existing chemical libraries. These libraries may include synthetic compounds, natural products, approved drugs, covalent agents, protein-protein interaction modulators, and allosteric regulators collected from various chemical suppliers. The entire workflow - from deep model training to virtual screening, and from database construction to target-focused de novo library generation - can be accomplished through DeepScreening. We believe that such deep learning-based web servers will greatly benefit biologists and chemists in fundamental research and drug discovery.

CONCLUSION
Pancreatic cancer, especially PDAC, remains a significant global health challenge due to delayed diagnosis, invasive metastasis, and stubborn treatment resistance. The current standard treatment regimens, based on gemcitabine (e.g., the AG regimen) or FOLFIRINOX, are limited by both intrinsic and acquired drug resistance, resulting in limited survival benefits for patients. This review systematically discusses the multidimensional mechanisms of chemotherapy resistance and highlights the potential of plant-derived compounds in overcoming these resistance barriers. Despite significant preclinical evidence, translational applications still face challenges, including improving compound bioavailability, systematically evaluating synergy with standard chemotherapy, and validating efficacy in advanced models (e.g., PDOs and genetically engineered mouse models). Future studies should focus on determining optimal dosing regimens, developing combination strategies, and establishing patient stratification biomarker systems. Plant-derived compounds provide valuable resources for overcoming challenges in pancreatic cancer therapy and show great potential as novel adjunctive strategies to improve clinical outcomes.